The Second Global Diabetes Summit

November 15-17, 2012; Columbus, OH Full Report - Draft

Executive Highlights

Below is our full coverage of the Second Global Diabetes Summit, which was held at the Ohio State University in Columbus, OH and began the day after World Diabetes Day (November 14, 2012). Dr. Kwame Osei (The Ohio State University, Columbus, OH), the summit’s organizer, did a remarkable job bringing KOLs together in an intimate environment to discuss diabetes and obesity on both the domestic and international levels. The impressive line-up of speakers covered a broad range of topics, as evidenced by the many (eight) different sections we have divided this report into to cover its breadth of scope: 1) Prevention of Diabetes, Obesity, and Complications, 2) Special Populations/Global Perspective, 3) Drugs, 4) Devices, 5) Type 1 Diabetes Therapies (Cure Related), 6) Gestational Diabetes, 7) Healthcare Structures, and 8) Opening Remarks (our coverage of the opening remarks made by leaders of diabetes organizations). We have highlighted our top six talks in blue (it was very difficult to narrow the list to six, and we were not able to get the list down to a round five).


  • The Second Global Diabetes Summit featured numerous thoughtful presentations on how to prevent obesity, type 2 diabetes, and reduce mortality associated with these diseases. In his comments to us prior to the beginning of the conference, Dr. Kwame Osei (The Ohio State University, Columbus, OH) identified the plenary session on type 2 diabetes prevention studies as one of his most anticipated sessions. In this session, the famous Dr. Jaakko Tuomilehto (University of Helsinki, Helsinki, Finland) reviewed the results from the Finnish Diabetes Prevention Study, noting that lifestyle intervention delayed progression to type 2 diabetes by five years. Dr. Peter Bennett (NIDDK, Phoenix, AZ) relayed results from the 2009 23- year follow-up on the Chinese Da Qing Diabetes Study, demonstrating that six years of lifestyle intervention had a legacy effect in delaying the onset of type 2 diabetes, reducing the incidence of severe retinopathy, and reducing all-cause and cardiovascular mortality in women – more evidence that early intervention seems to be the key in improving outcomes from diabetes complications. In a different session, Dr. Rury Holman (Oxford University, Oxford, United Kingdom) presented a statistical model of the legacy effect in UKPDS, which he now believes to be explained predominantly by patients’ historical A1c values. According to his model, if a 50-year- old man were diagnosed with type 2 diabetes with an A1c level of 8.0%, his mortality risk at age 70 would be reduced by 18.6% if his A1c were dropped to 7.0% immediately at diagnosis. Thus, he called for physicians to start treating their patients early and try to keep their A1c close to normal.
  • A theme that emerged from drug-related talks, as has been the case at numerous recent conferences, was the case for individualizing care. ADA President, Dr. Vivian Fonseca (Tulane University, New Orleans, LA), provided his evaluation of a number of oral drug classes under clinical investigation, stressing that despite the extensive array of drugs currently available, we still need more options because today’s drugs present a multitude of challenges for many people. We also heard a strong argument for initial combination therapy in type 2 diabetes by Dr. Bernard Zinman (University of Toronto, Toronto, Canada), as well as Dr. Ralph DeFronzo’s (University of Texas, San Antonio, TX) case for triple combination therapy (namely metformin, pioglitazone, and a GLP-1 agonist) to target the multiple underlying causes of the disease. We also attended two presentations on pharmacogenetics – the utility of predicting an individual’sresponse to a drug by using his/her genotype is surely enticing; while we don’t see this becoming common place any time soon, we are very glad to see the focus here, especially as it holds the hope of really improving the ROI of various drugs used. There are likely some respects in which genotypes are easier to use and some more challenging – for example, Dr. Eric Topol of Scripps argues that over 20% of patients on metformin have a genotype that will prevent metformin from working. Even if those patients could be identified and singled out as needing a non-metformin approach, that would be a major positive from our view – as now, those patients probably blame themselves to some extent for not seeing results from metformin. As for individualizing care, we see this prospect as exciting for patients and families but we continue to be very curious about the execution and how healthcare providers are being trained to individualize therapy given the enormous cost and time pressures they are under. We worry at this point that individualization of care is more theoretical than anything although we highly support the merits of this approach.
  • On the device front, we attended a noteworthy session on diabetes devices where the trio of Dr. William Tamborlane (Yale University, New Haven, CT), Dr. Richard Bergenstal (International Diabetes Center at Park Nicollet, Minneapolis, MN), and Dr. Aaron Kowalski (Assistant Vice President, Treatment Therapies, JDRF) spoke on CGMs, sensor-augmented pumps, and closing the loop, respectively. Dr. Tamborlane appealed to industry and the FDA to introduce easier-to-use CGMs to encourage increased patient wear time. We were particularly excited by Dr. Bergenstal’s update on the ambulatory glucose profile (AGP), a standardized, simplified means of displaying CGM data intended to provide a common platform for data analysis. He is currently working with the Helmsley Charitable Trust to develop an AGP software that CGM manufacturers could license and implement as a “title page” in their retrospective softwares; as we understand it, a white paper is being developed, which we believe will be very useful for the field. Finally, Dr. Kowalski provided a thoughtful update on progress toward the closed loop, identifying the slow on/off profile of injected insulin and the often under-recognized dysregulation of glucagon in people with type 1 diabetes as the biggest obstacles to full closed-loop control. Notably, he also disclosed that in the next few months JDRF hopes to fund the development of a bihormonal pump.
  • The conference emphasized taking a global approach to tackling diabetes – from the pre-conference reception where a Nigerian dance troupe performed a war dance to support all of the “warriors against diabetes” in attendance to the very last conference session on lessons learned from prevention trials around the world (see the “Prevention of Diabetes, Obesity, and Complications” section of this report), we were constantly reminded of the global nature of this pandemic. Dr. Jean-Claude Mbanya (President of the IDF, Yaounde, Cameroon) provided the conference’s opening presentation, reviewing the perilous state of diabetes around the world and urging attendees to lead by example in living a healthy lifestyle. We learned from various speakers about drivers of diabetes and obesity in the developing world; familiar mentions such as urbanization, increased wealth, and decreased physical activity topped the list, but speakers also cited less-frequently discussed causes such as low birth weight and childhood undernutrition, increasing gestational diabetes, obesogenic food economics, and even standards of female beauty favoring plump bodies over thin ones. Finally, the conference included workshops to address population-specific issues for Asians, Latinos, Blacks and West Indians, and youth with type 2 diabetes.
  • The conference placed a heavy emphasis on gestational diabetes, and we appreciated the opportunity to delve deeper into this important topic, especially given more recent interest in and focus on epigenetics. Dr. Thomas Buchanan (University of Southern California, Los Angeles, CA), discussed interventions in women with prior gestational diabetes mellitus (GDM). His group found that for these women, the strongestpredictor for beta cell failure is weight gain; thus he advocated for interventions that either reduce fat accumulation or modify fat metabolism (TZDs showed disease-modifying potential in the TRIPOD and PIPOD studies, though their side effects made them non-ideal). Dr. Donald Coustan (Brown University, Providence, RI) detailed the recent gestational diabetes (GDM) guidelines developed by the International Association of Diabetes in Pregnancy Study Group (IADPSG) and published in Diabetes Care in 2010. These guidelines identify roughly 18% of women as having GDM (a higher percentage than with previous US guidelines). While Dr. Coustan acknowledged that some oppose any criteria that recognize such a high prevalence of GDM, he argued that this rate simply mirrors the US epidemic of prediabetes and diabetes.
  • The summit also hosted a unique community event for about 1,000 people from Columbus, Ohio. The event afforded attendees the opportunity to hear KOLs (such as Drs. Francine Kaufman and Aaron Kowalski) discuss their areas of interest within the fields of diabetes and obesity – an opportunity we fear far too few patients get. During lunch (an assortment of wraps, fruits, and vegetables provided by the organizers), Oscar Joyner, President and CEO of REACH Media (Tom Joyner Morning Sow, gave a motivational speech during which he called for audience members to start preventing or improving their diabetes by making small incremental changes in their lifestyles. He detailed how he focused on substituting a healthier food he liked (fish) for unhealthy food he had been eating (red meats) and how his girlfriend got a weave to reduce her obstacles to working out. Later in the day there were cooking demonstrations (e.g., “Low Budget, Good Nutrition”), fitness classes (e.g., “Walking at Work”), and health screenings. Additionally, endocrinologists, dieticians, PCPs, podiatrists, pharmacists, and cardiologists were available for attendees to speak with. The event ended with NBA Hall of Famer and Novo Nordisk Diabetes Ambassador, Dominique Wilkins discussing his experiences living with type 2 diabetes, the disease both his father and grandfather died of. We applaud Dr. Kwame Osei (The Ohio State University, Columbus, OH) for organizing such an innovative event, and all of the KOLs who so graciously shared their time with the community.
  • We learned at the conference that a bill is currently in front of Congress to have Medicare cover the Diabetes Prevention Program (DPP), which would be very positive from our view. Dr. Ann Albright (Director, Division of Diabetes Translation, CDC, Atlanta, GA) updated the audience on the development of the National DPP. We think that it has the potential to help curb the diabetes and obesity epidemics as participation in the program is associated with an average 4.9% weight loss. For your reference, the Senate’s bill can be seen at and the House’s at Both bills have been referred to committee. Unfortunately, states that the House and Senate bill have only a 6% and an 11% chance, respectively, of getting past committee and only a 2% chance of being enacted. For comparison, only 13% of House and Senate bills made it past committee and only 4% of House bills and 3% of Senate bills were enacted in 2009-2010. We encourage you to assess these bills and consider contacting your Congresswomen and Congressmen.
  • The Second Global Diabetes Summit brought together so many notable speakers that we thought we would highlight some of our favorite talks to give you a place to start. Six of Close Concerns’ favorite presentations included (in no particular order)
    • Dr. Rury Holman’s (Oxford University, Oxford, United Kingdom) “Legacy Effects of Cardiovascular Outcome Studies: Metabolic Memory – UKPDS to DCCT,” which came with a clear message: act ASAP for type 2 diabetes;
    • Dr. Richard Bergenstal’s (IDC, Minneapolis, MN) very strong discussion of current/near- term issues in sensor-augmented pumping with his opinions for where the field ought to head;
    • Dr. Bernard Zinman’s (University of Toronto, Toronto, Canada) argument that there currently is no reason for the conventional “treat to failure” approach and proposal for an initial combination of metformin and TZD or metformin and DPP-4 inhibitor in lieu of initiation on metformin monotherapy;
    • Dr. Bernadette Melnyk’s (The Ohio State University, Columbus, OH) presentation “COPE TEEN (Thinking, Emotions, Exercise, and Nutrition): An Intervention to Promote Cognitive Behavior Skills and Healthy Lifestyle Behaviors in Overweight Adolescents,” which was new to us (we think she is very smart for focusing on behavior change as the key skill to teach people at risk for diabetes with related mental health conditions); and
    • Drs. Kevan Herold (Yale University, New Haven, CT) and Jay Skyler’s (University of Miami Miller School of Medicine, Miami, FL) presentations on the immunology of type 1 diabetes – we couldn’t choose just one of these incredibly comprehensive presentations on the treatment and prevention of type 1 diabetes using immunological approaches.
Table of Contents 

Prevention of Diabetes, Obesity, and Complications

Plenary Session VI: Lessons from Legacy Studies in Diabetes Prevention Trials


Jaakko Tuomilehto, MD, PhD (University of Helsinki, Helsinki, Finland)

The Finnish Diabetes Prevention study found that a four-year lifestyle intervention program was able to reduce risk of type 2 diabetes by 58% at three years follow up. The intervention asked mid-aged, overweight people with impaired glucose tolerance (IGT) to do five things: 1) lose >5% of their weight; 2) increase physical activity to >30 min/day; 3) increase dietary fiber consumption to ≥15 g/1,000 kcal; 4) decrease total dietary fat to <30% of total energy intake; and 5) decrease saturated fat to <10% of total energy intake. Dr. Jaakko Tuomilehto remarked that these intervention guidelines were much less intensive than the US’ Diabetes Prevention Program (DPP). He did not hypothesize as to why the Finnish Diabetes Prevention study’s intervention, despite being less intense than the DPP’s, delayed diabetes progression by an additional year beyond what was seen with the DPP intervention (five vs. four years) and reduced risk of progression by the same amount as the DPP intervention (58% reduction). He did, however, suggest that part of the reason the Finnish Diabetes Prevention study was successful was because it asked for people to make small (but permanent) changes, such as eating one less biscuit a day (worth 50 kcal/day or 2.5 kg/year [5.5 lb/year]) or walking an additional kilometer (0.6 miles) a day (also worth 2.5 kg/year [5.5 lb/year]). Dr. Tuomilehto thinks that because these changes were easy for people to implement and maintain, the intervention was more durable. Thus, nine years after the intervention’s start, the hazard ratio (HR) for developing type 2 diabetes was 0.61 (95% CI: 0.48-0.79). Dr. Tuomilehto explained that this means the intervention was able to delay the progression to type 2 diabetes by about five years.

  • Participants in the Finnish Diabetes Prevention study were asked to reach five targets: 1) weight reduction >5%; 2) increasing physical activity to >30 min/day (participants were offered a free gym membership, but only half accepted it); 3) increasing dietary fiber consumption to ≥15 g/1,000 kcal; 4) decreasing total dietary fat to <30% of total energy intake; and 5) decreasing saturated fat to <10% of total energy intake. Dr. Tuomilehto noted that these intervention guidelines were much less intensive than the US’ Diabetes Prevention Program (DPP).
  • Notably, when A1c rather than fasting plasma glucose was used to diagnose diabetes, the Finish Diabetes Prevention study’s intervention was not found to be significantly efficacious. During Q&A, Dr. Tuomilehto hypothesized that this could be because A1c takes longer to respond to an intervention than fasting plasma glucose does.
  • As might be expected, patients in both the intervention and control groups had zero incidence of diabetes if they achieved four or five of these targets. Those who achieved zero targets had a very high risk of developing diabetes.
  • Curiously, Dr. Tuomilehto noted that Finns who drank more coffee had a lower risk of developing type 2 diabetes than those who did not. Dr. Tuomilehto did not have an explanation for why this is the case. He noted during Q&A that this phenomena is probably not due to confounding healthy behaviors, because Finnish coffee drinkers are more likely to smoke and consume large amounts of alcohol.
  • Finland applied the lessons learned from this study in the form of The Development Programme for the Prevention and Care of Diabetes (DEHKO) from 2000-2010; the program is now being continued via the One Life projects. According to its website ( the main objectives of the One Life projects are to provide lifelong support for vascular health, prevent and delay vascular diseases, and include rehabilitation and peer support in care.
  • Dr. Tuomilehto noted that in a Japanese study, overweight people (n=641) with impaired glucose tolerance (IGT) but not isolated impaired fasting glucose (IFG) benefited from lifestyle intervention. For people who were assigned to the intervention, the hazard ratio for developing type 2 diabetes was reduced to 0.41 (95% CI: 0.24-0.69) among people with IGT. In contrast, no significant risk reduction was observed among people who had IFG and not IGT.

Questions and Answers

Q: You have some great data here. Have you done any modeling on the cost savings?

A: Well you can look at how much a diabetic patient will cost – I can’t remember exactly how much that is. There are costs also for the individual, because they have to spend their days going to the clinic. So I think if you can delay diabetes for five years, it would be a large savings.

Comment: You mentioned that coffee consumption reduces a person’s risk for diabetes. However, these people who drink large amounts of coffee might also be carrying out other healthy lifestyles.

A: People who drink more tea practice healthy lifestyle, but the coffee drinkers actually have a higher prevalence of smoking and alcohol consumption.

Q: When you looked at an A1c greater than 6.5%, why did you fail to find an effect?

A: I do not have an answer for why it is that way. The main answer is that when you have a high-risk individual with IGT and we measure their glucose levels regularly we are picking people up when their glucose levels are going up. It may be that it takes some more time for the A1c to go up. We did not follow these individuals in such a way that we could have the A1c measurements later on. A1c picks up diabetes later than glucose levels.

Q: I am just amazed how you do all of this work as a nation. Can you elaborate on how the lifestyle intervention was done?

A: Mainly through nurses and dieticians.

Q: Does it matter if the coffee is decaf or regular?

A: The Finns do not like decaf; we only like the real stuff. Normally we are drinking normal filtered coffee. Those who drink non-filtered coffee have lower health benefits from the coffee than those who do filter.

Q: What about blood pressure for coffee drinkers?

A: We know that blood pressure goes up for about one hour after drinking coffee, but it did not have an impact on hypertension incidence.


Peter Bennett, MBChB, FRCP, FFPHM (Scientist Emeritus, NIDDK, Phoenix, AZ)

Dr. Peter Bennett relayed results from the 23-year follow-up on the Chinese Da Qing Diabetes Prevention Study demonstrating that six years of lifestyle intervention had a legacy effect in delaying the onset of type 2 diabetes, reducing the incidence of severe retinopathy, and reducing all-cause and cardiovascular mortality in women. In the study, 530 people with impaired glucose tolerance (IGT) were randomized to diet, exercise, diet plus exercise, or control. After the six-year intervention, the cumulative incidence of diabetes was reduced by 31% in the diet group, 46% in the exercise group, and 42% in the diet plus exercise group compared to the control group (whose cumulative incidence was 15% at six years). In 1996, 20 years after the intervention initiation, a 43% risk reduction was observed in the progression to diabetes. With regard to macrovascular outcomes, no significant differences were observed for incidence of first cardiovascular (CV) event, CV mortality, or all-cause mortality. However, for the 23-year follow-up, the analysis split the cohorts by sex, revealing that the women who received the lifestyle intervention benefitted substantially in terms of CV outcomes: women saw a 53% risk reduction in all-cause mortality, and a 71% reduction in cardiovascular mortality (6.8% vs. 18.8% cumulative incidence in intervention vs. control groups). Men who received the intervention did not benefit in terms of all-cause mortality or cardiovascular mortality – Dr. Bennett speculated that this might be because smoking rates were dramatically higher in men than in women, potentially masking any protective effects of the lifestyle intervention. As was found in the UKPDS legacy study, it seems to us that the key to improving CV outcomes is early intervention, and the Da Qing 23-year follow-up suggests that intervening as early as the IGT stage has dramatic benefits.

  • In 1996, 20 years after the intervention initiation, investigators were able to follow-up on a remarkable 98% of study participants to examine whether a legacy effect persisted in reducing the incidence of diabetes, onset of microvascular and macrovascular complications, or mortality in the intervention groups. Even after 20 years, a 43% risk reduction was observed in the progression to diabetes (80% vs. 93% cumulative incidence in the intervention vs. control groups), representing an average 3.6-year delay in onset. Incidence of severe retinopathy was also significantly lower in the intervention group compared to the control group (9.2% vs. 16.2% a 47% risk reduction), while rates of nephropathy were similar, but very low (5.6% in the intervention group and 5.2% in the control group). With regard to macrovascular outcomes, no significant differences were observed for incidence of first cardiovascular (CV) event, CV mortality, or all-cause mortality, though there was a trend favoring a benefit in the intervention group for CV mortality and all-cause mortality.
  • For the 23-year follow-up, the analysis split the cohorts by sex, revealing that the women who received the lifestyle intervention benefitted substantially in terms of CV outcomes: women saw a 53% risk reduction in all-cause mortality (16.2% vs. 29.3% cumulative incidence in intervention vs. control groups), and a 71% reduction in cardiovascular mortality (6.8% vs. 18.8% cumulative incidence in intervention vs. control groups). Men who received the intervention did not benefit in terms of all-cause mortality or cardiovascular mortality – Dr. Bennett speculated that this might be because smoking rates were dramatically higher in men than in women (60% of men and 15% of women were smokers at baseline), potentially masking any protective effects of the lifestyle intervention.


Questions and Answers

Q: The mortality data you showed was quite interesting because though men and women in the control group had similar rate, it was just the intervention group where women benefited. To what extent could smoking explain that?

A: Among the baseline factors we had, it’s far from a complete list of what we’d like to have, but smoking was the only thing that stood out as being different among men and women. If you look at simply plugging these factors into a model, smoking certainly comes out as a determinant of mortality and CVD mortality in the men. It’s a very borderline significance in just women, but of course the power there is somewhat limited as well. It’s clearly an important determinant of overall mortality among the men. We can’t actually prove that smoking made the difference.

Q: Did you go back and look at the percentage of women who smoked during the follow-up? Women may have caught up over 20 years.

A: We do have data on survivors in terms of smoking. In those surviving up to 20 years, the percentage of women who smoked had actually fallen a little bit. But these women were also 20 years older than when they started, and there’s a tendency for smoking to decrease with age.

Q: In the first few months of follow-up there was an increase in mortality in the females. We’ve seen now some interventions where we do too much and there is an increase in mortality. Can you speak to that?

A: There were actually three accidental deaths in the men [sic] within the first six-year period. That’s where you see that little blip. It’s more in the intervention than in the control.


Enrique Caballero, MD (Joslin Diabetes Center, Boston, MA)

Dr. Enrique Caballero, one of the original co-investigators of the Diabetes Prevention Program (DPP), reviewed the initial and interim results from the DPP, highlighting that the study’s intensive lifestyle intervention delayed progression to diabetes by about four years (one year less than the Finnish Diabetes Prevention study’s lifestyle intervention). He also noted that when people in the control and metformin arms were switched to a lifestyle intervention, after the initial treatment period, they appeared to progress to diabetes more slowly. Dr. Caballero stated that visits for the DPP’s observation phase (DPPOS) end in December 2012, and that results from the full follow are expected in 2014.

Questions and Answers

Q: Maybe we should not call these trials diabetes prevention studies, because we are delaying diabetes not preventing it. How do you see us trying to delay diabetes in the real world? Should we be trying metformin with lifestyle?

A: I think that is the most important question nowadays. TZDs show the most dramatic reduction in the progression to diabetes, but they cost and have side effects that might limit the ability to use them. I think that around the world the most common recommendation is to use metformin, because it is not too expensive and the side effect profile is not bad. However, I think that it is starting to be overused in some places around the world. Based on our data it should be used in people with a BMI <30 kg/m2, who are younger than 60 years, and who have IGT. Lifestyle modification should be offered to everybody, but the challenge is implementing it in a scalable way.

Q: I just wonder how do we get public awareness?

A: It is a social and political issue. Fast food is inexpensive and it is available everywhere. I think there should be campaigns against that. It is very very difficult to do. I am amazed about what they have been able to do in Finland and Europe.


Jean-Claude Mbanya, MD, PhD (Institution, City, State/Country)

Dr. Jean-Claude Mbanya closed the symposium on diabetes prevention studies by describing the work that the IDF is doing to ensure that these trials’ findings are translated into regional and national programs. The IDF’s main approach for implementing these findings is through its Bringing Research in Diabetes to Global Environments and Systems program (BRIDGES). BRIDGES is supported by Lilly Diabetes, and provides money for research projects in primary and secondary prevention. Dr. Mbanya remarked that many of these projects have been very successful (e.g., in Vietnam and the Philippines). Dr. Mbanya then made some general comments on how communities need to begin addressing the rise of non-communicable diseases (NCDs), including diabetes, remarking that NCDs have revealed our weaknesses in public health practices. He argued that financial, legal, regulatory, and trade policies must be modified in order to encourage healthy choices. He also recommended that the food and beverage industry should be included in these conversations, because they have the money to win the fight if they do not like a policy (the recent defeat of a sugar-sweetened beverage tax in Richmond and El Monte, California would suggest as much – see our Closer Look email at On the health services front, he said that we need to change our framework for looking at healthcare. He emphasized that the focus should be on health not the disease, and that doctors who specialize in prevention should be as valued as much as – if not more than – those who focus on intervention. We look forward to seeing how the IDF works with policy makers to implement this vision.


Peter Bennett, MBChB (Scientist Emeritus, NIDDK, Phoenix, AZ); Enrique Caballero, MD (Joslin Diabetes Clinic, Boston, MA); Mahmoud Ibrahim, MD (Center for Diabetes Education, Atlanta, GA); Jean-Claude Mbanya, MD, PhD (President, IDF, Yaounde, Cameroon); Jaakko Tuomilehto, MD, PhD (International Diabetes Epidemiology Group, Helsinki, Finland)

Q: A question for Dr. Ibrahim – some Arab countries, like the UAE, are some of the wealthiest countries in the world. How are those governments involved with diabetes and obesity prevention, and do you now what percent of their GDP is devoted to turning this epidemic around?

Dr. Ibrahim: I’m not the one to talk on behalf of any government in this area. I understand there are some efforts in these areas. For example, Qatar is going to adopt a prevention program. The only thing I note in this area is that these activities are actually scattered. Despite the fact that the Gulf States have similar ethnic backgrounds or cultures, there’s not a unified system.

Dr. Mbanya: Three weeks ago I was in Qatar, Oman, and the UAE – I think they are conscious of what is happening and they know that unless they put in programs they will fail. I was in Oman to initiate their national diabetes program, and they’ve invested a lot of money. They have the money. The problem is the know-how for how to change the knowledge and culture of the people. One thing I noticed during my tour of the Gulf Countries was that primary care didn’t exist. It was all tertiary care. Now they’re starting to invest in primary care. Right now they have the Johns Hopkins, the Joslins, the Harvards, but they’ve never had the primary care. But now within the primary care system they’re putting prevention at the top of the list.

Dr. Tuomilehto: In September, the huge non-communicable disease global meeting was a big movement forward together with the WHO. There are issues in these countries that are to my understanding developing very nicely and very well, but as Jean-Claude mentioned, the know how is missing and the people have to first learn a couple of basic issues and then the situation will definitely improve. For you who are not aware, these countries are fantastic supporters of sports. So there are good signs that good things are developing. When I was in Saudi Arabia there were ladies’ gyms and things like that, so things are changing but it’s taking time.

Dr. Caballero: I can add too that from my perspective on medical education. I have the privilege of directing Joslin’s programs. What I have been able to identify is that, as you’re saying, it’s a reactive system that’s focused on the disease. They solve the acute problems but don’t see the chronic care model for diabetes care and prevention. So that’s where there’s a huge need first to educate HCPs so we’re all on same page for what we need to do. But to change the healthcare system, that’ll take longer. It’ll have to be tackled from different angles, so in conjunction with policies, governments, physicians, community – all need to move together in the right direction. There’s no healthcare system to my knowledge ready to face the challenge of diabetes and obesity. We’re going to be in big trouble everywhere unless some of the things presented today are implemented.

Q: (Dr. Tuomilehto): Peter you’ve been working in China on Da Qing, and we’ve seen the publications in China that actually prevalence is higher than previously believed. So how do you see the Chinese situation now? Also, the one-child policy was implemented in 1980, there’s been 30 years of that policy has it had an impact as well?

Dr. Bennett: The Chinese population is going to be an aging population because of that policy. On the other hand, at least it moderates the numbers to some degree, but the aging is going to be a big problem unless appropriate prevention activities can take place. The Ministry of Health has become fairly acutely aware of the problem. Certainly on the table a year or so ago were ideas about national programs. I don’t know if they have been implemented or not. But I think there are still a number of very important questions to be answered at the basic level as well as the public health level.

Q (Dr. Bennett): I’d like to pose this question to you – why do these interventions conducted over a relatively short period of time work so well over a long period of time? One thing that puzzled me in a DPP slide was that the metformin group and other groups basically were parallel – maybe there’s something I’m missing but I thought that if you continued metformin over a protracted period of time you might see divergence rather than parallel lines. So is it possible that the four years of metformin in the beginning produced the effect, and the rest of it is not important?

Dr. Caballero: I think the main difference was really seen in the first phase, as I’ve showed. The metformin group has really become a hybrid. They’ve continued with metformin but have also implemented changes. The metformin group attended the education classes more regularly in the DPPOS, showing those in the metformin are really improving their lifestyle – it’s really a combination group, so maybe that’s why they’ve reduced progression to diabetes. I think intensive lifestyle continues to be very successful, but its just that the metformin group has now combined both interventions. A key question is whether it is better to do 10 years of intensive lifestyle intervention or to take metformin and do a little lifestyle modification. I think we could argue it both ways. I still think the benefit we’re seeing at the beginning of lifestyle modification is similar to what you saw in the Finnish study as well. It continues to provide benefit in the long term.

Q (Dr. Bennett): Is it still the two kg loss you got that causes the effect or does the initial lifestyle change their behavior forever? Is it the dietary effect or the exercise or both? Jaakko’s and my data show maybe the exercise is just as important as the diet.

Dr. Tuomilehto: I think physical activity is really very important because the majority of these people have too little physical activity and if you improve it a little bit you can improve the metabolic state very much. At the same time I’d still like to emphasize what we’ve seen both in clinical trials and in obesity studies – you have to take care of several lifestyle traits at the same time. It’s not enough to look at one or two, but one must try to be on the healthy side on five or six different critical lifestyle factors. And actually I didn’t show this data here but after I think six or seven years when we asked people, it was quite clear that people had kept the healthy lifestyle. There were very few people who had relapsed. I think this is probably the most important. What comes then to the duration? What is needed to improve the situation? When I looked at our results, the main things happen during the first year when people are advised to do something. And nothing much happens after the second year. Maybe we should have programs that offer intensive intervention for two years and that’s it. If people don’t learn in those two years then they won’t ever learn, and you don’t need the ten years. Our data supports very much that probably two years is good, because one year might be enough for most but for some they would like to have consultations after two years. After that I don’t see any point to invest very much.

Dr. Bennett: Clearly if that is true then that would have an important role for how these community interventions are implemented.

Dr. Caballero: You don’t need a lot of weight loss to have an impact – that’s an important message. In DPP data what we found was that a one kg change in weight was related to a 16% reduction in type 2 diabetes. So it’s not a lot that you need to lose to have an impact.

Dr. Tuomilehto: Yes a few kg difference has a significant effect.

Q: I don’t see a lot of interest in funding studies for prediabetes for drugs with the potential to protect beta cells, like GLP-1 agonists. Why not invest in looking at different treatments?

Dr. Bennett: I think you’re jumping the gun a little – I know there are large studies for DPP-4 inhibitors in people without diabetes, and I also understand there are trials with GLP-1 agonists, but as Jaakko pointed out we don’t have the data yet. I think eventually we will. Something Jaakko mentioned – at least in my mind – we don’t have good evidence that people with isolated IFG benefit from these interventions at this time.

Q: I think every effort we do is worth it, but what happens when the public become aware of the clinical data like from the Look AHEAD trial? If they know the real data they will say, “Why should I invest in exercising and eating well if after four years I’ll be back to square one?” I think part of the problem is we’re intervening in people where it’s impossible to intervene because we’re not going to change their behavior. If you intervene to change health behaviors, we have to do it in elementary school. We’re dreaming with these studies that it’s possible to change an adult’s behavior.

Dr. Mbanya: Yes and that’s why we’re talking about investing in children.

Dr. Bennett: I substantially agree, I think the true primary prevention efforts on large scale in communities should probably focus on people below 18 years of age for practical purposes. The other thing – I’m not quite as pessimistic as you seem to be. The fact that we can delay diabetes for three to five years has an enormous impact on what happens 20 years later.

Plenary Session II


Rury Holman, FRCP, FMedSci (Oxford University, Oxford, United Kingdom)

Dr. Rury Holman presented a thorough and fascinating analysis of the UKPDS “legacy effect” – the benefits on myocardial infarction and all-cause mortality that became statistically significant only during the 10-year follow-up analysis after the intervention had completed. A new mathematical model suggests that over 80% of the legacy effect can be explained by time-weighted A1c. According to this model, if a 50-year-old man were diagnosed with type 2 diabetes with an A1c level of 8.0%, his mortality risk at age 70 would be reduced by 18.6% if his A1c were dropped to 7.0% immediately at diagnosis. By contrast, if A1c were not dropped to 7.0% until 10 years after diagnosis (i.e., age 60), then his mortality risk at age 70 would be only 6.6% lower than if no intervention had occurred. The key message, Dr. Holman said, is that “you cannot play catch-up”; he added that the clinical implication of the legacy effect is “not ‘treat early and stop,’ but ‘treat early and continuously.’” While we are unsure of the model’s exact relevance to modern-day patients (given changing health demographics, new drugs, etc.), we could hardly agree more with Dr. Holman’s closing advice: “Start early. Stay normal.”

  • Dr. Holman reviewed the UKPDS legacy effect: i.e., that the intensively managed group experienced a 15% reduced risk of myocardial infarction and 13% reduced risk of all-cause mortality that became statistically significant only after the 10-year follow-up period (even though treatment regimens and A1c values were equivalent between the intensive and control groups during this follow-up period). Dr. Holman said that he and his fellow UKPDS investigators were initially wary of assigning a cause to this phenomenon, since it could potentially have been due to unmonitored behavior changes that persisted during the follow-up period, rather than historical glucose control. For this reason they chose the vague term “legacy effect” rather than “metabolic memory,” which was used in DCCT/EDIC.
  • Based on a new statistical model, Dr. Holman showed that historical A1c does indeed seem to be the main driver of the UKPDS legacy effect. The model, generated by an analysis of 3,849 UKPDS individuals, uses a time-weighting function to assign different implications to A1c values based on how recently they were measured (Lind et al., ADA 2012 146- LB). For example, five-year-old A1c values were more correlated with risk than 10-year-old A1c values, so the model assigns more “weight” to the five-year-old A1c values. The weighting function peaked for A1c values 3.7 years in the past, but the cumulative effect of all the historical A1cs means that glycemic improvements would take roughly five-to-10 years before they translate to risk reduction in death or myocardial infarction. (Dr. Holman noted that an earlier study used similar methodology to show the effects of historical glycemic control on risk for proliferative diabetic retinopathy [Takao et al., Diabetes Res Clin Pract 2011]) When the model was applied to the entire UKPDS dataset, its predictions closely matched the actual result. Age and sex were also statistically significant factors in the model, but the effects of treatment assignment were no longer statistically significant. (We understand this to mean that the differential risks between the intensive and control group were indeed mediated mainly by historical glycemic control. We think that the role of past A1c in the UKPDS legacy effect had been widely assumed already, but we were nonetheless quite engaged by Dr. Holman’s clear and thorough presentation.) Dr. Holman then showed what the model implied for a UKPDS-participant-like patient’s risk at15 and 20 years after diagnosis, depending on whether an A1c reduction of 1.0% occurred immediately after diagnosis or 10 years after diagnosis. The model suggested that treating immediately led to much-greater risk reduction than waiting for 10 years. Even 20 years after diagnosis, the benefits of immediate vs. deferred A1c reduction would be apparent on a population scale.

All-Cause Mortality Risk Reduction in UKPDS, by Timing of A1c Decrease


Relative risk reduction 15 years after diagnosis

Relative risk reduction 20 years after diagnosis

Decreasing A1c by 1.0% at time of diagnosis



Decreasing A1c by 1.0% 10 years after diagnosis



Myocardial Infarction Risk Reduction in UKPDS, by Timing of A1c Decrease


Relative risk reduction 15 years after diagnosis

Relative risk reduction 20 years after diagnosis

Decreasing A1c by 1.0% at time of diagnosis



Decreasing A1c by 1.0% 10 years after diagnosis



Questions and Answers

Dr. Vivian Fonseca, MD, FRCP (Tulane University, New Orleans, LA): What are the implications for screening? Studies such as ADDITION have suggested that there is no value in screening, but I am not sure that this is correct.

Dr. Holman: ADDITION was not really a very intensive intervention. Though there was early screening, the A1c separation achieved was not great. If you don’t see a difference in the glucose, you are not likely to see a benefit. The argument about screening depends on cost and utility over time. If you use highly expensive pharmaceutical products from day one, the costs are so high that you can’t really balance the equation. But generic drugs are available. If we use metformin early – not necessarily before diagnosis; that is another question – but I think people have had the view that we don’t need strict control from day one. I think that view is completely wrong. Start early. Keep low. Stay low. Additionally, I think that the opportunity for screening and treating in the prediabetes phase really does need to be explored.

Symposium: Evidence-Based, Personalized and Lifestyle Approaches to Diabetes


Ann Albright, PhD, RD (Director, Division of Diabetes Translation, CDC, Atlanta, GA)

Dr. Ann Albright passionately argued that the US must begin to prevent type 2 diabetes, no matter how difficult it is to do so, because if the country does not, one-in-three American adults will have diabetes in 2050 (vs. one-in-ten now, which is already stretching the US’s resources). She began her talk with a review of the original Diabetes Prevention Program (DPP) study. She reminded the audience that those in the intensive lifestyle intervention group had a 58% reduction in the incidence of type 2 diabetes, but that the program cost over $3,000 per participant over a three-year period. However, in a real world application using the same curriculum, the cost has been reduced to $275-325 in the DPP YMCA program. Dr. Albright next discussed the National DPP, which aims to scale the DPP study’s intervention for high-risk people. She urged physicians to refer patients to the National DPP, emphasizing the high completion rate (~80% after the first four sessions) and average 4.9% weight loss. Over 9,000 people have participated in the National DPP at over 300 YMCA sites and 200 other sites in 30 states. These groups are required to submit data every six months for two years to receive full recognition. We’re very glad to see this ongoing data collection and hope it will help further refine and improve an already great initiative. Importantly, four third party payors cover the ongoing program (the CDC provides startup funds), of which United Healthcare was the first. During the panel discussion Dr. Albright cautiously alluded to a bill currently in front of Congress that, if approved, would allow Medicare to cover the National DPP. The Senate’s bill, the “Medicare Diabetes Prevention Act” can be found at and the House’s “Diabetes Prevention Act of 2012” is at We encourage you to assess these bills and consider contacting your Congresswomen and Congressmen. Unfortunately, states that the House and Senate bill have only a 6% and an 11% chance, respectively, of getting past committee and only a 2% chance of being enacted. For comparison, only 13% of House and Senate bills made it past committee and only 4% of House bills and 3% of Senate bills were enacted in 2009-2010.

  • Dr. Albright began her talk with a review of the original Diabetes Prevention Program (DPP) study. She reminded the audience that those in the intensive lifestyle intervention group had a 58% reduction in the incidence of type 2 diabetes, but that the program cost over $3,000 per participant over a three-year period. However, in studies of real world applications using the same curriculum the cost has been reduced to $275-325 in the YMCA program and $550 in a CDE study in Montana. Additionally, these studies found that weight change was similar whether the program was delivered by clinically trained professionals or lay educators, suggesting that it would be more cost effective to have the latter run it. Dr. Albright quickly emphasized, however, that HCPs are absolutely necessary for the National DPP to be a success, since the United States is in an “all hands on deck” situation; there are 79 million people who are at high-risk for diabetes (>30% risk of developing diabetes within ten years), and only 7% of Americans with prediabetes realize they have it.
  • Dr. Albright next discussed the National Diabetes Prevention Program, which aims to scale the DPP for high-risk people through four elements: (1) training to increase the workforce; (2) a recognition program to assure quality and develop a program registry; (3) intervention sites to deliver the program; and (4) health marketing to increase program use.


Linda Jaber, PharmD (Wayne State University, Detroit, MI)

Dr. Linda Jaber reported results from a DPP-like intervention, adapted to be culturally appropriate, in an Arab American population in Dearborn, MI. She found that such an intervention, especially when incorporating educational material from the National Diabetes Education Program (NDEP) and family support, was feasible and effective in promoting the adoption of preventative lifestyle changes. In order to tailor the DPP program to this population, it was translated into Arabic, provisions were made to allow female-only physical activity group sessions, and an educational module on fasting during Ramadan was added to the program. Initially, 116 individuals were identified as eligible for the study but 63 refused participation in the lifestyle intervention. After participating in the NDEP Small Steps, Big Rewards program, 49 converted to being willing to participate in the lifestyle intervention. After 24 months of DPP-model lifestyle intervention, 44% of achieved at least 7% weight loss, 59% achieved at least 5% weight loss, and 78% achieved the 150 min/week physical activity goal. Those who had family support during the intervention were significantly more likely to achieve the ≥7% weight loss goal than those who did not have family support (70% vs. 30% reaching goal). Dr. Jaber highlighted the fact that six out of the ten countries with the highest rates of diabetes are Arab (the UAE, Saudi Arabia, Bahrain, Kuwait, Oman, and Egypt) – a point we think may often be overlooked in discussions on diabetes in the developing world where the discourse focuses on the plight of Asia (where the sheer numbers are massive) and very low-income countries (where the cost of diabetes care can be overwhelming).


Samuel Dagogo-Jack, MD, DM (University of Tennessee Health Science Center, Memphis, TN)

Dr. Samuel Dagogo-Jack explained that he had sent in his original slide deck before the announcement of Look AHEAD’s suspension due to futility. After learning about the Look AHEAD study’s early completion he decided to do a different presentation that he characterized as “Looking back on Look AHEAD.” Dr. Dagogo-Jack warned against jumping to either of the stereotypical responses to a failed trial: purely accepting the findings or solely questioning the findings. He argued that on the one hand, Look AHEAD demonstrated the difficulty of losing enough weight to reduce cardiovascular risk, but that there are also reasons to question if the study was ideally designed. In particular, he questioned why the control arm was more likely to be on statins, despite there being no significant difference in LDL cholesterol, and he hypothesized that the discrepancy in statin use could partially account for the study’s negative finding. Dr. Dagogo-Jack also pointed to DCCT/EDIC as evidence that significant differences in cardiovascular outcomes are more likely to be recognized in very long trials, and suggested that if Look AHEAD had been continued that it would have eventually found a reduction in cardiovascular outcomes. Quoting Albert Einstein, “Not everything that counts can be counted; not everything that is counted counts,” Dr. Dagogo-Jack pondered if there were unknown factors increasing cardiovascular risk in people with type 2 diabetes that were not accounted for in Look AHEAD. Additionally, he noted that the intensive lifestyle intervention arm did have significant improvements in some comorbidities, such as sleep apnea, and that it will be interesting to see if the intervention was cost effective. Finally, during the panel discussion Dr. Dagogo-Jack remarked that it is possible that care for myocardial infarctions has improved greatly in the past decade making it more difficult for people to die from cardiovascular disease. He hypothesized that this might mean that cardiovascular death may now be too infrequent for many treatments to have a significant impact on this endpoint. We are interested in seeing how the study’s authors address these questions in the eagerly anticipated publication.

  • Dr. Dagogo-Jack began this presentation by remarking that new diabetes medications have not supplanted the need for lifestyle interventions. He explained that although many new diabetes medications have reached the market in the past couple decades, we should not mistake this as a golden age in pharmaceutical treatment for diabetes, because the incidence of diabetes has kept pace with these advancements.


Samuel Dagogo-Jack, MD (University of Tennessee, Memphis, TN); Ann Albright, PhD, RD (Director, Division of Diabetes Translation, CDC, Atlanta, GA); Linda Jaber, PharmD (Wayne State University, Detroit, MI)

Q: How does someone get enrolled in the National DPP, and what exactly happens once they enter?

Dr. Albright: you can come into the program through a physician referral, but also because if you’ve done a risk test. If you’re a woman with GDM your ticket’s punched – all that’s written in the standards. Once people show up, they go through 16 core sessions and a monthly follow-up. Updates are provided to the physician on how the participant is doing if they came in through the physician referral mechanism.

Q: Where do they show up?

Dr. Albright: Wherever programs are being delivered in their communities. Sometimes in a health care facility, the YMCA, a senior center, a church, a car dealership - literally it’s being delivered at a Ford dealership in some of these cities. Another thing I didn’t get to mention that is critical – we want to stay consistent with content, but need to look at the channels best available for people. We just completed advising for a study that United did – they took the content of DPP and filmed it as the content of a reality TV show. Now through Comcast and Sony, we have tested whether you can deliver this intervention in homes on people’s televisions. Results are going to be published soon – we reached millions. We really do need to study how effective website applications are and how to best reach people. It’s going to be critical for us to give people a place to go in person or technological and electronic options. We’re in continued conversations with Weight Watchers. Such groups are close to what we want done, but we just need it in a lot of places where a lot of people can access it and where what they’re going to get will be consistent and meet our standards.

Q: All of these places sound fine, but where does the public find these places? Are they listed somewhere?

Dr. Albright: They are all listed at This is an example of how we need the help of HCPs. We need help getting people to connect with our sites. We have a lot more work to do but we are only two years in, and nothing like this existed before the CDC began this project.

Q: What age does the National DPP target?

Dr. Albright: Adults, those over age at 18, because that is where the research has been done. We are looking to our colleagues at the NIH to do the research in adolescents and children, and we will begin to work on implementing the National DPP in this age group once the data is there.

Q: I’m a nurse practitioner from Washington, DC. We have organized three or four diabetes screening programs through the community. How do those kinds of sporadic events fit into the national DPP?

Dr. Albright: If you look at the standards of care, people can come in through a lot of different avenues. They have to be 18 years old, have a BMI of 24 kg/m2 or greater and 22 kg/m2 or greater if they are Asian. The third criterion is either a blood test confirming any one of the three acceptable diagnostic criteria, or a risk test indicating they have prediabetes. In that case they can self-refer but then they may have to self- pay. But they can go to the YMCA where payment is based on a sliding scale so that there’s no barrier for entry due to inability to pay. We’re trying to make it an all payor-paid so people aren’t deterred by inability to pay. I would encourage you to identify where the available programs are in the area, and when you do these screening events, be sure you’ve got the information to make it as easy as possible to get signed up for the classes. The National DPP will be a nice way to partner your efforts.

Q: I work at an urban center, and in central Ohio it looks like the cost is $160 for YMCA members and $250 for non-members. We do not see anything about a sliding scale and are concerned about our patients being able to afford this.

Dr. Albright: The sliding scale is something that we are working on. As I mentioned that are four payors that have come on to cover this right now, United being one. We are working to get payors on board. Getting Medicare and Medicaid is very important. We have been dedicating a lot of energy to getting payors to come on board. There is no question we need to have more insurance companies. People ask a lot about the uninsured, and we are getting there. When you do this you do not just flip a switch. To be where we are after two years is pretty remarkable, but we still have a long way to go. People need to push on the policies for this. I cannot talk about it largely, but I can mention that there is a bill in front of the legislature for Medicare reimbursement. So I encourage you to go look up that bill.

Q: From an international perspective – I’m from Toronto – I have an observation for Dr. Jaber. In Toronto I’m involved in a small group where we are doing a community-based health promotion. We also found that family support has been very helpful, and the interesting observation we’ve seen is that if the woman is the index subject, the family participation is less. For us if the man is the index subject, the family support is very forthcoming. Was that your observation too?

Dr. Jaber: Family support was significant for both men and women. In fact, women that involved their family were more energetic. When I talk about using peer mentoring and coaching within the groups, actually it was the women that took that curriculum, made copies on their own, and actually went back and coached other people in their families to follow the same rules, and even to friends and the greater community. So family support was relevant to both.

Q: You did not see any patriarchal dynamics?

Dr. Jaber: We did not see that, we expected to, but we did not see it.

Q: My question is about the dissemination of the DPP program into communities. I know you said it could be done by different people in different settings – will there be data to show where the programs are more effective? Or do we assume that wherever we meet people, if we have the right tool it will be effective?

Dr. Albright: We don’t have data showing whether it’s more effective in a healthcare setting. Other data have shown the intervention is equally effective when delivered by various different people. What appears to be more important is that it is readily available so there are fewer barriers for people to engage in the program. The coaches need to be excellent facilitators. Successful coaches were not necessarily incredibly knowledgeable about nutrition, but what matters was that they could help facilitate and draw information out from people. We talk about this as a lifestyle change program, honestly, not as a weight loss program. It is diabetes prevention mediated by weight loss. We talk about people needing to embrace these changes, problem solve amongst each other. It comes down to those interventions being individualized. What you eat, what your preferences are, those can be individualized within the content.

Q: I am dying to ask a question. I work with diabetes patients and I cannot believe the Look AHEAD results. You mentioned how great statins were. Did the patients that were in the intensive lifestyle intervention arm take less statins?

Dr. Dagogo-Jack: There was about a 10% difference for new starts in statins, and that is unexplained so far because the LDL levels looked similar. If you were not on any statins at the time you joined the standard arm you were about 10-20% more likely to be put on statins. This needs to be looked into. After my talk I was told that I neglected to mention that the natural history of cardiovascular disease literally changed during the study, because it is becoming increasingly difficult to die from a heart attack, because we now have stent placements etc. So nobody can design a study and say that you must never have intervention. It therefore might be increasingly difficult to show differences in cardiovascular death. I thought that was a very good point.

Comment: I’d like to say that it’s all of our responsibility out here to keep pushing the payors to pay, not for patients to go to the Y and see somebody for $10/hour. We need to get this paid for to make it easier for physicians to manage patients. So it’s important for the rest of us to push for this in the ACA to push Congress to make sure HCPs are getting paid to work collaboratively to make this a healthier world. [Applause]

Q: I am an insurance worker at a hospital. When working with type 1 and type 2 patients, an issue I see is that the ability to access fresh fruits and vegetables is very difficult. There are not even grocery store chains in those poorer populations. You can educate them but if they cannot access it, then it doesn’t do any good.

Dr. Albright: We have to work along the continuum. There are efforts on a number of fronts to improve access. They are going on in a myriad of places and ways, and are being organized by a number of different people. I think what we need to do is to look at the contribution they make, and are we getting the kinds of outcomes we want. We need both education and access efforts going forward.


Symposium: Pharmacology and Type 2 Diabetes


Vivian Fonseca, MD (President, ADA, New Orleans, LA)

Dr. Vivian Fonseca stressed that, even with the extensive array of drug options currently available, we still need more options because patients continue experience a variety of challenges with currently available therapies: patients cannot sustain good long-term control with today’s agents but have to continually add and change therapies; hypoglycemia remains a concern; few agents address postprandial glucose; many agents cause weight gain in patients for whom weight maintenance is already a challenge; and current options have not been able to eliminate the excess risk for cardiovascular disease experienced by people with diabetes. Additionally, he stressed the need to conduct comparative effectiveness trials (e.g. GRADE, ADOPT) in order to inform physicians which of the multitude of available drugs performs the best in real world settings. Dr. Fonseca then provided his evaluation of several drug classes currently under investigation: PPAR agonists, GPR-40 and -119 agonists, anti-inflammatory agents, GLP-1 agonists combined with insulin, SGLT-2 inhibitors, and 11β- HSD inhibitors. In general, Dr. Fonseca suggested that no one drug class would be able to offer all desired anti-diabetes effects (e.g. glucose lowering, beta-cell preservation, insulin sensitization, enhancing lipid profiles, weight loss, cardiovascular protection, etc.) but that it would be necessary to utilize combinations of drugs to address the multiple pathophysiological abnormalities in type 2 diabetes.

  • Dr. Fonseca stated that the holy grail of PPAR agonist development would be to develop a PPAR agonist that did not have the current agents’ negative side effect profile. Current PPAR-γ agonists exhibit great glucose lowering, insulin sensitizing, beta cell preservative, anti-inflammatory, and lipid profile enhancing effects but are also associated with edema, cardiac problems, fractures, and weight gain. He stated that for the most part, PPAR-α/ γ dual agonists have not been very successful because they raise LDL cholesterol, but that Roche’s phase 3 aleglitazar has been able to lower LDL (in contrast, the PPAR-γ agonist pioglitazone raised LDL in the same trial) while also raising HDL and providing more powerful triglyceride lowering than pioglitazone. On the other hand, Dr. Fonseca went on to say that perhaps lipid lowering is not that vital of a characteristic of a diabetes drug since effective lipid drugs alreadyexist. However, we would counter that a convenient “all-in-one” drug would be of great value in promoting patient adherence. He concluded that an insulin sensitizer that does not cause weight gain and fractures would be an important advancement.
  • GPR agonists stimulate insulin secretion from beta cells in a manner distinct from that of sulfonylureas. Dr. Fonseca stated that TAK 875, Takeda’s phase 3 GPR-40 agonist candidate, lowered glucose on par with glimepiride and with less hypoglycemia. He acknowledged that we do not fully understand the mechanism, but that this could be a promising approach to efficiently increase insulin secretion if this effect is confirmed in other trials.
  • Inflammation has been an area of interest because low-grade inflammation has been recognized to be associated with obesity and often precedes diabetes and cardiovascular disease. Dr. Fonseca believes trials to-date examining the treatment of inflammation to prevent diabetes have not been well conducted and that so far treatments with antibiotics have also not worked well. He highlighted the TINSAL-T2D study for salsalate – a de- acetylated dimer of salicylic acid used in the treatment of rheumatoid arthritis that is insoluble in acid and so does not cause stomach bleeding, as aspirin does. In the dose-ranging stage of this study, a modest A1c reduction of about 0.4% over one year was observed, but he stated that this was in the context of many patients going off of sulfonylureas due to excess hypoglycemia.
  • Dr. Fonseca stated that combining GLP-1 agonists and insulin made a lot of sense, although there appears to be a solid consensus among KOLs at this point, that has not yet translated to a high number of PCPs. Insulin’s ability to lower fasting glucose with no GI side effects but with weight gain complements the improvements in postprandial glucose, body weight, and hypoglycemia produced by GLP-1 agonists.
  • Dr. Fonseca seemed generally positive about the potential for SGLT-2 inhibitors. Dr. Fonseca stated that SGLT-2 inhibitors robustly reduce A1c (in the 0.8-1.0% range), fasting glucose, and postprandial glucose while also reducing glucose fluctuations because the amount of glucose excreted in the urine is proportional to the blood glucose level. Additionally, he noted that the glucose lowering effect is quick, does not produce hypoglycemia, and that the insulin independent mechanism makes combination therapy with other agents very effective. He cautioned the audience about various safety concerns including the increased risk of genitourinary infections, as well as the cancer imbalances seen in dapagliflozin trials and the slight rise in LDL cholesterol seen in canagliflozin trials.
  • Dr. Fonseca briefly commented on 11β-HSD inhibitors stating that the pituitary- adrenal axis is so important that any interference must be done very selectively. 11β- HSD is an enzyme that converts cortisone to cortisol that is often elevated in people who are obese.

Questions and Answers

Dr. Bernard Zinman, University of Toronto, Toronto, Canada: One could argue with you and say we have 9-10 different classes of drugs. Is the problem really that we need new drugs or is the problem that we aren’t using our therapies effectively?

A: Let me counter that by saying in the ACCORD study some very good investigators, except for me, used all these agents and what did we get? A lot more questions. So it’s good we have a toolbox with a lot of tools, but do we not need better tools?

Dr. Zinman: It seems to me that the bar is higher, and I think that clearly we want therapeutic agents that don’t cause weight gain, hypoglycemia, and have a robust drop in A1c. But we also possibly want added value: lowering LDL, blood pressure, and if one could prove CV benefit, ultimately then that new agent would go to the top of the list.

A: Rosiglitazone lowered blood pressure and had all these wonderful things. I’m interested in the net outcome. I have drugs for blood pressure, and I have drugs for lipids. I also again argue against this robust reduction in A1c. The average A1c is 7.6%. We don’t need robust lowering in new agents; we need safety and addressing targets in pathophysiology.

Dr. Zinman: What about durability?

A: Durability is important because it’s [e.g., the absence of durability] frustrating to patients.


Bernard Zinman, CM, MD (University of Toronto, Toronto, Canada)

Dr. Bernard Zinman made an emphatic case for initial combination therapy in type 2 diabetes, as opposed to initial monotherapy with a subsequent step-wise “treat-to-failure” approach. In particular he advocated for fixed-dose combinations of metformin with newer oral agents (he specifically mentioned TZDs and DPP-4 inhibitors). Compared to initial monotherapy, such fixed-dose combinations have potential to enable lower dosage (and thus perhaps fewer side effects), confer greater A1c reductions, and pre-emptively overcome clinical inertia; compared to taking two separate pills, fixed-dose combinations are likely to facilitate better adherence. Dr. Zinman said that he and other researchers in the upcoming GRADE Study tried to convince the National Institutes of Health that the study should compare initial combination therapy with a step-wise approach. Indeed, he lamented that the final study design instead compares different secondary agents added to background metformin: “an old paradigm,” in Dr. Zinman’s view.

  • Dr. Zinman criticized the ADA/EASD type 2 diabetes position statement’s recommendation to initialize therapy with metformin and then add other agents, saying that it sounds like a “treat-to-failure” approach; in this regard he was somewhat more positive on the AACE/ACE and Canadian Diabetes Association documents. He acknowledged that he could see why the AACE/ACE 2009 treatment algorithm’s flowchart might be overwhelming to some, but he said that he liked the suggestion to immediately begin dual therapy in patients who present with A1c in the range of 7.6% to 9.0%. However, he criticized the 2008 Canadian Diabetes Association (CDA) clinical practice guidelines for setting A1c ≥ 9.0% as the threshold to “consider initiating metformin concurrently with another insulin from a different class” (or simply to initiate insulin). He indicated that the 2013 CDA clinical practice guidelines will lower this threshold to ≥ 8.5%, but he believes that the threshold for dual therapy should be more like ≥ 7.0%.
  • In lieu of initial metformin monotherapy, Dr. Zinman suggested initializing therapy with fixed-dose combinations (FDCs) of metformin and a non-sulfonylurea oral agent (DPP-4 inhibitors seem the likeliest candidate for such an FDC in our assessment of the clinical practice landscape, though Dr. Zinman noted that half-doses of TZD have a more favorable risk-benefit profile than the full doses). Dr. Zinman said that some clinicians might say, “But Dr. Zinman, I have a patient who has been stable on metformin monotherapy for 20 years,” to which he would respond: “Yeah, I have ONE of those patients too.” He presented evidence that metformin’s rate of secondary failure (A1c climbing to 7.5% or higher) is 12% per year, and higher in patients with higher pre-metformin baseline A1c.
    • As for what agent to use in initial combination therapy with metformin, Dr. Zinman cautioned strongly against glyburide (“a bad drug”) and sulfonylureas generally (their only three advantages are that “they are cheap,” “they are cheap,” and “they are cheap,” he quipped). By contrast, he said that there was good logic behind a low-dose combination of metformin and TZD to achieve good efficacy with few side effects, and he reviewed that DPP-4 inhibitors can be used to boost metformin’s efficacy with minimal side effects (all of their characteristics appear favorable except for price, Dr. Zinman suggested). Indeed, one study has even suggested that DPP-4 inhibitors can reduce metformin’s gastrointestinal side effects (which are “more common than appreciated”). Dr. Zinman said that an adjunctive agent to improve metformin’s efficacy and tolerability would be quite welcome, though he emphasized that this finding had yet to be reproduced to his knowledge and that he therefore didn’t yet believe it.

Questions and Answers

Q: Low-dose combination therapy is generally a good strategy, but if you do get idiosyncratic side effects then you might not know which agent is causing them.

Dr. Zinman: Yes, but we generally know which idiosyncratic effects occur with which drug. If we saw pancreatitis in combination therapy, we would be worried about the DPP-4 inhibitor or GLP-1 agonist, for example.


Ralph DeFronzo, MD (University of Texas Health Science Center, San Antonio, TX)

Never one to be upstaged, Dr. Ralph DeFronzo began his presentation by warning that “everything [he was] going to say is in total contradiction with the ADA guidelines, which [he] thinks are totally worthless.” He went on to argue that effectively treating type 2 diabetes requires multiple drugs to correct pathophysiology, specifying that he thinks using metformin, pioglitazone, and a GLP-1 agonist (the members of his triple therapy), would work well despite being different from the ADA/EASD’s position statement. Speaking about the specific members of the triple therapy, he said that anyone who thinks that metformin’s efficacy is more durable than that of a sulfonylurea is “absolutely incorrect.” However, Dr. DeFronzo appreciates that metformin suppresses hepatic glucose production, which is why he includes it in the triple therapy. For thiazolidinediones (TZDs) he highlighted their strength as insulin sensitizers and their ability to preserve beta cell function. In addition he argued that they should still be used despite their association with bladder cancer. Dr. DeFronzo also spoke very positively of GLP-1 agonists, highlighting their ability to inhibit glucagon production, stop hepatic glucose production, help people feel more satiated, and improve muscular insulin sensitivity via weight loss. Finally, Dr. DeFronzo described how sulfonylureas (SUs) do not act to correct the pathophysiology of type 2 diabetes, do not produce durable results, and have an unacceptable safety profile, potentially accelerating atherosclerosis. Therefore, Dr. DeFronzo said that he has not used an SU since 1998.

  • Dr. DeFronzo argued that one of the “great misconceptions carried out in the diabetes literature” is that metformin is a good insulin sensitizer in muscles. He went on to explain that when he performed the research “to get metformin passed in the United States” he was unable to find any association between metformin and insulin sensitization in the muscles beyond what could be accounted for by weight loss.
  • He expressed great frustration with the idea that pioglitazone should not be prescribed because of its bladder cancer risk. First, he highlighted that the PROactivestudy found that pioglitazone reduced breast cancer risk at a similar magnitude to its association with elevated bladder cancer risk. Sarcastically, he suggested that all the women in the audience should demand free pioglitazone to prevent breast cancer, which he then said would be as weak an argument as saying nobody should take pioglitazone. Dr. DeFronzo concluded his presentation by presenting eight-year data from the Kaiser Permanente Northern California Registry ( Identifier: NCT01637935). He explained that the study’s author, Dr. James Lewis, has refused to publish the “because it makes him look foolish for rushing to publish the four-year data”; the eight-year data (n= 193,099 type 2 diabetes patients older than 40) found the HR for bladder cancer amongst those treated with pioglitazone to be insignificant at 0.98. We note that Dr. DeFronzo expressed many of the same frustrations in more detail during his presentation on the cancer risk of oral anti-diabetic medications at CODHy (for these more detailed opinions see pg. 2 of our CODHy Day 2 report at


Soren Snitker, MD, PhD (University of Maryland School of Medicine, Baltimore, MD)

Dr. Snitker reviewed evidence on various forms of monogenic diabetes and other genetic variants that might cause inter-individual variability in the response to various diabetes drugs. To date most of the research seems inconclusive and/or in insulin secretagogues. However, we share Dr. Snitker’s optimism that someday, with sophisticated algorithms and large genetic registries, such analyses could enable highly personalized diabetes therapy.

Questions and Answers

Dr. Kwame Osei (The Ohio State University, Columbus, OH): What do you think will be a general application of this technology for type 2 diabetes in the future?

Dr. Snitker: Treatment algorithms. For a particular variant, when you have modern computer systems in clinical practice settings, you could have the system tell people ‘if a patient has X variant, consider dose escalation or reduction or a different agent.’” But there aren’t really any specific algorithms, now. As we’ve seen, there are also many reasons you might not want to relay on sulfonylureas in the long run, so that might be out right from the beginning. I think it is a matter of getting new drugs and getting more information about them.

Dr. Osei: Yes, it seems we will need studies in large populations to better understand these effects.

Q: Your paper shows that sulfonylureas’ effects are long lasting, especially in monogenic diabetes. In type 2 diabetes, Dr. DeFronzo was showing high rates of secondary failures with sulfonylureas and even with metformin. So is there a difference between how monogenic and polygenic diabetes respond?

Dr. Snitker: I think that Dr. DeFronzo put it exactly right when he showed the number of ways that different agents address the underlying defects. For type 2 diabetes, sulfonylureas address none of the issues. But for some types of monogenic diabetes, sulfonylureas actually do target the sole pathophysiological defect.

Plenary Session III


Charles Rotimi, PhD (National Institutes of Health, Bethesda, MD)

Dr. Charles Rotimi began his presentation by emphasizing that the majority of health disparities are not due to genetics but rather, are due to the circumstances to which people are born. He continued to explain that while genetics may not be the dominant cause of health disparities, it could still help elucidate why some of the inequalities exist. For example, the incidence of end-stage renal disease is four times higher for African Americans than for whites, partially because a risky variant of the APOL1 allele occurs more frequently in people of African descent. Genetics can also help explain differing responses to medications. To this end, he argued against the assumption that beta-blockers are not effective in all African Americans. He explained that this assumption emerged because African Americans are more likely to have a GRK5 polymorphism that is cardioprotective; the polymorphism naturally inhibits the beta-adrenergic receptor. As a result, people with this variant do not respond as well to exogenous beta-blockers. However, exogenous beta-blockers are effective in African Americans who do not have this variant. Dr. Rotimi fears that the assumption that beta-blockers are ineffective in African Americans is preventing physicians from prescribing beta-blockers for their African American patients that would respond. He envisions future individualization of drug therapy to depend on individual patient’s genotype as opposed to imprecise indicators like race and ethnicity.

  • Part of the reason why African Americans have a four times higher incidence of end-stage renal disease (ESRD) than whites is because they are more likely to have a variant of the APOL1 allele that is higher-risk for ESRD. If a person carries two copies of the APOL1 risk allele, their odds ratio for developing ESRD is an incredible 10.5. The frequency of this risky variant is more prevalent in people of African descent, thereby increasing the likelihood that they will have a risky genotype. Dr. Rotimi explained that this allele is probably more frequent in people of African descent because the allele protects against the lethal form of sleeping sickness.


Frequency of the APOL1 risk variant

Yoruba (a West African ethnic group)


African Americans






  • The APOL1 nephropathy risk variant also helps explain why populations of African ancestry tend to have higher rates of cardiovascular disease despite having a healthier lipid profile. For background, among African Americans, increased average African ancestry is associated with lower triglyceride levels and higher HDL levels. However, a recent study found that people of African descent, unlike people of either Han Chinese or European descent, have an inverse relationship between HDL cholesterol (HDLc) levels and estimated glomerular filtration rate (eGFR). Yet, upon deeper analysis, the researchers found that the negative interaction between HDLc and eGFR among people of African descent was only seen in people with the risky allele (Bentley et al., International Journal of Nephrology, 2012). Thus, this allele might help explain why a paradoxical association of high HDL levels and poor eGFR is seen in people of African descent.
  • Dr. Charles Rotimi argued that because people of non-European descent are underrepresented in genome wide association studies (GWAS), these studies have not provided as much learning on health disparities as they have the potential to. He presented data showing that 96% of people enrolled in GWAS are of European descent; meaning only 4% are of non-European descent.
  • One striking example of the impact that the circumstances one was born into have on a person’s health is the puzzle of hypertension in African Americans – hypertension is more prevalent in African Americans than in Africans. Furthermore, studies have found that African American men who lived with both parents between the ages of one and 12 years had a systolic blood pressure that was 6.5 mmHg lower than people who never lived with both parents. This translated into a 43% reduced risk of developing hypertension.

Questions and Answers

Dr. Jean-Claude Mbanya: Is there not going to be a probe that you can put in a person’s blood and have it tell you if a beta-blocker will work in the person?

A: That would be an absolute dream. We would like to be able to determine who is going to respond to beta-blockers, and we are beginning to get there. At some point, we are all going to have our genome in our wallet that we bring in to the doctors office.

Dr. Mbanya: That actually scares me. [Laughs.]


Ewan Pearson, MD (University of Dundee, Dundee, UK)

Dr. Ewan Pearson believes that the heterogeneous group of diseases we currently call type 2 diabetes will, in the future, be distinguishable by sub-type, potentially based upon distinct genetic foundations. Such sub-typing could allow for more precise predictions for how well different patients would respond to a drug or what the likelihood of an adverse reaction would be based on factors such as age, sex, BMI, waist circumference, insulin levels, lipid levels, genotype, and biomarker panel. Dr. Pearson highlighted how monogenic forms of type 2 diabetes are successful treated using an understating of the underlying genetic defect. For example, in patients with HNF1A MODY, sulfonylurea (SFU) treatment is very successful because the drug acts downstream of the HNF1A genetic defect. He suggested that in the future a similar approach to making treatment decisions could be used more broadly in type 2 diabetes. Dr. Pearson then discussed various genetic variations that are already known to cause differential drug responses in people with type 2 diabetes; people possessing a slow CYP2C9 enzyme clear SFUs more slowly out of their system and are thus ultra-sensitive to SFU treatment. Another example is that some people have a loss of function mutation in OCT1 (a cation transporter) that makes them less responsive to metformin. However, Dr. Pearson stated that the magnitudes of these effects are not large enough to justify the cost of genotyping patients for OCT1 or CYP2C9 prior to deciding to initiate metformin or SFU therapy. Notably, Dr. Pearson stated that Lilly and Sanofi are funding the DIRECT trial, which will aim to stratify patients with diabetes by how quickly the disease progresses and by their response to therapies.

Questions and Answers

Q: We heard SFU bashing by Dr. DeFronzo yesterday, but I kept hearing SFU, SFU, SFU in your presentation. So that is interesting. The other interesting point is that you are still trying to understand the acute insulin response to SFU, which you are probably using in your new study to predict who is a good responder and who is not a good responder. I don’t want to hear about SFU treatment of MODY, but I want to hear from you how you value for typical patient with type 2 diabetes.

A: I think the debate about SFUs has gone on for years. You could have many good discussions about whether SFUs are good or not, as you could about TZDs. The future is about risk/benefit. If we could say someone could benefit really well to SFUs then their benefit is higher. The same is true for TZDs. TZDs are the perfect example of where pharmacogenetics could save them. If we could find a subgroup of people who respond well or don’t get the bad side effects, we could treat them. On another note, another big bugbear of mine is that we don’t stop drugs in people who don’t respond. I think we should be trialing all these drugs in individual patients. If someone doesn’t respond to an SFU or TZD then why carry it on? I think all the drugs have a significant future and we will be needing to use them and find the best drug for each patient.

Comment [Dr. Jean-Claude Mbanya]: No drug is bad as long as you are using it on the right patient. It’s not about the drug but the patient’s response to the drug.

Q: There is an increasing body of evidence that genotype might play a role in effectiveness of various prevention methods. I’m particularly interested in metformin – any comments on use of genotyping in prevention studies? Not just limited to metformin but other interventions as well.

A: Yes, like in the DIRECT study – a lot of money has been spent on prediabetes and trying to identify biomarkers that will predict rapid progression to diabetes. So if you’re going to have a preventative measure, clearly one approach would be to target preventative therapeutics to those who are at high risk. So metformin is an obvious one. Jose Flores has done a lot of work in looking at how metformin delays progression to diabetes and has some data showing that some of the OCT genes I talked about seem to have an interaction between metformin, genotype, and progression to diabetes. I think it’s a definite area of interest. If we’re going to use pharmaceutical therapies to treat prediabetes, which I guess is a controversial point, it makes sense in my mind that we do that in a high risk subgroup.

Plenary Session I


Griffin Rodgers, MD, MBA (Director, NIDDK, Bethesda, MD)

Dr. Griffin Rodgers provided an overview of the role the NIDDK has played, and will play, in translating clinical research into real-world applications (the so-called “T2” translation in the translational research paradigm) as well as touching upon two current “T1” translation efforts (translating basic science into new treatments). He focused on findings and implications of the DPP and DCCT/EDIC studies. Dr. Rodgers also briefly introduced three future studies that will focus heavily on primary prevention or disease reversal: D2D will examine whether optimizing vitamin D levels in patients at high risk might be able to prevent diabetes onset; RISE (Restore Insulin SEcretion) is a collection of studies independently testing various interventions that show promise in augmenting insulin secretion and preserving or expanding existing beta cells in people with type 2 diabetes; and GRADE will examine the best way to utilize the five second-line drugs in the 2012 ADA/EASD position statement. Most interestingly, in our perspective, Dr. Rodgers also highlighted two promising areas of bench research that are now being translated into potential new therapies: brown adipose tissue activation as a means of elevating energy expenditure and manipulating populations of gut microbiota to promote weight loss. Dr. Rodgers explained that lean individuals have more active brown fat than obese individuals and that the hormone irisin has been identified as inducing thermogenic activity of brown adipose tissue. Therefore, the possibility of pharmacologically inducing irisin in obese individuals is an attractive prospect. With regard to gut microbiota, he cited the result where gut microbiota of obese mice were transplanted into lean mice and caused them to gain weight. He stated that recently the G-protein coupled receptor-41 (GPR-41), which is regulated by short-chain-fatty-acids, has been identified as a mediator of this effect, so this may be a new target for obesity.


Plenary Session II


William Tamborlane, MD (Yale University, New Haven, CT)

Dr. William Tamborlane said that real-time CGM has great promise to increase the effectiveness and safety of intensive insulin therapy in children and adults, but he acknowledged that realizing the technology’s potential “has been more challenging than we may have expected” – especially for pediatric patients. In the JDRF CGM randomized controlled trial, only the patients under 25 years old who wore the sensors for at least six days every week accrued A1c benefits. Dr. Tamborlane explained that less- frequent wear time could not simply be blamed on poor compliance, because the people who wore their CGM least had a reason for doing so. Indeed, though less-frequent wearers tended to report the same benefits as more-frequent wearers, they reported a greater number of hassles. Therefore Dr. Tamborlane’s message to industry (and the FDA) is that patients need sensors that are better and easier to use (and/or CGM-based systems with expanded functionality, such as low-glucose-suspend pumps). He believes that cost-effectiveness analyses, a kind of “voodoo science” in his view, will become more favorable with new technologies (e.g. sensors that do not require confirmatory fingersticks). Dr. Tamborlane also mentioned that in two recent DirecNet pediatric studies enrolling children ages <4 years and 4-9 years, respectively, no A1c benefit was shown; he attributed this result in part to parents’ focus on minimizing hypoglycemia. We certainly agree that people have different goals for diabetes technology and thus were glad that the FDA appears open to approving artificial pancreas device systems that reduce hypoglycemia, even if they slightly increase A1c.

Questions and Answers

Dr. Richard Bergenstal (International Diabetes Center at Park Nicollet, Minneapolis, MN): When you are thinking about a sensor or sensor-pump combination, do you use any methods to determine beforehand who might be a good candidate?

Dr. Tamborlane: We did look at predictors in JDRF study and found only a few. One predictor was frequency of use within the first three-to-four weeks of the study. This got us to thinking about uses of CGM in clinical trials. But overall we need better devices; the motivation will clearly be better if we can have sensors initiate automatic changes in basal rates even when the patient is sleeping.

Q: Do you think that the results might have been different if different sensors were assigned?

Dr. Tamborlane: All three of the sensors, including the Navigator, were available. We thought about this – it was sort of the dilemma we faced when planning the DCCT. We wanted to know if the availability of sensor data improved control. We decided in DCCT that we weren’t going to randomize patients to insulin pumps or multiple daily injections; we would let them choose. It was similar here. Each of these sensors had their own little quirks; Navigator fell off, for example. We even allowed patients to switch. Sensors are fortunately better now.

Comment: We currently use predominantly one company’s sensors, and I think that the retention might be better with a different company.

Dr. Tamborlane: No single study answers everything. Do you any of you watch the show NCIS? Just like Gibbs has his rules, I have Tamborlane’s rules of clinical trials. One of them is, “You don’t do things that might impair your ability to show the primary outcome.” Our study was to see if using CGM could decrease A1c. The study you are proposing is, “Is one sensor different from another?” That is a different question.


Richard Bergenstal, MD (International Diabetes Center at Park Nicollet, Minneapolis, MN)

Dr. Richard Bergenstal reviewed studies of sensor-augmented insulin pumps and shared his perspective on how greater, more widespread benefits can be achieved in the future. To illustrate the benefits of sensor-augmented pumping he reviewed the STAR 3 randomized controlled trial and observational data from the Helmsley Charitable Trust, but he highlighted two distinct problems: sub-optimal adoption of insulin pumps and (especially) CGM, and sub-optimal clinical results even in STAR 3 (mean A1c 7.5% and 13 severe hypoglycemic events per 100 patient-years, as compared to an “optimized sensor-augmented pump” target of 6.5% A1c with no severe hypoglycemia). Dr. Bergenstal outlined several ways to improve patients’ uptake of and experience with the technology, including better pumps, sensors, and infusion sets; better education (e.g., on carb counting and pump settings); and wider understanding of and agreement on CGM-based glycemic targets (on the part of patients and providers, but also the FDA and payers). Dr. Bergenstal stated that another area ripe for improvement is the way that retrospective CGM data are presented. He advocated for all CGM companies to adopt a standardized, simplified “cover page” in their data reports, so that a clinician (or even a patient) could identify clinically relevant glycemic patterns within 30 seconds. The Helmsley Charitable Trust is funding development of the International Diabetes Center’s Ambulatory Glucose Profile (AGP) to become such a standard. Dr. Bergenstal indicated that when finished, the AGP software (called capturAGP) would be available for licensing to CGM manufacturers, who (hopefully) would incorporate it into their own download programs.

  • The Ambulatory Glucose Profile (AGP), developed by Dr. Roger Mazze and refined by Dr. Bergenstal; Ellie Strock, NP; David Wesley; and other colleagues at the International Diabetes Center, is designed as a standardized, simplified “title page” for every CGM’s data-reviewing software: “a common platform for simple discussions.” (Dr. Bergenstal noted that every one of the dozens of EKG manufacturers display their data in a common format, which makes EKG technology on the whole more usable for clinicians and patients.) The AGP’s main graphic is a modal day that aggregates all recent CGM data into five traces that represent the median, the boundaries of the 25th and 75th percentiles, and the boundaries of the 10th and 90th percentiles. Because these lines do not cross and are easy to distinguish from one another, glycemic patterns can be analyzed quickly by HCPs (and patients, Dr. Bergenstal noted). Above this modal day graphic are a series of numerical statistics (e.g., percentage time in range), and below the modal day graphic are thumbnails of each individual day’s CGM trace (which expand if the user hovers over his or her cursor over them). The Helmsley Charitable Trust is supporting development of capturAGP, a software version of the AGP that could be integrated into any CGM company’s proprietary software. We hope for the sake of patients and HCPs that the AGP (or a similar system) does indeed become the industry standard for the first page of downloads; we think that this would simplify HCP and patient training, allow CGM companies to devote their innovative efforts elsewhere, and reduce the precious visit-time that gets spent on pattern analysis. For more background on the International Diabetes Center’s and Helmsley Charitable Trust’s work on this front, see our coverage of an expert meeting dedicated to glycemic data display in March 2012:
  • Noting the need for common definitions of glucose control as measured by CGM, Dr. Bergenstal proposed his own “best attempt” to define various relevant ranges. He defined time in range as 70-180 mg/dl (though he hopes that as glucose control becomes more safe and widespread across the patient population that the widely accepted definition of good control will change to 70-140 mg/dl – we hope so too!). Sensor values above 400 mg/dl correspond to severe hyperglycemia, values >250-400 mg/dl are moderately hyperglycemic, values >180-250 mg/dl are mildly hyperglycemic, values <70-60 mg/dl are mildly hypoglycemic, values <60-50 mg/dl are moderately hypoglycemic, and values <50 mg/dl are “potentially severe hypoglycemia” or a “dangerously low glucose level.” The definitions of DKA and severe hypoglycemia would be based on established clinical criteria rather than any particular sensor values.


Questions and Answers

Q: Is the software that you showed proprietary? Can it be used with any CGM?

Dr. Bergenstal: It won’t be proprietary once it’s finished. With Helmsley Charitable Trust support, we will work with manufacturers of both CGM devices and SMBG devices to incorporate an AGP into their glucose monitor device outputs. It is really to develop a common platform for simple discussions.

Dr. William Tamborlane (Yale University, New Haven, CT): I think that most of us agree that 70-180 mg/dl is the target range and that it would be nice to get the 180 mg/dl threshold lower. The ADA has defined <70 mg/dl as being hypoglycemia. In our research with CGM, it is not uncommon for non-diabetic individuals to have sensor glucose in the range of 60-70 mg/dl. I would say that 60-70 mg/dl is nothing – neither in target nor hypoglycemic.

Dr. Bergenstal: I agree with you. I know that 70 mg/dl is valid for venous blood, but perhaps not for sensors. This is why with the AGP we display data below three different thresholds: 70, 60, and 50 mg/dl.

Q: One bit of data that I get from CGM and not from fingersticks is rate of change. If the glucose is falling more rapidly, can I use that additional information?

Dr. Bergenstal: Yes. Dr. Tamborlane, let’s say you have a particular numerical glucose value – how do you change your interpretation based on the trend information?

Dr. Tamborlane: Bolus calculators use carbs and static glucose – CGM trend data is a third dimension for bolus calculators. We devised an algorithm to adjust bolus calculators’ recommendations by 10-20% based on the rate of change.

Dr. Bergenstal: But all of that will be put to rest if Dr. Kowalski has his way and we put some brains into the pump.


Aaron Kowalski, PhD (Assistant VP, Treatment Therapies, JDRF)

Dr. Aaron Kowalski gave a valuable update on progress toward systems for closed-loop glucose control. The next step beyond Medtronic’s low glucose suspend (LGS) system (currently under FDA review) will be Medtronic’s predictive LGS system (available within the next year or two, Dr. Kowalski hopes). In a recent pilot study by Dr. Bruce Buckingham, a “home-grown” version of a similar system was reported to reduce nocturnal hypoglycemia by ~50%, and prolonged nocturnal hypoglycemia by ~70%. Meanwhile, control-to-range insulin delivery systems are moving toward large outpatient studies, while bihormonal control also has great promise, pending the resolution of some obstacles (e.g., the lack of stable soluble glucagon or a dual-chambered pump). In Dr. Kowalski’s view, the biggest obstacle to fully closed-loop control with current tools is the slow on/off profile of subcutaneous insulin, and another major (underappreciated) problem is the dysregulation of glucagon in people with type 1 diabetes; fortunately, several research projects are underway to address both of these challenges.

  • Pump manufacturers do not currently have commercial incentives to develop dual- chambered pumps designed for bihormonal closed-loop systems; Dr. Kowalski stated that to address this problem, JDRF hopes to fund one company to build a dual-chambered pump, in the next few months. We are extremely hopeful for this deal to go through so that the field of bihormonal control can be pushed forward – especially since the development process for such a pump is likely to be long because of regulatory issues, as Dr. Kowalski noted.
  • Dr. Kowalski’s said that the biggest obstacle to true closed-loop glucose control is the slow onset and slow offset of current subcutaneous insulins. Potential improvements include novel formulations or insulin analogs and the InsuPatch (InsuLine’s tool to heat the subcutaneous infusion site). Looking outside of subcutaneous delivery, Dr. Kowalski mentioned intradermal microneedles. BD has developed an operative prototype infusion set that lasts for multiple days; Dr. Kowalski noted that the set’s faster uptake may involve the lymphatic system. Dr. Kowalski also mentioned that patients from clinical trials of Roche’s Accu-Chek DiaPort “speak about it religiously” due to the greater responsiveness that intraperitoneal insulin enables. He reminded the audience that the delivery system is currently being studied for closed-loop control.
  • Several drugs are being studied to manage the glucagon dysregulation that occurs in type 1 diabetes, which is not addressed by current insulin-only delivery systems. (For example, people with type 1 diabetes have significant glucagon spikes after eating, as opposed to the glucagon suppression that occurs in people without diabetes.) Dr. Roger Unger of UT Southwestern, who controversially believes that glucagon dysregulation is the primary defect in type 1 diabetes, is starting to perform clinical trials of leptin for people with type 1 diabetes. Though leptin has performed well in Dr. Unger’s preclinical research, Dr. Kowalski said that he personally believes amylin deserves more attention. (As a reminder, amylin is normally co-secreted with insulin but is absent in type 1 diabetes; its effects include increasing satiety and inhibiting glucagon secretion.) The synthetic version of amylin (BMS/AZ’s pramlintide) is difficult to dose with subcutaneous injections, and it comes with risks for GI side effects and hypoglycemia. However, Dr. Kowalski believes that pramlintide could be quite beneficial if dosed more physiologically. (He mentioned that an endocrinologist in Philadelphia instructs some of his patients to wear two pumps simultaneously, one for insulin and one for pramlintide.) The artificial pancreas researchers at Yale are studying amylin as a potential adjunct to closed-loop control, and they plan to study liraglutide in closed-loop control as well (because GLP-1 agonists also inhibit glucagon secretion).
  • Dr. Kowalski concluded by asking for audience members to contact their Congressmen and -women about a $150-million NIH special appropriation for diabetes research; because this is a special appropriation, it will disappear unless it is renewed during the “lame-duck sessions” of the next two-to-three months. He urged people to visit the ADA and JDRF websites to learn more and add their voices to those already supporting the appropriation’s renewal. “Everyone here should be making those calls,” Dr. Kowalski said: “We all benefit from it.”

Questions and Answers

Q: Will patients still need to do fingerstick monitoring? How frequently?

Dr. Kowalski: In this regard we expect the treat-to-range system to be very similar to what patients are doing now. The idea is, “Can we pick up the slack for when people are missing the boat now?” I think that calibration will be very important. I generally use my CGM and only calibrate it [i.e., perform minimal confirmatory fingersticks]. That is off-label, obviously. The mathematicians designing closed-loop algorithms are very good at detecting bad sensor readings, but if you calibrate based on a bad blood glucose reading, then that is a problem. You will need good calibration. I don’t think there will need to be much calibration, though: maybe two-to-three times a day.

Dr. William Tamborlane (Yale University, New Haven, CT): One area that still needs a lot of attention is the patient-system interface. For example, the closed-loop system might accept only calibrations that are directly transmitted by RF from the actual glucose meter. This would be to address the concern that an adolescent might just make up a number and use that as the calibration, rather than performing a fingerstick. At that level of granularity we must try to prevent injury to patients due to insulin over-delivery.

I would add to Dr. Kowalski’s point that glucagon regulation is important to address. In people without diabetes, glucagon secretion is stimulated by pure protein feeding and suppressed by pure glucose feeding. Glucagon secretion at meal times is much higher in type 1 diabetes. In a recent DirecNet study of children recently diagnosed with type 1 diabetes, we saw a rise in glucagon with mixed-meal feeding equal to that seen in non-diabetes during a hypoglycemic clamp study. I think we can get around the problems associated with pramlintide by using Bydureon or liraglutide. We are very interested in this area; we are going to do a study of closed-loop control with liraglutide.

Special Populations/Global Perspective

Plenary Session I


Jean-Claude Mbanya, MD, PhD (President, International Diabetes Federation, Yaounde, Cameroon)

Dr. Jean-Claude Mbanya gave a rousing speech calling for members of the audience to ensure that the children and grandchildren in their communities have a better future. He reminded the audience, however, that the rising tsunami of type 2 diabetes is not only an issue of wealthy countries. There are currently 371 million people with diabetes in the world, 80% of whom are in low or middle income countries. Dr. Mbanya explained the prevalence of diabetes in these countries by reminding the audience that there are two faces of diabetes: over nutrition and under nutrition – an oft forgotten cause. Additionally, 50-80% of those who have type 2 diabetes in low- or middle-income countries are undiagnosed. As a result, by the time they are diagnosed they are already facing severe, and expensive, complications. Strikingly, in Cameroon – where Dr. Mbanya is from – it costs about $500/person to treat diabetes, which, despite being less than half the world median of $1,270/person, is still 1.5 times greater than the country’s GDP and 50-fold larger than the mean health expenditure per person. Dr. Mbanya effectively conveyed the large opportunity cost of diabetes in low- and middle-income countries by describing how difficult it was for the family of one of his pediatric diabetes patients to pay for the child’s healthcare. In order for the family to cover the child’s care they had to stop sending the daughters to school, and when the child died the family expressed some relief as they felt it was better for the child to die so that the rest of the family could live. Dr. Mbanya also described how in some communities families will try to hide a daughter’s diabetes, because they fear people will not want to marry her if they learn she has the disease. Finally, Dr. Mbanya said that professionals in the diabetes field have to set the example. He expressed his disappointment that the Second Global Diabetes Summit served high-fat and high-sugar pastries for breakfast, saying that he expects such meetings to serve fruits and vegetables (there were some, of course, but not enough). We note that the majority of conferences we attend provide high-fat and high-sugar options, and that we would love to see Dr. Mbanya’s request enacted.

  • Dr. Mbanya called for the promotion of lifestyle changes, and questioned why the current obesogenic environment is not being more rapidly modified when data has shown that primary prevention works. He reminded the audience that trials around the world – “from Finland to China” – have demonstrated that solutions to the obesogenic environment exist. Additionally, these trials, he remarked, have shown that primary prevention is possible; lifestyle changes prevent new cases of obesity and type 2 diabetes. Furthermore, thanks to the body’s metabolic memory the effects of prevention are long lasting, as the benefits of lifestyle changes are seen years after the trials that enacted them are completed.
  • In particular, Dr. Mbanya emphasized that we must be tackling the issue of childhood obesity. He argued that there are easy ways to prevent childhood obesity, including increasing the amount of physical activity children get during the school day and the healthiness of the meals they eat. Dr. Mbanya explained that the IDF selected “Diabetes: Protect our Future” as the theme for this year’s World Diabetes Day, because while it might be too late for some adults to avoid type 2 diabetes, it is not too late for their children and grandchildren. He called for the audience to ensure that later generations do not make the same mistakes adults have.

Plenary Session IV: International Diabetes Forum


Naomi Levitt, MD, MB, ChB (University of Cape Town, Cape Town, South Africa)

Dr. Levitt discussed drivers of the global increases in obesity and diabetes. While urbanization is strongly tied to increased prevalence of obesity and diabetes, she noted that diabetes prevalence is also on the rise in rural areas in all parts of the world. Globally, a staggering 1.5 billion people are now either overweight or obese. Dr. Levitt identified urbanization, wealth, food economics (the most obesogenic foods are the cheapest), and decreasing physical activity (due to mechanization of transportation and occupational activity) as drivers of increasing global obesity. Additionally, she highlighted several factors that were more specific to developing countries. In some developing countries, the cultural standard of beauty favors heavier bodies because obesity is associated with affluence and fertility while thinness is associated with HIV/AIDS. Additionally, it has been found that for women, early life experience with hunger is associated with adult presence of obesity. Furthermore, low birth weight and childhood undernutrition, especially if accompanied by future overnutrition, is associated with increased risk of impaired fasting glucose and diabetes. Dr. Levitt argued that the most important thing we can do to curb rising obesity and diabetes is to attack obesity at the population level by changing policy – namely, she called for regulation of the food industry in developing countries where sales of soft drinks are still steeply rising. She concluded that both individuals and governments have the obligation to create an environment promoting health and wellbeing for future generations.

Questions and Answers

Q: What can governments do at a global level with regard to the multinational food industry?

A: I think governments can do a great deal. It depends on whether one wants to focus on children. If we want to focus on them, a number of policy measures could be implemented. The first relates to schooling, physical activity in schooling, opportunities for activity, foods at schools, and subliminal advertising that takes place could all be addressed at the policy level. Additional policy could relate to food pricing. If we think about it, the cheapest foods are sugars and rice. There’s good data showing that with increasing consumption of soft drinks diabetes rises. As people consume more rice and more calories they also increase their risk of diabetes. Cheap foods are the most obesogenic foods. What we need to do is create policy whereby there is much more subsidization of healthy foods rather than unhealthy foods.

Workshop: Diabetes in Special Populations – Continental East Asians


Abhimanyu Garg, MD (Southwestern Medical Center, Dallas, TX)

Dr. Abhimanyu Garg presented data suggesting that currently known genetic factors do not explain why Asian Indians have a higher prevalence of type 2 diabetes compared to Caucasians. Instead, this increased prevalence could be explained by Asian Indians’ predisposition to increased subcutaneous truncal obesity and insulin resistance, along with the low birth weight and childhood undernutrition frequently encountered in India. This assertion was interesting to us since we more commonly hear that visceral abdominal fat, rather than subcutaneous abdominal fat, is most closely associated with insulin resistance. Additionally, Dr. Garg presented data demonstrating that, in a euglycemic hyperinsulinemic glucose clamp study, even at the same level of truncal skinfold thickness, Asian Indians were more insulin resistant than Caucasians.

Questions and Answers

Q: Those guidelines for BMI recommendations, are they generally accepted? In my office we are still using the guidelines for Caucasians.

A: I don’t know whether they’re accepted in the US, but clearly in India and in Asia they are used. Generally, in India, people are recommending a BMI below 23 kg/m2.

Q: I have a question about risk increasing when a child is born underweight and then gains weight. Earlier there was talk about an in utero period of three years – is that a peak time when the metabolic programming occurs?

A: So they are looking at the birth weight and onwards, but maybe they’re right that epigenetic changes occurring during pregnancy may be important. We haven’t gone back that far. So it is possible that some metabolic programming occurs, and that’s what we’re supposed to live with. And if there’s more of a load later on, the body can’t take it.

Q: So you looked at abdominal fat in internal organs as well? Like liver fat?

A: Not liver fat. In some earlier studies I published, we did this first in Caucasians. What we found, as opposed to what other people have said in the past, was that intraperitoneal fat wasn’t the best predictor of insulin resistance as measured by a clamps study. In fact, subcutaneous fat on the abdomen was a better predictor. We didn’t just take just one slice of the abdomen, but we took several. So we quantified the entire intraperitoneal fat and subcutaneous fat, and subcutaneous fat had the best regression. Even skin-fold thickness was better correlated than intraperitoneal fat. Obesity itself is a question of what are you predisposed to.

Q When do you see acanthosis nigricans (AN)?

A: I told you what they’re looking at in these studies; they probably didn’t even look at AN at all. But this is nonetheless relevant for people who deal with Asian Indians, and remember, the second generation of Asian Indian children are quite obese and may be developing all sorts of insulin resistance and changes like that.

Q: I have a clinical question on managing Indians – many are vegetarian or just don’t eat meat very often and have a very high carb diet. I see a lot of men and women predominantly in their thirties who are very sedentary, in the software industry with low HDL, high triglycerides, and are not obese by US standards. My question is based on their increased insulin resistance, what would you advise on their diet? How do you increase proteins and vegetables?

A: So what you’re saying is they have a vegetarian diet, but that’s not synonymous with high carb, because we eat a lot of cheese and oils. So they may also be eating a high fat diet. What is more important is activity. The only thing is if you ask a person who is sedentary to exercise he’s not going to lose weight. His fat will go down, muscle mass will increase, but weight will remain same. They will also have to eat less to lose weight. I usually don’t prescribe a low carb, low fat, or high protein diet; I just tell them work on total energy intake. Reduce portions and see what the results are. But the critical thing is activity. They have to pick up some activity or other.

Comment: I read a couple studies recently that rice alone will increase risk of diabetes with all other risk factors being similar.

A: Again when we blame rice we have to realize that for centuries the Chinese and other people have been eating a rice-based diet. Poor Chinese people who didn’t have any meat were eating mainly rice-based diet and didn’t have diabetes. We have to realize that when we start criticizing a particular nutrient in the diet; we have to look at it from all angles and see what the real problem is.

Q: Do you have any specific class of treatments like metformin or TZDs, or any special recommendations for Indian patients? Should we go with the BMI cutoffs of the Indian standard or the US standard?

A: I don’t have any specific insights into how to treat Indians differently than Caucasians. My feeling would be that, as we have discussed, we can start with metformin and then add a second agent; my bias is always sulfonylurea. What you heard from Ralph is Ralph’s opinion but sulfonylureas are cheap and I think what he said was, “Kwame you put three or four people on the podium, and we are all talking about the same language. You should have put someone on the podium who has another different opinion.” I still like sulfonylureas. When I look at the data that have been collected from various studies, the best data were from UKPDS, there was no harm with SFU. The second best data were from the trial in Australia – ACCORD here and another study in Australia. In ACCORD if you look at what drug we used in the intensive vs. control arms, there was a lot of TZD. In Australia, the use of SFUs was very light, and if anything they saw a reduction in mortality, but not in ACCORD. The same treatment algorithms will apply. The other thing I’d say about metformin and sulfonylurea - we’ve used these two for a long time. These newer drugs we don’t know very much about. I’m comfortable with drugs I’ve used. Unless someone says we missed the fact that sulfonylureas were causing lung cancer, then I’d stop.


Abhimanyu Garg, MD (Southwestern Medical Center, Dallas, TX)

Dr. Shashank Joshi was not able to attend the conference, so Dr. Abhimanyu Garg presented results from the Indian Diabetes Prevention Programs 1 and 2 (IDPP-1 and IDPP-2). In IDPP-1, about 500 patients with impaired glucose tolerance were randomized to control (standard health care advice), lifestyle modification (LSM), metformin, or LSM plus metformin. Mean baseline A1c was 6.1-6.2% and mean baseline BMI was ~26 kg/m2, across treatment groups. After three years, a strikingly high 54% of those in the control group progressed to diabetes compared to about 40% of those in the LSM, metformin, and LSM plus metformin groups. There was no added benefit of combining lifestyle and metformin. For context, ~30% of participants in the placebo arm of the US DPP progressed to diabetes after ~three years, and ~50% had progressed after ten years – this corroborates a statement made by Dr. Shashank Joshi during the 2012 Keystone Conference indicating that Asian Indians progressed from prediabetes to diabetes three to four times faster than Caucasians (we are still waiting on a reference for this statistic; for more detail, please see our Keystone 2012 full report at Making direct comparisons to the US trial and generalizations to the whole Indian population is a bit tricky – the LSM program was not as rigorous as that of the US DPP, and on average, participants in the intervention groups did not actually lose weight from baseline. In the IDPP-2, 407 patients were randomized to pioglitazone plus LSM or placebo plus LSM. Mean baseline BMI again was ~26 kg/m2 and mean baseline A1c was 5.8%. In both groups, about 30% of patients progressed to diabetes after three years; thus, pioglitazone did not confer any additional benefit over lifestyle modification for reducing risk progression from impaired glucose tolerance to diabetes. Dr. Garg noted that both the IDPP-1 and IDPP-2 had to use lower than maximal doses of metformin and pioglitazone (500 mg/day and 30 mg/day, respectively) because patients that were initiated on higher doses experienced symptoms of hypoglycemia (excessive hunger and dizziness). Therefore, it was concluded that Indian participants with impaired glucose tolerance might be more sensitive to metformin and pioglitazone.

Questions and Answers

Q: Would you expect different responses to metformin at a lower BMI?

A: Metformin at end of day causes just about 5 lbs of weight loss over 2-3 years. It’s not a great weight- reducing drug. People still use it. I’ve seen a lot of physicians use it for patients with IGT or someone who doesn’t like their body shape. I’ve also seen people use liraglutide or exenatide in normal, healthy non- diabetic women as a weight loss drug. These are not very powerful weight loss drugs anyway.

Q: The mean baseline A1c was 5.8%, so definition-wise they were not diabetics. But then we were making comparisons to see if A1c was coming down, and it didn’t come down. Have any studies seen reductions in A1c below 5.8% in people with just impaired glucose tolerance?

A: ADOPT would be the one, but those patients had diabetes not IGT. I’m trying to remember. There may have been a study with rosiglitazone and IGT. The critical question still remains unanswered, which is whether, when I stop that drug, the improvement will go away or not. Nobody has answered that critical question. If you are calling it prevention of diabetes and I have IGT and you tell me to take 3 years of pioglitazone so that I won’t develop diabetes, maybe I’d listen to you and do that. But if after three years I stop the drug and I develop diabetes, that critical question hasn’t been answered.

Q: I am wondering about the hypoglycemia on metformin and pioglitazone – was it documented? I assumed they tested and they were truly low. Or was it just symptomatic? Any thoughts about why they would have been low with metformin?

A: Yes, people say metformin doesn’t cause hypoglycemia. I don’t know whether they had given them glucometers to check their sugars when they were dizzy, but at least you can say there is something there.

I don’t think they would have given them glucometers to document, but I think there is something there to suggest metformin given to people with impaired glucose tolerance may cause hypoglycemia.

Comment: In the Indian diet, the distribution of carbohydrates toward the evening meal is higher compared to many western diets. And Indian families tend to eat very late. So those might be two factors for the increased prevalence of diabetes.

Q: The IDPP-2 was based on a pioglitazone study previously conducted in the US, and they had very different results in the States. The conclusion was that pioglitazone significantly reduced development of diabetes. So would you actually attribute this to ethnic differences?

A: That’s a possibility. Some people tell us that TZDs may work better in people who are more obese since they are PPAR-gamma agonists and increase differentiation of fat, and make people gain weight and fluid. I don’t know whether they have done a study looking at lean and obese people with diabetes to compare whether it is more efficacious in obese or lean patients. That would tell us at least if it doesn’t work as well in leaner patients. The other thing is that we always like to have ratification of studies. If two studies show the same results then your confidence in that particular drug or intervention is much more certain.

Workshop: Diabetes in Special Populations – Diabetes in Latinos


Enrique Caballero, MD (Joslin Diabetes Center, Boston, MA)

Dr. Enrique Caballero shared experiences and lessons from the Joslin Diabetes Center’s Latino Diabetes Initiative. He noted that he always invites a family member to patient visits, since diabetes is a family issue; for example, if he asks a mother to improve her meal plan, he knows that she will need to get buy-in from her husband and children. Another important aspect of care is to have frank conversations that address myths and misconceptions; Dr. Caballero told a terrifying story about one patient who had let his diabetes get far out of control because “I don’t want to go blind from taking insulin.” He also emphasized the importance of cultural awareness in all patient interactions. As an example, he told a story about a patient who nearly died from an overdose of an antihypertensive medication because her physician had written a prescription as “once/d”; once means 11 in Spanish.

Symposium: Diabetes and Youth

COPE TEEN (thinking, emotions, exercise, and nutrition: an intervention to promote cognitive behavior skills and healthy lifestyle behaviors in overweight adolescents

Bernadette Mazurek Melynk, PhD, CPNP, PMHNP, (The Ohio State University, Columbus, OH)

Mental health is a critical and underappreciated factor that limits the success of behavior change, Dr. Bernadette Mazurek Melnyk emphasized. Moreover, symptoms of depression and anxiety are higher in people (including adolescents) who are overweight or obese. To improve both the mental and physical health of overweight teenagers, Dr. Melnyk and her colleagues developed a 15-week program of weekly, group sessions (COPE TEEN). Each 50-minute session included 20 minutes of mild physical activity. Notably, for the first seven sessions, the program’s educational material focused entirely on cognitive behavioral skills; health and nutrition training did not come until the last eight sessions. The program was found successful in three pilot studies, which led to an NIH-funded randomized controlled trial that is currently underway in Arizona (n=779 teens aged 14-16, mean BMI 24-25 kg/m2). The researchers have collected 12-month follow-up data in two cohorts so far; preliminary analysis suggests that students who received COPE TEEN achieved lower mean BMI, higher average scores in assessments of social skills, higher grades in school, and less use of alcohol and marijuana. We think that this study is one of the world’s most important for the future for diabetes care, and we emphatically agree with Dr. Melnyk that mental and emotional factors deserve much, much more attention among clinicians and researchers. 

  • Speaking to her fellow researchers interested in targeting social and behavioral changes, Dr. Melnyk emphasized that new interventions will be adopted only if they can be shown to improve the factors that are already of interest to the healthcare and/or education establishments. Dr. Melnyk recalled the first application of COPE, in the parents of premature children. Through nine randomized controlled trials, the program was shown to reduce parental depression and PTSD, and to improve parent-child interactions. Nonetheless, no hospital adopted the program until Dr. Melnyk published findings that the program reduced hospital stays for premature babies by four-to-eight days (which could translate into $2.5-to-$5 billion in annual savings nationwide). Having learned from this experience, Dr. Melnyk designed the endpoints of the COPE TEEN randomized controlled trial to include academic grades, so that schools would be more likely to adopt the program if it is shown to be successful.


Francine Kaufman, MD (Medtronic and Children’s Hospital Los Angeles, Los Angeles, CA)

Dr. Francine Kaufman reviewed: 1) the presentation, diagnostic criteria, and recommendations for screening for type 2 diabetes in children; 2) causes of type 2 diabetes in youth; 3) treatment strategies; comorbidities and complications; and 5) strategies for prevention. Children who develop type 2diabetes typically have multiple risk factors compounded by living in an obesogenic environment – Dr. Kaufman stated that it is almost “predestined” that these children develop the disease given the environment in which they live. Dr. Kaufman also touched upon the approach that should be taken for screening; she indicated that attempting to screen in a public health setting, outside of the hospital, would be inefficient as the prevalence of type 2 diabetes in children is extraordinarily low: in baseline information collected on high-risk 6th graders for the HEALTHY trial (a youth diabetes prevention trial), asymptomatic diabetes was essentially not found in the 6,367 participants. Therefore, she stated that screening should be limited to the context of health care visits in patients presenting with symptoms and meeting criteria established by the ADA/American Academy of Pediatrics Consensus Statement. Dr. Kaufman also stressed the need to focus on comorbidities and complications, stating that 92% of youth with type 2 diabetes have two or more risk factors for cardiovascular disease and that this prevalence of risk factors is typically higher than is seen in the type 1 diabetes population. Lastly, and most urgently, Dr. Kaufman stressed the need to address the obesogenic environment the breeds childhood type 2 diabetes, highlighting several points of intervention where audience members could make a difference including: community building, the food and beverage industry, the media, and schools and child care. Notably, she highlighted hospitals’ hypocrisy in offering candy and soda in their gift shops, stating that the hospital environment certainly needs improvement.


Sonia Caprio, MD (Yale School of Medicine, New Haven, CT)

Dr. Sonia Caprio described how fatty liver disease might prelude the development of the metabolic syndrome in obese adolescents. She noted that though obesity is important for the onset and rise of risk factors for the metabolic syndrome, the location of excess fat is probably one of the leading causes, if not the leading cause, of insulin resistance and subsequently, the metabolic syndrome. Having excess fat deposits in the liver is particularly detrimental. As evidence of this, she presented MRI images of the torso of two different obese adolescents who both had a BMI over 35 kg/m2 and had similar amounts of visceral fat. One, however, had more fat deposits in the liver than the other. Dr. Caprio found that the individual with a fatty liver had higher triglyceride levels, lower levels of adiponectin, and greater insulin resistance. However, Dr. Carpio explained during the panel discussion that this finding currently has limited clinical applicability because liver biopsies are invasive and risky, and a fatty liver diagnosis does not translate to better therapy.


Silva Arslanian, MD (University of Pittsburgh, Pittsburgh, PA)

Dr. Silva Arslanian reviewed the progression of type 2 diabetes in children and adolescents with a focus on the many, poorly understood pathophysiological differences between people who are black and white. Compared to white children, black children tend to have higher rates of diabetes progression, lower rates of lipolysis (i.e., low ability to mobilize fat stores), higher insulin resistance, and much- higher insulin secretion (an “overcompensation” for the insulin resistance, according to Dr. Arslanian’s understanding of causality in diabetes progression) – all despite lower values of commonly accepted risk markers like LDL cholesterol and visceral adiposity. Related to these biological differences are socio-cultural trends, such as black children’s overall higher intake of sweets and fats and, for girls, a greater self-reported desire to be “on the heavy side.”


Silva Arslanian, MD, PhD (University of Pittsburgh, Pittsburgh, PA); Sonia Caprio, MD (Yale University, New Haven, CT); Francine Kaufman, MD (Children’s Hospital Los Angeles, Los Angeles, CA); David Repaske, MD, PhD (Nationwide Children’s Hospital, Columbus, OH)

Q: Joseph Shivers (Sansum Research Institute, Santa Barbra, CA): Dr. Arslanian, you presented data on how insulin secretion is higher in black children then white children. I am wondering if the higher insulin secretion precedes insulin resistance rather than the other way around.

Dr. Arslanian: It may be a valid theory, but it is very hard to test it. I think that to address that question one has to start soon after birth to see what comes first, the chicken or the egg.

Q: Dr. Kaufman you showed data on microalbuminuria levels in children with type 2 diabetes, I was curious whether we should be screening these children for retinopathy also?

Dr. Kaufman: Definitely, they should be screened. The guidelines are a little different than for children with type 1 diabetes. We adapted the screening for adults with complications so children could get it earlier. And even children with type 1 would need to wait five years. The rates in Australia show more retinopathy in type 1 than type 2 youth, but we should screen for all diabetes complications in type 2.

Q: Dr. Arslanian, you presented great data on African Americans and whites. My understanding is that Hispanics beat both of them for type 2 diabetes in youth. Do Hispanics behave more like whites or African Americans?

Dr. Arslanian: I think they behave more like Hispanics [Laughter]. Though the rates are higher in Hispanics their body composition is very different. They have much higher rates of liver fat. It is intriguing that you have two minority populations that are both insulin resistant, but with different phenotypes. The difficulty with pediatric populations is that we cannot do liver biopsies and muscle biopsies. So I am not sure if we are ever going to have answers to these questions.

Q: I think you convinced me that African American females bear a large burden of type 2 diabetes amongst youth. The question is, and I may be wrong but, don’t they get puberty a little earlier? When you are doing these studies are you excluding any hormonal contraceptives?

Dr. Arslanian: Absolutely yes.

Q: But is there a rule of hormonal contraception? I know some teenagers are put on medroxyprogesterone for irregular cycles or contraception. Does that change the incidence of diabetes amongst them?

Dr. Arslanian: Fran, Sonia do you know? All of our studies are done without contraceptions, it is an exclusion.

Q: It seems to me that differences in metabolic activity could be explained by different types of fat. Are you dealing predominantly with brown or white fat in these studies? Especially in the studies in children. We know there is a difference in adults.

Dr. Caprio: We have not measured brown fat – those techniques are very expensive and complicated. It’s done with PET and you have to infuse something radioactive, so we’re not allowed to do that in children. What we do is look at expression of UCP1 in adipose tissue. Even there I don’t have great numbers so, right now we have no clue at all.

Q: Dr. Caprio, you said that a fatty liver is a significant risk factor for the metabolic syndrome. What is the approach for utilizing this information clinically?

Dr. Caprio: We look for markers and if those come back positive then we send the patient for an echo. In my opinion I would stay away from doing a liver biopsy. It is very invasive and it is risky, and once we know they have NASH we don’t have a treatment. Are we just going to biopsy them again and again?

Dr. Repaske: The best treatment for NASH is weight loss.

Q (Dr. Kaufman): In the HEALTHY trial we asked children what their race or ethnicity was, and there was a group that had no idea. It was especially difficult in the Hispanic group. A lot would say, “You look at me and you decide – my mom’s this my dad’s that.” What’s the future? If we all became one great beige color then what?

Dr. Arslanian: For 20 years now we have done it based on self-identity over three generations, but mixing is increasing; 15 to 20 years ago it was much easier for us to find three generations of blacks that identified as black. Now it’s harder and harder because every time someone says my grandfather is white or somebody is Hispanic, so it’s getting difficult. Even genetically there are studies showing that even in those identifying as pure black, there are 30% white genes in the mixture. Those are the stairs behind the kitchen.

Q: Dr. Caprio, you showed us very interesting MRI images of the abdomens of two individuals. You said that the subcutaneous fat was markedly reduced in one of the subjects compared to the other one. Are we looking at subtle lipodystrophy in these subjects?

Dr. Caprio: The paper those images are from states in its title that they have partial lipodystrophy. Either their lipogenesis is down regulated or the formation of fat in the superficial layer is down regulated in this phenotype. Silva has shown you that there is also decreased lipolysis, so something is different there.

Q (Dr. Repaske): I’ve got another practical question for you, Fran, your first step in deciding treatment was to test for antibodies. What do you do with a blood glucose of 250 mg/dl before the antibodies comes back?

Dr. Kaufman: So you get the antibodies, and if you think its someone that could have type 1, you start them on insulin. If you’re gutsy, and they’re obese, not that dysglycemic, with a family history, you could put them on metformin. Most of the time I would start on insulin, get everything stabilized, and then readjust.

Workshop: Diabetes in Special Populations – Blacks and West Indians


Ayesha Motala, MB, ChB (University of KwaZulu-Natal, Durban, South Africa)

Dr. Ayesha Motala described the current state of type 2 diabetes and the metabolic syndrome in South Africa, particularly amongst the Black population. She explained that while diabetes used to be very rare on the African continent it has become a fertile ground for the diabetes epidemic, because of its triple burden: resource depletion, communicable diseases, and non-communicable diseases. In contrast to many regions of the world, Africa is not only facing a crisis in terms of growing diabetes and obesity rates, but is also still suffering from extremely high rates of communicable diseases, like malaria and HIV (in 2009 the 5.6 million cases of HIV in South Africa accounted for 68% of all cases of HIV; the country also had 69% of all new infections and 72% of all AIDS deaths). She described how the South African health system is already overwhelmed by communicable diseases and struggles to address diseases like diabetes and obesity. Additionally, Dr. Motala stated that many people depressingly believe that adults in South Africa will die from HIV/AIDS before they can become diabetic. However, research has shown that despite the impact of AIDS deaths on population growth, the total number of people with diabetes is expected to continue to increase.

Symposium: Immunology of Type 1 Diabetes


Kevan Herold, MD (Yale University, New Haven, CT)

Dr. Kevan Herold began his review of immunotherapies to preserve beta cell function by somberly remarking that the field probably would not be where it is today without Dr. George Eisenbarth, who passed away only three days earlier. Dr. Herold characterized microbiome approaches to beta cell preservation as being unproven, but still an important area to study. He then hypothesized that the reason antigen-specific therapies (with the exception of oral insulin) have been unsuccessful is that once a person is diagnosed with type 1 diabetes, that individual has many different autoantibodies, which he doubts a single antigen can effectively address. Turning to specific immunological trials, Dr. Herold suggested that if the initial stricter inclusion criteria had been used for the DPT-1 trial, then oral insulin probably would have been found to be significantly efficacious. Similarly, he highlighted that while the Protégé trial did not meet its primary endpoint, the high dose of teplizumab was found to preserve C- peptide levels significantly better than placebo. It is very challenging to hear that potential treatments for type 1 diabetes like teplizumab and oral insulin might be more widely available today if trials had been better designed, given that the current market offers few second chances to test a therapy. We hope that lessons learned from these trials will be applied to future studies, such that we can celebrate their findings and not regret their design. Looking to the future, he suggested that research should focus on selecting the right patients for a therapy, designing combination therapies, and improving tolerance. Additionally, he recommended treating people who are at high risk (~90%) for type 1 diabetes as if they already have the disease, since the diagnosis guidelines are based on microvascular risk, not the disease’s pathology, and early treatment could prevent or slow disease progression. Finally, Dr. Herold passionately asked providers to refer eligible patients to a trial on teplizumab use for the prevention of type 1 diabetes (a rare “second chance” for teplimuzab); he predicts that teplizumab will significantly delay, if not prevent, the development of diabetes ( Identifier: NCT01030861).

  • Dr. Herold acknowledged that the search for a cure for type 1 diabetes has been difficult (both emotionally and intellectually), and that it will probably remain challenging. He then reminded the audience of what his mother wisely told him to not forget: type 1 diabetes has been around for thousands of years, during which time it has had a chance to evolve. In contrast, humans have tried to intervene to prevent the disease only relatively recently.
  • When people are diagnosed with type 1 diabetes they probably still have 30-35% of their normal beta cell function. (Dr. Herold told the audience that this is in contrast to the 90% loss of function that he had been taught during his medical training.) He continued to explain that retaining this beta cell function is important because if people have more than 0.2 C- peptide pmol/l, then they have significantly lower long-term risk for complications.
  • He hypothesized that if the DPT-1 had maintained its original strict inclusion criteria, then it would have found oral insulin to be significantly efficacious. Dr. Herold explained that in an attempt to increase enrollment, the inclusion criteria was made more flexible, and that doing this might have diluted any effect that would have been seen with the original trial design.
  • Dr. Herold detailed three biologics he believes have been successful: 1) abatacept (anti-CTLA4 antibody), 2) teplizumab (anti-CD3 monoclonal antibody), and 3) rituximab (anti-CD20 antibodies). All three therapies initially had statistically successful preservation of beta cell function. Eventually beta cell deterioration occurred for all three, even when there was continued administration of the antibody. Dr. Herold called this the “type 1 diabetes problem” and proposed that it is due to a failure of these therapies to modulate or eradicate pathogenic cells and/or a preprogrammed death pathway being present in the beta cells of people with type 1 diabetes that cannot be modulated by an immune therapy.
  • Abatacept has been found to delay C-peptide decline by ~9.6 months. Additionally, the proportion of subjects who retained at least 0.2 pmol/ml C-peptide was significantly higher in the abatacept arm than the control.
  • Rituximab showed initial improvements in C-peptide preservation and insulin use; however, efficacy decreased over time, potentially because rituximab does not induce immunological tolerance. Dr. Herold presented data showing that the T cell’s proliferative response to antigens is similar in rituximab- and placebo-treated subjects, suggesting that the treatment does not actually eliminate autoreactive T cells. He explained thatmany people with type 1 diabetes have defective central (in the bone marrow) and peripheral B cell tolerance checkpoints, which are supposed to eliminate autoreactive B cells. Because these checkpoints are defective, and because rituximab appears to only eliminate autoreactive cells, not prevent their creation, the cells that replace those destroyed by rituximab are also autoreactive. Therefore, Dr. Herold advocated for focusing on therapies that induce immunological tolerance, rather than just kill the autoreactive cells.
  • In contrast to rituximab, Dr. Herold explained that teplizumab appears to induce immunological tolerance. Teplizumab treatment appears to cause autoreactive T cells to migrate to the gut and differentiate into regulatory T cells, which potentially could help mitigate the autoimmune attack.
  • Dr. Herold noted that though the Protégé trial did not meet its primary endpoint, it did show that there was C-peptide preservation and a reduction in the need for insulin in people treated with the drug. For more on the Protégé study’s one-year and two- years results, see our October 21, 2010 Closer Look at and our EASD 2012 Type 1 Diabetes Treatments and Cure-Based Therapies report at, respectively. We believe, that it is now fairly widely felt that Protégé’s design was not ideal – this is perhaps most challenging to see for patients given that few second chances are given to assess a drug’s efficacy. Dr. Herold said that the AbATE trial (a study of two 14 day courses of teplizumab in children with recent onset type 1 diabetes [n=83]), found that teplizumab delayed the decline of C-peptide levels by 15.9 months. For more details on the AbATE trial’s results, see our ADA 2011 Type 1 Therapies Cure Related report at
  • Interestingly, he detailed a new assay his lab has developed that measures beta cell death through quantification of demethylated insulin DNA. We are excited to see further results using this assay, as we think it could mark a major advance in quantifying beta cell viability, which thus far has only been measured indirectly in humans.
  • Looking to the future, Dr. Herold proposed that more work needs to be done on matching people with the right therapy, as “nothing works on everybody.” For example, people in AbATE who responded to teplizumab (defined as losing less than 40% of their beta cell function) were more likely to have a lower A1c level and use less insulin at baseline than non-responders (i.e., people whose beta cell function was reduced by at least 40%). Responders had C-peptide levels that were an average three-fold higher than non-responders.
  • He also pressed for more research focused on developing combination therapies, such as an anti-CD3 monoclonal antibody (like teplizumab) in conjunction with either anti-IL-1beta or IL-RA. He said that when mice models are treated with a combination of an anti-CD3 monoclonal antibody and either anti-IL-1β (an antibody that blocks the inflammatory protein IL) or IL-1RA (a natural inhibitor of IL), there is a synergistic reversal of diabetes, such that more reversal occurs when the drugs are given in combination than the sum of each of their effects as a monotherapy.


Jay Skyler, MD (University of Miami Miller School of Medicine, Miami, FL)

Dr. Jay Skyler opened with a very touching acknowledgement to the late Dr. George Eisenbarth, whose contributions to the field of immunology in type 1 diabetes have been unparalleled. Dr. Skyler discussed efforts to prevent diabetes at various stages of disease progression. He identified the use of non-cow’s milk formula and the removal of bovine insulin from cow’s milk as two primary prevention strategies (avoiding new cases of diabetes in people at risk for diabetes but who have no autoantibodies) that have shown promise in clinical trials. In contrast, there is not strong evidence that delaying exposure to gluten, supplementing vitamin D, or preventing early inflammation block antibody formation or type 1 diabetes progression. Notably, Dr. Skyler expressed frustration with recent efforts to promote the BCG vaccine as a promising means of preventing type 1 diabetes. One study showed that BCG vaccination prior to age three months may increase the likelihood of developing type 1 diabetes. Most efforts for secondary prevention (delaying disease progression in those with autoantibody formation) have been unsuccessful, though Dr. Skyler noted that some hope remains for oral insulin. Tertiary prevention efforts, in those clinically diagnosed with type 1 diabetes, have had mixed results. Anti-CD3, anti-CD20, and co-stimulation blockade have shown transient benefit. Other agents have exhibited a pattern of showing early promise but failing to remain effective over time. For example, in pilot studies GAD-65 vaccine and heat shock protein 60 (DiaPep277) was said to preserve C-peptide better than their respective controls, but these were not confirmed in subsequent trials. Dr. Skyler concluded by proposing a potential combination therapy approach: 1) use an anti-IL1B or anti-TNF to diminish the inflammatory response; 2) use an anti-CD3, anti-CD20 or co-stimulation blockade to modulate the adaptive immune system; 3) add GAD vaccine and/or oral insulin as a specific antigen; 4) use GCSF to stimulate recovery of regulatory T cells (T regs) or infuse T-regs themselves; and 5) use an agent such as exenatide, GLP-1, HIP2b, or INGAP to stimulate beta cell function.

  • Dr. Skyler first reviewed the natural history of type 1 diabetes, identifying risk factors for the development of type 1 diabetes and possible points of intervention. The late Dr. George Eisenbarth determined that people who develop type 1 diabetes begin with a genetic predisposition that outweighs genetic protection. This genetic predisposition by itself, however, is not sufficient for precipitating type 1 diabetes given that 40% of Caucasians have one of the two most strongly predisposing HLA haplotypes, and less than 40% of Caucasians actually have type 1 diabetes. Hence, there have been efforts to identify a potential environmental trigger. The next stage of the disease is the onset of T cell autoimmunity causing beta cell injury, which can be monitored by the appearance of autoantibodies. Subsequently, as beta cell function declines, insulin responsiveness to glucose is lost, resulting in dysglycemia. Finally, at the clinical onset of type 1 diabetes, some beta cell function still persists, which could be harnessed. An intervention could potentially be performed at any of these stages of disease development to prevent or delay progression.
  • Primary prevention trials (where an intervention is applied to those genetically at risk for type 1 diabetes but who have no evidence of disease progression) have focused on manipulating various putative environmental factors. These include cow’s milk, gluten exposure, inflammation, vitamin D insufficiency, insulin auto-reactivity, and the BCG vaccine.
    • Cow’s milk: The TRIGR pilot study found that infants weaned from breast-feeding to cow’s milk formula were more likely to exhibit antibody formation than in infants weaned onto non-cow’s milk formula. The full TRIGR study is ongoing. Additionally, the FINDIA study found that removal of bovine insulin from cow’s milk reduced the likelihood of developing antibodies.
    • Gluten exposure: The BABYDIET study demonstrated that early exposure to gluten did not predispose children to type 1 diabetes.
    • Inflammation: The NIP pilot study suggested that early prevention of inflammation did not block formation of antibodies, but Dr. Skyler noted that had there been a longer follow-up, an effect might have been observed.
    • Vitamin D insufficiency: In a meta-analysis of case-control studies, vitamin D supplementation in infancy did not seem to reduce risk for type 1 diabetes, but a randomized controlled trial is lacking.
    • Insulin auto-reactivity: The JDRF’s PrePoint study, a dose escalation safety study for oral or intranasal insulin to prevent the formation of insulin autoantibodies and diabetes, is currently ongoing.
    • The BCG vaccine: Dr. Skyler criticized the recent championing of the BCG vaccine as an intervention for type 1 diabetes, stating that the appearance of diabetes, if anything, is higher for those receiving the BCG vaccine before the age of three months (Huppman, et al., Diabetes Care 2005), and that previous studies later in the course of the disease had failed to show a benefit. Our commentary of one such study (see PLos One, 2012) can be found at
  • Secondary prevention trials target those who have already developed autoantibodies. Therapies that have been tested for the ability to perform secondary prevention are nicotinamide, intranasal insulin, parenteral insulin, and oral insulin.
    • Several therapies have failed to reduce progression to type 1 diabetes compared to controls. These include nicotinamide in the ENDIT and DENIS studies, intranasal insulin in the DIPP study, and parenteral insulin in the Diabetes Prevention Trial-Type 1 (DPT-1).
    • Some hope, however, remains for oral insulin. In the DPT-1, the subgroup with insulin autoantibodies (IAA) of ≥80 nU/ml at the beginning of the study (263 of the 372 study participants) experienced a projected 4.5-5 year delay in the onset of type 1 diabetes. However, this was only found in a post-hoc analysis; Dr. Kevan Herold (Yale University, New Haven, CT) explained that the DPT-1’s inclusion criteria based on autoantibodies was expanded in order to expand enrollment. A TrialNet prospective study in people with IAA ≥80 nU/ml is currently ongoing in hopes of replicating that result in a prospective study.
  • Tertiary prevention involves intervening in newly diagnosed patients. Many candidates for tertiary prevention have been researched including cyclosporine (an immunosuppressant), teplizumab (an anti-CD3 monoclonal antibody [mAb]), otelixizumab (another anti-CD3 mAb), rituximab (an anti-CD20 antibody), abatacept (an anti-CTLA4 antibody), heat shock protein 60 (Dia Pep 277), the GAD 65 vaccine (Diamyd), rapamycin (an immunosuppressant) and IL-2, canakinumab (an IL-1-antibody), anakindra (an IL-1 receptor antagonist), and thymoglobulin.
    • Dr. Skyler reviewed various tertiary prevention candidates that briefly reversed or halted the progression of type 1 diabetes but then ceased to work. These included cyclosporine, teplizumab , otelixizumab , ,rituximab, and abatacept.
    • Some other treatments have shown promise, but with limitations. Heat shock protein 60 (DiaPep 277) appears to preserve C-peptide better than placebo in response to an intravenous glucagon challenge, but not in response to the more standard mixed-meal tolerance test. In one initial study, the GAD-65 vaccine (Diamyd) appeared to preserve C- peptide levels after 30 months, but only in people receiving the treatment within 6months after diagnosis, and in those responders it did not reduce their insulin requirement (Ludvigsson et al., NEJM 2008). Yet, two larger subsequent studies with GAD-65 vaccine showed no effect (Wherett et al, Lancet 2011; Ludvigsson et al., NEJM 2012).
    • Finally, various candidates have failed to perform tertiary prevention. These include the combination of rapamycin and IL-2; the combination of mycophenylate and anti-CD25; canakinumab; and anakindra. Thymoglubulin, in the START trial, showed no significant effect after one year, but Dr. Skyler speculated that one might become apparent after more time.
  • Dr. Skyler listed several agents that are subject of ongoing or planned studies: alefacept (which would cause apoptosis of memory-effector T cells), sitagliptin plus the proton pump inhibitor lansoprazole for beta cell proliferation, alpha-1 antitrypsin (a protease inhibitor that protects tissues from inflammatory cells’ enzymes), secukinumab (an anti-IL-17A mAB), tocilizumab (an IL-6R mAb), and proinsulin peptides.
  • Finally, Dr. Skyler proposed a potential combination therapy approach. He proposed to first diminish the inflammatory response with an anti-IL1B or an anti-TNF, then modulate the adaptive immune system with an anti-CD3 or an anti-CD20 or co-stimulation blockade, then add a GAD vaccine or oral insulin as an antigen-specific approach, then stimulate the recovery of regulatory T cells (T-regs) with GCSF (granulocyte colon-stimulating factor) or infuse T-regs themselves, and finally use exenatide, GLP-1, HIP2b or INGAP to stimulate beta cell function. Dr. Skyler concluded by stating that in spite of recent failures in the field, he thinks we will be able to halt type 1 diabetes in the future.


Kevan Herold, MD (Yale University, New Haven, CT); Jay Skyler, MD, MACP (University of Miami Miller School of Medicine, Miami, FL)

Q: With regards to the effect of BCG vaccination, the slide showed that among children who had been vaccinated under three months of age there was a high incidence of type 1 diabetes. WHO recommends vaccination within a week of birth in some countries. This seems like a paradox and that the incidence of type 1 diabetes might be quite related to the vaccination of BCG.

Dr. Skyler: The BCG story is complicated. There have been at least three studies done in new onset diabetes, two that showed no effect and one that might have shown a slight acceleration. The study I showed was conducted in people who had first-degree relatives with type 1 diabetes. They were given the BCG vaccine with the idea to prevent type 1 diabetes. You can see that it did not do this. I do not think that there is reason to think it is protective. I also do not think that we should not use it in countries where there is Tb [tuberculosis] risk due to fear of it causing type 1 diabetes. The highest rate of type 1 diabetes is in northern European Caucasians and that is not an area with much Tb, so the BCG vaccine is not very prevalent. I don’t think that public health practices should be modified in this respect. In another study, the BCG vaccine was used in three people with type 1 diabetes. The report study was published in PloS One this summer and it suggested that you might be able to protect against type 1 diabetes. However, the control only had two patients. It had such smaller numbers that I don’t think I would have made any conclusion based on it. However, the authors concluded that BCG is going to be the future. They have a very active website and are raising a lot of money saying this.

Q: In clinical practice we see a small group of kids who have this prolonged honeymoon period. Has that group of diabetics been looked at? Are they genetically different or different in their immune response, and could that be instructive to further studies in terms of preservation of beta cells? And also for responders and non-responders, could it be that those are your outliers? That if you took the patients with the lower A1cs and the lower insulin at the beginning of the study and put some of them in the control group would that have affected the outcome?

Dr. Skyler: This was a randomized, placebo-controlled trial so the fact that you look at subgroups doesn’t take away the randomization.

Dr. Herold: It’s a randomized controlled trial, and we do correct statistically for all the various parameters you mentioned, including basal C-peptide and A1c. So I think that’s dealt with, at least in the statistical analysis. The first part of your question is interesting, though. What’s different about those who tend to maintain the C-peptide response? I have to say I don’t know at all. There have been a variety of studies looking at differences in autoantibodies, and so on, that have from time to time come up with subtle differences and I don’t find any of that data convincing.

Dr. Skyler It’s not very convincing. But those are picking parameters we know about; there may be parameters we don’t know about.

Dr. Herold: It also depends if you’re a lumper or a splitter. If you’re a lumper you say everyone follows the same course, and if you go out long enough then you’ll see beta cell function decline. If you’re a splitter then that question is important. There could be differences we haven’t looked at yet.

Q: The microbiome was mentioned as a potential modulator of immune function. Is there any work being done to see if modifying the microbiome has an effect on the natural history of the disease?

Dr. Herold: I will start and then Jay will say the exact opposite. I think that the question you are asking is a critical question that must be answered. There is circumstantial evidence that modifying the microbiome might be effective. So the question is very valid. My own view is that the idea that it will have a direct effect of type 1 diabetes is still unproven but needs to be done.

Dr. Skyler: There was a NOD mouse study done earlier this year. It used a bacteria that had been modified to express human proinsulin and IL-10, and they gave this bacteria to the mice after giving a low dose of anti-CD3 so that the anti-CD3 could be given at a lower dose than Kevan was using. They found that it had a profound effect. It was a way of doing four things at once. Giving proinsulin (an antigen), IL-10 (an anti- inflammatory), modifying the microbiome, and using an immunomodulatory therapy. They will hopefully be able to human studies in the next year or two.

Dr. Herold: I absolutely agree. There was one other piece of data pertaining to your question. Another study looking at a mouse model. If you put the mice in a facility where there are no endogenous bacteria, parasite, or viruses, the mice will develop rip-roaring diabetes. However, if you take feces from the mice raised in a conventional facility and put them with the mice in this other facility then they do not develop diabetes.

Gestational Diabetes

Plenary Session V: Diabetes and Pregnancy


Donald Coustan, MD (Brown University, Providence, RI)

Dr. Donald Coustan provided his thoughts on the potential impact of the recent gestational diabetes (GDM) guidelines developed by the International Association of Diabetes in Pregnancy Study Group (IADPSG) and published in Diabetes Care in 2010. These definitions will label roughly 18% of women as having GDM (a higher percentage than with previous US guidelines). While Dr. Coustan acknowledged that some object to any criteria that give such a high prevalence of GDM, he argued that this rate simply mirrors the US epidemic of prediabetes and diabetes. Two randomized controlled trials based on similar criteria have demonstrated significant improvements in maternofetal outcomes (ACHOIS Trial Group NEJM 2005; Landon et al. NEJM 2009). Furthermore, the IADPSG guidelines have been found cost- effective as long as post-delivery care is also given (Werner et al., Diabetes Care 2012). The guidelines were adopted by the ADA in January 2011 but rejected (so far) by the American College of Obstetricians and Gynecologists; the NIH will offer its own assessment at a consensus development conference scheduled for March 2013 (rescheduled from October 2012 due to Hurricane Sandy).

  • As a reminder, the recommendations of the International Association of Diabetes in Pregnancy Study Group (IADPSG) were motivated by the large multinational Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study (n≈23,000), which revealed a continuous association between maternal glucose tolerance at 28 weeks and adverse pregnancy outcomes such as macrosomia, preeclampsia, and preterm birth (NEJM 2008). The new guidelines diagnose GDM based on having at least an abnormality in fasting, one-, and/or two-hour glucose values during a 75-g OGTT (defined by values ≥92 mg/dl, ≥180 mg/dl, and ≥150 mg/dl, respectively). Previous guidelines required two abnormal glucose values for GDM classification. These criteria were determined by the IADPSG based on an average odds ratio of 1.75 for three adverse outcomes: birth weight > 90th percentile, cord C-peptide > 90th percentile, and percent newborn body fat > 90th percentile. (Dr. Coustan acknowledged that because these risk relationships were continuous, the choice of thresholds was somewhat arbitrary.) By these measurements, 17.8% of pregnant women in the HAPO study had GDM. Dr. Coustan believes that this prevalence, while alarmingly high, is reasonable given unpublished NHANES 2010 data indicating that 31% of US women ages 18-44 have either diabetes (4.5%) or prediabetes (26.4%).


Thomas Buchanan, MD (University of Southern California, Los Angeles, CA)

In a fascinating discussion of diabetes prevention in women with previous gestational diabetes (GDM), Dr. Thomas Buchanan reviewed data from randomized controlled trials, outlined molecular mechanisms for GDM, and suggested several clinical implications. Dr. Buchanan noted that the rate of diabetes progression matches that the rate of decreasing beta-cell compensation. Further, he proposed that the loss of beta-cell function seems mainly to be mediated by insulin resistance and its associated changes in free fatty acids, adipokines, etc. In this framework, interventions to preserve beta-cell function can occur at three levels: treating the beta cell itself, changing fat metabolism (e.g., with TZDs), and changing fat accumulation (e.g., with weight-loss interventions). In the TRIPOD and PIPOD trials in women with prior GDM, Dr. Buchanan and colleagues found that troglitazone and pioglitazone significantly delayed progression to type 2 diabetes. Notably, the women with the most dramatic reductions in insulin secretion during intravenous glucose tolerance testing at baseline saw the greatest benefits from troglitazone and pioglitazone treatment. Thus, Dr. Buchanan hypothesized that “unloading” the beta cell seems to be an important therapeutic strategy in diabetes prevention. (He noted that an alternative approach to preserving beta-cell function is simply to increase the rate of beta- cell growth, which has been shown in preclinical studies of GLP-1 agonists.)


Mark Landon, MD (The Ohio State University, Columbus, OH), Donald R. Coustan, MD (Brown University, Providence, RI), Celeste Durnwald, MD (University of Pennsylvania, Philadelphia, PA), Thomas Buchanan, MD (University of Southern California, Los Angeles, CA), Steven Gabbe, MD (The Ohio State University, Columbus, OH)

Q: Could you briefly discuss the choice of insulin during pregnancy, especially with the new category-B labeling for insulin detemir?

Dr. Durnwald: In our practice, we use NPH as our basal insulin. I never start a woman on Lantus during pregnancy. There is some concern that when you use Lantus you may have to use BID dosing because it doesn’t quite act with 24-hours like it does in someone with type 2 diabetes. I no longer use regular human insulin; I think the rapid-acting analogs have been shown to be safe and more effective for meal coverage. I don’t use Levemir (detemir).

Dr. Coustan: The study on detemir in pregnancy has been published. Unfortunately, they haven’t published levels of Levemir on cord blood. I understand from Dr. Lois Jovanovic that the researchers have these levels; I am not sure whether they have been analyzed yet. Since insulin syringes hold either half or one CC, woman must take two injections if she needs more than 100 units at a time. They only make U- 500 of regular insulin, so under those circumstances we use U-500 regular insulin simply because it saves the patient discomfort.

Q: A member of the audience was piqued by Dr. Buchanan’s suggestion that bariatric surgery has a place in the management of morbidly obese individuals. Could you discuss this?

Dr. Buchanan: We are not really preventing diabetes in most people through lifestyle interventions. Our approaches to obesity have two components, much like cigarette smoking. For people who are not yet smoking we use various public health messages and incentives; with people already addicted to cigarettes, we have to take fairly drastic measures. From what I can see, there is nothing that broadly and reliably fixes the problem of obesity in people who already are obese. We are being funded by NIH to do a study comparing gastric banding to oral medication – it was supposed to be TZDs – for diabetes prevention as an early intervention. The reason we are doing the study is to move the field toward the idea that this is a progressive disease and that we need to intervene as early as possible, especially if the disease continues to progress. Unless someone comes up with a medication to take a patient from BMI 32 kg/m2 to 22 kg/m2, I think we have to use metabolic surgery. If you can get someone with BMI 32 kg/m2 to lose 40-50 pounds, they will absolutely stop progressing to diabetes. If you look at the health statistics across the BMI range, you can see that obesity in the mid-30-range causes a lot of health problems. I think that one way or another, we need to be more aggressive about reducing body fat; right now, bariatric surgery seems to be the best we have.

Dr. Coustan: Obesity is about more than diabetes. It is a life-threatening condition, and I don’t think that we view it that way often enough. I think that when I see patients in pregnancy, it may be a teachable moment about obesity not only on its risk of diabetes but the risk to their life. I think that there are randomized trials showing benefit to bariatric surgery for making diabetes go away, even before the weight loss. Maybe there is a disruption of the track between the digestive system and the pancreas. I don’t think it is any longer a radical approach; if we consider obesity a life-threatening illness as dangerous as cancer, then I think we need to treat it.

Q: Do you have a theory on why the treatment of gestational diabetes reduces the risk of preeclampsia?

Dr. Coustan: I could not give you precise mechanisms. The association is there with insulin resistance, elevated insulin levels, and hypertension, but I can’t give you more than that. From the HAPO study database, we are going to analyze the insulin levels at the time of glucose analysis. Whether that will shed any light, I am not sure.

Dr. Buchanan: I don’t think we even understand why obesity causes hypertension. We had a program funded on this, but we still don’t know.

Dr. Coustan: It’s not just obesity – when we controlled for BMI, we still saw an independent effect of glucose control on hypertension.

Dr. Buchanan: I am not sure that statistically controlling for BMI really controls for fat mass, though.

Healthcare Structures

Workshop: Value of Cost Sharing in Diabetes Self-Management


Iris Sanchez, DNP, FNP, BC-ADM (University of Texas School of Nursing, Houston, TX)

Iris Sanchez believes that in general, the net value of attending shared medical appointments (SMAs) is positive for patients, but that more rigorous studies must be conducted in order to definitively make this case to payors and administrators. Given that 50% of the costs associated with diabetes care stems from inpatient hospitalization, Ms. Sanchez argued that using SMAs to intervene earlier, before a health issue becomes urgent, could prevent the need for inpatient services and reduce healthcare costs for patients. She then examined the economic value for patients with regard to medications, clinician visits, ER visits, and hospital admissions, concluding that SMAs may increase access to medications, reduce numbers of clinician and ER visits, and reduce rates of hospital admissions compared to usual care. She noted that most data comes from the adult type 2 diabetes population, and we do not have good data on the value of SMAs to patients with type 1 or gestational diabetes. She urged audience members to collect data on outcomes and economic value in their patients participating in SMAs to supplement the sparse data currently available.


Sandra Burke, PhD (President, American Association of Diabetes Educators, Chicago, IL)

President Sandra Burke began the last presentation of the day by describing the current state of diabetes and lamenting, “The scope of the problem is beyond belief.” She went on to describe how current projections have the US spending ~20% of its GDP on healthcare in 2016, thus the increasing cost of healthcare has become a recurring theme of healthcare reform. One of the strategies that has been proposed for containing costs is cost-sharing, or having patients pay for a portion of their healthcare services. The unintended consequence of cost sharing is that it appears to cause people to under-use medical services. An unpublished retrospective analysis (n=80,524) by the ADA found that people, other than the most at risk, were more likely to participate in diabetes self-management if they were in a low cost-sharing healthcare system than a high cost-sharing system. Additionally, people of all risk levels were more likely to comply with medications when they had low cost-sharing. Ironically, people with low-cost sharing had reduced costs during hospital admissions. President Burke then reviewed two potential models for chronic care: the accountable care organizations (ACOs) and the patient-centered medical home. Dr. Burke seemed to be concerned about the bundling of expenses that often occurs within ACOs, but she seemed somewhat more optimistic about the potential of the patient centered medical home for diabetes patients because of its emphasis on coordinating care.


Ruth Lipman, PhD (American Association of Diabetes Educators, Chicago, IL); Sandra Burke, PhD (President, American Association of Diabetes Educators, Chicago, IL); Iris Sanchez, DNP, FNP, BC-ADM (University of Texas School of Nursing, Houston, TX)

Q: Iris how do you see the shared medical visits comparing and contrasting with diabetes self-management education programs [DSMEs]?

Ms. Sanchez: The first question I would have is, do we even know the definition of a shared medical appointment? That’s one of the barriers. There’s no one governmental office or association saying this is the standard definition, how to conduct one, who can conduct one, and what are the qualifications. Right now there are lots of practices conducting them, and we don’t know how they’re doing it. What we do know is that usually there is a physician involved and patients can receive point-of-care and medication changes. Compared to a DSME program – that is comprised of diabetes educators, but we don’t know what information is actually communicated to that individual’s healthcare provider.

Comment: I have been doing combined visits with physicians at point-of-care and changing drugs as a pharmacist working with a dietician and nursing staff. All of sudden they are calling it the medical home. There is a way to do it, it is just how do you define it for the bill. If you talk to the right person at CMS you get the wrong answer. Physicians are nervous that if they get audited they are going to fail. I think that you need the number one priority is to get a clear definition. How do you identify the benefits of all of this? I did a study with an integrated HMO who had information on compliance and drug use. We turned in the clinical piece, and a pharmaceutical company did the analysis and found that there was huge savings per year. They said they could not attribute it to the shared appointments. There is a bill in Congress right now for CDE payment. Each and every one of the non-physician providers have to be able to bill. If you want to report out your stuff in smaller numbers there are ways to publish all of this information out, but at the end of the day everyone else has to get recognized as a provider and get paid. I would just say keep pummeling them and get as information out as possible.

Dr. Burke: I would add to that, from an evidence based practice standpoint, we really need randomized controlled trials if we can make that work. Case studies are important, but are fairly low-level evidence. We need to see more than case studies to change practice. I think building the evidence base over the next few years will become very important, but I could not agree with you more that we have to be able to bill for services.

Q: I am a primary care provider and CDE from Dayton, Ohio. The question is how do you work the education in for a short visit. My manager keeps saying we have to get volume. My question is how to implement the managed care system. The government sector is so different; they can do things we can’t do in the private sector. Where are the models for implementing these in the private sector?

Dr. Burke: The one that comes to mind first is the one at Emory Memorial Hospital in Atlanta. Amparo Gonzalez was initially involved in that. That is probably one of the best models I can think of. It has not been published yet.

Ms. Sanchez: That is why I wanted to know who is doing shared medical appointments. Talk to your peers who are doing them. You are right, the VA does not have to worry about billing codes, but they are useful to speak to.


Opening Remarks

Opening Remarks


Kwame Osei, MD (The Ohio State University, Columbus, OH)

Dr. Kwame Osei rousingly introduced the leaders of various national and international diabetes organizations – it was very powerful to see all of these mighty individuals gathered together in one space and passionately call the audience to action.

  • CEO of the Ohio State University Wexner Medical Center, Dr. Steven Gabbe, expressed gratitude for how far we have come (today’s sensors, pumps, and the A1c test were not available 40 years ago when he diagnosed himself with type 1 diabetes) but also highlighted how far there is to go, especially for today’s children. Dr. Gabbe stated that diabetes is now one of the most common chronic diseases among children, remarking that he even has teenage patients developing gestational diabetes.
  • President of the International Diabetes Federation, Dr. Jean-Claude Mbanya, also expressed the sentiment that we must work today to children of protect future generations, stating, “tomorrow will be a better future because of what we decide today.”President of the American Diabetes Association, Dr. Vivian Fonseca, noted that while average A1cs are falling and rates of retinopathy, end-stage renal disease, and amputations are declining, overall diabetes incidence is still rising. He remarked that, if even a relatively wealthy country like the US has difficulty affording diabetes, then it will certainly bankrupt other countries. He implied that, if for no other reason, this alone should fuel efforts to address the disease from a global perspective.
  • Dr. Aaron Kowalski, Assistant VP of Glucose Control Research at the JDRF, echoed the sentiment that we have come a long way but still have much work to do. Notably, he remarked that even at arelatively good A1c of 7% the average person still spends more than 10 hours/day hyperglycemic and one hour/day hypoglycemic – he stated, as well, that we clearly need to do better.



-- by Jessica Dong, Hannah Deming, Kira Maker, Joseph Shivers, and Kelly Close