The Impact of Sleep on Cardiovascular Disease

It is difficult to overestimate the importance of sufficient quality and duration of sleep on nearly every body system, and multiple recent trials emphasize the critical role it plays in cardiovascular health, including implications for causality to both peripheral and coronary artery disease.

The first of these was recently published in the European Heart Journal Open and includes analyses of the association between sleep and peripheral artery disease (PAD) in multiple cohorts. This included a Swedish cohort of approximately 50,000 participants (the Swedish Infrastructure for Medical Population-based Life-course and Environmental Research (SIMPLER) cohort), case-control data from over 150,000 participants in the Million Veteran Program (MVP), and over 450,000 participants from the UK Biobank cohort. In addition to analyses of observational data from these cohorts, Mendelian randomization was used to further assess the observational associations. Mendelian randomization (MR) has the advantage of limiting several types of bias, including reverse causation. For example, because the DNA variants associated with sleep-related traits were randomized at birth, MR can be thought of as a type of randomized trial, evaluating the effect of inheriting variants known to influence sleep.

The results of both the observational and MR analyses indicated that a bidirectional relationship between sleep and PAD is likely, with a U-shaped curve best fitting the association between sleep duration and PAD. Using between 7-8 hours of sleep per night as a reference, the SIMPLER cohort found a 74% higher risk for PAD among those sleeping < 5 hours, and a 24% increase in risk for those sleeping 8 or more hours per night. Additionally, taking daytime naps was associated with a 32% greater risk for PAD. Similarly, the MVP data revealed a 38% greater risk for PAD among people sleeping 5 hours per night vs. 7, and a 48% increase in those sleeping 10 hours per night. The UK Biobank observational data indicated a similar connection, with an 87% greater risk for 5 hrs/night, and an increase of 14% for 8 or more hours, along with a 32% greater risk associated with daytime napping.

The MR analysis found the strongest association for short sleep; for each additional hour of genetically predicted sleep, there was a 30% reduction in PAD risk. While longer sleep duration and napping did not appear to be detrimental in the MR analysis, a bidirectional effect was observed between sleep duration and PAD, i.e., a genetic risk for PAD increased the risk for shorter sleep duration. The MR makes a stronger case for causality, with shorter sleep likely to be a causal factor for PAD and vice versa.

A second recent study was published in the Journal of the American Heart Association, providing an analysis of the relationship between sleep and cardiovascular disease (CVD) in the Multi‐Ethnic Study of Atherosclerosis (MESA) Sleep Study, a cohort of nearly 2,000 older adults (mean age 68.5) with extensive sleep data, including polysomnography. This analysis was done because the American Heart Association currently uses the “Life Simple 7” (LS7) score, which includes 7 metrics associated with CVD, such as smoking, diet, exercise, BMI, etc., but does not include any sleep-related measures, and the authors wanted to test the inclusion of sleep-related parameters to help predict CVD risk.

Surprisingly, this analysis found that although the existing LS7 score did predict CVD prevalence, it was not predictive of CVD incidence. Alternatively, several different scores which incorporated sleep were predictive of both CVD incidence and prevalence. For example, those with a sleep score in the highest tertile of a metric that only included sleep duration, there was a 43% lower risk of CVD. Using a score that incorporated additional measures, including duration, sleep efficiency, daytime sleepiness, etc., revealed a 47% lower risk of incident CVD with more optimal sleep. This study makes a strong case that an evaluation of sleep should be incorporated into the assessment of CVH risk, especially given that other standard measures in the LS7, including BMI, blood pressure, cholesterol, etc., were less predictive than metrics that included sleep.

A third recent study that also supports the plausibility of sleep as playing a causal role in CVD was published in Frontiers in Cardiovascular Medicine. In brief, this was also a Mendelian randomization, whose study participants had diabetes, nearly 4,000 that also had coronary artery disease (CAD), and 12,000 controls with only diabetes. Five sleep traits were analyzed, and a significant 16.3% increase in risk for CAD was found among those with a genetic predisposition for insomnia, with borderline associations for sleep duration and snoring.  Because of the nature of a Mendelian randomization analysis, insomnia as a causal factor for CAD, at least in diabetics, appears likely.

Of course, poor sleep is a risk factor for not only CVD, but also for obesity, diabetes, overall mortality, etc., but given that CVD is the leading cause of death among men and women in the US, and 30-50% of adults may be getting insufficient sleep, assessment (as well as treatment to improve sleep) should be routine practice. But it’s worth mentioning that the impact of poor sleep has quite broad implications. For example, a recent meta-analysis published in Current Biology found that less than 6 hours of sleep per night (objectively assessed) was associated with a robust decrease in vaccine-induced antibody response, though the effect was limited to men. This analysis included 7 viral vaccine studies (influenza, hepatitis A and B), and provides another motivation to assess and improve sleep.