Pharmacologic induction of weight loss to treat type 2 diabetes

OBJECTIVE - Most individuals with type 2 diabetes are overweight, and weigh t loss for them is an important therapeutic objective. However, usual weigh t-loss strategies have generally not produced sustained weight loss. Pharma cologic agents to assist weight loss might be useful, but no long-term data on their effectiveness and safety in patients with type 2 diabetes are ava ilable. We therefore initiated a 2-year placebo-controlled trial of the wei ght-loss medications fenfluramine and phentermine in type 2 diabetic subjec ts. RESEARCH DESIGN AND METHODS - A total of 44 overweight (>120% ideal body we ight) subjects with type 2 diabetes were enrolled in a randomized, placebo- controlled, double-blind trial of fenfluramine and phentermine. All subject s received intensive nutrition counseling, an exercise prescription, and in struction in behavior modification. Subjects were randomly assigned to 20 m g fenfluramine three times a day and 37.5 mg phentermine daily (n = 23) or dual placebos (n = 21). Diabetes medications were adjusted as necessary to achieve glycemic goals. Changes in weight, glycemia, lipemia, and blood pre ssure were assessed every 2 months, as were adverse events. In September 19 97, when fenfluramine was withdrawn from the U.S. market, fenfluramine was stopped in all subjects. Thus the length of drug treatment varied, but 16 s ubjects (8 in each group) reached 12 months of treatment. Only data obtaine d before the withdrawal of fenfluramine are included in this report. RESULTS - At study termination, diabetes medications had been reduced in 1 subject in the placebo group (5%) and 11 subjects in the drug treatment gro up (52%) (P = 0.005). Drug treatment resulted in significant reductions in body weight, BMI, and HbA(1c) at all time points through 6 months. Changes in weight at 6 months were -2.7 +/- 1.4 kg (mean +/- SEM) with placebo trea tment and -9.6 +/- 1.5 kg with drug treatment (P = 0.003). Even though more subjects in the drug treatment group required reductions in diabetes medic ations, at 6 months, changes in HbA(1c) were -0.3 +/- 0.2% with placebo tre atment and -1.6 +/- 0.3% with drug treatment (P = 0.002). Easting plasma gl ucose and triglycerides were significantly reduced at some time points with drug treatment. No serious adverse events attributable to study medication s were observed. CONCLUSIONS - Premature study termination decreased the power of our study at later time points. However, our data suggest that weight loss medication s are an effective treatment for type 2 diabetes during active weight loss. Whether the benefit persists after weight loss has stopped remains to be d etermined.