Adding Sleep to Cardiovascular Health Metrics Predicts Disease Risk—The Original Framework Didn't
Sleep is one of the three pillars of health. Yet unlike diet and physical activity, the American Heart Association's cardiovascular health framework never included it. A study in 1,920 older adults tested whether that was a mistake.
The findings were striking: adding sleep to cardiovascular health metrics improved disease prediction. Even more notable was what happened when researchers tracked participants over time. The original cardiovascular health score failed to predict who would develop heart disease. Scores that included sleep succeeded.
This wasn't about optimization. It was about whether cardiovascular health assessment was missing something fundamental.
The Framework That Forgot Sleep
The American Heart Association's Life's Simple 7 has been the standard for measuring cardiovascular health since 2010. Seven metrics define ideal health: diet quality, physical activity, smoking status, body mass index, blood pressure, cholesterol, and blood glucose. A higher composite score across these metrics has been linked to lower cardiovascular disease risk in multiple studies.
But sleep was absent. Despite evidence linking sleep duration, quality, and disorders to cardiovascular outcomes with effect sizes comparable to diet and physical activity, no sleep metric was included. Approximately 35% of US adults sleep fewer than seven hours per night. Twenty percent report excessive daytime sleepiness. The prevalence of sleep disorders like insomnia and obstructive sleep apnea rivals that of other cardiovascular risk factors.
Researchers examined whether an expanded definition of cardiovascular health that includes sleep would be associated with cardiovascular disease risk. They tested this in the MESA Sleep Study—a subset of 1,920 participants from the Multi-Ethnic Study of Atherosclerosis who underwent comprehensive sleep assessment including overnight polysomnography, seven-day wrist actigraphy, and validated questionnaires.
The Sleep Landscape in Older Adults
The sleep data revealed how common poor sleep was in this population. Actigraphy showed 63% of participants slept less than seven hours per night, with 30% sleeping less than six hours. Sleep efficiency below 85% affected 10% of participants. Excessive daytime sleepiness was present in 14%, and 36% had clinically significant insomnia symptoms. Moderate-to-severe obstructive sleep apnea affected 47% of participants.
These sleep problems clustered. Short sleepers were significantly more likely to have low sleep efficiency, high night-to-night variability in sleep duration and timing, excessive daytime sleepiness, and obstructive sleep apnea. They also had higher prevalence of overweight and obesity, diabetes, and hypertension, along with lower cardiovascular health scores overall.
The cooccurrence of multiple sleep problems suggested that screening for one dimension of sleep health might identify individuals at higher risk across multiple domains.
Four Versions of the Essential Eight
Researchers computed the Life's Simple 7 score using standard criteria, then created four separate expanded cardiovascular health scores. Each version added the original seven metrics plus a sleep component scored from 0 to 2, making the total possible score range from 0 to 16 instead of 0 to 14.
CVH Score 1 included only sleep duration. Seven to nine hours was considered ideal, six to seven hours intermediate, and less than six or more than nine hours poor. This represented the simplest possible addition to cardiovascular health assessment.
CVH Score 2 incorporated sleep characteristics strongly linked to cardiovascular risk in existing literature: sleep duration, insomnia symptoms, daytime sleepiness, and obstructive sleep apnea. Meeting all four criteria earned an ideal score, meeting two to three earned intermediate, and meeting zero to one earned poor.
CVH Score 3 used sleep characteristics previously associated with cardiovascular disease in this specific cohort: sleep duration, sleep efficiency, daytime sleepiness, and obstructive sleep apnea. This represented commonly studied objective sleep metrics.
CVH Score 4 added measures of sleep regularity to Score 3, including night-to-night variability in both sleep duration and sleep timing. This represented a multidimensional approach to sleep health assessment.
The mean age of participants was 69 years, with 54% female. Racial and ethnic distribution was 40% White, 27% Black, 23% Hispanic, and 10% Chinese American. Mean follow-up was 4.4 years.
Prevalent Disease: All Scores Performed Similarly
Among the 1,920 participants, 95 had prevalent cardiovascular disease at the time of the sleep assessment. Researchers used logistic regression to examine associations between cardiovascular health scores and odds of having cardiovascular disease.
The Life's Simple 7 score was strongly associated with prevalent disease. Those in the highest versus lowest tertile had 75% lower odds of cardiovascular disease. All four expanded scores that included sleep showed similar or stronger associations. CVH Score 1 with sleep duration showed 71% lower odds. CVH Score 2 with literature-based sleep characteristics showed 80% lower odds. CVH Score 3 with cohort-specific sleep metrics showed 68% lower odds. CVH Score 4 with multidimensional sleep health showed 67% lower odds.
For cross-sectional associations with existing disease, adding sleep to cardiovascular health metrics did not substantially change performance compared to the original framework. All scores effectively identified who already had cardiovascular disease.
Incident Disease: Sleep Made the Critical Difference
The longitudinal analysis revealed something different. During 4.4 years of follow-up, 93 participants developed new cardiovascular disease. Cox proportional hazards models examined which cardiovascular health scores predicted who would develop disease.
The Life's Simple 7 score was not significantly associated with incident cardiovascular disease. The hazard ratio comparing highest to lowest tertile was 0.62 with a 95% confidence interval of 0.37 to 1.04. The traditional cardiovascular health framework failed to predict future disease in this cohort of older adults.
CVH Score 1, which added only sleep duration, was significantly associated with incident disease. Those in the highest versus lowest tertile had 43% lower cardiovascular disease risk. The hazard ratio was 0.57 with a 95% confidence interval of 0.33 to 0.97.
CVH Score 2 with literature-based sleep characteristics showed a hazard ratio of 0.66 but did not reach statistical significance. CVH Score 3 with cohort-specific sleep metrics showed a hazard ratio of 0.63, also not statistically significant.
CVH Score 4 with multidimensional sleep health was significantly associated with incident disease. Those in the highest versus lowest tertile had 47% lower cardiovascular disease risk. The hazard ratio was 0.53 with a 95% confidence interval of 0.32 to 0.89.
Even the simplest addition—sleep duration alone—improved cardiovascular disease prediction beyond the original seven-metric framework. The most comprehensive sleep assessment provided the strongest risk prediction.
Why Sleep Duration Alone Captured Risk
The finding that adding sleep duration alone improved disease prediction was initially surprising. But the clustering of sleep problems explained this. Short sleepers in this cohort had higher prevalence of poor sleep efficiency, high sleep variability, excessive daytime sleepiness, and obstructive sleep apnea. Screening for sleep duration identified individuals likely to have multiple concurrent sleep health problems.
This has practical implications for clinical assessment. While comprehensive sleep evaluation provides the most information, simply asking about sleep duration during routine visits may identify individuals at elevated cardiovascular risk when more detailed sleep assessment is not feasible.
The researchers noted that sleep duration represents the most widely measured and easily assessed aspect of sleep health, making it accessible to practitioners and individuals alike for monitoring over time.
The Disconnect Between Cross-Sectional and Longitudinal Findings
Why did the Life's Simple 7 predict prevalent but not incident cardiovascular disease? The researchers suggested several possibilities. The cross-sectional analysis reflected lifetime exposure to cardiovascular risk factors, while the longitudinal analysis captured risk over a relatively short 4.4-year period in older adults who had already accumulated decades of exposure.
Sleep may contribute uniquely to disease development through mechanisms not fully captured by traditional cardiovascular risk factors. Sleep duration, quality, and disorders have been linked to sympathetic nervous system activation, reduced parasympathetic activity, inflammation, oxidative stress, and endothelial dysfunction. Short sleep and irregular sleep patterns can disrupt circadian rhythmicity and cause circadian misalignment, leading to metabolic dysfunction.
Sleep also influences the other cardiovascular health metrics. Evidence shows that short sleep, poor-quality sleep, insomnia, and obstructive sleep apnea are associated with obesity, diabetes, and hypertension. In this cohort, short sleep, low sleep efficiency, and high sleep variability were related to higher body mass index and overweight or obesity. Sleep variability predicted metabolic syndrome. Poor sleep has been associated with higher caloric intake, unhealthy food choices, and lower physical activity.
By failing to account for sleep, the original cardiovascular health framework may have missed a fundamental contributor to disease risk that operates both independently and by influencing other health behaviors and risk factors.
What This Means for Cardiovascular Health Promotion
The criteria used to define cardiovascular health metrics include face validity, consistency with guidelines, simplicity and accessibility for practitioners and individuals, actionable targets, applicability across populations, and ease of measurement for monitoring over time.
Sleep meets all these criteria. The evidence base linking sleep to cardiovascular outcomes is substantial. National guidelines recommend seven to nine hours of sleep per night for adults. Sleep duration is simple to assess and monitor. Improving sleep is actionable through behavioral interventions and treatment of sleep disorders.
The approach to promoting cardiovascular health has traditionally focused heavily on diet and physical activity. This should be expanded to encompass behaviors across the 24-hour period, including sleep. Healthcare providers should assess patients' sleep patterns, discuss sleep-related problems, and educate patients about the importance of prioritizing sleep for cardiovascular health.
The formal integration of sleep into cardiovascular health promotion guidance would provide benchmarks for surveillance and ensure that sleep receives attention and resources comparable to other lifestyle behaviors in public health policy.
The Research That Still Needs to Happen
This study had limitations. The modest number of cardiovascular disease cases attributable to the 4.4-year follow-up period limited statistical power. The researchers could not test for subgroup differences by sex or race and ethnicity. The sleep health scores included important sleep-related cardiovascular risk factors but may not capture the full picture of an individual's sleep health.
Additional studies in external cohorts with larger sample sizes and longer follow-up are warranted to confirm these findings and examine this expanded definition of cardiovascular health in relation to subclinical cardiovascular disease and lifetime risk of cardiovascular outcomes. Clinical trials are needed to evaluate the influence of screening for sleep problems and improving sleep hygiene on cardiovascular outcomes.
The value of screening for self-reported sleep duration during clinic visits as a feasible and time-efficient approach for assessing sleep and improving cardiovascular disease risk prediction warrants further investigation.
The Bottom Line on Sleep and Heart Health
Adding sleep to cardiovascular health metrics improved disease prediction in older adults. The original seven-metric framework failed to predict who would develop cardiovascular disease over 4.4 years. Scores that included sleep succeeded.
Even the simplest addition of sleep duration alone improved risk prediction. The most comprehensive assessment of multidimensional sleep health provided the strongest prediction. For women in midlife and beyond, sleep is not optional for cardiovascular health. It belongs in the framework alongside diet, physical activity, and other established risk factors.
The evidence supports treating sleep as an equal counterpart to other health behaviors for cardiovascular disease prevention. Training that works addresses recovery as part of the program, not as an afterthought.
Study: Makarem, N., Castro-Diehl, C., St-Onge, M. P., Redline, S., Shea, S., Lloyd-Jones, D., Ning, H., & Aggarwal, B. (2022). Redefining cardiovascular health to include sleep: Prospective associations with cardiovascular disease in the MESA Sleep Study. Journal of the American Heart Association, 11, e025252.
