Healthy People 2020 and the World Health Organization maintain that social determinants of health, including social, economic, and physical environments, shape people’s opportunities to achieve optimal health (1,2). Furthermore, the 5-tier health impact pyramid suggests that interventions related to the improvement of socioeconomic factors may have the greatest potential impact to change the social context to enable an individual’s default decisions to be healthy (3). Housing insecurity and food insecurity have been identified as 2 important social determinants of health (4,5). The definitions of housing insecurity vary and include having high housing costs relative to household income, living in environments of poor quality and unstable neighborhoods, living in overcrowded housing, or being homeless (6). In 2010, more than a quarter of US households (20.2 million households) paid more than 50% of their incomes for housing (7). Food insecurity can be defined as uncertainty about one’s ability to access safe and nutritious foods because of restricted financial resources (8). Nearly 15% of US households (approximately 48.8 million Americans) were food insecure at some time during 2010; 10.5% of US households were food insecure during 2000 (9). Research suggests that food insecurity is associated with the consumption of more high-energy–dense foods that may lead to weight gain, diabetes, and poor physical health (10,11). Studies also indicate that food insecurity is associated with psychological distress, anxiety, and depression among low-income women and children (12,13). Psychological distress may, in turn, adversely affect people’s physical health, mental health, and sleep (14).

Emerging data suggest that insufficient sleep is associated with obesity, diabetes, cardiovascular disease, stroke, high blood pressure, arthritis, asthma, anxiety, and depression (15–20); the rate of poor sleep is higher among women than men and is correlated with low household income, poverty, and unemployment (21–23). In 2008, 37.9% of US adults reported 14 days or more of insufficient sleep (24). Few studies have addressed the relationship between housing insecurity and insufficient sleep (21,22). The Behavioral Risk Factor Surveillance System (BRFSS), the largest telephone surveillance system in the world, recently began collecting data on insufficient sleep, defined as the number of days that the respondent felt that he or she did not get enough rest or sleep in the previous 30 days. The survey also began collection data on housing insecurity and food insecurity among US adult respondents (25). Frequent mental distress (≥14 days of mental distress during the previous 30 days) is a reliable indicator of psychological distress (26) and is also highly associated with insufficient sleep (18–22). The objective of this study was to assess the relationships between housing insecurity, food insecurity, frequent mental distress, and insufficient sleep.

The BRFSS, a state-based, random-digit–dialed telephone survey of noninstitutionalized, civilian US adults aged 18 or older, consists of a core set of questions on public health issues that are asked in all states and optional modules that states may elect to include. In 2009, an optional module on social context, which included questions on housing insecurity and food insecurity, was administered in 12 states (Alabama, Arkansas, California, Hawaii, Illinois, Kansas, Louisiana, Nebraska, New Mexico, Oklahoma, South Carolina, and Wisconsin). Data from 68,111 (90.7%) of 75,103 respondents to this module were analyzed after excluding those who did not respond to questions about sleep (n = 1,320) and housing insecurity or food insecurity (n = 5,672). The response rate in these 12 states ranged from 42.8% to 66.9% (median, 58.2%) (27).

Of the 68,111 survey respondents in this study, 50.8% were aged 45 or older, 62.2% were non-Hispanic white, 61.9% had more than 12 years education, 28.5% reported housing insecurity and 19.3% reported food insecurity during the previous 12 months, and 10.8% reported frequent mental distress and 26.4% reported frequent insufficient sleep during the previous 30 days (Table 1). In addition, a significantly higher percentage of non-Hispanic blacks, Hispanics, and other non-Hispanics than non-Hispanic whites reported food insecurity (P < .001) and housing insecurity (P < .001) (Figure). A significantly higher percentage of respondents who had 12 years or less of education reported food insecurity (P < .001) and housing insecurity (P < .001) than those with more than 12 years of education.

Bivariate analyses (Table 2) showed that the prevalences for both frequent insufficient sleep and frequent mental distress were significantly higher among women than among men and among people younger than 65 than among those aged 65 or older (P < .001). A significantly greater percentage of non-Hispanic blacks reported frequent insufficient sleep (P = .02) and frequent mental distress (P < .001) than non-Hispanic whites. The percentage of Hispanics reporting frequent insufficient sleep was similar to that reported by non-Hispanic whites (P = .17), but a significantly greater percentage of Hispanics reported frequent mental distress (P < .001). The prevalence of frequent insufficient sleep did not differ significantly by years of education, but the prevalence of frequent mental distress was negatively associated with years of education (P < .001). The unadjusted prevalence of frequent insufficient sleep was significantly higher among respondents who reported either housing insecurity or food insecurity than among those who did not (P < .001). Respondents who reported either housing insecurity or food insecurity were about 3 times as likely to report frequent mental distress than were those who did not, and those who reported frequent mental distress were more than twice as likely to report frequent insufficient sleep as those who did not (59.7% vs 22.3%).

Unadjusted results showed that respondents who reported housing insecurity or food insecurity were each more than 70% more likely to report frequent insufficient sleep than those who did not report the insecurity (Table 3, Model 1). Housing insecurity and food insecurity remained significantly associated with frequent insufficient sleep after adjustment for sociodemographic covariates for housing insecurity (PR = 1.67; 95% CI, 1.55–1.79) and for food insecurity (PR = 1.75; 95% CI, 1.62–1.88) (Table 3, Model 2). Although further adjustment for respondents’ frequent mental distress status reduced the magnitude of the relationship by 27% (Table3, Model 3), both housing insecurity and food insecurity remained significantly associated with frequent insufficient sleep, suggesting that frequent mental distress was only a partial mediator of the relationships of these social context variables with insufficient sleep.

Our results demonstrated that both housing insecurity and food insecurity were associated with frequent insufficient sleep among US adults in 12 states. This positive relationship was modestly attenuated but not completely explained by frequent mental distress. Although the mechanisms underlying the association are not clear, one potential explanation is that stress caused by housing insecurity or food insecurity could lead to prolonged psychological distress or depressive symptoms (13,33–35). Previous studies have also shown that housing insecurity, particularly crowding and multiple relocations, were associated with psychological distress, poor health, and developmental risk among children (33,34), and food insecurity was highly associated with mental distress among women (13,35) and school-aged children (12) even after adjustments for sociodemographic characteristics. In addition, poor housing conditions may also be related to poor sleep through concerns about personal safety, exposure to higher noise levels, and inadequate heating or cooling (36,37).

Other studies reported that low levels of food security were significantly associated with short sleep duration and suggested that the significant relationship might be due to racial/ethnic minority status and low socioeconomic position (21,22), findings that are also consistent with ours. Furthermore, hunger may acutely affect the ability to sleep well. Educational and racial/ethnic disparities in these social contexts may also play an important role in perceived insufficient sleep and frequent mental distress.

Chronic stress or worry related to poor housing and poor nutrition may interact with endocrine systems such as the sympathoadrenal medullary and the hypothalamo–pituitary–adrenocortical systems and are associated with elevated levels of cortisol, adrenocorticotropic hormone (ACTH), and corticotropin-releasing hormone (CRH). Research indicated that an elevated level of cortisol was highly associated with frequent microarousals (38). In addition, elevated ACTH levels could increase the frequency of morning awakening (39). Furthermore, the stages of slow-wave sleep and rapid eye movements in the sleep cycle could be reduced by excessive release of CRH (40). These physiological changes may cause sleep disturbances with increased cognitive arousal or cortical arousal or both (41).

Our results were consistent with previous studies in which either “housing instability” (definitions of which have included frequent moves, difficulty paying rent, spending more than half of household income on housing, being evicted, or living in overcrowded conditions) or “food insufficiency” (respondents reported that their family sometimes or often did not get enough food to eat) was used to assess the relationship with mental health (12,13,33). As in these earlier cross-sectional studies, which examined the relationship of each social context with frequent insufficient sleep, we cannot assess causation.

Although this is the largest study to date to address the relationship between housing insecurity and food insecurity and insufficient sleep among US adults, it has several limitations. First, because it was a cross-sectional study, we cannot establish causation. Second, because our results are based on the analysis of information provided by BRFSS in only 12 states and BRFSS does not include people residing in institutions such as nursing homes and prisons, our results are not generalizable to the entire US adult population. Third, the dichotomous measures of housing insecurity and food insecurity used in this study were each derived from responses to a single survey question on respondents’ confidence in their financial capacity to acquire adequate housing and food and therefore did not reflect all respects of housing insecurity and food insecurity. For example, the measure of housing insecurity did not address whether respondents resided in housing of poor quality, in overcrowded housing, or in unstable neighborhoods (6), and the measure of food insecurity did not reflect the extent to which respondents experienced hunger (9). However, the influence of the different definitions of housing insecurity and food insecurity on our results may be limited given the findings from previous studies that are consistent with ours (9,33,34). In addition, frequent insufficient sleep is a subjective measure of sleep health and was not corroborated by polysomnography. However, frequent insufficient sleep may partly reflect some sleep problems such as obstructive sleep apnea and is highly related to sleep duration (20). Efforts are needed to address the mediating role of other social determinants such as marriage, income, employment, accessibility of health services, and geographic variation in social context in the relationship of housing insecurity and food insecurity with poor sleep (22,23,42). Finally, our results could have been influenced by residual variations because other cofounders were not taken into account.

Fundamental societal transformations may be required to achieve social and economic changes that affect health (3). Potential improvements in housing and access to healthy food include community-level projects to provide environmentally healthful and safe housing for low-income families, food subsidy programs, and educational technology programs to enable low-income individuals to re-enter the workforce in new careers.

In summary, our findings, which are based on general-population data, add to a growing body of evidence that shows housing insecurity and food insecurity are associated with psychological distress and insufficient sleep. Our findings are consistent with previous findings showing that stress related to housing insecurity and food insecurity may sustain and adversely affect sleep and precipitate long-term adverse consequences to physical and mental health. Furthermore, our findings support the idea that sleep health and mental health are embedded in the social context. Research is needed to assess whether interventions that reduce housing insecurity and food insecurity will also improve those outcomes.