New Data Show Antipsychotics May Increase Risk for Diabetes in Children: What Now?

Would it surprise you to learn that antipsychotics like Risperdal, Abilify, and Seroquel are the second most prescribed class of medications for behavioral problems in children and adolescents in the Medicaid program, second only to stimulants like Concerta or Adderall, which are commonly prescribed for ADHD? In recent years, these medications have become the in-vogue approach to managing serious disruptive and aggressive behaviors in children due to their immediate and powerful sedative and mood stabilizing properties. They’ve also become a frequent add-on to treating children diagnosed with depression or ADHD; by 2008, 1 in 3 youths receiving antidepressants, and 1 in 5 youths receiving stimulants in the Medicaid program were receiving an antipsychotic at the same time. For some of our highest risk children, like youth in foster care who so often exhibit disruptive and oppositional behavior as a manifestation of their trauma, the dependence on antipsychotics to control their behavior has become epidemic.

Antipsychotics are prescribed to children in foster care fifty percent of the time they are prescribed medications, principally to address disruptive or oppositional behaviors.

Over the last decade, our interdisciplinary research team at PolicyLab has examined the growing use of antipsychotics among children and adolescents in our nation’s Medicaid program with the intent of shining a spotlight on the extent of such use and the resultant practice and policy implications. This month, in JAMA Pediatrics, we released the largest study to date documenting the significant health risks associated with these powerful drugs. Our research suggests that the initiation of antipsychotics among our nation’s youth may elevate their risk not only for significant weight gain, but also for Type II diabetes by nearly 50 percent. Moreover, among those who were also receiving antidepressants, the risk may double.

These new findings should give us pause.  With such vast numbers of children being exposed to these medications, the implications for potential long-lasting harm can be jarring. The inclination might be to blame the doctors or pharmaceutical companies and to impose strict restrictions on the use of these drugs. However, as a pediatrician who has traversed this issue clinically with children and families at CHOP and views this from a population health perspective, I advocate for a more balanced approach.

Mainly, we should acknowledge that the very high level of use of these medications reflects a growing demand to address very challenging behaviors in children.

These behaviors are often not well addressed within public systems that lack the appropriate services or workforce to provide supportive and evidence-based counseling as an alternative. Even in an environment of readily available counseling services – so we would hope – there will always be children and adolescents who need access to these types of medications, even if only for a short period of time. This is why when I’m asked what the magnitude of antipsychotic use should be among this population, my answer is “less,” but it is never zero. For the same reason, state responses that have employed strict preauthorization policies may be too blunt a response for an issue that needs a more fine point.

My growing sense is that we need to incorporate these new revelations about the risk for diabetes and its potentially long-lasting complications through a more thoughtful consideration of the true risks and benefits of prescribing an antipsychotic to a child. Although adding an antipsychotic to a drug regimen of an individual child who is already receiving an antidepressant may double the child’s risk for diabetes, at a population level, the risk still remains relatively small. That is because diabetes is still uncommon overall in this young population. Among children 11-18 years of age, the underlying risk for diabetes for those who were not using antipsychotics in our study was 1 in 400, rising to 1 in 260 among those who initiated antipsychotics, and 1 in 200 among those who initiated while also receiving an antidepressant. For some children in immediate crisis, the benefit of the antipsychotic for acute management may still outweigh that risk.

At the same time, our research shows that the risk is not zero, and may ultimately be greater if we had followed children for longer periods of time. For that reason, we should absolutely try to reduce the number of children and adolescents exposed to these powerful medications. If possible, these medications should not even be considered until more appropriate and evidence-based counseling services, such as trauma-focused cognitive therapy, are employed to address underlying emotional trauma, which is often the root cause of these behaviors. Clinicians and families who are making medication decisions should consistently revisit the strategy by which they are addressing the challenging behaviors they are seeing. If they plan to prescribe antipsychotics for a child, they should start cautiously with the lowest dose possible, even as they more strictly monitor for early evidence of weight gain or abnormal lab tests that often predict later onset of diabetes. Once a child is on the antipsychotic drug, a plan should be agreed upon and periodically revisited to ensure that other therapeutic strategies are being maximized. That same periodic review would also seek to transition the child off the antipsychotic as soon as possible once these problems are more suitably addressed through counseling. 

This will ultimately require state Medicaid programs to offer more regular guidance and education to their providers about expectations for quality in their delivery of behavioral health services to at-risk children. They might also offer supportive resources such as consultative psychiatrists and crisis managers, who can regionally direct families and clinicians to additional resources.

While our research findings are certainly provocative, they should not be read to support a one-size-fits-all solution. We must be ever mindful that in trying to address the over-prescription of antipsychotics we should not tilt so far that we under-treat children who have serious problems for whom medication is beneficial.