Temporal trends of psychiatric disorders incidence by sex, education and immigration status among young and middle-aged adults in Sweden, 2004–2019

Objective: To explore temporal changes in incidence of major psychiatric disorders across sociodemographic subgroups in Sweden. Methods: This population-based open cohort study included all individuals born during 1958–1994 and residing in Sweden at any time during 2004–2019. We identified psychiatric disorders registered in inpatient and outpatient specialist care. We calculated person-years from the inclusion until diagnosis of psychiatric disorder of interest, death, emigration or December 31, 2019, whichever occurred first. Combining sex (female, male), education (low, medium, high) and immigration status (first- or second-generation immigrant, native), we created a variable with 18 strata. Average annual percent changes (AAPCs) in age-standardized incidence rates (ASIRs) were estimated using Joinpoint regression. Results: A total of 5,051,875 individuals aged 25–61 years were followed for ≈ 56–58 million person-years. First-generation immigrants generally had lower overall ASIRs than second-generation and natives with more pronounced differences among persons with low education and females. While ASIRs of autism spectrum and other pervasive developmental disorders (AAPC 11.8%, 95% CI: 9.5, 15.8), as well as attention deficit hyperactivity disorder and conduct disorders (18.8%, 16.6, 25.0) rose over time, other psychiatric disorders were stable or had decreasing temporal changes (AAPC ranged from 0% for substance use disorders to -5.7% for schizophrenia/acute and transient psychotic disorders). First-generation immigrants generally experienced more favourable changes (i.e. more decreases or less increases) in ASIRs and this was most evident among those with low education. Conclusions: While incidence of psychiatric disorders in inpatient and outpatient specialist care generally declined during 2004–2019, there were important sociodemographic variations in temporal changes. Clinical trial number: Not applicable.