Sleep and Women's Health: Sex- and Age-Specific Contributors to Alcohol Use Disorders

Abstract

Prior research on alcohol use disorder (AUD) has focused primarily on men, but psycho-social-cultural changes have led to more women drinking or binge drinking, thus highlighting sex differences observed in alcohol use. In parallel, recent evidence indicates bidirectional links between alcohol use and sleep disruption, offering a burgeoning field of research for the study of sex differences in sleep–alcohol interactions. As part of the 2018 Research Conference on Sleep and the Health of Women at the National Institutes of Health, three presentations focused on the intersection between alcohol and sleep in women, including links between disrupted sleep and the risk of AUD. The literature to date hints at sex differences in the relationships between sleep and alcohol use that may be relevant to prevention and/or intervention. For example, insomnia is more prevalent among women, yet men may be more likely to self-medicate insomnia with alcohol and may benefit more from alcohol's sedating effects. Sex differences in sleep timing and duration that begin during adolescence may also be relevant, as later and/or shorter sleep appear to be risk factors for binge drinking. Preliminary data also implicate circadian timing as a sex difference potentially relevant to alcohol use. Limited extant data suggest complex relationships between sex, sleep, and alcohol problems, but defy easy summary. Relevant studies sufficiently powered to test sex differences are needed.

Introduction

Historically, a significant amount of attention has been placed on studying alcohol use disorder (AUD) in men as they are more likely to drink, be heavy drinkers, binge drink (i.e., show hazardous drinking behavior), and to have an AUD and alcohol-related deaths than women.^1–3 However, with psycho-socio-cultural changes, there are more women drinking and binge drinking, such that male–female differences in alcohol use are decreasing.⁴ Several reasons account for the sex differences in alcohol use and impact, including differences in alcohol-metabolizing enzymes, higher body fat and lower proportion of body water in women, and psychosocial factors: women with AUD are more likely to have comorbid psychiatric disorders than men with AUD (e.g., mood disorder), and depressive symptoms in women with alcohol problems predict heavier alcohol consumption. What is less clear, however, is whether alcohol has the same effect on sleep in women as in men, as most studies have been done in men or are underpowered to investigate sex differences. In 2018, the National Institutes of Health convened the Research Conference on Sleep and the Health of Women to highlight research in this space. The purpose of this summary is to specifically showcase presentations on the intersection between disrupted sleep in women and the risk of AUD.

Sleep Disturbances and Risk for Alcohol Problems: Different Pathways for Men and Women?

Decades of research have shown that sleep disturbance/insomnia is a pathway toward alcohol consumption with alcohol being used as a sedative (reviewed in Refs.^5,6). However, with prolonged use, alcohol increases sleep disruption, and in AUD, chronic alcohol use is associated with substantial changes in sleep architecture that persist into abstinence, and are risk factors for relapse.^5,6 In turn, relapse into alcohol use again leads to sleep disturbances and the vicious circle continues. Some of us have proposed that differential pathways to alcohol problems may occur via positive and/or negative reinforcement pathways.⁷ For example, with respect to positive reinforcement, later sleep timing is associated with more impulsivity and sensation seeking,⁸ particularly among men.⁹ Men tend to exhibit later sleep, as well as more impulsivity and sensation seeking.¹⁰ Furthermore, preliminary evidence suggests that later sleep timing in adult men may also be associated with a more stimulating response to alcohol.¹¹

Whether sex differences exist in negative reinforcement pathways linking sleep and alcohol remains unclear. On the one hand, published data on adolescents and adults suggest that men may be more likely to self-medicate their insomnia with alcohol,^12,13 thus putting them at more risk of subsequent alcohol problems. On the other hand, insomnia is more prevalent among women than in men, a difference that emerges postpuberty and persists through adulthood,^12,14 which would presumably put women more at risk for alcohol problems related to using alcohol as a sleep aid and/or to reduce anxiety at bedtime.

Adults with insomnia appear to benefit from acute alcohol use in terms of increases in deep sleep and reductions in anxiety,¹⁵ but such studies have not investigated sex differences. One experimental study that considered sex differences in the acute effects of alcohol found that alcohol relative to placebo led to increased deep slow wave sleep (SWS) and decreased rapid eye movement sleep in both men and women, but that alcohol reduced latency to SWS more in men than in women, which could be due to sex differences in alcohol pharmacokinetics.¹⁶ However, women reported even more bedtime sleepiness after alcohol than men.

Using data from the SRI laboratory comparing sleep in abstinent men and women with AUD,¹⁷ no significant sex × diagnosis interaction effects were demonstrated; however, women with AUD relative to controls tended to show less reduction in SWS than men with AUD versus controls. Because this group was a mixed group of recovery patients, with varying lengths of sobriety, and different lifetime alcohol exposures, it is unclear whether women show a greater chance of recovery, or if SWS regulatory systems in women may be less impacted by alcohol use.

Sleep in Women: Effects of Alcohol and Age

It is well known that age impacts sleep structure differently in men and women,¹⁸ so in exploring the relationships between sleep, sex, and alcohol, it is important to consider age, as it may modify the impact of alcohol on sleep. Across adolescence, the timing of sleep becomes later due to biological factors—circadian timing system and sleep–wake homeostasis—and changes in the psychosocial milieu.¹⁹ Many young people begin college toward the end of these developmental changes, and their impact is evident during the first semester particularly due to the “relief” from schedule timing for school. For example, it has been shown in a large sample of first-year students, that the transition to college was associated with an average 2-hour delay of sleep timing.²⁰

Beginning in high school, the chief sex difference was that women tended to report earlier bedtimes and rise times than men. Multiple studies report that adolescent women are less likely to achieve a sufficient sleep duration,^21–23 whereas the evidence is mixed whether adult women get more or less sleep than adult men.^23–25

Typical adolescent sleep changes include later sleep and circadian timing, worsening of sleep continuity (more insomnia), shorter sleep duration (particularly on school days), and increased daytime sleepiness.²⁶ A growing number of longitudinal studies suggest that all of these sleep characteristics are prospectively associated with the onset and/or escalation of alcohol use and alcohol-related problems.^27–30 Based on data from the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA) study, both worse sleep quality and later weekday bedtimes at baseline were associated with more severe binge alcohol use 1 year later.³¹ This is further supported by unpublished data from NCANDA that suggest that although female adolescents with poor sleep quality are more likely to report binge alcohol use, only male adolescents with poor sleep quality are at risk of worsening binge alcohol use a year later.

Although studies clearly show sex differences in circadian measures, few studies have investigated whether these differences are related to different alcohol use patterns or alcohol impact in men and women. One recent study examined relationships between sleep and circadian factors with sex differences in college drinkers. The percentage of women who reported being a “light or nondrinker” was higher than men at 61%, but the male:female ratio for heavy reported drinking was about 50:50.²⁰ In terms of sleep duration, the only observed gender difference related to alcohol use was significantly shorter reported sleep duration in women whose heavy drinking began in college. “Veteran” heavy drinkers (i.e., those who reported drinking before college) of both sexes reported later bedtimes than students who were nondrinkers or light drinkers. Heavy drinking was also associated with greater sleep irregularity in both sexes. Finally, women who began drinking heavily in college had a longer interval between dim light melatonin onset and bedtime than men and women at other levels of drinking.³² Another analysis of these data showed that poor sleep quality moderates the association of psychiatric symptoms and alcohol-related consequences, with no sex difference noted.³³

Conclusions

Women are chronically under-represented in alcohol-sleep research studies. Based on these observations, future studies should consider both sex and age effects when examining the role of sleep as a risk factor for alcohol problems, as well as effects of chronic alcohol use on sleep itself. It will be important to (1) identify sleep and circadian-related factors distinguishing alcohol initiation and AUD risk in women versus men; (2) examine whether the effects of alcohol use on sleep differ between men and women; (3) explore whether and/or how the sex differences in the timing of sleep and circadian rhythms that may vary developmentally, particularly in adolescence and early adulthood, affect alcohol use; (4) assess the roles of biological versus behavioral factors in sleep and alcohol use and where sex differences arise. The evidence for sex-specific pathways from sleep to alcohol problems remains mixed, but intriguing, and offers opportunities for future sleep-alcohol research.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

Grants from the National Institutes of Health: R01AA026249 (Hasler/Pedersen), U01AA021690 (Clark), R01AA011873 (Molina), K01 AA021135 (Pedersen), Foundation Grant from ABMRF/The Foundation for Alcohol Research (Pedersen). AA021696; U01DA041022 (Baker). MH079179 (Carskadon).