Metformin has been used off-label for weight loss in psychiatry for many years, much of it for help with adult weight gain as a side effect of atypical antipsychotics. However, most trials of metformin for weight loss were actually conducted on the child and adolescent population.
This article will look at the quality of the data in our pediatric population of three studies, and then come up with some recommendations for your practice.
The first study was a randomized, double-blind, placebo-controlled trial of 38 children between 10–17 years old with various psychiatric disorders and who had experienced more than 10% weight gain in less than 1 year of olanzapine, risperidone, or quetiapine treatment. Participants were randomly assigned to either metformin (850 mg BID) or placebo. After 16 weeks, patients on placebo gained an average of 4 kg, while those on metformin lost a little weight (-0.13 kg) (Klein D et al, Amer J of Psychiatry 2006;163(12):2072–2079).
In a second study, a 12-week double-blind placebo-controlled trial of 32 children and adolescents with schizophrenia or schizoaffective disorder on risperidone received either metformin (500 mg BID) or a placebo. This study showed no benefit to the use of metformin for weight control (Arman et al, Saudi Medical Journal 2008;29(8):1130–1134).
A third study was a 16-week double-blind, randomized controlled trial of metformin (500 mg BID for kids 6–9 years and 850 mg BID for kids 10–17 years) vs placebo in a group of 61 children and adolescents with autism spectrum disorder, who were on various atypical antipsychotics. After 16 weeks, they found that metformin lowered BMI z-scores significantly more than placebo (-0.08 for metformin vs +0.02 for placebo). In terms of raw weight scores, those on placebo gained an average of 2.80 kg (1.90 kg to 3.70 kg, 95% CI), while those on metformin gained on average only 0.07 kg, with some noted variability (-0.88 kg to 1.02 kg, 95% CI). Metformin was generally well-tolerated, but some participants experienced significantly more GI side effects compared to those taking placebo (Anagnostou E et al, JAMA Psychiatry 2016;73(9):928–937).
It is important to note that when using metformin with children, the goal is not weight loss, but rather weight control—to maintain only a small amount of weight gain and to keep them on their BMI growth curve trajectories. Although there is a need for more research, the preliminary data above so far supports the use of metformin for weight control in children and adolescents using atypical antipsychotics.
So, should you use metformin at this point?
CCPR got some advice from Bradley Engwall, MD, a private practice child psychiatrist in Berkeley, CA, who often uses metformin to control weight while prescribing atypical antipsychotics. The following are his answers to some questions you might have:
When should you use metformin in children?
Dr. Engwall: About 50%–75% of my patients develop weight issues while on atypical antipsychotics, and I regularly prescribe metformin. Before going to the metformin, however, my initial strategy is to try and use an antipsychotic that is least likely to cause weight gain. I start with aripiprazole as it’s usually on the formulary and has less documented weight gain impact than other atypicals. If significant weight gain ensues, lurasidone is more weight-neutral, and if an atypical is needed, it is often the next choice. Depending on the case/indication and insurance restrictions around lurasidone access, we may use ziprasidone or risperidone. I also will try to have patients go through diet modification with an exercise plan before prescribing metformin.
How should you use metformin in children?
Dr. Engwall: I recommend that, when available, you use the once-a-day extended release metformin tablets, dosing with dinner or at bedtime. Start at 500 mg, then have parents titrate up to 750 mg, and then 1000 mg at 1-week intervals over the course of a month, and as they see it benefiting their child. I generally use 500 mg pills. They are also available at 750 mg and 1000 mg, but these are typically too large for most children to swallow. The highest dose I will titrate towards is 1000 mg/day for younger children, and 1500 mg–2000 mg/day for tweens and adolescents.
What are typical side effects, and what parameters do you monitor?
Dr. Engwall: Typical initial side effects are nausea, upset stomach, and loose stools. I follow height, weight, BMI, fasting blood sugar, HbA1c, triglycerides, and LFTs—typical parameters for monitoring anyone on atypical antipsychotics. Due to metformin use, I also often find problems with B12 absorption and folate absorption. It is useful to get baseline B12 and methylmalonic acid levels, as well as folate and homocysteine levels.
If you are starting out with a folate deficit, it will become even more important to supplement with folate. I recommend using l-methylfolate instead of just folate, as l-methylfolate is the only form of folate to go through the blood-brain barrier. As tolerated, you can supplement with l-methylfolate at 400 micrograms/ day or 1000 micrograms/day. At about 2 months in, you should recheck the levels of B12, methylmalonic acid, homocysteine, and folate to confirm that these patients are receiving sufficient amounts.
If I see a change, I adjust as needed and add B12 and l-methylfolate. A serious potential side effect to metformin, especially at higher doses, would be hypoglycemia. It’s uncommon, and I have never seen it with our ASD kids—since they are eating so much, this is a very rare issue.
Any concluding thoughts on using metformin for children?
Dr. Engwall: Yes. Consult with the patient’s pediatrician or primary care physician early on in treatment. Most pediatricians and family practice doctors think of metformin as only for use in cases of diabetes. They are typically unaware of the emerging literature supporting the use of metformin for these kids on atypicals. So, it’s important to let them know about this up front—there is an educational piece here, and I typically explain the reasons for starting metformin and share the supporting literature. The judgment call to start metformin is not hard—you usually see massive weight gain quickly with atypicals. It is infrequent, but unfortunately some patients’ weight gain simply does not respond to the metformin. For those patients, I consider the risks and benefits of continuing the atypical antipsychotic vs other alternatives.