Suboptimal sleep, characterized by sleep disturbances and short (< 7 hours per 24-hour period) or long (> 9 hours per 24-hour period) duration, has been associated with several adverse health effects, such as obesity, type 2 diabetes, dementia, coronary heart disease, and premature mortality.^1–4 Yet, approximately one-third of the US adult population report insufficient sleep.⁵ Given the established associations between poor sleep and adverse health conditions, it is important to identify determinants of sleep quantity and quality, including the neighborhood environment.

Characteristics of the study sample overall and by race and ethnicity are presented in Table 1 weighted to the US population. Approximately 17% were Hispanic/Latino, 63% were White, 11% were Black, 6% were Asian, and 3% were other non-Hispanic adults (data not shown separately for this group). Approximately 60% of the study population had at least some college education, were married or cohabiting, and reported excellent or very good health.

In this nationally representative study of US adults, perceived neighborhood environments and sleep patterns both differed by race and ethnicity, and the associations between the perceived neighborhood environment and sleep also varied across racial and ethnic groups, with magnitudes of association inconsistently larger or smaller across groups as well as some differences in direction. Among Hispanic/Latino adults, unsafe walking conditions were associated with waking up unrested and having trouble staying asleep, while available walking paths were associated with a lower prevalence of short sleep duration and trouble falling asleep (Table 3). The significant finding for unsafe walking conditions and some sleep disturbances in Hispanic/Latino adults corresponds to evidence suggesting poorer sleep quality in young Puerto Rican adults living in neighborhoods with high stressors—which included measures of crime and safety—though the study sample was limited only to this ethnic subgroup living in two sites.¹²

Among White adults, unsafe walking conditions were associated with short sleep duration, waking up unrested, and trouble falling and staying asleep; walking paths and sidewalks were associated with a lower prevalence of trouble staying asleep, while walking paths were linked to increased sleep medication use; transit stops were positively associated with waking up unrested; and places to relax were negatively associated with trouble staying asleep and waking up unrested. The significant association between unsafe walking conditions and short sleep duration observed in this analysis somewhat contradicts exploratory results showing no association between safety and sleep duration among White adults found in Johnson et al.²⁴ However, the sample in that study was not nationally representative and sleep duration was measured objectively while safety was assessed through different items.

Among Black adults, having sidewalks on most streets and access to transit stops were associated with a lower prevalence of short sleep duration and sleep medication use, respectively. The lack of association between safety measures and short sleep duration for Black adults in the present study also contrasts with previous findings linking safety and sleep duration as well as violence and short sleep in African American samples^24,37; but again, the studies differ in their populations and measures used.

Lastly, among Asian adults, traffic was associated with short sleep duration and waking up unrested but with a decreased prevalence of frequent sleep medication use; available walking paths were negatively associated with trouble falling asleep and frequently using sleep medication; and places to relax were associated with a lower prevalence of long sleep duration while access to entertainment/services was negatively associated with waking up unrested. Results from the Johnson et al²⁴ analysis did not demonstrate an association between safety and sleep duration in Asian/Chinese participants, which may not necessarily support the current findings.

Our estimates of perceived neighborhood characteristics by race and ethnicity align with those reported by Soto et al.³⁸ The lower availability of certain amenities within walking distance observed for White adults in this study—such as sidewalks, stores, and transit stops—likely reflects the underrepresentation of the White population in urban areas³⁹ where such features may be integrated more closely within residential neighborhoods than in nonurban areas. Regarding sleep measures, our findings were consistent with previous literature that suggests shorter sleep duration among Black populations.^14,40 While studies have also indicated greater sleep disturbances among racially minoritized groups, especially by objective measures,^14,41 several analyses have found higher levels of sleep complaints, reported insomnia symptoms, and medication use among White populations,^{14,23,40,42,43} which is also supported by our estimates. These discrepancies may originate from cultural differences in reporting and stigmatization of sleep problems among historically minoritized groups, leading to under- or over-reporting of sleep complaints along racial and ethnic lines,²² as well as potential differences in attitudes toward and access to medication.⁴⁴

While we assessed variations in associations between the perceived neighborhood environment and sleep, it could be likely that prevalence differences in sleep health indicators may also be exacerbated by differences in other neighborhood characteristics among racial and ethnic groups, which could be further examined in future studies. Previous work suggests that sleep duration disparities by race and ethnicity do not exist or are attenuated in lower-income areas and housing where racially minoritized groups reside in similar living conditions to White populations,^45–47 and a recent study found that neighborhood social vulnerability partly mediates the effect of race on severity of insomnia.⁴⁸

However, in contrast to previous work suggesting that associations between neighborhood and similar facets of sleep health are stronger among some racially minoritized populations,^7,24 our results did not tend to confirm this. The strength of associations differed across relationships. These previous studies have focused on the social environment, but regarding perceived neighborhood walkability and other outcomes, Billings et al⁴⁹ also observed the strongest association between low walking environment and apnea-hypopnea index—a measure of obstructive sleep apnea—among Black study participants; however, their indicators of perceived walkability did not pertain to specific built environment features. Evidence relating to BMI based on the same data as the present study indicates that race and ethnicity does function as a moderator, with walkability associated with decreased BMI in White, Black, Hispanic, and Asian adults and only modest differences in magnitude among them.⁵⁰ Yet, unlike the present analysis, the study only assessed walkability using a sum score for the exposures and not the individual variables, so it is not possible to determine whether relationships may vary depending on the aspect of walkability.

Overall, however, the strongest association in our analysis was observed among Asian adults for traffic and sleep medication use. The finding of a lack of safety for walking due to traffic being linked to less frequent use of sleep medication may be a proxy for some other condition that cooccurs with traffic in these neighborhoods. This could perhaps be the case if the increased traffic is related to the presence of a valuable neighborhood asset that promotes well-being, such as a community or cultural center or other supportive community services. In addition, the direction of associations differed for some relationships, suggesting that some neighborhood features may be both favorable and unfavorable to sleep health depending on the racial and ethnic group. For example, walking path access was associated with lower sleep medication use among Asian adults but greater use among White adults. This may reflect potentially different uses of walking paths within the neighborhoods of Asian and White adults, such as if they are used more in one setting and not used as much in the other or have differing times or volumes of use that can impact sleep through noise. Additionally, the strength of associations did not necessarily correspond to the exposure or outcome burden. For example, although the strongest association for White adults was between crime and short sleep duration, they had the lowest burden of unsafe crime conditions in their neighborhoods; and similar situations were observed for other exposure-outcome pairs.

While the present study’s results may represent true differential effects by race and ethnicity, some findings may also be affected in part by other factors. One is the fact that the NHIS is based on self-report, and sleep and neighborhood characteristics are therefore assessed subjectively rather than objectively. In addition to potential racial and ethnic differences in reporting, there may also be differences in interpretation of survey items, such as what constitutes a place to relax, being within walking distance, or most streets having sidewalks. Although perceptions of the neighborhood are independently—and sometimes more strongly—associated with health compared to objective measures,⁵¹ these possible perceptual differences by race and ethnicity may limit the ability to make comparisons across groups.

Furthermore, the associations between the neighborhood environment and sleep may vary depending on other social stressors, which could also partly account for unexpected or null findings in associations by race. For example, in the community-based African American Sleep & Health (AASH) Study, Williams et al^52,53 found that the effects of greenspace redevelopment on sleep health depended on whether or not individuals within the communities were experiencing financial strain. Those living under financial strain in redeveloped communities had worse rather than better sleep health, which may be due to increased stress or fear of displacement because of rising costs associated with redevelopment. Thus, examining associations between the neighborhood environment and sleep in the context of other stressors may help to clarify the intricacies of these relationships.

Relatedly, race and ethnicity may interact with other identities and conditions in impacting the relationships between the perceived neighborhood environment and sleep health. While it was beyond the scope of this analysis, an intersectional approach can be useful in better understanding the nature of these associations by considering the role of other factors such as sex, age, or disability status. This approach has been adopted in studies of neighborhood social cohesion and sleep health using pooled data across multiple years.^7,54

We also acknowledge the impacts of the COVID-19 pandemic on sleep and neighborhood conditions in 2020, as well as on data collection and sample composition in the 2020 NHIS compared to other years, as this may have also influenced some results.²⁷ Studies have observed changes in sleep patterns during the pandemic, though sleep duration changes appear to be consistent across race and ethnicity.⁵⁵ A further limitation is the fact that some of the smaller racial subgroups (e.g., Asian) are underpowered while using only 1 year of data given that some outcomes have a low prevalence when stratified, which results in a reduced ability to detect significant associations of the same magnitude across racial groups. This also limits our ability to further adjust for additional covariates, especially those that may not have as much variation in some of the exposure variables, such as urban-rural residence. We likewise recognize the cross-sectional nature of the NHIS as an additional limitation.

Despite the limitations, these findings demonstrate the uniqueness of relationships between neighborhood factors and sleep, as results varied depending on the specific feature of neighborhood walkability, sleep outcome, as well as racial and ethnic group. Moreover, the study further documented the persistence of racial and ethnic inequities in sleep health based on the prevalence differences in sleep outcomes by group, and it similarly illustrated disparities in access and exposure to favorable and unfavorable features of the built and social environment. In particular, the results of this study underscore the need to consider exposure and outcome distributions in conjunction with effect measures to fully capture disparities,²⁶ as associations alone or distributions alone may not provide a complete picture. These findings highlight the complexities of conducting and interpreting health disparities research regarding the built environment with health outcomes.

The findings also have implications for clinical practice. Namely, they suggest the importance of considering neighborhood factors such as safety and land use types and resources in screening for sleep conditions, as well as racial and ethnic groups that may be at greater risk. This can also extend to designing sleep interventions, which may likewise benefit from considering neighborhood circumstances and their relation to race and ethnicity.