Neighborhood spikes in drug and alcohol overdoses and subsequent suicide: A serial cross-sectional study of California ZIP codes.

Objective: The overdose crisis potentially increases risk factors for suicide while decreasing resources available to help those in need. This study sought to determine whether acute increases in drug or alcohol overdoses increase suicide rates. Methods: This serial cross-sectional study used comprehensive emergency department, hospital discharge, and mortality data for ZIP-code-months in California, March 2010-December 2018 (n = 163,240). Overdose spikes were defined as ZIP-code-months with rates >2 standard deviations above expected. We used confounder-adjusted Bayesian Poisson space-time misalignment models to estimate whether overdose spikes, lagged 0-6 months, were associated with the rate of suicide (overall and by firearm involvement). Results: Overdose spikes were identified in 6728 (4.1%) ZIP-code-months. In unadjusted models, overdose spikes were associated with same-month increases in suicide (IRR: 1.07, 95% CI: 1.02, 1.14) and non-firearm suicide (IRR: 1.09, 95% CI: 1.02, 1.17). The latter association persisted in adjusted models (IRR: 1.08, 95% CI: 1.00, 1.15), which also showed reduced firearm suicide four months following a spike (IRR: 0.90, 95% CI: 0.82, 0.99). No other associations were significant. Conclusions: We did not find robust evidence of an association between spikes in ZIP-code-level overdoses and suicide. However, associations in unadjusted models point to a potential syndemic of overdose and suicide.