A number of studies presented at the European Association for the Study of Diabetes (EASD) meeting in Athens, Greece, discussed topics related to obesity, including not only its treatment but also its hormonal mediators, relationship to the metabolic syndrome and diabetes, and complications.

Therapeutic Approaches to Obesity:

Gastric Banding

Hanusch-Enserer and colleagues[1] studied 32 persons with morbid obesity before and 6 and 12 months after gastric banding in order to examine the relationship between weight loss following bariatric surgery and cardiovascular risk factors. Subjects experienced a mean weight loss of 32 kg, which resulted in reduced blood pressure, waist circumference, fasting glucose, and triglyceride levels. In addition, vascular cell adhesion molecule-1 and E-selectin were lowered. There was no change in CD40 ligand, lipoprotein-associated phospholipase A2, or matrix metalloproteinase-2, suggesting that the degree of weight loss attained with such procedures should not be expected to normalize the proinflammatory state seen in individuals with morbid obesity.

Rimonabant

A great deal of interest has been generated by the cannabinoid receptor 1 (CB1) antagonist rimonabant. Julio Rosenstock, MD,[2] Dallas Diabetes and Endocrine Center, Dallas, Texas, presented a pooled analysis from the Rimonabant in Obesity (RIO)-Lipids, RIO-Europe, and RIO-North America studies of the effects of rimonabant in 1290 overweight individuals with impaired fasting glucose, defined as fasting glucose between 100 mg/dL and 125 mg/dL. Those receiving placebo had a 1-year weight loss of 1.7 kg. Patients receiving rimonabant 5 mg had a weight loss of 3.2 kg, and those receiving rimonabant 20 mg had a weight loss of 6.9 kg, with 2-cm, 4-cm, and 7-cm reductions in waist circumference, respectively. Of the 3 groups, 4.9%, 4.0%, and 3.6%, respectively, progressed to type 2 diabetes, whereas 39%, 42%, and 47% reverted to normal fasting glucose.

There is fascinating evidence from the RIO studies that in obese individuals, rimonabant has greater effects on glucose and lipids than can be extrapolated from the metabolic changes seen in placebo-treated individuals for a given degree of weight loss. Knerr and colleagues[3] noted that CB1 is predominantly expressed in the brain, adipose tissue, and immune cells, and is involved in the regulation of immunity, appetite, and body weight. CB1 expression is upregulated in the adipose tissue of obese rats, whereas mice not expressing the CB1 receptor show decreased food intake. Food intake is also reduced with the administration of rimonabant. In a study of 77 individuals with normal glucose tolerance receiving a glucose load, those ≤ age 47 years had a 12% decrease in CB1 mRNA expression in peripheral blood leukocytes, whereas those > age 47 years had a 24% increase, which was not explained by differences in sex, body mass index (BMI), waist circumference, blood pressure, triglycerides, uric acid, or fasting or 2-hour blood glucose levels. If similar processes occur at central nervous system CB1 receptors, this may contribute to the weight gain that occurs with aging by reducing the degree of satiety that may otherwise accompany nutrient intake. At peripheral metabolic sites, decreases in insulin sensitivity, increases in glucose, and abnormality of lipids may occur, and there may also be immune effects, including those potentially related to the inflammatory processes accompanying atherosclerosis. Thus, the effects of age on CB1 receptor response to nutrients may play important roles in a variety of central and peripheral changes that are known to accompany the aging process.

Sibutramine

Wang and colleagues[4] studied 60 overweight (baseline BMI approximately 27 kg/m2) type 2 diabetic women receiving sulfonylurea plus metformin. Thirty subjects were randomized to receive sibutramine 15 mg daily, and 30 were randomized to placebo (in addition to their previous medications) in this 12-week study. Subjects in the sibutramine group had a 2.5-kg weight loss, compared with a .1-kg weight loss in the placebo group. Sibutramine-treated patients had a 1.7% decrease in A1C, compared with a .2% decrease in the placebo group. These changes were accompanied by improvements in lipids in the sibutramine group, but no significant differences in blood pressure and heart rate were found between the 2 groups. These data confirm that sibutramine remains a useful approach, with the caveat that for patients with poor glycemic control, such as the study participants, sibutramine may not truly be the appropriate intervention.

Incretin Hormones

In addition to agents that are specifically directed at the treatment of obesity, several glycemic therapies that have interesting effects on weight were described. McCormack and colleagues[5] administered the dipeptidyl peptidase IV inhibitor PSN9301 (which is in phase 2 clinical development), metformin, or both in a diabetic rodent model. Similar weight loss and glucose-lowering effects were shown for PSN9301 and metformin, and there was additive benefit when the 2 treatments were administered together.

In an open-label extension study of 87 obese (BMI, 35 kg/m2) type 2 diabetic persons currently treated with lifestyle modification alone or with metformin, Wintle and colleagues[6] reported that the addition of exenatide 10 mg twice daily resulted in a reduction in A1C from 7.8% to 6.9%, which was seen after 12 weeks, and a 3.8-kg weight loss, which occurred steadily through 34 weeks of observation for both groups.

Robert E. Ratner, MD, MedStar Research Institute, Hyattsville, Maryland, and colleagues[7] similarly reported that an open-label extension study of 195 obese (BMI, 34 kg/m2) type 2 diabetic persons treated with metformin, a sulfonylurea, or both who were given exenatide 10 mg twice daily sustained a reduction in A1C of 1.1% over the course of 2 years, and showed a progressive 5.2-kg weight loss during the period of observation. Weight loss at 2 years was 4 kg, 5 kg, and 6 kg for those with baseline BMI < 30 kg/m2, 30-40 kg/m2, and > 40 kg/m2.

A caveat of both studies is that only a self-selected subset of patients chose to participate -- 87 of 127 in the first study and 195 of 354 in the second study. This leads one to speculate that those who experienced particular benefit were overrepresented, and average weight loss may not be as great in general clinical practice.

Insulin Detemir

Hermansen and Tame[8] reported a fascinating analysis of 476 type 2 diabetic individuals who were inadequately controlled with oral therapy who initiated treatment with either the new long-acting insulin analog detemir or NPH insulin twice daily. The investigators designed a study that employed an aggressive basal insulin titration algorithm to achieve excellent glycemic control. Insulin detemir is thought to exhibit a lower degree of day-to-day biological variability than NPH insulin, and there is evidence of fewer episodes of hypoglycemia in patients using detemir. It has been speculated that detemir may be associated with less weight gain as well, either because there is a lack of hypoglycemia-driven nutrient consumption or, interestingly, because of greater penetration of the blood-brain barrier, allowing it to exhibit the well-characterized hypothalamic insulin effect of decreasing appetite. Even more speculatively, detemir may exhibit a greater relative hepatic action due to albumin-related uptake.

Both insulin detemir and NPH insulin achieved similar decreases in A1C -- from 8.36% to 6.46% for detemir and from 8.42% to 6.58% for NPH -- but detemir was associated with less weight gain than NPH insulin. Categorizing patients by BMI, those ≤ 25 kg/m2, > 25-27 kg/m2, > 27-29 kg/m2, > 29-31 kg/m2, and > 31 kg/m2 had weight gains with detemir of 2.25 kg, 1.81 kg, 1.10 kg, .54 kg, and .63 kg, respectively, whereas those treated with NPH had respective weight gains of 3.15 kg, 2.73 kg, 3.08 kg, 2.79 kg, and 2.60 kg. These results suggest that detemir may be of particular benefit in those with the highest weight at baseline.

Hormonal Mediators of Obesity

Pavlatos and colleagues[9] studied the appetite-stimulating hormone ghrelin. Ghrelin is produced in the gastric fundus and activates neuropeptide Y and agouti-related protein neurons in the arcuate nucleus of the hypothalamus. Levels increase prior to eating, perhaps promoting meal initiation, and decline postprandially, suggesting that ghrelin decreases rapidly to improve satiety. In 9 lean vs 9 obese women, fasting ghrelin levels were 708 pM/L vs 354 pM/L, respectively, with an inverse relationship to BMI, body fat, and waist circumference. Protein-rich meals led to a decline in ghrelin levels at 1 hour and 2 hours postprandially in lean persons, but to no change in obese women. There was no change in either group with a high-fat meal. This suggests a mechanism whereby meals relatively high in protein may favorably affect satiety, although paradoxically offering less benefit to those who most particularly require such an aid to the achievement of satiety.

Nannipieri and colleagues[10] measured adiponectin mRNA in subcutaneous and visceral fat obtained during abdominal surgery from 6 lean nondiabetic persons, 10 obese persons with diabetes, and 10 obese persons with normal glucose tolerance. There was no effect of diabetes, but adiponectin levels were 41% to 53% lower at both sites in obese compared with normal-weight persons, confirming the evidence that this adipocyte hormone, which produces improvement in insulin action, is paradoxically diminished with increasing fat mass.

Invitti and colleagues[11] studied 162 obese children aged 8-18 years and demonstrated in a multivariate analysis that there was a negative correlation between adiponectin levels and the homeostasis model of assessment of insulin resistance (HOMA-IR), and there was no additional effect of BMI, triglycerides, high-density lipoprotein cholesterol, plasminogen activator inhibitor-1 (PAI-1), fibrinogen, 2-hour glucose, and insulin response to glucose. One quarter of the children had the metabolic syndrome, which was associated with 31% lower adiponectin and 91% greater HOMA-IR. Thus, among children, the degree of obesity was more strongly associated with adiponectin level than with insulin resistance. This effect appeared to be related to the metabolic syndrome, although one cannot be certain from this study about the direction of causality.

Diet Characteristics

Lindström and colleagues[12] discussed the effect of dietary macronutrient composition on the development of type 2 diabetes in 522 overweight, middle-aged persons with impaired glucose tolerance who participated in the Finnish Diabetes Prevention Study. In this study, which demonstrated that lifestyle intervention reduced the development of type 2 diabetes, individuals from both the lifestyle and control groups with the highest vs lowest quartiles of fiber intake, after adjustment for weight, weight change, and physical activity, showed a 59% reduction in diabetes incidence. For total fat intake, the highest quartile had a 2.5-fold increase in incidence over the lowest quartile, and for saturated fat intake, a 1.9-fold increase. Compared with individuals on either a low-fat/low-fiber diet or a high-fat/high-fiber diet, a low-fat/high-fiber diet led to approximately half the incidence of diabetes, whereas compared with persons ingesting both high fat and low fiber, a low-fat/high-fiber diet led to a diabetes incidence approximately one third as great. When offered with sufficient support to facilitate real patient acceptance, such an approach may be tremendously helpful in reducing the likelihood of diabetes.

Rizkalla and colleagues[13] studied the effects of low vs high glycemic index diets. The study authors noted that a low glycemic index diet may be more beneficial for healthy overweight individuals than for overweight diabetic individuals, who may not show the same degree of reduction in fat mass. In 11 healthy, overweight persons, postprandial leptin levels were lower following a 1-month period of low vs high glycemic index diet, whereas no modification in leptin levels could be detected in 12 men with type 2 diabetes. The study authors suggested that the lack of a leptin-lowering effect may be related to a smaller reduction in body fat in the diabetic group. This would, however, tend to suggest a greater satiety-promoting effect of the low glycemic index diet in persons with diabetes, suggesting that our understanding of the relationships between diet, obesity, and adipokines remains incomplete.

Rivellese and colleagues[14] treated 10 type 2 diabetes patients with high monounsaturated vs saturated fat diets for 3-week periods. Although a euglycemic hyperinsulinemic clamp did not show improvement in insulin sensitivity with the monounsaturated fat diet, following a standard fat-rich meal this approach reduced the postprandial increases in both cholesterol and triglycerides of small very low-density lipoprotein particles -- the more atherogenic components.

Radulian and colleagues[15] randomized 64 obese, hypertensive, hypertriglyceridemic persons to a carbohydrate-restricted diet or a calorie- and fat-restricted diet. Patients on the low-carbohydrate diet lost more weight during the 6-month study (7 kg vs 5 kg). There was also a greater decrease in mean triglyceride levels in the low-carbohydrate group (from 229 to 163 vs from 206 to 186). The mean fasting glucose level also decreased more in the low-carbohydrate group. These data suggest that there is an advantage to dietary carbohydrate restriction. Ginis and colleagues[16] treated 58 diabetic persons with a Greek Mediterranean diet, which is high in fiber, mono- and polyunsaturated fatty acids, and complex carbohydrates, for 28 days. Subjects had a decrease in A1C from 7% to 6.5%, a decrease in ferritin levels from 110 ng/mL to 85 ng/mL, a decrease in fibrinogen levels from 393 to 374, and increased magnesium levels, suggesting that this diet may potentially benefit diabetics.

In a study involving 274 adolescents, Basu and colleagues[17] measured urinary 8-iso-PGF2 alpha (an isoprostane derived from the free-radical oxidation of arachidonic acid, and reflecting oxidative stress in the nonenzymatic oxidation of cell-membrane lipids) and 15-keto-dihydro-PGF2 alpha (a PGF2 alpha metabolite reflecting cyclooxygenase-mediated inflammation). Both indicators correlated with BMI, waist circumference, and fasting insulin. Thus, obesity already shows a relationship with oxidative stress and inflammation in adolescence, and dietary (as well as overall lifestyle) intervention should begin in childhood to reduce the development of obesity and the metabolic syndrome.

Fatema and Miller,[18] noting that there is an association between moderate alcohol consumption and reduced cardiovascular disease risk, studied the effect of the consumption of 20 g of alcohol (2 standard drinks) 1 hour before a meal on the glycemic and insulinemic response to that meal. For 8 healthy normal-weight persons consuming beer, wine, or gin vs water 3 hours following a standard breakfast, the area under the response curve for plasma glucose after the subsequent meal was reduced 25%, 23%, and 19% for the alcoholic beverages, without reduction in insulin levels. This suggests a mechanism by which alcoholic beverages may contribute to a reduced risk for cardiovascular disease, as well as a potential treatment for persons with diabetes!

References

1. Hanusch-Enserer U, Cauza E, Penka M, et al. Effect of severe weight loss on non-traditional cardiovascular risk factors in morbidly obese subjects. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 1136.

2. Rosenstock J. The potential of rimonabant in prediabetes: pooled 1-year results from the RIO-Lipids, RIO-Europe and RIO-North America studies. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 37.

3. Knerr K, Herder C, Rose B, et al. Glucose-induced cannabinoid receptor 1 repression in peripheral blood leukocytes. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 735.

4.Wang T-F, Pei D, Li J-C, Kuo S-W. Effects of sibutramine in overweight poorly controlled Chinese female type 2 diabetic patients: a randomized, double-blind, placebo-controlled study. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 802.

5. McCormack JG, Kuhn-Wache K, Freyse E-J, et al. Effects of the short-acting dipeptidyl peptidase IV inhibitor PSN9301 and metformin alone and in combination on glucose tolerance and body weight in the fa/fa Zucker rat, and in a polygenetic rat model of diabetes. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 789.

6. Wintle M, Kim D, Poon T, et al. Improved glycemic control and reductions in body weight in exenatide-treated subjects with type 2 diabetes with modest hyperglycemia. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 793.

7. Ratner RE, Stonehouse A, Gao H-Y, Poon T, Kim D. Exenatide maintains glycemic control for 2 years in patients with type 2 diabetes: data from an ongoing, open-label study. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 795.

8. Hermansen K, Tame SC. Insulin detemir results in less weight gain than NPH insulin when added to oral agents in type 2 diabetes, with this advantage increasing with baseline obesity. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 5.

9. Pavlatos S, Kokkinos A, Tentolouris N, et al. Changes in plasma total ghrelin levels after high-protein and high-fat isoenergetic meals in lean and obese women. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 222.

10. Nannipieri M, Bonotti A, Anselmino M, et al. Adiponectin expression in subcutaneous and visceral adipose tissue in severely obese subjects with or without diabetes. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 88.

11. Invitti C, Gilardini L, Mc Ternan P, et al. Adiponectin is a marker of insulin resistance and increased risk of atherogenesis in obese children. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 90.

12. Lindstrom J, Peltonen M, Uusitupa M, Tuomilehto J. Dietary composition and development of type 2 diabetes in high-risk subjects -- The Finnish Diabetes Prevention Study (DPS). Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 743.

13. Rizkalla SW, Kabir M, Boillot J, Vinoy S, Slama G. Differential regulation of leptin levels by carbohydrate foods: control trials in diabetic and non diabetic men. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 736.

14. Rivellese AA, Annuzzi G, Giacco R, et al. Effects of saturated and monounsaturated fat rich diets on insulin sensitivity and meal lipid tolerance in type 2 diabetic patients. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 740.

15. Radulian G, Rusu E, Constantin C. A low carbohydrate compared with a low fat diet in elderly patients with type 2 diabetes mellitus. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 741.

16. Ginis A, Tzioras C, Panayotou N, Phenekos C. The effect of Greek mediterranean diet on trace elements and blood coagulation factors in type 2 diabetic patients. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 744.

17. Basu S, Steffen LM, Vessby B, et al. Obesity in 15-year-old adolescents is related to oxidative stress and inflammation. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 366.

18. Fatema K, Miller JB. The effect of 'pre-dinner drinks' on postprandial glycaemia and insulinaemia. Program and abstracts of the European Association for the Study of Diabetes 41st Annual Meeting; September 12-15, 2005; Athens, Greece. Abstract 737.

Zachary T. Bloomgarden, MD Mount Sinai Medical School
Posted Nov 2005