Examining relationships between the Danish Composite Deprivation Index and risk of developing schizophrenia: A national multilevel analysis

Background
Few population-based multilevel analyses examining individual- and neighborhood-level risk factors for schizophrenia have been conducted.

Methods
A study cohort of all persons born in Denmark from 1990 to 1999 was followed for diagnosis with schizophrenia. Follow-up was initiated at 10th birthday and terminated at death, emigration, incident diagnosis, or 31st December 2018, whichever came first. A Danish Composite Deprivation Index was derived using 10-year weighted average neighborhood-level indicators in 1990–1999 categorized into five domains: Income; Employment; Education, Skills & Training; Health & Disability; and Crime. By fitting multilevel log-linear Poisson regression models, neighborhood-level deprivation indicators were examined with and without adjustment for individual-level covariates.

Results
Four neighborhood-level deprivation domains, Employment, Education, Skills & Training, Health & Disability, and Crime, as well as the Danish Composite Deprivation Index (adjusted IRR 1.14; 95 % credibility interval 1.10–1.17), were associated with elevated risk independent of individual-level deprivation measures. The specific neighborhood-level indicators linked with the highest adjusted elevations in risk were: Proportion of inhabitants aged 18–22 years who did not complete primary school before age 18 (adjusted IRR 1.23; 1.20–1.27); Proportion of inhabitants convicted for any violent crime (adjusted IRR 1.19; 1.16–1.23); and Proportion of inhabitants convicted for any crime resulting in a custodial sentence (adjusted IRR 1.15; 1.12–1.18).

Conclusion
This novel population-based multilevel analysis has evidenced the independent associations of neighborhood-level deprivation indicators on schizophrenia risk elevation. Replication is needed in other populations to inform the refinement of preventive strategies.