Association of baseline steroid use with long-term rates of infection and sepsis in the REGARDS cohort

Background: Prior studies associate steroid use with infection risk but were limited to select populations and short follow-up periods. The association of steroid use with long-term risk of community-acquired infections is unknown. We sought to determine the association of steroid risk with long-term risks of community-acquired infections and sepsis. Methods: We used data on 30,239 adults aged >= 45 years old from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohort. The primary exposure was oral or injectable steroid use, determined from medication inventory obtained at baseline in-home visit. The primary outcome was time to first infection event during 2003-2012, determined through adjudicated review of hospital records. We determined associations between baseline steroid use and first infection hospitalization events using Cox proportional hazards models, adjusting for demographics, health behaviors, chronic medical conditions, and medication adherence. Among the first infection hospitalization events, we also determined the association between baseline steroid use and sepsis. Results: Steroid use was reported in 2.24% (n = 677) of the study population. There were 2593 incident infection events during the 10-year follow-up period. Infection incidence rates were higher for steroid than non-steroid users (37.99 vs. 13.79 per 1000 person-years). Steroid use was independently associated with increased risk of infection (adjusted HR 2.10, 95% CI: 1.73-2.56). Among first-infection events, steroid use was associated with increased odds of sepsis (adjusted OR 2.11, 95% CI: 1.33-3.36). The associations persisted in propensity matched analyses as well as models stratified by propensity score and medication adherence. Conclusions: In this population-based cohort study, baseline steroid use was associated with increased long-term risks of community-acquired infections and sepsis.