Register-based studies of sex steroid hormones and psychiatric disorders

Sammanfattning: The focus of this thesis is the interplay between sex steroid hormones, psychiatric- and neurodevelopmental disorders and adverse behavioral outcomes. Sex steroid hormones are mostly known for their role in sexual differentiation and reproduction, but have been suggested to be involved the etiology of several psychiatric disorders and in adverse behaviors that show prominent sex differences in prevalence. These prevalence differences suggest that sex steroid hormones could play a role in the pathogenesis of psychiatric disorders. A greater understanding of causal risk factors for psychiatric disorders and adverse behaviors can eventually lead to better possibilities for prevention and treatment. In Study I-III, we explored the association between conditions affecting androgen levels, criminal behavior and psychiatric disorders. All three studies were based on Swedish population-based registers and had matched cohort designs. In study I, we identified individuals diagnosed with congenital adrenal hyperplasia (CAH) and polycystic ovary syndrome (PCOS). CAH is a condition where androgens are increased prenatally, whereas PCOS is associated with increased circulating androgen levels. The outcome was criminal convictions. We found that PCOS and was associated with an increased risk of criminal convictions, but there was no difference between individuals diagnosed with CAH and the general population in prevalence of convictions. In study II, we examined the association between Klinefelter syndrome and bipolar disorder, schizophrenia, autism and attention-deficits/hyperactivity disorder (ADHD). Men with Klinefelter syndrome have an extra X-chromosome resulting in a 47,XXY karyotype, which often cause hypogonadism. Compared to men from the general population, men with Klinefelter had almost four times higher risks of schizophrenia and bipolar disorder and an approximately six times higher risk of autism and ADHD. In Study III, we compared individuals diagnosed with congenital hypogonadotropic hypogonadism or delayed puberty with controls from the general population. Hypogonadotropic hypogonadism is a rare genetic disorder characterized by incomplete or absent puberty and infertility, whereas delayed puberty are diagnosed in the 2.5 percent of the population that has the latest puberty onset. The outcome was autism, ADHD and intellectual disability. Both individuals with hypogonadotropic hypogonadism and delayed puberty had increased risk for being diagnosed with autism, ADHD and intellectual disability. In study IV, we instead studied the risk of adverse outcomes in individuals with autism and ADHD-symptoms. Some previous studies have indicated that neurodevelopmental disorders increase the risk of being sexually victimized. We investigated this association in a cohort of twins using a genetically informative prospective design. We found that parent-reported autism- and ADHD-symptoms was associated with an increased risk of reporting sexual victimization. However, when controlling for other neurodevelopmental symptoms, the unique part of autism and ADHD did no longer predict sexual victimization. Quantitative genetic analyses showed that the association between the general neurodevelopmental symptoms and sexual victimization risk was due to shared genetics. Study V and VI concerns the relationship between hormonal contraception (HC) use and risk of suicidal behavior. Hundreds of millions of women worldwide use HC, but there are reports about potential mental side effects. In Study V, we assessed associations between different types of HC and the risk of suicidal behavior in a nationwide cohort study. In the main analyses, we found a protective association between most types of HC use and suicidal behavior. However, non-oral progestin only contraceptive use was associated with an increased risk. In analyses stratified on age, use of any class of HC was associated with a higher risk of suicidal behavior during adolescence, but not in older women. In women using non-oral progestin-only, we found an increased risk of suicidal behavior in all ages. To control for unmeasured confounders, we also performed within-individual analyses, comparing the risk in exposed to unexposed periods within the same individual. In these analyses, the protective effect of HC use on suicide was attenuated and for oral combined formulations, there was a small increased risk of suicide attempts. The risk associated with non-oral progestin-only contraceptive use with suicidal behavior was similar in the population-based and intra-individual analysis. In study VI, we wanted to better understand the mechanisms behind the association between HC use and suicidal behavior. Therefore, we explored how a range of risk factors for suicide identified in population-based registers was associated with HC use at different ages. At age 13-14 years, eleven of the thirteen risk factors were associated with an increased risk of oral combined HC use. The associations diminished with age and were essentially non-existent for HC use above age 26. Compared to the estimates for oral combined formulations, there were a consistent pattern of higher OR for all risk factors on progestin only formulations use at all ages, especially for non-oral formulations. We also performed quantitative genetic analyses. Overall, additive genetic effects accounted for the main part of the correlation between risk factors and all forms of HC use. The main findings from this thesis suggest that sex steroid hormones can influence adverse behaviors, psychiatric and neurodevelopmental disorders and that neurodevelopmental disorders can increase the risk of negative outcomes. Conditions associated with increased circulating androgens, but not prenatal androgen exposure, might increase the risk of criminal behaviors. Klinefelter syndrome and hypogonadotropic hypogonadism, both conditions leading to decreased androgen levels, are associated with psychiatric and neurodevelopmental disorders. Neurodevelopmental symptoms are in turn associated with increased risk of sexual victimization, due to shared genetics. Use of HC (especially progestin-only formulation) might increase the risk of suicidal behavior, although the association could partly be explained by differences in pre-existing risk factors between young women using HC and non-users.