Abstract
STUDY OBJECTIVES: Evidence is limited on the role of caffeine intake in the relationship between sleep quality and the incidence of major adverse cardiovascular events (MACE) particularly in patients with sleep breathing disorders. Therefore, this study's primary aim was to determine the potential confounding effects of total caffeine consumption on the relationship between sleep quality parameters (total sleep time [TST], sleep efficiency [SE], sleep latency [SL], daytime sleepiness, and wakefulness after sleep onset [WASO]) and MACE. METHODS: This study is a secondary analysis of data from the Sleep Heart Health Study (SHHS). Sleep assessments (TST, SE, SL, daytime sleepiness, and WASO) were performed objectively using in-home polysomnography. Caffeine was measured using a survey asking about the average number of cups/cans/glasses of tea, soda, and coffee consumed per regular day and during the last night before polysomnography. RESULTS: A final sample of 5628 participants was included in SHHS Visit 1 (78 per cent White/Caucasian; 54 per cent female). Cumulative incidence rates measured over 10.9 ± 2.8 years were 15.1 per cent for MACE and 19.7 per cent for all-cause mortality. In univariate models, all sleep measures except SL were associated with MACE; but after adjustment, only TST remained a significant predictor (odds ratio [OR] = 1.122, p = .011). No confounding effect of caffeine was observed in the associations between sleep measures and MACE. Moderate-high intake attenuated MACE risk among individuals with greater daytime sleepiness (OR = 0.91, p = .011). CONCLUSIONS: Caffeine was not a confounding factor in the relationship between sleep measures and MACE. While exploratory analyses suggested potential modification of the association between hypersomnolence and cardiovascular outcomes, these effects were attenuated after statistical adjustment and correction for multiple testing and should be interpreted cautiously.