Antipsychotics are effective at many things, including inducing weight gain. For example, a study of 90 adolescents started on atypical antipsychotics for a variety of diagnoses found significant weight gain in 70% of patients, on average about 12 pounds (Mareno C et al, Bipolar Disord 2010;12(2):172–184). In addition to the metabolic risks, obesity confers additional risks of low self esteem and behavior problems, asthma, irregular or early menarche, polycystic ovarian syndrome, slipped capital femoral epiphyses, obstructive sleep apnea, gallstones, and increased mortality 20 and 32 years later (Jain A, What works for obesity? London: BMJ Publishing Group; 2004). Furthermore, adolescents hate weight gain, and it is one of the reasons they often give for not complying with prescribed medication.
Lifestyle interventions are often a hard sell, and the possibilities of a pharmacologic intervention are enticing. But do they work? Several options are reviewed below.
Given the fact that a large proportion of our patients with ADHD on stimulants are failing to gain weight, adding a stimulant to an antipsychotic seems like a logical approach to control weight gain. Unfortunately, most studies do not bear the practice out. Penzner et al used the naturalistic model in the SATIETY study to look at patients without mood or psychotic disorders who were started on antipsychotics to control aggression and oppositional behavior in the context of ADHD, oppositional defiant disorder (ODD), pervasive developmental disorders (PDD), or conduct disorder. Seventy-one of these patients were on stimulants and 82 were not. There was no significant difference in weight gain between groups given antipsychotics alone and those given antipsychotics and stimulants, although those on stimulants prior to starting antipsychotics were more likely to be at a lower weight to start with (Penzner et al, J Child Adolesc Psychopharmacol 2009;19(5):563–573).
Metformin (Glucophage) is an oral antiglucose agent used in the treatment of diabetes. It is generally well-tolerated but may cause GI distress. Metformin is not approved for weight loss, but has shown some effectiveness when used for it.
A review of studies on metformin for obesity (not related to antipsychotic use) found six studies in adolescents, five of them randomized controlled trials (RCTs). They concluded that metformin, 500 to 1,000 mg twice daily, was effective in five out of six of trials, and resulted in 0.8 to 2.7 kg/m2 reduction in BMI (3.2 kg/m2 for the open label study) (Rogovik et al, Drugs 2010;70(3):335– 346).
Another study reports a case of an adolescent girl maintained on clozapine (Clozaril) who was treated with metformin, 500 mg twice daily, resulting in a five-pound weight loss. However, total weight gain after treatment with clozapine for 10 months was 49.6 pounds and her weight had remained on “a steady trajectory upwards,” so the case report is more significant for the deceleration of weight gain after starting metformin than weight loss itself (Weaver et al, J Child Adolesc Psychopharmacol 2010;20(2):153–157).
In a 16 week RCT of weight control in adolescents on antipsychotics using metformin doses up to 2,000 mg daily, a significant difference in weight gain and a decrease in waist circumference were found in the metformin group compared with the placebo group (Klein DJ et al, Am J Psychiatry 2006;163(12):2072–2079). In this study, too, metformin was started after the antipsychotic—it was helpful in reducing the rate of weight gain and stabilizing the weight, rather than weight reduction. It may (or may not) also prevent glucose intolerance associated with antipsychotics.
An open label study found no significant weight loss after eight weeks on metformin, but the study was small and the metformin was added after the fact (Shin et al, J Child Adolesc Psychopharmacol 2009;19(3):275–279). Studies in adults on antipsychotics have been mixed, with some finding significant benefit and some finding no benefit. (For example, see Jain op.cit or Maayan L et al, Neuropsychopharmacol 2010;35(7):1520–1530.)
It may make sense to start metformin, particularly in patients with additional risk factors for metabolic syndrome. Weight loss may not be the most important outcome: rather, deceleration of weight gain and weight stabilization, decreased triglycerides, and potentially (although not demonstrated) reduction in the risk of metabolic syndrome. It would be nice to see a RCT of metformin started concurrently with antipsychotics to see if weight gain can be prevented.
Orlistat (marketed over the counter as the weight-loss drug Alli) is now the single FDA-approved antio-besity agent available in the U.S. and is approved for patients 12 and over. It impairs fat absorption by inactivating gastric and pancreatic lipases. A meta- analysis of orlistat and sibutramine (which was recently pulled from the market) for weight reduction in obese adolescents (but not those on antipsychotics), found two RCTs comprising 573 adolescents, dosed daily with orlistat 120 mg for three and six months. The authors calculated a modest reduction of 0.83 kg/m2 in BMI, consistent with other analyses. The effect size was 0.24, and therefore of questionable clinical significance. There were no significant differences in serum lipids, glucose, or insulin. The patients suffered relatively frequent GI side effects, no doubt from the presence of undigested fat in the lower GI tract (Viner RM et al, Obes Rev 2010;11(8):593–602).
Conjugated Linoleic Acid
A natural fatty acid found primarily in meat and dairy products, conjugated linoleic acid (CLA) may be of use in the treatment of obesity. Adult studies have been mixed. In an RCT of pediatric patients ages six to 10 with obesity, 28 kids were given 3 g of a proprietary mixture of CLA called Clarinol daily for around seven months. Thirty-one kids were given a fat based placebo (sunflower oil). The study authors found a reduced accretion of fat mass in the CLA group—that is, the children gained weight, but not as much as the placebo group did, and not as much as their weight trajectory would have predicted. There was no effect on fasting glucose, lipids, or insulin. One subject had an elevation in liver enzymes (Racine LM et al, Am J Clin Nutr 2010;91 (5):1157–1164).
In a review of an Italian study of 53 obese children ages five to 18, those who were given two to three grams of glucomannan fiber daily were found to have decreased excess bodyweight, cholesterol, and triglycerides after four months. This review also reported a second positive study on glucomannan in 60 children under age 15 (Rogovik op.cit).
With the fairly modest effects of the pharmaceuticals described previously, most experts agree that “lifestyle interventions” are the mainstay of obesity treatment. (They also enhance the weight loss achieved with pharmaceutical interventions.) It appears that both low-carbohydrate and low-fat diets are effective and safe for adolescents. Additional interventions we can make as clinicians are as follows (Rao G, Am Fam Physician 2010;81(12):1449–1455):
1. Before starting antipsychotics, make sure our patients understand that they increase hunger and that they should plan accordingly, using the following suggestions:
a. eat breakfast daily
b. limit sweet beverages (soda, fruit juice, etc) to one serving daily
c. limit fast food to less than once a week
d. eat dinner together as a family
e. limit TV viewing
f. get daily exercise
g. switch to low calorie snacks
2. Regularly measure weight, BMI, and waist circumference.
3. Assist our patients in making and maintaining these choices by referring to specialists as needed: nutritionists, physical therapists, personal trainers, and/or counselors specializing in weight loss. The USDA has two excellent websites as well: www.mypyramind.gov and www. nutrition.gov, which includes an interactive program to analyze the individual’s diet.