Behavioural outcomes of treatment with selective serotonin reuptake inhibitors

Sammanfattning: Mood and anxiety disorders are some of the biggest contributors to morbidity worldwide, and may be lethal. Appropriate treatment is therefore paramount. Antidepressant medications constitute the primary pharmacological treatment for these disorders, with selective serotonin reuptake inhibitors (SSRIs) as the most common type in several Western countries. While developed to treat disorders that increase the risk of violence and suicide, there is concern that SSRI treatment may in itself increase the risk for these behavioural outcomes, especially among young people. The overarching aim of this thesis is therefore to contribute to the understanding of the risks and benefits of treatment with SSRIs in relation to severe behavioural outcomes in different age groups, including when SSRIs are combined with other central nervous system (CNS) drugs. We also document antidepressant prescription patterns in young individuals – the age group where the balance between benefits and risks of antidepressant treatment is least clear. In study I, we described the prevalence of antidepressant use and polypharmacy of CNS drugs with antidepressants over time in children, adolescents, and young adults living in Sweden. We found that, over time, there was an increasing trend in antidepressant use and an increase in the co-prescription of antidepressants with other CNS drugs. We also found that antidepressant users had higher likelihood than population controls of collecting other CNS drug classes additionally to antidepressants. In Study II, we investigated the hazard of conviction for violent crimes during treatment with SSRIs, including in different time periods since start and end of treatment. In a follow-up of up to 8 years, we found that the hazard of violent crime was statistically significantly elevated throughout treatment periods, and for up to 12 weeks after the end of treatment. This was true in youths as well as older adults, which adds to prior research that has found elevated risk of aggression outcomes during SSRI treatment in young adults but not older individuals. In Study III, we explored the incidence rate of suicide attempts or deaths (suicidal behaviour) in time periods before and after initiation of SSRI treatment. We found that the month immediately prior to SSRI treatment initiation was associated with the greatest incidence rate of suicidal behaviour, that treatment periods up to one year after treatment initiation were associated with lower incidence rate compared to the month immediately before initiation, and that the incidence rate gradually decreased over treatment time. However, all treated periods had higher incidence rates than the month one year before treatment start. These patterns were consistent across age categories, including among children and young adults. In Study IV, we applied a data-driven screening approach to compare the incidence rate of suicidal behaviour in periods after and before initiation of additional CNS drugs during continuous SSRI treatment. We found several drugs that were associated with statistically significantly reduced incidence rate of suicidal behaviour when initiated during SSRI treatment, and only two associated with increased risk of suicidal behaviour. We found no evidence of harmful effects of combining SSRIs with additional CNS drugs. Many of the signals of reduced suicidal behaviour correspond to prior evidence; novel signals could be further investigated to evaluate the use of these drugs concurrently with SSRI treatment. In conclusion, the presented thesis has documented: the increasing prevalence of antidepressant use and polypharmacy of antidepressants with other CNS drugs in young individuals resident in Sweden; the associations between SSRI use and violent crime and suicidal behaviour; and the impact of initiating other CNS drugs during SSRI treatment on the risk for suicidal behaviour. The findings are expected to help guide future research and clinical decision making.