Abstract Study Objectives This study examines whether caffeine is independently associated with sleep outcomes in the general population. Methods Data from the National Health and Nutrition Examination Survey (NHANES) from 2005 to 2018 were analyzed, including adults aged ≥18 years who completed both dietary interviews and the Sleep Disorders Questionnaire. Caffeine exposure was estimated from a 24-hour dietary recall and expressed as milligrams per day, modeled both continuously and categorically (0, 1–99, 100–399, and ≥400 mg/day). Sleep outcomes included self-reported sleep duration (6 hours vs. ≥6 hours). Survey-weighted multivariable logistic regression models were applied to examine associations, adjusting for age, sex, race/ethnicity, education, smoking status, alcohol use, income-to-poverty ratio, and depression. Results The sample included 37,469 adults (mean age 46.4 years; 51.7% female). Compared with 0 mg/day, the adjusted odds ratios (aORs) for short sleep were 1.0 (95% CI, 0.9–1.2) for 1–99 mg/day, 1.1 (95% CI, 0.9–1.2) for 100–399 mg/day and 1.5 (95% CI, 1.2–1.8) for ≥400 mg/day. Caffeine intake of 101 to 400 mg taken within 8 hours of bedtime was associated with short sleep. Although significant in unadjusted analyses, the association between caffeine intake and self-reported trouble sleeping was no longer significant after multivariable adjustment (low intake: aOR 1.0 (95% CI, 0.9–1.1); moderate intake: aOR 1.0 (95% CI, 0.9–1.2); high-intake: aOR 1.0 (95% CI, 0.9–1.2)). Conclusions In this U.S. cross-sectional sample, higher caffeine intake was modestly associated with shorter self-reported sleep duration particularly when taken within 8 hours of bedtime.
Nguyen et al. (Tue,) studied this question.