Elucidating the alcohol-sleep-hangover relationship in college students using a daily diary approach.

This daily diary study expands knowledge of the pharmacological alcohol-sleep relationship using a multilevel modeling approach. The interplay between alcohol and sleep on hangover susceptibility is also explored. College students (n = 337; 52 % female) provided 2976 days of self-reported alcohol use. We regressed sleep duration onto accumulated sleep debt, prior night sleep duration, and estimated blood alcohol concentration (eBAC) at bedtime; linear mixed models disaggregated day and person-level effects. Binomial models, assessing days after drinking when eBAC = 0 % versus when eBAC>0 % at waketime, regressed hangover susceptibility onto the same predictors plus sleep duration. More accumulated sleep debt predicted slightly longer same-night sleep. Greater than average bedtime intoxication predicted longer than average same-night sleep when drinking ceased early, but later drinking attenuated the relationship. People who typically stopped drinking later in the night reported typically shorter sleep durations on drinking nights. When waketime eBAC = 0 %, higher eBAC at bedtime and drinking later on a given night predicted greater next-day hangover susceptibility. Typical bedtime eBAC and typically later drinking predicted typically greater hangover susceptibility. When waketime eBAC>0 %, longer sleep duration predicted more likely hangovers. Bedtime eBAC and sleep debt interacted, such that more sleep debt attenuated the positive association between intoxication and next-day hangover susceptibility. Late-night drinking appeared to reduce sleep duration and increase hangover susceptibility. Accumulated sleep debt complicated the alcohol-sleep-hangover relationship. External factors influencing sleep behaviors were not assessed, but the results highlight the need to deconstruct sleep into acute and chronic processes. Future studies should better subdivide physiological processes related to hangovers.