The Obesity Society Annual Meeting

September 20-24, 2012; San Antonio, TX Full Commentary – Draft

Executive Highlights

We traveled to San Antonio, Texas in late September for five days of excellent learning at the 30th Annual Scientific Meeting of The Obesity Society (TOS). The optimism for weight-loss drugs seen at last year’s TOS (see our report at was more tangible than ever this year in light of the recent approvals of Vivus’ Qsymia and Arena/Eisai’s Belviq. Vivus certainly dominated the conference’s advertising, discussion, and exhibit hall given the US launch of Qsymia just two days prior to the start of the meeting (see our report on the launch at

On the device and surgery front, there were few new learnings or exciting data at TOS 2012. Roux-en-Y gastric bypass was touted for its valuable physiologic benefits beyond just weight-loss, though we continue to hear questions regarding long-term side effects and complications.

Notably, next November, the American Society for Metabolic & Bariatric Surgery (ASMBS) and TOS will co-locate their respective yearly conferences and jointly host “Obesity Week” in Atlanta, Georgia in November 2013. We also hope future TOS meetings bring more discussion of the vast array of obesity devices in development, especially ones that have significant potential to reduce diabetes and diabetes complications, such as the GI Dynamics’ EndoBarrier.

Relative to prior years, TOS 2012 had more focus on the potential of mobile health, though like devices and surgery, there were few big or exciting takeaways besides “we need more research.” By contrast, there were a number of interesting policy discussions by a number of opinionated speakers. Speakers and attendees lauded New York City Mayor Michael Bloomberg’s recent soda ban (see the September 15, 2012 Closer Look), which came up in more presentations than we could count and was generally characterized as a justified public health move. Although perhaps it isn’t surprising given the audience, we were glad to hear this support, given that several years ago, Mayor Bloomberg’s proposed tax on sodas didn’t appear to stand a chance of passing (and it didn’t, as Pepsi threatened to depart for North Caroline, taking thousands of jobs with it – that was the end of the soda tax movements. Optimism was also high concerning the potential revenue and health benefits that targeted government taxes could have, though it’s clear that there are still very formidable barriers that need to be overcome.

This report contains our full coverage of TOS 2012, organized into six categories: 1) Drug therapies; 2) Mobile Health; 3) Devices and Bariatric Surgery; 4) Policy and Practice; 5) Other Notable Talks; and 6) Exhibit Hall (Eisai and Vivus’ booths). Below, we discuss the major themes and our big-picture takeaways from the conference, followed by coverage of individual presentations. Talk titles highlighted in yellow were not previously published in our daily reports.

  • Many speakers were energized and optimistic about pharmacotherapy given the recent approvals of Arena/Eisai’s Belviq (lorcaserin) and Vivus’s Qsymia (phentermine/topiramate ER). Notably, both academics such as Dr. Ken Fujioka (Scripps Health, San Diego, CA) as well as company execs seemed energized; Dr. Fujioka expressed optimism that the approvals would “change the landscape,” (can’t get much better than this if you’re Vivus, we would say) and Dr. Preston Klassen (Orexigen, La Jolla, CA) characterized the present as a “fantastic time for the obesity space.” Key opinion leaders also provided thoughtful commentary on the new drugs throughout the conference. At the Vivus-sponsored CME nearly 350 attendees (very packed!) listened to Drs. Robert Kushner (Northwestern School of Medicine, Chicago, IL), Donna Ryan (Pennington Biomedical Research Center, Baton Rouge, LA), Samuel Klein (Washington University, St. Louis, MO), and Louis Aronne (Weill Cornell Medical College, New York, NY) discuss the benefits of weight loss and the use of Vivus’ Qsymia and Arena/Eisai’s Belviq. Speakers showed strong interest in using both drugs to treat diabetes, especially on top of anti-glycemic agents like metformin or GLP-1 agonists. There was not much new data at TOS, though Vivus had a poster suggesting its proposed treatment algorithm (which has non- responders discontinue or increase the dose) helps maximize the benefit/risk profile of the drug. As a reminder, the “recommended dose” of Qsymia (phentermine 7.5 mg/topiramate ER 46 mg) is associated with a much lower incidence of adverse events leading to discontinuation of the drug (4.0%) than in either non-responders (29.3%) or even those on placebo (8.4%) – this should really help confidence in the compound, we would think. The poster hall also featured a poster by Arena - analyzing data from the phase 3 program for lorcaserin, it demonstrated that a higher percentage of people (both with and without type 2 diabetes) receiving lorcaserin decreased the use of concomitant medications for type 2 diabetes compared to placebo over a one-year treatment period.
  • Qsymia’s launch only two days before the conference was outstanding timing for Vivus, of course, and the “platinum plus” sponsor (one level higher than Novo Nordisk!) dominated the conference’s marketing. The company stole the show with white and purple Vivus lanyards securing every attendee’s badge, a full back page ad on the conference’s final program (plus a logo on the cover), a CME, a product theater, and even a massive white and purple carpet at the entrance to the exhibit hall (beckoning attendees to the only booth taller than eight feet, which, given that it was only about 25 feet away from the entrance, was virtually impossible to miss). By contrast, Eisai’s presence was understated. Belviq is still being scheduled by the Drug Enforcement Agency and was not mentioned at Eisai’s booth, and the company’s Saturday night CME had heavy competition with Novo Nordisk’s CME and TOS’ party “Fiesta in ‘Ol San Antonio.”
  • Cost-effectiveness and reimbursement were the focus of a smattering of sessions – we believe such sessions will continue to increase in number. There was one cost- effectiveness session featuring a particularly striking presentation from Dr. William Dietz (Centers for Disease Control, Atlanta, GA) on the economic burden of obesity. Dr. Dietz emphasized the disproportionate cost burden imposed by those with a BMI >35 kg/m2 – in one study, they claimed 40% of medical costs yet just 8% of prevalence. This is disturbing given that although obesity is plateauing overall, BMIs over 35 and over 40 and even over are increasing at a disturbing rate. Also valuable was Dr. William Herman’s (University of Michigan, Ann Arbor, MI) discussion on the cost-effectiveness of obesity treatment strategies. We saw almost universal cost effectiveness (characterized by <$50,000 per QALY gained) across studies of lifestyle interventions, bariatric surgery, and even orlistat, although we point out that with such approaches, it is not really the QALY that is in dispute, it is how to get patients to adhere to challenging lifestyle programs or medicines with very difficult side effect profiles. In our view, this underscores that while cost-effectiveness may always be that challenging to demonstrate technically, especially when obesity therapies are used in populations with particularly staggering costs (e.g., BMIs >35 kg/m2 and people with type 2 diabetes), the point is to find approaches that can actually be used successfully in large groups of patients. Ultimately, Mr. Theodore Kyle (ConscienHealth, Pittsburgh, PA) cautioned that there is indeed a long way to go on the reimbursement front, most notably because Medicare Part D excludes coverage of weight loss drugs.
  • This year’s TOS left us with a good sense of some key factors in demonstrating cost- effectiveness for new therapies like Qsymia and Belviq: effects on comorbidities, adherence, and tolerability. As noted, adherence and tolerability will be absolutely critical as cost pressures continue to heighten – we think payors will hone further in on adherence and persistence, primarily because these are chronic therapies (which have historically low adherence) and because there’s potential for inappropriate use. We also note that obesity is still not universally recognized as a medical condition by many, and this also limits reimbursement. On the plus side, we expect adherence might initially be higher for obesity drugs relative to other chronic drugs because the benefits will be more noticeable (i.e., weight loss vs. a reduction in cholesterol or blood pressure) and patients are likely to be highly motivated to lose weight (vs. a high motivation to reduce one’s cholesterol). However, the big question on our minds is whether adherence and persistence will remain high once weight loss plateaus with Qsymia and Belviq – we would actually assume that if patients begin to gain the weight lost, however, due to a quite tolerable side effect profile for Qsymia in particular, patients would not mind going back on the drug.
  • We also heard a lot at TOS 2012 about the importance of treating the comorbidities associated with obesity. We believe this is where significant cost saving potential of obesity pharmacotherapy will come in, especially given the ongoing emphasis on how losing just a little bit of weight can vastly improve one’s health - this was proven, of course, by DPP – although this is a good example of where lifestyle programs are challenging to maintain. With all this in mind, we think Qsymia is by far the best positioned to garner early payer interest given its strong efficacy, minimal side effect profile, a titration algorithm that seems to maximize benefit/risk (see our coverage of Vivus’ poster), and stopping rules. Belviq has a better tolerability profile than other obesity drugs have, though its efficacy (and thus effects on comorbidities) is less compelling. However, considering Belviq has a more aggressive stopping rule than Qsymia (Belviq: <5% weight loss at week 12 vs. Qsymia: <3% weight loss at week 12 with an option to escalate to the higher dose), patients that stay on Belviq will be very strong responders to the drug and this may enhance payer interest. As a reminder, in Belviq's trials, about 56% of all people taking the drug did not achieve a weight loss of 5% - still, 44% of patients did, and some minority of that would have been “super-responders” who lose 10% or even 15% weight loss – it is that smaller group that may most interest payors, and we hope doctors continue the strong interest in individualizing therapy and identifying those patients.
  • As in diabetes, interest in building mHealth for obesity is high – but “there’s no there, there” yet as there appeared to be very few, if any, mHealth solutions that have been termed so successful that they are scalable for widespread use. Notably, TOS recently added an eHealth/mHealth business section to the society to further the use and promote the study of mHealth in the obesity arena. One approach that garnered great interest was Dr. Steve Cole’s (HopeLab, Redwood City, CA) presentation on the results of the Zamzee Randomized Controlled Trial certainly spoke to the potential of mHealth solutions in this field – after six weeks, children randomized to the Zamzee intervention (a two-part solution that included a kid-friendly accelerometer and online website) averaged 42 more minutes per week of moderate and vigorous physical activity than children in the control group! We were very impressed by this, and believe design of the Zamzee will continue to improve. Overall, however, as Dr. Alain Labrique (Johns Hopkins University, Baltimore, MD) aptly pointed out, the mHealth landscape is generally under-evaluated and overcomplicated. The definition of a successful mHealth solution changes, of course, depending on whether the intervention is considered from the perspective of a patient, regulator, provider, or industry. Moving mHealth from its current state of impressive creativity - Dr. Labrique said there are over 1,000 weight related applications for Apple and Android phones - to a period where established interventions are widely available will require appropriate efficacy studies that can keep up with the pace of innovation. And, then, there is a timing problem – it would be wonderful to see the impact of large trials, but the early trials must be done such that payors and governments and employers are interested in investing. We have been testing Zamzee internally at our patient newsletter diaTribe since mid-summer – overall, although we’ve had a few problems associated with losing or breaking Zamzees, the early reviews have been very high, and at $29/Zamzee, this could certainly be a solution used with very large groups of children.
  • Sugar sweetened beverages (SSBs) were a major topic of discussion at TOS, and there was widespread support for enacting policies that will tax or ban them. The issue was of course quite timely given the passage of New York City Mayor Michael Bloomberg’s soda ban in the prior week (see the September 15, 2012 Closer Look at TOS 2012 started with a stimulating debate on this very topic, featuring researchers Drs. David Allison (University of Alabama, Birmingham, AL) and Frank Hu (Harvard School of Public Health, Cambridge, MA) at this high-profile session. Indeed, NEJM had a major piece on the impact of SSB that week and the topic is certainly receiving larger national notice. Both Dr. Allison and Dr. Hu honed in on the very much two-sided research data linking SSB consumption to changes in body weight. While we don’t believe there was a clear takeaway or winner, and while a concluding straw poll indicated that most audience members were unchanged in their views, that doesn’t necessarily tell the whole story, since we suspect more went into the debate already in favor of a tax. Indeed, across the meeting, opinions on the topic definitely favored SSB tax and ban policies, with many remarks in sessions and Q&As heavily supporting Mayor Bloomberg’s work.
  • More on policy - we particularly enjoyed an illuminating talk by Dr. Frank Chaloupka (University of Illinois at Chicago, Chicago, IL), an expert on tobacco taxes. He explained that excise taxes are the way to go (e.g., a penny per ounce soda tax) since they have a bigger impact on reducing consumption and lead to more stable streams of revenue. By his math, a penny per ounce tax on SSBs would generate approximately $15 billion in additional revenue per year. Excise taxes are, of course, politically much harder to pass – Mayor Bloomberg tried this first several years back and as noted, Pepsi objected and tied it instantly to jobs that would be lost if the tax were instituted (in which case the soda giant said it would transfer tens of thousands of employees to North Caroline. We think the ban on beverages over 16 ounces is a very astute way to achieve the same thing; it is basically just a tax of another kind given that people who really want 32 ounces, for example, will certainly pay more by purchasing two 16-ounce drinks than they would have otherwise paid by purchasing a 32-ounce drink. Dr. Steven Gortmaker’s (Harvard School of Public Health) forecast from a SSB tax was similar (~$12.7 billion in tax revenue per year), though he also forecasted 22.9 million disability adjusted life years (DALYs) averted and a national reduction in health care costs of $1,700 per dollar spent– wouldn’t that be something to put towards investing in obesity education and prevention,marketing physical activity, and perhaps subsidizing fruits and vegetables?! Now, just to get it passed…
  • Presenters emphasized the physiological benefits of gastric bypass; however, a number of bariatric surgery studies appear to be replete with design and recruitment difficulties. Many presenters, like Dr. Jonathan Purnell (Oregon Health & Science University, Portland, OR), added to the body of evidence that Roux-en-Y gastric bypass (RYGB) confers physiological benefits beyond weight-loss. Dr. John Kral (SUNY Downstate Medical Center, Brooklyn, NY), Mickey Stunkard Lifetime Achievement Award recipient, even suggested that the name “bariatric surgery” was inappropriate for gastric bypass as it overemphasized the weight-loss component of the surgery. However, a talk by Dr. Edward Livingston (Deputy Editor, JAMA, Chicago, IL) asserted that the studies evaluating bariatric surgery thus far have not been robust. Dr. Livingston emphasized that poor follow-up and a lack of long-term studies leave many unanswered questions in this field. Randomized controlled studies for bariatric surgery intervention could be a new standard these already exist! Enteromedics is RCT etc., but as Dr. Philip Schauer (Cleveland Clinic, Cleveland, OH) pointed out based on his experience with the STAMPEDE study (which randomized patients to intensive medical therapy or bariatric surgery), while RCTs are feasible for bariatric surgery, they are plagued with recruitment challenges. Certainly, with bariatric surgery being seriously considered for obese adolescents, as Dr. Thomas Inge (University of Cincinnati, Cincinnati, OH) would recommend, questions about long-term effects, patient characteristics that predict success or failure, and mechanism of action need to be answered. As to the latter, in an interesting talk on vertical sleeve gastrectomy (VSG), Dr. Randy Seeley (University of Cincinnati, Cincinnati, OH) suggested that both VSG and RYGB operate through a mechanism connected to bile acids, independent of GLP-1.
Table of Contents 


1. Drug Therapies

Preconference Workshop: Early Development Pharmacologic and Device Session For Up and Coming Obesity Treatments


Ken Fujioka, MD (Scripps Clinic, San Diego, CA)

Opening the session, Dr. Ken Fujioka noted that since last year’s meeting, two new medications (phentermine/topiramate [Vivus’ Qsymia] and lorcaserin [Arena/Eisai’s Belviq]) have been approved for obesity, expressing optimism that they would “change the landscape” and “open up this area.” He also pointed out that bupropion/naltrexone (Orexigen’s Contrave) and (Novo Nordisk’s Victoza) are well into final phase 3 trials, and anticipated approval in the next few years (Contrave, of course, is actually well past phase 3, has been submitted, received a CRL from the FDA, and is now working on a cardiovascular outcomes trial that the FDA is requiring prior to approval). Subsequently, he reviewed the recently released consensus report “Obesity Drug Outcomes Measures,” which summarized the findings and conclusions from a series of discussions between clinicians, consumer advocates, industry representatives, pharmaceutical representatives, public health organization representatives, and officials from the FDA, CDC, and NIH on the regulatory roadblocks for obesity medications. Importantly, the report emphasized the need to view drugs as treatments for obesity (and presumably prevention of co-morbities) as opposed to weight-loss drugs. The report identified two overarching challenges the FDA faces: determining the benefits of obesity medications beyond weight loss, and evaluating risks when there are a huge number of potential patients. It recommends the FDA include benefits beyond weight loss in its evaluation, supports approvals for subsets of the obese population for which benefits outweigh the risks, and encourages the use of REMS to limit inappropriate use and, notably, even for the FDA to have the power to make an obesity drug unavailable for off-label use. Finally, Dr. Fujioka identified pediatric obesity as a huge unmet need. Presently, orlistat (Roche’s Xenical) is the only medication approved for children; however, most are unwilling to take the drug because of its troubling side effect profile, leaving no options between diet/exercise and surgery. As such, the report encourages companies to include children and adolescents in clinical trials once safety concerns have been addressed. The committee that wrote the report met four times over the course of a year; it is available free at

  • “Obesity Drug Outcomes Measures” emphasizes that the FDA needs to delineate how companies can demonstrate benefits beyond weight loss. Currently “feel and function” or symptomatic improvements of obesity related disease are not considered by the FDA during the approval process. The FDA did, however, indicate that it intends to move toward a more patient-centered process and include “feel and function” in its evaluation of drugs. Dr. Fujioka recommended that the FDA update its guidelines for obesity medications to reflect this. To this end, he also recommended that the FDA should include a quality of life questionnaire to increase the measurability of a patient’s “feel and function,” and consider other benefits of weight loss such as mobility, sexual function, improvements in mood/depression, and sleep apnea. This is very important, of course, not because “convenience” or “happiness” should be the basis for approval or reimbursement, but because improvements in either factor often weigh heavily into adherence, how the patients do on treatments, etc.
  • The FDA has a challenge in evaluating risks, given that a huge number of potential patients could use obesity medications. There is a large jump in patient number from clinical trials to the general obese population, which the FDA must consider when determining if a drug is safe. Thus, Dr. Fujioka emphasized that it is necessary to ensure that the drug is not being inappropriately used (which is defined as a medication being given to a patient for whom the risks outweigh the benefits). To limit or mitigate the risk of medically inappropriate use of obesity drugs, the report encourages companies to use REMS. (While we certainly agree this should be done, we also believe there is a risk of not making drugs available to people that need them.) The report also recommends enabling the FDA to make an obesity drug unavailable for off-label use. Dr. Fujioka noted that this has been done before – the FDA has been able to make certain antibiotics unavailable for off-label use.
  • The consensus report recommends viewing obesity as a spectrum of health states (without taking BMI into consideration). The authors propose the following categories: obese and well, obese with risks (e.g., prediabetes), obese and sick (e.g., diabetes). Currently the FDA has to approve a drug for all of these groups, but the risks and benefits of a drug vary from one group to another. Thus, Dr. Fujioka said it would be “very good medicine” if the FDA could approve a drug for specific subsets of the obese population. He commented that currently, the best way to categorize individuals into the aforementioned states is the Edmonton Obesity Staging System, a tool created by the increasingly well-known and highly regarded obesity expert, Dr. Ariel Sharma (see our interview with Dr. Sharma from April, 2012 at
  • Dr. Fujioka emphasized that an enormous unmet need exists in pediatric obesity, as orlistat is currently the only obesity medication approved for this population. Dr. Fujioka explained that many children and adolescents are averse to orlistat because of its side effects, and that most extremely obese children and adolescents need more than diet and exercise. Unfortunately, besides orlistat, there is no treatment option between diet/exercise and bariatric surgery, and surgery is far from ideal. To address this need, “Obesity Drug Outcomes Measures” encourages companies to include children and adolescents in their clinical trials once safety concerns have been addressed, in order to make medications available for the pediatric population in a timelier manner.
  • The consensus report states that it is unclear how much weight loss is needed in the pediatric population in order to improve health. Dr. Fujioka noted that some obese children may not actually need to lose weight, but just maintain weight, since they are growing. He said that the government or researchers should develop a registry of pediatric patients being treated with obesity drugs to assess the long-term outcomes and side effects. On an optimistic note, we learned that in his clinic, Dr. Fujioka has seen pre-teens who are able to maintain the weight they lose from obesity medications even several years after discontinuing therapy, whereas adults often need to stay on drug to maintain the weight loss.



Sanjay Bhanot, MD (Isis Pharmaceuticals, Carlsbad, CA)

Dr. Sanjay Bhanot briefly reviewed antisense technology before describing Isis Pharmaceutical’s antisense drug for obesity (ISIS-FGFR4Rx), which targets the fibroblast growth factor receptor 4 (FGFR4). He opened his presentation by explaining that antisense technology uses single-stranded oligonucleotide molecules that bind to an mRNA strand and trigger its degradation, thus reducing levels of the mRNA’s protein product. Isis’ obesity drug markedly reduces FGFR4 levels and has been shown in preclinical studies to reduce body weight, adiposity, and lipogenesis in DIO mice. When applied to mouse and monkey hepatocytes in vitro, FGFR4 antisense oligonucleotides (ASO) also increase fatty acid oxidation. Dr. Bhanot then highlighted that FGFR4 ASO results in an increase in the expression and plasma levels of the mouse fibroblast growth factor 15 (FGF15). This finding prompted Isis to conduct preclinical studies using FGF19 – the human ortholog of mouse FGF15 – which found that FGF19 increases oxygen consumption in lean animals and increases fatty acid oxidation when applied to hepatocytes. To Dr. Bhanot, these results suggest that FGF15/FGF19 are not only markers of FGFR4 reduction, but may also be mediators of FGFR4 ASO’s anti-obesity effects. The FGFR4 ASO drug is currently in phase 1 development and Dr. Bhanot expects to reveal clinical data within the next 12 to 18 months.

  • Dr. Bhanot gave a brief description of antisense technology. While traditional drugs block the activity of a protein, antisense drugs bind to mRNA and trigger its degradation, thus resulting in a decrease in protein level. Specifically, antisense drugs are typically single-stranded molecules roughly 20 oligonucleotides in length that bind to a certain mRNA strand with high specificity and activate RNase H, which destroys the mRNA strand.
  • Antisense drugs are typically administered as simple saline formulations given once a week or less frequently. Such drugs remain in plasma for only roughly one hour before their uptake by tissues. Mr. Bhanot emphasized that antisense drugs do not cross the blood brain barrier, and thus do not elicit the CNS side effects that have historically hindered the development and use of obesity medications. The drugs’ neutral CNS impact also suggests that antisense technology may be an effective complement to centrally acting obesity medications. Mr. Bhanot noted that because the antisense parent drug and its nuclease products are eliminated via urinary excretion, their use poses little risk of drug-drug interactions. Furthermore, he highlighted that antisense drugs are not distributed to the heart and are thus not expected to trigger cardiovascular side effects.
  • Isis has 24 antisense drugs in development, five of which are in metabolic disease. The company has studied antisense technology in its clinical safety studies of over 22 drugs. Notably, the presentation slide indicated that Isis has treated over 150 patients with type 2 diabetes with an antisense drug for up to 13 weeks, and did not find any safety concerns unique to this patient population.
  • Isis is developing an antisense drug for obesity (ISIS-FGFR4Rx) that targets the fibroblast growth factor receptor 4 (FGFR4) in liver and fat tissues. Fibroblast growth factors are a family of proteins linked to growth effects and believed to play a role in maintaining metabolic homeostasis. Preclinical studies have shown that FGFR4 antisense oligonucleotides (ASO) reduce body weight and adiposity in a dose-dependent manner in DIO mice with or without caloric restriction, with no observed effect on physical activity. In DIO mice, FGFR4 ASO also resulted in reduced lipogenesis and decreased levels of key enzymes including ACC1, FAS, and DGAT-2. Reducing levels of FGFR4 also increased fatty acid oxidation in mouse and monkey hepatocytes in vitro, indicating that the drug has a direct effect on lipid metabolism.
  • Fibroblast growth factor 15 (FGF15) may be a key mediator of FGFR4 ASO’s anti- obesity effects. Preclinical studies show that FGF4 ASO reduces leptin and adiponectin expression, increases plasma PYY levels, has no impact on plasma levels of FGF21 or GLP-1, and most notably, increases ileum FGF15 gene expression and plasma FGF15 levels. Dr. Bhanot explained that the increase in bile acids triggered by FGFR4 ASO may lead to the upregulation of FGF15. Subsequent preclinical studies performed by Isis showed that infusing FGF19 – the human ortholog of mouse FGF15 – into lean animals results in greater oxygen consumption, suggesting that FGF19 increases metabolic rate and energy expenditure. Interestingly, these effects are sustained even in the presence of FGFR4 ASO (which results in a ~80% reduction in FGFR4), indicating that FGF19 may signal through other FGF receptors, or that some FGFR4 may remain in the liver. Furthermore, applying FGF19 directly onto hepatocytes results in increased fatty acid oxidation. To Dr. Bhanot, these results suggest that FGF15/FGF19 are not only markers of FGFR4 reduction, but may also be mediators of its favorable weight effects.

Questions and Answers

Q: Since you’re increasing the metabolic rate and you say there is no increase in motor, muscle, or thyroid activity, where is the extra energy going?

A: You lose a lot through heat – the temperature is going slightly up. But the main question is if there is no change in thyroid activity, why is there this big rate? Because you reduce ACC1 and the other ACC molecules, which influence fatty acid oxidation. And over time, you reduce lipid synthesis. That also contributes in the overall lipids that accumulate in the animals.

Q: With the increased metabolic rate do you see any changes in heart rate or blood pressure with the drug?

A: We have not measured those yet, and they can be deceptive when you look at them in animals. But they are something we will look at.

Q: Do you have any data on the glucose tolerance of these animals? Or on specific measures of insulin sensitivity?

A: Yes, we’ve done euglycemic clamp studies and have demonstrated that you improve hepatic insulin resistance and observe a small, but significant effect on peripheral insulin sensitivity. It turns out we’re primarily affecting gluconeogenesis.

Q: I heard that there’s a difference in this fibroblast growth factor receptor 4 in the fat and muscles between African Americans and Caucasians that affects the response of African Americans to weight loss. Do you have any idea about what the risk differences are between the races? Do you know if the gene expression is different?

A: Is there a difference in the fibroblast growth factor receptor 4 in Caucasians versus the other races? I don’t know of any studies that have looked at this so far.

Q: You noted that there was an increase in bile acid pool. Did you look at cholesterol?

A: Yes, there was about a 10% to 15% change in bile acids. We looked at the entire pathway including cholesterol. There was no change in cholesterol or triglycerides. We followed these animals for several months looking at if there were any changes in this pathway, and we did not see anything.

Q: You showed that the drug increases PYY. Is this the true PYY or PYY 3-36? Because PYY increases with food intake, and PYY 3-36 decreases.

A: This is the true PYY. I should also add that this seems to be a phenomenon that is only present in rodents. When we’ve given the drug to primates, we have a 65-70% reduction of receptors and we’ve observed increased FGF19, but no change in PYY. So this effect is localized to rodents.


Preston Klassen, MD (Orexigen, La Jolla, CA)

Dr. Preston Klassen began his presentation by applauding the recent approval of new obesity medications, stating that it is a “fantastic time for the obesity space,” (what a difference a year makes!) and hoping that the recent approvals will encourage more investment in early-stage obesity drug development. For the majority of his presentation, he discussed the challenges of obesity drug development, using Contrave as an example. He noted that the ultimate challenge is more about the FDA accepting the benefits of weight loss than it is about understanding the theoretical risks, emphasizing that an opportunity exists to expand recognition of benefits beyond the cardiometabolic benefits of weight loss. In addition, companies have the challenge of figuring out how to incorporate the evaluation of low frequency adverse events (e.g., cancer, CV events, birth defects) into their development programs; Dr. Klassen hoped this would be clarified in future guidance, and stated that the FDA will first address the issue of cardiovascular safety. As an example of how a company could undertake such an evaluation, he described the Light Study (the cardiovascular outcomes trial for Contrave), reviewing the study framework, targeted background event rate, enrollment criteria, and rules for therapy discontinuation (see our Closer Looks on the study at,, and As the press release from Orexigen earlier this month stated, Dr. Klassen said that in the 14 weeks since study enrollment began, 4,500 patients have entered the study – we assume there have been further growth but that Orexigen is not reporting it here. After reflecting on Contrave’s development and regulatory history, Dr. Klassen suggested that other pharmaceutical companies perform early evaluation of cardiometabolic endpoints, focus part of their phase 3 programs on specific populations of interest, and explore the clinical benefit of weight loss beyond the current efficacy guidance (e.g., sleep apnea, quality of life, pain).

  • Dr. Klassen explained that the current regulatory guidance implies that the primary benefit of weight loss is to reduce cardiovascular disease, as the FDA’s assessment of efficacy focuses on weight loss and its surrogate cardiometabolic benefits. However, no weight- loss medication has been shown to improve cardiovascular outcomes to date. Unfortunately, other potential benefits of weight loss are either not validated (e.g., joint pain and quality of life), or are limited to narrow patient segments (e.g., sleep apnea and diabetes). Dr. Klassen pointed to the FDA’s presence at the roundtable discussions at George Washington University as evidence that the agency is working to expand its view of the benefits of weight loss.
  • He noted that the challenge is more about the FDA accepting the benefits of weight loss, e.g., about understanding the need, the clinical benefit, and the overall importance of weight loss and the degree to which weight loss is related to improving other co-morbidities. As an example of this, he pointed to bupropion (a component of Contrave), which is approved for major depressive disorder (MDD), seasonal affective disorder (SAD), and smoking cessation, despite its modest increases in heart rate and blood pressure. Bupropion, Dr. Klassen argued, has an identical hemodynamic-related risk profile when used as a component of Contrave for treating obesity, or as monotherapy for other conditions like MDD, SAD, and smoking cessation. The benefit of Contrave’s associated weight loss, however, he said, is not as established for the FDA as is bupropion’s value in treating other conditions.
  • Dr. Klassen explained the challenges of selecting the patient population for the Light Study. Ideally, one would use the general obese population to make the trial representative of the treatment population, but then one would have a very low event rate (~0.2%) and it would take too long to obtain results (i.e., enough events). On the other hand, if one selects a high background rate (~3-5%), then the findings from the high-risk study population might not be translatable to the treatment population. Thus, Orexigen elected for the “Goldilocks approach” and are targeting a 1.5% event rate, enrolling patients with a history of cardiovascular disease, or diabetes with at least one risk factor (for more information, please see our July 11, 2012 Closer Look at It will be interesting to see what happens on this front.
  • Importantly, after 16 weeks of treatment, anyone in the Light Study who has not lost 2% of their baseline weight or who has a clinically meaningful increase in blood pressure will be discontinued from treatment. Orexigen will still include these patients in their intent-to-treat analyses.

Questions and Answers

Q: I would like to clarify one thing about your discussion of high-risk populations. Having heard some of the Advisory Committee meetings, one of the main criticisms of why the overall adverse events were so low was that high-risk patients were not included. Therefore, there was a definite suggestion to include them in phase 3 programs to show no harm. Did you discuss this with the FDA? If so how did it go?

A: I completely agree. Back when our original trials were designed, this guidance was not out there. So when you do not have guidance telling you to enroll high-risk patients, who do you enroll? People who want to enroll – predominantly Caucasian women, around the age of 45, generally speaking. The paradigm has changed. Companies need to think what proportion needs to be at high cardiovascular risk. What I can tell you is that with a mid level of risk (1.5% per year), we’re conducting a seven-to-ten thousand-patient study. It has to be done in a streamlined way.

Q: How are you measuring the blood pressure and the heart rate for these patients?

A: Resting blood pressure and heart rate are measured in standard clinic visits – they’re measured three times, then averaged. But, this is a streamlined trial that is designed for clinical events. We already know a lot about the drug’s impacts on blood pressure and heart rate. I now need to know what’s happening in regard to cardiovascular events. So we’re not collecting a lot of lab data – we did that in phase 3. It’s really looking at whether the patients are on the drug, whether they’re using it, whether they’re coming in for visits, and whether they’re having events.

Q: Doing a non-inferiority trial with responder analysis where therapy is discontinued gets really tricky. So in your trial, if people reach week 16 and do not respond, they are still included in the intent-to-treat analysis as if they were on drug. Isn’t that a problem in a non-inferiority study, since individuals not exposed to drug are still counted?

A: You’re spot on. This was a huge discussion we had with the FDA, and they’re in full agreement that that is appropriate for the design. If you want to do the most rigorous statistical analysis, you do it like SCOUT, and have patients stay on the study drug no matter what. What you end up with is 70% of patients on sibutramine not losing enough weight to be clinically meaningful. At the end of the day, what does that mean? You don’t know much, because it doesn’t reflect real world use of the drug. Or, you can take the approach of doing a responder analysis, which mimics real world use, but then you need to be more flexible in how you do the analysis. The intent-to-treat analysis will be primary, but of course they would also look at per-protocol analysis.


Chris Adams (Compellis Pharmaceuticals, Boston, MA)

In his presentation, Compellis Pharmaceuticals CEO and founder Chris Adams revealed new phase 1 data for the company’s novel obesity medication CP404 – a repurposed, intranasal formulation of diltiazem, a calcium channel blocker that has been on the market as an antihypertensive drug for roughly 30 years. CP404’s development rests on the theory that blocking a person’s sense of smell results in decreased food intake and increased weight loss. The drug is administered nasally, passes over the olfactory epithelium, and blocks odorant signal transduction pathways to the brain. Mr. Adams noted that because CP404 provides roughly 10% of diltiazem’s typical oral dose (resulting in a ≤4 mg nasal dose), he believes any systemic effects will be minimized. In preclinical rodent studies, CP404 prevented weight gain and decreased smell perception, with no effect on motor capability or activity. Compellis has also conducted one pilot study that assigned 12 participants to 2 mg, 4 mg, or 8 mg doses of CP404 or to placebo. Results of the vision analog scale of smell suggest that CP404 decreases smell acuity and may influence patients’ sense of satiety. Regarding safety, CP404 had no effect on systolic or diastolic blood pressure. While some participants reported nasal irritation, Mr. Adams stated that this side effect is likely due to the CP404’s formulation rather than its intrinsic effects. Compellis plans to administer CP404 in combination with topically applied lidocaine (a local anesthetic; a sodium channel blocker that has been shown to decrease smell acuity and food intake) and expects the combination therapy to provide at least a 5% weight loss in patients. We are eager to know what the drug could do in combination, since we assume it would be primarily additive.

Questions and Answers

Q: We’ve given these drugs to patients for hypertension and they have been given at pretty high doses. You’d think that at those doses, some of the drug reaches the areas in the CNS that your compound does. But are there weight-loss effects reported with these compounds? Also, knowing that many of these compounds have side effects, have you had a chance to evaluate which is the best one? Was diltiazem the best choice among the calcium channel blockers? Have you done comparative studies?

A: We have not. We wanted to first demonstrate the effects of blocking the calcium channel. Once we’ve generated that data, we’ll go back and screen more proprietary compounds that have more potent effects. Regarding your first question, you’ll need far too much drug to have that effect [on the CNS].

Q: I was wondering whether you looked at side effects such as depression or other quality of life measures?

A: We have not seen changes in depression to date, though we have a very small sample size. Clearly quality of life is impacted by olfactory nerve damage. Fortunately people with olfactory nerve damage tend to have a healthy weight, which we think is because of satiety.

Q: Why didn’t you measure food intake in those subjects for which you measured the VAS [visual analog scale]? Also, the efficacy of this drug is predicated upon the lasting inhibitory effect in the long run. Do you have any data on this?

A: We don’t have any data on the long-term effects – we haven’t done those studies. The reason why we didn’t look at food intake in that study was because it was so small. We wanted to show the FDA that the drug was safe and that people weren’t affected by the drug.

Q: I ask because VAS has a lot of problems, so using that data is a large leap of faith.

A: That was our phase 1a study. So our phase 1b study will obviously measure food intake. The object here is to look at a combination of lidocaine and diltiazem and target both channels at the same time.


Keith Gottesdiener, MD (CEO, Rhythm)

Dr. Keith Gottesdiener discussed the potential of the melanocortin-4 receptor (MC4R) agonist RM-493 for the treatment of obesity, highlighting key findings from animal studies to date. He noted that MC4R is a particularly interesting target, because it is the final common pathway for multiple divergent pathways involved in feeding and maintenance of body weight. In animal studies, RM-493 has demonstrated promising weight loss (13% during an eight-week treatment period in primates), increased energy expenditure, and improved insulin sensitivity. While the efficacy and safety of RM-493 in humans are not yet well characterized, Dr. Gottesdiener seemed confident that the drug would not cause adverse effects on heart rate and blood pressure, based on primate study data. Rhythm is currently finishing phase 1 studies for RM-493, and the company intends to provide a clinical update in early 2013. Dr. Gottesdiener noted that the early data have been “very promising” with respect to the drug’s cardiovascular safety profile.

  • In a dog model, RM-493 treatment suppressed food intake, and resulted in an approximate 10% weight loss. The dogs were administered subcutaneous injections of RM- 493 at a dose of 4.5 umol/kg/day three times daily for periods of three days, with four-to-five day washouts between doses. During washout, the dogs stopped losing weight and their weight stabilized to some degree, but never returned to baseline. Over a number of treatment/washout periods, the dogs in the study lost around 10% of their initial body weight.
  • In an eight-week primate study, subjects (n=12) lost 13% of their weight. The obese, insulin-resistant, eight-to-ten-year-old male rhesus monkey subjects were treated with vehicle for four weeks at baseline, then with 450 nmol/kg/day of RM-493 for eight weeks, followed by vehicle for an eight-week washout period. Dr. Gottesdiener noted that there was a consistent treatment effect, as every animal in the study demonstrated profound weight loss. The 13% weight loss didn’t occur at the end of the eight-week treatment period; in general, maximum weight loss was achieved two-to-three weeks, or even sometimes four weeks after the end of the actual treatment period. In addition, improvements in glycemic control were also sustained after the end of treatment for as long as 12 weeks. Subjects in the study decreased food intake by as much as 40% for the first three weeks of treatment, then returned to baseline, yet the animals continued to lose weight; Dr. Gottesdiener stated that the significant weight loss they experienced was attributable to fat loss, increased activity and energy expenditure, and improved insulin sensitivity.
  • The efficacy and safety of MC4R agonists in humans have yet to be fully characterized; however, Dr. Gottesdiener seemed optimistic RM-493 does not have adverse effects on blood pressure and heart rate. Merck’s phase 2a and food intake studies were not successful at demonstrating efficacy; however, it is possible that the candidate they selected was not ideal for these purposes. In terms of safety, data to date suggest that MC4R agonists could cause erections in humans, effects on heart rate or blood pressure, nausea, and/or off-target effects (e.g., MC1R agonism could result in skin tanning). Lilly’s MC4R agonist LY2112688 was shown to increase blood pressure by an average of 9.3 mmHg 24 hours after dosing, and other side effects were noted, including increased heart rate, yawning, stiffness, erections, and nausea (Greenfield et al., NEJM 2009). LY2112668 was subsequently withdrawn from clinical development, and some came to the conclusion that increased blood pressure was an unavoidable class effect of MC4R. Encouragingly, in Rhythm’s obese primate model, subjects did not vomit or show signs of nausea, had no changes erectile activity, and no skin darkening with RM-493 treatment. In terms of cardiovascular effects, heart rate decreased, as did diastolic blood pressure; meanwhile, systolic blood pressure was somewhat variable. Investigators subjected the same primates to LY2112688 treatment; the primates experienced increases in blood pressure with that drug, versus no increase with RM-493.

Questions and Answers

Q: Why do you think your compound is so different than Lilly’s compound?

A: I’m not going to speculate a lot at this point. Clearly one of the most important things for us to do is to help nail down the mechanisms of these compounds. It’s not entirely surprising to me that we saw different effects with RM-493 and previous MC4R candidates. I worked on the Merck compound. We put it into rhesus monkeys, and it worked wonderfully, but it did little in humans. We pushed doses higher than anticipated, and there was no weight loss, no erections, nothing. Meanwhile, Pfizer said their candidate was as good as Viagra as an erectogenic. Lilly saw few excess erections. Clearly there are important ways these compounds differ, and one of the most critical things is to work on the science behind it.

Q: Is there something inherently different about your compound than previous candidates in the class that would explain the differential blood pressure effects?

A: It’s something we’ll be looking into closely. We saw 13% weight loss, and an approximate 10% decrease in blood pressure in our eight-week primate study. It’s very hard to know whether the blood pressure effect of RM-493 was tied to the weight loss, or if the drug had an intrinsic ability to decrease blood pressure.


Nozer Mehta, PhD (Unigene, Boonton, NJ)

In his presentation, Dr. Nozer Mehta discussed the rationale and progress of Unigene’s satiety peptide program, which aims to develop an oral peptide therapy for obesity. Following the discovery that salmon and eel calcitonin are agonists of the amylin receptor, Unigene selected analogs of the calcitonin/amylin family of peptides that exhibit increased binding to the amylin receptor, decreased binding to the calcitonin receptor (to minimize bone effects), and reduced food intake and body weight in preclinical models. The program’s lead compound UGP281 exhibits a 50% reduction in calcitonin receptor binding, but demonstrates greater potency for the amylin receptor compared to salmon calcitonin. In rat studies, UGP281 given via injection (5 μg/kg to 20 μg/kg) resulted in dose-dependent reductions in body weight and food consumption. Similar results were observed during in vivo studies in dogs. Dr. Mehta then described Unigene’s oral delivery technology: an enteric coating prevents the tablet from opening in the stomach but dissolves within the neutral pH environment of the small intestine, releasing the peptide drug along with protease inhibitors and permeability enhancers. The peptide is then absorbed across the intestinal wall via paracellular transport. Dr. Mehta noted that oral dosing of UGP281 in dogs resulted in an 8-9% reduction in body weight over a five-week period. Dr. Mehta then ended his presentation by reviewing the recombinant manufacturing technology that he believes will enable Unigene to make large quantities of UGP281 at a relatively low cost. While Dr. Mehta did not mention the initiation of clinical development for UGP281, it is our understanding that Unigene expects to begin phase 1 studies in 1H13.

Questions and Answers

Q: Calcitonin has been used for years in humans to treat osteoporosis. There are several formulations. I’ve never observed weight loss in humans, and I presume it’s never been published. Can you explain why it has never been observed?

A: First, people weren’t looking for it. Most calcitonin used is nasal calcitonin and the dose used is much lower and much more variable. There are some early reports that showed that when injectable calcitonin is given, you do see weight loss. But also, the analogs we’re developing are much more potent that typical calcitonin.

Q: How much higher are the calcitonin doses that you are using? Have you seen any adverse effects on the bone?

A: We still need to do appropriate dose-finding studies. At the doses we were using, we were already getting a reduction in food intake. We don’t know what the optimal dose will be. The nasal dose is about 200 IU, which is roughly 33 micrograms.

Q: Can you go over your mechanistic rationale for choosing calcitonin?

A: Studies published in the literature found, for reasons that I am not entirely clear on, that calcitonin binds to the amylin receptor. There is a lot of variation between calcitonin in different species. There is about a 50% difference between human calcitonin and salmon calcitonin. For some reason that 50% is enough that the animal calcitonin binds to receptors the human calcitonin doesn’t.

Q: Are there any effects on glucose control?

A: We haven’t looked at that very rigorously. We have done an OGTT experiment and we do see an effect on glucose – there may be some beneficial effects in that regard. But we want to compare the molecule to others and look at different doses before we publish any data.

Q: Where do you think calcitonin is working?

A: Classically amylin receptors are in the brain. There is also some evidence that there are receptors, not just amylin receptors, in the GI tract. We have not yet investigated that, and have no conclusions on if that is true.

Oral Abstract Session: Co-Agonism of GLP-1 and Glucagon Receptors


James Plumer (Imperial College, London, UK)

James Plumer reviewed the development and preclinical testing of a peptide that simulates both the GLP-1 and glucagon receptors – this was one of the sessions to which we were looking most forward at TOS, although the fact that all the research is still pre-clinical reinforces the very early stage – we were a bit disappointed it was not further evolved. The rationale for a dual GLP-1/glucagon agonist is based on oxyntomodulin, a naturally occurring dual agonist released from intestinal L cells that reduces food intake and body weight. In Mr. Plumer’s lab’s attempt to develop a peptide that promotes appetite suppression and insulin secretion (via the GLP-1 receptor) while also increasing energy expenditure (via the glucagon receptor), they created a chimera peptide by combining sequences of glucagon and exendin-4 (a peptide analogous to GLP-1 that is currently marketed as exenatide [Byetta]). Mr. Plumer explained that the glucagon receptor selectively binds to the N terminus of glucagon while the GLP-1 receptor recognizes the C terminus of exendin-4. Thus, scientists can combine the receptor-binding ends of the glucagon and exendin-4 peptides to create a viable dual agonist that is more resistant to protein degradation due to the structural differences between the two parent peptides. The lab’s lead dual agonist GX7 has similar potency to GLP-1/exendin-4 and glucagon at the GLP-1 and glucagon receptors, respectively. In rodent models, single and repeated administration of GX7 produced a sustained anorexic effect. Subsequent pair-feeding studies showed that GX7 has effects on body weight that are independent of food intake and that are likely due to the increased energy expenditure caused by stimulation of the glucagon receptor. To end his presentation, Mr. Plumer discussed future directions for his lab’s work, noting that tests measuring glucose tolerance and insulin sensitivity are needed to better understand the seemingly conflicting effects of GLP-1 and glucagon on blood glucose levels (GLP-1 agonism stimulates insulin secretion while glucagon agonism raises glucose levels). His lab also plans to study the effects of chronic GX7 administration in diet-induced obese animals. As a reminder, other companies developing GLP-1/glucagon dual agonists include Zealand Pharma (ZP2929; phase 1 recently initiated) and Transition/Lilly (TT-401; phase 1).

Questions and Answers

Q: What cells did you use with the overexpressing receptor for performing the cyclic AMP assay?

A: Human embryonic kidney cells, which overexpressed the human receptors.

Q: When you performed the pair feeding, did you measure energy expenditure? Is it at night, is it during the day, is it activity? This is only pair feedings – you should use calorimetry tests to measure the energy expenditure.

A: The effects of oxyntomodulin have been shown to increase energy expenditure through oxygen consumption, and oxyntomodulin was shown through investigations in London to increase spontaneous activity (fidgeting) in humans. This effect of glucagon on energy expenditure has been shown. But I agree, we do need to show the mechanism of how this glucagon receptor activity is translating into increased weight loss. It needs further investigation. To be honest, it’s not well understood at the moment.

Q: The logic of combining these two peptide hormones is that they’re evolutionarily related. Did you try to build an evolutionary consensus peptide?

A: GLP-1 and glucagon are cleaved from the same gene and have evolved individually to have two separate effects. What has actually happened is that the glucagon receptor is dependent on action on glucagon’s N terminal region while the GLP-1 receptor selects for GLP-1 receptor ligands by binding to the C terminals of the molecules. So it’s fortunate that you can combine the ends of these two sequences to form a peptide that binds to both receptors.

Q: Do you expect that the chimera peptide will stay stable in circulation, or will it get cleaved into two parts that will act on their individual targets? If it stays stable, you’d imagine that the molecule would be very large. Most of the effects on food intake and energy expenditure will come from actions at the brain. How will the large molecule get across the blood brain barrier? How does that compare to individual peptides?

A: We’ve currently got exendin-4 on the market as exenatide (Byetta), a daily treatment, and we know that exendin-4 is resistant to degradation. It has evolved to be resistant to protein degradation. The fact that we’ve combined glucagon with GLP-1 means that resistance to protein degradation is already there. We’ve shown in rodent PK studies that the chimera peptide does last over eight hours at functional levels. Regarding access in blood to the brain, both glucagon and exendin-4 function to inhibit food intake at the brain, so we know that they cross the blood brain barrier. The glucagon effect is also thought to be through actions at the liver and adipose tissue. So we’ll see effects through the blood brain barrier – that’s not an issue with these compounds.


Oral Abstract Session: Exercise, Metabolic Disease Risk, and Obesity


Arya Sharma, MD, PhD (University of Alberta, Edmonton, Canada)

On behalf of SCOUT (Sibutramine Cardiovascular Outcome Trial) investigators, Dr. Arya Sharma presented new results from an analysis showing that individuals randomized to sibutramine, who had the greatest falls in blood pressure had a higher incidence of cardiovascular events (nonfatal myocardial infarction, nonfatal stroke, resuscitated cardiac arrest, and cardiovascular death). Since sibutramine has the potential of increasing blood pressure, the investigators assumed that participants who experienced the greatest elevations in blood pressure would have been at the highest risk of experiencing an event. On the contrary, people receiving sibutramine whose systolic blood pressure (SBP) remained unchanged or increased slightly (-2.5 to 5.5 mmHg) had the lowest incidence rates (10.0%) of primary outcome events (POE) over five years, while those on sibutramine who had the greatest decreases in SBP (<-11.5 mmHg) had the highest POE incidence (13.8%). In contrast, the placebo arm had a U-shaped risk distribution, where patients who had the greatest SBP increases (≥5.5 mmHg) and greatest SBP decreases (<-11.5 mmHg) had a higher five-year incidence of events than those with modest changes in SBP. Acknowledging he did not have an explanation for the findings, Dr. Sharma sought the help of the audience – while several suggestions were proposed, none adequately explained the counterintuitive findings.

  • During the six-week lead-in period, both the placebo arm and the sibutramine arm experienced decreases in systolic (SBP) and diastolic blood pressures (DBP) and a slight increase in pulse. Overall changes in blood pressure and pulse were balanced between the two arms at the end of the lead-in period.

Mean Change in Vital Signs from Lead-in Period Baseline Through Randomization Phase Baseline


Overall (n=9,804)

Placebo (n=4,898)

Long-term sibutramine (n=4906)

BMI (kg/m2)




SBP (mmHg)




DBP (mmHg)




SBP/DBP ≥140/90

mmHg (%)




Pulse (bpm)




  • Counterintuitively, the more systolic blood pressure decreased from the lead-in period baseline for patients in the sibutramine arm, the more likely they were to have a primary outcome event (POE) through year five. This was a surprising finding for the investigators, as they had thought that sibutramine caused POEs by elevating blood pressure. For comparison, the placebo arm had a U-shaped curve of risk – while the placebo arm also had an elevated POE incidence (11.4%) in those with the greatest drop in SBP, the POE incidence was also high in those who had the largest increase in SBP (10.8%).

Primary Outcome Event (POE) Incidence in the Placebo and Sibutramine Arms by SBP Change Quartiles

Change in SBP (mmHg)

Placebo POE Incidence (%)

Sibutramine POE Incidence (%)




-11.5 to -2.5



-2.5 to 5.5






Questions and Answers

Q: Did the people who had the biggest fall in blood pressure have the highest baseline numbers?

A: There were no obvious statistical differences, but it is hard to do a lot of that sub-analysis because the numbers get very small.

Comment: Older people with many comorbidities tend to have stiff arteries and it is thought to be a bad idea to reduce their blood pressure.

A: The puzzling piece with that explanation is that you do not see the same trend in the placebo group. So simply losing weight and blood pressure was not enough in the placebo group to increase the POE incidence.

Q: Could it be because of co-interventions?

A: Probably not. If anything, there was a down titration of antihypertensive medications in both groups.

Q: There was a women’s health study that looked at the impact of hormone therapy on plaque, which suggested that particularly in older women it might be problematic due to destabilizing their plaque. Could sibutramine be impacting plaque stability?

A: Your guess is as good as mine.

Q: I am wondering if there was a similar trend in heart rate, since sibutramine increased heart rate.

A: We looked at that. There was a pretty consistent increase by one to two beats per minute in the sibutramine group, but beyond that it didn’t seem to matter.


Oral Abstract Session: Drugs/Surgery


Brian Murphy, PhD (Bristol Myers Squibb, Princeton, NJ)

Dr. Brian Murphy presented disappointing early stage clinical results for Bristol Myers Squibb’s (BMS’s) melanin concentrating hormone receptor 1 (MCHR1) antagonist, BMS-830216 (which is then modified within the body to become the active ingredient BMS-819881). Preclinical studies had found that intracerebroventricular (ICV) injection of melanin concentrating hormone (MCH) caused rats to increase food intake by 8-11%, while mice MCHR1 knockouts were lean, had reduced fat stores, and increased resistance to diet induced obesity. Based on these positive results, BMS performed two clinical trials. The first, a phase 1 randomized, double-blind, placebo controlled, single ascending dose study in healthy participants (n=6 for each dose and 12 for placebo) found BMS-830216 to be well tolerated, and successfully converted into BMS-819881. The second phase 1/2 trial, however, found that in obese individuals the drug was associated with no effect on food intake and, if anything, an increase in weight. This trial was also randomized, double-blind, and placebo controlled, but was performed in obese individuals, and had two parts: a multiple-ascending dose (MAD) portion (n=6 for each dose and 10 for placebo) and a proof-of-concept component (POC; n=24 for 300 mg BMC-30216, 600 mg BMC- 30216, and placebo). In the MAD arm, mean body weight was similar to baseline throughout the study, while in the POC study the 300 mg arm gained 0.74 kg over the placebo (p=0.06) and the 600 mg arm gained 0.81 kg (p=0.13). Dr. Murphy hypothesized that the discrepancy between the preclinical and clinical results may be because rodents do not have MCHR2 receptors but humans do. Though BMS- 819681 is specific for the MCHR1 receptor it is possible that this redundancy in MCH receptors in humans may have altered its effect.

  • The phase 1 study of BMS-830216 ( Identifier: NCT00878020) found the drug to be well tolerated and that it was successfully converted into BMS- 819881. Healthy participants (mean age was 29.8 years; 92% were male; mean BMI was 25.8 kg/m2) were randomized to one of five treatment doses (10 mg, 30 mg, 100 mg, 300 mg, 600 mg; n=6 for each dose and 10 for placebo).
  • The phase 1/2 trial ( Identifier: NCT00878020) found that in obese individuals the drug was associated with no effect on food intake and, if anything, an increase in weight. This trial consisted of two randomized, double-blind, placebo-controlled parts: a multiple ascending dose study and a proof-of-concept study. Both studies enrolled adults (mean age was 39.8 years; 40% were male) who were obese (mean BMI was 34.24 kg/m2). Participants of both studies received a loading dose before the treatment phase began. The most commonly reported adverse events were fatigue (17.7%), nausea (15.2%), headache (11.4%), and photosensitivity reduction (11.4%). Participants reported no changes in anxiety or suicidality while on BMS-830216.
    • Participants of the multiple ascending dose study received doses of BMS- 830216 daily for four weeks, the first two of which were inpatient. Six people received each dose (from 10 mg to 600 mg daily), and ten received a placebo.
    • Enrollees in the proof-of-concept study received their dose daily for four weeks outpatient. The two BMS-830216 doses tested (300 mg and 600 mg) and the placebo arm each had 24 participants.

Questions and Answers

Q: Did you take CNF levels?

A: CNF was not collected.

Q: We have seen that some of these antagonists affect serotonin receptors and can cause anxiety or depression, did you look at that?

A: There were no signals of changes in physical activity levels. We also did not have any adverse events come up about anxiety or depression.


Poster Presentations


R Dvorak, C Peterson, and W Day

To better characterize the benefit/risk profile of phentermine 7.5 mg/topiramate ER 46 mg (Vivus’ recommended dose of Qsymia) and the drug’s titration algorithm, Vivus conducted a post-hoc, retrospective analysis of the CONQUER, EQUIP, and OB202/DM230 studies. The study compared weight loss, cardiometabolic parameters, adverse events (AEs), and discontinuations due to AEs among 12-week responders to the recommended dose (>3% weight loss; n=348), 12-week non-responders (<3% weight loss; n=150), and those on placebo (n=1561). Not surprisingly, at week 56, those who responded to the recommended 7.5 mg/46 dose lost more weight (-11.4%) than those who did not respond (-2.2%; p-value not provided). Responders also had substantially greater improvements in blood pressure, HDL, triglycerides, and fasting insulin. Most notably, however, was the adverse event data – AEs leading to study drug discontinuation were actually lower in 12-week Qsymia responders (4%) than in those on placebo (8.4%), and substantially lower than in Qsymia non-responders (29%). This suggests that the FDA approved titration algorithm does indeed maximize benefit and minimize risk, which we think is very encouraging from a regulatory, payer, and patient perspective (i.e., it seems like those who respond well to Qsymia are also those who experience fewer side effects, which should help improve adherence in the long-term as well as filter out non-responders and save payers money).

  • As a reminder, Vivus’ proposed treatment algorithm (PTA) for Qsymia is as follows: 1) patients start on an initial starting dose (3.75 mg phentermine/23 mg topiramate ER) once daily for two weeks; 2) then, the recommended dose (7.5 mg/46 mg dose) is taken for 12 weeks; and 3) at 12 weeks, those with >3% weight loss (responders) stay on the recommended dose, while those with <3% weight loss (non-responders) either discontinue Qsymia or increase the dose (to the titration dose of 11.25 mg/69 mg for two weeks followed by the top dose of 15 mg/92 mg for 12 weeks).
  • Investigators performed a post-hoc analysis of clinical trial data to characterize the benefit/risk profile of Qsymia’s recommended dose (phentermine 7.5 mg/topiramate ER 46 mg). The analysis included patients from the CONQUER, EQUIP, and DM-230 (a 28-week extension of a 28-week phase 2 trial of overweight/obese people with type 2 diabetes) studies (n=498). Weight loss, cardiometabolic parameters, and adverse events leading to study drug discontinuation were compared among those receiving placebo (n=1,1561), responders to the Qsymia recommended dose (>3% weight loss; n=348) and non-responders (<3% weight loss; n=150). Only participants in the CONQUER trial (n=488) were used to determine efficacy of the recommended 7.5 mg/46 mg dose.
  • There were no differences in baseline demographics or clinical characteristics between the placebo group, responders, and non-responders. The average BMI of the phentermine 7.5 mg/topiramate ER 46 mg arm was 36.2 kg/m2, compared to 36.7 kg/m2 for placebo (n=994). Participants in both groups had a mean age of 51 years, were 70% women, and 86% Caucasian.
  • Of the 417 people who received the recommended dose of phentermine 7.5 mg/topiramate ER 46 mg and remained in the study through week 12, 83.5% (n=348) were responders. According to the poster, however, 71 non-responding participants withdrew from the study prior to week 12. This group was not included in the calculation of how many recipients of the recommended 7.5 mg/46 mg dose responded at week 12. According to our calculations, if these individuals were included (a total n=488), 71% of participants were responders over the 12-week period.
  • At week 56, responders lost much more weight (-11.4%) than non-responders (-2.2%; p-value not provided), and had greater improvements in cardiometabolic measures. Particularly dramatic was responders’ decline in triglycerides (-15.2%) compared to an increase in non-responders’ (+0.7%). In this case, non-responders included those who did and did not stay in the study through week 12 (n=140). No p-values were provided for comparison.

Weight loss and changes from baseline in cardiometabolic measures in responders and non-responders after 56 weeks of phentermine 7.5 mg/topiramate ER 46 mg treatment.

Measure Responders Non-responders
Least square mean percent weight loss (%) -11.4 -2.2
SEP (mmHg) -6.1 01.6
DBP (mmHg)  -3.4 -0.8
HDL-C (%) +8.0 +0.2
Triglycerides (%) -15.2 +0.7
Fasting insulin (ulU/ml) -4.5 -1.7
  • Notably, adverse events leading to study drug discontinuation were actually lower in 12-week Qsymia responders (4%) than in those on placebo (8.4%), and substantially lower than in Qsymia non-responders (29%). We’re glad to see that the FDA approved treatment algorithm appears to maximize benefit and minimize side effects, a big plus for patients and certainly a good sign as payers consider covering the drug. We would also be interested to see this same analysis conducted for responders and non-responders at the highest dose of Qsymia.

Incidence of adverse events leading to drug discontinuation over 56 weeks

Adverse event (%)











Nervous system disorders




Psych disorders




GI disorders




Renal and urinary disorders




Cardiac disorders





E Vargas, M Sanchez, W Shanahan, C Anderson

Analyzing data from the phase 3 program for lorcaserin, this poster demonstrated that a higher percentage of patients (both those with and without type 2 diabetes) receiving lorcaserin decreased the use of concomitant medications for dyslipidemia, hypertension, and type 2 diabetes compared to placebo over a one-year treatment period (the authors did not state whether any of the reductions were statistically significant). The vast majority of patients in the trial on concomitant medications did not increase or decrease the doses of their medications over the course of the studies.

  • In the phase 3 program for lorcaserin, a higher proportion of patients receiving study drug reduced their daily dose of antidyslipidemic, antihypertensive, and antihyperglycemic medications compared to placebo over the one-year treatment period.

Patients with Type 2 Diabetes


Non-Diabetic Patients (Pooled



Placebo (N=252)

Lorcaserin 10 mg

BID (N=256)



Lorcaserin 10 mg

BID (N=3,195)

Change in Total Antidyslipidemics Daily Dose n (%) [%]








6 (2.4) [3.8]

14 (5.5) [8.9]

23 (0.7) [3.8]

43 (1.3) [7.0]


No Change

137 (54.4) [85.6]

136 (53.1) [86.1]

474 (14.9) [78.2]

484 (15.1) [79.3]



17 (6.7) [10.6]

8 (3.1) [5.1]

109 (3.4) [18.0]

83 (2.6) [13.6]

Change in Total Antihypertensives Daily Dose n (%) [%]








15 (6.0) [9.3]

21 (8.2) [11.9]

54 (1.7) [7.3]

70 (2.2) [9.5]


No Change

131 (52.0) [80.9]

138 (53.9) [78.4]

595 (18.7) [80.0]

594 (18.6) [80.9]



16 (6.3) [9.9]

17 (6.6) [9.7]

95 (3.0) [12.8]

70 (2.2) [9.5]

Change in Total Antihyperglycemic Daily Dose n (%) [%]








29 (11.5) [11.6]

43 (16.8) [16.9]

0 (0.0) [0.0]

1 (0.0) [5.3]


No Change

165 (64.5) [66.3]

177 (69.1) [69.7]

8 (0.3) [57.1]

14 (0.4) [73.7]



55 (21.8) [22.1]

34 (13.3) [13.4]

6 (0.2) [42.9]

4 (0.1) [21.1]

*Values in parentheses use total N (overall study population) as the denominator; values in brackets use n (total number of subjects who took any medication between Week 0 and Week 52) as the denominator. For medication (antidyslipidemics and antihypertensives) with missing doses, no interpretation of change was made.

  • Patients in the BLOOM-DM trial on lorcaserin experienced reductions in the use of concomitant antidiabetic medications over the yearlong trial, whereas patients on placebo experienced an average increase in daily dose for metformin, SFUs, and TZDs. Those on placebo on a DPP-4 inhibitor decreased their daily dose of the concomitant medication by an average 6.9% over the course of a year.


Lorcaserin 10 mg BID



% Change in Daily Dose


% Change in Daily Dose





















  • Over the one-year treatment period in the BLOOM and BLOSSOM trials, a numerically smaller percentage of patients on lorcaserin initiated medications to treat dyslipidemia, diabetes, or hypertension than patients on placebo.






Lorcaserin 10 mg BID


Lorcaserin 10 mg BID


Lorcaserin 10 mg BID

N (%)








38 (2.4)

28 (1.8)

42 (2.6)

33 (2.1)

80 (2.5)

62 (1.9)


5 (0.3)

4 (0.3)

1 (0.1)

0 (0)

6 (0.2)

4 (0.1)


20 (1.3)

13 (0.8)

24 (1.5)

22 (1.4)

44 (1.4)

35 (1.1)


D Kim, J Vath, A Chen, J Marjason, T Hughes

In this four-week phase 1b study in obese non-diabetic women (BMI 30-50 kg/m2; average baseline weight 92-99 kg [202-218 lbs]), twice-weekly subcutaneous (SC) beloranib (Zafgen’s MetAP2 inhibitor – please see our Closer Look at for more on beloranib) provided significant weight loss of approximately 1 kg (2.2 lbs)/week for those treated with 1.0 mg or 2.0 mg SC beloranib (an average 4.25 kg [9.3 lbs] and 4.23 kg [9.3 lbs] weight loss at four weeks); those treated with 4.0 mg SC beloranib achieved even greater reductions in weight (an average 6.1 kg [13.4 lbs]) from baseline (those on placebo achieved an average 1.2 kg [2.6 lbs] weight loss from baseline). In addition, SC beloranib treatment led to significant improvements in LDL-c, triglycerides, and CRP versus placebo. Participants receiving SC beloranib experienced a higher frequency of adverse events; four patients receiving SC beloranib (one on the 2.0 mg dose, three on the 4.0 mg dose) withdrew from the study due to sleep disturbance. In general, adverse events were mild and self-limiting in nature, and occurred more frequently in the higher-dose arms. As such, the authors concluded that doses less than 4.0 mg seem to have the most favorable efficacy/tolerability profile.

  • In the four-week phase 1b study, subcutaneous (SC) beloranib treatment led to significant improvements in weight, LDL-c, and triglycerides, and CRP versus placebo. Obese (BMI 30-50 kg/m2) non-diabetic adult females (18-65) were administered placebo (n=6), or 1.0 mg (n=6), 2.0 mg (n=4), or 4.0 mg (n=4) SC beloranib twice weekly, without any accompanying lifestyle intervention. At baseline, this per-protocol population was on average 46-53 years old, 92-99 kg (202-218 lbs), and had a mean BMI of 35-36 kg/m2. Patients receiving 1.0 mg or 2.0 mg SC beloranib experienced a steady weight loss of approximately 1 kg (2.2 lbs) per week, culminating in respective 4.25 kg (9.3 lbs) and 4.23 kg (9.3 lbs) losses at the end of the four-week study. Hunger (as assessed by a visual analog scale) decreased significantly from baseline with SC beloranib treatment versus placebo, suggesting reduction in hunger as a possible mechanism for the observed weight loss. Meanwhile, those receiving 4.0 mg beloranib achieved a 6.1 kg (13.4 lbs) weight loss over four weeks. For comparison, those on placebo experienced an average 1.18 kg (2.6 lbs) weight loss from baseline. In addition, those receiving SC beloranib experienced reductions in LDL-c from baseline (21-30% decrease from baseline means of 110-132 mg/dl). Patients on select doses of SC beloranib had statistically significant improvements in HDL-c, and leptin from baseline (not versus placebo). SC beloranib treatment also showed favorable changes in systolic and diastolic blood pressure from baseline.
  • SC beloranib treatment was associated with an increase in treatment-emergent adverse events, including sleep disturbances that led four patients to withdraw from the trial – three patients in the 4.0 mg arm and one patient in the 2.0 mg arm. Most adverse events were mild and self-limiting in nature.

Adverse Event

Placebo (n=6)

1.0 mg (n=6)

2.0 mg (n=6)

4.0 mg (n=7)

Psychiatric Disorders

1 (16.7%)

5 (83.3%)

6 (100%)

6 (85.7%)

GI Disorders

4 (66.7%)

3 (50%)

3 (50%)

5 (71.4%)

Skin and Subcutaneous Tissue Disorders

0 (0%)

1 (16.7%)

3 (50%)

5 (71.4%)

Infections and Infestations

0 (0%)

3 (50%)

0 (0%)

1 (14.3%)

Metabolism and Nutrition Disorders

0 (0%)

4 (66.7%)

1 (16.7%)

2 (28.6%)


Symposium: When Drugs Get in the Weigh


Xavier Pi-Sunyer, MD, PhD (St. Luke’s-Roosevelt Hospital Center Columbia University, New York, NY)

Dr. Xavier Pi-Sunyer reviewed the medications currently approved by the FDA for weight loss (orlistat [Roche’s Xenical], phentermine monotherapy, lorcaserin [Arena/Eisai’s Belviq], phentermine/topiramate [Vivus’ Qsymia]) and their clinical data, as well as naltrexone/bupropion (Orexigen’s Contrave), which he believes will be approved soon as long as the Light Study goes well – we are inclined to think it will have positive results although the likely sheer number of events continues of course to be a topic of discussion. During his presentation, Dr. Pi-Sunyer noted that patients with diabetes on lorcaserin or phentermine/topiramate have poorer weight loss then patients without diabetes. In a conversation we had with him after the session, he said he would prefer using Qsymia in people with diabetes due to its increased efficacy (except for women of childbearing age who did not take the need for contraceptives seriously). During Q&A, it was great to hear Dr. Pi-Sunyer’s opinions on specific therapies. When asked about using generic phentermine and topiramate instead of Qsymia (due to cost), Dr. Pi-Sunyer strongly cautioned about the difficulty of getting the dosing right with the generics. He explained that Vivus uses an extended form of topiramate, which is not generic, and allows lower doses to be used with success. He said that while the generic combination of bupropion and naltrexone has not been used widely yet, he suspects people will try it (instead of Orexigen’s Contrave). In response to another question, he stated it is likely that lorcaserin and phentermine will be tried in combination; when asked about evidence that patients on lorcaserin regain some weight after their first year on the drug, he hypothesized that the trend is largely due to patients thinking they have done the work and lost the weight, and lose focus on diet and exercise. Dr. Pi-Sunyer also made several comments throughout his presentation about obesity pharmacotherapy in general, including that the FDA is particularly concerned about the safety of all obesity medications because it assumes many people will take them. On the individual level, he highlighted the wide variability of weight loss all drugs have, and that initial weight loss is the best predictor of long-term success. Dr. Pi-Sunyer therefore advised that if a patient is not losing weight on a drug after six weeks, it should probably be discontinued. Additionally, he stated that HCPs need to ensure that their patients have realistic expectations for how much weight they are going to lose on an obesity medication, otherwise they could be disappointed by the results of any current pharmacotherapy.

Questions and Answers

Q: With regards to the weight regain that we see after the second year of patients being on lorcaserin, how much do you think is due to background lifestyle changes?

A: I think that is a lot of it. I think people believe they have done the work and lost the weight. They then don’t think about how much they are eating or working out, and they stop going back to see their doctor.

Q: Would you use any of these drugs in a patient who has had bariatric surgery?

A: None of them have been tested in that population. It certainly is not FDA approved. We need some randomized controlled trials to see if it works.

Q: The current cost of Qsymia is about $150 a month. Can you comment on using the generics instead of Qsymia, given the cost?

A: With phentermine, there is no problem – you can use the generic version alone and it is fine. If you want to put phentermine and topiramate together however, you have to realize that J&J had to use far larger doses [of topiramate] in order to get the weight-loss effect. But Vivus is using slow release [topiramate], so it can use a low dose. The generic is not slow release, so you wont get that effect. With the combination of bupropion and naltrexone, however, nobody has done that yet, but I suspect people will do that.

Q: Do you think people will probably use lorcaserin with phentermine?

A: Probably. When the FDA stopped fenfluramine, it specifically did not stop phentermine too, because it thought that the valve problems were due to fenfluramine.

Q: Could you please be more specific about the psychological contraindications for lorcaserin? Are you referring to severe disorders?

A: Lorcaserin has not been tested in patients with psychological disorders, so there really is not data. However, with sibutramine there was a similar concern that people would have serotonin problems, and those were not seen.

Q: With Qsymia, patients would take up to ~90 mg of topiramate. We all have lots of diabetics, and a lot of them have mild renal impairment. It has been 27 years since I have seen metabolic acidosis, but with Qsymia, are we going to start seeing that again?

A: I think that could be a problem with some diabetics particularly if they are on metformin. I think that you have to monitor them closely to make sure they don’t have acidosis. [Editor’s note: when we spoke to Dr. Pi-Sunyer after the session, he said he was not greatly concerned about the risk of acidosis due to Qsymia].


Symposium: What Outcomes Matter: Prevention and Treatment of Pediatric Obesity to Reduce Diabetes Risk


Jack Yanovski, MD, PhD (NIH, Bethesda, MD)

Dr. Jack Yanovski was careful to base his presentation on clinical findings and not his own speculation, so phentermine/topiramate ER (Vivus’ Qsymia) was only briefly discussed during Q&A (he thinks it has potential but emphasized the need for long-term studies), and lorcaserin was not mentioned. Instead, Dr. Yanovski described the benefits and risks of several treatments that have been tested in children, including several that have been removed from the market (fenfluramine, sibutramine, and ephedrine plus caffeine). Throughout his review Dr. Yanovski emphasized that a drug must be as safe as possible to be used in children because they will tend to take it longer than adults – maybe for the rest of their lives. Thus, he said that one should limit pharmacotherapy to children who have a BMI greater than the 95th, or maybe even the 97th, percentile. Dr. Yanovski said that the only drug approved for weight loss in children, orlistat (Roche’s Xenical) cannot be routinely recommended for treatment of obese adolescents (saying that there was not sufficient clinical data for it to have been approved in the first place) because it is only minimally effective in this population. Notably, Dr. Yanovski said that a study (the only one he is aware of) looking at the use of exenatide in obese children found that it improved BMI and insulin use significantly, and he said GLP-1 agonists are an exciting area for future development.

Questions and Answers

Q: I have had a number of patients who have been put on Adderall for appetite suppression. What do you think about this as a treatment option?

A: There are insufficient clinical trial results to make that recommendation. Adderall and Ritalin have been shown to suppress appetite for a little while but they tend to wear off towards the end of the day resulting in many kids eating large amounts at night. It has not really been sufficiently studied to make a recommendation.

Q: What do you think about using topiramate in children?

A: Topiramate alone in adults was stopped due to a bad side effect profile. There is only very short-term data – about 12 weeks – of using it with phentermine, but it did have some efficacy. Long-term studies are needed and have not been performed yet.

Q: What about probiotics for treating obesity?

A: I can’t tell you much about probiotics for obesity but I think it is a fascinating idea.

Corporate Symposium: The New Treatment Paradigm in Obesity: Proper Patient Selection, Assessment of Comorbidities and Stepped Interventions (Sponsored by Vivus)


Robert Kushner, MD (Northwestern University Feinberg School of Medicine, Chicago, IL)

Dr. Robert Kushner opened the session by administering an audience response pretest to the packed room of ~350 attendees. The crowd included over half MDs and nearly 20% PhDs. The post-test at the end of the session showed an improvement on every single question. Most interesting was the final question, “As a result of attending this symposium, how confident are you in your ability to manage a patient with obesity?” At the end of the session, 38% were very confident and 33% were somewhat confident, while another 17% said it reinforced what they already do in their practice; unfortunately, we did not capture the figures before the presentations – we are attempting to get those figures.



Donna Ryan, MD (Pennington Biomedical Research Center, Baton Rouge, LA)

Dr. Donna Ryan discussed the risks and benefits of obesity and the decision making process HCPs should use when selecting an obesity treatment. To start, Dr. Ryan described how obesity increases one’s risk of having over 80 medical conditions. In particular, she emphasized how BMI and risk of type 2 diabetes are inextricably linked – while we may feel we hear this frequently already, presumably those at TOS may benefit from having this reinforced. We know, of course, that there is also more interest from a regulatory and reimbursement perspective of avoiding diabetes or managing it better, compared to reducing weight but having no other effects on health. Dr. Ryan also described how powerful modest weight loss is, highlighting that it can reduce pain and diabetes risk, and that greater weight loss (such as that seen with bariatric surgery) may reduce mortality and even cause diabetes remission. Dr. Ryan repeatedly emphasized that when attempting to balance the benefits and risks of weight loss interventions, practitioners must use their best clinical judgment. Currently, practitioners only have the NIH’s 1998 federal guidelines, and in Dr. Ryan’s view, these are far from an algorithm. New guidelines are expected in 2013, but Dr. Ryan expects that they will contain the same BMI cut off points of 25 kg/m2 for being overweight and 30 kg/m2 for being obese. She also stressed that BMI is not enough for a clinical decision, pointing to the George Washington University’s report “Obesity Drug Outcomes Measures” – the report recommends that the overweight and sick should be viewed differently than the overweight and well when making clinical decisions. Additionally, Dr. Ryan highlighted that in NHANES III, the Edmonton Obesity Staging System (which includes factors other than BMI to determine how ill an obese individual is) had better predictive power for mortality risk than BMI alone. We think it’s very positive that HCPs are hearing more and more about EOSS at meetings like TOS. Dr. Ryan concluded her presentation by reemphasizing that HCPs cannot prescribe an obesity treatment only because a patient asks for it – although this is certainly true, with new alternatives, we also think HCPs should be ready to hear more directly from more patients.



Samuel Klein, MD (Washington University School of Medicine, St. Louis, MO)

Dr. Samuel Klein gave a comprehensive research overview of the benefits of weight loss on the metabolic complications of obesity, noting that weight loss “is more powerful than any medication we’ve developed.” He emphasized that just two days of calorie restriction and as little as 2% weight loss leads to marked improvements in several metabolic parameters, including hepatic insulin sensitivity, HOMA IR, and reductions in intrahepatic triglyceride content. The metabolic benefits of weight loss also increase progressively as more weight drops off, although a plateau is typically reached around 15-20% body weight loss. Dr. Klein mentioned Qsymia two or three times, highlighting the drug’s weight loss mediated improvements in HDL, triglycerides, blood pressure, and ALT concentrations. Last, we enjoyed Dr. Klein’s review of his 2004 NEJM study that looked at liposuction in obese people with normal glucose tolerance and diabetes. Interestingly, there was no improvement in obesity-related comorbidities associated with liposuction, leading Dr. Klein to conclude, “removing body fat itself does not improve the metabolic complications of obesity.” Instead, a negative energy balance is required. It is easy to understand, hearing about liposuction’s impact on comorbidities, why regulatory and reimbursement authorities are looking for impact on co-morbidities rather than just treatments that prompt weight loss.


Louis Aronne, MD (Weil Cornell Medical College, New York, NY)

Dr. Louis Aronne gave the final presentation of the symposium in which he reviewed the mechanisms, dosing, package insert, clinical trials, benefits, and risks of phentermine/topiramate (Vivus’ Qsymia) and lorcaserin (Eisai/Arena’s Belviq). He began by reminding the audience that the central nervous system (CNS) is highly redundant. Thus, one must be very cautious when using a single drug for weight loss, as it will often require higher doses for efficacy and result in serious side effects. Consequently, he believes it is more effective to use small doses of a compound – indeed, Dr. Aronne hypothesized that one-day, triple fixed dose combination drugs may exist. He then explained how Qsymia uses this multi- compound strategy by combining phentermine (a sympathomimetic to suppress one’s appetite) with topiramate (an anticonvulsant to block the body’s counter response to weight loss that would otherwise negate phentermine’s effect). In contrast, lorcaserin produces satiety by potently and selectively binding to 5-HT2C, but does not have a second agent to oppose the CNS’s counter response. Dr. Aronne stated this might be why lorcaserin is associated with less net weight loss (4.8 kg [10.6 lbs.]) than phentermine/topiramate (12.2 kg [26.9 lbs.]), and why recipients in the BLOOM study began to regain weight in the second year of treatment. Notably, when reviewing the BLOOM-DM study result that lorcaserin was associated with a 0.5% reduction in A1c, he said he would be happy to use lorcaserin with metformin in his patients with type 2 diabetes. In closing, Dr. Aronne suggested that obesity medications can help fill the current treatment gap that exists between diet and exercise (resulting in <5% weight loss) and bariatric surgery (>15% weight loss but invasive) until less invasive surgical treatments are developed.

  • Dr. Aronne hypothesized that in the future, we may use triple compound drugs to treat obesity, since the central nervous system (CNS) is highly redundant. As a result of this redundancy, high doses of a weight loss drug are needed, but they can cause severe side effects through direct CNS interactions or counter effects by the CNS. Lower doses of agents – especially in combination – are more tolerable and better for treatment. In his opinion, past use of drugs with high doses of single active ingredients is probably why obesity medications were riddled with negative safety profiles. Additionally, Dr. Aronne said that a single agent may be effective for some period, but that a counter effect may build over time and lessen its long-term efficacy. With this in mind, we look forward to better understanding the long-term durability of weight loss with lorcaserin.
  • The maximum dose of phentermine used in Qsymia is less than one third of the average dose of phentermine that has traditionally been prescribed in the past. On the clinical trial front, Dr. Aronne was particularly impressed with phentermine/topiramate’s demonstrated efficacy in morbidly obese individuals (who he said were suited for bariatric surgery) during the SEQUEL study, as well as the average 27 lbs. of weight loss seen on the top dose of phentermine/topiramate in EQUIP.
  • Lorcaserin (Eisai/Arena’s Belviq) is a potent and selective 5-HT2C serotonin agonist that activates POMC neurons. Importantly, lorcaserin has nearly no affinity for 5-HT2B, the serotonin receptor stimulated by fen-phen and implicated in many of the drug’s issues. Dr. Aronne reviewed several of lorcaserin’s clinical trials including BLOOM and BLOOM-DM. When reviewing the BLOOM longitudinal data, he did point out the trend for participants to begin regaining weight during the second year – Dr. Aronne attributed this finding to lorcaserin’s use as a single weight loss agent.
  • Dr. Aronne would be happy to add lorcaserin onto metformin in his patients with type 2 diabetes. To support his view, he emphasized that the 0.5% reduction in A1c (after 52 weeks) associated with lorcaserin was a good result for a drug that is not intended to directly treat diabetes. We wonder if Arena/Eisai or Vivus will pursue indications for either drug as a treatment for diabetes – we assume they will, especially as metformin will be submitted to FDA soon as a treatment for pre-diabetes as we understand it. Additionally, we hope to see studies of Qsymia or Belviq in combination with a DPP-4 inhibitor and/or metformin – we suspect such a fixed dose combination would have quite compelling efficacy, not to mention potential for use in prediabetes.
  • Dr. Aronne reviewed previously published data comparing weight loss outcomes from pharmacological treatments. He noted that lorcaserin was associated with substantially less weight loss (4.8 kg [10.6 lbs.]), stating it was likely because lorcaserin is a single agent and therefore is more limited in its treatment power. Dr. Aronne cautioned that this data is not intended to be used for cross-drug assessments (i.e., different trials, different baselines, etc.), though it did put into context the stronger efficacy of Qsymia relative to the other products in development.


Net Weight Loss


12.2 kg (26.9 lbs.)


6.2 kg (13.7 lbs.)


4.8 kg (10.6 lbs.)


4.4 kg (9.7 lbs.)


Robert Kushner, MD (Northwestern University Feinberg School of Medicine, Chicago, IL); Samuel Klein, MD (Washington University School of Medicine, St. Louis, MO); Louis Aronne, MD (Weill Cornell Medical College, New York, NY); and Donna Ryan, MD (Pennington Biomedical Research Center, Baton Rouge, LA)

  • Case 1: 28 years old with a BMI of 32 kg/m2 and depression. Family history of obesity and type 2 diabetes. Currently on escitalopram [an anti-anxiety medication]. Skips breakfast 50% of the time and her portions are large at dinner.
    • Audience Response:
      • 1. Counsel on lifestyle changes and/or refer to a commercial weight reduction program: 54%
      • 2. Prescribe Qsymia and lifestyle change: 21%
      • 3. Prescribe phentermine and lifestyle change: 11%
      • 4. Prescribe Belviq and lifestyle change: 8%
      • 5. Reassure her, no further intervention indicated: 4%
      • 6. Prescribe orlistat and lifestyle change: 2%
    • Dr. Klein: She is not under a rush, and she does not have an ideal diet. So I think intense behavioral modification could be enough.
    • Dr. Ryan: There are definitely things that she could work on so I think there is the option to do strict lifestyle intervention. However, she does have a family history of type 2 diabetes, and she has had success with phentermine, so you could have her use that again.
    • Dr. Aronne: That she responded to phentermine in the past suggests that she will respond to it again. However, I would recommend cross titrating her on bupropion [currently approved for treating depression] because she has depression. However, you would not just discontinue her escitalopram and start the bupropion. You would start the bupropion at a very low dose and see how she responds. Then you would reduce the escitalopram slightly and see how she does, and you would continue in that manner.
  • Case 2: 72 years old with coronary artery disease and a BMI of 36 kg/m2. Progressive weight gain, hypertension, hyperlipidemia.
    • Audience Response:
      • 1. Counsel on lifestyle changes and/or refer to a commercial weight reduction program: 46%
      • 2. Prescribe Qsymia and lifestyle change: 17%
      • 3. Prescribe Belviq and lifestyle change: 13%
      • 4. Prescribe orlistat and lifestyle change: 13%
      • 5. Reassure him, all lab values controlled: 6%
      • 6. Prescribe phentermine and lifestyle change: 5%
    • Dr. Klein: His BMI is 36 kg/m2 – that is not so big [laughter]. Having excess weight when you are older makes it even more difficult to be functional than when you have excess weight and are younger. So if he is having difficulty moving around, that could be a good motivator to get him to lose weight. Explaining this to him may make him willing to pay for a commercial program. I agree with the audience to focus on #1, and not to forget the quality of life for this person.
    • Dr. Ryan: I don’t think he could be hurt by 5% to 10% weight loss. What you don’t want to do is reduce his lean body mass, so I don’t think he is a good candidate for bariatric surgery.
    • Dr. Aronne: I would look for sleep apnea in a man of this weight, something you often see. As a first step, I might consider a low dose of metformin as part of the intensive behavioral program. When people come and see us, they have already tried diet and exercise, and we have to give them other ideas.
  • Case 3: 52 years old with a BMI of 31 kg/m2 and an A1c of 10.4%. Currently on glargine 38 units at night and metformin for diabetes.
    • Audience Response:
      • 1. Add a GLP-1 agonist: 42%.
      • 2. Prescribe Qsymia and lifestyle change: 19%
      • 3. Refer to bariatric surgery: 15%
      • 4. Prescribe Belviq and lifestyle change: 9%
      • 5. Intensive lifestyle change and refer to dietitian: 9%
      • 6. Intensify insulin therapy: 5%
    • Dr. Klein: This patient is in awful glycemic control. That’s what I would treat most aggressively. Some extra insulin at night would help. GLP-1 agonists are not cheap, but that would be nice to help glucose control. We always say lifestyle therapy, but we have to really appreciate that it is really difficult to do. That might be a step that could be tried.
    • Dr. Aronne: I strongly agree that a GLP-1 agonist is the first thing. These are the kinds of cases we see all the time in our clinic. Often it’s a patient that cannot control his or her diet. Having some type of low fat protein, yogurt for breakfast might help. We find that those little things, that attention to detail, can make a very big difference. In this patient, we would start the GLP-1 at the low dose, cut insulin slightly, and recommend a low fat, protein breakfast. With the addition of two weight loss medications that also reduce glucose, we now have other lines of treatment. Using lorcaserin and phentermine/topiramate, we’ll be able to keep going along that line.
    • Dr. Ryan: I would take a weight centric approach – the Look AHEAD approach. I would immediately cut the insulin in half and leave them on metformin. I would put them on two-meal replacement per day, and one portion controlled meal a day. That’s about 1200 calories, so a lot of negative energy balance. You must cut the insulin. I would also use phentermine and topiramate to maximize the weight loss. This is someone 52 years old with poorly controlled diabetes. You need a concerted effort, a full course charge on this. As loses weight, may stop insulin altogether. My approach not glucose centric as weight centric.
  • Case 4: 41 years old with hypothyroidism, depression, and elevated blood pressure. BMI of 42 kg/m2. Eats out 75% of time.
    • Audience Response:
      • 1. Prescribe Qsymia and lifestyle change: 31%
      • 2. Counsel on lifestyle change and/or refer to a commercial weight reduction program: 28%
      • 3. Refer to bariatric surgery: 17%
      • 4. Prescribe Belviq and lifestyle change: 11%
      • 5. Prescribe phentermine and lifestyle change: 7%
      • 6. Prescribe orlistat and lifestyle change: 5%
    • Dr. Aronne: At this weight and depressed, I think of sleep apnea. I see it so often. If she has sleep apnea, that could help with weight loss and help her to get off citalopram. Phentermine and topiramate give the kind of weight loss that is good for patients in this BMI range – it’s 14%, approaching lap band surgery. The kind of patients we see are looking for an alternative to surgery. You could have surgery, but many don’t want to have that.
    • Dr. Klein: I agree completely. Qsymia has changed the paradigm for initiating bariatric surgery. It’s worth trying before undergoing bariatric surgery. Secondly, it’s the only medication that has been studied people with no limit to the BMI. The highest person in the clinical trials had a BMI of 68 kg/m2. This is really a good potential candidate for that.
    • Dr. Ryan: It comes across when talking to the individual. Part of the treatment decision depends on patients’ willingness to undertake any of the recommendations. Engaging the patients in these decisions is very important. Patients have to be committed to whatever treatments you recommend.



Robert Kushner, MD (Northwestern University Feinberg School of Medicine, Chicago, IL); Samuel Klein, MD (Washington University School of Medicine, St. Louis, MO); Louis Aronne, MD (Weill Cornell Medical College, New York, NY); and Donna Ryan, MD (Pennington Biomedical Research Center, Baton Rouge, LA)

Q: There are two new medications that have been approved, how do you choose which one to use?

Dr. Aronne: That is a great question, because we have experience with phentermine and topiramate so we have something to choose on and the other we haven’t used yet so what I say I not based on clinical experience. It is going to be based on a number of factors. If someone is on a serotonin activator that will be an easy answer – they should not be on lorcaserin. Someone with a higher BMI I might use phentermine/topiramate because we have data using it in that population. Those may be superficial reasons but as we move along we might get a better idea of which one to go with overall.

Q: In individuals undergoing liposuction, you might hypothesize that if they gain weight, they would store more fat in their liver since you’ve taken away the adipose cells. Is that a risk for liposuction?

Dr. Klein: That hasn’t been shown but there is potential for that. There is one study showing a bit of weight gain with modest liposuction. There was an increase in visceral fat. Once you have liposuction, you destroy the architecture of fat cells. You gain weight and it goes to another area. In women, that’s the breasts and under the arms.

Q: Dr. Ryan, what was long-term maintenance in your study Look AHEAD?

Dr. Ryan: The strategy in Look AHEAD was a very intensive six-month effort, and the weight loss was generally maintained at 12 months. At some time in the next six months the eight-year data will be released, and it is good. The biggest predictor of success was the amount of initial weight loss a person had. So there is a subgroup that does really well with this strategy. Also, the meal replacement strategy was repeated if people needed it. The long-term strategy for keeping weight off was monitoring people and if people began to regain the weight then we sent them on a campaign, we reintroduced the meal replacements, we put them on an exercise program etc. Not that that is a strategy everyone needs, or that works for everyone. If someone is starting to regain weight we need to act early. Also, all of the interventions we used are in the public domain and can be found online [

Q: What about patients that have a contraindication to phentermine due to heart rate or blood pressure. It is okay to prescribe topiramate only without phentermine? Since Qsymia has a lower dose of phentermine, should we be concerned about using higher doses?

Dr. Aronne: Topamax is not approved for weight loss, so that would be off label use. That doesn't restrict you, but it’s off label. At the Qsymia low dose, which is not the dose people are supposed to stay on, it’s one tenth of a standard phentermine tablet. I would predict that that has little impact on blood pressure. I talk to our cardiologists all the time. They admit that shouldn’t have an impact. The other thing to do is to treat blood pressure more aggressively. We tend to do things we think are really safe first and move from there. We might get someone to lose weight first. Perhaps try a meal replacement first for a short time, then phentermine/topiramate.

Q: Is there a burnout component to obesity? Does the duration of obesity have an impact on the severity of comorbidities? What does the literature say?

Dr. Klein: The longer you have obesity the worse potential outcomes you have. That is why someone who has been obese since childhood is at greater risk for comorbidities than somebody who gets to that same BMI as an adult. For example, the longer a person has obesity, the greater the pancreatic beta cell damage they have.

Q: When are the new guidelines coming out?

Dr. Ryan: I want to say – the road to hell is paved with good intentions. We started this in January 2008. It’s been a real odyssey. I think that people’s expectations about what these guidelines are going to be is different from what they are. There’s lots of knowledge to inform PCPs and providers on how to manage obesity, blood pressure, and lipids. The methodology we’re using is a very stringent methodology. It’s recommended by the Institute of Medicine. I don’t want to go into the why’s and wherefores. They are due to come out in 2013. When I write my memoirs, you’ll hear every gory detail.

Q: What are some of the combinations of pharmaceutical therapies that you think are provocative or interesting to think about?

Dr. Aronne: A GLP-1 with lorcaserin would make a lot of sense. You would have to look out for nausea. Or phentermine/topiramate plus a GLP-1; people with diabetes who use those get really good weight loss. Lorcaserin/phentermine, I think is an obvious combination. I was involved when the FDA was sorting out the fen-phen out. The problem really was the fen [fenfluramine] not phentermine; phentermine just made it effective enough. I know of several people who have plans to study phentermine’s safety when used with topiramate. I think trials are going to get organized quickly to study phentermine/topiramate, rather than what happened last time with fen-phen when nothing happened until there was a problem.

Comment by Dr. Kushner: I think that there is a tendency to have a boom-and-bust trend with these kinds of drugs. It is provocative to think about using these therapies in combination but I think we have a responsibility to use these drugs correctly until we know what works and is safe.

Dr. Aronne: I completely agree. The misuse of medications has really bothered the FDA for a really long time. One of the reasons we heave not seen medications is because the FDA feels like only way they can prevent a medication’s misuse is by preventing it from coming on the market. So we have to make sure that these drugs are used appropriately. I think that if we use these drugs appropriately then we will set the standard. We might need enforcement at the state level, so that if someone is misusing the drug they are held accountable, but that is for another conversation.

Q: On the negative liposuction results, is there any evidence that an omentectomy leads to metabolic improvement?

Dr. Klein: Intra-abdominal fat consists of omental fat and mesenteric fat. It can be removed laparascopically. You can remove an amount of the omentum – one kg of omental fat is about 25% of intra-abdominal fat. Removing that fat surgically has no effect on insulin sensitivity or beta cell function.

Q: Do higher levels of obesity increase the mortality of all illnesses, or does it vary with illness?

Dr. Ryan: There is a lot of actuarial evidence that as BMI increases the risk of mortality increases across the entire US population, but if you look at subsets of the population the risk is not the same. African

American women appear to be able to tolerate the excess weight (in terms of mortality) until their BMI hits 40 kg/m2 and then we see that excess mortality. But there is a difference between individuals and the population. We all know people who are obese in their 80s and have tolerated it well. We have to understand that these are population figures.

Q: You’ve been a major proponent of changing medications to weight neutral or weight loss medications. What about individuals that are stable on their medications – like antidepressants of antipsychotics? What’s the potential risk of taking them off medications they’ve been stable on? How do you weigh that?

Dr. Aronne: I always ask myself the question, ‘Why is the patient in my office? Why is someone there today?’ We have a big box of tissues in the room. We call it the trail of tears. People are so upset with their inability to lose weight. They are frustration with the things they have been through. I think that the difficulty of losing weight, the pain of it, is underestimated. We try to do this in the most parsimonious way. We have a discussion and talk to people – here are your options. Here are the risks. We talk to their psychiatrist, if that applies. We don’t undertake things like that lightly. You don’t just cut someone’s antidepressant. As we become more comfortable, using something like phentermine/topiramate may be a good strategy – at a low dose – if someone starts gaining weight on a medication.


Corporate Symposium: Managing Obesity & Comorbid Conditions: A New Era of Treatment Strategies (Sponsored by Eisai)


Caroline Apovian, MD, FACP, FACN (Boston University School of Medicine, Boston, MA)

Dr. Caroline Apovian began the session and her two-part presentation with a question for the audience of ~100: what is the prevalence of type 2 diabetes that is attributable to obesity: (a) 28%, (b) 35%, (c) 46%, or (d) 61%? Originally, only 49% of the audience selected the right answer: (d) 61%. Dr. Apovian then reviewed the close connection between BMI and type 2 diabetes (saying that over 61% of type 2 diabetes is attributable to obesity), but that there is significant variation in risk on the personal level. As an example, she noted that there are patients with very high BMIs who never develop diabetes, while some other people with only slightly elevated BMIs (24 kg/m2) develop type 2 diabetes. Additionally, Dr. Apovian reviewed how one’s risk of cardiovascular disease mortality goes up with inflection points (where risk begins to increase more quickly with weight) at 25 kg/m2 (the conventional BMI threshold used to define overweight) and 30 kg/m2 (the conventional BMI at which one is defined as obese) – Dr. Apovian said the presence of these inflection points are how the thresholds were selected. She proposed that in the future it may be possible to study a person’s fat tissue to determine their specific risk for comorbidities. When Dr. Apovian again asked the audience her original question on what percent of type 2 diabetes is attributable to obesity, 94% of the audience selected the right answer.



Caroline Apovian, MD, FACP, FACN (Boston University School of Medicine, Boston, MA)

Following her overview of risk stratification, Dr. Caroline Apovian changed topic slightly to discuss the internal regulators of weight gain and loss. She stated that when people gain weight and fat tissue becomes dysfunctional, it is infiltrated by macrophages, which release cytokines that attract B cells. The proliferation of this inflammatory response results in the proliferation of fatty acids and other molecules that go to the liver, musculature, and other organs, promoting diseases such as hypertension, heart disease, and type 2 diabetes. Additionally, she explained that our understanding of adipose tissue has changed – it is no longer viewed as “just sitting there storing fat,” but rather, as an endocrine organ that releases leptin (which suppresses appetite) and other hormones that regulate weight. The gut similarly secretes many appetite regulators including ghrelin (which stimulates hunger) and PYY (which promotes satiety). Meanwhile, insulin from the pancreas causes one to feel more full. All of these hormones and more, Dr. Apovian said, interact redundantly in the hypothalamus to regulate one’s weight and ensure we don’t starve to death. Furthermore, when one loses weight, these hormone levels are altered in a way that promotes regain of the weight that was lost, making weight maintenance extremely difficult and not just a matter of will power. Dr. Apovian stated that an NEJM article (Sumithran et al., 2011) reported that one year after overweight and obese patients (n=50) lost about 14% of their baseline weight, they still had altered levels of hunger signals “torturing them” (Dr. Apovian later apologized for saying weight loss is tortuous and said she meant to say challenging – personally, we think having to battle with one’s own hormones to not eat sounds like a mild form of torture). Dr. Apovian said that it is possible that five years after weight loss, these hormones might be back at their original levels.


Donna Ryan, MD (Pennington Biomedical Research Institute, Baton Rouge, LA)

Dr. Donna Ryan emphasized that behavioral intervention should be the foundation for any obesity treatment, and should not be abandoned even if a patient is on a medication or has undergone bariatric surgery. Early in her presentation, she expressed optimism that the recent decision by CMS to reimburse obesity screening and intensive behavior therapy (though currently only for primary care physicians) would be followed by a further expansion of obesity services and reimbursement. She promoted the behavioral intervention delivered in the Look AHEAD study as a model for behavioral intervention, emphasizing that what worked in the trial can work in the office setting: motivational interviewing, self-monitoring, goal setting, stimulus control, problem solving, and relapse prevention. Dr. Ryan noted that even after four years, participants in the Look AHEAD study maintained an average of close to 5% weight loss from baseline. Subsequently, she highlighted the strategies used by those in the National Weight Control Registry who successfully achieved weight loss and maintenance: a low-fat diet, watching total calories, high levels of physical activity (about 400 kcal/day, or the equivalent of one hour of brisk walking), and frequent self-monitoring. Dr. Ryan explained that the strategies for weight loss and weight maintenance are not the same – diet and monitoring of food intake are very important for weight loss, whereas exercise and monitoring of weight are very important for weight maintenance. Dr. Ryan stated that the physician’s role is more one of an enforcer during weight loss, and a supporter during weight maintenance. In closing, Dr. Ryan stressed that we not only need tools to support healthy behaviors, but also medications and devices to augment behavioral therapy.



Harold Bays, MD (Louisville Metabolic and Atherosclerosis Research Center, Louisville, KY)

After providing a quick review of orlistat’s and phentermine’s profiles, Dr. Harold Bays discussed the mechanisms, efficacy, safety, and dosing of lorcaserin (Arena’s Belviq), phentermine/topiramate (Vivus’ Qsymia), as well as naltrexone/bupropion (Orexigen’s Contrave) and liraglutide (Novo Nordisk’s Victoza). Throughout his presentation, Dr. Bays emphasized the importance of looking not only at weight loss, but also at improvements in metabolic parameters when assessing the benefits of obesity medications. He highlighted data from the BLOOM trial showing gradual weight regain with lorcaserin after one year of treatment, and pointed out that there is a Risk Evaluation and Mitigation Strategy (REMS) program in place for phentermine/topiramate due to its risk of birth defect.

  • Lorcaserin (Arena’s Belviq): Dr. Bays highlighted that lorcaserin is a selective 5HT-2c receptor agonist that did not show increased valvulopathy risk over the course of two years in clinical trials. Although the drug’s DEA schedule is not yet complete, Dr. Bays predicted it would most likely be classified as a Schedule IV drug. Commenting on the drug’s longer-term efficacy in the two-year BLOOM study, Dr. Bays noted that there is gradual weight regain with lorcaserin after the first year of treatment. He commented that the most important part is not how much an obesity drug improves weight, but rather, how much it improves metabolic health. He emphasized improvements in glycemic parameters – A1c and fasting glucose – seen with lorcaserin, and noted that there was some improvement in blood pressure with the drug, and only “marginal” improvements in total cholesterol.
  • Phentermine/topiramate (Vivus’ Qsymia): In terms of safety, Dr. Bays highlighted that the topiramate component of the drug could contribute to secondary angle-closure glaucoma, in addition to increasing the risk of cleft palate. He pointed out that Vivus has a Risk Evaluation and Mitigation Strategy (REMS) program in place to educate both providers and patients on the potential risk of birth defects in taking phentermine/topiramate while pregnant. When he asked the audience how many people had already gone through the healthcare provider training program, about 20 people raised their hands (in a room of approximately 100 attendees). He explained that the phentermine/topiramate combination in Qsymia uses lower doses of phentermine and topiramate compared to the maximum approved doses of these drugs individually for weight loss and other indications. Specifically, the dose of phentermine in phentermine/topiramate is ≤40% the maximum dose, and topiramate is <25% of the maximum dose. Dr. Bays stated that since the phentermine/topiramate combination uses these lower doses, the combination drug has better tolerability than the combined use of the two generic components would. Dr. Bays also cited CONQUER data, commenting that phentermine/topiramate brings about “robust” weight loss, as well as improvements in A1c and blood pressure (we note that he did not show SEQUEL data or comment on the two-year durability of weight loss with phentermine/topiramate).
  • Naltrexone/bupropion (Orexigen’s Contrave): Dr. Bays emphasized that naltrexone/bupropion is still an investigational agent, with an ongoing cardiovascular outcomes trial, dubbed the Light Study.
  • Liraglutide (Novo Nordisk’s Victoza): Dr. Bays stated that liraglutide, while approved as a treatment for type 2 diabetes, remains an investigational agent for the treatment of obesity. He noted that while there is a decent amount of discussion on the role of GLP-1 on the muscle and on the liver, the effects of GLP-1 agonism on adipose tissue has largely been overlooked, and should be explored further.



Caroline Apovian, MD (Boston University School of Medicine, Boston, MA)

Dr. Caroline Apovian began the final presentation of the symposium by reviewing the current National Heart, Lung, and Blood Institute’s (NHLBI) obesity guidelines on using BMI and waist circumference to estimate a patient’s disease risk. Similar to what Dr. Donna Ryan (Pennington Biomedical Research Center, Baton Rouge, LA) had said the night before at Vivus’ CME (see our Day #2 report at, Dr. Apovian told the audience that the guidelines are currently being updated but that they probably will not change very much. She said there may be some changes based on different ethnicity’s disease risk (we assume this relates to findings that Asian Americans are at an increased risk of developing comorbidities at lower BMIs than other ethnicities), that there will be a stronger emphasis on evidence from clinical trials, and stricter measures to prevent conflicts of interest. Dr. Apovian then described the Edmonton Obesity Staging System (EOSS), which stages a patient’s mortality risk based on their medical, functional, and mental characteristics on a scale of zero to four where stage zero describes a person with an elevated BMI but no comorbidities, and stage four is a person with severe comorbidities. Dr. Apovian recommended that practitioners reserve more aggressive obesity treatments for people in EOSS stage three or four. Like Dr. Ryan the night before, Dr. Apovian highlighted research finding that the EOSS is better at predicting increased mortality due to obesity than BMI alone, and said that “BMI measures only how big the patient is, not how sick the patient is.” Looking to the future, Dr. Apovian said it would be great to have a biomarker that successfully predicts who is at an increased cardiometabolic risk. In closing, she reemphasized that central obesity is the key criteria for cardiometabolic risk, and people who are not able to store a lot of subcutaneous fat instead store the excess fat in their liver, muscles, around their heart, etc., resulting in more deleterious effects.



Caroline Apovian, MD (Boston University School of Medicine, Boston, MA), Donna Ryan, MD (Pennington Biomedical Research Institute, Baton Rouge, LA), and Harold Bays, MD (Louisville Metabolic and Atherosclerosis Research Center, Louisville, KY)

Case Study

Case: 54-year-old woman with hypertension, hyperlipidemia, migraine headaches, and depression. Weighs 242 pounds; has a BMI of 38.2 kg/m2. Has an A1c of 6.9%; on metformin. Mother had type 2 diabetes.

Dr. Apovian: How would you classify the patient’s risk?

Dr. Bays: She is at high risk.

Dr. Apovian: We do not have a measurement of her waist circumference yet. Should we go back and get a measurement of her waist circumference?

Dr. Ryan: She already has hypertension, sleep apnea, depression, and metabolic syndrome. We already know she is at high metabolic risk. She has a BMI over 35 kg/m2. We do not need to put her in a gown and measure her waist circumference at this point, since this patient is already known to have comorbidities. There is no sense in subjecting her to getting her waist circumference measured when we won’t have any additional information to gain. She knows that she is overweight and I think that doing a waist circumference measurement is going to be a barrier to building a relationship with her. I generally don’t measure waist circumference in patients who have a BMI over 30 kg/m2 because it doesn’t add much information. Sometimes, waist circumference is measured at an endpoint in trials. Waist circumference is useful for patients with BMIs of 31 or 26 kg/m2, and even at lower BMIs in Indian and Southeast Asian populations.

Dr. Apovian: I agree. Normally, I would like to see the waist circumference measured, however, she already has type 2 diabetes, hypertension, etc. Waist circumference helps the most for patients with BMI between 25 and 35 kg/m2.

Dr. Bays: I believe that when adipocytes are dysfunctional, they contribute to metabolic disease, and that is what you have here. We don’t need a waist circumference to know that is what is happening in this woman. Nobody denies that there are not genetic components, but for most patients we see in our clinic, everyday metabolic disease, like diabetes, is an adipose dysfunction. If you are unable to store fat in the legs and arms, then that excess energy has to be stored in the liver or around the heart. The adipose tissue surrounding the vessels transmits signals that damage them.

Dr. Apovian: What would your treatment strategy be for this patient? Which drug would you use to treat her?

Dr. Ryan: Before developing a treatment plan, I would look at the patient globally, and find what would thwart effective treatment. Does she have symptoms of GERD? Of urinary incontinence? I would want to know more about her mobility as well. If she’s taking an anti-depressant that causes weight gain, I would taper use of that drug, and start her on one that is weight-neutral or causes weight loss. Much of how I would develop a treatment strategy for this patient couldn’t be decided based on this case information – it depends on interactions with the patient. I think that ideal care for this patient would be what we provided to patients in Look AHEAD, and use of a medication for additional weight loss. Which medication to use would be something to discuss with the patient – I’d want to discuss the pros and cons of each drug with the patient. Is she sexually active? Does she have plans on becoming pregnant? If so, I would probably not recommend Qsymia. Lorcaserin is another option – it has been shown to decrease A1c 0.5% in clinical trials.

Dr. Apovian: I don’t think she can get pregnant [since she’s 54].

Questions and Answers

Dr. Apovian: A few people asked questions about the use of HCG [human chorionic gonadotropin, aka the pregnancy hormone]. The trouble is that it hasn’t been studied in the way that the FDA wants in randomized controlled trials, with large sample sizes and placebo comparators. There are studies in literature, but they all documented a strong placebo effect. Also, patients were instructed to be on a 500-600 kcal/day deficit on top of receiving HCG injections. What we would need is for patients in a control arm to receive a placebo injection. There are patients who swear by HCG, but how do they know it’s not just a placebo effect with the injection, along with the 500 kcal deficit? The FDA does its due diligence to make sure that a drug is safe and effective.

Q: Is there data on the use of Belviq and Qsymia with anti-depressants (e.g., SSRIs)?

Dr. Bays: These are newly approved drugs; these questions will be a whole lot easier to answer a year from now.

Q: In Look AHEAD, was weight loss predicted by BMI?

Dr. Ryan: Good question. Those with BMI in the 25-30 kg/m2, 30-35 kg/m2, 35-40 kg/m2, and above 40 kg/m2 had roughly the same percent weight loss from baseline. Those who started out with higher baseline weight lost more kilograms. The modest weight loss that was achieved in Look AHEAD benefited people with higher BMI and people with lower BMI. The weight-loss benefit seen in higher and lower BMI categories underscores the benefits of the intervention across the BMI spectrum. The people in Look AHEAD were obese when they started and they were obese when they finished; they were just a lot healthier.

Q: Should women who become pregnant while using Qsymia discontinue the medication slowly if they are on the high dose?

Dr. Bays: It’s recommended that people on the high dose of Qsymia who discontinue the medication do so at a gradual rate. For pregnant women, how do we balance the risk of seizure with the risk of cleft palate? I have no idea, since we don’t have any randomized controlled trials. My opinion, however, is that we should treat acute risks acutely, and chronic risks chronically. I don’t know that a gradual tapering of Qsymia would reduce the risk of cleft palate. Stopping the drug acutely, however, would exacerbate the risk of seizure. One could take the middle ground. It’s more the art of medicine than the science. I would say that the dose should be tapered, but maybe more quickly.

Q: How do we keep patients on medications when the goal is weight maintenance, not weight loss?

Dr. Ryan: When patients stop taking their obesity medications, weight regain occurs. What happens in the clinic is patients get tired of taking the medications. They’re happy with the initial weight-loss trajectory, but when their weight loss plateaus, they think the drug is not working. Patients do not associate these drugs with weight control like they do with blood pressure management and anti-hypertension medications. I try to get my patients to commit to one year of therapy, then intermittent treatment afterwards. We monitor their weight, and put them back on medications and meal replacements as appropriate.

Dr. Apovian: We do the same thing. We slowly taper the drug down, and we decide ahead of time how much weight regain they are going to tolerate before they go back onto the therapy.


Corporate Symposium: Translating Science into Strategy: Contemporary Clinical Approaches in the Management of Obesity (Sponsored by Novo Nordisk)


Robert Kushner, MD (Northwestern University Feinberg School of Medicine, Chicago, IL)

Dr. Robert Kushner highlighted significant events that have occurred in the field of obesity prevention and discussed the future outlook. He lauded the Centers for Medicare and Medicaid Services (CMS) for offering coverage for intensive behavioral therapy for obesity that is provided by primary care practitioners, describing the measure as a start toward improving coverage. However, he found it unfortunate that this coverage did not apply to registered dieticians. Dr. Kushner also drew attention to the CMS’ mandate that EMR software be capable of calculating BMI. To him, such software would alert physicians to overweight and obesity and hopefully initiate more weight counseling to improve treatment. Dr. Kushner recognized the recent FDA approvals of lorcaserin (Eisai’s Belviq) and phentermine/topiramate (Vivus’ Qsymia) as landmark decisions that could substantially influence obesity therapy (see our coverage of Belviq’s approval at and of Qsymia’s recent launch at He noted however, that our society is currently facing a generation of physicians who obtained their medical training with no experience in prescribing anti-obesity medications. To that end, he said the challenge going forward would be to educate physicians on the appropriate ways to use these medications. The American Board of Obesity Medicine’s initiative to provide an obesity certification exam (the first exam will be offered this November) represents progress on this front. During the Q&A discussion, Dr. Kushner expanded on this topic and noted that if enough doctors took the board exam to become “local champions” in combatting obesity, and if insurers would take notice of this movement, then obesity would become a true medical specialty. This comment summed up Dr. Kushner’s conclusion to his talk – that physician engagement in the obesity arena and insurance coverage of obesity therapies would be two substantial driving factors in stemming the obesity epidemic. Insurance coverage would be largely dependent on how the Affordable Care Act is implemented in each state – Dr. Kushner commented that while obesity falls under preventative and wellness services, a very strong lobbying effort will likely be needed for states to include obesity services as an essential benefit.

  • Under the Affordable Care Act (ACA), each state has the ability to determine its own essential benefit list and health care exchange; thus each state can decide whether to include obesity prevention within its health care exchange. Dr. Kushner reminded the audience that state-based exchanges are designed to allow individuals and/or small businesses to come together to buy insurance coverage similar to that of large companies. Each state determines the core list of services offered in each exchange. The ACA lists “preventive and wellness services and chronic disease management” as an essential benefit, and while obesity prevention certainly falls under this umbrella, the ACA does not require states to include it.
  • The December 2011 Health and Human Services Guidance stated that states would choose one of four health insurance plans as a benchmark for their own essential health benefits list: 1) one of the three largest small group plans in the state; 2) one of the three largest state employee health plans; 3) one of the three largest federal employee health plan options; or 4) the largest HMO plan offered in the state’s commercial market.
  • Discouragingly, when looking across the top-three enrolled small-employer health plans in each state, only four of the 153 plans included coverage for weight-loss programs; one hundred twenty four plans excluded coverage, seven offered limited coverage, and 17 did not have data available (further details available at While it is possible that coverage using the other possible benchmarks (see above bullet) could offer a more positive outlook, Dr. Kushner concluded that extensive lobbying would be required to see changes in obesity coverage when the ACA is fully implemented in 2014.



George Bray, MD (Louisiana State University, Baton Rouge, Louisiana)

The esteemed Dr. George Bray gave a thoughtful, high-level overview of the factors that influence obesity. He opened by listing the many clinical conditions that are associated with obesity, citing type 2 diabetes as the most important co-morbidity. He then referenced the Framingham study to emphasize that social networks influence obesity and that people who are obese are more likely to have obese friends. To him, the social aspect of obesity suggests that social networking strategies could potentially be useful tools to reverse the obesity epidemic. Dr. Bray then turned to environmental factors and noted that while the role of personal responsibility in weight maintenance is currently controversial, he believes that personal responsibility does not occur within a social vacuum, but rather operates within diverse environmental conditions that influence people’s eating behaviors. After listing several environmental factors, including weight-promoting medications (more details below), Dr. Bray mentioned twin studies to introduce the topic of genetics and body weight. He noted that studies have found a strong, positive correlation between BMI and the number of obesity-related alleles. To him, the ability of genetic markers to predict a person’s treatment response represents an especially fascinating field of study. Dr. Bray then cited two social trends that can each nearly account for the increased prevalence of obesity and that combined, certainly explain the obesity epidemic: 1) the increase in sedentary activity between 1960 and 2010, which has resulted in a 100 calorie/day reduction in energy output; and 2) the steady increase in food intake since the 1970s. For the remainder of his presentation, Dr. Bray discussed the biological pathways that regulate body weight, with a focus on leptin, gut signaling (via GLP-1), and gut microbiota. He ended his presentation by recapitulating that obesity is associated with several clinical conditions and that obesity management programs, such as those found in the DPP, can reduce these risks.

  • Dr. Bray listed several environmental factors that influence obesity: food abundance, exercise, temperature regulation, medications, associative sorting, infective agents, smoking, intrauterine environment, maternal age, and sleep debt. Notably, he focused on the role of medications in weight gain, stating that within most drug classes, certain medications produce more weight gain than others, and that it is the physician’s ethical responsibility to choose the drug that more favorably influences weight. Using type 2 diabetes as an example, he explained that sulfonylureas cause weight gain, that DPP-4 inhibitors are generally weight neutral, and that GLP-1 agonists such as liraglutide result in weight loss.
  • Dr. Bray noted that gut microbes are ten times more numerous than other body cells and influence body weight in several ways: 1) they contribute to energy homeostasis by harvesting nutrients; 2) they can increase gut permeability in people who are obese or who have type 2 diabetes; and 3) microbial lipopolysaccharides may promote low-grade inflammation



Ken Fujioka, MD (Scripps Clinic, San Diego, CA)

Dr. Ken Fujioka gave his perspective on the two recently approved weight loss drugs (lorcaserin [Eisai/Arena’s Belviq] and phentermine/topiramate ER [Vivus’ Qsymia]), as well as the two weight loss drugs he thought were next in line for FDA approval (naltrexone/bupropion [Orexigen’s Contrave] and liraglutide [Novo Nordisk’s Victoza]). Dr. Fujioka targeted his presentation to the HCPs in the audience and reviewed the things HCPs should consider before prescribing lorcaserin or phentermine/topiramate ER (unfortunately, few HCPs heard Dr. Fujioka’s comprehensive overview, as there were easily less than 50 total attendees at the symposium).

  • Lorcaserin (Eisai/Arena’s Belviq): Dr. Fujioka gave a systematic review of the FDA Briefing Documents and the package insert. He highlighted that lorcaserin intervention led to a decrease in heart rate. Dr. Fujioka explained that he categorizes obese patients as either “obese sick” (i.e., where treating the obesity will help resolve other health issues such as diabetes) or “obese well” (i.e., where weight is an isolated issue). Because Dr. Fujioka saw no cardiac signal associated with the lorcaserin, he would consider using it in both “obese sick” and “obese well” patients. Interesting, he found the Pregnancy category X categorization of lorcaserin surprising; to his knowledge, the drug led to no teratogenicity effect on the fetus. Dr. Fujioka suggested the categorization was likely due to the FDA having a “different set of rules for weight-loss drugs.” Furthermore, Dr. Fujioka emphasized the warning against giving lorcaserin to patients with congestive heart failure (CHF) – lorcaserin is a serotonin agonist selective for the 5HT2c receptor at the recommended 10 mg dose; however, patients with CHF may overexpress the 5HT2b receptor, making the 5HT2b receptor a more likely drug target than in patients without CHF.
  • Phentermine/topiramate ER (Vivus’ Qsymia): In his review of phentermine/topiramate ER, Dr. Fujioka seemed to consider the dosing and titration scheme more difficult than that for lorcaserin since Qsymia comes in four doses. He noted that unlike with lorcaserin, there was sufficient reason to label phentermine/topiramate ER as a Pregnancy Category X drug, as it wasclearly associated with oral clefts. He noted that topiramate at a high dose was associated with suicidal behavior, but emphasized that the dose included in this formulation was much lower. He also explained the interaction between phentermine/topiramate ER and birth control pills (BCP). While the drug affects how BCPs work, it does not do so in a way that affects fertility.
  • Naltrexone/bupropion (Orexigen’s Contrave): Dr. Fujioka highlighted the cardiovascular outcomes trial (CVOT) required by the FDA for this drug (for details on the progress of this trial, please see our July 11 Closer Look at As a reminder, in March, an FDA Advisory Committee voted in favor of requiring obesity drugs without a theoretic risk or signal for cardiovascular harm to rule out a certain degree of excess CV risk (see our March 30 Closer Look at Dr. Fujioka stressed that the CVOT for naltrexone/bupropion would not be another SCOUT Trial (as a reminder, this trial led to the removal of sibutramine [Abbott’s Meridia] from European and US markets). In the SCOUT trial, patients took sibutramine regardless of whether the drug was effective. In Orexigen’s CVOT, the study drug will be discontinued if the patient does not lose weight, or if he/she experiences a negative impact on blood pressure. To Dr. Fujioka, this study design more closely models real-world medicine.
  • Liraglutide (Novo Nordisk’s Victoza): Dr. Fujioka saw great potential in liraglutide for treating both obesity and prediabetes. Liraglutide is currently approved for the treatment of type 2 diabetes and is being investigated in a phase 3 program as an anti-obesity medication (further details are available in our Novo Nordisk 2Q12 report at



Robert Kushner, MD (Northwestern University Feinberg School of Medicine, Chicago, IL); George Bray, MD (Louisiana State University, Baton Rouge, Louisiana); and Ken Fujioka, MD (Scripps Clinic, San Diego, CA)

Q: Say you have a patient and you want to use a weight loss medication. Their BMI was 35 kg/m2. Now their BMI is 26 kg/m2 but they’re struggling to keep their weight down so you want to use a weight loss medication again to help maintain the weight. Can you use a weight loss medication?

Dr. Kushner: That’s a good question. The current guidelines don’t dictate or address this situation. I think all of us that use pharmacotherapies think of it as a wonderful use for weight maintenance. Putting aside the FDA indication, the potential of using a medication to clamp the weight loss of someone who was obese is an excellent use for the drug. You could use it following bariatric surgery. Those of us who are caring for patients with bariatric surgery are often seeing weight regain two to five years later. These are really interesting questions, but the current indications don’t cover it.

Dr. Bray: Can I challenge you with an alternative? Because in the Look AHEAD trial, some people lost weight and then maintained it without pharmacotherapy. That group was quite surprising

Dr. Kushner: I think it can be used patients who are unable to maintain their weight loss. For example, the post-surgery patient.

Dr. Bray: Would you be willing let that patient go with no therapy for three months and see whether they maintain the weight with the best advice?

Dr. Kushner: I would.

Dr. Bray: I think I would too because there are a substantial number of people who maintained quite a substantial amount of weight loss. This really has change my view of whether we need lifelong treatment.

Dr. Fujioka: But say they’re regaining weight now.

Dr. Kushner: I would say that when you embark on obesity management, you don’t just look at weight, you look at a range of health outcomes that you want to achieve – good lipid levels, etc. I take all that into consideration, not just what’s happening to the BMI.

Q: Does lorcaserin make you drowsy?

Dr. Fujioka: Yes, it could. The reported side effect is low, one to two percent, but this class can do that.

Q: How did you arrive at the conclusion that obesity is largely caused by the food supply with physical activity being less important?

Dr. Bray: Food available for consumption went up by about 425 calories beginning in the 1970s. Data from Tim Church showed a physical activity decrease by about 100 calories per day in the same time. So it would suggest 20% is attributable to decreased physical activity and 80% is due to an abundance of food supply.

Q: If primary care physicians assume the responsibility to simply screen and then refer patients, do we have resources to care for the large patient population?

A: I would argue that we won’t be able to care for the large population unless we screen and refer. I had originally hoped that all physicians would have the capacity to handle all the obesity care in their practice, but now I don’t think that this is possible. There’s too much on the plate for physicians, so I think that screening and referring is reasonable. There are a lot of referrals, such as registered dieticians, weight loss programs (including Weight Watchers and Jenny Craig), multiple internet-based programs that are dynamic, and some commercial programs. There are programs like the YMCA and like TOPS. I think there are many resources available. I think the key is knowing what is in your community, and what your patients’ needs are. Is it group therapy or individual therapy? Are they tech savvy? Do they like group dynamics? You have to be aligned philosophically with what the program does and what the patient needs. So the referring doesn’t take you out of the picture. You remain in the picture, but you’re not doing the treating yourself.

Q: What do you do when you have a patient you want to put on lorcaserin, but he is already on SSRI?

Dr. Fujioka: You wouldn’t want to mix the two together, so you would have to transition them. From a pharmacological point, you can’t abruptly withdraw someone from an SSRI. I would probably wean the patient off of SSRI, then start them on lorcaserin.

Q: Lorcaserin has a beneficial effect on A1c, I believe. How might that influence its use? For example, if a patient didn’t reach their weight loss goal but reached the A1c reduction goal, should lorcaserin be discontinued?

Dr. Kushner: I think this gets to the drug’s mechanism of action. The improvement in diabetes control is weight related. I think the average A1c reduction is 0.5%, which is a pretty good reduction, but it’s unlikely to occur without a reduction in body weight. I think it’s an interesting theoretical question. And currently, we’re not prescribing lorcaserin for diabetes; we’re prescribing it for obesity and the A1c reduction is an added effect.

Dr. Fujioka: You would only see that if they had very out of control blood sugar and they’re losing weight. Once you start to control the blood sugar, they will actually maintain weight because they’re not spilling glucose and water. At best, it’s a 0.5% drop in A1c and you wouldn’t see that without weight loss because that’s the mechanism of the drug.

Q: Will having Qsymia around change how phentermine is used?

Dr. Fujioka: Phentermine is approved for a 12-week span. You need to inform the patient if you are going off label. I asked Dr. Eric Colman (Deputy Director of the Division of Metabolism and Endocrinology Products, FDA, Silver Spring, MD) this particular question. It’s become almost a standard-of-care question. In some states, it is very tightly regulated; in other states, you can go off label.

Q: What is the effect of topiramate on contraceptives?

Dr. Kushner: Topiramate does affect contraceptives, but not in regards to fertility, only in the regulation of menstrual control. So they’ll get some spotting, but fertility will be in tact.

Q: How long will it be before obesity medicine is accepted as a medical specialty?

Dr. Kushner: I hope it’s soon. The whole idea of starting this board is to raise the competency and awareness around obesity and to increase the workforce to tackle this epidemic. If enough individuals take the board to become local champions and if insurers start paying attention, I hope obesity will be recognized as a true medical specialty.

Q: If an insurance company will not pay for Qsymia, do you worry that physicians will then write two prescriptions, one for phentermine and one for topiramate, and that the dose will be too high?

Dr. Kushner: I don’t know for sure. My thought is that it will not happen. Only 2% of physicians are currently writing prescriptions for obesity medications, and we have a whole generation of young doctors who have never seen anyone write an obesity medication. Given the lack of understanding of obesity medications, I think doctors would be very hesitant to write two prescriptions for two drugs they are not familiar with off label if they could just write one prescription.

Dr. Bray: I agree. Topiramate has several side effects and when you try to balance these, you’ll run into some trouble. I think Dr. Kushner is right. Physicians will not do that. It’s not an easy strategy. It may be cheaper for patients, but it will be difficult to implement.

Q: Is there a difference between GLP-1 agonists for weight loss?

Dr. Fujioka: There is one trial comparing the diabetic doses of Byetta [exenatide] and liraglutide – the curves were pretty much the same for weight. But the current study for liraglutide uses a higher dose.


Product Theater: New Treatment Options for Obesity (Sponsored by Vivus)


Timothy Garvey, MD (University of Alabama at Birmingham, AL)

Dr. Timothy Garvey gave the only talk during this moderately attended 8 am product theater. Unlike like the previous night’s more general CME event sponsored by Vivus, the focus was very clearly on Qsymia: “We really want to hone in on Qsymia as a pharmaceutical agent and how to use it judiciously and safely to get the best outcomes for your patients.” Dr. Garvey gave a comprehensive overview of the new drug, including how it works, data from the pivotal trials, an exhaustive review of safety information and side effects, and a refresher on dosing and management. Overall, he emphasized that Qsymia is a good option between lifestyle intervention and surgery, has a low rate of side effects due to the drug’s small doses of phentermine and topiramate, has demonstrated durable weight loss over one and two years in clinical trials (and lower dropout rates than placebo), and is a simple once-daily drug that improves cardiovascular risk factors. We most enjoyed the Q&A, which included questions on off- label use and use of Qsymia in people with diabetes. The session featured handouts on the HCP training program (the same brochure given out in the Qsymia exhibit hall booth and at the CME event last night; said Dr. Garvey, “It’s not mandatory, it’s pretty simple, and it’s kind of fun. In fact, you can do it right now because you know the answers”), a prescription form for home delivery of Qsymia (it only listed CVS and Walgreens; as a reminder, Kaiser is also a certified pharmacy – see our report on the drug’s launch at, and the prescribing information in two separate brochures.

  • Vivus representatives handed out a one-page prescription fax form for Qsymia, the first time we can recall seeing the document. It’s extremely simple, consisting of a prescriber contact information section, a patient information section, a checkbox to select either CVS or Walgreens, and faxing instructions. There is a tear off section to hand to the patient, which notes: 1) the pharmacy will contact you for payment information; and 2) after payment has been received, it will take an average of three to five business days to receive your medication. We look forward to understanding what providers think of the REMS, if they find it burdensome, and how patients like the prescription/mail order process (especially those new to mail order). As a reminder, Vivus is also submitting a protocol to FDA this month to potentially expand Qsymia distribution into select retail pharmacies. We look forward to more updates on this front – the expansion would be a boon for Vivus considering that mail order only accounts for ~30% of US prescriptions.

Questions and Answers

Q: How long can Qsymia be used since phentermine is only approved for 12 weeks?

A: The short-term restriction for phentermine does not apply to Qsymia. It’s only for higher doses of phentermine when it’s used as a single agent. There is no contraindication to chronic use. We have two- year data – we don’t have 10-year data. Obesity is a life long disease, so we’re committing these patients to long-term treatment. Like any new medication, we must develop an experience over time as we figure out the strategies to most effectively maintain weight loss. We’ve got good data for two years and we’re optimistic that this can serve our patients over the long haul.

Q: Can Qsymia be used with SSRI medications?

A: Yes. That was not an exclusion criteria in the clinical trials. Up to 20% of patients in these trials had depression and there was no exacerbation. It can definitely be used with SSRI medications.

Q: Were there depressed patients in the clinical trials?

A: Yes. There were no interactions with anti-depressant medications.

Q: Can phentermine counteract the negative cognitive effects of topiramate?

A: With higher doses, topiramate can be associated with cognitive slowing. With the doses in Qsymia, this is much less of a problem. In terms of cognitive function and phentermine counteracting a topiramate-like drug – it’s possible. It’s a theoretical construct that makes sense, but we don’t have hard data.

Q: Is it easy to split the generics and use them off label?

A: It can be done. Physicians have done that up until now. It’s problematic getting the right doses, the right dose escalation strategy, and you risk greater side effects. You can’t get these same dosages. There is no clinical trial data to use to base that therapeutic intervention on. It’s not FDA approved. Medically, legally, you need to think about that. You’re placing yourself at risk. Qsymia, with its accompanied trial data and REMS, has been tested and is FDA approved. You could use generics if you wanted to, but it’s going to be a lot easier and better for your patients if you use the FDA approved approach.

Q: Has Qsymia been studied in patients under 18?

A: That’s a really good point. These patients need help as well. Phentermine is a sympathomimetic agent. If you think about it, we throw sympathomimetic agents like Ritalin at kids all the time. Topiramate is improved in children with epilepsy. So it’s not like children and adolescents haven’t seen these types of drugs. But there is no data and no FDA approval. Hopefully, there will be studies done at some point. This is a really severe problem in our country and around the world. [Editor’s note: One of Qsymia’s post- approval requirements is for Vivus to conduct a study using Qsymia in a pediatric population (see our July 18, 2012 Closer Look for more on the required post-marketing studies at

Q: Are you required to start at the lower dose of 3.75/23 mg?

A: Well, it’s not required. Physicians can use it off label. But it makes a whole lot of sense. This is a slower dose escalation than was done in the clinical trials. There, patients were at the initial titration dose for one week. This is two weeks. It’s done so patients can safely accommodate themselves to Qsymia. I think it’s a really good idea to minimize side effects.

Q: Is there any hormonal interaction between Qsymia and contraceptives?

A: I don’t know what that interaction is. With the weight loss that occurs, any abnormalities in sexual hormones will tend to improve. Women who weren’t ovulating, particularly those with polycystic ovarian syndrome, may start ovulating with weight loss. That’s why it’s important to have effective contraception. You must disregard it when patients say, ‘I have been trying to get pregnant for 20 years.’ I’m not aware of any interactions with hormonal contraception.

Q: In the preliminary trials, how high was the incidence of seizures in patients going off Qsymia?

A: There were no seizures in the clinical trials. The recommendation when stopping the drug is to take it every other day for one week. It comes back to the prescribing information for topiramate. It’s generally a good idea to wean patients off in that way. There were no seizures associated with Qsymia.

Q: Is there any caution with use of Qsymia in patients with metformin?

A: Qsymia is not approved for treating diabetes. However, it is approved for weight loss, including patients with diabetes. Weight loss will favorably affect glucose and in many cases cause a decrease in diabetes. In CONQUER, 16% had diabetes. So there is some experience there for use of the drug. They did very well and there was no interaction there. There is a possibility for hypoglycemia, even though metformin is not highly associated with hypoglycemia. You need to have home blood glucose monitoring.

Q: How does Qsymia affect dyslipidemia or other metabolic disorders?

A: Weight loss attacks the cardiometabolic disease process. Insulin resistance gets better. It happens with weight loss due to Qsymia. You see a decrease in blood pressure comparable to many antihypertensive drugs. Triglycerides go down dramatically. HDL levels increase. LDL levels decreased marginally, but you don’t expect to see a change with weight loss. There is a decrease in glucose as well as insulin – that’s indicative of an increase in insulin sensitivity. Inflammation, CRP, goes down. Fibrinogen levels go down. Adiponectin – a market of insulin sensitivity and a good adipokine associated with leanness and reduced risk for diabetes and heart disease – goes up. Diabetes got better even though medications were reduced. You see a favorable effect on all cardiometabolic syndromes.

Q: Is Qsymia okay in patients with kidney stones?

A: It’s not contraindicated in patients with kidney stones. But it is something you have to monitor. If a patient is having frequent kidney problems, it may be a patient you would not want to put on Qsymia.


2. Mobile Health

Preconference Workshop: Physical Activity Monitoring Methodologies


Peter Katzmarzyk, PhD (Pennington Biomedical Research Center, Baton Rouge, LA)

Dr. Peter Katzmarzyk provided an overview of physical activity monitoring methods, which include questionnaires, direct researcher observation, pedometers, heart rate monitoring, accelerometry, arm bands, GPS/cell phones, and doubly labeled water (the gold standard). He detailed the numerous pros and cons of each method, though he seemed most positive on accelerometers (“one of the most popular methods to measure physical activity”). As measured via accelerometry, only 6% of 16-19 year olds, 3.5% of 20-59 year olds, and 2.5% of 60+ year olds meet the recommended physical activity recommendations in the United States. Further, he noted that via an NHANES questionnaire, about 40% of the population sits for six or more hours per day. Dr. Katzmarzyk concluded with a review of physical activity monitoring interventions. He emphasized that 1) on average, pedometer interventions can increase physical activity by about 2,500 steps per day (Bravata et al., JAMA 2007); 2) physical activity monitoring studies should use an outcome measurement device that is separate from the intervention device; 3) physical activity monitoring should be embedded in a behavioral intervention; and 4) there is an optimal balance between better monitoring technologies and human/financial limitations – combining accelerometry with questionnaires is one promising avenue to give the information greater context.



Catrine Tudor-Locke, PhD (Pennington Biomedical Research Center, Baton Rouge, LA)

Dr. Catrine Tudor-Locke discussed the use of physical activity monitoring data in the National Health and Nutrition Examination Survey (NHANES), emphasizing that it was “pretty exciting” to have such rich data on a national scale. Her focus was on research methods and data interpretation rather than population wide statistics, which was disappointing. However, she did mention a few startling stats: as measured by accelerometers in NHANES, less than 6% of adolescents and less than 5% of adults meet physical activity guidelines (reiterating the previous speaker). Further, participants in the 2005-2006 NHANES spent about five waking hours per day not moving at all and only seven minutes per day at a walking speed of 100 steps per minute or more (the threshold for moderate or vigorous physical activity). She and colleagues have published an extensive number of papers on this topic, including Tudor-Locke et al., Prev Med 2011; Tudor-Locke et al., J Phys Act Health 2011; Tudor-Locke et al., Prev Chronic Dis 2012; and a number of others.

  • Physical activity monitoring data (using an ActiGraph AM-7164 accelerometer) was added to NHANES in 2003. Participants 6-85 years old in the NHANES panel were recruited at mobile examination centers and were asked to wear the instrument for up to seven days. Data is recorded in one-minute time intervals in both activity counts and steps. Thresholds of 2,020 activity counts and over 100 steps per minute were set for moderate and physical activity. Activity counts are available from 2003-2006 and steps data is only available for 2005-2006.


Symposium: Do You Hear Me Now? Using Mobile and eHealth for Weight Control


Susan Woolford, MD, MPH (University of Michigan, Ann Arbor, MI)

Dr. Susan Woolford provided a research overview of mobile and internet-based technology for pediatric and adolescent weight control, a technology she said is “moving at warp speed.” She was most optimistic on the potential for texting since 80% of adolescents own a cell phone, 75% of them say that they text, about 50% say they can text blindfolded, and one in three teens sends more than 100 texts per day. Despite many pilot projects and published studies in the last year, however, the fact still remains: weight outcomes in these trials have been disappointing. Interventions have generally been well received broadly, although ultimately an inability to change behavior has been challenging. Given more positive results in adult studies, Dr. Woolford believes the potential in pediatrics is there, but there are a number of facets to consider: 1) changing behavior is not that simple; 2) the content of messages is critical (e.g., saying don’t go to McDonalds as someone walks by a McDonalds might not be helpful – in fact, it may encourage them to go there!); 3) privacy and security; 4) ease of use (“they won’t use anything that will require more effort on their behalf); 5) the depth of individual tailoring; 6) incorporating behavioral therapy; 7) novelty/cool factor (“some of our messages sounded like an English person from the 1950s”); and 8) external factors (baseline level of motivation, parents, etc.).

Questions and Answers

Q: I don’t remember – do you actually have teens step on scales to enter their weights?

A: For our studies, we do measure weight in teens on a weekly basis. We have some creative ways of our team measuring things. The limitation with pedometers, for instance, is they put them on their pets, in dryers, etc.

Q: Do you show them graphs?

A: Yes, we do. They can see their weight over time.

Q: I’m amazed by how many teens text. It’s really brief, like a nice little gift. Maybe you get a microburst of dopamine – in other words, the texting is medicating them! It seems plausible. What else would explain why that behavior would be so much more prevalent?

A: Absolutely.



Alain Labrique, PhD (Johns Hopkins University, Baltimore, MD)

Dr. Alain Labrique spoke on the potential of mHealth to change the landscape of health care; however, he warned that we may still be in a phase of inflated expectations with regards to mHealth and that careful research will be needed to shepherd ideas into scalable, realizable solutions. Dr. Labrique’s talk was littered with small vignettes of the use of mobile technology in global health. For example, in a study of 168,231 women in Gaibandha, Bangladesh, only 23% of those surveyed had electricity in their homes, while a surprising 71% had household phone ownership and 20% had used a phone in the past 30 days for an emergency health purpose. This reinforced to us that the global potential of this field, though vast, is hard to wrap one’s arms around – Dr. Labrique noted that there are over 1,000 weight- related mobile apps available for Apple and Android phones, and determining which apps are actually efficacious is no small task. Magazines like Forbes, Shape, and International Business Times have put out their own lists of the top diet and weight-loss apps (“Lose It!,” “Fooducate,” “Nike Training Club,” and “Eat This Not That! The Game” seem to be top contenders) and the company Happtique has created an mHealth certification process to help users identify apps that meet high standards and are based on reliable content. Still, for many patients and physicians, our sense is that the mHealth landscape remains difficult to navigate. Dr. Labrique further explained that whether an app “works” can change according to the quality of evidence sought and from whose perspective the app is considered (i.e., patient, regulator, industry, etc.). He closed his presentation with reference to the Gartner “Hype” Cycle (a graphic representation of technology adoption from idea to commercial availability: and seemed to suggest that the jury was still out on whether mHealth would end up in the “trough of disillusionment” or make it to the “plateau of productivity” – an apt question and one we hope will be addressed to a greater degree in the next year.

Questions and Answers

Q: Have you looked at populations that would really jump on mHealth because this is a life preserver or populations who have access to the technologies for reasons that aren’t specific to health care?

A: The exciting part of this technology is that people aren’t going out to buy a phone for health care; they’re buying it for other reasons. There’s a fundamental need for technology that is driven by all sectors of the market. It’s an opportunity and we have to use this platform to deliver strategic intervention. There are two facets: in much of the world we don’t yet have as much penetration of smart phones as we would like. But we also do have to realize these aren’t just phones, these are small computers. We need to frame research questions in such a way to so as to identify whether mHealth is an appropriate solution.

Q: Your title was originally about social media for weight loss. Have you looked at the data here? I’ve had a website for overweight kids for 12 years and only a tiny fraction actually interact on this site. Are we thinking this social media is going to be wonderful? The data I’ve accumulated suggests otherwise.

A: I apologize for the miscommunication on the title. We have looked to some extent at social media. It’s important to define the social network, like whether it is open or closed, as well as who comprises that network. One of the things to consider is also where the individuals are in their change. They have to be at that stage of readiness to accept input in order for the technology to work.



Michael Lin (Johns Hopkins Medical School, Baltimore, MD)

Mr. Michael Lin presented preliminary results from the Tailored Rapid Interactive Mobile Messaging (TRIMM) study. This 12-month weight loss trial randomized 130 individuals from inner city Baltimore (a generally underserved population) to either lifestyle education (control) or lifestyle education + TRIMM messaging (intervention). Lifestyle education consisted of a 20-minute meeting with a dietitian at baseline to discuss goal setting and behavior changes and education material at three and six month follow-up appointments. The lifestyle education + TRIMM messaging arm received three to four health related text messages a day for the six month intervention period on top of lifestyle education. These text messages were semi-personalized (i.e., used the participants’ first name and were specific to the patients’ weight loss goals) and interactive (e.g., one text per day would ask a question and request an answer: “Please respond! Michael, how much do you weigh today?”), but still automated (e.g., “You’ve gained five pounds, but don’t be discouraged. If you reduce your portion sizes…”). Data from the study’s initial assessment suggests the TRIMM intervention amplifies weight loss – at six months, the control group experienced two to three pounds of weight loss (0.5-1.5% body weight loss), while the intervention group experienced six to eight pounds of weight loss (3.0-4.0% body weight loss). However, a more robust statistical analysis will be done that makes intent-to-treat adjustments (at six months ~70% of patients returned for assessment) and a final assessment will be done at 12 months – we assume this will reduce the observed weight loss benefit of the intervention. Mr. Lin emphasized the simplicity, and thus, scalability of this solution, but recognized that there is no “one-size-fits-all” solution. The goal moving forward will be to better understand the type of person this intervention works for and why it works. We look forward to seeing more data on the approach, especially its cost, provider time, and patient engagement in the long haul.

Questions and Answers

Q: Did you help patients set their goals?

A: Yes, in both the control and intervention group.

Q: Did you have them set specific goals or general goals like eating less fat?

A: In the initial assessment with the dietitian, she defined specific goals. In the education material provided, advice was more general and the text messages themselves were more general.

Q: How accurate were the self-reported weights?

A: I can only speak anecdotally, but I think they were relatively accurate. We will be doing that analysis soon. With the anonymity of the text messages, individuals didn’t seem to have fear of gain of weight.

Q: Did the patients enrolled in the study know whether a computer or person was generating the text messages?

A: We’ve experienced anecdotally that some participants think they are talking to the dietitian. Even if they don’t think about it concretely, there is still that subliminal thought in their head. We do tell them initially that it is computer generated.

Q: Engagement predicts weight loss. Children use text messaging more than adults, but it seems from preliminary studies that engagement seems to be much higher in adults then in children, anecdotally. What is your take?

A: I do think that many of these adults have more motivation and more understanding of the consequences of being overweight or being obese. I think its worth noting our participants average age is 51 and some hadn’t texted before so we gave them a five minute workshop – these patients had very high engagement as well.

Q: So there were not that many young adults?

A: I only know the average age.

Q: Were initial sessions one-on-one or in a group?

A: One-on-one.



Steve Cole, PhD (HopeLab, Redwood City, CA)

To conclude the session, Dr. Steve Cole presented the results of the Zamzee Randomized Controlled Trial and Biomarker study. The trial examined the physical activity and health benefits of a kid-friendly accelerometer and online motivational website in 448 kids over six months. Participants were randomized to a control group (passively wearing the Zamzee device but without access to the online website) or the Zamzee group (wearing the Zamzee device with access to the online website). After six months, those in the Zamzee group averaged 118 minutes per week of moderate and vigorous physical activity, compared to 74 minutes per week in the control group (59% more time; p<0.001). Baseline activity levels were not reported, so we cannot say by how much Zamzee increased physical activity. Impressively, the researchers also measured A1c, cholesterol, and C reactive protein. A1c and LDL cholesterol were significantly better in the Zamzee group vs. the control group at month six, though the difference was not clinically significant. Disappointingly, there were no changes in BMI, though the mean at baseline was a modest 22 kg/m2, so this was challenging to demonstrate. Overall, we think the technology is quite cool and are glad to see such a rigorously designed, real world trial of the device. We also enjoyed hearing the thoughts of Dr. Cole during Q&A, who described his job not as a public health professional, but one of “marketing physical activity to kids.” Given the tremendous benefits of exercise, we think the approach and intervention certainly has a lot to offer America’s youth, though the company must still prove it can keep kids engaged with the technology over time.

  • Zamzee combines a wearable accelerometer with an online module that encourages kids to be more active. The accelerometer has a built-in thumb drive and clips on to clothing to track activity. It does not have an on-device screen, so users must upload their data to the Zamzee website to view data. Once uploaded, the activity is converted into points that can be used to purchase gift cards, access online content, and donate to charity. The site has a big social component as well, which incorporates leaderboards, tweet-like functions, and an ability to friend other users. Zamzee is now operating as a full operational business at The Zamzee meter is $29.95 and access to the website is provided free of charge. HopeLab, a nonprofit research organization, developed Zamzee with support from the Robert Wood Johnson Foundation.
    • We think the developers have done a good job of making the website very kid friendly and well designed. Since the accelerometer is quite cheap and made of plastic, we have heard of kids breaking it (not to mention losing it). Of course, the company must make the tradeoff between a lower end, cheaper device that gets the job done and a higher end, more durable and expensive device that fewer people can afford (or are willing to invest in) and that parents would worry about getting lost. Interestingly, the site does not incorporate any health-related educational messaging about the importance of activity, diet, etc. Dr. Cole noted in Q&A that Zamzee has approached this problem from a marketing perspective: how do we sell physical activity to kids? It is certainly an interesting take on this age old problem, though we wonder how engaged users will be in the long term with the system.
  • The Zamzee system was tested in a six-month randomized controlled trial in 448 middle school children at six diverse study sites across the US ( identifier: NCT01433679). Kids were instructed to carry the Zamzee accelerometer during waking hours for six months. Half the kids were randomized to a passive monitoring control condition (the control group), where they wore the Zamzee but did not have access to the online module. The other half were randomized to wear the Zamzee and upload to the full website (the Zamzee group). We thought this was a clever and very fair way to conduct the controlled trial. The trial included two suburban sites, two rural sites, and two urban sites.
  • The trial’s 448 participants had a mean age of 12.7 years and a mean BMI of 22 kg/m2. The study had a relatively small population of overweight and obese children – 25% had a BMI >25 kg/m2 and 8% had a BMI >30 kg/m2. The sample was quite ethnically diverse (45% white, 25% Hispanic, 14% African American, 12% multiethnic, and 3% Asian). More than 90% of study participants supplied enough physical activity data to include in intention-to-treat analysis. About two-thirds of kids took part in the blood biomarker study.
  • After six months, those in the Zamzee group averaged 118 minutes per week of moderate and vigorous physical activity, compared to 74 minutes per week in the control group (59% more time; p<0.001). This translated to a 4.5-fold increase in the number of days that kids in the Zamzee group reached the CDC recommended level of 60 minutes of moderate or vigorous physical activity. Baseline activity levels were not reported, so we cannot quantify how much Zamzee increased activity from baseline to six months. Given the study’s size and randomization, we assume baseline levels were somewhat comparable between the two groups. Encouragingly, the increased activity in the Zamzee group was sustained over time and was slightly stronger in the second half of the study. Meanwhile, the control group’s activity declined over time. There’s no doubt that sustainability is a critical component of any successful lifestyle intervention, especially with kids that are liable to lose interest once the novelty wanes over time. Six months of stable engagement strikes us as encouraging, though we wonder whether kids or parents would lose interest after a longer period of time.
  • Compared to the control group, those in the Zamzee group had statistically significant reductions in A1c and LDL cholesterol, though there was no significant change in BMI. We were somewhat disappointed not to see a change in BMI, though given the low prevalence of obesity in the study, this was not too surprising. We note that the reductions in LDL and A1c, while statistically significant, were not clinically significant. Still, the fact that they trended down or stayed constant in the Zamzee group and increased in the control group certainly suggests a trend in the right direction. The decline in total cholesterol was borderline insignificant (p=0.057) and there was no significant difference in C reactive protein between the groups. A gene expression analysis is in progress.



Change in A1c


Change in LDL Cholesterol




No Change^



(73 mg/dl)



+5 mg/dl^



(73 mg/dl)

*p=0.012 for group by time interaction

^p=0.034 for group by time interaction

  • Dr. Cole noted that Zamzee incorporates three fundamental principles. First, it incorporates rapid and personalized feedback. He noted that it does a good job of letting kids know how active they are and how to get in the zone of moderate and vigorous physical activity. Second, Zamzee uses incentive-based motivation – kids using the system can earn points from engaging in physical activity that can be used to purchase gift cards, access online content, donate to a cause, etc. Third, Zamzee incorporates intrinsic motivation to convert induced behavior change to self-sustaining habits. In an ode to Daniel Pink’s excellent book Drive, Dr. Cole explained how Zamzee’s features target the core elements of intrinsic motivation: autonomy (customizing what the website experience is like, customizing the Zamzee device itself), mastery (continued growth over time, seeing one’s graphical progress, getting closer to a goal), and purpose (ability to contribute to charitable causes, engaging in a meaningful pursuit, social engagement).

Questions and Answers

Q: I’m curious how critical the tangible rewards were. Did they have that throughout the six months. Did you phase those out? I’m just wondering if it’s feasible to do this when it’s not funded externally.

A: Tangible rewards aren’t absolutely necessary at all. In some pilot studies, we had no rewards and you can get a kick up in physical activity. It’s obviously not as much as you would see with $5 per month or $20 per month. Tangible rewards make a significant difference but it’s not the whole story. There is version of this that is scaling in a corporate way at You log in and get points in virtual rewards. It’s only if the parent buys in or a sponsor does that you get tangible rewards. In terms of developing this stuff without philanthropic funding – it’s impossible.

Q: I’m most impressed by this innovative use of technology and interactions. You mentioned that the participants had some opportunity to contribute to the design. Is there a community being built up among participants?

A: They definitely have an opportunity to contribute to charitable causes. This pretested as significantly engaging for this demographic. In terms of the second thing, there is lots of complexity for social networking, particularly with young teenagers. We have created a closed ecology social network that does not involve socialization. There are generated messages that can be sent to one another. It would be unscalable to have to curate, so there is lightweight messaging. Basically, they can graphically high five one another – “Great job!” There is much more of this on the website, which is running as a business. In the trial, when you randomize individuals, there was much less opportunity for kids to engage with one another without contaminating experimental conditions. We’re working on expanding that now that it’s a business. We also think communication within a family context is probably the place to do it. There are collective goals and a tremendous opportunity there.

Q: Very elegant presentation – the methodology and a lucid finding backed up by biology. This is one of 140 RCTs in testing mHealth in a rigorous way. Some say that in mHealth, using rigorous methodologies is to deny the rapid pace of change. Do you feel that you’ve demonstrated fundamental principles that would apply to future technology?

A: While the technology is new, the basics of human psychology are at least 50 years old – arguably 10,000 or one million years old. The skin of Zamzee will change and the technology will move forward. When we started working on Zamzee five years ago, we had to put together an accelerometer pack because kids weren’t carrying smart phones. Already, the accelerometer pack is somewhat outdated. But it turns out that it doesn’t matter. The magic is not in the accelerometer, it’s in website. And it’s not website itself, but the mindset you create and the things you avoid. We’ve eradicated any mention of exercise, health, and education from this. These are generally not aspirational for kids. We don’t think of ourselves as public health professionals. We are marketers. We are selling physical activity to kids.

Q: I have a question about the treatment group and the degree of increase in moderate and vigorous physical activity – did you stratify those teens regarding the level of BMI? Did you see any difference in the degree of increase between those at the lower end of an elevated BMI and those that were more obese?

A: We’ve got thirty people on the high end of overweight, barely across the line of obese. They showed fewer minutes of increment, but still what looked to be a detectable increment. It was too small to say anything stat significant. I have no doubt that Zamzee has a harder job moving somebody that finds it harder to move. I would not pretend that this is a magic bullet or an effective weight loss intervention. This is a useful component to plug into the broader context of weight loss programs.

Q: Are we motivating people the wrong way by saying, “If you don’t engage in physical activity, you will have health consequences?”

A: I think that there is a tremendous opportunity there. Historically, engaging high BMI people in the context of obesity has been done through negative psychology. We have a tremendous opportunity to do that with positive psychology.


3. Devices and Bariatric Surgery

Oral Abstract Session: Drugs/Surgery


Philip Schauer, MD (Cleveland Clinic, Cleveland, OH)

Dr. Philip Schauer offered his view on whether randomized controlled trials in surgery were feasible by discussing the recruitment of the STAMPEDE trial. STAMPEDE was a randomized controlled trial (RCT) that compared the effectiveness of bariatric surgery (gastric bypass and sleeve gastrectomy) versus intensive medical therapy in achieving diabetes remission (A1c <6.0%) in obese patients (BMI 27- 43 kg/m2). The trial investigators prescreened 9,589 patients in order to achieve STAMPEDE’s 150 patient enrollment target, a process that took 46 months (28 months longer than the anticipated 18 months). Dr. Schauer cited patients’ perceived notion that surgery had excessive risk compared to medical treatment as the major cause for patient drop out during the enrollment process. While Dr. Schauer concluded that RCTs were feasible, he added the caveat that they were very challenging. Nonetheless, he was hopeful that lessons learned from STAMPEDE might pave the way for a definitive, large, multicenter randomized control trial in bariatric surgery that assessed hard clinical endpoints. Notably, Dr. Edward Livingston (Deputy Editor, JAMA, Chicago, IL) chaired this symposium, leading us to think of his TOS talk on the standards for bariatric surgery with regard to allowable attrition rates in bariatric surgery (see page 8 of our Day #2 report at We wonder how Dr. Livingston viewed this presentation and whether he believes randomized controlled trials can (or should, given ethical considerations) become the gold standard for therapeutic comparisons involving bariatric surgery.

  • The STAMPEDE study found that bariatric surgery was more effective than intense medical therapy in achieving glycemic control. The study randomized 150 patients in a 1:1:1 fashion to three possible treatment arms: intensive medical treatment (n=41; seven withdrew consent upon randomization and two were lost to follow up), medical therapy plus gastric bypass (n=50), or medical therapy plus sleeve gastrectomy (n=49; one withdrew consent prior to surgery). For an in-depth discussion of the study design and results, please see page 2 of our American College of Cardiology 2012 report at

Questions and Answers

Q: How does this experience differ from recruitment in other surgical randomized control trials? Is recruitment very difficult in any kind of surgical study?

A: When comparing Drug A vs. Drug B, the patient’s perception of differences is minimal, but here we have an invasive procedure. It’s difficult for patients to understand the risk. We did share with patients that there was potential for the surgery to cause diabetes remission and enable them to go off medications, but I think that the perceived risk of surgery is the underlying challenge in comparing surgical procedures versus pharmacological interventions.

Q: For people randomized into the medical arm, was there the expectation that they could not change their mind for five years and want surgery in the future?

A: We indicated that as new drugs and lifestyle interventions came out over the five years, we would add those. We fully expected that some participants would drop out.



Lance Davidson, PhD (University of Utah School of Medicine, Salt Lake City, UT)

Dr. Lance Davidson presented results from a prospective analysis of factors influence weight maintenance and regain following gastric bypass surgery. Specifically, the study examined if weight loss could be maintained with lifestyle choice, and how much weight regain is dictated by fat free mass and metabolic rate. At the time of surgery, participants’ (n=140) demographics (80% females, 88% Caucasian, mean age of 46.3 years) and baseline weight (mean baseline weight was not reported) were obtained. Two years post-surgery participants’ weight, body composition, diet, and physical activity were determined by a questionnaire and the time they were able to remain on a treadmill. Participants’ weight was measured again six years post-surgery. In his analysis, Dr. Davidson controlled for age, sex, and initial absolute weight loss (at year two). We think that his decision to use absolute weight loss rather than percent weight loss may have skewed his results as it does not account for a person’s baseline weight. Still, after controlling for these factors, weight regain at year six was significantly predicted by both fat free mass (p=0.018) and resting energy expenditure (REE; p=0.0012). On average, each additional kilogram of fat free mass (β=0.33) a person had at year two increased the amount of weight regain they had at year six, while each kcal/day of REE (β= -0.016) at year two slightly decreased the amount of weight regain. Dr. Davidson said that he had hoped that one of the lifestyle choice variables (diet, physical activity, or treadmill time) would be show a significant effect to provide an option that an individual could control and do themselves to maintain their weight loss. He hypothesized that these variables may not have been significant because there was little change in these measures during the course of the study to impact weight regain significantly, and he concluded that his results show the need for individuals to work harder to overcome the effect of non-modifiable characteristics such as either a low REE or large fat free mass.

Questions and Answers

Q: Did you look for correlation between fat free mass and REE?

A: Yes I did look at that and they were correlated. I did an analysis to see if the correlation was having a significant impact and it came close to significance, but wasn’t quite significant.

Q: I am not surprised that you didn’t find anything with diet modification. What percent did each of those variables play in weight regain?

A: I am going to have to get back to you on that.

Q: Did you look at the drop in RMR [resting metabolic rate] and weight regain?

A: I used absolute RMR, I looked at drop in RMR and it was not significant.

Q: Did you look at stratified fat free mass?

A: I did not do any stratified analysis.

Q: Why do you think you saw the difference in REE and fat free mass?

A: I do not know that, and I would really like to know. Dr. Gallagher at this meeting showed that organ mass drops could account for changes in REE, which could explain it.

Q: When you used your model did you use kilos or percent weight regain?

A: I used kilos of weight regain.

Comment: We do percent because you can’t really compare a big person and a small person. So I think that was a flaw in the analysis, and it could have impacted your conclusions. I would have used percent.

A: Thank you I appreciate that.

Q: I was wondering if you looked at the subset of high responders?

A: I have not done any stratification within this analysis.

Comment: We have done similar studies on short-term weight loss using dietary restriction and have found that changes in the oxidative phosphorylation genes predict who is going to respond the most. So I think that this is in support of your RMR findings.


Oral Abstract Session: Metabolic Mechanisms of Gastric Bypass


Niels Vrang, MD, PhD (The Institute of Chemistry University of Copenhagen, Copenhagen, Denmark)

Dr. Niels Vrang detailed his lab’s animal study of how the gut endocrine system changes following roux- en-y gastric bypass (RYGB). Dr. Vrang opened by championing the use of stereology (a mathematical method for modeling in three dimensions) as a time-efficient, systematic, uniform, and mathematically unbiased method for analyzing gut morphology. In their study, Dr. Vrang and colleagues retrieved RNA and histology samples from four segments of rats’ GI tract (biliopancreatic, alimentary, common channel, and colon) following RYGB. Data showed that after RYGB the length of the alimentary channel increased. Total mucosa volume greatly increased, with the most predominant effect observed in the alimentary channel. The number of L cells (cells found in the GI tract that secrete glucagon) more than doubled following surgery, though the constant L cell density suggested that the increase in L cell number was driven by the proliferation of the gut mucosa. To Dr. Vrang, this finding suggests that the density of L cells is coded in the different segments of the gut. Data also showed that the expression of preproglucagon – which codes for GLP-1, GLP-2, and Oxyntomodulin – increases in the common channel following surgery.

Questions and Answers

Comment: We have a paper in press that finds exactly the same result that you found. But I didn’t use stereology. We just simply took a one-centimeter piece from the ruling, one from the common limb, one from the biliopancreatic. It’s mainly the same results – a doubling of L cells, with no change in L cell density. So you can actually do it another way.

Q: What do you think is promoting this significant increase in mucosal hyperplasia? It seems that you found it throughout the length of the gut. It’s not simply becoming more similar to duodenal villous morphology.

A: I can’t really say what’s driving it. Maybe it’s some stimulus to the upper gut. There’s also an increase in the growth of the gut, an increase in the number of L cells, and increase secretion of GLP-1 as well as GLP- 2, which is known as a growth factor which may explain a growth in the epithelium where the food is coming in. I do know that when you bombard the jejunoileum, you get a massive growth in the epithelium.

Q: What age were the rats? Maybe there’s an effect of aging on these results.

A: I’m not sure. They were weighting roughly 500 grams, so I would assume that they were roughly three or four months of age. Then they were taken down five months after the surgery. So we’re talking about relatively old rats – in terms of rat lifespans.

Q: Does the gut also grow in humans after gastric bypass?

A: I would love to investigate that. I’m pretty sure it does. I don’t think the growth is just an inward growth. It’s a very dynamic organ. If you overeat, it grows in order to absorb all the nutrients. It makes sense that if you fast, the gut mucosa shrinks. I’m pretty sure it changes. It’s a very difficult thing to investigate. People would have to donate their entire intestines. It would be hard to get a control group.

Q: Microbiota is a hot topic. Did you look at that to see if it explains any of the results?

A: I’m not a big fan of microbiota, but that’s my personal thinking.



Clifton Jackness, MD (Columbia University, New York, NY)

Dr. Clifton Jackness’ presentation suggested that the rate of weight loss following Roux-en-Y gastric bypass (RYGB) surgery was responsible for the insulin sensitivity and secretion benefits observed in patients. Dr. Jackness reviewed his study comparing insulin secretion outcomes in patients with type 2 diabetes undergoing RYGB (n=11) to patients on a very low calorie diet ([VLCD]; n=14). The VLCD (500 kcal/day in an in-patient setting) was designed to replicate the rapid weight loss that is observed in RYGB surgery (seven to ten percent body weight reduction over a three to four week period). Baseline characteristics did not differ significantly between the two groups (mean A1c: VLCD 8.5% vs. RYGB 8.2%). Both the VLCD and RYBG arms in Dr. Jackness’ study had similar, statistically significant improvements in insulin sensitivity, acute insulin response to glucose, and disposition index after intervention; however, differences between the RYGB and VLCD groups at baseline and at follow-up were not significant. Previous literature has shown that RYGB induces insulin secretion benefits that weight loss alone does not confer; however, Dr. Jackness’ study implied that if the rate of diet-induced weight loss was increased to match the rate of weight loss in RYGB, weight loss alone may in fact be responsible for the benefits in insulin secretion. While we were interested to hear Dr. Jackness’ hypothesis that RYGB benefits could be attributed to the rate of weight loss and thus, could be replicated by diet, we note that the type of VLCD diet used in this study would be beyond challenging to implement in a real-world setting. Furthermore, we assume this type of rapid, diet-induced weight loss was difficult for the individuals in this study to maintain after leaving the tightly-controlled, in-patient environment. As such, we question the clinical relevance of the findings in the VLCD arm in this study. We would be interested for a follow-up study to compare weight maintenance (and consequently, insulin sensitivity and secretion) between the VLCD and RYGB arm after the initial intervention.

Questions and Answers

Dr. Samuel Klein, MD (Washington University School of Medicine, St. Louis, WA): The intravenous approach doesn’t challenge in the same way as oral tests. Do you have any data on this?

A: We’re still analyzing the data for all of that.

Dr. Klein: The changes seem quite identical in both groups, so one really wouldn’t expect to see a difference in the oral data. I think that this is very important. Can you speculate why rapid vs. slow weight loss matters?

A: That’s a good question. We didn’t expect to find this. This was the antithesis to our initial hypothesis. It could have to do with decreasing leptin levels or the degree of caloric restriction such that the gastric bypass forces the restriction whereas someone who goes on a diet when they are not watched in the same manner as in an inpatient study does not have as rapid weight loss.

Comment: You could have presented this as a study with rapid weight loss vs. slow weight loss and see how those compared. You should look at the rate of weight loss on these parameters.



Blandine Laferrere, MD (Columbia University, New York, NY)

Dr. Blandine Laferrere described a study that investigated the short-term and long-term mechanisms of improved glucose control in people with type 2 diabetes following gastric bypass surgery. She opened by explaining that the two main mechanisms that mediate the improved glucose metabolism observed after surgery are: 1) the “gut effect”: following surgery, an acceleration of nutrient transit leads to increased GLP-1 levels, which then results in improved beta cell function; and 2) the “weight loss effect”: calorie restriction and weight loss can directly improve beta cell function, but can also improve hepatic insulin sensitivity and increase peripheral insulin sensitivity. Her study aimed to distinguish the “gut” vs. “weight-loss” effects to better understand whether gastric bypass has weight-independent contributions to diabetes remission. People with type 2 diabetes who underwent gastric bypass (n=9) were compared to lean and obese healthy control individuals. In the study, participants with type 2 diabetes who underwent surgery experienced a remission in their disease, characterized by normal levels of A1c, fasting glucose, and two-hour glucose. Following surgery, these participants exhibited enhanced and sustained levels of GLP-1 and GIP in response to a 50-gram oral glucose test, as well as a restored incretin effect. Furthermore, participants who underwent gastric bypass exhibited normal insulin secretion in response to both oral glucose and IV glucose infusion. These participants also experience rapid, complete, and sustained normalization of beta cell glucose sensitivity in response to oral glucose (the “gut” factor). However, they exhibited impaired beta cell sensitivity in response to IV glucose, and this effect persisted even three years after surgery, when patients had lost roughly 30% of their baseline body weight.

Q: I’m a little concerned about the insulin resistance measurement, as a lot of the conclusion rides on whether it changes or not. HOMA fails when beta cell sensitivity changes. In this case, it may not with respect to glucose, but there may be another component. You could probably calculate insulin sensitivity from glucose infusion clamps and it may be more accurate. Have you tried that rather than HOMA?

A: We’re in the process of looking at all of that. We now have 15 people in the study and we’re looking at insulin sensitivity in different ways.

Q: Regarding the marked improvement in insulin sensitivity when the BMI was still quite high (around 30 kg/m2), did you look at central fat? It could be that there was a huge loss of central fat.

A: We did not look at central fat in this cohort, but we’re looking at it in another cohort.

Q: Can you speculate on the role of diet in improving some aspects of beta cell function? Other than gut hormones, what could be responsible for improvements in beta cell function? I’m wondering, in regard to regular weight loss, whether a change in diet could have an effect on beta cell function similar to that which you see with surgery.

A: The beauty of using both oral glucose and IV glucose is that, if the diet and weight loss had a huge impact on beta cell function, you would have seen it during the IV glucose stimulation. I think the beta cell remains impaired in those patients.


Poster Presentations – Intervention Studies (other adult)


A Escalona, F Pimentel, A Sharp, C Munoz, D Turiel, C Gomez, M Slako, L Ibanez, S Guzman

In this Chilean study, GI Dynamics’ EndoBarrier was implanted in 43 subjects (baseline weight 110.8 kg [244 lbs]; BMI 44.1 kg/m2; age 35 years; 76.7% women). During the one-year treatment period, the EndoBarrier was removed from 16 patients due to device movement (9), line obstruction (3), abdominal pain (2), acute cholecystitis (1), and subject request (1). The 27 patients who completed one year of treatment with the EndoBarrier experienced an average 22.8 kg (20.2%) weight loss from baseline. The subset of participants for whom one-year post-explant data is available (n=8) retained 16.1 kg (14.4%) weight loss one-year post-explant (after one year of implantation, they had achieved an average 21.6 kg [16.1%] weight loss from baseline). LDL-c remained significantly lower one-year post-explant than at baseline (HDL appeared to stay elevated, although not significantly above baseline). In contrast, both systolic blood pressure and diastolic blood pressure rebounded from the time of explant to one-year post-explant (no significant difference from baseline); triglycerides and fasting glucose one-year post- explant were numerically higher than at baseline, though not statistically significant. It is encouraging to see that the weight-loss effect of the EndoBarrier is durable out to one-year post-explant; we’re curious to see longer-term data, and whether re-implantation of the EndoBarrier could be considered for those who regain some of the weight they lost.


Symposium: What Outcomes Matter – Prevention and Treatment of Pediatric Obesity to Reduce Diabetes Risk


Thomas Inge, MD (University of Cincinnati, Cincinnati, OH)

Dr. Thomas Inge’s presented his view that bariatric surgery was beneficial in the treatment of obesity and diabetes in adolescents. He proposed that obesity and diabetes in adolescence is a more virulent disease than in adulthood because adolescent weight status predicts adult disease and because more adolescents are likely to fail diabetes therapies than are adults. He asked, “Shouldn’t we be thinking about more aggressive treatment?” Dr. Inge proposed gastric bypass as a solution because it reduces fat mass, improves insulin resistance, and improves beta cell function. Echoing Dr. John Kral (SUNY Downstate Medical Center, Brooklyn, NY) from the day before, Dr. Inge emphasized that gastric bypass wasn’t just “bariatric surgery,” it was “metabolic surgery”, as it conferred metabolic benefits beyond weight loss. Dr. Inge concluded with his belief that surgical models will provide important clues to disease development, which he hoped clinicians and researchers would soon be able to exploit to get more tools for the obesity and diabetes treatment toolbox. Given Dr. Inge’s favorable attitude towards bariatric surgery, we wished he would have provided more specificity as to when and for whom he thought bariatric surgery should be used. For example, we are curious what Dr. Inge’s opinion is on the BMI threshold for considering bariatric surgery in adolescents. Additionally, during the Q&A discussion, Dr. Inge referred to pulling out risk factors from adult literature for bariatric surgery failure to apply to adolescent practice – it was our understanding that even in adult populations these risk factors were not well established and we would be interested to learn which should be considered most important or most applicable to a younger population.

Questions and Answers

Q: In our area we do 800 to 1200 bariatric surgeries a year. One observation I’ve made in our clinical practice is that younger patients are quicker to gain back weight post operation. I was surprised not to see anything on psychological outcomes in your talk. My concern is not just whether patients are losing weight or improving insulin resistance, but what the psychological implications are on young minds from this surgery. In general, if we are going to investigate in younger children and if you are recommending we do it in younger patients, do we need to focus on pre- and post-surgical psychological evaluations?

A: It’s not a lack of interest. In the Teen-Longitudinal Assessment of Bariatric Surgery [a study which has recruited 242 adolescents to undergo RYGB or sleeve gastrectomy] we are focusing a lot on behavior on the front and back end. I firmly believe that we have to evaluate the family on the front end and sort out what the risk factors are for failure of surgery. At least, we can borrow from adult literature and make sure they are addressed in adolescents. I don’t know what to say about weight regain. We are not appreciating weight regain in adolescents any greater than in adults. But we are concerned about missing data (i.e., when adolescents go off to college and get married and stop thinking about their bariatric surgery) and we are dedicated to going and getting that missing data. Do adolescents have a higher risk of failure? I don’t believe its true.

Comment: Follow up is very poor and short in bariatric surgery.

Comment: I’m sure at your site you have a psychologist who gives approval for surgical intervention.

A: Absolutely. That is the standard.

Comment: I want to commend you on the excellent work you are doing. I know the care you are taking before you operate. This is the big dilemma – there is no substitute for having the family involved.


Symposium: Beyond Weight Loss – Metabolic Changes Following Bariatric Surgery (Jointly Sponsored by American Society for Metabolic & Bariatric Surgery [ASMBS] and The Obesity Society [TOS])


Jonathan Purnell, MD (Oregon Health & Science University, Portland, OR)

Dr. Jonathan Purnell presented results from the Longitudinal Assessment of Bariatric Surgery (LABS) 3 Diabetes Substudy (manuscript in preparation) which looked at various measures of beta cell response before and at six months after gastric bypass (GBP) in people with and without type 2 diabetes (n=39 and 22, respectively). He began by reviewing two contributing factors of hyperglycemia: insulin resistance (beta cell workload) and diminished beta cell response (insulin deficiency). Study results showed that 1) GBP improved insulin sensitivity in people with and without diabetes (which, he explained was largely due to weight loss); 2) GBP increased insulin secretion and disposition index in those with diabetes (which, he said was likely independent of weight loss); and 3) GBP reduced pro- insulin levels in those with and without diabetes. Notably, GBP also altered glucose absorption patterns, such that glucose absorption was enhanced and occurred earlier compared to baseline. Dr. Purnell concluded that this change could predispose some patients to hyperinsulinemic, hypoglycemia.

  • The LABS 3 Diabetes Substudy measured beta cell response in patients with and without diabetes before (n=41 and 22, respectively) and six months after (n=39 and 22, respectively) gastric bypass surgery (GBP). Both groups had 26% weight loss with GBP. Measures of beta cell response were taken over the course of 240 minutes after mixed meal ingestion.
  • Insulin sensitivity improved six months after GBP in those with diabetes (p=0.004) and in those without diabetes (p=0.02). The improvement in insulin sensitivity was unsurprising, and Dr. Purnell explained that previous studies have shown that the primary determinant in improved insulin sensitivity is the weight loss itself (Nanniperi et al., JCEM 2011).
  • Basal insulin secretion in those with and without diabetes was significantly decreased six months after GBP, which Dr. Purnell said was reflected by the patients’ lower blood sugars. Total islet cell secretory capacity also improved in patients with type 2 diabetes (p=0.008); however in those without diabetes, secretory capacity did not change significantly. Broken down between dynamic islet cell secretion (i.e., the pool of insulin available) and static islet cell secretion, Dr. Purnell found that dynamic secretion significantly decreased in people without diabetes (p=0.002), but remained unchanged in people with diabetes, whereas the reverse was true for static islet cell secretion – static secretion improved in people with diabetes (p=0.002) but not in those without diabetes.
  • The disposition index significantly improved in people with diabetes, but not in those without diabetes. Dr. Purnell explained that this meant individuals without diabetes were reducing their insulin secretion in proportion to their increase in insulin sensitivity, whereas those with diabetes had actual changes in this ratio and were not appropriately reducing insulin secretion in response to increased insulin sensitivity.
  • Pro-insulin levels fell in both patients with and without diabetes six months after GBP; however, prior to surgery pro-insulin levels were higher in patients with diabetes than in those without (p=0.001) and six months after surgery there was no significant difference between groups (p=0.05).
  • Interestingly, GBP led to enhanced and earlier glucose absorption such that glucose absorption occurred ~30 minutes earlier than at baseline. Dr. Purnell posited that this could predispose some patients to hyperinsulinemic hypoglycemia. Prior to GBP, peak glucose absorption occurred ~60-90 minutes post mixed meal stimulation. During the Q&A that followed, Dr. Purnell explained that while he has had patients who have felt hypo, he has not had patients who have had clinically defined hypoglycemia (however, some of his colleagues have). Also highlighted during Q&A was that the altered glucose absorption pattern introduces significant dysglycemia, a possible point of concern (for both patients with and without diabetes). While Dr. Purnell acknowledged this possibility and noted it was something to be aware of, he emphasized that in patients with diabetes A1c went down; his “official line” would be that GBP was beneficial. We’ve learned in the past that glycemic variability is also very important for health outcomes, so we are curious for additional studies to illuminate whether the benefits in terms of weight loss, insulin sensitivity, and insulin secretion for the patient undergoing GBP outweigh the potential negative consequences introduced by the altered post-meal glucose absorption pattern.

Questions and Answers

Q: This was an oral not an intravenous load?

A: Correct.

Q: Did you adjust mixed meal loads for changes in weight?

A: We kept it constant

Q: Did you rescue the beta cells? Do they really retain their normal secretory pattern?

A: Our study has the same issue that others have in that we don’t have a weight loss control group or normal control group to compare to. I don’t know if it gets all the way back to normal; I would suspect it doesn’t.

Q: You said your patients felt hypoglycemic, but it appeared to me they had a high rate of change of glucose. Did you ever see real documented hypoglycemia?

A: For the patient I showed, when she felt hypoglycemic, she had neuroglycopenia, but we think it was due to the rate of fall. There are some who get down around a clinically defined hypoglycemia level; my colleagues have had real hypoglycemia, but I haven’t.

Q: For mechanisms besides the effect on GLP-1 and GIP, do these mechanism relate to insulin sensitivity in the muscle or liver?

A: We did not do clamps, which would have allowed us to look at specific organs.

Q: I’m concerned about the degree of hyperglycemia. What are your feelings in terms of cardiovascular disease risk?

A: The A1cs get better. My official line would be that this is beneficial. The role of dysglycemia has been associated with heart rhythm issues. It’s something to be aware of it.

Comment: To answer an earlier question, we measured insulin secretion rate and beta cell sensitivity and showed full recovery of beta cell in oral stimulation but no change in up to three years in beta cell response under intravenous stimulation.

Comment: I would be cautious about the lack of improvement in disposition index in the group without diabetes. The trends look identical even though they were not significant.

Q: What do you think of the glucose excursion in non-diabetic patients? It looks harmful; increased glucose excursion has been implicated in CVD.

A: Our study clearly shows the excursion. If you look at long-term studies though, CVD goes down and mortality goes down.

Comment: At least we should be aware that the pattern may not be such a beautiful thing.

A: I agree; we are in unchartered territory.

Q: Do you have anything on insulin clearance?

A: I assume it improved, but I don’t have the data yet.



Dympna Gallagher, EdD (Columbia University, NY, NY)

In front of a packed audience (with at least twenty people standing along the wall) Dr. Dympna Gallagher examined the mystery that bariatric surgery decreases people’s resting energy expenditure (REE) relative to their fat free mass (FFM). In an attempt to determine if changes in a person’s body composition might be the explanation, Dr. Gallagher and her team performed a LABS (Longitudinal Assessment of Bariatric Surgery) ancillary study (n=105, 87 of whom were women) to identify and measure components of weight change in patients who have undergone bariatric surgery (73% had RYGB, 13% had a gastric sleeve, 14% had other forms of bariatric surgery) and to determine which components impacted REE. Specifically, Dr. Gallagher hypothesized that though one’s organs are light relative to the body, because they are so metabolically active (accounting for over 50% of REE) their shrinking due to bariatric surgery could account for the drop in REE relative to FFM. She found that organ mass as a percentage of FFM decreased significantly after bariatric surgery (from 6.0% pre-surgery to 5.4% both one and two years post-op). When the researchers analyzed REE relative to FFM adjusted for organ mass they found that there was no significant difference between the surgery and control group at time of surgery (p=0.287), or one (p=0.765) or two years (p=0.265) after surgery. Thus, Dr. Gallagher concluded that, at least in this cohort, adjusting for the decrease in organ mass after bariatric surgery explained the difference in REE between the treatment and control arms.

  • In addition to changes in organ size, Dr. Gallagher noted shifts in the composition of adipose tissue, particularly subcutaneous adipose tissue (SAT), intermuscular adipose tissue (IMAT), and visceral adipose tissue (VAT). All three decreased, but some dropped more relative to total adipose tissue (TAT) than others. In the female participants (n=87) SAT as a percentage of TAT increased significantly from baseline to one year after bariatric surgery (89.8% to 91.2%). Similarly, IMAT as a percentage of TAT was significantly increased one year post-op (3.3% to 4.3%; p=0.0125), while VAT as a percentage of TAT was significantly higher a year after surgery (6.8% to 4.6%; p<0.001).


Symposium: Bariatric Surgery: Mechanistic Insights (Jointly sponsored by the American Society for Metabolic & Bariatric Surgery [ASMBS] and The Obesity Society [TOS])


Randy Seeley, PhD (University of Cincinnati, Cincinnati, OH)

Dr. Randy Seeley packed a punch in his presentation exploring the metabolic effect of vertical sleeve gastrectomy (VSG). He argued that VSG confers benefits through a physiological change similar to Roux-en-Y gastric bypass (RYGB), as opposed to a mechanical change (restriction), the mechanism through which gastric banding works. Dr. Seeley commented that he no longer subscribes to either of the two thought camps in the ongoing debate as to why RYBG works – one camp argues that the effect is due to undigested chyme being pushed further down the gastrointestinal tract (and thereby stimulating GLP-1 secretion, amongst other effects) and the second camp argues that RYGB works because there are no nutrients touching part of the stomach and duodenum. He reasoned that since VSG is shows the same metabolic effects on glucose regulation as RYGB and is neither a hindgut nor a foregut procedure (i.e., the surgery doesn’t cause nutrients to bypass the foregut nor does it put undigested chyme in the hindgut), it is possible that RYGB’s metabolic effects are not produced through either of these mechanisms. Perhaps to the chagrin of some audience members, he went on to argue that the physiological mechanism of VSG (and thus, likely RYGB as well) was not dependent on GLP-1, but instead was tied to bile acids. While Dr. Seeley made a compelling argument, his conclusions should be interpreted cautiously as the studies were all conducted in murine models, and may not directly translate to humans.

  • Dr. Seeley stated that restricted food intake is a poor explanation for why vertical sleeve gastrectomy (VSG) is effective. He walked the audience through findings from three rat studies as support. First, VSG produced a transient reduction in food intake in rats. Three weeks following VSG, the animals’ food intake returned to pre-operation levels. Second, animals actively defended their new post-VSG body weight, even when this meant consuming 25% more calories than pre-operation. Third, when female rats post-VSG became pregnant and gave birth to their pups, they were able to double their caloric intake to replace the calories drained through lactation. Dr. Seeley explained that the natural tendency for rats’ food intake to return to normal levels following VSG and the impressive ability for the rats to consume calories even beyond this level made gastric restriction an unlikely mechanism behind VSG.
  • It is unlikely that the behavioral changes after VSG and RYGB are just due to patients becoming more compliant. Rats, explained Dr. Seeley, are reasonably good at mixing their diet appropriately when presented with jars of carbohydrate, fat, and protein feed. Dr. Seeley said that following VSG, the sensible response of an animal with a restricted stomach should be to increase fat intake so as to get more calories from less food. However, VSG rats actually shifted their food preference away from fat and towards carbohydrates. Dr. Seeley suggested that it was unreasonable to think that patients are willing to change their diet so dramatically after surgery primarily as the result of the surgeon’s instruction (as many have suggested) when they were not able to change before. It was more reasonable, he said, that biology was impacting dietary preference, as in the case of the rats.
  • The effects of VSG on glucose regulation parallel RYGB. Dr. Seeley emphasized that patterns of insulin secretion in both RYGB and VSG changed in similar ways and, importantly, that these changes were not dependent on weight loss. Furthermore, rats that underwent RYGB had an enormous increase in GLP-1 post-operation compared to rats that lost the same amount of weight by diet. This occurred in VSG as well.
  • Because VSG is neither a hindgut nor a foregut procedure and confers the same benefits in glucose regulation as RYGB, Dr. Seeley suggested that the effect of RYGB was not due to changes in nutrient flow. The two predominant theories to explain the mechanism of RYGB start with a trigger associated with the altered digestive route. One camp argues the mechanism of action relates to nutrients bypassing the stomach and duodenum; the second camp argues that RYGB works because undigested chyme is pushed into the hindgut (and stimulates GLP-1 secretion, etc.). Dr. Seeley posited that neither is true, and instead, a change in the gene expression pattern of the greater curvature of the stomach is mediating the effects of VSG. (As a reminder, VSG excises part of the stomach such that the lesser curvature of the stomach is left intact, whereas the greater curvature of the stomach is altered.)
  • Neither ghrelin nor GLP-1 seems to mediate the effect of VSG on fat mass or glucose regulation. Dr. Seeley compared ghrelin knockout and wild-type mice. Both showed the same magnitude of change in fat mass and glucose regulation pre and post-VSG, albeit the starting points were different, underscoring that ghrelin does play a role in fat mass and glucose regulation, just not a role in conferring the benefits of VSG. Similarly, when comparing GLP-1 receptor knockout mice and wild-type mice, the delta is the same for fat mass and glucose, with the latter trending towards a bigger effect in the GLP-1 receptor knockout mice.
  • Bile acids appear to play an integral role in conferring the benefits of VSG. Dr. Seeley’s group hypothesized that a component of bile acid could be responsible for changes in signaling that trigger the metabolic effect observed post-VSG. Dr. Seeley compared wild-type mice to whole body knockouts for the FXR bile acid receptor to show that VSG gives decreased fat mass and glucose regulation benefits in rats deficient for FXR; thus, bile acid signaling through the FXR receptor appeared to be implicated in mediating the mechanism behind VSG.

Questions and Answers

Q: How soon after surgery do changes in food preference occur?

A: We have not done a systematic study probing them at different times, so I don’t have a good answer. It clearly lasts well after they lost the weight, but I can’t tell you how soon after they lost the weight it occurred.

Comment: You suggested GLP-1 has nothing to do with decreased food intake, but there are a lot of people who still think it does.

A: I guarantee you that.

Q: What about the timing of these measurements? You would want to measure GLP-1 when food intake is really suppressed.

A: The high GLP-1 levels occur at a time when the food intake is totally normal. We’ve done it three weeks out and three months out. That’s the point here. The levels of GLP-1 are not the only determinant. In FXR knockouts, the only response intact in VSG-treated rats is that GLP-1 is increased. You can have high

GLP-1 that doesn’t seem to confer a benefit.

Q: Would the stomach have digestive enzymes that could mean something?

A: Obviously you change a whole mess of things when you cut away part of the stomach. We’ve measured pH, and the pH is way more basic in the stomachs of animals with VSG. Could that matter? Absolutely.

Q: Which means processing of glucose, lipid, and protein could be different.

A: Intestinal nutrient handling is altered. You’ve given a reasonable hypothesis about what contributes to nutrient handling.

Q: Can you put together the FXR findings and gene expression in the greater curvature of the stomach?

A: I could try, but it would be a futile effort. I don’t know the degree to which those hormones matter to the outcome and how that would relate to changes in bile acids. I don’t have a good way to put it together for you yet.

Q: What’s happening with gastric emptying in VSG?

A: It’s very high. Even though you have an intact pylorus, gastric emptying is very high. To make one last note here, the high GLP-1 is not dependent on the high gastric emptying.



Lee Kaplan, MD, PhD (Massachusetts General Hospital, Boston, MA)

After reviewing the effects of obesity and Roux-en-Y gastric bypass (RYGB) on intestinal microbiota and the effects of microbiota on obesity, Dr. Lee Kaplan suggested that the appropriate manipulation of microbiota could induce the same physiological changes as bariatric surgery. However, because Dr. Kaplan used murine microbiota transplant experiments to demonstrate these principles, the conclusions can, of course, not be directly applied to humans. Nonetheless, Dr. Kaplan’s series of experiments provide excellent fodder for discussion and reason for additional investigation into the role of intestinal microbiota in regulating adiposity and obesity in humans.

  • Host and microbiota is a two-way street: diet can shape intestinal microbiota, and microbiota can provoke host responses. To substantiate the former, Dr. Kaplan showed intestinal microbiota from mice that switched from a low-fat diet to a high-fat diet. The change (and obesity that followed) led to an increased concentration of firmicutes and decreased concentration of bacteroidetes in the gut. To substantiate the latter, Dr. Kaplan presented murine microbiota transplant experiments. Mice were raised in three ways: 1) mice were born and maintained in a germ-free environment (i.e., no intestinal microbiota [GF]); 2) mice were conventionally raised (i.e., normal microbiota [conventional]); and 3) mice were born in a germ- free environment, but then administered bacteria (i.e., manipulated microbiota [conventionalized]). GF mice conventionalized with bacteria from lean mice showed less increase in body fat than GF mice conventionalized with bacteria from obese mice.
  • Intestinal microbiota can play a metabolic role in the host by influencing a number of signaling pathways including: 1) inflammatory signals, which in turn can regulate insulin sensitivity; 2) short-chain fatty acids, which affect adiposity, and which can also regulate inflammatory signals; and 3) bile acids, which can regulate adiposity, influence GLP-1 production, and regulate energy expenditure.
  • Roux-en-Y gastric bypass (RYGB) treatment affected gut microbiota, independent of weight loss, and gut microbiota from RYGB-treated mice independently induced leanness in obese mice (i.e., with no concomitant surgery). Dr. Kaplan detailed the following experiment. Diet-induced obese mice were randomized to one of three treatment arms: 1) RYGB (surgery with weight loss); 2) sham surgery ([SHAM] surgery without weight loss); or 3) restrictive dieting weight-matched with sham surgery ([WMS] weight loss without surgery). Fecal sampling pre- and post-surgery revealed a dramatic change in intestinal microbiota in the RYGB group that was not seen in the SHAM group and not seen to the same degree in the WMS arm. The most dramatic difference in the RYGB group was the increase in enterobacteriales. Next, microbiota from mice in the RYGB and SHAM arm were transplanted into GF mice. GF mice conventionalized with RYGB microbiota experienced a decrease in body weight and adiposity whereas GF mice conventionalized with SHAM microbiota experienced an increase in adiposity and no change in body weight.
  • Interestingly, when GF mice were conventionalized with RYGB microbiota, food intake did not decrease, suggesting that the transplanted microbiota were affecting a signaling pathway that regulated energy expenditure.


Symposium: Losing Weight and Not Your Patients: Maximizing Adherence and Minimizing Attritions


Edward Livingston, MD (Deputy Editor, JAMA, Chicago, IL)

Dr. Edward Livingston’s presentation on attrition in bariatric surgery studies incited a heated Q&A discussion. One of his main points was that bariatric surgery works well for some patients in some circumstances – and that it should not be viewed as a panacea. Generally, we think people already agree on this but we understand why Dr. Livingston would disagree with those who advocate for use of bariatric surgery very broadly. Dr. Livingston spoke from his perspective as Deputy Editor for Clinical Content at JAMA, which we felt made his perspective quite interesting, given the number of studies he has likely seen; specifically, he pointed out that most bariatric literature did not meet high enough standards for JAMA in terms of patient retention in particular. For JAMA, he said, the rule of thumb for any study is that retention should be 80% or higher; however, for bariatric surgery, he said the journal had lowered the requirement to 75% retention at two years of follow-up because few papers actually reached the 80% threshold. (This does not seem like a hugely meaningful change from our view, but we certainly agree about the importance of follow-up; in today’s mobile society, the notion that 25% or more would be lost to follow up may not be too surprising for any study.) Indeed, Dr. Livingston and JAMA associates put together a systematic review of bariatric surgery literature that found that only 8% of all published studies fulfilled these criteria. (We’re following up on the denominator – it is unknown how many studies they examined.) Attrition introduces bias, he explained, as most people drop out of studies when the therapy is not working or when they have adverse events. He pointed to a study (Riele et al., British Journal of Surgery 2010) that tracked down patients who had been lost to follow-up in a study with a gastric banding intervention. As it turned out in this particular study, the patients who had been lost in follow-up did not have as good weight loss as those who had been included in the follow-up – we’re uncertain the degree to which that finding is generalizable. The Q&A for this session was troubling in our view, with some particularly generalized points made about obesity trial participants – Dr. Livingston made a remark during the early part of Q&A that we found an unsettling way to characterize trial participants: “ …it’s the patients themselves. Many are not very reliable people...”  Furthermore, Dr. Livingston put forward his view that the FDA’s allowable attrition rates for obesity medications were too high. He explained that the FDA requires 5% total body weight loss, but to his understanding, only requires one measurement before and after pharmaceutical intervention and has no threshold of allowable attrition – a major concern in his eyes when most obesity drug studies have 40-60% attrition: “It’s good enough for the FDA, but not good enough for our journal.” We are curious if Dr. Livingston and JAMA have discussed this directly with FDA, which he did not mention. It is also true that more recently, there are certainly pharma studies that have had considerably higher attribution rates; Vivus’ Qsymia trials, for example, have routinely shown attrition rates at 60% and higher and there are other examples as well. Dr. Livingston concluded by reminding the audience to consider studies with low follow-up with a grain of salt. These studies, he said, should not be considered definitive. While we agree that definitive information is hard to come point, it is seems logical that the less efficacious treatments and the treatments with more troubling side effect profiles would have higher attrition; with better treatments being tested and easier ways of communicating with patients emerging, ideally attrition will improve in trials moving forward. In terms of assessing past trials, we agree that attrition can be a major weakness, though would caution not to throw out the baby with the bathwater, given the relatively few number of obesity therapies and technologies being tested, relative to the size of the problem. We did find this session valuable in understanding more about JAMA’s guidelines – we continue to appreciate JAMA as among the highest quality journals to which authors can aspire.

Questions and Answers

Q: Why do you think for this population in particular [those receiving bariatric surgery], the literature has been limited by very poor follow-up?

A: The standards in surgical literature are not particularly great. One, it has to do with where these studies have been published. And two, it is the patients themselves. Many are very unreliable people. When things go wrong, they often disappear. It’s a combination of the standards applied to the field and the people themselves.

Q: The strategies for increased follow-up require extra resources and run-in periods. Can you comment on how giving surgery for free might change the generalizability of those results?

A: It’s exactly generalizable to a patient getting their care for free. It’s a very significant restriction; the study is only generalizable to the condition in which the trial was performed. So these therapies may or may not work.

Q: If 8% of bariatric literature passes an acceptable threshold, what should we think about bariatric surgery?

A: I’m considered a pariah because I’ve spoken negatively about the field, but the literature is getting better. We’ve seen a couple studies in the New England Journal of Medicine and we have three in this issue of JAMA. As they get better, we know more. I do perform bariatric surgery in practice, but we should keep in mind this surgery works extremely well for some patients in some circumstances. I do agree the current literature does not support what a lot of my colleagues advocate in terms of bariatric surgery.

Comment: You’re wrong, at least for certain populations. In our recent trials we’ve used all the strategies possible for retention. They drop out from clinical care at the same percentage they drop out of studies. There are certain populations where we’d be lucky to have 80% retention.

A: In psychiatric studies, for examples, it changes. If you are studying schizophrenia and lose a lot of patients because of the nature of disease, you have to interpret that result in the context of there being a subpopulation that is compliant and a subpopulation that is noncompliant. It doesn’t change the science; it changes how you interpret the results.

Q: I completely disagree that we can blame the patients for poor follow-up. Every patient already has a lot of stigma. To lump them into the notion of “they’re obese and are probably not going to follow up” is probably against the notion that everyone in this room is trying to put forward. I’ve read a lot of the journals you are discrediting and there is good information in there. The societies [TOS and ASMBS] are coming together, and hopefully you won’t be the ultimate pariah once again.

A: I think you’ll notice that most insurance companies have a schizophrenic policy on bariatric surgery and the reason for that is they consider the evidence for bariatric surgery to be inadequate. The only way we can convince them bariatric surgery should be paid for it to provide evidence that is unassailable. (Editor’s note – Dr. Livingston didn’t address the original question.)


Mickey Stunkard Lifetime Achievement Award


John Kral, MD, PhD, FACS (SUNY Downstate Medical Center, Brooklyn, NY)

Dr. John Kral, the recipient of the Mickey Stunkard Lifetime Achievement Award for a lifetime of outstanding contributions to the field of obesity in terms of scholarship, mentorship, and education, presented a “smorgasbord” of lessons learned from his career. His overall presentation (besides being an homage to “Metabolic Mickey” and “Surgical Stunkard”) suggested that bariatric surgery was not the appropriate term for the surgical treatment of chronic over-nutrition. First, he argued that bariatric surgery is “truly not about the weight”, but rather health, and second, that bariatric surgery includes types of operations that are “too different to lump together.” Dr. Kral emphasized that pure gastric restriction and gastroinstestinal diversion are very different physiologically – he pointed to a number of studies that showed a “whopping difference” between the effects of gastric banding and gastric bypass. Dr. Kral also urged researchers to be careful in translating lessons learned from studies in precocial species (i.e., rats that do no eat in the absence of deprivation) to altricial species like us. Furthermore, he warned that misconceptions plague studies of energy balance and that readers should not trust energy balance studies that are not done in steady state conditions. Dr. Kral concluded with a message of clinical relevance, explaining that cognitive behavioral therapy likely has a role in post-surgical management and that providers should talk to their patients about “satiety” and reasonable expectations after surgery.


4. Policy and Practice

Preconference Forum: How Evidence Can Drive Obesity Policy


William Dietz, MD, PhD (Former Director, Division of Nutrition, Physical Activity, and Obesity, CDC)

After reviewing several statistics on the prevalence of obesity in the United States, Dr. William Dietz discussed the progress made in various strategies for obesity prevention and management. Overall, he seemed optimistic that we as a nation are trending in the right direction with regards to most target behaviors for obesity prevention and management. In particular, he was encouraged by recent progress to address childhood obesity; meanwhile, he highlighted that more attention needs to be directed to addressing severe obesity, as current initiatives are unlikely to focus on this subpopulation that generates a disproportionate amount of healthcare costs. Finally, Dr. Dietz emphasized that prevention and treatment of obesity must be prioritized in the medical setting – clinicians should be trained, and bias against obese individuals should be addressed, such that patients in need can obtain quality care.

  • Dr. Dietz emphasized that we need to be much more aggressive in addressing severe obesity. He noted that if current trends continue, 11% of the adult population in the US will be severely obese by 2030, meaning a BMI over 40. In 2005, the 3% of people with self-reported BMI above 40 kg/m2 accounted for 20% of the costs of obesity, and the 5% of people with BMI between 35-40 kg/m2 generated another 20% of the costs of obesity. Unfortunately, current initiatives are not likely to specifically address severe obesity; Dr. Dietz believes changes in the care delivery system will be required to move the needle.
  • He seemed optimistic about recent progress to address childhood obesity. He noted that the energy deficits required to hit the Healthy People 2010 and 2010 goals would be relatively easy to achieve in the pediatric population through changes such as substituting water for juice in after-school programs, or non-sugared cereal in place of sugary cereal. He highlighted that in a September 2012 issue brief released by the Robert Woods Johnson Foundation, the prevalence of obesity in the pediatric population decreased in Philadelphia, New York City, Mississippi, and California.
  • Dr. Dietz reviewed the progress made in strategies for obesity prevention and management:
    • Reducing obesity risk factors during pregnancy: Dietz stated that limited progress has been made to address weight gain, smoking, and gestational diabetes during pregnancy.
    • Decreasing consumption of high-energy-density foods: Dr. Dietz highlighted the Healthy Food and Sustainability Policy (HFSP), a policy adopted by the Department of Health and Human Services and General Services that established guidelines and restrictions for foods purchased and served in their facilities as a good example of the progress made in this strategy. Specifically, the HFSP called for: 1) limits on trans fat, saturated fat and sodium; 2) no added sweeteners to fruits; 3) daily offerings of at least one raw and one prepared vegetable; 4) whole grain options for bread and pasta as a standard choice; 5) all milk 2% fat or less; 6) low-fat protein entrees and vegetarian entrees at least twice per week; 7) at least 50% beverages other than 100% juice less than 40 calories per serving; and 8) deep-fried options limited to one choice per day (! we are not sure we heard this correctly?). At TOS 2011 in Orlando, Dr. Dietz stated his belief that the policy has the power to be transformative, and hopes that it trickles down to strategies from state and local governments, thereby potentially affecting foods purchased in public spaces, such as parks and recreational facilities.
    • Increasing fruit and vegetable intake: Dr. Dietz noted that this strategy has received a lot of attention, and that it could move us toward a model of regional food systems since some parts of the country make this much easier than others.
    • Reducing consumption of sugar-sweetened beverages (SSBs): Dr. Dietz commented that there has been a lot of movement on this front throughout the country. For example, the city government in Boston eliminated SSBs from all municipal services (e.g., police departments). Meanwhile, hospitals in the city have labeling systems (red, yellow, and green) for different drinks in their cafeterias, which has helped shift consumption toward healthier options. In addition, Dr. Dietz stated that he thinks that taxes on SSBs will eventually be implemented; it’s just a matter of when.
    • Reducing screen time: Dr. Dietz said that there has been limited progress in this strategy, and that it is a dead issue at the moment. He thinks that an upcoming report from the Federal Trade Commission that will reexamine the costs of advertising directed at children is going to provide us with another opportunity to open the conversation on food marketing to children.
    • Breastfeeding: Dr. Dietz stated that enormous progress has been made in the promotion of breastfeeding. The CDC provided funding for 90 Baby Friendly hospitals in 29 states, and there is other support for Baby Friendly hospitals in seven states.
    • Promoting physical activity: Dr. Dietz commented that some notable progress has occurred on this front. For example, ChildObesity180 has initiated the Active Schools Acceleration Project (ASAP) and the Healthy Kids Out of School Project (HKOS), which promote physical activity (as well as consumption of water, fruit, and vegetables over less healthy alternatives).
  • Dr. Dietz was pleased that the Affordable Care Act was upheld, given its provisions for prevention. Specifically, Title IV focuses on the prevention of chronic disease and improving public health, establishing a fund for community transformational grants projected to grow to $1.5 billion by 2015, and supporting innovation in the space (e.g., employer-based wellness programs). Title VI provides approximately $70 million in funding for the Patient Centered Outcome Research Institute (PCORI) for assessment of prevention, diagnosis, and treatment options in efforts to improve healthcare systems.



Scott Kahan, MD (Director, STOP Obesity Alliance)

Dr. Scott Kahan reviewed several recent advances in obesity policy (CMS coverage of intensive behavioral therapy and the approval of two obesity drugs – Arena’s Belviq and Vivus’ Qsymia), subsequently discussing several underlying factors (e.g., expansion of the evidence base, ongoing advocacy) that contributed to these welcome developments. While much progress has been made, Dr. Kahan noted that there remains much to be done, including expansion of coverage for obesity treatments, the incorporation of patient centeredness into the policy conversation, and the implementation of physician training for obesity management and treatment.

  • Citing several historical examples, Dr. Kahan asserted that policy intervention is essential for progress in public health. He emphasized that both top-down and bottoms-up interventions are needed, noting how this approach was successful in significantly reducing both infectious disease mortality and tobacco use in the 20th century.
  • Dr. Kahan highlighted two recent advances in obesity policy: coverage of intensive behavioral therapy for the management of obesity by the Center for Medicare and Medicaid Services (CMS), and the approval of Arena’s Belviq and Vivus’ Qsymia. With regards to the recent drug approvals, he commented that they provide a significant opportunity to improve health – in clinical trials, one-third to one-half of patients on Belviq achieved ≥5% weight loss, and over 60% of patients on Qsymia achieved ≥5% weight loss. Dr. Kahan stated that the approval of these two agents is particularly notable, as they are the first obesity medications to be approved in 13 years.
  • Subsequently, he discussed several factors underlying these policy advances: 1) expansion of the evidence base; 2) ongoing advocacy to interpret the evidence and inform policymakers; and 3) a shift toward policy decisions informed by evidence as opposed to outdated assumptions and biases.
    • Expansion of the evidence base: Dr. Kahan noted that in the past decade, there has been much more data on the efficacy of obesity treatments and their ability to improve health. In the Louisiana Obese Subjects Study (LOSS), intensive medical intervention in primary care led to substantial weight loss, and was maintained over the course of a year (Ryan et. al, AIM 2010). In the landmark Diabetes Prevention Program, modest weight loss led to significant improvements in preventing the development of diabetes in obese individuals with prediabetes. Recently, the USPSTF recommended that clinicians screen all adult patients for obesity and offer behavioral interventions to obese adults, based on studies that showed modest weight loss on the order of 8.8-15.4 pounds improves health (e.g., blood glucose, cardiovascular risk factors).
    • Ongoing advocacy: Dr. Kahan commented that ongoing advocacy is making a huge difference, in particular highlighting the recent obesity drug outcome measures roundtables held by George Washington University. Multiple stakeholders – obesity experts, researchers, clinicians, patient advocates, consumer advocates, and representatives from the FDA, CDC, and NIH – convened to discuss what role drugs play in obesity management, and how to approach the uncertain regulatory environment. The working group’s key findings were that: 1) medications for obesity must be viewed as obesity treatments and not merely weight-loss agents; 2) the drugs should be available for clinical use for individuals for whom use is safe and medically appropriate; 3) individuals should be characterized by their level of health, feeling, and functional impairment to inform the benefit/risk analysis; and 4) given concerns of medically inappropriate or unsafe use, the FDA could potentially employ a mechanism such as REMS or another alternative to allow the approval, distribution, and use of medications solely by those for whom the drugs are indicated.
    • A shift away from antiquated assumptions and biases to a more informed and objective position on obesity: Previously, obesity was predominantly viewed as an individual issue, whereas now it is seen as a population issue; obesity has evolved from being a matter of personal responsibility to being recognized as worthy of multidisciplinary interventions. Whereas obesity was formerly considered more a matter of appearance, it is now being increasingly recognized as a health issue. Success now focuses on incremental and sustainable weight change, as opposed to returning to normal weight. Instead of only focusing on prevention, attention is now on both prevention and intervention (for those who are already obese). This recent shift in how obesity is viewed should help to move policy forward to improve the management and treatment of obesity.
  • Dr. Kahan outlined a number of upcoming priorities in obesity policy: 1) obtaining Medicaid coverage for obesity medications; 2) defining Essential Health Benefit standards in the Affordable Care Act; 3) incorporating patient centeredness into the policy conversation; and 4) providing physician training for obesity management and treatment. Regarding Medicaid, only 12 states currently cover drug treatment for obesity, while 33 states explicitly exclude coverage. Meanwhile, the Affordable Care Act will require health plans to meet Essential Health Benefit standards by 2014; how these standards are defined could greatly impact obesity management and treatment. Currently, 20 of 153 state benchmark plans cover bariatric surgery, while 89 explicitly exclude coverage; four plans include coverage of weight-loss programs, while 124exclude such coverage. Dr. Kahan noted that the ultimate goal would be for obesity not to be treated any differently than any other health condition.

Questions and Answers

Q: I was wondering if you had any reps from third-party payers at your roundtables. I’m a bariatric surgeon, and based on my experience, there has been a lot of variability as to how and who can get covered for surgery.

Dr. Kahan: What our bottom-line belief was, as a result of our discussions, was that policymakers will cover treatments as long as there is good reason to do it. Insurance companies will be willing to cover surgery or medications as long as we can give them good reason to show patients get better, and that ultimately their costs can go down. One of the ways that we conceptualized this was to look more at the immediate benefits of these treatments, as opposed to health improvements down the line. Things like improvements in feeling and functioning can happen quite quickly. Patient-centered outcomes that make a big difference in terms of people’s lives (sometimes requiring minimal obesity treatment) are things that we think third-party payers will pay for if we can have a discussion and show that those things are improving.



Donna Ryan, MD (Pennington Biomedical Research Center, Baton Rouge, Louisiana)

Dr. Donna Ryan’s presentation focused on the current state of clinical practice surrounding obesity therapy. She began with a look at the barriers preventing adequate obesity treatment, including stigmatization and reimbursement issues; however, she noted that the cultural, social, and political environment surrounding obesity has been changing for the better (highlighting of course, the recent approval of two obesity medications, Belviq and Qsymia). Going forward, she hoped that improved physician education and better approaches to obesity treatment (i.e., ones that could be successfully implemented by a time-pressed physician) would greatly improve care. Dr. Ryan concluded with an optimistic outlook on obesity treatment in clinical practice – “We don’t have a bonfire, but we have a little bit of a fire and we are going to get better.”

  • Dr. Ryan reviewed the barriers that exist in clinical practice that prevent providers from adequately treating obesity. To start, diagnosis requires significant clinical judgment, as BMI and waste circumference vary by race, gender, and other factors (e.g., bodybuilders with lots of muscle and little body fat can be healthy but still technically obese according to BMI). There are also reimbursement challenges in obesity, both for office visits and medications. For comparison, hypertension office visits and the 200+ hypertension medications available are always reimbursed, she said. Stigmatization of patients prevents many from seeking treatment, whilst stigmatization of doctors prevents adequate professional education and thus, adequate treatment of patients (we were particularly dismayed to hear one obesity doctor in the audience say he is often labeled a “quack” by his peers). Obesity treatment is further complicated by infomercials and supplement advertisements touting quick weight loss fixes and of course, by patients seeking weight loss for cosmetic reasons as opposed to health reasons.
  • Encouragingly, Dr. Ryan felt that recent environmental changes surrounding obesity (i.e., by regulatory agencies and third party payers) would help providers embrace obesity treatment in the office. She highlighted the recent approval of two obesitymedications by the FDA; Dr. Ryan noted that modest weight loss (5-10%) can produce major health benefits, a fact that will be instrumental in driving the FDA’s approach towards obesity drugs and devices. (This comes from DPP historically – see details at the NIH website at She also pointed to changes in medical education that would improve obesity treatment – Dr. Ryan explained that 2012 will be the first year that the American Board of Obesity Medicine offers an obesity medicine certification process, and she remains hopeful that an obesity certification will soon achieve subspecialty status.
  • Dr. Ryan suggested that better HCP training (and support for the obesity certification program) as well as better systems approaches to help physicians provide obesity care in their offices, would greatly improve clinical practice. She pointed to the Louisiana Obese Subjects Study (LOSS) as evidence that with a minimal amount of training for PCPs, medical intervention in office settings can lead to weight loss (Ryan et al., Archives of Internal Medicine 2010).


Preconference Forum: Food Industry Outreach – How Easy Is It to Match Policy With Practice on Nutrition issues: Unintended Consequences?


Xaviar Pi-Sunyer, MD, PhD (Columbia University, New York City, NY)

Dr. Xavier Pi-Sunyer gave an overview of considerations that go into creating nutrition guidelines. He opened his presentation with a slew of statistics demonstrating how diet could play a role in preventing chronic disease. Many of the most prevalent chronic diseases in the US have diet as a major causative agent, including: 1) overweight and obesity (US adult prevalence: 60%); 2) diabetes (11%); 3) prediabetes (35%); 4) hypertension (34%); 5) cardiovascular disease (37%); 6) osteoporosis (50% of women and 25% of men over 50 years of age); and 7) cancer (41% over lifetime). Dr. Pi-Sunyer explained that people should eat in a way to prevent and treat chronic disease, which was a major consideration in crafting the 2005 and 2010 guidelines (Dr. Pi-Sunyer played a part in creating both). One challenge to creating nutrition guidelines, he said, was that dietary guidelines primarily seek to meet nutrient needs through consuming food (not supplements), and often, the evidence for food is not robust. He concluded by explaining why guidelines were apt to change over time. Guidelines change, he said, because knowledge about foods, nutrition, physical activity, health, and disease change. Moreover, changing patterns of health and disease lead to necessary changes in guideline focus. For example, he said, the 2010 guidelines made obesity a major theme.



David Mela, PhD (Unilever, Vlaardingen, The Netherlands)

Dr. David Mela’s presentation detailed the methodology that Unilever uses to translate global food recommendations into products (Unilever’s brands include Country Crock and Ben & Jerry’s). He gave the example of Unilever’s goal to keep children’s ice cream products under 110 calories as the quintessential example of Unilever’s process. The company begins with WHO’s mean energy recommendation for children aged six to 12 years old, then assumes that ~20% of a child’s caloric intake comes from in-between-meal snacks, and that a child will have three snacks a day; thus, Unilever determined that a single ice cream snack (i.e., one of a child’s three snacks in a day) should have at most 110 calories. The presentation took a provocative turn when Dr. Mela cast the food industry as the victims of “black hat bias,” whereby papers published with involvement by food industry employees are quickly stereotyped by academia as being less reputable. Similarly, while the food industry wants to work with academics, he explained that academics who work closely with industry risk being called “appeasers.” He asked whether the food industry was “categorically contaminated” and urged the audience to remember that the industry does not speak with a single voice – he said he disagreed with many of the nutritional messages from other big names in the food industry. Furthermore, he argued that healthcare providers react too emotionally to technical nutrition solutions (like artificial sweeteners to reduce sugar consumption), suggesting that HCPs are not considering the totality of evidence. Overall his talk seemed to place the blame on groups other than the food industry, certainly, groups other than Unilever, including: 1) consumers (“unless people change, there isn’t a sustainable business model” for industry to provide healthier, greener products); 2) healthcare providers (who often have an “emotional reaction” to technical solutions); 3) media and experts (who give conflicting messages about nutrition); and 4) regulatory agencies (e.g., a product has to have at least a 30% reduction in fat, sugar, or energy in order to claim a reduction on the label, which discourages industry from making smaller changes in products). While the obesity epidemic is complex, with many organizations and people contributing to its severity, we couldn’t help but have an “emotional reaction” to Dr. Mela’s portrayal of the problem.


Oral Abstract Session: The Influence of The Children’s Food and Beverage Advertising Initiative


Monique Kent, PhD (University of Ottawa, Ottawa, Canada)

Dr. Monique Kent reviewed her study on the impact of Canada’s Children’s Food and Beverage Advertising Initiative (CAI) on actual television advertising as well as children’s exposure to the advertising. For context, the CAI was created in 2007 and consisted of 16 large food manufacturers who each made pledges to decrease unhealthy food and beverage advertising targeted at children. By comparing advertising data from May 2006 to May 2009 or May 2001 for 11 unhealthy food and beverage categories, the study found that overall advertising volume increased on all stations, but decreased on children’s stations by 4.5% in May 2009 vs. May 2006. However, exposure to food and beverage advertisements among children aged two to 11 increased. Encouragingly, based on the categorization by the UK Nutrient Profiling system (where foods are categorized either as “less healthy” or “healthier”), the percentage of “less healthy” foods advertised on children’s stations decreased from 56% in May 2006 to 50% in May 2011. Interestingly, most of the improvements in advertising were by non-CAI companies, leading us to wonder what other incentives might have existed over this time for these companies to decrease “less healthy” advertising.

  • For background, in 2007 16 large manufacturers joined Canada’s food and beverage advertising initiative (CAI). Half of these companies pledged to only market healthy food options to children under 12 years old; the other half pledged to stop advertising to children under twelve years old. Of note however, each manufacturer developed its own nutritional guideline. Television advertising in Canada is also regulated by the Broadcast Code for Advertising to Children (e.g., no advertisement can be targeted at preschoolers), the Code Interpretation Guidelines (e.g., advertisements cannot show children eating quantities of food that are larger than normal), and by station specific policies.
  • To determine the effect of CAI, Dr. Kent and colleagues examined advertising data for May 2006, 2009, and 2011 from The Nielson Company, which provided data on 27 stations’ advertising 6:00 am to 12:00 am; these stations included 25 general stations and two children’s stations. The study considered advertisements for 11 unhealthy food categories (candy, chocolate bars, cookies, cheese, beverages, etc.) and examined the exposure to these advertisements in children aged two to 11 in two regions, Toronto and Vancouver.
  • Comparing May 2006 to May 2009, on all stations, there was a 38% and 39% increase in advertisements aired in Toronto and Vancouver, respectively, for the 11 food categories. Encouragingly there was 4.5% volume decrease on children’s stations in both markets; sugar and energy ads in particular were down on children’s stations. In terms of exposure, the number of advertisements viewed per month by children aged two to 11 years on all stations increased by 16.8% in Toronto and 6.4% in Vancouver. On children’s stations specifically, exposure increased 5.4% in Toronto and 2.5% in Vancouver.
  • According to the UK Nutrient Profiling system which categorizes food as “less healthy” or “healthier”, the nutritional quality of food and beverages advertised on children’s stations improved in May 2011 vs. May 2006; In May 2006, 56% of foods advertised were “less healthy” (44% were “healthier”) vs. 50% of foods in May 2011 (50% were “healthier”). Looking at average nutrients per 100 g of food advertised on children’s stations in both Vancouver and Toronto, the study found a decrease in carbohydrates and sugar in 2011 vs. 2006; however, discouragingly, total fat and saturated fat increased and fiber decreased. Of course, as Dr. Kent said, there are still big improvements to be made.
  • Interestingly, the companies responsible for improving their advertising in terms of nutritional quality were not manufacturers in the CAI, but were non-CAI companies. During Q&A, Dr. Kent offered that this could have been due to increasing pressure in the environment causing non-CAI members to change their products as well.
  • The study was limited in a number of ways: 1) it compared data from only one month; 2) a 2010 change in The Neilson Company’s methodology prevented a comparison of exposure in 2006 to 2011; 3) only 11 food categories were considered; and 4) nutritional data for children’s exposure on non-children’s stations was not available.

Questions and Answers

Q: I noticed the companies that were not in the initiative started with a higher “less healthy” percentage than those who joined the initiative.

A: Exactly. I don’t know whether it’s coincidence or pressure in the environment that caused them to change their products.

Q: I am curious as to why you included yogurt in this study. I would think we’d want to increase this.

A: We didn’t just want to look at foods like chocolate and cookies. We really wanted to see what was going on with a wide variety of foods. Some yogurts do fall in the less healthy category. That’s why the nutritional analysis was so important at the end.

Q: How many of the channels monitored were US channels? Is there spillover from US to Canada?

A: The Nielson Company in Canada doesn’t tape any of the US stations. Certainly Canadian children would be watching TV from the US because the media border is permeable, but we only looked at Canadian stations. What is really interesting is that the initiative exists in both Canada and the US, but the companies are different and made different commitments. For example, in the US they’ve homogenized their nutritional criteria.


Poster Presentations – Policy: Environmental/Economic/Health Policy


G Srivastava, D Guanaga, L Kaplan

This study aimed to determine the frequency of obesity documentation by medical residents during the inpatient admission evaluation or discharge planning process among patients with obesity. The investigators conducted a retrospective review of 1,158 patients’ EMRs from stays at Massachusetts General Hospital. The overall prevalence of obesity was 35% (400/1,158) and 40% in patients 26-70 years old. Surprisingly, the retrospective EMR analysis revealed that not a single patient with obesity had a weight or BMI documented in any of the resident admission notes or discharge summaries. Further, residents addressed obesity in their assessments and plans only 6% of the time overall, and a mere 14.5% of the time in patients with a severe BMI >40 kg/m2. We wonder why the rates of documentation were so low, especially at a renowned teaching hospital like MGH – perhaps it has to do a lack of focus on obesity in medical school, residents being desensitized to obesity since it is so common in the United States, or underoptimized EMR processes. The one bright spot was that residents were twice as likely to document obesity if it was previously noted in the outpatient EMR problem list. To us, this represents a clear case of why good record keeping through smart EMR systems is especially critical – we could imagine a reminder system that asks residents if the patient was obese each time they complete an inpatient admission note. The poster’s authors note that the resident inpatient admission note “plays a central role in defining the overall clinical history, status, needs, and plans for the patient.” As such, we hope this study draws attention to what seems like a low-hanging-fruit, high- bang-for-the-buck area to address, especially given the inpatient costs associated with obesity. More broadly speaking, we hope obesity draws more and more attention in medical schools around the country, both to educate students about the problem and to inspire more young leaders to enter primary care and endocrinology.


Symposium: Pay Now or Later


Theodore Kyle, RPH, MBA (ConscienHealth, Pittsburgh, PA)

After Mr. Theodore Kyle provided an overview of the CMS guidelines on obesity treatment, he highlighted a number of remaining gaps in coverage, and discussed the potential impacts of the Affordable Care Act. In conclusion, he noted that while federal policy on obesity coverage has had “a tortured history,” significant progress has been made in the past decade, but nonetheless, we have “miles to go before we sleep.”

  • Mr. Kyle described the current state of coverage for obesity treatment by Medicaid and Medicare. Currently, 47 states cover surgery in some fashion, 13 cover behavior therapy, and 10 cover pharmacotherapy, all requiring pre-authorization. Regarding behavior therapy, Mr. Kyle was optimistic that more Medicaid programs would start to cover the services, given the recent decision by CMS to cover the services for Medicare. In Medicare, surgery is covered at Centers of Excellence for patients with BMI above 35 kg/m2 and comorbidities, behavior therapy is covered, and unfortunately, drugs are excluded from Part D. Mr. Kyle pointed out that the language in Part D specifically excludes weight-loss drugs, not obesity drugs.
  • He highlighted a number of remaining gaps in Medicaid and Medicare coverage that he hopes will be resolved in the upcoming years. First, medical nutrition therapy is not yet covered for obesity – Mr. Kyle noted that there is a precedent with diabetes for this. Regarding the Part D exclusion, he suggested that reframing medications as treatments for obesity and its consequences instead of weight loss could hopefully bring about coverage; if not, then advocates would have to take the lengthier route and try to get a legislative change. In addition, Mr. Kyle proposed that the CMS develop or identify a way to certify physicians who treat obesity, so that they could also get reimbursed for the time they spend providing intensive behavior therapy (since, under the current CMS guidelines, only primary care physicians are reimbursed for these services). Finally, Mr. Kyle hoped that the CMS would cover innovative procedures, drugs, and devices for the treatment of obesity.
  • Mr. Kyle discussed the potential impacts of the Affordable Care Act (ACA) on reimbursement. In terms of the good, the ACA: 1) provides for significant investment in prevention through the Prevention and Public Health Fund; 2) includes a menu labeling initiative; 3) requires coverage at no copay of services with USPSTF A or B grade recommendations, including obesity screening and intensive counseling for adults, and obesity screening and moderate-to-intensive counseling for children (both B grade recommendations); and 4) mandates that exclusion of those with pre-existing conditions is no longer permitted. In terms of the bad, the ACA: 1) allows employers to penalize workers up to 30% of healthcare costs for failure to meet specific outcomes measures including BMI, which introduces a backdoor way for employers to penalize those with health risks; 2) such penalties may increase up to 50% with the permission of the Secretary of Health and Human Services; and 3) with the Supreme Court limiting the Medicaid expansion, coverage of obesity services through such means for low-income individuals may not come to fruition. In terms of the unknown, it remains to be determined what essential benefits will be in each state. Mr. Kyle noted that in recent times, there has been a power shift from the federal government to state governments, as the December 2011 Health and Human Services guidance allows states to determine essential health benefits based on various benchmarks.



Louis Aronne, MD (Weill Cornell Medical College, New York, NY)

Looking at reimbursement from the perspective of physicians who treat obese patients, Dr. Louis Aronne offered his advice for obtaining reimbursement: 1) have confidence in what you are doing; 2) seek and ye shall find (look for comorbidities and rationalize treatment of weight in relation to those complications); 3) document, document, document; and 4) negotiate at the group level and at the patient level. In particular, he emphasized framing weight treatment as a way of managing other comorbidities, since employers and insurers are willing to provide reimbursement for such purposes.

  • Dr. Aronne outlined a number of principles for physicians to get reimbursement for treating obesity:
    • Have confidence in what you are doing. Dr. Aronne emphasized that those providing real healthcare with accepted treatments delivered by accepted standards should be able to get reimbursement (whereas, those who are operating pill mills, e.g., selling HCG out of their desk drawers shouldn’t and won’t get reimbursed).
    • Seek and ye shall find. Dr. Aronne noted that a strikingly high percentage of obese patients have at least one comorbid condition, and that insurers are willing to reimburse for the treatment of weight in relation to those complications. As such, physicians need to be sure to look for comorbidities – e.g., impaired fasting glucose, hyperlipidemia, sleep apnea.
    • Document, document, document. Dr. Aronne explained that in the current system, employers (not insurance companies) decide whether or not to cover obesity treatment. Given that employers cover the management of comorbid conditions, physicians need to not only (as stated above) look for comorbid conditions, but also document them so that they will get credit for treating them. Things to document include patient evaluation, risk/benefit assessment of the treatment, and the treatment plan.
    • Negotiate at the 20,000-foot (group) level and at the ground (patient) level. Dr. Aronne stated that a large multi-specialty medical group he joined managed to negotiate reimbursement with four of five major payers to reimburse medical therapy (i.e., meal replacements, dieticians, prescription medications) for patients who fit the indication for surgery for obesity treatment (BMI ≥35 kg/m2 with at least one comorbidity), so patients would have the option to choose between surgery and medical care. Meanwhile, if one thinks a patient has been wrongly denied coverage, one could appeal, write to a medical director, and even quote obesity literature on the impact of weight loss on a particular illness. Dr. Aronne reminded audience members that coverage is a business decision for the insurer; even if one thinks it’s obvious that treatment should be covered, the case often still needs to be made for insurers.


Symposium: Price Policy and Health


Julian Alston, PhD (University of California Davis, Davis, CA)

Dr. Julian Alston’s presentation explored whether USDA’s farm policy was contributing to the obesity epidemic. In contrast to our expectations, he suggested that the subsidies and taxes legislated by the Farm Bill probably have little impact on American’s diet and may even be fighting obesity. He noted that the $20 million per year spent on farm subsidies represents less than 3% of total US food consumption expenditure. Furthermore, Dr. Alston argued that the growth in high fructose corn syrup (HFCS) consumption has been caused more by restrictions on sugar imports (which increase the cost of sugar leading consumers to buy cheaper corn-based sweeteners such as HFCS) than corn subsidies. To that end, he reviewed an econometric study, which showed that eliminating corn subsidies would only cause a 9% to 10% decrease in corn production. Dr. Alston said that corn subsidies also do not impact the cost of other food products greatly because commodity costs represent only a small share of food prices (~20%). For example, if the corn subsidy were eliminated, the retail price of meat would only increase by ~0.3%. In another econometric model, Dr. Alston showed that without current grain and oilseed policies the average American adult would consume 1,200 less calories a year or about a third of a pound. Dr. Alston argued that the calorie deficit was negligible, but we think it is better than the status quo and could accumulate to result in decent weight improvements. Dr. Alston suggested that if all food policies (both subsidies and taxes, including the current ~100% tax on sugar) were eliminated then the average American adult would consume 3,000 more calories a year (which equates to about a pound a year). Finally, Dr. Alston reviewed studies on different potential food taxes and concluded that a fat tax (i.e., $5/kg fat in the food) would cause the average American adult to lose 5.61 lbs (vs. 5.5 lbs via a calorie tax and 5.50 lbs via a sugar tax); however, a calorie tax would be the most cost-effective, costing $0.90/lb. (vs. $1.34/lb for a fat tax and $0.93/lb for a sugar tax).

Questions and Answers

Q: Are you saying that subsidies had nothing to do with the dramatic rise in calorie consumption? What would be an explanation for this dramatic rise since the 1970s?

A: You are not the first person to ask me that, it is a good question. I am reluctant to give the answer – that is why I am stammering [laughter]. The answer is agricultural R&D productivity has been going up by about 2% a year. Farm subsidies are negligible in the supply demand balance compared to this. So even if we had this doubling of the population, the real cost of price of food has decreased because of research and development. I don’t like saying this answer though because I fear people will respond by cutting farm R&D, which is the opposite of what we need. Agricultural R&D is a good thing; it has kept people from starving.

Q: I don’t understand why you don’t like a sugar tax when you showed it reduced weight.

A: Its not that I don’t like a sugar tax, it is that I like a calorie tax better than a sugar tax because it is more efficient. Based on current research it addresses the problem of weight gain – consumption of too many calories – more directly than a sugar tax. I hedged that statement though because I am not sure if there is more of a cost to health from sugar than calories. We need to do research on that.



Frank Chaloupka, PhD (University of Illinois at Chicago, Chicago, IL)

Dr. Frank Chaloupka gave an excellent overview of taxes on tobacco, noting, “Taxes are a very effective tool for promoting public health.” He showed slide after slide linking reductions in the overall use of tobacco with higher taxes – we found these quite persuasive overall. Interestingly, about half of the effect of taxes on tobacco consumption is in reducing the overall number of people smoking – indeed, in months where tobacco taxes increased, the number of monthly calls into smoking quit lines spiked quite precipitously. Dr. Chaloupka next turned to a report from the WHO on the Best Practices in Tobacco Taxation to show general principles that might apply to obesity. In his view, excise taxes are the way to go (e.g., a penny per ounce soda tax): they uniquely apply to specific products, have a bigger impact on reducing consumption, lead to more stable streams of revenue, have a greater public health impact, and are easier to administer than ad valorem taxes (e.g., a 1% tax on soda). A penny per ounce tax on sugar sweetened beverages would generate approximately $15 billion in additional revenue per year, a healthy sum that Dr. Chaloupka argued could be put towards public health and mass media campaigns. And economically speaking, taxes could be used to overcome the significant negative externalities created by obesity (a cost to society of ~$147-210 billion per year). Dr. Chaloupka concluded by addressing the four major counterarguments against taxation. We found this to be one of the best policy presentations we’ve seen at this year’s TOS and hope to see such a tax emerge – as we noted in our September 15, 2012 Closer Look (which you can read here:, Richmond, CA City Councilman Jeff Ritterman is campaigning in California for a penny per ounce soda tax. If it passes on this November's ballot, the tax would be the first of its kind in the nation and the first to be approved by voters.

  • Dr. Chaloupka also addressed the four major counterarguments to tobacco and obesity-related taxation:
    • Taxes won’t generate the revenues that are expected. “False,” said Dr. Chaloupka. He noted that when you raise taxes, revenues go up – plain and simple. Since taxes only account for a fraction of the retail price, there is only a small decline in consumption of the taxed good. This decline is outweighed by the gain in tax revenues.
    • Taxes will negatively impact jobs. Dr. Chaloupka explained that tobacco taxes have actually been associated with a net increase in jobs. While there has been some reduction in jobs in tobacco dependent sectors, the money that used to be spent on tobacco goes to other sectors. The government also spends the tax revenues on new initiatives that create jobs.
    • There is potential for tax avoidance and tax evasion. Dr. Chaloupka acknowledged that there is some tax avoidance, but you still see the real public health benefits of tax increases. This is also less likely given the magnitude of taxes being considered and the costs of transport.
    • Taxes are regressive and have a disproportionate impact on the poor. Dr. Chaloupka acknowledged that this was the hardest counterargument to refute. Generally speaking, there is greater price sensitivity among the poor, meaning you see the largest reductions in tobacco use among the lowest income populations (vs. small relative reductions in higher income groups). As a result, low income populations end up paying a small amount of the increase in tax despite the health benefits that they gain.

Questions and Answers

Q: I have run into economists that believe there is no argument that justifies taxing sugar sweetened beverages. How would you respond to these individuals? Is there any role of behavioral economics?

A: Some economists say there is no rationale. I argue there is. The behaviors we look at are behaviors that begin at young ages. We don’t understand the long term consequences and the potential for addiction. People are uninformed or poorly informed when making decisions at an early age. Kids are the most responsive to these kinds of taxes and price changes. With tobacco, every increase in the tobacco tax leads to an increase in price. When it comes to beverage taxes, until someone adopts one, we won’t know how the beverage industry will respond. But when you start having broad based taxes, it will show up in terms of real increases in prices.


Symposium: Show Me the Money! The Economics of Obesity


Steven Gortmaker, PhD (Harvard School of Public Health, Boston, MA)

Dr. Steven Gortmaker emphasized the importance of conducting cost-effectiveness studies for obesity intervention programs – “it’s crazy,” he said, to do all these obesity effectiveness studies but not to measure intervention costs. Ironically, he believes the latter is actually easier to measure. While cost effectiveness data is limited, Dr. Gortmaker believes the empirical evidence for certain government policies is starting to grow. He pointed to a Lancet series evaluating obesity interventions in Australia – encouragingly, the data showed that an unhealthy food and beverage tax, reduction of advertising of junk food and beverages to children, and school programs to reduce sugar sweetened beverages, proved cost-saving in terms of net cost per DALY saved (Gortmaker et al., Lancet 2011). Dr. Gortmaker then presented a case study using technical analysis to model and evaluate a national excise tax of one cent per ounce on sugar-sweetened beverages (SSBs). Richmond, CA City Councilman Jeff Ritterman, for example, is currently campaigning for a similar tax. The hypothetical impact was impressive: 22.9 million DALYs averted, a national reduction in health care costs of $1,700 per dollar spent, and a median estimated $12.7 billion in tax revenue per year generated in Dr. Gortmaker’s model – wow! We hope the day actually comes when such a policy is passed and we can have the luxury of asking these questions. Dr. Gortmaker concluded by drawing comparisons to the anti-tobacco movement. What really made a difference in reducing smoking, he said, was implementing simple, cost-effective strategies like raising taxes, reducing marketing, and restricting consumption in public places.

  • Going forward, Dr. Gortmaker was hopeful that the ongoing CHOICES study, a cost- effectiveness modeling evaluation of 40 US childhood obesity interventions, would grant more insight as to what policies are cost-saving in the US. The research team will build on previous work funded by the Robert Wood Johnson Foundation to develop a cost- effectiveness model for obesity prevention in the US. For more information on the study, visit
  • Dr. Gortmaker believes that an SSB tax should be central to US policy and is a good place to direct efforts. We would have been very interested to hear Dr. Gortmaker’s perspective on the recent approval of Mayor Bloomberg’s soda ban, especially considering that prior to the prosed soda ban, Mayor Bloomberg had long tried to get a soda tax passed in New York City to no avail. We are curious to see whether this ban discourages SSB consumption overall, reduces overall caloric intake, or even has a ripple effect around the US.
  • Dr. Gortmaker highlighted a number of possible implementation considerations that could affect the success – some positively, some negatively – of a sugar- sweetened beverage excise tax including: 1) the potential of the intervention to shift public awareness and social norms; 2) that potential regressivity could be offset with earmarking; 3) an additional impact beyond BMI-mediated health effects; and 4) industry-aided backlash against public health intervention due to the current anti-tax sentiment in the US. The latter point seems particularly relevant, considering that when Mayor Bloomberg had tried to levy a soda tax, Pepsi threatened to leave New York and take many jobs with it.



William Dietz, MD, PhD (Former Director, Division of Nutrition, Physical Activity, and Obesity, Centers for Disease Control, Atlanta, GA)

Dr. William Dietz gave a broad review of the costs of obesity ($147 billion per year according to Finkelstein et al., Health Affairs 2009), focusing his presentation on a number of key themes. First, Dr. Dietz showed a number of studies suggesting that the costs of obesity (medical, absenteeism, and presenteeism) increase by degree of severity, and those with BMIs >35 kg/m2 claim a disproportionate share of costs (in one study, they claimed 40% of medical costs yet just 8% of prevalence). Second, obesity appears to be more debilitating for women than men – in some cases, women experienced triple the costs of men within the same BMI category. Third, ethnic differences exist with respect to years of life lost due to obesity, as African Americans lose fewer years relative to whites at the same BMI. Lastly, bias and stigmatization from medical providers may also contribute to the costs of obesity – this was disappointing to see from a patient perspective and we hope it is changing as obesity becomes more recognized as a medical condition.

  • Those with BMIs >35 kg/m2 represent 8% of the population, yet they account for 40% of costs attributable to overweight and obesity (Arterburn et al., Int J Obesity 2005). By contrast, those who are overweight (36% of the population) account for 31% of costs, while those with BMIs 30-35 kg/m2 (15%) generate 29% of the overweight and obesity-related costs.
  • While there are indications that obesity may be leveling off, Dr. Dietz noted that severe obesity trends have not plateaued in recent years, jumping from 6.2% in women in 1999 to 8.1% in 2010, and 3.1% in 1999 to 4.4% in 2010 for men (at May’s CDC Weight of the Nation conference, Dr. Justin Trogdon forecasted that severe obesity would grow 130% between 2010 and 2030; see pages 4-5 of our report at Given the disproportionate share of costs commanded by the severely obese, Dr. Dietz believes we need to be much more effective at addressing these higher grades of obesity. At the same time, since most individuals are in the 25-35 kg/m2 BMI category, he believes there’s an opportunity to address this group in the primary care setting.
  • Reducing costs related to severe obesity can take two forms: moving individuals to a lower BMI or reducing the number of associated complications and risk factors. Dr. Dietz showed unpublished data from General Motors (n=222,933) to look at medical costs by number of complications and BMI category. As expected, costs increase linearly with an increase in BMI. What is less discussed (though also expected) is that costs also increase linearly at a given BMI as the number of complications increase (i.e., the highest costs are found in those with six or more complications and a very high BMI). Dr. Dietz believes we have not adequately explored reducing the number of risk factors/complications whilst keeping people at the same BMI. He explained that physical activity is one good way to do this, especially for cardiovascular disease.
  • Data on medical costs, absenteeism, and presenteeism (attending work while sick) support the notion that obesity negatively affects women more than men. Dr. Dietz showed a study indicating that within the same BMI category, women typically have double or triple the medical costs seen in men (Finkelstein et al., Journal of Occupational & Environmental Medicine 2010). The data is similar for measures of absenteeism and self-reported presenteeism in the workplace (the latter was assessed by self-reported answers to the question, “On a scale from 0 to 10, during the past seven days, how much did health problems affect your productivity while working?”). Dr. Dietz cautioned that presenteeism is the least reliable measure because it is self-reported and very few studies have measured it directly. Nevertheless, data indicate that presenteeism may account for costs that are at least as much, if not more, than medical costs.
  • In terms of years of life lost, obesity appears to affect Caucasians much more than African Americans. Dr. Dietz highlighted an interesting study (Finkelstein et al., Obesity 2009) illustrating that for class I (30-34.9 kg/m2), class II (35-39.9 kg/m2), and class III obesity (>40 kg/m2), African Americans experience fewer years of life lost relative to Caucasians. Dr. Dietz confessed, “I have no idea why.” He called this another “ripe area of research.”
  • “The very patients who need our care the most are those who are most likely to avoid medical care.” Dr. Dietz noted that patients with obesity face weight bias, discrimination, and stigmatization, which negatively impact the healthcare they receive as well as their use of healthcare. Studies show that providers who interact with obese patients spend less time in appointments, have fewer discussions with patients, and intervene less often. Additionally, patients with obesity are less likely to obtain preventative services like breast cancer screening (Amy et al., International Journal of Obesity 2006). Dr. Dietz also highlighted that obese people are more likely to cancel appointments, are more likely to avoid healthcare, and are more likely to experience discrimination from providers.
    • We appreciated a review of Lee Kaplan et al.’s surprising poster, “Underrecognition and Underreporting of Obesity By Medical Residents (761-P).” Astonishingly, in 1,100 patients at Mass General Hospital, medical residents noted obesity as a problem 0% of time. This was despite the fact that 15-20% of thesepatients had BMIs over 40 kg/m2. We think this underrecognition speaks to the challenge of convincing society that obesity is indeed a medical condition. Dr. Dietz noted, “It’s essential that we focus on education of physicians.”

Questions and Answers

Q: White men live about five years longer than African Americans. If obesity affects whites more in terms of years of life, is it possible that they might come out even?

A: That’s a plausible hypothesis and a great question. I’ll provide you Eric Finkelstein’s email – he might be able to answer that.

Q: Do you have data on injuries? There was a paper out of Duke that found higher workers compensation in obese patients. Is that a hidden bucket of costs we’re not accounting for?

A: Yes it is. There are data, though not as strong, about work-related injury and disabilities. Overall, it is increased. It is another bucket of costs that is worth including.

Q: I’m interested in presenteeism and I’m thinking about the fact that it’s self-reported. Are there distinct differences in men and women?

A: Good point. None of the work that I have read suggests gender differences.

Comment: Blacks at the same level of BMI have a lower body fat percentage, less visceral fat, better triglycerides, and a lower risk of dyslipidemia. They are more susceptible to reverse causation. We need to account for that.

A: I also didn’t include QALYs (quality adjusted life years), which probably differ significantly due to the grade of obesity. There might be big differences that might well exceed years of life lost.



William Herman, MD (University of Michigan, Ann Arbor, MI)

Dr. William Herman summarized a systematic review of the cost-effectiveness of three approaches to treating obesity: lifestyle intervention, orlistat, and bariatric surgery. For the purposes of this presentation, he used a cutoff of $50,000 per QALY gained for cost-effectiveness. Out of 14 lifestyle intervention studies, four were cost saving (including one study using front of package labeling and a junk food tax), eight were cost effective, and only two were deemed not cost-effective. Orlistat (120 mg three times per day) was cost-effective in three out of six studies in healthy overweight or obese people and highly cost effective (<$20,000 per QALY gained) in obese patients with IGT or type 2 diabetes. However, Dr. Herman noted that orlistat has “modest efficacy,” “poor long-term adherence,” and the manufacturer supported five out of the seven included studies. Last was bariatric surgery (gastric banding or Roux-en-Y gastric bypass), which was cost-effective in six out of six studies in the non- diabetic population. In five studies in people with type 2 diabetes, the procedure was cost saving in three studies and cost-effective in two studies. Dr. Herman concluded with a summary of the litany of important assumptions inherent in all these analyses, which somewhat muddied our takeaways from his talk. We look forward to seeing cost-effectiveness data on Vivus’ Qsymia, especially given its ability to cause improvements in obesity-related comorbidities.

Questions and Answers

Q: When it comes to bariatric interventions, they're targeted to the more severely obese. John Kral and I recently worked on a project looking at this group, and we concluded that lifestyle intervention and pharmacology is not cost effective in severely obese populations. Can you comment?

A: In general, I would agree with respect to the available pharmacology treatments not being cost- effective. But with new treatments, I wouldn’t rule out cost-effectiveness.

Q: You mentioned several cost savings interventions in lifestyle and surgery. These are no brainers – they save money on the intervention. But those cost savings interventions are not implemented. What are the barriers?

A: Some of the more cost saving interventions are environmental interventions. There are a lot of objections to taxes and regulation. With respect to bariatric surgery in high risk individuals, it is very effective and cost saving and should be implemented more widely. But bariatric surgery is not going to fix this country’s obesity problem. We need a range of interventions.

Q: On Dr. Dietz’s slide, class three and four obesity accounted for 40% of costs. In bariatric surgery, could the estimates be lower than they should be? Might there be greater cost savings?

A: The bariatric surgery procedures are expensive, though the cost savings don't take too long. For obesity at that level, weight loss doesn’t take too long. The costs without treatment are so enormous that you can justify an expensive intervention.

Q: The list of limitations is quite formidable. With so many in these types of analyses, especially in the absence of long-term outcomes data, how much do they add?

A: There are limitations to all studies. In reality, we cannot wait for perfect data. We need to take action. The total financial cost of an intervention is a function of per capita cost and the size of the target population. Lifestyle interventions are inexpensive and effective. Bariatric surgery may have a smaller total cost burden to the country and be applied to a smaller group of people. Lifestyle interventions are a no brainer.


Symposium: Implementing Obesity Treatment in Primary Care: Is It Possible?


Thomas Wadden, PhD (University of Pennsylvania School of Medicine, Philadelphia, PA)

Dr. Thomas Wadden explored the role of primary care physicians (PCPs) in obesity care, describing two models in which the treatment of obesity is delegated to additional healthcare staff (e.g., medical assistants) or a remote lifestyle program (e.g. the telephone-based Healthways program). While Dr. Wadden recognized that it is possible for primary care physicians to provide lifestyle counseling themselves, he didn’t think that that meant they should necessarily do it, since it could be more cost effective to deliver such programs through auxiliary healthcare professionals, in the community, or through commercial programs. Dr. Wadden was thrilled that the CMS now covers intensive behavioral therapy for obese patients, describing the development as a positive step in the right direction. In terms of lifestyle interventions for weight management, Dr. Wadden thought the lifestyle management program in the DPP as a good model for what he wished could be done in primary care (however, he noted that primary care has its limitations and likely could not implement such a program).

  • Dr. Wadden outlined three models of how primary care physicians (PCPs) could be involved in the treatment and management of obesity: 1) PCPs could treat both obesity and its related comorbidities; 2) PCPs could just treat the comorbidities of obesity, and have additional staff (nurses, medical assistants, dieticians) treat obesity; or 3) PCPs could just treat the comorbidities of obesity, and have obesity treatment provided out of practice (e.g., referral to community or commercial programs, electronic interventions, or obesity specialists). Dr. Wadden emphasized the importance, in any case, for PCPs to keep relationships with whomever is treating obesity for their patients.
  • He noted that treating obesity in primary care is possible, citing a study in which PCPs provided tailored interventions for their patients. In the study (n=144), participants received either standard care or tailored intervention. PCPs provided six monthly individual 15-minute visits for the tailored intervention group, supported by dieticians, psychologists, and exercise specialists, while standard care participants did not receive any special treatment for obesity. Patients receiving the tailored intervention achieved an approximate 1.5 kg (3.3 lbs) weight loss, but then began to regain weight after one year (Martin et al., Obesity 2006). Dr. Wadden commented that this weight loss was not in clinically meaningful range and that prescribing medications in addition to lifestyle modification could increase weight loss. As an example, he noted that sibutramine (Abbott’s Meridia) has been documented to increase weight loss beyond brief lifestyle counseling alone (Wadden et al., NEJM 2005; [Editor’s note: Meridia is no longer available in the US due to cardiovascular safety concerns]).
  • In the two-year POWER-UP trial, enhanced lifestyle counseling (LC) provided significantly more weight loss than standard care. Enhanced LC consisted of quarterly PCP visits, brief monthly lifestyle counseling sessions with medical assistants, and the use of meal replacement or obesity medications (sibutramine or orlistat). At the end of two years, patients receiving standard care achieved 1.7 kg (3.7 lbs) weight loss, while those who received enhanced LC achieved 4.6 kg (10.1 lbs) weight loss from baseline. Dr. Wadden noted that there was some weight regain in the enhanced LC group in the second year, partially attributable to sibutramine being removed from the market, the FDA issuing an alert about potential liver problems with orlistat, and Slim Fast being unavailable for three to four months during the study. Dr. Wadden commented that the POWER-UP enhanced LC intervention would likely have been more effective if lifestyle intervention had been delivered on the schedule suggested by the CMS, given that there would have been more interaction between HCPs and patients in the early months of intervention. Nonetheless, he believed that the POWER-UP trial demonstrated that lifestyle intervention delivered by medical assistants could provide a more economic approach than having PCPs directly treat obesity.
  • In the two-year POWER Hopkins trial, remote lifestyle counseling and in-person support brought about similar reductions in weight for study participants. The telephone-only remote intervention (Healthways) consisted of weekly 20-minute calls in the first three months, followed by monthly 20-minute calls for the remainder of the two years. In-person support consisted of weekly group or individual visits with Johns Hopkins staff in the first three months, followed by three contacts per month in months four through six, and two contacts per month for the remainder of the study. Participants were recommended to reduce their calorie intake, follow the DASH diet, and self-monitor weight, calories, and exercise. In the remote intervention, patients were given scripted email feedback on the aforementioned. At baseline, patients (n=415) were on average 54 years of age, 103 kg (227 lbs) in weight, with a BMI of 37 kg/m2. At the end of two years, those who received the remote intervention achieved an approximate 4.8 kg (10.6 lbs) weight loss from baseline, while the in-person support group achieved an average 5.2 kg (11.4 lbs) weight loss. Dr. Wadden commented that this trial shows that primary care referrals to call-center programs can be effective in bringing about weight loss and cost less than in-person programs.

Questions and Answers

Q: Do you think the additive effects of sibutramine and lifestyle modification can be generalized to other medications?

A: I do, but of course we need to wait to see definitively. In most studies where a drug is added to lifestyle, the amount of weight loss increases. The only exception is when patients are on very low calorie diets and have lost 20% of their weight. I don’t think medications would add any weight loss beyond that.

Q: In your model for your study, how would you expect the lifestyle intervention to be billed?

A: The problem is that CMS is not going to be reimbursing medical assistants. I don’t know if we could have a medical assistant see a patient, then have a physicians say “hi” to the patient and then bill it under the provider’s name.

Q: Often, lifestyle counseling is only as good as the person doing the counseling. Did the trials use any validated tools or protocols for lifestyle intervention?

A: We chose medical assistants based on personality – outgoing, people pleasers who like working with people. Initially, we trained them on the DPP protocol for six to eight hours on how to deliver the intervention. We certified them using the same method as in the Look AHEAD study to see if they could deliver the intervention appropriately. Your point is well taken that we have to make sure that people are trained to delivery lifestyle intervention. My concern about CMS reimbursement is what lifestyle interventions primary care physicians are supposed to use.



Kurt Stange, MD, PhD (Case Western Reserve University, Cleveland, OH)

Dr. Kurt Stange described the challenges of obesity treatment and prevention in the primary care setting given the competing demands that primary care providers (PCPs) constantly face, suggesting that perhaps they should take on the public health perspective and focus on incremental improvements over time on a population level, instead of trying to remedy each and every obese individual’s weight issues. He explained that PCPs have too limited time at each patient visit to focus in depth on any particular disease or condition (e.g., obesity), especially since patients often present with multiple issues and concerns. For example, in one study, patients with diabetes seen at a clinic presented with an average 25 problems (a range of 13-32 problems), including biomedical, behavioral, social, and environmental health issues (Bolen et al., under review). Dr. Stange noted that this is both the strength and the weakness of primary care – the ability to address a multitude of issues in a short period of time, and the inability to address any specific issue in depth. In addition, regular disruptions in the continuity of care (he stated that in one study, 24% of patients with managed care insurance were forced to change family physicians over the course of two years because of changes in insurance coverage) have caused both physicians and patients to invest less in doctor-patient relationships, which ultimately results in a lower quality of overall care. In terms of emerging opportunities, Dr. Stange suggested that perhaps PCPs should focus on teachable moments (when they recognize salient patient concerns that would motivate patients to change) as a way to encourage improved obesity management. He emphasized that the relationships between physicians, patients, their families, communities, and the healthcare system are an important part of the strategy to create a tipping point toward healthy eating and activity.

Questions and Answers

Q: In our frenetic society where many patients come in talking on their phones, and we only have a few minutes to try to connect with them in the right way to try to make a difference, how do we do so? How do we get something meaningful for this patient, at this moment, at this time?

A: What we can do is to be a bit mindful, and recognize that we don’t have to do everything at once. What seed can you plant that you can water next time?

Q: It seems that a lot of primary care physicians have to rely on self-reports in the 10 minutes they see each patient. Do you see a role for technology to help patients track their diet and physical activity? Could primary care physicians use that data to enhance their discussions?

A: Yes. The hard part is, how do we get that integrated with electronic medical records? It could be incredibly powerful if that kind of thing gets integrated with electronic medical records. It would be a big whoop.


Symposium: Developing Recommendations on Best Practices in Obesity Prevention


Michael Jensen, MD (Mayo Clinic College of Medicine, Rochester, MN)

During one of the last presentations of the conference, Dr. Michael Jensen, co-chair of the National Heart, Lung, and Blood Institute’s expert panel assembled for the update of the “Clinical Guidelines on the Identification and Treatment of Overweight and Obesity in Adults,” described the strategy behind the updated guidelines and what to expect. The current set of federal guidelines (which defines being overweight as BMI >25 kg/m2 and obese as BMI >30 kg/m2), astonishingly, is nearly fifteen years old (it can be found at Unfortunately, Dr. Jensen said the new guidelines would not cover pharmacotherapy for obesity, given that when they decided to stop accepting abstracts (around 2009) no new pharmacotherapies existed. Instead, Dr. Jensen said the new guidelines will focus on answering five critical questions for primary care physicians: 1) Who should work to lose weight? 2) What health improvements are associated with different degrees of weight loss? 3) What diets can be recommended for weight loss? 4) What results can be expected from good lifestyle intervention programs? And 5) What is the safety and efficacy of surgical procedures for weight loss? When asked during Q&A why the guidelines could not be updated to account for recent progress in pharmacotherapies (i.e., the recent approval of Arena/Eisai’s Belviq (lorcaserin) and Vivus’ Qsymia (phentermine/topiramate ER), Dr. Jensen said the panel decided that one section could not be updated without the entire report being updated – which would have been unfeasible. Considering the substantial impact that recently approved obesity drugs could have on treating obesity, we find the inflexibility on this point to be most surprising and disappointing, particularly given how long we have waited for this update. As we understand it, the decision may have more to do with funds available – there was some controversy on how the piece was put together and we don’t sense uniform agreement on the ultimate value. The update was originally supposed to be published in 2012, but during the Vivus- sponsored CME event, Dr. Donna Ryan (Pennington Biomedical Research Center, Baton Rouge, LA) said it would not be published until 2013.

Questions and Answers

Q: Why could you not update part of the guidelines during this writing process to account for recent progress?

A: We didn’t feel like it would be right to select certain sections to update without updating all of them. We wanted the whole report to be based on the same research timeframe. If we updated the entire report every time a new paper came out that we wanted to add (and this did happen on several occasions) the guidelines would be continuously updated, rather than discontinuously. Maybe they should be continuously updated but that would take massive man-hours and is not up to a single panel.


Symposium: Front of the Package Labeling: Helpful or Confusing


David Ludwig, MD, PhD (Children’s Hospital Boston, MA)

Dr. David Ludwig gave a striking presentation on the evils of processed foods, noting that they are high in calories, low in micronutrients and fiber, contain unhealthful fats, and “bear primary responsibility for diet related chronic diseases.” Regarding the subject of his talk’s title, Dr. Ludwig asserted that there are indeed “good” foods (those that protect against diet-related chronic disease, like fruits and vegetables) and “bad” foods (those that are ultraprocessed and undermine health and promote diet- related chronic disease). In his view, the challenge is not to completely eliminate food processing, but to move consumers from ultraprocessed products to more traditionally processed, whole foods (e.g., those in a Mediterranean style diet). Dr. Ludwig strongly argued against front of package labeling, citing four fundamental flaws in the strategy (see bullet below). Rather, he believes we should 1) restrict front of package labeling; 2) strengthen the nutritional facts panel to provide a more comprehensive view of food quality; 3) encourage consumers to read ingredient lists; and 4) focus educational efforts on the promotion of whole and minimally processed foods. We enjoyed his passionate talk and couldn’t agree more with driving focus and efforts towards encouraging less processed food consumption. The question on our mind is how to overcome the very powerful Big Food industry lobby. Still, we think momentum is building right now, especially with the commitments from the First Lady, the Partnership for a Healthier America, and the recent bad of sodas larger than 16 ounces in NYC.

  • Dr. Ludwig presented four fundamental flaws in front of package labeling:
    • 1. Health claims based on individual dietary factors are misleading. Front of packaging labels have a selective focus, ignoring potentially harmful aspects of a food.
    • 2. Front of packaging labels restricted to nutrient content can be deceptive by presenting information out of context. For instance, an eight-ounce sugary drink has less calories than a few ounces of nuts, but no one would argue that the former is healthier.
    • 3. “Healthier” processed foods are not necessarily healthy. Manufacturers can manipulate snack food ingredients by replacing fat or sugar with refined starch.
    • 4. Front of packaging labels and claims confer an aura of healthfulness that tends to encourage consumption.
  • Dr. Ludwig shared three statistics we found quite compelling:
    •  1. An isotopic analysis of Dr. Sanjay Gupta’s hair found that 69% of his carbon could be traced to corn – this speaks to the fact that much of our food supply to corn based.
    •  2. Fast food has increased from 2% to 20% of the calories in children’s diets since the 1970s.
    • 3. In a preclinical rat study, 94% demonstrated a preference of saccharine over IV cocaine. Saccharine is the artificial sweetener found in Sweet ’N Low.

Questions and Answers

Q: Thanks for a provocative and yin-yang talk. You talked a lot about processed food, but obesity in children is almost certainly done with healthful foods as well. I think we need to be careful about our language.

A: I am familiar with a number of studies that consistently show very low quality diets among kids and low serum levels of key nutrients. I have to disagree that the overall quality of nutrition in children’s diets is high. It is extraordinarily low. And perhaps low in ways we have not observed. No one is measuring antioxidant and anti-inflammatory properties. When food is so processed, the calories are separated from the protective factors.

Q: I work for Kraft foods. I would agree with you about trans fats – by and large, the food industry has done a good job in western nations. You also mentioned the glycemic index, rapidly absorbed carbohydrates. I’m wondering about polished white rice. It accounts for 35-80% of calories in Asia. That’s high glycemic index and does not have many nutrients. How do you fit that into your view of processed foods?

A: White rice is undermining the nutritional quality of that region of the world. Frank Hu is working on a study looking at substituting brown rice for white rice. For people working in the fields for 12 hours per day, they have limited access to calories and white rice is giving you the calories. They’re not at risk for obesity. But when you take the white rice diet and the Chinese peasant moves to the city, the Indian moves to Delhi, then obesity rates skyrocket along with the risk for diabetes. We’re seeing that happen throughout Asia. This is one of the most important chronic nutritional issues in the developing world.

Q: It seems as though as a species, we’ve seen an acceleration of this sensory input – taste, smell, visual stimuli, multitasking, video games. Through this process of intensity and stimulation, we are becoming sense junkies.

A: I will leave it to appreciation for your point. It’s a topic for all of us to discuss in another session.


Opening Session: President's Address, George Bray Award, Atkinson-Stern Award, Journal Awards and Grant Presentations


Patrick O’Neil, PhD (Medical University of South Carolina, Charleston, SC)

It’s been a busy and productive year at TOS,” said Dr. Patrick O’Neil as he began his President’s Address. He highlighted the small but significant improvements TOS has made both in putting on the conference and in advancing the organization’s mission and advocacy message. This year there were 891 abstracts submitted to the conference, 61 exhibitors (vs. 52 in 2011), and five corporate sponsored symposia (vs. two in 2011). As for TOS’ mission, he spoke to the Treat Obesity Seriously Campaign, TOS’ efforts to influence policy makers, and the vision to become a natural resource of information about obesity. Dr. Jennifer Lovejoy (University of Washington, Seattle, WA) and Dr. Eric Ravussin (Pennington Biomedical Research Center, Baton Rouge, LA) each took turns on the podium to present a plethora of grants and awards for new pilot studies, excellent articles in Obesity (TOS’ research journal), and contributions to obesity prevention and intervention. Notably, Mr. Theodore Kyle (ConscienHealth, Pittsburgh, PA) received the Atkinson-Stern Award for distinguished public service in improving the lives of those affected by obesity, and Dr. James Hill (University of Colorado Denver, Denver, CO) was awarded the George Bray Founders Award for contributions to the understanding and treatment of obesity.



James Hill, PhD (University of Colorado Denver, Denver, CO)

As the recipient of the George Bray Founders award, Dr. James Hill gave an impassioned talk on obesity policy and the role of science in influencing policy. He began by reflecting on his own career. Since beginning to study obesity in 1981, Dr. Hill has come to learn that is the environment, not the biology, which is broken. He emphasized that behavior is influenced by the environment, and the environment is in turn influenced by policy. Although obesity policy is complex and it is sometimes difficult to bring the right kind of science to the problem, he said, it doesn’t mean we shouldn’t try. Part of this difficulty is achieving the appropriate balance between acting now and waiting for the necessary studies to prevent unintended (and potentially negative) consequences of policy intervention. Dr. Hill drew an interesting parallel between energy balance and policy, calling both “complex adaptive systems. For example, he reminded the audience of the common compensatory behavior whereby individuals who increase their physical activity through an exercise program will increase their energy intake. Policy often acts the same way he explained, by means of “policy resistance” – the tendency of the intervention to be defeated by the systems’ response to the intervention itself. Just as a scientist cannot study the affect of an exercise program in isolation (i.e., not considering food intake), policy makers cannot look at interventions in isolation. In one fascinating example, Dr. Hill summarized a six- month field study conducted by Dr. Brian Wansink (Cornell University, Ithaca, NY) and colleagues in a small town. Half of the households faced a 10% tax on soft drinks and half did not. The 10% tax resulted in a short-term (one-month) decrease in soft drink purchases, but there was no decrease in purchases over a three-month or six-month period. Moreover, in beer-purchasing households, the tax led to increased purchases of beer (Wansink et al., 2012; “From Coke to Coors”). Dr. Hill concluded with a call to action. He urged scientists and audience members to play a more active role in crafting obesity policy and determining the balance between acting now and preventing unintended consequences: “We need to use policy to reduce obesity and everyone needs to be engaged with how that happens.”


Keynote Debate: The Role of Sugar-Sweetened Beverages in the Obesity Epidemic: Plague Rat or Red Herring?

This debate concerns the following proposition: “Resolved: There is sufficient scientific evidence that decreasing sugar sweetened beverage consumption will reduce the prevalence of obesity and obesity- related diseases.”


Frank Hu, MD, PhD (Harvard School of Public Health, Cambridge, MA)

To an attentive and packed audience, epidemiologist Dr. Frank Hu focused his affirmative arguments on large studies associating sugar sweetened beverage consumption with weight gain. His presentation’s pace was lightning fast compared to Dr. David Allison (who argued the negative, see below), though it was also a bit less structured. To begin, Dr. Hu reviewed the results of Schulze et al. published in JAMA in 2004: in 91,249 women in the Nurses Health Study II, higher consumption of sugar-sweetened beverages was associated with a greater magnitude of weight gain and an 83% increased risk for development of type 2 diabetes in women compared to those with less consumption. The diabetes data – which we thought was his most compelling point – also surfaced in a 2010 meta- analysis (Malik et al., Diabetes Care) that associated sugar sweetened beverages with a 26% increased risk of type 2 diabetes for one or two servings per day (n=310,819). Using the Bradford and Hill criteria for causality, Dr. Hu demonstrated that there is indeed strong evidence that sugar sweetened beverage consumption is causally linked to risk of diabetes (Malik and Hu, Curr Diab Rep 2012). He also covered a meta-analysis of RCTs on sugar sweetened beverages – in six trials, adding these beverages to participants’ diets showed dose-dependent increases in weight. Last, Dr. Hu provided a more experimental view, noting that sugar sweetened beverages are associated with insulin resistance, beta cell dysfunction (Malik et al., Circulation 2010), higher triglycerides, CRP, IL-6, TNF-alpha, and leptin, plus lower LDL and adiponectin. Dr. Hu seemed to be on the defensive from the start of his talk, and indeed, we think he had the harder side of this debate.



David Allison, PhD (University of Alabama, Birmingham, AL)

Speaking charismatically and confidently, Dr. David Allison argued that while it is reasonable to hypothesize that reducing sugar-sweetened beverage consumption will reduce obesity, the current body of scientific evidence does not support this premise. He began by reminding the audience that the proposition asks whether there is sufficient evidence to show an effect of reducing sugary beverage consumption, and does not consider whether enough evidence exists to support public health action. He then evaluated the current data, stating that while studies have found an ecological correlation, such correlations are among the weakest forms of evidence. Observational studies have shown a correlation between sugary drinks and obesity, but have not proven causation. Furthermore, while short-term feeding studies indicate that people compensate less for the liquid calories in sugary beverages, data show that compensation still occurs. Dr. Allison’s main argument centered on a meta-analysis (of which he is an author) that aggregated data from the six randomized controlled trials that have publically released data on this topic (Mattes et al., Obesity Reviews 2011). He noted that when looking only at the primary analyses on continuous endpoints (BMI, weight, and body fat), the meta-analysis did not show a statistically or clinical meaningful effect of reducing sugary beverage consumption. Dr. Allison also emphasized that while each trial showed a significant effect through the analysis of a secondary endpoint, the lack of consistency among such analyses across the six trials makes it difficult to draw any meaningful conclusions. Dr. Allison then postulated why scientists believe that reducing sugary beverage consumption would reduce obesity despite the lack of convincing evidence. He cited the distortion of scientific information by the media, the wave of emotion-raising advocacy against sugary beverages, and the fact that observational studies that fail to find an effect are less likely to be published. Dr. Allison ended his presentation by noting that several organizations share his viewpoint (including the AHA and ADA) and that he would change his opinion upon the publication of randomized, controlled trials that are well designed, executed, and analyzed, and whose results are substantial enough to outweigh the existing data.



Frank Hu, MD, PhD (Harvard School of Public Health, Cambridge, MA)

Dr. Hu focused his rebuttal on the difference between efficacy trials and effectiveness trials. Efficacy trials, he explained, are highly controlled. Such studies would compare regular soda to diet soda, and you would expect those on regular soda to gain weight. These trials have strong results and establish a causal link between soda consumption and weight gain. Conversely, effectiveness trials test different intervention methods (for example, classroom education to teach kids to stop drinking soda). The results from these trial are mixed, suggesting that these methods may not be very effective in fostering behavior change. In Dr. Hu’s view, the inconsistent results from effectiveness trials do not provide evidence against causality. In the case of tobacco as an example, studies show that tobacco is related to lung cancer. However, there is no randomized controlled trial to demonstrate that reducing tobacco smoking will cause lung cancer risk down. But based on the causal evidence linking tobacco smoking and lung cancer, if everything else remains equal, a decline in cigarette smoking will decrease lung cancer. Since the relationship between sugar sweetened beverages and weight is causal, reducing consumption of sugar sweetened beverages will reduce obesity.



David Allison, PhD (University of Alabama, Birmingham, AL)

In his rebuttal, Dr. Allison made two points. First, he agreed with Dr. Hu on the existence of efficacy vs. effectiveness data, stating that the majority of previous studies have been effectiveness studies. These studies haven’t shown that reducing sugary beverage has no effect, but have failed to prove that such a measure would actually have an impact on obesity. Second, he disagreed with Mr. Hu’s use of the smoking analogy, noting that the relative risk of lung cancer doubles with smoking, a far stronger relationship than the 26% increased risk of diabetes associated with consumption of sugary beverages. Dr. Allison reminded the audience that with such a small percentage, the likelihood of confounding effects is very high.


Frank Hu, MD, PhD (Harvard School of Public Health, Cambridge, MA) and David Allison, PhD (University of Alabama, Birmingham, AL)

Q: Despite weak evidence in support, can we move forward, or must we wait?

Dr. Allison: As a scientist, I can offer nothing. I can offer my own opinions as a citizen and an individual. Moving forward, we have to assess the likely benefit of an action against one’s moral beliefs. If the action would markedly impair freedom, some would say no. On the other hand, if the action had no cost or little cost and didn’t impair freedom, we might say yes. I have weak evidence to say that buying my wife a gift on Valentines’ day is beneficial, but I do it every year anyways.

Dr. Hu: I think the evidence is very strong and compelling. We have to be vigilant about a couple of things. One, there is a tendency to ask for perfect proof in terms of the relationship between foods and beverages and outcomes. Is this realistic? Is this possible? Those RCTs David mentioned – the best scientists designed them, they had the best resources, and still the control group decreased consumption. The intervention group’s change in behavior was not as big as expected. For real life interventions, it’s wishful thinking we will have a perfect design. I don’t think that’s realistic to demand absolute proof from science.

Q: Because sugar-sweetened beverages may increase weight gain, can it be assumed that reducing sugar-sweetened beverage consumption can produce weight loss and decrease chronic disease?

Dr. Hu: I’m not talking about weight loss – weight loss is a different issue than weight gain. Obesity doesn’t occur overnight. The weight gain is very gradual and insidious. The average amount of weight gain for US adults is one to one and a half pounds per year. I think prevention of weight gain in both children and adults is more important than weight loss in terms of changing in the overall obesity landscape. I think that at this point, the evidence is clear that sugar-sweetened beverages will increase weight gain and contribute to obesity. So if we reduce the consumption of such beverages, the weight gain will be reduced and the risk for obesity will be reduced. Regarding weight loss, I think that for people who are overweight and obese who include sugar-sweetened beverages as a large part of their diet, it could be a good idea to reduce sugary beverage consumption to decrease caloric intake. Longer studies are needed to clarify their impact.

Dr. Allison: I agree with Dr. Hu. On purely logical grounds, we can’t assume that this statement is true. This goes back to the issue of action. Some actions are imminently reasonable knowing what we currently know. When a patient says, “I’m having trouble with my weight and I drink six liters of sugar-sweetened beverages a day,” what clinician wouldn’t say, “maybe you should stop that”? Of course we’ll take those actions, and of course that’s plausible. Of course we could say to people that stopping sugar-sweetened beverages will help. Do I have the evidence to show it will help? Not today. Might it help? Sure, give it a try.

Q: Does Michael Phelps need to worry about his intake of sugar-sweetened beverages?

Dr. Allison: I have not met Michael so I cannot comment.

Dr. Hu: That doesn’t apply to the general population.

Q: Is there good evidence showing an association between artificially-sweetened drinks and weight gain?

Dr. Hu: There are a lot of studies on artificial sweeteners and body weight, but the results have been very confusing. Some show a positive association. However, there’s a major methodology issue in studying this relationship because people switch from regular soda to diet soda. Many people are already overweight and have metabolic disease, and are using artificially sweetened drinks as a dieting tool. In our studies, we did an age adjustment analysis and there’s a strong positive association between diet soda and risk of type 2 diabetes. However, after adjusting for baseline weight and metabolic disease at baseline, this association went away. This finding suggests that the positive association may be due to the reverse causation problem. It’s a difficult problem to tackle. In clinical trials that compare diet soda to regular soda, diet soda leads to much better weight status, in terms of weight gain and weight loss.

Q: What about potential confounders like diet, BPA packaging, etc. How do we isolate the effect?

Dr. Allison: It’s very difficult. We say, ‘We adjusted for all the variables as best we could, and the association held up.’ But ‘as best we could’ is not good enough. Our ability to control is limited by our awareness of it, being able to modify it, and measuring it. Unfortunately, we don’t measure things very well in nutritional epidemiology. I think these are plausible hypotheses. In some cases, we can do randomized trials. But in others, such as with BPA, we will have to tease it out with toxicology and epidemiology.

Dr. Hu: We just finished an analysis of BPA in control subjects and in people with diabetes. In the next few weeks I can tell you whether the association between sugar sweetened beverages and diabetes is independent of BPA or not.

Q: Why is the policy debate only on regulating sugar-sweetened beverages and not candies, sugar, or fruit syrups? Should we be debating the bigger issue?

Dr. Hu: That’s certainly a legitimate question. The main issue is that policy has to be evidence based. If you list all the evidence for the different categories of food you just mentioned, the evidence is strongest and most abundant for sugar-sweetened beverages. And that’s a simple fact. That’s the reason: evidence- based policies have to look at how much evidence is out there. Of course, another consideration is that sugar-sweetened beverages have no nutritional value. They are just empty sugars and calories with no redeeming nutritional qualities. In that sense, sugar-sweetened beverages can be considered low hanging fruit. Of course it’s not a fruit [laughter]. That’s a reason why it’s rightly justifiable to single out sugar- sweetened beverages as a policy intervention. I think what Mayor Bloomberg is doing in NYC is fully justified.

Dr. Allison: I think that it’s not clear what the redeeming nutritional value of donuts is either. I think many of these things [that Dr. Hu said about sugar-sweetened beverages] could be said about other foods too. I think the reason for singling out sugar-sweetened beverages is not scientific, it’s social. Public health advocates know that social change happens when you have a villain. What’s a better villain than sugar- sweetened beverages? You have corporate goliaths, you don’t need sugar-sweetened beverages, and you can build a scientific story. There’s also the mere-exposure effect. Familiarity is hard to differentiate from truth. So if you repeat things enough times, people start to believe them. I think we’ve had a ground swell of feverish response around sugar-sweetened beverages. There’s a villain and that’s why there is a single- minded association with sugar-sweetened beverages.

Dr. Hu: I completely disagree with that. I’m not a policy person. I’m not obsessed with sugar-sweetened beverages. My job is to provide the best evidence so that the policy makers can make their decisions. My job is to build the evidence base, and clearly the evidence base is strongest for sugar-sweetened beverages. I don’t think sugar-sweetened beverages are equal with other foods. I think the main thing is that it’s unique because it accounts for 8-10% of total calories in the US diet. It’s the largest source of added sugar in the US diet, it’s extremely cheap, and the science is clearly much stronger than anything else we’ve looked at in terms of dietary factors and health outcomes.


5. Other Notable Talks

Oral Abstract Session: Exercise, Metabolic Disease Risk, and Obesity


Stephanie Mayer, MD (Duke University School of Medicine, Durham, NC)

Dr. Stephanie Mayer presented a type 2 diabetes subgroup analysis of a 2010 study comparing a 48- week low carbohydrate diet (<20 g per day) to a 48-week low fat diet plus 120 mg orlistat three times daily. In 46 people with type 2 diabetes, both diets led to comparable reductions in BMI (-2.4 kg/m2 in the low carb group and -2.7 kg/m2 in the low fat diet plus orlistat group). However, the low carb diet group experienced a 0.7% reduction in A1c (baseline: 7.6%), compared to a rise of 0.1% in the orlistat group (baseline: 7.6%; p =0.045). We thought this was pretty notable given the low baseline. Overall, Dr. Mayer spent most of her presentation discussing a novel way to evaluate changes in anti-glycemic medication intensity associated with weight loss: the medication effect score. The measure takes a patient’s current dose level for a particular medication (e.g., 30 mg of pioglitazone), divides it by the maximal dose level (e.g., 45 mg for pioglitazone), and multiplies that by the expected decline in A1c (for pioglitazone, a decline of 0.95%). This analysis is then repeated after the intervention to determine if the anti-glycemic medication intensity has declined. The measure can also be used for insulin using baseline weight (her group uses one unit of insulin per kg as the threshold for insulin resistance). Dr. Mayer believes that the Medication Effect Score will allow comparison across medication regimens, which should be helpful in evaluating the anti-glycemic benefits of weight loss in clinical trials.

Questions and Answers

Q: I like this Medication Effect Score idea. It’s really helpful in clinical settings and for pharma studies. Have you discussed this with the FDA?

A: We’re hoping to present this as a novel mechanism for comparing between patients and groups and the impact different agents have. We hope to publish this data and if we are approached by FDA we certainly discuss it with them. (Editor’s note – we hope someone can proactively approach FDA on their behalf.)

Q: Have you thought about this for other therapeutic areas? Blood pressure lowering?

A: Yes, we will bring it forward shortly for blood pressure.

Q: Responses are not proportional. Taking the percentage of the maximal dose may not be ideal.

A: The PK is individual to each agent. We substantially drew on the consensus algorithm that we based a lot of our adjustment factors on. It’s based on the average at the maximum dose.

Comment: I love that as a very easy way to do it.


Oral Abstract Session: Intervention


Tricia Leahey, PhD (Brown University, Providence, RI)

Dr. Tricia Leahey presented the results of a 12-week study showing that an internet behavioral weight loss program (IBWL), and an IBWL in combination with optional group sessions were able to enhance the weight-loss efficacy of Shape Up Rhode Island (SURI), a 12-week community-based weight-loss program. At the end of the study, participants in the SURI, SURI + IBWL, and SURI + IBWL + optional group sessions group achieved statistically significantly different weight losses of 0.9 kg (2.0 lbs), 3.1 kg lbs), and 4.9 kg (10.8 lbs) from baseline, respectively. Ms. Leahey suggested that IBWL and optionalgroup sessions likely enhanced weight loss by providing additional behavior change skills and increasing social support. The incremental cost effectiveness ratio (ICER) of SURI + IBWL and SURI + IBWL + optional group sessions was $42/kg and $110/kg, respectively. While initial weight loss was maintained at six months, there was no significance between groups at the 12-month mark – this reinforces to us that lifestyle alone is really hard to make work - there are legions of programs that show maintenance of weight loss is simply very challenging. Ms. Leahey noted that treatment adherence was associated with weight outcomes – but either way, since there was no difference at six months, we’re not sure this matters.

  • Shape Up Rhode Island (SURI) is a 12-week online program open to all adult residents in Rhode Island. Participants join teams of five to 11 members; these teams compete with each other on physical activity. SURI participants receive: 1) access to a reporting website that displays personal and team progress; 2) self-monitoring logs to record progress; 3) pedometers; 4) a weekly newsletter; 5) community workshops on healthy eating and physical activity; and 6) prizes. Ms. Leahey commented that programs with wide reach such as SURI provide an opportunity to disseminate behavioral weight-loss strategies to help address the obesity epidemic; however, they tend to yield only modest weight loss on the order of 1-3 kg (2.2-6.6 lbs). As such, she noted that the challenge is to enhance such programs without substantially increasing the cost.
  • This 12-week study aimed to test whether adding an internet-based behavioral weight-loss (IBWL) program or IBWL plus optional group sessions would enhance the weight loss outcomes seen with SURI. In the study, participants were randomly assigned to SURI (n=46), SURI + IBWL (n=90), or SURI + IBWL + optional group sessions (n=94). While the IBWL was primarily delivered via the internet, it included a one-time in-person kick-off group session: 1) to set weight, diet, and activity goals; 2) to teach self-monitoring skills; and 3) to orient participants to the IBWL website. The IBWL consisted of 12 multimedia lessons based on the DPP, self-monitoring, and weekly personalized automated feedback. The 12 weekly optional group sessions included a private weigh-in plus one hour group sessions led by a dietitian, with materials supplementing the internet program. The vast majority of participants were Caucasian females, with average age of 46.9 years and BMI of 34.3 kg/m2. After 12 weeks, those in the SURI, SURI + IBWL, and SURI + IBWL + optional group sessions groups achieved respective weight losses of 0.9 kg (2.0 lbs), 3.1 kg (6.8 lbs), and 4.9 kg (10.8 lbs) from baseline (differences between each group were statistically significant). Compared to SURI alone (7%), a higher proportion of those in the SURI + IBWL (41%) and SURI + IBWL + optional group sessions (54%) groups achieved 5% weight loss after 12 weeks. Ms. Leahey noted that treatment adherence was associated with weight outcomes. The incremental cost effectiveness ratio (ICER) of SURI + IBWL and SURI + IBWL + optional group sessions was $42/kg and $110/kg, respectively. While initial weight loss was maintained at six months, there was no significant difference between groups at the 12-month mark.



Chandra Jackson, PhD (Harvard University, Boston, MA)

Dr. Chandra Jackson presented a study demonstrating that the association between BMI and mortality appears weaker in black compared to white adults. In providing a brief background, she noted that current Body Mass Index recommendations are based largely on data from Caucasian American or European populations, though studies show that the prevalence of obesity differs substantially by race and sex. Excess body weight may confer a greater mortality risk in whites compared to blacks because data show that at a given BMI, blacks tend to have more lean body mass, lower body fat percentages, smaller waist circumferences, more superficial fat, less visceral fat, and a decreased risk of dyslipidemia compared to their white counterparts. Dr. Jackson’s study pooled data from the National Health Interview Survey (1997-2002, with mortality follow-up to 2006), which surveyed a large (n=30,000/year), nationally representative sample of the US population. The study included self- reported data from roughly 33,000 never-smokers, and during the median nine-year follow-up period, 1,205 deaths occurred within the study cohort. Stratification by BMI quintile revealed that in black women, the 3rd quintile (25.0-27.4 kg/m2) for BMI was associated with the lowest risk of death from all- causes. In contrast, data from white women revealed that the typically observed 2nd quintile (23.0-24.9 kg/m2) for BMI was associated with the lowest mortality risk. In black men, the 2nd and 3rd quintiles (23.0-27.4 kg/m2) conferred a similarly low absolute risk for mortality while in white men, the 2nd quintile (23.0-24.9 kg/m2) conferred the lowest death rate. In ending her presentation, Dr. Jackson concluded that while obesity remains a strong risk factor for both racial groups, the data support a careful reconsideration of current BMI recommendations in hopes of more accurately reflecting mortality risk for blacks. Dr. Jackson also noted that additional studies are needed to address the important methodological limitations of this study.

Questions and Answers

Q: My understanding of pediatric literature is that yes, there are body composition differences, but those amount to one BMI unit in kids with a BMI below the 95th percentile. Once you get above the 95th percentile, the difference isn’t a real “difference” – excess weight is fat. Is the same true in adults? If it’s a difference of one BMI unit, it may not account for the mortality difference.

A: I think it’s a great point. I think we rely on the literature that suggests that body composition differs. I do think that black women hold more weight in the gluteal femoral area, which is associated with a lower risk. I think your point is very well taken. I can’t answer that question because I don’t know. I don’t think the literature has answered that question. Though we know that after a BMI of 35 kg/m2, waist circumference is less predictive of mortality.

Q: Do these differences in mortality between African Americans and Caucasians take into account the overall difference in mortality among African Americans and Caucasians?

A: They don’t, because of the nature of the data. This is the exact kind of data that past studies have relied on – cross sectional data with one BMI data point in the long term. It’s particularly susceptible to reverse causation. For example, blacks are more likely to have undiagnosed conditions. So it contributes to the idea that higher BMI is more likely to contribute to mortality because they’re ill and their BMI is increased because they’re ill.



Joseph Tam, DMD, PhD (National Institutes of Health, Bethesda, MD)

Dr. Joseph Tam presented promising preclinical data (published in Cell Metabolism this July []) on Jenrin Discovery’s CB1 receptor antagonist JD-5037. Dr. Tam began by reminding the audience that until rimonabant (Sanofi’s Acomplia) was found to have severe psychological side effects, CB1 receptor antagonists were among the most promising anti-obesity drug candidates. Dr. Tam then presented a 28-day proof-of-concept murine study (initial results were presented at last year’s TOS, see page 9 of our 2011 TOS report at The study showed that JD-5037 doesn’t penetrate the blood-brain barrier to a significant extent, and thus may have a better safety profile. Dr. Tam showed that JD-5037 conferred comparable reductions in food intake, body weight, and adiposity to SLV-319 (Solvay-BMS’ ibipinabant [a CB1 receptor antagonist that JD-5037 was derived from). Furthermore, the two drugs had comparable effects on normalizing insulin and blood glucose levels, and a similar ability to attenuate insulin resistance and glucose intolerance. The changes in metabolic parameters were seen in diet-induced-obesity (DIO), ob/ob (leptin deficient), and db/db (leptin receptor-defective) mice. However, the drops in food intake, body weight, and adiposity were only seen in the DIO mice, suggesting that these changes were due to DIO mice becoming re-sensitized to leptin. Proposing this leptin-dependent mechanism may be how JD-5037 causes weight loss without interacting with the CNS, Dr. Tam presented new data suggesting JD-5037 reverses hyperleptinemia, as it was associated with the elimination of more than 70% of the diet- induced leptin plasma levels. Further investigation found that JD-5037 accomplishes this by suppressing leptin secreted by adipose tissue and increasing the clearance of leptin through the kidneys. Finally, JD-5037 was associated with a decrease in leptin resistance, probably due to the drop in leptin. We hope to see clinical data confirming these results in humans; at TOS 2011, Jenrin Discovery said it would be filing an IND in 2012. One interesting implication of JD-5037’s mechanism is that it appears to be ineffective in obese mice with low levels of leptin. If true in humans, a person would need to have their leptin levels tested to better determine if they would respond to treatment – this may be a positive in that it may be a way to “individualize” treatment, i.e., people with higher leptin levels would be the most logical ones for this drug and would potentially see the best outcomes.

Questions and Answers

Q: Have you tried to combine this with a central nervous system antagonist?

A: No we have not.

Q: One of the things my lab was thinking of trying to do to overcome leptin resistance was to increase levels of a compound like leptin. Your model seems to be the opposite. What do you expect to have happen if you used your compound with leptin treatment?

A: We have tried to do use JD-5037 with leptin treatment, but when we do that we increase the amount of leptin in the body. Thus, the hyperleptinemia remains and so does the leptin resistance.

Q: It is a very interesting target. In both obese humans and monkeys you can have a wide range of circulating leptin levels. Some have interpreted the high ones being leptin resistant and the low ones not being leptin resistant, however we have not been able to tell a difference in leptin resistance by leptin level.

A: We are trying to see if the drug is effective in mice with very low leptin, and so far we have found that it is not effective if they have low leptin on either body weight or leptin resistance.

Oral Abstract Session: Pregnancy and Early Life

Liping Pan, MD (Centers for Disease Control and Prevention, Atlanta, GA)

Dr. Liping Pan reviewed a CDC study that examined trends from 1998 to 2010 in the prevalence of extreme obesity among US low-income children ages two to four years old. The study included data from the CDC Pediatric Nutrition Surveillance System (PedNSS), which monitors the nutrition status of low-income children and collects information on their height and weight during clinic visits. Data was included from approximately 27 million children in 30 states plus Washington, DC, and the analytic sampling consisted of a random selection of one weight/height data point per year for each child. Overall, the mean BMI of the entire study cohort increased from 16.5 kg/m2 in 1998 to 16.7 kg/m2 in 2003, and then held steady at 16.7 kg/m2 until 2010. We’re unsure if this difference was statistically significant. Similarly, the prevalence of obesity increased from 13.1% in 1998 to 15.2% in 2003, after which it remained stable at 14.9% in 2010. Encouragingly, while the prevalence of extreme obesity (defined as BMI for age and sex ≥120% of the 95th percentile according to 2000 CDC growth charts) increased from 1.75% in 1998 to 2.22% in 2003, Dr. Pan explained that it decreased from 2003 onwards to reach 2.07% in 2010. Broadly, we were hoping that Dr. Pan would have explained the statistical definition of a “decline” in prevalence, since the prevalence rate for obesity also declined between 2003 and 2010 – although overall obesity prevalence may be declining, “extreme” obesity may be increasing, and we would like to better understand these statistics.. To conclude her presentation, Dr. Pan offered possible reasons for the recent decline in extreme obesity, citing state- and local-level childhood obesity prevention efforts, national recommendations for childhood obesity prevention, and the implementation of obesity prevention initiatives within the Women’s, Infants, and Children (WIC) program.

  • This switch from an upward trend to a slight downward trend in extreme obesity was observed across all age groups and both genders, as well as all racial/ethnic groups, with two exceptions: 1) the upward trend from 1998 to 2003 was not observed in Asians and Pacific Islanders; and 2) the downward trend from 2003 to 2010 was not seen in American Indians and Alaskan Natives.

Questions and Answers

Q: You cited, as a possible reason [for the decline in extreme obesity], the policy recommendations that are given at the national level. Why haven’t those affected American Indians and Native Alaskans?

A: We see a difference by racial/ethnic group, but it’s hard to speculate the reason. For example, the breast-feeding rate is lower in this group, and there are environmental factors such as access to food and physical activities. There are studies looking at this, but I don’t feel comfortable speculating on this. That’s a great point.

Comment: I also wanted to suggest another possible reason [for the downward trend of extremely obesity], which is that birth weight has gone down since 2003.

Q: I was wondering if you also tracked food and security in the low-income families. Has that changed over this period in the US?

A: We just published information in food and security and obesity among 12 states in the US and found a strong trend. We’re looking at more data.

Q: It appeared that the data for the American Indian/Alaskan Native group was much noisier than the other ethnic groups.

A: That’s because the sample size for this group was much smaller than that of the other racial/ethnic groups. There was much more variance, as we expected.


Symposium: What Outcomes Matter – Prevention and Treatment of Pediatric Obesity to Reduce Diabetes Risk


Gary Foster, PhD (Temple University, Philadelphia, PA)

Dr. Gary Foster, a veteran psychologist in obesity prevention and intervention studies, gave an impartial overview of a number of school-based programs targeting obesity. While there have been pockets of success with different strategies or among certain subpopulations, overall, he thinks “results are not all that impressive.” As an example of one such inconclusive trial, he presented results from the HEALTHY study, which assessed an intervention designed to reduce overweight, obesity, and diabetes in sixth graders (Foster et al., NEJM 2010). The intervention consisted of four components: behavior, nutrition, physical education, and communications and was implemented in 21 middle schools with >50% minority students (black or Hispanic) and/or >50% students eligible for free/reduced lunch. BMI and diabetes risk factors were assessed at baseline (the beginning of sixth grade) and at the end of eighth grade. Additionally, 21 control schools with similar baseline characteristics received no intervention. In the 42 schools included in this study, the number of overweight and obese sixth graders was disheartening – approximately 50% of sixth graders were either overweight (~20%) or obese (~30%). Overall, there was no difference in primary outcome (prevalence of combined overweight and obesity) in the intervention (45.8%) vs. control group (45.2%); however, some secondary outcomes were achieved. For example, the number of students with waist circumference > 90th percentile was significantly lower in the intervention group (p=0.04). Interestingly, the prevalence of obesity (BMI >95th percentile) was reduced considerably in the control group (8.5% reduction), but was still less than that of the intervention group (11.5% reduction; p=0.05). Dr. Foster posited that perhaps just measuring BMI at baseline influenced children’s later behavior. Going forward, Dr. Foster suggested that school- based interventions be designed specifically for children who are already overweight or obese (as he thought this subpopulation, generally, benefits most from intervention), but also warned against the potential stigmatization associated with offering programs only to this subpopulation. We were disappointed that Dr. Foster could not offer more concrete suggestions to improve school-based interventions, and we think this reinforces just how multifaceted the problem of childhood overweight and obesity is. He summed this complexity up perfectly in his conclusion – schools are just one piece to the puzzle, he said, and we’ve got work to do.


Symposium: Losing Weight and Not Your Patients: Maximizing Adherence and Minimizing Attritions


John Foreyt, PhD (Baylor College of Medicine, Houston, TX)

To a standing-room-only audience, Dr. John Foreyt gave a fast-paced and very thorough review of successful adherence strategies to promote long-term weight management. He strongly emphasized that extended treatment beyond six months, ideally through weekly or biweekly contact, is essential for promoting adherence and success in long-term weight management. Other helpful factors include providing a multi-component program, using a multidisciplinary intervention team, and maintaining therapist contact with individuals. We certainly support the idea of long-term contact, though we wonder if there is a way to scale it cost-effectively. Dr. Foreyt, a Look AHEAD investigator, also provided a valuable review of the adherence strategies employed in the now 11-year long trial (“the best model we have”) – monthly individual sessions, weight loss competitions with other trial sites, group gatherings, special functions, etc. Aside from lifestyle intervention, Dr. Foreyt also seemed pretty enthusiastic about the potential for pharmacotherapy, noting, “Thank god. We’ve got some drugs now. Pharmacotherapy can do a terrific job of helping with maintenance.” His talk concluded with a review of four things clinicians can do to improve adherence: 1) have a comprehensive initial assessment; 2) use multiple indicators of success; 3) frame goals for post-treatment in terms of behaviors than can be controlled; 4) adopt a continuous, long-term care approach to management.

  • There is research supporting a broad array of strategies to promote better adherence to long term weight maintenance: telephone counseling; internet, email, and e- counseling; skills training and relapse prevention training; providing patients with food as a maintenance strategy (i.e., meal replacements or any kind of structured eating); physical activity; pharmacotherapy (“Thank god. We’ve got some drugs now. Pharmacotherapy can do a terrific job of helping with maintenance.”); and multi-component programs (e.g., behavioral and dietary).
  • The three biggest success factors predicting weight loss in Look AHEAD were self- reported physical activity, treatment attendance, and use of meal replacements. Dr. Foreyt showed slides for each success factor, broken down by quartile of usage – greater usage of each was associated with a perfectly linear increase in weight loss. For self-reported physical activity, individuals in the top quartile of physical activity (an average of 287 minutes per week) lost 12% of their body weight, compared to 9% in the third quartile, 7% in the second quartile, and 4% in the first quartile. The same pattern existed for use of meal replacements – top quartile users lost 11% of their body weight, vs. 6% for those in the bottom quartile. Finally, those in the who attended every class lost 11% of their body weight, compared to less than 5% for those who came to half the classes.
  • Look AHEAD, now in its eleventh year, continues to employ a variety of adherence strategies. Participants engage in monthly individual sessions, open groups with other members of the trial, national campaigns and competitions among all 16 Look AHEAD sites, and other retention strategies, e.g., those in the “10% club” receive invitations to special functions. Dr. Foreyt highlighted the impressive 94% of individuals that completed the self-assessment at year four. By comparison, he noted that this number is usually around 50% after one year in drug treatment studies.
  • Dr. Foreyt concluded with a summary of four things clinicians can do to improve adherence. First, clinicians should make sure to conduct a comprehensive initial assessment that includes physical and psychological issues. For instance, a patient may not be ready for weight loss or might have depression or other risk factors that need treatment. Second, clinicians should use multiple indicators of success beyond just weight loss: reducing cardiovascular risk factors (“5% weight loss is all it takes”), improved quality of life, an improved dietary pattern, and improved physical activity. Third, post-treatment goals should be framed in terms of behaviors that can be controlled. Last, a long-term, continuous approach to treat obesity should be adopted. Dr. Foreyt noted that newer intervention methods – like the Internet – may offer the ability to intervene at lower costs than traditional treatments and to extend the reach of future treatment and maintenance programs.


Joseph Skelton, MD (Wake Forest Baptist Medical Center, Winston-Salem, NC) and John Foreyt, PhD (Baylor College of Medicine, Houston, TX)

Q: Your Look AHEAD outcome data – especially who comes back – is impressive to say the least. What is it that you are screening for beforehand? Are you finding the most conscientious people in Houston?

Dr. Foreyt: We make them jump through hoops. They have to do a food diary for two weeks and come back four times. We put them through the rigors before they even get to the point of randomization. That’s how we got a good group. And then once you develop a relationship with patients, that keeps them coming. It’s a family and our team is fantastic. They’re in their eleventh year of treatment.

Q: The challenge now is how to integrate these programs into regular healthcare. Have you performed any cost effectiveness analyses?

Dr. Skelton: We don’t get paid very much.

Dr. Foreyt: Look AHEAD has a whole economic committee and they will publish the data. I haven’t seen it yet.

Q; I’ve been conducting a study based on the addiction model. There are forty youth and they are losing weight with this model. But they want to lose more weight than 5-8% of their body weight. I’m wondering – are we really addressing the root cause of obesity? Should we look at addictive type causes?

Dr. Foreyt: As far as losing more weight – it’s really tough. There’s that big standard deviation. Some patients do lose more weight, but they must work harder because of physiological and biochemical and psychological factors. In terms of looking at the root causes, it’s a good question. As a clinician, I suppose. As a behaviorist, we just look at behavioral factors. How do you get them to adhere to a diet and exercise program? Are we missing the root causes? Those are empirical questions and that’s very interesting.

Q: Behavioral economics is the new kid on the block. We are nudged from different areas. There’s this model of readiness to change. But sometimes we no longer have the luxury of waiting. Perhaps we should look at it from a perspective of us being the nudger?

Dr. Foreyt: Ask the economic guys. Those are issues I don’t know the answer to. You may be right. Let’s see what the publications are with respect to Look AHEAD.

Comment: As clinician and researchers, we often undersell what we do. Attrition can be helped and reduced. Look AHEAD is really important – if shows that if you do this, you will fix your diabetes. I fix people’s diabetes in medical clinic. My most gratified patients are the ones that come off insulin. Hopefully, we can turn this research data into marketing what we do.


Symposium: Weight Loss and Cancer: Helpful or Harmful (Sponsored by the American Institute for Cancer Research [AICR] and The Obesity Society [TOS])


Anne McTiernan, MD, PhD (University of Washington, Seattle, WA)

Dr. Anne McTiernan opened by briefly reviewing several observational studies demonstrating that cohort data suggest intentional weight loss is associated with a reduced risk for certain cancers. She followed with a succinct overview of studies that examined the effects of anorexia nervosa, famine, and bariatric surgery on cancer risk, concluding that while the data show a favorable effect for bariatric surgery, the studies have provided mixed results. For example, she mentioned that one systematic review aggregated 34 publications on weight loss and cancer incidence and found that 16 (just under half) found a relationship between weight loss and cancer risk. Furthermore, five of the six studies on intentional weight loss showed a negative association with cancer risk. In general, Dr. McTiernan said the positive effects of weight loss appear to be stronger in women than in men. After noting the lack of information on weight cycling and cancer incidence, Dr. McTiernan described her institute’s randomized controlled trial that examined the effect of weight loss on several cancer biomarkers. The 12-month study randomized roughly 430 post-menopausal women (average age of 58 years, BMI of 31 kg/m2, and weight of 83 kg [183 lbs.]) to a dietary weight loss program, an exercise program, a combination of diet and exercise, or a control. Participants on the diet or diet plus exercise programs

exhibited statistically significantly lower levels of several hormones (including estrone, estradiol, the sex hormone binding globulin, and free estradiol) compared to controls. Those who lost ≥5% of their baseline weight exhibited the largest effects on these biomarkers. Furthermore, the diet program (± exercise) appeared to provide favorable effects on the levels of insulin, C reactive proteins (a marker of inflammation), interleukin-6, and leukocytes. In ending her presentation, Dr. McTiernan concluded that weight loss has a significant effect on cancer biomarkers, though randomized trials are needed to determine whether weight loss can prevent the incidence of cancer.

Questions and Answers

Q: I’m worried about the effect of weight cycling on cancer risk because most people who lose weight end up regaining it.

A: We have to realize that there’s very little information on this topic. The two studies suggest that weight cycling increases the risk of renal cancer and non-Hodgkin lymphoma. Those are rare cancers as well. It would be helpful to have replication of the data and to look at more common cancers. Some of these large cohort studies – if some asked participants about weight cycling – could provide better information in the future.


Symposium: Intergenerational Transmission of Obesity

Karin Michels, ScD, PhD (Harvard University, Cambridge, MA)

Epigenetics, the “science of non-genetic, mitotically heritable variability in gene expression potential” (Waterland and Michels, Annual Rev Nutr 2007), that is, how DNA methylation, histone modification, and non-coding RNAs affect gene expression, has become a hot topic in the last several years. In her presentation, Dr. Karin Michels discussed the epigenetic links between maternal nutrition and infant weight, reviewing studies of the agouti mouse model, and the Dutch Hunger Winter phenomenon in humans. When agouti mice (yellow in color; usually overweight) are bred with non-agouti mice (dark in color; lean) and the mother’s diet is supplemented with a methyl donor (e.g., folic acid), their offspring’s coat color and weight is closer to agouti than would be expected solely with Mendelian genetics, providing evidence of the epigenetic influence of maternal nutrition on offspring weight (Waterland, Mol Cell Bio 2003; Waterland, Int J Obes 2008). Phytoestrogens have also been shown to have similar effects (Dolinoy et al., Env Health Pers 2006), while BPA has been demonstrated to have the opposite effect (that is, it reduces DNA methylation, and as such, offspring of agouti/non-agouti mice more closely resemble the non-agouti parent in phenotype). While data are sparser in humans, the Dutch Hunger Winter phenomenon provides a well-documented example. In the Dutch Hunger Winter at the end of World War II (1944-45), there was a substantial deprivation of food in the area around Amsterdam – most were only eating 500-1,000 calories per day. Individuals who were in utero during the first or second trimester of pregnancy to the Dutch Hunger Winter were found to have a 60% increase in adult obesity later in life (Ravelli et al., NEJM 1976). Researchers have suggested that caloric deprivation may have affected DNA methylation patterns later in life – methylation of the IGF2 gene has been shown to be altered in adults who were exposed in utero to the Dutch Hunger Winter (Heijmans et al., PNAS 2008). In conclusion, Dr. Michels stated that maternal diet affects intrauterine conditions, which impact both short-term and long-term growth.

Questions and Answers

Q: One of your messages was that changes during the first trimester of pregnancy are the most important. Do you think there that epigenetic effects can occur postnatally as well?

A: Yes. Even in the third trimester, in the postnatal environment, during breastfeeding, and throughout the entire life the epigenome can be modified. It might take more to modify the epigenome later in life, but if you have a major stressor, I think it could operate any time in life.



Jane Wardle, PhD (University College London, London, UK)

Dr. Jane Wardle began her presentation by explaining that obesity is quite heritable. She referenced a study of adopted children which found a that the children’s BMI were more strongly associated with that of their biological parents (with whom they have never lived) than that of their adopted parents. Additionally, twin studies have shown that roughly 70% of the variation in people’s BMI can be explained by genetics. Dr. Wardle argued that genes promote obesity by influencing one’s appetitive processes and response to food cues. In general, genes associated with weight are expressed in areas of the brain linked to appetite, and appetitive characteristics (even in one’s infancy) are prospectively associated with weight gain. She positioned this finding as good news, since genes could influence weight through pathways over which people have less control, such as the development of fat mass. Dr. Wardle then posited ways to prevent at-risk children from going down their genetically-influenced (not genetically determined) trajectory to becoming obese. She listed several potential targets within the home microenvironment that could influence childhood obesity: food availability, parental eating, desserts, TV viewing, and bedtime rules (in a twin study she found that short sleep duration is associated with weight gain and that sleep duration is less genetically regulated compared to other weight-related factors). Dr. Wardle also revealed that no parents volunteered for her study, which investigated how changes in the home environment influence healthy energy balance behaviors in children. Common parental responses to the study were: “I cant change my own eating habits”, “I don’t want arguments at mealtimes”, and “[my children] just argue with one another if they don’t watch TV”. Based on these responses, she stated that researchers must focus on identifying easy changes that parents can make. Dr. Wardle concluded that currently, the best way to reduce intergenerational transmission of obesity is to choose a thin partner.

Questions and Answers

Q: You focused on appetitive characteristics being heritable. Surely physical activity is also heritable?

A: I do not mean to imply that physical activity is not important. There is quite a strong heritable component for your day-to-day preferences for what activities you like to do. There is a lot less heritability for high levels of activity.

Comment: I agree with your ideas of home interventions, but I wanted to bring attention to a couple components of TV viewing reduction. We placed TV limiting devices in households and found that it was associated with less BMI gain in adolescents. Today, in the New England Journal of Medicine, a study was published showing that replacing sugary drinks helps reduce BMIs in children [Editor’s note details on the study are available at

Comment: You seemed to want to say something about parenting and sleep, and I want to hear what you have to say [laughter].

A: In our twin studies, we have found that short sleep duration is associated with weight gain, and that sleep duration is not very heritable. The strong influence on sleep duration is bedtime, not genetics. So sleep duration is something parents can truly control – beyond their kid’s genes – to help prevent obesity.


Symposium: Integration of Metabolic Signals by the Brainstem


Suzanne Appleyard, PhD (Washington State University, Pullman, WA)

In her presentation, Dr. Suzanne Appleyard focused on basic science research investigating the response and regulation of catecholamine neurons in the nucleus of the solitary tract (NTS; a structure within the brainstem). NTS catecholamine neurons project to the hypothalamus and to reward nuclei, and are thought to be involved in several biologic processes including food intake, the regulation of motivation, memory formation, and cardiovascular control. Dr. Appleyard’s lab showed that catecholamine neurons have mostly monosynaptic connections to the vagus nerve and are strongly driven by incoming afferents from the vagus nerve, including gastric signaling. Dr. Appleyard concluded that catecholamine neurons lie downstream of GI signaling, including CCK-sensitive afferents. Her lab also demonstrated that ghrelin (a hormone released from the stomach that stimulates food intake) decreases the basal activity of NTS catecholamine neurons, as well as their stimulated activation. In contrast, serotonin increases the firing rate of catecholamine neurons by increasing the excitatory input, glutamate.


Symposium: Obesity and Cancer

Christos Mantzoros, MD (Harvard Medical School, Boston, MA)

Dr. Christos Mantzoros gave an excellent overview of the physiology and pathophysiology of adiponectin, as well as current efforts to develop an adiponectin-based cancer therapy. He opened by reminding the audience that obesity is strongly associated with a greater risk for several cancers and that reducing visceral fat even without removing obesity can decrease this cancer risk. Turning to adiponectin, he noted that adiponectin has a strong inverse correlation with central obesity, with high amounts of visceral fat resulting in lower adiponectin levels. Dr. Mantzoros then explained that adiponectin is also inversely associated with several malignancies that have been linked with obesity and insulin dysfunction. Furthermore, previous studies have found that the adiponectin receptor is highly expressed in tumors and that the addition of adiponectin in vitro to endometrial, breast, and colon cancer cells limits cell proliferation. Dr. Mantzoros then discussed a recently-published mouse study in which adiponectin treatment reduced the growth of colorectal cancer tumors and, more importantly, increased the necrotic area within the tumor. Adiponectin treatment also led to favorable changes in insulin sensitivity, inflammation biomarkers, and angiogenic factors. To Dr. Mantzoros, these findings suggest that adiponectin may have a role in the chemoprevention of cancer. Further studies have also indicated that combining adiponectin with metformin may confer an additive effect in attenuating malignancies. After discussing possible therapeutic approaches, Dr. Mantzoros described the selective, non-TZD PPARγ modulator INT131, which is in phase 2 development. Data from 600 participants showed that INT131 (1 mg to 10 mg dosing) significantly increased adiponectin levels between five to 15 nanograms. A phase 2 study in people with type 2 diabetes revealed that INT131 provided similar glycemic control compared to pioglitazone 45 mg. Notably, INT131 1 mg leads to less weight gain and edema compared to pioglitazone 45 mg, and does not cause osteoporosis. Dr. Mantzoros concluded by announcing that the FDA has cleared INT131 for phase 3 trials, and that INT131 will be investigated in people with diabetes.

  • The hormone adiponectin is secreted by fat cells and circulates in the bloodstream at levels higher than any other hormone. The molecule binds to two receptors and promotes insulin sensitivity. Previous studies showed that adiponectin is strongly linked to central obesity, with high amounts of visceral fat resulting in lower adiponectin levels. Data show that adiponectin mediates the link between central obesity and insulin resistance, and is also associated with markers of inflammation. Aside from visceral fat, several other factors regulate adiponectin levels including a person’s diet, exercise, sleep, and genetic profile.
  • Low adiponectin levels are associated with an increased risk for several cancers that have been linked with obesity and insulin dysfunction. Dr. Mantzoros first cited a study that showed that among young women, the risk of endometrial cancer in women with high adiponectin levels was 40% of the risk in women with low levels. Data from the Health Professions follow up study (1994-2002) corroborated this finding and showed that low adiponectin levels is a strong and independent predictor of cancer.
  • Metformin has been shown to down-regulate the proliferation of cancer cells. Dr. Mantzoros briefly described an in vitro experiment where interleukin-1 beta induced malignant potential was attenuated by both adiponectin and metformin individually. Notably, combining both agents produced a greater effect than either alone, which to Dr. Mantzoros indicates that scientists could potentially develop both adiponectin and metformin for the treatment of certain cancers. He then emphasized that adiponectin and metformin activate both overlapping and distinct signaling pathways in humans and animals.
  • Dr. Mantzoros posited several therapeutic approaches involving adiponectin. He first noted that the adiponectin molecule itself is too complicated to develop a therapy for human use. However, companies have exerted significant efforts to create adiponectin receptor agonists. Dr. Mantzoros also postulated that scientists could manipulate the expression and/or secretion of adiponectin to increase plasma adiponectin levels.

Questions and Answers

Q: In the Health Professions study, was there a control with the BMI or body weight?

A: Yes. We controlled for BMI or waist:hip ratio. If one controls for BMI, the association with adiponectin remains. If one controls with waist:hip ratio, the association with adiponectin becomes borderline significant. This indicates that most of the effect through adiponectin is through waist:hip circumference, but not 100%.

Q: Is there an effect of adiponectin on any of the nutrient transporters that are important in nutrition?

A: I’ve never looked into it.

Q: So adiponectin is produced and circulates at extremely high levels. At those levels, theoretically, adiponectin would saturate all the receptors. So the big question is how does the decrease from 10 micrograms to 5 micrograms (which is still a very high level) have such potent biological effects?

A: We have talked about this. It’s a philosophical question that can apply to other hormones.

Q: In your mouse study, when you give 5 micrograms of adiponectin, how much does that influence the adiponectin level in the mouse itself? Were you able to see an actual increase?

A: We did. I don’t remember the concentration. We increased adiponectin to supraphysiological levels. It didn’t go from a low to high physiological level. It was a proof of concept study so we wanted an effect.

Q: Did you look at the relationship between adiponectin and cortisol? I didn’t see anything to show that you looked at cortisol.

A: There is an association with cortisol, but it’s much weaker than adiponectin’s association with insulin or estrogen or testosterone.

Q: You mentioned that adiponectin has receptors on muscles. Does low adiponectin levels increase or decrease muscle mass?

A: I don’t think that there is any major change in muscle mass that could be detected. Adiponectin influences metabolism at the level of the muscle, but there’s no increase or decrease in muscle mass.

Q: How about androgen levels? Does adiponectin increase or decrease androgen levels when it’s low?

A: Androgen levels decrease adiponectin, and adiponectin pushes androgen levels up

Q: We’ve shown an association between neck circumference and markers of the metabolic syndrome, and some data show that neck circumference is correlated with androgen levels. Have you looked at this?

A: No, I’m aware of the recent papers on the association with neck circumference but we have not included it in our epidemiological studies. If you go back to the 1990s or the 1980s, the decision was to measure the waist circumference.


Symposium: Environmental Obesogens


Dana Dolinoy, PhD (University of Michigan, Ann Arbor, MI)

After reviewing the basics of epigenetics, Dr. Dana Dolinoy described her group’s work investigating the epigenetic effects of early bisphenol A (BPA) exposure and early lead (Pb) exposure. Normally, when agouti mice (Avy/a) are crossed with wild-type mice, the 50% of their offspring who are agouti mice follow a bell-shaped distribution in coat color and phenotype, from yellow (generally obese) to brown (lean; healthy). In the presence of exposure to high doses (50 mg) of BPA in utero, the color coat distribution skews toward the yellow phenotype, whereas in the presence of exposure to low doses (50 g or 50 ng) of BPA in utero, the color coat distribution skews toward the brown phenotype – evidence of non-monotonic effects of BPA. Female offspring exposed to BPA in utero had lower food intake, increased activity, and increased energy expenditure. Similarly, when parent agouti mice were exposed to Pb and crossed with wild-type mice, the color coat distribution of their agouti offspring skewed toward the yellow phenotype in the presence of high Pb exposure, and toward the brown phenotype in the presence of low Pb exposure. At three months, female offspring who were exposed to higher doses of Pb in utero experienced a decrease in activity levels; there was no difference in males. Given these findings, Dr. Dolinoy emphasized the importance of exploring the dose response. In terms of future directions, she noted that more research needs to be done to link epigenetic changes to changes in phenotype and human health outcomes.

Questions and Answers

Q: Which tissues did you use to look at the epigenome in mouse studies? Are you in the process of comparing miRNA data?

A: The data you saw is from liver tissue, although we’re also looking at other tissues. We’ve designed our study to be able to look at other epigenetic marks.

Q: Is the dose response of coat color related to the other phenotypes that you measured – e.g., activity? Is color coat a good measure of food intake?

A: No, we purposely separated those two experiments. We know why the agouti mouse gets fat. It’s not directly translatable to humans. Really, you just use the agouti mouse as a biosensor for whether nutrition or the environment can change methylation. That’s all it can truly tell you.


Symposium: The Importance of Circadian Rhythms and Sleep in Maintaining Metabolic Homeostasis


Fred Turek, PhD (Northwestern University, Evanston, IL)

Dr. Fred Turek gave a thoughtful presentation on the importance of sleep and circadian rhythms. He opened by noting that studies on sleep disturbances have focused on humans rather than animal models, and that the reverse is true for studies on disrupted circadian rhythms. After describing the traditional and contemporary views of circadian rhythms (details bel0w), Dr. Turek discussed the molecular basis of the circadian clock. He explained that scientists’ understanding of circadian rhythms has rapidly evolved over the past six to seven years, beginning with the discovery of the clock gene in the late 1990s and the finding that a single point mutation within this gene changes the free running period. Within four to five years of this breakthrough, scientists discovered several additional genes that influence circadian rhythms, including bmal, cry, per, and tim. The field has now moved to focus not only on the role of these genes in circadian rhythms, but on their interactions with genes that regulate metabolism and the immune system, as well as their influences on obesity and metabolic disease. Animal studies show that mutations in the clock gene can lead to greater weight gain, hyperinsulinemia, and diabetes. Dr. Turek then focused on environmental models of sleep disruption, noting that the timing of eating appears to be a key factor in regulating weight gain. Human studies that altered participants’ sleep cycles show that sleep debt is associated with insulin resistance, altered production of leptin and ghrelin, and increased energy intake. Notably, a literature review including 84 epidemiologic studies (up until 2010) found that 76 of the 84 studies show an associated between sleep duration and obesity. Dr. Turek then ended his presentation by championing for circadian clinics.

  • The traditional view of circadian rhythms posits that humans require both external and internal synchronizations to regulate our internal clock, which is located only in certain brain areas. Disruption of this clock (through jet lag, shift work, etc.) may result in a higher risk for disease. While this viewpoint remains valid, several fundamental changes in our understanding of sleep regulation have led to the contemporary view, which states that chronic sleep loss leads to obesity, diabetes and cardiovascular disease, that circadian dysregulation results in disease and metabolic dysfunction, and notably, that the molecularly machinery for humans’ circadian clock is located in all cells in the body, not just in the brain. Previous studies have revealed that this clock regulates the timed expression of roughly 10% of all genes in a given tissue.

Questions and Answers

Q: Can you allude to how you see the interface of clock and the regulation of networks in the context of metabolism or food ingestion? You showed that the time of day can affect or interfere with overall the energy balance. Have you identified specific networks that seem to be specifically driven by the discrepancy?

A: I would say no. I took out that slide, but I wanted to show you how the circadian system and the sleep cycle are interacting with each other. In all the mutant animals that people have worked with, they find major change in the sleep cycle. Many people are looking at the relationship between clock and sleep at the molecular level, but tying that into metabolism – I don’t think we’re there yet.

Q: If you gain weight depending on when you eat, can manipulating the timing of feeding help with weight loss?

A: That’s a very good question. If someone says to me, “If I eat at a different time of day, I’ll lose weight” I would say, “You probably should drop calories too.” People have said, “Fred, this defies the law of thermodynamics: energy in vs. energy out.” But the sort of nutrient content can have an effect. Even though you’re taking in the same amount of calories, whether you take on weight depends on the nutrient content of the calories. It might also concern the microbiome, which changes if you eat at one time vs. another. Remember, it’s made up of bacteria with their own circadian clocks. How have their clocks evolved along with our clocks? That’s a very exciting area.


Key Lectures


Barbara Kahn, MD (Harvard Medical School, Boston, MA)

Dr. Barbara Kahn described her lab’s process of finding a novel phosphorylation site on the AMP- activated protein kinase (AMPK) that elucidates leptin’s effect on AMPK and leptin’s ability to suppress food intake. She began by explaining that AMPK activity in the hypothalamus influences food intake and metabolic effects, including energy expenditure and glucose homeostasis. Dr. Kahn’s lab had previously demonstrated that leptin-induced inhibition of AMPK activity in the hypothalamus is necessary and sufficient for leptin’s effect on food intake and body weight. Subsequent studies confirmed that the PI3K- AKT and mTOR-p70S6 kinase pathways converge to mediate leptin signaling at AMPK (PI3K, AKT, and p70S6 kinase appear necessary for leptin-induced AMPK inhibition). Specifically, p70S6 kinase (S6K) phosphorylates AMPK at serine 491 on the α2 catalytic subunit to inhibit AMPK activity, thereby mediating leptin’s effect on AMPK. Dr. Kahn then discussed negative regulators of leptin signaling, focusing on the protein tyrosine phosphatase PTP1B, whose gene is located on a region of the human chromosome where several mutations have been linked to obesity and diabetes. Blocking PTP1B expression in different areas of the brain leads in inconsistent effects on leptin levels, energy expenditure, insulin sensitivity, and core body temperature. The inconclusive results reflect the intricacies of the signaling pathways involved in energy homeostasis, and Dr. Kahn noted that additional studies using compound knockout rodents will hopefully clarify how different modulators of leptin activity and insulin action work together in the hypothalamus.

  • Dr. Kahn’s research sought to determine how leptin inhibits AMPK activity in the hypothalamus. Previous data showed that leptin’s ability to inhibit AMPK activity in the hypothalamus was critical for leptin’s effect on food intake. Dr. Kahn’s research found that animals with mutant, constitutively active AMPK failed to exhibit leptin’s anorexigenic effects, indicating that inhibition of hypothalamus AMPK activity is not only necessary for leptin’s mechanism of action, but also necessary and sufficient to regulate food intake and body weight. Dr. Kahn’s lab subsequently showed that DIO rodents had increased leptin levels but reduced AMPK inhibition, suggesting that an inability to modulate AMPK activity likely contributes to leptin resistance.
  • Dr. Kahn then researched the molecular mechanism through which leptin inhibits AMPK in the hypothalamus. She explained that the phosphorylation of AMPK at threonine 172 of the α1 catalytic subunit is essential for activating AMPK. While research had revealed that phosphorylation of AMPK also occurs on serine 485 on the α1 subunit and on serine 491 on the α2 subunit, the functional consequences of this phosphorylation in vivo was not known, though at the time Dr. Kahn hypothesized that serine phosphorylation inhibited AMPK activity. Her lab showed that leptin stimulates serine485/491 phosphorylation and inhibits AMPK activity in the hypothalamus. Preventing this phosphorylation in vivo blocks leptin’s effect on AMPK and led to a lower degree of weight loss, as well as greater food intake in AMPK mutant mice.
  • In researching pathways responsible for the serine phosphorylation of α2AMPK in the hypothalamus, Dr. Kahn focused on PI3 kinases, which mediate insulin signaling. Her lab found that inhibiting PI3K blocks leptin-induced serine phosphorylation of AMPK. Furthermore, a molecule downstream of PI3K in the pathway, AKT, is necessary for AMPK serine phosphorylation and the inhibition of AMPK activity. However, the amino acids flanking serine491 suggested that it would be difficult for AKT itself to bind to and phosphorylate the site.
  • The p70S6 kinase (S6K) lies downstream of AKT and inhibits AMPK. Previous finding showed that preventing S6K expression blocks leptin’s anorexigenic effects, and that leptin stimulates S6K. These results prompted Dr. Kahn to investigate whether S6K mediates leptin’s effect on the serine phosphorylation of AMPK. Constitutively active S6K resulted in the phosphorylation of AMPK at serine485/491, while mice not expressing normal S6K failed to exhibit leptin’s effects on food intake and body weight. Her lab also demonstrated that p70S6 kinase forms a complex with AMPKα2 within the brain. Based on these results, Dr. Kahn concluded p70S6 kinase is a novel inhibitor of AMPK that phosphorylates α2AMPK at serine 491. She also hypothesized that under some circumstances, one linear pathway (PI3K → AKT → p70S6 kinase) mediates leptin-induced AMPK inhibition, rather than multiple parallel pathways.

Questions and Answers

Q: Is the binding of S6K and AMPK regulated by nutrient sensitivity?

A: Is it dependent on leptin? I don’t think we’ve done that experiment, because it takes a lot of material to do that. We couldn’t do it in the hypothalamus per se and would have to use a bigger part of the brain. At that point it didn’t seem relevant if we used a large part of the brain.

Q: Have you checked the neuronal activity with the deletion of PTP1B? It seems like you have an opposite effect on PTP1B knockout mice in POMC neurons. Maybe the deletion caused different changes in neuronal activity in response to leptin.

A That’s a really good point. We’re looking at that, but we haven’t done electrophysiology studies. That would be interesting to do.

Q: You mentioned that AMPK was important for a variety of functions. Does that mean that the AMPK kinases are really the regulatory substances rather than AMPK?

A: It’s very interesting about these upstream kinases. For example, LKB1 [which phosphorylates and activates AMPK] is actually not regulated in vivo. It’s constitutively on, and people are trying to figure out its gene regulation, is there sometimes less of it. There’s a lot of it around a lot of cells, and it’s constitutively on. That’s why people started looking at phosphorylation, because it’s more likely that turning AMPK off or on is the regulating factor. So I don’t think the upstream kinases are the whole system because they are not exquisitely regulated.



Robert Whitaker, MD (Temple University, Philadelphia, PA)

Addressing a full room, Dr. Robert Whitaker reviewed the findings of the Neighborhoods, Obesity, and Diabetes randomized INTsocial experiment, which investigated the effect of neighborhood poverty levels on obesity and diabetes. Impressively, this study is the largest social experiment that has been conducted in the US. While results of the study were published almost a year ago (Ludwig et al., NEJM 2011), Dr. Whitaker kept the audience engaged for his allotted hour with an excellent presentation. The intervention came in the form of section 8 housing vouchers. Four thousand six hundred families in five US cities living in high poverty census tracts (poverty rate ≥40%; mostly single women with children [n=4,498]) were randomly assigned to receive a low poverty voucher (the treatment group [LPV]) which had to be used to rent housing in a low poverty census tract neighborhood (i.e., a better neighborhood; poverty rate <10%), a traditional voucher which could be used to rent housing in any census tract neighborhood, or no voucher (the control group). While the study was meant to assess the effect of neighborhoods on economic self-sufficiency, the study actually found no effect on income or employment. Rather, when low-income women with children were randomly assigned the chance to move from a high-poverty to a lower-poverty neighborhood (48% took advantage), they experienced a lower prevalence of extreme obesity (13% relative reduction in BMI ≥35 kg/m2 and 19% relative reduction in BMI ≥40 kg/m2) and diabetes (22% relative reduction). One important study limitation was that BMI and diabetes were not assessed at baseline, as neighborhoods’ influence on health was not the original study objective. Dr. Whitaker drew particular attention to the difference in overall wellbeing between the control and LPV group at follow up: 77% of the LPV arm reported being “very happy” or “pretty happy” vs. 72.5% of the control group. Taking the benefits in health and wellbeing together with the lack of change in economic self-sufficiency, Dr. Whitaker posited that psychological wellbeing may be necessary (even sufficient) to change many physical health outcomes. We appreciated the heartwarming message underlying Dr. Whitaker’s presentation – that you can be healthier and happier without being wealthier. We certainly hope this is true.

Questions and Answers

Comment: This was the last of around 30 huge social experiments conducted in the US between 1965 and the 1990s. We did a whole series of studies between the 1960s and 1990s when we had a different government interested in empirical basis. No agency has tried to start one since, except the USDA with some small studies on food; we’ve really not been able to get Congress behind studies like this. It’s kind of sad because we learned an enormous amount from these studies and have implemented many of the lessons we learned.

Q: I read that the results were seen in people who moved to safer neighborhoods, but not racially segregated neighborhoods?

A: The question is about the recently published article in Science and whether the impact on subjective wellbeing was driven by differences in the proportion of minorities. The short answer is that it has almost everything to do with poverty rate as opposed to proportion of minorities. Neighborhood segregation is actually nearly at an all time low. At the same time, economic segregation is getting more severe in the US. Even though there’s not as much – though, still a lot – of racial segregation, there continues to be increasing economic segregation.

Q: Were addresses traced during the entire follow-up period?

A: I meant to convey that indeed, there was continuous tracking and we took into account the duration in different tracts. With regards to moves, the average control family made two moves and the average intervention family made 2.5 moves.

Q: Was the duration spent in low poverty neighborhoods associated with outcomes at all?

A: That’s a question of dose-response. The short answer is yes. There was a sense that there was a dose- response relationship across the whole intervention in terms of the percent of extreme obesity.

Q: Did you see all of the difference in morbid obesity between groups in those who actually took up the voucher in the treatment group?

A: If economists were here, they’d say that’s not something that would be easy to tell.

Comment: It would be interesting if you found nothing among the people who did not move.

A: Indeed that was the case, but there could have been characteristics among the non-movers that confounded the outcome.

Comment: But the people who took the voucher were randomly assigned to that group.

A: Yes, but there are differences in compliers and non-compliers.

Comment: For BMI analysis, it looks like there was a difference between the control and traditional voucher groups, but for diabetes there is no difference at all. Usually, the difference in BMI translates into a difference in diabetes.

A: That is astutely observed and we are humbly uncertain. If that leaves you with an ember of skepticism, blow on that ember in the morning if you want. Do you have any idea?

Comment: You may have missed a lot of cases of diabetes because you didn’t use fasting glucose or OGTT. The data is not highly inconsistent, but it is important to note discrepancies between the obesity and diabetes trend.

Q: Are there observable characteristics between the individuals assigned to intervention who took the vouchers versus those who did not take the vouchers? Could you have done an analysis of just the individuals who took voucher against the control group?

A: In the interest of time, I would tell you to look at the supplementary tables in NEJM. There are definite differences within the treatment group between those who did and did not use the vouchers, but controlling for those in the analysis takes us out of the randomized control trial paradigm.

Comment: A potential mediator for the difference in diabetes and obesity could be depression, which has been linked to diabetes.

A: That could be.



Paul Estabrooks, PhD (Virginia Tech, Blacksburg, VA)

The charismatic Dr. Paul Estabrooks began his key lecture by commenting that the best description of the current translational research paradigm is “bench to bookshelf” (as opposed being “bench to bedside,” the tagline commonly used to describe translational research) – the research gets published and stops there rather than actually benefitting patients. He explained that the current pipeline for translational research has four transitions: transition 1 (T1) is translating basic science (i.e. preclinical trials) to humans; T2 is then using clinical insights with patients (i.e. phase 2 and 3 trials); T3 is implementing these findings into practice (i.e. phase 4 trials and health service research); and finally, T4 is translating these practices into population health (i.e. outcomes studies). This pipeline, however, Dr. Estabrooks argued, rarely works due to its preferences for: (1) simple over complex answers, to which Dr. Estabrooks responded with H.L. Mencken’s quote, “To every complex question, there is a simple answer…and it is wrong”; (2) control over context – his response: Yogi Berra’s line, “In theory there is no difference between theory and practice, in practice there is”; (3) randomization over reality, “If we want more evidence-based practice we need more practice-based evidence” – Larry Green; and (4) best possible over best available evidence. Dr. Estabrooks also stated that the pipeline is too slow, that not enough effort is placed in replicating successful programs, and during Q&A he argued that too little funding is available for later stage translational research. He called for the audience to push the paradigm, shifting it to emphasize attention to intervention features that can be adopted and delivered broadly and that have the ability for sustained and consistent implementation at a reasonable cost. He also said that interventions that reach large numbers of people deserve substantial attention – even if they are not effective – because programs that are effective but only impact a few people (of which there are many) can learn how to have broad reach from these ‘ineffective’ interventions. Dr. Estabrooks concluded by arguing that society does not need more efficacy trials looking at the impact of changing eating and activity behaviors because it is already well established that improving these behaviors is effective. What should be studied, he said, is how to implement programs that reach more people and are maintainable.

Questions and Answers

Comment: I wonder if we can push the paradigm even further. The problem I see with your approach is that it is still steeped in the trial mentality. We need to move away from attribution to adaptation. I think we need more mechanisms for continuous improvement. We need data collection feedback mechanisms to allow people on the ground to learn from what they are doing. Maybe it is still too early to move this kind of audience in that direction, but I argue we need to go there.

A: I agree with that. When I say an intervention I do not necessarily mean a trial. My quick response would be I am not sure the degree to which, as scientists, we are ready to move there. I worry at this point that we don’t have the systems in place to do that.

Comment: My push back is that when we try to implement these interventions in the community we are told that we are trying to do too much, so I am not sure I could sell a ‘complex answer’ intervention in my community.

A: I would say there are different ways to apply this depending on what outcomes you are trying to get. And part of the reason we have difficulty selling complex programs is the current paradigm. The current paradigm has this idea that you first have to show efficacy for five years, and then you have to show effectiveness [efficacy in the target population] for five years, and then you get funding from the Robert Wood Johnson Foundation who will find a few nice communities for you to work in. By the time you are implementing your intervention broadly it has been 20 years.

Q: Could you talk about how you have gotten these complex interventions funded?

A: I think a lot of it is grandsmanship. Frankly we minimized the reach side of the study [part of the trial was randomized and controlled and part of the trial, the ‘reach’ portion, allowed participants to select their intervention]. We emphasized the randomized controlled portion of the trial. We did not talk about how we were going to evaluate these different structures, because first off you only have twelve pages for the grant proposal, and then you get into the complexity issue, which funders don’t like.


Special Session: TOPS Research Achievement Award


Thomas Wadden, PhD (University of Pennsylvania School of Medicine, Philadelphia, PA)

The session began with a thoughtful remembrance of Dr. Ahmed Kissebah, a renowned obesity researcher and former director of TOPS who passed away earlier this year. Dr. Kissebah’s colleague gave a heartfelt description of his work in obesity and contribution to TOPS. Afterwards, Dr. Thomas Wadden stepped to the podium to review the role of lifestyle modification in treating obesity, discuss the Diabetes Prevention Program (DPP) and the Look AHEAD studies, and offer options for disseminating lifestyle modification. Dr. Wadden began by describing lifestyle modification as the combination of reducing energy intake (by 500-1,000 calories per day), increasing physical activity (to ~180 minutes a week), setting realistic goals, and recording food intake, exercise, and weight using paper or electronic diaries. Dr. Wadden then gave a brief overview of the DPP, stating that the favorable effects of lifestyle modification on weight and diabetes risk prompted excitement in researchers, and led to the development of the Look AHEAD trial. After reviewing the one-year and four-year results from the Look AHEAD trial, Dr. Wadden then moved to discussing the Centers for Medicare and Medicaid Services (CMS’) recent decision to reimburse weight counseling (details below), as well as the many possible ways to translate ILI outside of the clinical trial setting. Specifically, he listed the YMCA, commercial programs such as Weight Watchers, and electronic weight-loss interventions as possible effective tools. Dr. Wadden ended his presentation by encouraging the audience to support the work of young researchers and, over a round of applause, stated that he will donate half of his TOPS prize to fund the Early Career Research grants.

  • Dr. Wadden briefly reviewed the Diabetes Prevention Program, which assigned overweight patients with impaired glucose tolerance to lifestyle modification, metformin, or placebo. Those on the lifestyle modification program maintained a weight loss of 3-4 kg (6.6-8.8 lbs) by the third to fourth year of the program, a result Dr. Wadden attributed to the regular group classes, as well as the individual visits. Lifestyle modification also reduced the risk of diabetes by 58% compared to placebo and by 31% compared to metformin, demonstrating that modest weight loss combined with physical activity can reduce the occurrence of diabetes.
  • The Look AHEAD trial is investigating whether a weight loss of7% with physical activity can reduce the risk of cardiovascular morbidity and mortality in overweight and obese people with type 2 diabetes. The study randomized 5,145 participants to diabetes support and education (DSE; control) or to intensive lifestyle intervention (ILI), which was given for the first four years of the planned 13.5-year study duration. Dr. Wadden explained that after conducting background research, the trial investigators modified the lifestyle program used in DPP to create Look AHEAD’s ILI program. Compared to DPP, Look AHEAD puts a greater focus on group therapy (vs. individual treatment) and uses a higher-intensity treatment program during the first few months of the trial. At year one, those on ILI lost an average 8.6% of their baseline body weight compared to a 0.7% in those on DSE. At year four, patients on ILI regained but stabilized their weight, and exhibited a 4.7% weight loss compared to the 1.1% weight loss observed in those on DSE. Furthermore, 46% of the ILI group kept off at least 5% of their baseline weight at year four compared to 25% of the DSE group. Over four years, those on ILI also experienced greater reductions in A1c, triglycerides levels, systolic blood pressure, and the use of diabetes drugs and insulin compared to those in the DSE group, as well as greater HDL levels and fitness.
  • Dr. Wadden emphasized that the Centers for Medicare and Medicaid Services (CMS) now covers high-intensity lifestyle counseling for obese seniors by PCPs, nurse practitioners, and physician assistants. Specifically, the program covers weekly sessions for the first month and twice-monthly sessions during the next five months. If the patient loses at least 3 kg (6.6 lbs) during the first six months, he/she is eligible for monthly sessions during the next six months. While Dr. Wadden applauded the CMS’ effort to combat obesity, he listed several considerations that question the practicality of the program. First, he asked whether PCPs are trained to provide lifestyle counseling, noting that they could potentially learn the treatment through workshops or certification programs. He then asked whether PCPs have the ability or time to provide lifestyle counseling, noting that a program that also includes registered nurses, dieticians, and medical assistants could be more time-efficient and economical. Third, he questioned whether PCPs can afford to provide counseling given the currently low reimbursement rate. Finally, he suggested that the CMS program allows PCPs to refer obese patients to external resources (such as TOPS, the YMCA, commercial programs, call centers, or internet-based programs) while communicating with these resources to monitor their patients’ progress.
  • Dr. Wadden ended his presentation by noting that community resources, commercial programs, and electronic weight-loss tools may be effective and cost- efficient ways to translate lifestyle modification to the real-world setting. He listed several electronic interventions that could reach large numbers of patients at potentially lower costs, including webcams and podcasts, social networking sites, email, and text messaging. However, he acknowledged that patient retention is typically low with electronic interventions, and recommended that patients supplement such programs with personalized feedback from a trained interventionist.


Lilly Scientific Achievement Award


Kevin D. Hall, PhD (NIH/NIDDK, Bethesda, MD)

Dr. Kevin Hall gave a truly outstanding review of his mathematical research to model human macronutrient metabolism. First, Dr. Hall explained the basic premise underlying his work: using inputs about food intake and physical activity to predict changes in energy expenditure, metabolic fluxes, body composition, and fuel selection. He then jumped into a variety of validation cases, where his model’s predictions were compared against actual data from various clinical trials. Most fascinating was his refutation of the 3,500 calorie per pound weight loss rule, “the most common mathematical model used in this field.” He showed multiple examples of how this rule overpredicts how much weight loss can be expected (Hall et al., Lancet 2011) because it fails to account for the dynamic weight changes that occur in the number of calories the body burns (i.e., the body’s total energy expenditure dynamically decreases with weight loss). Dr. Hall also discussed his model’s applications to a broad array of other research questions: low carb vs. low fat diets (low fat diets win in this case), why every trial seems to observe weight regain after a weight loss phase (people do not stick to a calorie deficit), modeling the rise of obesity and food waste since the 1970s, and the effect of a 20% soda tax on obesity (about four pounds lost per person over five years, far lower than the USDA originally predicted). We’re very glad to see that Dr. Hall and his team have also generated an online tool ( that calculates tradeoffs in exercise and diet to arrive at a goal weight, plus what permanent changes are required to stay at that goal weight.

  • Dr. Hall debunked the belief that a deficit of 3,500 calories equals one pound of weight loss. He reviewed an inpatient weight loss study by Leibel et al. (NEJM 1995). Study participants (n=18 obese and n=23 non-obese) were fed a liquid diet for many weeks and repeatedly had their 24-hour energy expenditure measured. They were studied at their usual body weight and after losing 10-20% of their body weight by underfeeding or gaining 10% by overfeeding. Using the study information, Dr. Hall’s model did a good job of reproducing the weight loss that was observed in study. However, the weight loss was much less than would be assumed under a 3,500 calorie deficit per pound rule. Dr. Hall explained that this rule really overpredicts how much weight loss is expected because it fails to account for dynamic changes that occur in the number of calories burned (Hall et al., Lancet 2011). In essence, total energy expenditure dynamically decreases over the course of a weight loss phase. Impressively, his model accounts for this dynamic change.
  • Dr. Hall’s model predicts that a 20% tax on caloric sweetened beverages would cause a total of about four pounds of weight loss per person over five years, far less than the twenty pounds originally predicted by the USDA in a 2010 report (“Taxing Caloric Sweetened Beverages”). The report used the 3,500 calorie per pound rule, estimating an average reduction of 40 calories per day per person. After Dr. Hall ran the data through his dynamic model, he obtained a far less impressive result and convinced the USDA to redo its previous analysis.
  • Weight regain following weight loss might not be explained by the metabolism slowing down, but rather, a failure to adhere to a low calorie diet. Dr. Hall applied his model to previously published data from a clinical study where participants lost weight over the first six to eight weeks and then slowly regained it over time. The model was used to predict total energy expenditure and energy intake changes, which were compared to the weight loss observed in the study. Dr. Hall showed that in the first six to eight weeks of the study, people stuck to an 800 calorie per day deficit and lost weight. But slowly, they relaxed their adherence to the diet over time and at the ten month mark, they were back to their starting weight. He cautioned that these assumptions might not be correct, but it paints a very different picture of weight loss and regain than is commonly assumed. He believes this is a fruitful area for further study.
  • Dr. Hall has also applied his model to the obesity epidemic to better understand 1) the calorie intake changes that have occurred since the 1970s; 2) how these changes compare to the US food supply; and 3) the astonishing amount of food that is wasted. Assuming physical activity has stayed constant since the 1970s (a big assumption in Dr. Hall’s view), his model found that a 250 calorie per day increase (per person) would have been needed to generate the average body weight increase seen in the past 40 years. Interestingly, applying this to food supply data reveals some startling findings: there’s been a 50% progressive increase in per capita food waste since the 1970s. This now totals more than 1400 calories per person per day, more than 25% of the total freshwater consumption, and ~300 million barrels of oil per year (Hall et al., PLoS One 2009). It was also unfortunate to hear that more than two- thirds of the additional food available since the 1970s has gone into the trash, with just one third of it consumed. And if physical activity has gone down since that time, these food waste figures are an underestimate. We found these numbers truly dispiriting and hope awareness of food waste grows among policymakers, producers, and consumers.
  • Dr. Hall discussed low carb vs. low fat diets, noting his model’s prediction that a reduced fat diet is associated with three-fold more fat mass loss relative to a low carb diet. This was despite a 24 hour increase in fat oxidation in the low carb group vs. no change in the low fat diet group. Dr. Hall cautioned that it is not clear if something else kicks in over time, but during this transient phase, the reduced fat diet “won” in terms of fat mass loss. An ongoing clinical study ( Identifier: NCT00846040) will compare reduced carbohydrate vs. reduced fat consumption while keeping other nutrients at standard levels to maintain an individual's weight. We will be interested to see what the clinical study shows, as most of the body composition evidence that we’ve seen seems to side with low carb diets.
  • Dr. Hall and colleagues have generated an online tool for researchers and consumers: The website allows a user to plug in demographic information and a goal weight, and the model will calculate tradeoffs in exercise and diet over time. It will also inform a user about what permanent changes are needed to maintain a goal weight. Dr. Hall noted that this has received a “tremendous response” and has been visited by more than 750,000 people since August 2011.


6. Exhibit Hall

The TOS exhibit hall first opened Thursday evening, luring attendees with an open bar, a myriad of free food stations with long lines, and 61 exhibits. As expected, Vivus’ booth (the only one taller than eight feet) stole the show and we include our impressions of the exhibit below, along with our trip to Eisai’s booth, which made no mention of Belviq (lorcaserin) – although it has FDA approval, it does not have the final DEA approval yet, which is bad timing.

  • Vivus: The platinum plus sponsor has taken conference marketing at TOS to a whole new level – white and purple Vivus lanyards securing every attendee’s badge, a full back page ad on the conference’s final program (plus a logo on the front cover!), and even a massive white and purple carpet at the entrance to the exhibit hall (beckoning attendees to booth 307, which somewhat humorously, was only about 25 feet away). The Qsymia (phentermine/topiramate ER) branding could absolutely not be missed (relatively speaking, it reminded us of Novo Nordisk’s frequent larger-than-life presence at diabetes conferences). The centerpiece of the booth was the healthcare provider training program, where visitors could enroll in the program on iPads and watch a narrated slide deck for more information. A concise eight-page handout had a summary of the REMS-focused training program, five true/false assessment questions, and a contact information form. We were somewhat surprised to see that completing the HCP training program only requires answering the five true/false questions (fairly easy considering the answers were clearly stated on the following page!), filling out one’s contact information, and faxing the form in. Still, we’re impressed, seeing Vivus devote so much attention to appropriate use and HCP education, and we look forward to better understanding the REMS and HCP training once Qsymia is on the market for longer.

Interactive touchscreen monitors along the periphery of the booth summarized the drug’s prescribing information in a 17-section interface, while a dosing and management checklist brochure highlighted the titration scheme and included the prescribing information in paper form. We asked about the free Q and Me online patient support program that is “coming soon.” The program will be made available to anyone and includes a food log, an activity log, a weight tracker, recipes, “supportive emails” (e.g., reminders, daily tips), a food analyzer, and a tool for calorie burn – Vivus is clearly continuing to emphasize the importance of diet and exercise along with use of Qsymia. A medical information booth rounded out the backside of the booth; we filled out a form and requested further information on use of Qsymia in people with diabetes.

  • Eisai: Eisai’s dark wood, professional looking booth was tucked in the back of the exhibit hall where there generally was less traffic. In stark contrast to Vivus’ booth, it featured only small, nondescript Eisai logos and no mention of Eisai/Arena’s new anti-obesity drug, Belviq. When we asked the three reps for more information or material on Belviq, they politely explained that the booth was not about the drug. Instead the exhibits theme was “Insights into the neuropathophysiology of weight control” and was dedicated to Eisai’s “Obesity Education Network”, launched yesterday at Bar height tables with tablets and large touch screen monitors (separated by small floral bouquets) allowed attendees to register and peruse the HCP-only educational website (upon visiting the site, access is only permitted if a user answers “Yes” to the statement “I am a US healthcare professional.”). The website is broken into five sections: Pathophysiology, Obesity Education Resources, Challenges in Obesity, Obesity in the News, and Take the Quiz. There was not a mention of Belviq to be found. There are a variety of downloadable reports, an ask the expert section, and educational pages on topics like “What is Satiety?” The quiz section piqued our interest, but when we opened it on one of the booth’s tablets, the quiz already had all of the answers filled in. When we pointed this out to the sales rep, she explained that this had been happening all evening, and asked us to use one of the touch screen monitors instead. We subsequently attempted to email ourselves copies of the manuscripts “Overweight and Obesity: A Rising Epidemic With Major Risk Factors for Many Chronic Diseases” and “The Pathophysiology of Obesity: Causes and Consequences” by following the touchscreen prompt to scan our name tag. However, the Eisai rep explained that their badge scanner had not arrived yet, and a manual email entry attempt left us with a frozen screen. With no paper copies of the manuscripts and the technical difficulties, we left the booth empty handed, but we’re glad to see that this is part of the plan.


-- by Adam Brown, Nina Ran, Hannah Deming, Vincent Wu, Kira Maker, Jessica Dong, and Kelly Close