TY - JOUR T1 - Antidepressant use and violent crimes among young people: a longitudinal examination of the Finnish 1987 birth cohort JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - 12 LP - 18 DO - 10.1136/jech-2016-207265 VL - 71 IS - 1 AU - Elina Hemminki AU - Marko Merikukka AU - Mika Gissler AU - Kristian Wahlbeck AU - Jukka Savolainen AU - Tiina Ristikari AU - Mikko Aaltonen Y1 - 2017/01/01 UR - http://jech.bmj.com/content/71/1/12.abstract N2 - Background The use of antidepressants, especially selective serotonin reuptake inhibitors (SSRIs), has been questioned due to poor efficacy and safety. We examined whether young violent offenders were more likely antidepressant users prior to their first violent offence than other young persons.Methods The study is a follow-up of children born in Finland in 1987 (n=59 120), linking national registers to each other using personal identity codes. Data on psychotropic drug use came from a register of reimbursed drugs and data on crimes from a register on court convictions (after the age of 14 years). Participants were followed until the age of 18 years, and for some analyses until the end of the follow-up (mean 21 years). To adjust for differences in background characteristics, regression analyses for antidepressant use were made, using the no-conviction group as the reference.Results Proportions of young people convicted by the age of 18 years were: 5% of boys (1.7% for violent crimes) and 1% (0.5%) of girls. Antidepressant use (both overall and for SSRIs) prior to violent crime was more common among those convicted than among those without convictions. Among boys with repeated violent crimes, it was also more common than among boys with non-violent crimes. Adjustment for differences in background characteristics decreased the associations between antidepressant use and violent crime, but did not eliminate them.Conclusions The results add further evidence for caution in prescribing antidepressants among young persons. It also calls for a reanalysis of violence measures in the original trial data. ER -