Abstract
Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in the United States for both men and women. The increasing appreciation of the gender differences in behavioral and physiologic interconnectedness of cardiometabolic risk factors has led to a growing literature evaluating relationships between sleep—an integral part of overall health and a source of physiological resilience—in relation to CVD and its risk factors in male and female cohorts. This article reviews the published research related to gender differences in the role of insufficient sleep in the etiology of CVD. Evidence on gender differences in the associations between sleep and other behavioral risk factors for CVD, namely diet and physical activity, is in the nascent stages of characterization but suggests that insufficient sleep may have a more pronounced detrimental impact on the diets of women. The literature on gender differences in the role of insufficient sleep in inflammation, dyslipidemia, glycemic control and insulin sensitivity, and obesity is inconclusive. Inadequate sleep duration is associated with elevated blood pressure, weight gain, hypertension, and obesity in both men and women, but associations may be stronger among women. Moreover, gender differences may vary by age group. Inflammation resulting from inadequate sleep may also be more pronounced among women. Inconsistent results were observed in studies on dyslipidemia with results ranging from null to detrimental associations in men and from detrimental to protective associations in women. Short sleep is also associated with poor glycemic control among men, with evidence of potential gender differences in associations with insulin sensitivity. Finally, insufficient sleep is convincingly associated with CVD etiology, but women appear to be more vulnerable to the effects of poor sleep on CVD risk and mortality. In conclusion, emerging data demonstrate that heterogeneity exists between men and women in how sleep duration influences behavioral and cardiometabolic risk factors in addition to CVD risk and mortality, but additional research is warranted to clarify associations and disentangle underlying mechanisms. In an era of genomics and personalized medicine, studying gender differences in the role of sleep in the risk for cardiometabolic disease is crucial for identifying innovative targets and tailoring public health interventions to focus on the most vulnerable population groups.
Introduction
Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in the United States for both men and women. 1 However, the absolute number of women living with CVD exceeds that of men, and since 1984 more women than men have died each year from heart disease. 1 Gender differences in CVD burden and its contributing risk factors in the United States are increasingly recognized, particularly due to the focus on including women in clinical investigation over the past 2 decades. 2 These observed gender differences could also be attributed to disparities in the prevalence of CVD risk factors. For instance the prevalence of high blood pressure (BP) is greater in women >65 years of age, 3 diabetes mellitus is more prevalent among women than men ≥20 years of age (8.3% vs. 7.2%), 4 and women might be more susceptible to inflammation than men. 5 Similarly, significant heterogeneity in susceptibility to metabolic syndrome exists between men and women. Abdominal obesity is the dominant characteristic in women as opposed to various risk factor combinations reported in men. 6 Another key explanation for gender differences in CVD burden is the observed variation in the prevalence of lifestyle risk factors. For instance, men have a higher prevalence of smoking 4 but a lower prevalence of physical inactivity 7 and adherence to CVD prevention guidelines. 8
In light of the growing burden of CVD, there has been an increasing interest in deciphering other modifiable risk factors to promote behavior change for disease prevention. Sleep is an integral part of overall health and a source of physiologic and psychologic resilience that is increasingly being recognized as an integral component of the lifestyle paradigm to ameliorate not only CVD risk and outcomes but also cardiometabolic risk factors, including body adiposity, measures of glycemic control, inflammation, and BP. 9 The increasing appreciation of gender differences in behavioral and physiologic interconnectedness of cardiometabolic risk factors has led to a growing literature evaluating relationships between sleep and the risk of CVD in male and female cohorts. 9
While ∼1/3 of Americans report insufficient sleep (< 7 h/night)9,10 with evidence of a decline in the quantity of sleep over the past few decades,11,12 women were more likely to report insufficient sleep than men. 10 Men and women may have a differential risk of nonoptimal sleep duration due to differences in a variety of factors, including lifestyle, circadian rhythm abnormalities, mood disorders, and coexistent medical conditions. 13 Women could also be uniquely at risk for getting nonoptimal levels of sleep due to changing reproductive hormone levels and caregiving responsibilities over the lifespan that may differentially affect women. 13 Finally, women may also be more vulnerable to the cardiometabolic health consequences of insufficient sleep. 5 Therefore, the purpose of this contemporary review is to summarize and evaluate critical gender differences in the epidemiologic evidence relating insufficient sleep duration to behavioral and cardiometabolic risk factors, and to CVD risk and mortality among adult men and women. Insights from this review could guide future research of potential relevance for prevention, diagnostics, and therapy for CVD across the genders.
Sleep and Cardiometabolic Risk Factors
Sleep and blood pressure
Early reports from acute laboratory sleep-deprivation studies reporting increased BP and sympathetic nervous system activity after sleep restriction14,15 suggested that a potential mechanism linking short sleep duration to CVD risk could be increased BP or risk of hypertension. Three recent systematic reviews and meta-analyses of cross-sectional and prospective epidemiologic studies have reported 20%–61% higher hypertension risk with short sleep duration.16–18 In a meta-analysis by Wang et al., 18 sleep deprivation among women (≤ 5 h/night vs. 7 h/night) was associated with a higher risk of hypertension (odds ratio [OR]: 1.68; 95% confidence interval [CI]: 1.39–2.03) than that observed among men (OR: 1.30; 95% CI: 0.93–1.83).
However only 3 of the reviewed prospective cohort studies presented findings stratified by sex.19–21 An analysis within the London-based Whitehall II study of 10,308 white-collar British civil servants ages 35–55 years examined both the cross-sectional and prospective associations of sleep duration with prevalent and incident hypertension. 19 While no cross-sectional associations were observed in men, women with short sleep duration (≤ 5 h/night vs. 7 h/night) had a 2-fold higher risk of hypertension (OR: 2.01; 95% CI: 1.13–3.58) and higher systolic BP (SBP). In longitudinal analyses, null associations persisted among men, but among women associations were attenuated. No significant differences in SBP between short and normal sleepers were detected in either sex. These null findings were corroborated by a longitudinal Spanish study, which also showed no association between short sleep duration and prevalent or incident hypertension and BP in either gender. 20 However, in an ancillary study to the Coronary Artery Risk Development in Young Adults (CARDIA) cohort study, 21 short sleep was associated with higher SBP and diastolic BP (DBP) and 37% higher odds of hypertension, but the findings did not show a gender-specific association in adults ages 18–30 years.
More-recent studies in this area have been conducted primarily within Asian populations and have reported conflicting results.22–26 In a cross-sectional analysis within the China Health and Retirement Longitudinal Study, compared with the reference group (≥ 7 h/night and <8 h/night), the groups who had <6 h/night and ≥6 h/night and <7 h/night had a 68%–69% higher risk of hypertension (OR: 1.68; 95% CI: 1.17–2.42 and OR: 1.69; 95% CI: 1.11–2.59, respectively) in middle-age men but not women. 24 Similarly in a cross-sectional study of Korean women, null associations were reported between sleep duration and BP. 23 In contrast to these null findings, other studies reported detrimental associations between short sleep and BP in women. In another Korean study, sleep <7 h/night, compared to ≥7 h/night, was associated with higher SBP and DBP in women but not in men. 22 Similarly in 2 Chinese studies, short sleep duration was associated with higher SBP and DBP and risk of hypertension in females only.25,26 These associations were particularly strong among early middle-age females (ages 35–44) in whom sleep duration <6 h/night vs. 7–8 h/night increased risk of hypertension by 77% (OR: 1.77; 95% CI: 1.02–2.78). 25
Finally, the most recent prospective cohort study, conducted in British men and women, ages 50 or older, showed that both men (OR: 1.73; 95% CI: 1.08–2.76) and women (OR: 1.44; 95% CI: 1.00–2.07) with short sleep (≤ 5 h) were at increased risk of incident hypertension, but associations were stronger among men. 27 However age-stratified analyses revealed that short sleep was predictive of a >2-fold higher risk of incident hypertension in men (OR: 2.27; CI: 1.01–5.11) and women (OR: 2.10; C.I. 1.08–4.09) younger than 60 but not in older people. 27 These findings suggest a potential complex interplay between sex and age group that might modify associations between sleep deprivation, BP, and risk of hypertension.
Sleep and pro-inflammatory biomarkers
A key mechanism through which sleep restriction can alter cardiometabolic risk is via its impact on levels of cytokines, such as interleukin (IL)–6, C-reactive protein (CRP), and tumor necrosis factor–α (TNF-α), which are collectively known to be important in regulating inflammation. 28 A 2015 review and meta-analysis of experimental and cohort studies investigating sleep duration in relation to inflammation showed that shorter sleep duration, but not the extreme of short sleep (due to insufficient data on this topic), was associated with higher levels of CRP (effect size = 0.09; 95% CI: 0.01–0.17) but not IL-6 or TNF-α. 28 However, gender differences in these associations were not explored in most studies. The literature on sex-specific associations between sleep duration and inflammation is limited and findings are primarily derived from cross-sectional studies.29–33
A cross-sectional study within the London-based Whitehall II cohort reported a null association between sleep duration and IL-6 and CRP in men. 33 In the same study, lower IL-6 levels were observed among women who slept 8 h (OR: 0.89; 95% CI: 0.83–0.96), compared to 7 h. Moreover, a nonlinear association was reported between sleep duration and CRP levels in women, such that CRP levels were significantly higher in women who were short sleepers (≤ 5 h/night vs. 7 h/night) (OR: 1.42; 95% CI: 1.02–1.96). 33 These cross-sectional findings suggest that associations of sleep duration with IL-6 and CRP are affected by gender. A follow-up longitudinal study within the same cohort later sought to examine if changes in sleep duration were associated with IL-6 and CRP levels from 2 measures obtained 5 years apart. 34 Findings showed that for every h/night decrease in sleep duration between 1991 and 1994 and 1997 and 1999, there was a concomitant increase in CRP (8.1%) and IL-6 (4.5%) averaged across measures in 1997–1999 and 2002–2004. Unlike the cross-sectional analysis, 33 no interaction between sex and sleep duration in relation to CRP or IL-6 was detected in the longitudinal study. 34
In contrast to findings from Whitehall II cohort, an American 32 and a British cross-sectional study 31 reported null associations for short sleep duration with CRP and IL-6 in both men and women. However, two clinical studies conducted in women suggested that associations vary between pregnant and nonpregnant women.29,30 One report concerning 43 community-dwelling women (mean age: 28 yr), reported null associations between sleep duration in relation to CRP, TNF-α, and IL-6. 30 However, among 19 pregnant women, short sleep was associated with higher levels of IL-6 but not TNF-α in mid- to late-pregnancy. 29 Finally, in men (n = 12), there was one experimental study that evaluated the effect of sleep restriction on inflammatory markers and showed that one night of acute sleep deprivation induced a significant increase in TNF-α (p < 0.01), but there were no significant changes in IL-6, cortisol, and CRP. 35 Taken together, these studies reported conflicting findings but suggested that associations between short sleep duration and inflammation may be stronger for women and may vary across the female life course. Additional prospective cohort studies are warranted to confirm these results.
Sleep and dyslipidemia
Emerging evidence suggests that short sleep may be associated with dyslipidemia as a consequence of insulin resistance. Short sleep has been linked to promoting an insulin-resistant state characterized by increased levels of triglycerides due to increased production of very low–density cholesterol (VLDL) and of intestinally derived lipoprotein. 36 Low high-density lipoprotein (HDL) cholesterol could also occur due to lower production of HDL from the degradation of VLDL as a result of reduced lipoprotein lipase activity and as a function of the production of HDL-depleted cholesteryl ester by the action of cholesteryl ester transfer protein. 36
Despite the biologic plausibility of this association, there is a paucity of literature examining short sleep in relation to markers of dyslipidemia in male and female cohorts, with most evidence derived from cross-sectional studies.23,37–39 Among men, results ranged from null associations to a detrimental effect of insufficient sleep. In Chinese men, short sleep was not associated with total cholesterol, HDL, low-density lipoprotein (LDL), and triglycerides 38 ; whereas in Japanese men, sleeping <5 h/night was associated with a higher risk of low HDL and high triglycerides, compared to sleeping >5 h/night. 37 Another Japanese cross-sectional study reported null findings between sleep duration in relation to serum triglyceride and HDL levels, but men who slept ≥8 h/night had ∼9.2 mg/dL lower LDL cholesterol levels than those of men who slept for 6–7 h/night. 39
However, studies in women reported findings ranging from a detrimental to a protective effect of short sleep on blood lipids.23,38,39 In one Chinese study, compared with women reporting 8 h of sleep, women who slept <6 h/night had up to 65% and 71% higher odds of high total cholesterol and LDL respectively. 38 Similarly, Japanese women with short sleep duration (< 5 h/night vs. 6–7 h/night) had a 51% higher risk of elevated triglyceride levels (OR: 1.51; 95% CI: 0.96–2.35) and a >5-fold higher risk of low HDL levels (95% CI: 2.29–14.94). 39 However, among 8505 Korean women, ages 25–70, short sleep duration was shown to have a protective effect on cholesterol levels. 23 In particular, sleep durations of ≤5 h/night and 6 h/night were found to be associated with 14%–16% lower odds for reduced HDL cholesterol (OR: 0.82; 95% CI: 0.69–0.97 and OR: 0.86; 95% CI: 0.75–0.98) and 19% lower odds of increased triglycerides (OR: 0.81; 95% CI: 0.49–0.99). 23 This evidence of a protective effect among women was contradictory to the general body of evidence and was attributed to higher socioeconomic status and physical activity levels in short sleepers. Nonetheless because adjustment for these factors did not change associations, other unknown physiologic mechanisms may be at play.
Prior research by one of the current authors (B.A.) has also shown a differential effect of short sleep duration on lipids among women versus men. 40 In this cross-sectional, population-based study of men and women participating in a CVD screening program in New York City, it was found that 1/3 of participants slept <6 h/night, and women were significantly more likely than men to report short sleep duration. 40 Short sleep duration was associated with increased LDL-cholesterol overall; however gender modified the association between sleep duration and LDL-cholesterol level. Sleeping <6 h/night was significantly associated with reduced LDL-cholesterol levels among women and increased LDL-cholesterol levels among men. The cross-sectional design did not allow for determination of cause and effect. However, these results suggested that it might be important to consider the influence of hormones on associations between sleep and blood lipids. These results should be prospectively explored further in future studies.
Sleep, glycemic control, and insulin resistance
Emerging evidence has shown that sleep is an important modulator of metabolic homeostasis.41–43 Insufficient sleep is increasingly recognized as a contributor to adverse metabolic sequelae, including reduced glucose tolerance, insulin resistance, and diminished insulin sensitivity. 9 However, much of the literature has focused on the effect of short sleep duration on glycemic control and insulin resistance and sensitivity in pediatric populations and among patients with type 2 diabetes. 41 Short sleep duration is associated with increased risks of metabolic syndrome and type 2 diabetes.41,43 In patients with type 2 diabetes, short sleep is associated with poorer glycemic control and higher HbA1C, 41 but associations are less clear in healthy adults, with few studies reporting associations stratified by sex. Several potential pathways may lead to insulin resistance and poor glycemic control after sleep restriction, including altered function of the autonomic nervous system, endocrine changes, and an altered inflammatory state but further research is needed to elucidate the exact causal pathways and to determine if they vary by gender. 42
Two cross-sectional studies within Asian populations have reported conflicting results on associations between short sleep and glycemic control.44,45 A cross-sectional analysis among Chinese adults investigating sleep duration in relation to components of metabolic syndrome reported no associations between sleep duration <7 h/night and blood glucose levels in both men and women. 45 However, shorter sleep durations were not examined in that study. In contrast to these findings, a Taiwanese cross-sectional study reported that short sleep duration (<6 h/night) was independently related to hyperglycemia (OR: 1.39; 95% CI: 1.10–1.74) in males only. 44 These findings are supported by clinical studies in men, which are consistent in reporting that sleep restriction of habitual bedtime or sleeping 5 h/night during one or more consecutive nights is associated with reduced insulin sensitivity and glucose tolerance in otherwise healthy male adults.42,46–48
Two studies reported evidence of gender differences in associations between sleep duration and insulin levels or insulin resistance.49,50 One study among rural Chinese adults showed that in women, short sleep duration (≤7 h/night versus >7 to ≤8 h/night) was associated with a higher homeostatic model assessment of insulin resistance index (HOMA-IR). 50 However a U.S. study, using 2005–2010 NHANES data, showed that women who were short sleepers (≤6 h/night) had lower concentrations of insulin, whereas men who were short sleepers had higher concentrations of insulin, compared to normal sleepers (7–9 h/night), but no gender differences were detected for HbA1c, fasting glucose, and 2-h glucose. 49 Notably, there are limited prospective cohort studies that have been performed to evaluate these associations, and, in particular, clinical and epidemiologic studies are lacking in females. Therefore, additional research is warranted among women to clarify the effect of insufficient sleep on the glucose and insulin response in this population group.
Sleep and body adiposity
Insufficient sleep has gained attention as a potential contributor to the obesity epidemic, as the rise in the prevalence of obesity has paralleled a decline in sleep duration in Western society. 51 Overall the existing data support an association between shortened sleep and obesity.51–53 A landmark systematic review and meta-analysis 52 of 18 cross-sectional and cohort studies among adults reported a 55% higher risk of obesity with short sleep duration (pooled OR: 1.55; 95% CI: 1.43–1.68) but pooled estimates were not reported separately for men and women.
Among the reviewed studies only 4 reported findings stratified by sex,54–57 but gender differences were only observed in half of the studies.56,57 An early report on Japanese workers showed that sleeping <5 h vs. 5–8 h was associated with an ∼2-fold higher risk of obesity in men (OR:1.96; 95% CI: 1.20–3.23), but associations were not significant among women. 56
Gender differences were also reported within a nationally representative U.S. sample using 1982–1984 National Health and Nutrition Examination Survey (NHANES) data, but inconsistent with the Japanese study, women who slept ≤5 h/night vs. 7 h/night had up to >2-fold higher odds of obesity, whereas null results were observed in men. 57 However, significant gender differences were not observed in other studies.54,55 One analysis within the Quebec Family Study showed that short sleep (5–6 h/night vs. 7–8 h/night) was a significant predictor of overweight and obesity in both men (OR: 1.72; 95% CI: 1.09–2.71) and women (OR: 1.63; 95% CI: 1.11–2.59). 54 In both sexes, lower adiposity indices were observed in normal vs. short sleepers but significant differences disappeared after statistical adjustment for plasma leptin levels, which were lower in short sleepers. Similarly in a Finnish cross-sectional study, short vs. midrange sleep duration was associated with 46% and 75% abdominal obesity in both men and women, respectively. 55
Beyond the cross-sectional evidence, associations between short sleep in relation to obesity have been evaluated in a number of longitudinal studies.58–65 Objective short sleep duration was not significantly associated with incident obesity in either men or women in a Penn State cohort. 60 In other studies, associations between short sleep and obesity were observed in men, but not in women.61,63,64 A prospective cohort study by Sayon-Orea et al. showed that sleeping <5 vs. 7 to <8 hours/night was associated with a 2-fold higher risk of incident obesity in men but not in women in a Mediterranean cohort. 63 Similarly in a study of Japanese adults, sleeping <5 or 5–6 h/night vs. 7–8 h/night was associated with a >2-fold and 80% higher risk of obesity in men but not in women. 64 These findings were corroborated by another Japanese study conducted in men only, which showed that sleeping <6 h/night vs. 7–8 h/night was associated with a >2-fold higher risk of incident obesity. 61
Other longitudinal studies have reported the detrimental impact of insufficient sleep in both sexes. Among middle-age to older adults within the National Institutes of Health–AARP® cohort, shorter sleep (< 5 or 5–6 h) was associated with more weight gain in both sexes, and sleeping <5 h/night vs. 7–8 h/night resulted in an ∼40% higher risk of developing obesity in men (OR: 1.45; 95% CI: 1.06–1.99) and women (OR: 1.37; 95% CI: 1.04–1.79). 62 Similarly in a Japanese cohort study, a sleep duration of <5 h/night vs. 5–7 h/night was associated with a 20% higher obesity risk in men and a 70% higher risk in women. 65
In contrast to these findings, a prospective cohort study conducted in Spain showed that women who slept ≤5 h/night vs. 7 h/night were more likely to gain ≥5 kg (OR: 3.61; 95% CI: 1.45–9.01) but no overall association between sleep duration and weight gain was observed in men. 59 These findings are consistent with the Nurses' Health Study, which showed that short sleep duration is associated with modest future weight gain and incident obesity. 58 In particular, women sleeping ≤5 and 6 h gained 1.14 kg and 0.71 kg more than women sleeping 7 h over a period of 16 years. Women who slept 5 and 6 h had 32% and 12% higher risks of 15-kg weight gain (relative risk [RR]: 1.32; 95% CI: 1.19–1.47 and RR: 1.12; 95% CI: 1.06–1.19) as well as 15% and 6% higher risk of obesity (RR: 1.15; 95% CI: 1.04–1.26 and RR: 1.06; 95% CI: 1.01–1.11), respectively. 58
The possible mechanisms underlying gender differences in these associations are not known in detail. In both men and women, short sleep affects some of the mediators of appetite control by lowering concentrations of leptin and increasing concentrations of ghrelin66,67 resulting in greater hunger and appetite. In addition to having an effect on energy intake, sleep may also affect energy expenditure, as it was also postulated that short sleep duration would produce daytime sleepiness, which would, in turn, lead to less physical activity.9,51 Finally experimental evidence has suggested that sleep may also modulate the basal metabolic rate 68 and non–exercise-activity thermogenesis. 69 However, the biologic mechanism underlying the potential differential susceptibility to obesity and weight with short sleep between men and women is a lacuna in the literature that remains to be addressed.
Sleep and Behavioral Risk Factors of Cardiovascular Disease
Sleep and diet
Emerging evidence suggests that sleep is associated with several metrics of diet, including energy intake, diet quality, dietary variety, meal frequency, and food intake.9,70,71 Insufficient sleep may lead to an increased number of meals and snacks, lower dietary variety, and higher overall food intake, resulting in a positive energy balance.71,72 Several mechanisms might underlie these associations, including more time and opportunities for eating, psychologic stress, greater sensitivity to food reward, additional energy needed to sustain extended wakefulness, and changes in appetite hormones, particularly leptin and ghrelin but also glucagon-like peptide 1 in women.9,70 However, it is notable that the limited epidemiologic studies in this area did not stratify by sex, thereby prohibiting the examination of gender differences in the associations between insufficient sleep and various dietary measures.
Among postmenopausal women, an analysis within the Women's Health Initiative showed that objectively measured short sleep duration is associated with higher caloric intakes and consumption of dietary fats. 73 The remaining evidence on sleep duration and diet in women is derived from clinical studies with limited sample sizes (n ≤ 14) that have focused primarily on energy intake as an outcome and demonstrated that sleep restriction is associated with reduced dietary restraint and increased caloric intake.74–76 Similarly among men, clinical studies also showed higher energy intakes with shorter sleep duration76–78 in addition to evidence of increased fat intake 79 and greater morning hunger. 78 In one analysis that compared the effects of short sleep (4 h/night) vs. habitual sleep (7–9 h/night) in men and women, women with short sleep durations had greater increases in caloric intake and total and saturated fat intakes, compared to men, 76 suggesting that the effect of sleep deprivation on diet may be more pronounced among females.
Sleep and physical activity and sedentary behaviors
Epidemiologic data linking short sleep duration to physical activity and sedentary behavior are relatively scarce. Due to the likely bidirectional association between sleep and measures of energy expenditure, most existing studies have focused on physical activity and sedentary time as exposures and insufficient sleep as an outcome. 9 Only one study conducted in healthy women reported null associations between sleep restriction (4–6 h/night) and level of physical activity. 75 In contrast one study showed that short-term sleep loss (4 h/night) was associated with decreased physical activity levels in healthy men. 79 Therefore, much remains to be explored in this area of research, as additional studies are warranted to examine these associations among U.S. adults and to confirm these preliminary findings of differential associations between the sexes.
Sleep and Cardiovascular Disease Risk and Mortality
Given the hypothesized associations between insufficient sleep and multiple cardiometabolic and behavioral risk factors enumerated in the previous sections, a significant body of evidence has linked short sleep to CVD risk and mortality. A systematic review and meta-analysis, published in 2011, of 15 studies investigated sleep duration in relation to coronary heart disease (CHD), stroke, and overall CVD risk and mortality encompassing a total of 474,684 participants from 8 different countries (United States, Japan, United Kingdom, Sweden, Germany, Singapore, Israel, and Taiwan). 80 Findings from the meta-analysis showed that short sleep duration was associated with a 48% greater risk of developing or dying of CHD (pooled RR: 1.48; 95% CI: 1.22–1.80), a nonsignificant 15% higher risk of developing or dying of a stroke (pooled RR: 1.15; 95% CI: 1.00–1.31; p = 0.047), but not total CVD (pooled RR: 1.03; 95% CI: 0.93–1.15). 80 However, any observed heterogeneity in findings of the reviewed studies was not attributed to gender differences in this meta-analysis.
A number of recent cross-sectional and prospective cohort studies have investigated short sleep duration in relation to CVD risk.81–85 In two European prospective cohort studies, insufficient sleep was associated with CHD and stroke in men only.81,82 The first study, conducted with British men and women, showed that sleeping ≤5 h/night was associated with an increased risk of CHD events, such as stroke, in men only (hazard ratio [HR]: 1.66; 95% CI: 1.23–2.25). 81 The second study included men and women, ages 25–74 in Germany, 82 and reported that, among men, short sleep (< 5 h) was associated with a 44% higher risk of total stroke (HR: 1.44; 95% CI: 1.01–2.06). However, associations between insufficient sleep in relation to fatal and nonfatal stroke were weaker and not statistically significant. In contrast, among women sleep duration was not related to any stroke outcome. 82
Other studies reported stronger associations between short sleep duration and risk of CVD events in women.83–86 A cross-sectional study within the National Health Interview Survey showed that sleeping ≤5 h/night vs. 7 h/night was associated with an 88% and a >2-fold higher risk of CVD in men and women, respectively (OR: 1.88; 95% CI: 1.37–2.56 and OR: 2.57; 95% CI: 1.94–3.39, respectively). 83 However, among women, sleeping 6 h was also associated with a 60% higher risk of CVD, suggesting that women may be more susceptible to higher CVD risk at a lower threshold of sleep deprivation. 83
In the Malmo Diet and Cancer Study cohort, short sleep was not associated with risk of CVD events (myocardial infarction [MI], stroke, or death due to ischemic heart disease) in men, but in women sleeping ≤6 h, compared to 7–8 h, short sleep was associated with a 30% higher risk (HR: 1.30; 95% CI: 1.10–1.50). 86 Concurrently, a prospective cohort study among middle-age German adults showed that women sleeping ≤5 h/night vs. 8 h/night had almost a 3-fold higher risk of MI (HR: 2.98; 95% CI: 1.48–6.03), but null results were observed in men. 84 These findings were corroborated by results from the Nurses Health Study, which showed that women who slept ≤5 h/night vs. 8 h/night had a 45% higher risk of CHD. 85 Collectively, these findings suggest that women who are short sleepers may be at higher risk of CVD, CHD, and MI, compared to men, and that the threshold at which short sleep is detrimental may be lower for women.
A number of studies also investigated short sleep duration in relation to CVD mortality.87–95 In multiple European and Asian cohorts, short sleep did not predict coronary artery disease and CVD-related mortality in either men or women.91–93,95,96 However in one analysis among Chinese adults, sleeping ≤5 h/night vs. 7 h/night was associated with 70% and 43% higher risks of CHD mortality in men and women, respectively. 90 Gender differences were detected in a Taiwanese prospective cohort study, which reported that women who slept 0–4 h/night vs. 8 h/night had an ∼2-fold higher risk of CHD (HR: 1.84; 95% CI: 1.03–3.29); no associations were reported in men. 87 Similarly in a Japanese cohort, short sleep (0–4 h/night) was not associated with total CVD mortality in both sexes but there was an increased risk of CHD mortality in women only (HR: 2.32; 95% CI: 1.19–4.50). 94 In a U.K. prospective cohort study, short sleep (≤ 6 h/night vs. 7–8 h/night) was also associated with an 81% higher risk CVD mortality among women only (HR: 1.81; 95% CI: 1.05–3.10). 89 While these studies suggest that women are more prone to the effects of insufficient sleep on CVD mortality, in the Copenhagen Male Study, men who slept <6 h/night vs. 6–7 h/night were also at increased risk of mortality from ischemic heart disease (HR: 1.46; 95% CI: 1.07–2.00); this study did not include women. 88
Conclusions
While it is certain that the physiologic status, hormone secretion, metabolic status, psychologic condition, and stress levels are different between males and females, the exact underlying mechanisms of the gender-specific relationships between sleep duration and various cardiometabolic risk factors are largely unknown. 97 Accumulated and emerging data convincingly demonstrate that significant heterogeneity exists between men and women in how sleep duration influences behavioral and clinical cardiometabolic risk factors in addition to CVD risk and mortality.
Women may be more vulnerable to the cardiometabolic consequences of poor sleep, as stronger associations between short sleep and CVD risk and mortality have been reported in this population group. However, gender differences in sleep in relation to other behavioral risk factors and cardiometabolic risk factors are in the nascent stages of characterization. Furthermore, due to the cross-sectional nature of much of the reviewed studies, a prerequisite for demonstrating causality is establishing a temporal sequence between short sleep and later risk of CVD, and its behavioral and clinical determinants that can only be convincingly demonstrated in prospective cohort studies. Moreover, differences in the definitions of insufficient sleep and differential adjustment for confounders limit comparability across existing studies. Therefore, additional longitudinal studies that use consistent definitions of short sleep and adjust for appropriate confounders are warranted; adding repeated assessments of sleep and cardiometabolic risk factors over time could refine this study design further.
It is notable that in many existing studies, associations between sleep, cardiometabolic risk factors, and CVD are treated as a singular entity irrespective of sex. In an era of genomics and personalized medicine, studying sex-related differences in the role of sleep in cardiometabolic risk is crucial for identifying innovative targets and tailoring public-health interventions to the most vulnerable population groups. Therefore, future epidemiologic studies are essential for disentangling the various mechanisms underlying sex-related differences in the risk and manifestation of CVD, and promoting optimal cardiometabolic health of both males and females at the clinical and population level.
Footnotes
Acknowledgment
This research was supported by an American Heart Association Go Red for Women Strategically Focused Research Network Award (Grant # AHA16SFRN27960011) awarded to Dr. Aggarwal.
Author Disclosure Statement
No conflicts of interest exist.
