Beverage Co-Consumption Profiles, Sleep, and Fatigue in Professional Army Personnel

Background/Objectives: Active-duty military personnel may use caffeinated and sweetened beverages to manage alertness and fatigue, but co-occurring beverage behaviours are rarely examined. We identified beverage-frequency co-consumption profiles in Spanish Army personnel and tested their cross-sectional associations with self-reported sleep and habitual physical fatigue. Methods: This cross-sectional study included 1155 active-duty Spanish Army personnel. Latent classes were estimated from ordinal frequencies of caffeinated beverages, sugar-sweetened beverages, and energy drinks; these indicators represented frequency categories rather than caffeine dose, sugar intake, or standardized serving size. Adjusted outcome models incorporated classification uncertainty and controlled for demographic, occupational, training, hydration, alcohol, smoking, and recruitment-cohort covariates. Results: A three-class solution identified lower-frequency (47.7%), intermediate-frequency (28.3%), and higher-frequency (24.0%) profiles. Compared with the lower-frequency profile, the higher-frequency profile was associated with 0.230 h less sleep per night (95% CI −0.416 to −0.045; q = 0.022), 2.129 percentage points lower sleep efficiency (95% CI −4.177 to −0.082; q = 0.041), and 0.395 points greater habitual physical fatigue (95% CI 0.077 to 0.714; q = 0.022). Sensitivity analyses were directionally consistent. Conclusions: Beverage-frequency responses clustered into three profiles. The higher-frequency profile was cross-sectionally associated with shorter self-reported sleep, lower sleep efficiency, and higher habitual physical fatigue. The sleep-duration contrast was approximately 13.8 min/night, but the clinical or perceptible relevance of the sleep-efficiency and fatigue differences cannot be established from these self-reported measures. Prospective studies with product-specific dose and timing data are required to establish temporality and practical relevance.