Biological pathways underlying the association between habitual long-sleep and elevated cardiovascular risk in adults

Abstract

Cardiovascular disease is the leading cause of death in the world. In addition to non-modifiable factors such as age and sex, cardiovascular risk is also driven by behavioral, and therefore somewhat modifiable, factors such as physical activity, diet, and sleep. It is well established that sleep duration has a U-shaped association with mortality and cardiovascular disease, with recent evidence suggesting that this association is observed even while controlling for the effects of comorbid conditions. Whereas several biological mechanisms mediating the association between chronic short sleep duration and cardiovascular risk have been established, the biological mechanisms underlying the relationship between habitual long sleep (≥ 9 h) duration and cardiovascular risk, in the absence of other chronic diseases, are not well understood. This review will focus on summarizing the literature investigating the mechanisms underlying the association between habitual long sleep duration and cardiovascular risk. We will also propose the mechanistic pathways, distinct from the ones for short sleep, by which habitual long sleep can increase cardiovascular disease.

1.1. Sleep duration and cardiometabolic health.

Sleep is a fundamental biological process needed to sustain life [1], with habitual sleep durations between 7 and 9 hours necessary to maintain cardiovascular (CV) and metabolic health in adults [2–4]. Sleep duration changes across the lifespan; toddlers and children sleep for much longer durations compared to midlife adults and this trend toward shorter sleep duration continues into old age [5]. The vital role of sleep in maintaining health has been elucidated through observing physiological changes associated with sleep restriction. Lack of sleep and disrupted sleeping patterns can adversely affect CV parameters, and these sleep-related changes can chronically increase overall CV disease risk [6]. For example, short sleep duration, typically defined as ≤ 6 hours per 24 hours, has been associated with increased sympathetic activity and suboptimal reduction in nocturnal blood pressure, both of which contribute to the pathophysiology of hypertension due to persistently increased stress on the heart and vasculature [7–9]. Acute and chronic sleep deprivation can also accelerate the process of atherosclerosis [10] by multiple pathways including increased oxidative stress hypercholesterolemia [4], and increases levels of pro-inflammatory cytokines interlukin-6 (IL-6), tumor necrosis factor-α (TNF-α) [11], and C-reactive protein (CRP) [12]. Short sleep can also increase CV risk by impairing glucose tolerance and insulin sensitivity [3], decreasing circulating leptin levels while elevating ghrelin levels (when caloric intake is clamped to maintain an individual’s energy balance at baseline) [13], and increasing the activation of brain regions associated with food intake [14], all of which can promote the development of obesity and type 2 diabetes [2]. Such metabolic dysregulation is associated with elevated carotid intima-media thickness and decreased vascular reactivity, which are both predictive of atherosclerosis [15–17]. Thus, via hypertension, atherosclerosis, inflammation, metabolic abnormalities, and vascular abnormalities, short sleep produces a systemic, pro-atherogenic milieu and leads to increased CV risk (Figure 1). Similar to short sleep duration (≤6 hours), excessive sleep duration (typically defined as ≥9 hours sleep per 24 hours) has also been associated with a plethora of chronic health conditions [2]. Both long and short sleepers are at increased risk for mortality (long sleep risk ratio 1.12, short sleep risk ratio 1.3) [18], incident type 2-diabetes [19], stroke [20], and CV disease [2]. Yet, only shorter sleep durations have been found to be associated with hypertension [21] and obesity [22] suggesting that the adverse health outcomes observed in individuals obtaining either short or long sleeping durations may be acting via different mechanisms.

Figure 1.

Adverse changes associated with short sleep duration. Short sleep (≤6 hours) has metabolic, hormonal, and inflammatory effects which contribute to the development of atherosclerosis and cardiovascular disease.

1.2. How is long sleep related to cardiovascular risk?

It is generally believed that habitual long sleep duration is associated with increased CV risk due to underlying co-morbid conditions [23, 24] which themselves lead to increased sleep/bed rest. Yet, emerging evidence suggests that the relationship between habitual long sleep duration and CV risk may hold even in the absence of co-morbid conditions (Table 1). For instance, a recent systematic review and meta-analysis has concluded that habitual long sleep duration in adults (≥9 hours) is significantly related to mortality (relative risk ratio (RR) 1.39), incident diabetes mellitus (RR 1.26), stroke (RR 1.46), CV disease (RR 1.25), and coronary heart disease (RR 1.24)) [25]. Similarly, a large (n=461,347) study which utilized data from the UK Biobank concluded that long sleep duration can increase the risk for incident myocardial infarction even when controlling for genetic predispositions, lifestyle variables, and depression [26]. Yet, studies evaluating the mechanistic underpinnings of long sleep and cardiovascular risk are rare, and to our knowledge, no experimental studies have been conducted in this area. Below, we present previously proposed biological and psychosocial explanations, and attempt to synthesize a pathway connecting long sleep duration with CV risk. Through our interpretation of the limited available literature, we theorize that habitual long sleep may elicit a pro-inflammatory metabolic state, which in addition to a generally unhealthy lifestyle may create optimal conditions for development and progression of CV disease.

Table 1.

Current knowledge on the association between long sleep and CVD.

Reference	Study design	Participant details	Measures	Results
Cappuccio et al. 2011 [2]	Meta-analysis of 15 prospective studies; follow-up >3 y; incident cases of coronary heart disease (CHD), stroke, or CVD	474,684 male and female participants with 4169 CHD events and 8420 CVD events in a follow-up of 6.9-25 y	Self-reported sleep duration and incident cases thru certification and event registers, risk ratios (RRs) for incident cases	Long sleep (>8-9 h) associated with greater risk of CHD (RR 1.38), stroke (1.65), and total CVD (1.41)
Jike et al. 2017 [25]	Meta-analysis of 137 prospecive studies; incidence of diabetes mellitus, hypertension, CVD, stroke, CHD, obesity, depression, and dyslipidemia.	5,134,036 participants	Self-reported sleep duration and incident health conditions at follow-up, risk ratios (RRs) for mortality and incident health conditions	Long sleep (>9 h or >10 h) associated with mortality (RR, 1.39), diabetes mellitus (1.26), CVD (1.25), stroke (1.46), CHD (1.24), and obesity (1.08)
Daghlas et al. 2019 [26]	Prospective observational analysis using Mendelian randomization to assess causality	461,347 UK Biobank participants free of CVD	Self-reported sleep duration and estimated multivariable adjusted hazard ratios (HR) for myocardial infarction (MI)I; examined joint effects with sleep disturbance traits and a coronary disease genetic risk score	Long sleepers (>9 h) had a 34% higher risk of incident MI, healthy sleep durations (6-9 h) mitigated MI risk even with high genetic predispositions
Zhao et al. 2019 [41]	Cross-sectional study using a sleep factor questionnaire (SFQ) and the prevalence of hypertension	1518 participants recruited from health examination population at primary care setting of Jiujiang University Hospital	SFQ to assess wide range of sleep characteristics and their association with the prevalence of hypertension	Compared to sleep durations of 7-8 h, sleep durations of 8-9 h resulted in greater odds for hypertension OR=1.60 and sleep durations >9 h had an OR= 2.39

2.1. Possible mechanisms connecting habitual long sleep duration to CV risk (Table 2)

Table 2.

Summary of possible mechanisms connecting long sleep with cardiovascular risk factors.

Pathway to Cardiovascular Disease	Correlates to long sleep
Metabolic	-low HDL [36, 37] -hyperglycemia [27] -hypertriglyceridemia [27] -insulin dysregulation and resistance [33]
Inflammation	-high CRP [37, 39] -high IL-6 [37, 39]
Behavioral/Lifestyle	-depression [43, 44] -low socio-economic status [43, 44] -unemployment [43, 44] -insufficient physical activity [43, 44] -increased alcohol consumption [37] -smoking [37]

^*[31][36][43] define long sleep as ≥8 h

^**[27][33][34][42] define long sleep as ≥9 h

2.1.1. Long sleep and metabolic dysregulation

Several studies have found a U-shaped association specifically between sleep duration and metabolic syndrome, but the mechanistic pathways can differ between short and long sleep [27–30].

Moreover, it has been observed that long sleep is associated with risk for incident diabetes and diabetes comorbid with coronary heart disease [31], but that the association between long sleep and coronary heart disease does not hold in the absence of diabetes mellitus [31]. This suggests that the relationship between long sleep and CV disease may be dependent on underlying metabolic dysfunction. Almost all patients with type 2-diabetes are dyslipidemic due to underlying insulin deficiency or resistance [32] and dyslipidemia coupled with the presence of proatherogenic lipoproteins, HDL deficiency, hypertriglyceridemia, endothelial dysfunction, and the chronic low-grade inflammation characteristic of diabetes mellitus provides the perfect conditions for the development of CV disease [32]. In one study investigating the association between extreme (≤6 hours or ≥9 hours) sleep durations and metabolic syndrome, chances of suffering from metabolic syndrome increased by more than 45% for both short and long sleepers compared to 7-8 hour-sleepers, even after adjustment for factors such as demographics, physical activity, and depression. National survey data from Korea has also revealed the characteristic U-shaped distribution between sleep duration and the various components of metabolic syndrome [27]. While abdominal obesity and hypertension were most prominent in short sleepers, long sleepers had more prevalent hyperglycemia and hypertriglyceridemia [27]. Furthermore, in the Prevalence, Prediction and Prevention of diabetes (PPP)-Botnia Study in Finland, long sleepers were shown to have elevated fasting insulin, 120-min insulin following glucose administration, insulin resistance, and a lowered insulin sensitivity index independent of sleep apnea or insomnia [33]. Such insulin dysregulation could lead to diabetes mellitus [33] and subsequently increase CV risk. The association between long sleep and diabetes mellitus is clinically relevant, as diabetes is closely linked with the upregulation of inflammatory pathways and atherosclerosis [34, 35]. Furthermore, long sleep has specifically been associated with decreased levels of HDL [36, 37], an anti-atherogenic biomarker. Thus, habitual long sleep could promote a specific subset of metabolic abnormalities, mainly related to lipid dysregulation and compared to short sleep, lead to increased CV disease via a distinct metabolic pathway.

2.1.2. Long sleep and inflammation

A study on the Health, Aging, and Body Composition in community-dwelling older adults found that sleep duration only had the U-shaped association with mortality when underlying inflammation or chronic health conditions were present, particularly diabetes mellitus [38]. Furthermore, the relationship between short sleep and mortality was more attenuated by inflammation than that of long sleep, which was predominantly explained by lifestyle factors. Yet, it has been consistently found that long sleep duration is associated with elevated levels of the pro-inflammatory cytokines CRP and IL-6. In the Cleveland Family Study cohort, there was a positive linear association between additional hours of habitual sleep duration and the levels of these two pro-inflammatory cytokines. For each additional hour of habitual sleep duration above the mean of 7.6 hours, CRP levels increased by 8% and IL-6 levels increased by 7% [39]. This same trend was detected, albeit more dramatically, in the Nurses’ Health Study, wherein women who habitually slept 9 or more hours had 44% higher levels of CRP than women who habitually slept 8 hours [37]. CRP and IL-6 levels typically increase simultaneously as IL-6 stimulates CRP production, and both these cytokines are predictors of risk for myocardial infarction and diabetes [39]. The endothelial expression of these cytokines can also contribute to the development of hypertension, a known risk factor for the development of vascular disease [40]. It has been recently shown that both short (6-7 h) and long (over 9 h) sleep durations are associated with hypertension [41]. Increased inflammation due to long sleep can also be mediated by the presence of comorbidities such as diabetes [31]. Indeed, disturbed glucose homeostasis, and metabolic syndrome more generally, are linked with the secretion of chemokines and cytokines indicative of endothelial dysfunction and subclinical, chronic low-grade inflammation [34, 35]. There are also strong correlations between impaired glucose homeostasis, pro-inflammatory biomarkers (CRP, IL-6, and TNF-α) and atherosclerosis [35]. A paradox to long sleep as a cause of inflammation is that pro-inflammatory cytokines are generally observed to be sleep inducers. It has been hypothesized that they may function to promote sleep in a diseased state to allow for better recovery [37]. The Cleveland Family Study attempted to correct for this issue of directionality by using the Epworth Sleepiness Scale to assess sleep propensity. They found that the strength of association between sleep duration and the levels of CRP/IL-6 was unaffected by sleep propensity adjustment [39]. This suggests that long sleepers who have elevated levels of these pro-inflammatory cytokines do not necessarily have a higher subjective sleep propensity, and while this finding does not definitively suggest that the cytokines are not responsible for lengthened sleep duration, it offers mild support for long sleep as a causal factor for the inflammation-mediated increase in CV risk.

2.1.3. Long sleep and exogenous factors

In order to obtain a comprehensive understanding of the influence of long sleep duration on CV risk, it is important to consider the behavior or lifestyle choices (demographics, health behaviors, family environment) of long sleepers even in the absence of overt comorbid conditions. While these factors are statistically controlled for in most of the aforementioned studies, it is possible that these lifestyle choices are at least in part contributing to adverse outcomes. One study used polysomnography measures to conclude that that people with longer self-reported sleep duration spent more time in bed (not necessarily sleeping) and tended to have more CV disease, and depression which could be indicative of social isolation [42]. Increased time in bed also reduced time spent being physically or socially active, which is predictive of negative health outcomes [42]. Several CV disease-promoting lifestyle and psychological factors including depressive symptoms, low socio-economic status, unemployment, and insufficient physical activity are more prevalent in long sleepers [43, 44]. Additionally, among women in the Nurses’ Health Study, increased alcohol intake and smoking were more common in long sleepers [37]. This, it is possible that unusually extreme habitual sleep duration reflects abnormalities at the biological, psychological, or social levels [45]. If long sleepers indeed engage in less healthy behaviors, this would help partially explain the association between prolonged habitual sleep and negative health outcomes. It has been shown that interventions in weight control, improved diet, and increased physical activity attenuated the association between long sleep and type 2-diabetes [46]. This again suggests that in people with habitual long sleep duration, lifestyle factors may be involved. It is therefore important for both researchers and practitioners alike to recognize that long sleepers may also have a higher propensity toward unhealthy behavior or may not have the means to improve their health naturally.

We also must consider the possibility that long sleep may contribute to negative metabolic and inflammatory outcomes in part due to the increased time being sedentary in bed. While it is difficult to separate these two factors without a proper experimental study, we can examine the cardiovascular implications of bed rest as an extreme to long sleep duration. It has been noted that long sleepers are less likely to get sufficient exercise [43, 44], and longer periods in bed, in addition to chronic lack of sufficient exercise, can contribute to a pro-atherogenic inflammatory state [47]. Similarly, five days of bed rest in previously healthy subjects has been shown to elevate blood glucose and insulin levels in both the fed (glucose loaded) and fasted state, indicative of early insulin resistance [48]. Furthermore, this duration of bed rest can also lead to elevated total cholesterol and triglycerides, which can in turn promote atherogenesis [48]. While bed rest experiments severely restrict activity, and cannot be directly compared to an increased sleep paradigm, it is worthwhile to experimentally differentiate the effect of long sleep duration from prolonged time in bed and study the effects on cardiometabolic physiology.

Finally, the use of certain medications has been shown to be associated with long sleep durations.. For instance, while those with a history of depressive symptoms are 1.3 times more likely to sleep long durations, those taking antidepressants are 3.5 times more likely to do so [49]. Similarly, the use of sleep-aids like Valium and Xanax make someone three times as likely to sleep long durations [49]. Sleep-aids are a commonly prescribed treatment for insomnia which can be caused by a variety of factors including lifestyle stressors or comorbid conditions like mental illness, chronic disease, and sleep-related breathing disorders [50]. Those taking sleep aids, and therefore sleeping long durations, likely have insomnia brought on by stress, chronic disease, or sleep disturbances, all of which contribute to the development of CVD. Thus, sleep aids are associated with cardiovascular disease both directly and indirectly and could be a possible link between cardiovascular disease and long sleep. Another possible mechanism underlying long sleep and CVD is is the higher prevalence of snoring [51] and moderate to severe obstructive sleep apnea (OSA) [52] among long sleepers. OSA is a risk factor for hypertension, coronary heart disease, heart failure, and arrhythmias [53]. This connection is often attributed to pro-inflammatory and pro-thrombotic factors that promote atherosclerosis, along with a general elevation in sympathetic activity [53]. Thus, the presence of (undiagnosed) OSA could also explain why long sleepers are more likely to suffer from cardiovascular disease.

3.1. Conclusions and clinical implications

There is currently a lack of experimental evidence for a causal relationship between long sleep and cardiovascular risk in adults, an association that is clinically and biologically important. From the available evidence, we hypothesize that the underlying mechanisms are metabolic in nature which work to promote CVD via an inflammatory pathway (Figure 2). Specifically, long sleep could lead to low HDL [36, 37], hyperglycemia, hypertriglyceridemia [27], and insulin resistance [33], all of which can lead to vascular endothelial dysfunction and subclinical inflammation which further promote atherosclerosis [34, 54]. This pro-atherogenic environment may be worsened by associated societal, lifestyle, and behavioral factors like substance use, low physical activity, or limited access to healthy foods [43, 44]. Regardless of the true direction of causality, we encourage assessment of sleep duration in clinical assessments as either short or long sleep duration could indicate risk for chronic conditions. Because perilous diseases like CV disease and diabetes mellitus are highly prevalent in our population, causing premature morbidity and mortality, and because sleep is a fundamental behavior, it is important to investigate the relationship between long sleep and chronic disease to inform optimal preventative strategies against CV disease, an important step toward a healthier population on the national and global scale.

Figure 2.

Proposed pathway assuming causal influence of sleep (≥9 hours). Long sleep promotes the development of metabolic abnormalities and is associated with less health-promoting behavior, which worsens the metabolic condition. Disturbed metabolic homeostasis results in chronic, low-grade inflammation which may contribute to long sleep in a cyclical fashion.

Highlights.

• Short and long sleep increase cardiovascular (CV) risk via distinct pathways

• Long sleep associated is with pro-inflammatory cytokines and metabolic dysfunction

• Long sleepers are more likely to have physically and mentally unhealthy lifestyles