Daily Step Count and Sleep

Physical activity and sleep are often presented as two separate pillars of cardiovascular health. This study encourages us to consider them together—not to determine which compensates for the other, but to understand how they shape daily risk.

Physical activity and sleep play a central role in cardiovascular prevention, yet they are still too often studied as two separate behaviors. On one hand, daily activity—often measured by step count—reflects a concrete and easily interpretable aspect of the movement accumulated throughout the day. On the other hand, sleep duration represents a major component of the 24-hour cycle, with potential effects on metabolic, autonomic, inflammatory, and hormonal regulation. From a public health perspective, these two dimensions are particularly interesting because they are measurable, modifiable, and understandable by both patients and professionals.

The difficulty stems from the fact that these behaviors are not independent in real life. A very active person may sleep better, while short or fragmented sleep may reduce the desire or ability to move. Some studies have even suggested that a high level of physical activity could mitigate, or even neutralize, the increased cardiovascular risk associated with insufficient sleep. But this idea remains fragile, as much of the available data relied on self-reported questionnaires—measurements thus subject to memory errors, reporting biases, and often imprecise estimates of activity or sleep…

In adults followed for several years, are daily step count and sleep duration measured by accelerometer independently and jointly associated with the risk of major cardiovascular events, and can a high step count offset the risk associated with short sleep?

The study

Researchers conducted a prospective analysis using the UK Biobank cohort. The final sample included 88,012 participants with valid accelerometer data, no history of major cardiovascular events at the time of analysis, and the necessary covariates available. Participants were 62.2 years old on average, and 58% were women.

Between 2013 and 2015, participants wore an accelerometer on their dominant wrist for seven days. Using the raw data, open-source machine learning algorithms estimated two main variables: the median number of steps per day and the median duration of nighttime sleep. Participants were then classified into three step categories: low, under 7,500 steps per day; intermediate, between 7,500 and 11,000 steps; and high, over 11,000 steps. For sleep, the categories were as follows: short sleep, less than 6.5 hours; intermediate sleep, between 6.5 and 7.5 hours; and long sleep, more than 7.5 hours.

The primary outcome measure was the occurrence of a major cardiovascular event, defined as cardiovascular death, myocardial infarction, non-fatal stroke, or a revascularization procedure. These events were identified from electronic health records. The median follow-up was 7.9 years, during which 3,817 major cardiovascular events were observed. Of these, slightly more than half were non-fatal heart attacks or strokes, about a quarter were revascularization procedures, and just under a fifth were cardiovascular deaths.

The statistical analysis was based on Cox models adjusted for several factors likely to confound the association, including age, sex, ethnicity, education level, deprivation index, smoking, alcohol consumption, and processed meat consumption. The authors also explored the effect of cardiometabolic variables such as body mass index (BMI), HbA1c, blood pressure, and total cholesterol, but these analyses should be interpreted as descriptive rather than as formal evidence of mediation.

Results & Analyses

The main finding is the absence of a statistical link between step count and sleep duration in predicting major cardiovascular risk. The study does not show that sleep depends heavily on daily step count, nor that the effect of steps depends heavily on sleep duration. In other words, these two behaviors appear to contribute to risk largely independently, rather than through a mechanism of complete compensation.

The first independent association concerns the number of steps. Compared to participants taking more than 11,000 steps per day, those taking fewer than 7,500 steps had a 52% higher risk of a major cardiovascular event after maximum adjustment. The intermediate group, between 7,500 and 11,000 steps, also had a higher risk, though to a lesser extent, with a relative increase of 24%. This is consistent with the idea that a higher level of daily physical activity is associated with better cardiovascular protection; however, these thresholds should not be interpreted as universal guidelines. They represent a statistical categorization specific to this cohort, not individualized clinical recommendations.

The second independent association concerns sleep duration. Participants sleeping less than 6.5 hours per night had a 24% increased risk of major cardiovascular events compared to those sleeping between 6.5 and 7.5 hours. In contrast, long sleep, defined here as more than 7.5 hours, was not associated with a statistically significant increase in cardiovascular risk. This qualifies the conventional wisdom that long sleep is systematically harmful (even though some studies link it to a higher risk of cancer). In this study, it is primarily short sleep that emerges as a risk signal, while the long sleep category shows no clear independent excess risk.

The joint analysis helps clarify the hierarchy of profiles. The reference group consisted of participants with a high number of steps and an intermediate duration of sleep. The most unfavorable profile was one combining fewer than 7,500 steps per day and less than 6.5 hours of sleep, with an increased risk of approximately 84%. Participants with few steps but intermediate sleep duration already had a 38% increased risk, indicating that sleeping within the reference range is not sufficient to offset the adverse association with low step count. Conversely, participants with a high number of steps but short sleep duration retained a slightly higher risk, around 15%, suggesting that a high step count does not fully neutralize the signal associated with sleep deprivation.

Secondary results provide several important nuances. First, BMI modestly attenuates the association between low step count and cardiovascular risk, as well as that between short sleep duration and cardiovascular risk. This suggests that part of the link may involve or be associated with adiposity, without allowing for a causal link to be established. In contrast, HbA1c, blood pressure, and total cholesterol have little effect on the main estimates. Furthermore, sensitivity analyses show that reverse causality remains a concern. When events occurring in the first few years of follow-up are excluded, the association between low step count and cardiovascular risk weakens significantly. This may indicate that a latent, undiagnosed cardiovascular disease already reduces daily activity before becoming clinically apparent.

This caution is even more pronounced for sleep. When the authors exclude participants with a history of cardiovascular disease or cancer, the association between short sleep duration and cardiovascular events is no longer statistically significant. This does not mean that short sleep is unimportant, but that its interpretation is more nuanced. Sleep can be simultaneously a behavior, a marker of health status, a consequence of underlying pathologies, or an indicator of physiological and psychosocial stress. The study therefore does not allow us to conclude that mechanically increasing sleep duration alone will reduce cardiovascular risk.

The proposed mechanistic interpretation must be viewed with caution. For the benefits, the authors cite plausible pathways already consistent with the literature: improved insulin sensitivity, reduced systemic inflammation, and improvements in certain lipid parameters and ambulatory blood pressure. For short sleep, possible mechanisms include increased activation of the sympathetic nervous system, immune dysregulation, and endocrine disturbances that may promote atherosclerosis. However, in this study, these mechanisms are not directly demonstrated.

Practical Applications

The first application is to avoid simplistic compensation reasoning. In middle-aged or older adults, walking a lot does not seem to completely offset the risk associated with short sleep. Conversely, getting an intermediate amount of sleep is not enough to neutralize the adverse signal of a low step count. For cardiovascular prevention, this study therefore supports a two-pronged approach: maintaining a sufficient level of daily physical activity and ensuring a sleep duration compatible with proper recovery.

For healthcare professionals, step count can serve as a simple indicator of daily physical activity. A daily step count below 7,500 appears here as an unfavorable signal, especially when accompanied by short sleep duration. This does not mean that every patient should immediately aim for more than 11,000 steps per day. For certain individuals—particularly the elderly, those who are deconditioned, in pain, or have risk factors—the practical goal will first be to gradually increase the amount of walking, reduce sedentary periods, and improve the regularity of movement.

For people who are already very active, the message is different. A high number of steps is associated with the most favorable risk profile, but it does not make sleep deprivation a negligible factor. For a recreational athlete, a highly active professional, or a patient who walks a lot but chronically sleeps less than 6.5 hours, the intervention should not focus solely on physical activity. It should also address sleep patterns, the regularity of schedules, work load, family demands, symptoms of sleep disorders, and stressors.

This study also urges against an anxiety-inducing interpretation of “long” sleep. In this cohort, sleeping more than 7.5 hours was not independently associated with a higher cardiovascular risk. Long sleep may sometimes reflect a medical condition, chronic fatigue, or insufficient recovery, but it should not automatically be interpreted as dangerous behavior. The clinical context, sleep quality, daytime symptoms, and overall health status remain essential. Note that in other studies, long sleep is defined as sleeping more than 9 hours and is associated with an increased risk of cancer.

Finally, it should be noted that the measurements are based on a single week of accelerometry. This week provides a more objective estimate than a questionnaire, but it does not necessarily capture habits spanning several years. In practice, healthcare professionals should prioritize repeated monitoring, trends over time, and overall lifestyle consistency rather than a single isolated value.

Reference

Yun J, Brocklebank L, Harper C & Doherty A. Joint associations of device-measured step count and sleep duration with incident major adverse cardiovascular events: prospective analysis of the UK Biobank. eClinicalMedicine, 92: 103769, 2026.