Context: The US FDA has issued an advisory warning of a possible link between antidepressant treatment for paediatric patients with major depressive disorder (MDD) and an increased risk of suicidal behaviour. A large database of paid health insurance claims for adolescents with MDD provided the opportunity to examine this possible relationship.
Objective: To examine the potential empirical link between antidepressant treatment and suicide attempts among adolescents aged 12-18 years using a community sample of managed care enrollees across the US.
Design: A retrospective longitudinal cohort using paid insurance claims for all healthcare and prescription fills for adolescents who were newly diagnosed with MDD and had at least 6 months of follow-up data. A multivariate Cox proportional hazards regression analysis was used to test the hypothesis that antidepressant use increased the risk of suicide attempt, adjusting for propensity for allocation to each treatment group and for demographic and clinical characteristics.
Setting: Managed care plans including both commercial and Medicaid plans in the east, midwest, south and western regions of the US from January 1997 to March 2003.
Participants: All adolescent insurance members aged 12-18 years at first diagnosis of MDD.
Main outcome measures: Suicide attempts as indicated by medical utilisation with International Classification of Diseases (9th edition) [ICD-9] or 10th edition (ICD-10) codes in any healthcare setting or by any covered provider.
Results: 24 119 adolescents met inclusion criteria (63% female). Crude suicide attempt rates ranged from 0.0-2.3% by index treatment group. Treatment with SSRIs (hazard ratio) [HR] = 1.59; CI 0.89, 2.82), other antidepressants (HR = 1.03; CI 0.43, 2.44), or multiple antidepressants (HR = 1.43; CI 0.70, 2.89) after index MDD diagnosis resulted in no statistically increased risk of suicide attempt. Treatment with antidepressant medication for at least 180 days (6 months) reduced the likelihood of suicide attempt compared with antidepressant treatment for <55 days (8 weeks) [HR = 0.34; CI 0.21, 0.55]. Other variables that were independently associated with greater risk of suicide attempts included female gender, severity of illness indicators, younger age at time of MDD diagnosis, and living in the midwest or west.
Conclusions: Antidepressant medication use had no statistically significant effect on the likelihood of suicide attempt in a large cohort of adolescents across the US after propensity adjustment for treatment allocation and controlling for other factors. The relationship between suicidal behaviour and antidepressant medication use is complex and requires further investigation.