The research team, led by investigators at the University of Texas Health Science Center at Houston, analyzed self-reported data from 11,039 adults aged 18 to 65 years. The sample had a mean age of 41 years, was slightly majority female at 50.7 percent, and reflected the country’s diversity: 59.8 percent of participants were non-Hispanic White, 17.8 percent Hispanic or Latino, 12 percent non-Hispanic Black, and 6.3 percent non-Hispanic Asian. Participants reported whether they engaged in moderate or vigorous physical activity at work, defined as tasks such as carrying light loads, household chores, and yard work, or in recreational settings, such as running, swimming, and team sports. Moderate activity was described as movement causing a small increase in breathing or heart rate sustained for at least ten minutes, while vigorous activity caused large increases over the same duration.

Sleep behaviors were captured through a battery of NHANES questions covering five domains: usual sleep duration on weekdays or workdays, frequency of snoring over the past year, episodes of snorting, gasping, or breathing cessation during sleep, whether a doctor had ever been told about trouble sleeping, and how often the participant felt excessively sleepy during the day in the past month. Short sleep duration was defined as fewer than seven hours per night, in line with recommendations from the American Academy of Sleep Medicine and the Sleep Research Society that adults obtain at least seven hours of nightly sleep. The researchers then applied weighted logistic regression models, adjusting for age, sex, ethnicity, body mass index categories, and socio-economic status proxied by the family income to poverty ratio.

The headline finding is striking. Moderate work activity increased the odds of short sleep duration by 42 percent, snoring by 45 percent, breath cessation during sleep by 37 percent, and daytime sleepiness by 63 percent. Vigorous work activity was even more strongly linked to short sleep, raising its odds by 72 percent, alongside a 42 percent increase in the odds of snoring, a 31 percent increase in breath cessation, and a 21 percent increase in daytime sleepiness. In contrast, vigorous recreational activity moved in the opposite direction, reducing the odds of short sleep duration by 19 percent and the odds of trouble sleeping by 17 percent. Moderate recreational activity showed no significant associations with sleep behaviors after adjustment, suggesting that the protective effect of leisure movement may require a sufficiently vigorous intensity.

The burden of poor sleep associated with work-related exertion was not distributed evenly across the population. In stratified analyses, non-Hispanic Black participants engaging in moderate work activity showed consistently elevated odds of all measured sleep behaviors, with the highest odds observed for daytime sleepiness. Hispanic participants engaging in moderate work activity had the highest odds of snoring within that group, while among those performing vigorous work activity, Hispanic participants showed the greatest odds of short sleep duration and non-Hispanic Asian participants the greatest odds of snoring and of trouble sleeping. These ethnicity-specific patterns emerged even though a formal interaction analysis found no statistically significant effect modification, indicating that the direction of associations was broadly similar but the magnitude differed meaningfully across groups.

The authors propose several mechanisms that could explain why work-related activity appears to harm sleep while recreational activity appears to protect it. Occupational physical exertion often coincides with long hours, shift schedules including overnight work, and physically demanding conditions that disrupt circadian rhythms and promote fatigue and exhaustion. Prior research has estimated that moderate-to-vigorous work activity is associated with roughly a 15 percent increase in the likelihood of disordered sleep. Stress is another candidate driver: elevated cortisol levels, mediated through the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system, are well-established predictors of poor sleep quality and insomnia, and occupational pressures, financial instability, and socio-economic disadvantage can sustain chronic stress responses that fragment sleep.

By contrast, the benefits of leisure-time exercise are extensively documented. Regular moderate-to-vigorous recreational activity promotes cardiovascular health, favorable genetic biomarker profiles, and musculoskeletal fitness, which in turn support glucose homeostasis, bone health, psychological well-being, and overall quality of life. Recreational exercise has also been shown in previous studies to protect against the development of sleep disorders, a finding consistent with the reduced odds of short sleep and trouble sleeping observed here. The distinction between the two contexts of movement may therefore reflect not the physiology of exercise itself but everything that surrounds it: autonomy, scheduling, stress, and recovery time.

Metabolic health added another layer of complexity. Participants with obesity, who made up 40.2 percent of the sample, showed increased odds of short sleep duration, snoring, and breath cessation, consistent with the established link between adiposity and obstructive sleep apnea, in which fat deposited around the neck contributes to airway obstruction during sleep. Hormonal disruptions associated with obesity, including elevated evening cortisol and ghrelin and reduced leptin, are known to be bidirectionally related to sleep loss. Intriguingly, underweight participants engaging in moderate recreational activity had more than three times the odds of experiencing breath cessation compared with those who did not, a finding the authors suggest may relate to reduced lung capacity and weaker respiratory muscles in underweight individuals, though they caution that further longitudinal research is needed to confirm this mechanism.

The study’s strengths include its large, nationally representative sample, the use of NHANES’s complex weighting methodology to account for the clustered, stratified sampling design, and its unprecedented attention to the type and intensity of physical activity alongside ethnicity and body weight. Still, the authors are careful to note the limitations inherent in the design. The cross-sectional data cannot establish temporality or causation, and both sleep behaviors and physical activity were self-reported rather than confirmed by polysomnography or measured objectively with accelerometers. The analysis also could not account for potential confounders such as occupation type, workplace conditions, marital status, stress, and environmental factors. The authors call for prospective longitudinal studies to clarify the mechanisms linking different forms of physical activity to sleep health and to examine the social and workplace factors that may drive the disparities observed across ethnic and weight groups.

For the public, the message is nuanced rather than simple. Moving more remains one of the most reliable routes to better health, and only about half of American adults currently meet federal guidelines of 150 to 300 minutes of moderate or 75 to 150 minutes of vigorous aerobic activity per week. But this research suggests that the clock and the context of that movement matter: exertion embedded in demanding, potentially stressful work was associated with shorter, more disturbed sleep, while vigorous exercise chosen freely in leisure time was associated with longer and sounder rest. As sleep scientists push to elevate sleep health on public health agendas, the findings underscore that interventions to improve sleep may need to look beyond exercise volume and address the conditions under which people are physically active, particularly for the ethnic and weight groups that bear the heaviest burden of poor sleep.

Subject of Research: Associations between work-related and recreational physical activity and sleep behaviors in US adults

Article Title: Unraveling the relationship between sleep behaviors and physical activity in the United States: insights from 2015 to 2020 NHANES data

Article References: Ernest, D. K., Singh, B., Shamna, F., Onugha, E. A., & Sharma, S. V. (2024). Unraveling the relationship between sleep behaviors and physical activity in the United States: insights from 2015 to 2020 NHANES data. Journal of Activity, Sedentary and Sleep Behaviors, 3(1), Article 31. https://doi.org/10.1186/s44167-024-00070-8

Image Credits: AI Generated

DOI: 10.1186/s44167-024-00070-8

Keywords: sleep health, physical activity, NHANES, work activity, recreational exercise, sleep duration, sleep apnea, ethnic disparities, obesity, daytime sleepiness, cross-sectional study, public health

Healthier SG represents a deliberate shift in how the city-state approaches health. Rather than paying clinics to treat illness after it appears, the programme asks residents to empanel, formally register, with a general practice clinic or a polyclinic, creating a continuity of care relationship intended to catch chronic diseases early and manage them better. It opened to adults aged 60 and above in mid-2023, and eligibility was extended to those aged 40 and older by late the same year. Voluntary empanelment of this kind has been adopted by health systems around the world as a strategy for strengthening primary care, but researchers note that its equity implications, particularly in multi-ethnic Asian populations, have rarely been examined in detail. The Singapore team set out to fill that gap, asking not just whether people enrol, but who makes it through each step of the journey, and who is left behind along the way.

To answer that question, the researchers turned to the Singapore Population Health Studies, drawing on a sample of 3,899 residents aged 40 and above. They framed the analysis using Andersen’s Behavioural Model of Health Services Utilisation, a longstanding framework in health services research that organises the drivers of care use into three categories: predisposing factors such as age, sex, education and ethnicity; enabling factors such as income, housing type and employment; and need factors such as the burden of chronic disease. Engagement with Healthier SG was classified into four sequential stages: people who were unaware of the programme and not enrolled, those who were aware but had not enrolled, those who had enrolled but not completed a screening, and those who had both enrolled and been screened. Using multinomial logistic regression, the team estimated odds ratios and marginal effects, the predicted probabilities of occupying each stage, for every factor.

The headline numbers are striking. Half of the respondents, 50.5 per cent, had never heard of Healthier SG, despite the programme’s nationwide promotion. Only 36.1 per cent had either enrolled or completed at least one of the preventive screenings the programme offers. In other words, barely more than a third of the eligible population had moved beyond the starting line of a scheme designed to reach everyone. For a programme that strips away most of the cost of preventive care, that level of disengagement is a signal that something other than money is shaping behaviour, and the statistical analysis pointed clearly at what.

Socio-economic advantage left a visible fingerprint on every stage of engagement. Older adults, people with higher educational attainment, retirees, and residents living in higher-tier housing were all significantly more likely to both enrol in the programme and complete their screenings. The housing gradient is particularly telling in Singapore’s context, where housing type serves as a widely used proxy for socio-economic status. Residents in larger, higher-value public and private flats navigated the programme more successfully than those in smaller units, suggesting that differences in health literacy, digital access, time flexibility and familiarity with the healthcare system continue to operate even when the price of care has been heavily subsidised. Education, similarly, appeared to function less as a route to affording care than as a route to understanding and trusting it.

Ethnic disparities emerged in a more nuanced pattern. After the researchers adjusted for socio-economic factors, respondents of Indian ethnicity showed lower odds of enrolling in the programme, a gap that persisted even when financial and educational differences were accounted for. Malay respondents, by contrast, showed higher odds of being aware of Healthier SG without actually enrolling, a pattern the authors describe as awareness without conversion. That distinction matters: it suggests that for some communities the barrier is not information but something that happens at the point of deciding to sign up, whether that involves trust in primary care institutions, cultural fit of communication materials, differing health beliefs, or practical frictions in the enrolment process itself. The findings point to the need for culturally responsive outreach rather than a uniform national campaign.

One of the study’s most consequential findings concerns need, the factor that health policy most wants to drive preventive care use. People carrying a higher burden of chronic disease were more likely to complete screenings once they had enrolled, which is exactly what the programme is designed to achieve. But a higher chronic disease burden did not make people any more likely to enrol in the first place. The residents with the greatest clinical need were not gravitating toward the programme on their own. In a voluntary system, the people who stand to gain the most from early detection and continuous management are, it turns out, no more likely to sign up than anyone else, and in some socio-economic groups considerably less likely.

Methodologically, the study’s strength lies in treating engagement as a sequence rather than a single yes-or-no outcome. By modelling four distinct stages with multinomial logistic regression and reporting marginal effects alongside odds ratios, the researchers could show where in the funnel different groups stall. Older, better-educated, more advantaged residents flow through the entire pipeline from awareness to screening. Disadvantaged residents and some ethnic minority groups drop out earliest, most often at the awareness stage itself. That granularity matters for policy, because the remedy for an awareness gap, targeted, culturally attuned communication through trusted community channels, is entirely different from the remedy for an enrolment gap, which may require simplifying registration, offering assistance with sign-up, or building trust through community partnerships.

The authors are careful about what their cross-sectional data can and cannot show. Because the study captures a single point in time during the programme’s early rollout, it cannot establish whether today’s engagement gaps will harden into tomorrow’s health disparities, and they call for longitudinal research to track whether early differences in enrolment and screening translate into differences in disease outcomes over the years ahead. The cohort also represents the population aged 40 and above, the group targeted by the programme, rather than the entire adult population. Still, the direction of the findings is consistent with a growing international literature showing that voluntary, incentive-based preventive care programmes tend to reproduce existing social gradients unless they are actively designed not to.

The broader lesson reaches well beyond Singapore. As health systems worldwide pivot toward prevention, from the United Kingdom’s emphasis on continuity of care to community-based primary care reforms across Asia, voluntary empanelment is becoming a default strategy. Singapore’s experience offers an early warning: universal eligibility and generous subsidies are necessary but not sufficient. Reaching the residents who rarely see a doctor, who live in smaller flats, who have less schooling and who may distrust or simply not encounter official health messaging, requires deliberate, targeted and culturally informed effort. Equity in preventive care, the study concludes, is not achieved at the moment costs are eliminated. It is achieved, or missed, in the accumulation of small frictions and quiet barriers that determine who walks through the door. The measurement of those barriers, stage by stage, is where reform begins.

Subject of Research: Socio-economic and ethnic disparities in preventive care engagement under Singapore's Healthier SG voluntary primary care empanelment programme

Article Title: Beyond financial access: socioeconomic and ethnic disparities in preventive care engagement under Singapore’s Healthier SG programme

Article References: Beyond financial access: socioeconomic and ethnic disparities in preventive care engagement under Singapore’s Healthier SG programme. (n.d.). https://doi.org/10.1186/s12939-026-03028-1

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03028-1

Keywords: Healthier SG, Singapore, preventive care, health equity, primary care reform, voluntary empanelment, Andersen Behavioural Model, socio-economic disparities, ethnic disparities, screening, health policy, public health