Rationale:

• When lifestyle treatments are ineffective or the patient is unable to lose more weight, the addition of drug therapy may be helpful.

Evidence:

• In meta-analyses, sibutramine and rimonabant produced more weight loss than orlistat (412; 413; 414).
• Rimonabant produces weight loss and increases risk of depression and anxiety (415).
• Metformin produces weight loss and with lifestyle intervention slowed weight gain in patients receiving antipsychotic drugs (416; 417).
• Sibutramine produces weight loss in obese adults (418; 419), obese diabetics (420), and obese children (421; 422) that is dose related (423).
• Sibutramine is effective in maintaining weight loss (418; 424).
• The therapeutic effect of sibutramine can be blunted by metoprolol or diuretics but not by ACE inhibitors or calcium-channel blocking drugs (425).
• According to a meta-analysis (426), sibutramine raises blood pressure or prevents it from falling, even in hypertensive patients (427; 428; 429; 430; 431), with a weighted mean increase of 3.2 mm Hg, but reduces left ventricular mass (432).
• Orlistat produces dose-related weight loss (433) and reduces systolic blood pressure and diastolic blood pressure by a weighted mean difference of -2.5 and -2.0, respectively (426).
• Orlistat produces more weight loss than placebo in 2-year trials (434; 435; 436; 437) and in one 4-year trial (419; 438).
• Orlistat decreases LDL cholesterol more than can be accounted for by weight loss (439; 440).
• In patients with diabetes, sibutramine (412; 420; 441), orlistat (412; 437; 442; 443; 444), fluoxetine (445), and exenatide (446) produce weight loss and improvement in diabetic control.
• In a meta-analysis of studies of interventions to treat weight gain in schizophrenics, orlistat (-2.7 kg [CI, -2.3 to 3.1]) and sibutramine (-4.3 [CI, -3.6 to -4.9]) in reducing weight gain was similar to that of behavior therapy (357).
• Combining sibutramine and orlistat did not enhance weight loss (447).
• Greater than 2 kg (4 lb) weight loss at 4 weeks (423), >5% weight loss at 6 months, and maintaining >5% weight loss at 1 year are important criteria for success (448; 449).
• Diabetics treated with pramlintide experience weight loss (450).

Comments:

• The American College of Physicians clinical guidelines for management of obesity include fluoxetine and bupropion in their list of recommended drugs (451; 414), but these drugs are not recommended in other guidelines. Fluoxetine at a dose of 20 mg/d is approved by the FDA for the treatment of depression, but it is not approved by the FDA for treatment of obesity. A new drug application for fluoxetine as a weight loss drug was withdrawn by the manufacturer because there was no durability for its weight loss effect. After 20 weeks, patients treated with 60 mg/d of fluoxetine (three times the recommended dosage for treatment of depression) began to regain weight, and by the end of 1 year, their weights were not significantly different from those of subjects taking a placebo (452). Clearly, fluoxetine is not an effective drug for treating obesity beyond a few weeks and does not meet the current FDA criteria for a weight loss drug.
• Bupropion is approved by the FDA as an antidepressant. The data showing weight loss are limited to three small clinical trials that do not meet FDA criteria for considering it a safe and effective drug to treat obesity.
• Sympathomimetic drugs (phentermine, diethylpropion, benzphetamine, phendimetrazine) produce weight loss but are only approved for short-term treatment in the U.S. by the FDA and are not available in Europe.
• Phenylpropanolamine was withdrawn from the OTC market because of alleged stroke risk in women (453).
• A meta-analysis of hormone replacement therapy showed a decrease in abdominal obesity, insulin resistance, lipids, new onset diabetes, and blood pressure in women without diabetes (454). However, the increased risks of cardiovascular disease and breast cancer associated with hormone replacement therapy should be weighed against other benefits.
• Randomized, placebo-controlled 1- and 2-year clinical trials with rimonabant show persisting weight loss of approximately 8.5% (455; 456; 457). A meta-analysis of four studies shows a 4.9-kg greater weight loss than placebo (458).
• Two studies suggest that conjugated linoleic acid produces modest fat loss (459) and may reduce weight gain over the holiday season (460).
• Aminophylline cream is a topical fat reducer when applied to the abdomen (461).
• Intraparietogastric administration of botulinum toxin produces more weight loss than placebo (462).