Associations between antidepressants and risk of suicidal behavior and violent crimes in personality disorder.

BACKGROUND: Despite uncertain benefits, antidepressants are used in the management of personality disorders (PDs). We investigated the association between antidepressants and two adverse outcomes - suicidal behaviour and violent crimes - in individuals with PDs. METHODS: We used nationwide Danish healthcare registries to identify all individuals with a diagnosed PD aged 18-64 years from 2007 to 2016. Antidepressant use was identified using dispensed prescriptions. Individuals were followed up for healthcare presentations of suicidal behaviour and separately for police-recorded charges of violent crimes. We applied a within-individual design comparing rates of suicidal behaviour and violent crimes during time periods of antidepressant treatment with periods without treatment. Subgroup analyses were performed according to PD clusters, individual antidepressants, specific PDs, psychiatric comorbidities, and history of suicidal behaviour and violent crime. RESULTS: The cohort included 167,319 individuals with a diagnosed PD, 19,519 (12%) of whom were prescribed antidepressants and presented at least one outcome event during follow-up, making them eligible for within-individual analyses. Overall, we found an association with lower rates of suicidal behavior during periods of antidepressant treatment, compared with periods when individuals were not on antidepressants (incidence rate ratio 0.86, 95% CI 0.84-0.89). However, this association was modified by specific PDs, individual antidepressants, comorbidities, and past history. For violent crimes, we did not observe consistent associations in any direction. CONCLUSIONS: Antidepressants were associated with lower rates of suicidal behaviour, but less clearly in violent crimes. Types of PDs, individual antidepressants, and comorbidities modified these associations.