Weight-loss drugs not enough in type 2 diabetes

OB/GYN News, Nov 15, 2004 by Timothy F. Kirn

Simply prescribing a weight-loss drug for a patient with type 2 diabetes, and doing nothing more, is not likely to be adequate management, according to an analysis of the published literature by researchers at the Centers for Disease Control and Prevention.

Drug treatment does produce weight loss in diabetic patients, but it is less than the weight loss in nondiabetic patients, reported Susan L. Norris, M.D., of the CDC's division of diabetes translation, Atlanta.

Moreover, the studies did not last long enough to show that the weight loss produced is more than temporary, offering real, long-term benefit.

In reviewing the literature for weight-loss drug studies involving persons with type 2 diabetes, Dr. Norris and her colleagues identified 14 studies involving three drugs that were randomized and placebo controlled and that contained enough data so that the results could be combined. The drugs were fluoxetine, orlistat, and sibutramine. The studies included a total of 2,231 patients.

Metformin was specifically not included in the analysis, even though it can produce weight loss, because it is usually prescribed for glycemic control, Dr. Norris said.

In the pooled data from the studies, weight loss with the drugs ranged from 2.6 kg to 4.5 kg, or 2%-3% of initial body weight. The report does not conclude that any one of the three drugs showed any superiority over the others.

That degree of weight loss is smaller than is reported for the drugs in nondiabetic populations, where the range is 2-10 kg, Dr. Norris said.

Fluoxetine and orlistat led to significant lowering of hemoglobin [A.sub.1c] levels, while sibutramine showed possible lowering. Fluoxetine, which had the greatest effect, reduced glycated hemoglobin by 1% at 8-16 weeks, and one longer study showed that this reduction was maintained at 1 year.

A 1% reduction in glycated hemoglobin has been shown to reduce the risk of microvascular complications in large trials. But those studies used hypoglycemic agents, and fluoxetine may act differently, Dr. Norris said.

Modest weight loss of the degree found in the analysis has been shown to have health benefit in general populations, Dr. Norris noted. But the studies of these drugs and weight loss had follow-up periods that lasted only as long as 52 weeks, and the risks and benefits of taking these agents longer are not known, either in diabetics or nondiabetics.

Studies of behavioral interventions suggest that many individuals put the weight back on when the intervention is stopped.

Given the degree of weight loss, it would appear that behavioral interventions are as effective as pharmacotherapy, Dr. Norris said. Moreover, there are studies with general populations that suggest combining lifestyle changes with medication improves treatment results.

Dr. Norris's article also noted that the majority of the studies in their analysis were sponsored by drug companies and therefore could be biased.

She and her colleagues tried to obtain unpublished studies from the manufacturers and researchers but received none, she said.

BY TIMOTHY F. KIRN

Sacramento Bureau