Evaluating the association between caffeine exposure and sleep duration and trouble sleeping in adults

This study examines whether caffeine is independently associated with sleep outcomes in the general population.

Data from the National Health and Nutrition Examination Survey 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-h 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 h vs. ≥6 h). 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.

The sample included 37 469 adults (mean age 46.4 years; 51.7 per cent female). Compared with 0 mg/day, the adjusted odds ratios (aORs) for short sleep were 1.0 (95 per cent CI, 0.9-1.2) for 1-99 mg/day, 1.1 (95 per cent CI, 0.9-1.2) for 100-399 mg/day and 1.5 (95 per cent CI, 1.2-1.8) for ≥400 mg/day. Caffeine intake of 101-400 mg taken within 8 h 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 per cent CI, 0.9-1.1]; moderate intake: aOR 1.0 [95 per cent CI, 0.9-1.2]; high-intake: aOR 1.0 [95 per cent CI, 0.9-1.2]).

In this U.S. cross-sectional sample, higher caffeine intake was modestly associated with shorter self-reported sleep duration particularly when taken within 8 h of bedtime.