Cannabis use and sleep disturbances among White, Black, and Latino adults in the United States: A cross-sectional study of National Comorbidity Survey-Replication (2001-2003) data.

OBJECTIVE:Research investigating cannabis use and sleep health is limited, and results are mixed. Few studies were nationally representative with racially-ethnically diverse samples or assessed potential modifiers. Our objective was to investigate cross-sectional associations between reported cannabis use and sleep disturbances by potential modifiers among non-Hispanic White, non-Hispanic Black, and Hispanic/Latino men and women in the United States. METHODS:We used nationally representative National Comorbidity Survey-Replication data collected from 2001 to 2003 among 3929 adults. Poisson regression with robust variance estimated prevalence ratios (PR) and 95% confidence intervals of patterns of sleep disturbances identified through latent class analysis. Models adjusted for sociodemographic, health behavior, and clinical characteristics were stratified by race-ethnicity and by race-ethnicity along with sex/gender, and age, separately. RESULTS:Over half of adults reported cannabis use (52%-ever/lifetime vs 48%-never). We identified two latent classes: multiple sleep disturbances with daytime sleepiness and no sleep disturbances with some daytime sleepiness. Prevalence of multiple sleep disturbances with daytime sleepiness was higher among participants reporting lifetime cannabis use (23% vs 20%). Associations did not vary by race-ethnicity or sex/gender. Lifetime vs never cannabis use was marginally associated with a higher prevalence of multiple sleep disturbances with daytime sleepiness only among adults aged 25-29years (PR=1.09 [95% confidence interval: 1.00-1.18]; eg, PRage 40+ years=1.00 [0.97-1.03], pinteraction=0.03). CONCLUSIONS:Associations between cannabis use and sleep may vary by age. Replication with more recent data and prospective studies that investigate intersectional identities among diverse populations with objective assessments are warranted.