Abstract
Background:
Sleep deprivation is a critical risk factor for physical and mental health problems. The current study examined whether sleep problems were related to housing status and physical and mental health among people experiencing homelessness (PEH).
Methods:
Data were collected via an anonymous survey completed by adults accessing services at a day shelter in Oklahoma City, OK. Demographic characteristics, sleep problems, days of poor physical and mental health, and self-rated health were assessed and compared across housing subgroups. Logistic regression analyses were performed to examine potential associations between sleep and health.
Results:
Participants (N=404) were predominantly male (71.0%) and racially minoritized (57.4%) with an average age of 47.0 (SD=11.6) years. Days of unintentional sleep over the past 30 days differed significantly by housing status. Logistic regression analyses indicated that short sleep duration of ≤6 hours was associated with poorer self-rated health and more days of poor physical health over the past 30 days compared to those who reported an optimal sleep duration of 7–9 hours. Additionally, reporting ≥1 day of unintentional daytime sleep was associated with more days of poor mental and physical health over the past 30-days. Reporting more days of inadequate sleep was associated with poorer self-rated health, and more days of poor physical and mental health over the past 30 days.
Conclusion:
Sleep problems were associated with poorer physical and mental health among PEH. Safe environments where adults experiencing homelessness can sleep comfortably and without interruption may be needed to support good physical and mental health.
Keywords: homeless, sleep, health, housing
1. Introduction
In 2020, it was estimated that more than 580,000 people in the U.S. experienced homelessness each night (HUD, 2020). Nearly four in ten people experiencing homelessness (PEH) slept in locations that were not suitable for human habitation, such as outside on the street (i.e., unsheltered; HUD, 2020). The remaining six in ten slept in sheltered locations such as overnight emergency shelters (“sheltered”) or lived with friends, family, or other temporary locations (i.e., unstably housed; HUD, 2020). According to the Centers for Disease Control and Prevention, one-third of U.S. adults reported getting less than the recommended 7 to 8 hours of sleep each day (CDC, 2016), and PEH are even more likely to experience sleep inadequacy due to harsh and often unsafe sleeping environments (e.g., due to poor weather, noise, or crime; Corning, 2010; Gonzalez & Tyminski, 2020; Ra et al., 2021).
Plausibly, several dimensions of sleep may be adversely affected in PEH; particularly sleep duration, inadequate sleep, and unintentional sleep (Chang et al., 2015). The U-shaped association between sleep duration, defined as the quantity of time a person sleeps over the course of a 24-hour day, and health is well-documented (Chaput et al., 2020). For example, both shorter (≤6 hours) and longer sleep duration (>9 hours) are associated with increased risk of all-cause mortality, cardiovascular disease, and diabetes (Kakizaki et al., 2013; Itani et al., 2017; Chaput et al., 2020). Inadequate sleep, a term used to refer to poor sleep quality or shorter sleep duration at least a few times a month, also has been linked to poorer health outcomes such as poor/fair self-rated health, and frequent physical and mental health distress (e.g. depressive symptoms, anxiety, pain; Strine & Chapman, 2005). Unintentionally falling asleep during daytime hours is also associated with stress and poor mental/physical health among PEH (Reitzel et al., 2017). Gonzalez & Tyminski (2020) have suggested that sleep deprivation among PEH may be an important barrier to overcoming homelessness.
Housing status may be linked with sleep problems among PEH. One study showed that compared to sheltered homeless adults, unsheltered homeless adults were twice as likely to report sleep inadequacy (30.7% vs. 15.3%, respectively; Ra et al., 2021). Unsheltered sleep is challenged by various factors including poor weather conditions, uncomfortable sleeping locations, and fear of victimization (e.g. assault); nonetheless, sheltered environments (e.g. emergency housing) are also challenged by other factors such as noise levels, issues with privacy and shelter overcrowding, which may lead many PEH to sleep outside (Gonzalez & Tyminski, 2020). However, research on the relationships between housing status (i.e., unsheltered, sheltered, unstably housed), sleep problems, and health is limited. Hence, the purpose of this study was: 1) to examine the potential link between housing status and sleep problems and 2) to evaluate the association between sleep problems and health in PEH.
Specifically, it was hypothesized that unsheltered housing status would be associated with worse sleep problems (i.e. shorter sleep duration and more days of inadequate sleep and unintentional sleep) than sheltered or unstable housing status, and that all sleep problems would be independently and significantly associated with adverse health outcomes including poorer self-rated health and more days of experiencing poor physical and mental health. Gaining a better understanding of the relationship between sleep and health may inform policy and program initiatives focused on increasing the availability of housing and improving the sleep environment among PEH.
2. Method
The present study was a secondary data analysis of anonymous survey data collected between May 2019 and March 2020 from 555 adults accessing day shelter services at the Homeless Alliance Day Shelter in Oklahoma City, OK. The following cases were excluded from current study analyses: 1) housed participants (n=27), 2) participants with missing data (n=34), 3) average sleep duration shorter than 2 hours and/or longer than 13 hours (n=92), and 4) responses greater than 30-days for questions asked about the past 30-days (unintentional sleep, inadequate sleep, physical health, and mental health; n=97).
Participants who reported a sleep duration of <2 or >13 hours were excluded from study analyses as outliers. Outlier thresholds were calculated using the interquartile range method (i.e. upper threshold = Q3 + [1.5*IQR] and lower threshold = Q1 − [(1.5*IQR]; Sullivan et al., 2021). The lower and upper fences of sleep duration were 1.5 and 12.5 hours, respectively. Study participants reported their sleep duration in whole numbers; thus, the outlier thresholds were rounded to 2 (lower) and 13 (upper). Also note that the majority of the participants excluded as outliers (n=61) reported that they slept 0 hours per night on average. Likewise, 6 additional participants reported that they slept an average of >24 hours per night. In total, 151 individuals were excluded from the analyses, leaving an analytic sample of 404 participants.
The study protocol, including a waiver of consent for an anonymous study, was approved by the Institutional Review Board at the University of Oklahoma Health Sciences Center (protocol #10124). Participants were provided with a brief letter containing study information (including contact information), and participants were asked for verbal consent before initiating the survey. Additional details of the parent study procedures are described elsewhere (Alexander et al., 2021; Ehlke et al., 2022; Boozary et al., 2022; Sifat et al., 2023).
Measures
Sociodemographic Characteristics.
Participants self-reported their demographic characteristics including age (years), biological sex (male vs. female), education (years), ethnicity (Hispanic vs. non-Hispanic), and race (White, African American, or other minoritized race). Other minoritized races included Asian, Native Hawaiian/Pacific Islander, American Indian/Alaska Native, Multi-racial, and other race. Note that years of age was dichotomized at the median (≤47 vs. >47 years) because the variable was not normally distributed.
Housing Status.
Participants responded to the question: “Where did you sleep last night?” Participant responses were categorized as follows: (1) sheltered location (homeless shelter, jail, hospital, and drug or alcohol treatment center), (2) unstably housed (a friend’s or family member’s house, hotel/motel, or other temporary location), and (3) unsheltered location (abandoned buildings, outside or on the street, car, or tents). The shelter categorizations were based on those used in the U.S. Department of Housing and Urban Development annual point-in-time counts to estimate the number of sheltered and unsheltered homeless individuals on a single night (HUD, 2020).
Sleep problems.
The Behavioral Risk Factor Surveillance System (BRFSS) Inadequate Sleep Questionnaire (CDC, 2017) is a self-report measure that assessed the following aspects of sleep: (1) sleep duration, the average number of hours of sleep in a 24-hour period; (2) inadequate sleep, the number of days of insufficient rest or sleep during the past 30-days; and (3) unintentional sleep, the number of days of unintentionally falling asleep during the day during the past 30-days. This study adopted sleep inadequacy categorizations that were used in previous research with PEH for sleep duration (≤6 hours [short], 7–9 hours [optimal], ≥10 hours [long]) and unintentional sleep (0 days versus ≥1 day) (Chang et al., 2015). Inadequate sleep days was divided in to three categories (0 days, 1–13 days, 14–30 days).
Current physical health and mental health problems.
Participants were asked to self-rate their overall health on a Likert scale from 0–4 (0=Poor, 1=Fair, 2=Good, 3=Very good, 4=Excellent). Participants also reported the number of days during the past 30-days on which they experienced: 1) poor mental health (stress, depression, or problems with emotions) and 2) poor physical health (physical illness or injury; BRFSS, 2021). Health variables were dichotomized based on the median value due to non-normal distribution or to meet the assumption of linearity of the logit in the logistic regression models: self-rated health (poor/fair vs. good/very good/excellent self-rated health); days of poor mental health (<3 days vs. ≥3 days); and days of poor physical health (<2 days vs. ≥2 days).
Statistical Analysis
Descriptive statistics were generated to describe participant characteristics (race, age, ethnicity, and education), sleep problems, and health status. Because the variables were not normally distributed, medians and interquartile ranges were reported for demographic characteristics (age and education) and days of poor physical health and mental health. Chi-square/Fisher’s exact analyses were conducted to compare participant characteristics across the three housing status subgroups. Differences in race and sex were found by housing status and thus, additional logistic, nominal, or ordinal regression analyses were conducted to control for potential effects of sex and race in the comparison of health and sleep problems across the three groups. For the main analyses, logistic regression analyses were performed to evaluate the potential associations between each individual sleep variable (sleep duration, inadequate sleep, and unintentional sleep) with each of the three health variables, adjusting for demographic characteristics including sex, race, and age. Although uncorrected values are presented in Tables 2–4, Bonferroni alpha corrections to adjust for multiple comparisons are provided in the Table footnotes, and differences in the statistical significance of findings between uncorrected and corrected p-values are noted. All analyses were performed using the “stats” package as part of R studio software, version 4.1.2 (2021–11-01; R Core Team, 2021) and SAS (version 9.4).
Table 2.
Associations between sleep problems and self-rated health.
| Dependent variable: Self-rated health (0=Fair/Poor vs. 1=Good/Very Good/Excellent) |
|||
|---|---|---|---|
|
| |||
| Independent variables | Odds ratio (OR) | 95% CI | p-value |
|
| |||
| Sleep duration | |||
| Short, ≤6 hours (ref. optimal, 7–9 hours) | 0.51 | [0.31, 0.80] | 0.004 * |
| Long, ≥10 hours (ref. optimal 7–9 hours) | 0.56 | [0.23, 1.41] | 0.206 |
| Age, >47 years (ref. ≤47 years) | 1.00 | [0.99, 1.02] | 0.718 |
| Sex, Male (ref. female) | 1.64 | [1.04, 2.59] | 0.032 |
| Race, Black (ref. White) | 1.18 | [0.70, 2.00] | 0.542 |
| Race, Other (ref. White) | 1.03 | [0.63, 1.68] | 0.909 |
|
| |||
| Inadequate sleep | |||
| 1–13 days (ref. 0 days) | 0.63 | [0.34, 1.13] | 0.130 |
| 14–30 days (ref. 0 days) | 0.42 | [0.21, 0.80] | 0.010 * |
| Age, >47 years (ref ≤47 years) | 1.00 | [0.99, 1.02] | 0.654 |
| Sex, Male (ref. female) | 1.55 | [0.98, 2.44] | 0.060 |
| Race, Black (ref. White) | 1.20 | [0.72, 2.04] | 0.490 |
| Race, Other (ref. White) | 1.04 | [0.64, 1.70] | 0.866 |
|
| |||
| Unintentional Sleep | |||
| ≥1 day (ref. 0 days) | 0.73 | [0.46, 1.15] | 0.182 |
| Age, >47 years (ref ≤47 years) | 1.00 | [0.99, 1.02] | 0.661 |
| Sex, Male (ref. female) | 1.65 | [1.05, 2.59] | 0.030 |
| Race, Black (ref. White) | 1.23 | [0.73, 2.08] | 0.445 |
| Race, Other (ref. White) | 1.15 | [0.71, 1.87] | 0.583 |
Note: Bolded p-values indicate statistical significance (p<0.05).
Relationship is statistically significant after Bonferroni correction (alpha set at p=0.016).
Table 4.
Associations between sleep problems and days of poor physical health.
| Dependent variable: Days of poor physical health (<2 days = 0 vs. ≥2 days = 1) |
|||
|---|---|---|---|
|
| |||
| Independent variables | Odds ratio (OR) | 95% CI | p-value |
|
| |||
| Sleep duration | |||
| Short, ≤6 hours (ref. optimal, 7–9 hours) | 1.71 | [1.11, 2.65] | 0.0157 * |
| Long, ≥10 hours (ref. optimal 7–9 hours) | 1.51 | [0.63, 3.61] | 0.348 |
| Age, >47 years (ref. ≤47 years) | 1.02 | [1.01, 1.04] | 0.009 * |
| Sex, Male (ref. female) | 0.64 | [0.41, 1.00] | 0.053 |
| Race, Black (ref. White) | 0.83 | [0.50, 1.37] | 0.467 |
| Race, Other (ref. White) | 1.39 | [0.87, 2.25] | 0.170 |
|
| |||
| Inadequate sleep | |||
| 1–13 days (ref. 0 days) | 3.46 | [1.91, 6.50] | <0.0001 * |
| 14–30 days (ref. 0 days) | 5.40 | [2.76, 10.98] | <0.0001 * |
| Age, >47 years (ref ≤47 years) | 1.03 | [1.01, 1.05] | 0.007 * |
| Sex, Male (ref. female) | 0.73 | [0.46, 1.15] | 0.177 |
| Race, Black (ref. White) | 0.78 | [0.47, 1.30] | 0.345 |
| Race, Other (ref. White) | 1.49 | [0.92, 2.44] | 0.106 |
|
| |||
| Unintentional Sleep | |||
| ≥1 day (ref. 0 days) | 3.17 | [2.02, 5.05] | <0.0001 * |
| Age, >47 years (ref ≤47 years) | 1.03 | [1.01, 1.04] | 0.007 * |
| Sex, Male (ref. female) | 0.64 | [0.40, 1.02] | 0.061 |
| Race, Black (ref. White) | 0.71 | [0.42, 1.18] | 0.187 |
| Race, Other (ref. White) | 1.17 | [0.72, 1.89] | 0.533 |
Note: Bolded p-values indicate statistical significance (p<0.05).
Relationship is statistically significant after Bonferroni correction (alpha set at p=0.016).
3. Results
Participant Characteristics
Participants (N=404) were divided into three housing subgroups: unsheltered (n=250 [outside/on the street, n=183; tent, n=35; car, n=28; abandoned building, n=4;], unstably housed (n=59 [friend/family house, n=48; hotel/motel, n=3; other temporary location, n=8]), and sheltered (n=95; [homeless shelter, n=91; hospital, n=3; drug/alcohol treatment center, n=1]). Participants were predominantly male (71.0%) and racially minoritized (57.4% [African American, 25.7%; Asian, 0.7%; Native Hawaiian/Pacific Islander, 0.5%; American Indian/Alaska Native, 16.8%; Multi-racial, 8.7%; other, 5.0%]), and 10.9% reported Hispanic ethnicity. The median age of participants was 47.0 years (IQR = 37.8, 55.0) and 32.2% reported having less than a high school education. Overall, participants reported a median of 2 days of poor physical health (IQR = 0.0, 8.0) and 3 days of poor mental health (IQR = 0,0, 10.0) over the past 30-days; and 35.6% self-reported fair/poor health. The majority of participants reported sleeping less than 7-hours per night (59.2%), having 1–13 days of inadequate sleep in the past 30-days (56.4%), and having at least 1 day with unintentional sleep during the daytime (68.1%) in the past 30-days. See Table 1.
Table 1.
Participant characteristics overall and by housing status.
| Overall sample N=404 |
Unsheltered n=250 |
Unstably housed n=59 |
Sheltered n=95 |
p-value | |
|---|---|---|---|---|---|
|
| |||||
| Age, above median (>47 years), n (%) | 198 (49.0%) | 131 (52.4%) | 27 (45.8%) | 40 (42.1%) | 0.202 |
| Sex, Male, n (%) | 287 (71.0%) | 184 (73.6%)a | 33 (55.9%)a,b | 70 (73.7%)b | 0.022 |
| Race | 0.017 | ||||
| White, n (%) | 172 (42.6%) | 119 (47.6%)a | 15 (25.4%)a,b | 38 (40.0%)b | |
| African American, n (%) | 104 (25.7%) | 54 (21.6%)a | 23 (39.0%)a b | 27 (28.4%)b | |
| Other, n (%) | 128 (31.7%) | 77 (30.8%) | 21 (35.6%) | 30 (31.6%) | |
| Ethnicity, Hispanic, n (%) | 44 (10.9%) | 28 (11.2%) | 6 (10.2%) | 10 (10.5%) | 0.966 |
| Education, <High school, n (%) | 130 (32.2%) | 78 (31.2%) | 23 (39.0%) | 29 (30.5%) | 0.477 |
| Poor mental health | 0.490 | ||||
| Above median (>3 days), n (%) | 179 (44.3%) | 105 (42%) | 28 (47.5%) | 46 (48.4%) | |
| Poor physical health | 0.461 | ||||
| Above median (>2 days), n (%) | 195 (48.3%) | 115 (46%) | 32 (54.2%) | 48 (50.5%) | |
| Self-rated health | 0.213 | ||||
| Fair or poor, n (%) | 144 (35.6%) | 84 (33.6%) | 19 (32.2%) | 41 (43.2%) | |
| Sleep duration | |||||
| Short, ≤6 hours, n (%) | 239 (59.2%) | 148 (59.2%) | 27 (45.8%) | 64 (67.4%) | 0.104 |
| Optimal, 7–9 hours, n (%) | 139 (34.4%) | 87 (34.8%) | 27 (45.8%) | 25 (26.3%) | |
| Long, ≥10 hours, n (%) | 26 (6.4%) | 15 (6.0%) | 5 (8.5%) | 6 (6.3%) | |
| Inadequate sleep days | 0.131 | ||||
| 0 days, n (%) | 74 (18.3%) | 55 (22.0%) | 7 (11.9%) | 12 (12.6%) | |
| 1–13 days, n (%) | 228 (56.4%) | 136 (54.4%) | 38 (64.4%) | 54 (56.8%) | |
| 14–30 days, n (%) | 102 (25.2%) | 59 (23.6%) | 14 (23.7%) | 29 (30.5%) | |
| Unintentional sleep days | 0.025 * | ||||
| ≥1 day, n (%) | 275 (68.1%) | 162 (64.8%)a | 49 (83.1%)a,b | 64 (67.4%)b | |
Note: All analyses presented in the table are unadjusted, and bolded p-values indicate statistical significance (p<0.05). Post-hoc analyses were performed when significant differences by housing status were identified. Differences on health and sleep variables by housing status were additionally evaluated in models adjusting for race and sex, and there were no significant differences on any variable with the exception of unintentional sleep days.
Unsheltered differed significantly from unstably housed.
Sheltered differed significantly from unstably housed.
Relationship remained statistically significant (p=0.018) after adjustment for sex and race.
Comparisons of Participant Characteristics by Housing Status
Participants in the three housing categories were relatively similar in age, ethnicity, and education level; however, the three groups differed on sex and race. Specifically, compared to females, males reported a higher proportion of sheltered or unsheltered housing status (>70%) than unstable housing (55.9%; or alternatively, women reported a higher proportion of unstable housing than sheltered or unsheltered housing status). When examining the distribution of race within housing status group, the highest proportions of Whites reported unsheltered (47.6%) and sheltered (40.0%) housing status, while the highest proportion of African Americans reported being unstably housed (39.0%). There were no significant differences in self-rated health, days of poor mental or physical health by housing status in the unadjusted and adjusted analyses. However, the three groups differed significantly on days of unintentional sleep in the unadjusted analyses, and the association remained significant after adjustment for sex and race in a logistic regression model (p=0.018). No differences were found for sleep duration or days of inadequate sleep by housing status in the unadjusted or adjusted models. See Table 1.
Associations of Sleep Problems with Self-rated, Mental and Physical Health
Self-rated health.
Analyses indicated that the odds of reporting good to excellent self-rated health was lower among those who reported short sleep duration compared to those who reported optimal sleep duration (OR=0.51, 95% CI: 0.31, 0.80). Likewise, the odds of reporting good to excellent self-rated health was lower among people who reported 14–30 days of inadequate sleep compared to people who reported no days of inadequate sleep in the past 30-day period (OR=0.42; 95% CI: 0.21, 0.80). See Table 2.
Mental health.
Sleep duration was not a significant predictor of the number of poor mental health days. However, having one or more days of inadequate sleep and/or unintentional sleep (compared to no days) in the past 30-days increased the odds of having ≥3 days of poor mental health. Specifically, participants were more likely to report ≥3 days of poor mental health if they reported 1–13 days or 14–30 days of inadequate sleep (OR=2.81; 95% CI: 1.55, 5.33 and OR=5.16, 95% CI: 2.62, 10.53, respectively). Additionally, reporting ≥1 day of unintentionally falling asleep during the daytime was associated with reporting ≥3 days of poor mental health (OR=2.93; 95% CI: 1.86, 4.71). See Table 3.
Table 3.
Associations between sleep problems and poor mental health.
| Dependent variable: Days of poor mental health (<3 days = 0 vs. ≥3 days = 1) |
|||
|---|---|---|---|
|
| |||
| Independent variables | Odds ratio (OR) | 95% CI | p-value |
|
| |||
| Sleep duration | |||
| Short, ≤6 hours (ref. optimal, 7–9 hours) | 1.50 | [0.97, 2.33] | 0.068 |
| Long, ≥10 hours (ref. optimal 7–9 hours) | 0.75 | [0.28, 1.83] | 0.535 |
| Age, >47 years (ref. ≤47 years) | 1.00 | [0.99, 1.02] | 0.676 |
| Sex, Male (ref. female) | 0.56 | [0.36, 0.88] | 0.011 * |
| Race, Black (ref. White) | 0.64 | [0.38, 1.06] | 0.086 |
| Race, Other (ref. White) | 1.31 | [0.82, 2.11] | 0.259 |
|
| |||
| Inadequate sleep | |||
| 1–13 days (ref. 0 days) | 2.81 | [1.55, 5.33] | 0.001 * |
| 14–30 days (ref. 0 days) | 5.16 | [2.62, 10.53] | <0.0001 * |
| Age, >47 years (ref ≤47 years) | 1.01 | [0.99, 1.02] | 0.566 |
| Sex, Male (ref. female) | 0.63 | [0.39, 0.99] | 0.047 |
| Race, Black (ref. White) | 0.61 | [0.36, 1.02] | 0.061 |
| Race, Other (ref. White) | 1.47 | [0.91, 2.39] | 0.120 |
|
| |||
| Unintentional sleep | |||
| ≥1 day (ref. 0 days) | 2.93 | [1.86, 4.71] | <0.0001 * |
| Age, >47 years (ref ≤47 years) | 1.01 | [0.99, 1.02] | 0.582 |
| Sex, Male (ref. female) | 0.56 | [0.35, 0.88] | 0.013 * |
| Race, Black (ref. White) | 0.55 | [0.32, 0.93] | 0.027 |
| Race, Other (ref. White) | 1.16 | [0.72, 1.88] | 0.545 |
Note: Bolded p-values indicate statistical significance (p<0.05).
Relationship is statistically significant after Bonferroni correction (alpha set at p=0.016).
Physical health.
Analyses indicated that people who reported a short sleep duration were more likely to have 2 or more days of poor physical health compared to those who reported optimal sleep duration (OR=1.71; 95% CI: 1.11, 2.65). Reporting at least 1 day of inadequate sleep or unintentional sleep was associated with a greater likelihood of reporting 2 or more days of fair/poor physical health. Compared to people who reported no days of inadequate sleep, those who reported 1–13 days and 14–30-days of inadequate sleep were more likely to report 2 or more days of poor physical health (OR=3.46; 95% CI: 1.91, 6.50 and OR=5.40; 95% CI: 2.76, 10.98, respectively). Likewise, people who reported 1 or more days of unintentional sleep were more likely to report 2 or more days of poor physical health (OR=3.17; 95% CI: 2.02, 5.05). See Table 4.
4. Discussion
The primary aims of the study were to evaluate the association of housing status with sleep and health outcomes and to examine whether sleep problems were associated with poorer health among PEH. Findings indicated that reports of unintentional daytime sleep varied by housing status, with unstably housed individuals more likely to report unintentional sleep relative to sheltered and unsheltered individuals. Short sleep duration (≤6 hours) was associated with poorer self-rated health and more days of poor physical health than among those reporting optimal sleep duration (7–9 hours). Reporting more days of inadequate sleep was associated with poorer self-rated health, and more days of poor mental and physical health. Finally, reporting any unintentional daytime sleep over the past 30-days was associated with more days of poor physical and mental health. Overall, findings suggest that poor sleep is associated with poorer physical and mental health among PEH, and that efforts are needed to improve sleep in this vulnerable population.
Overall, sleep characteristics were similarly poor across the housing status categories. However, unstably housed individuals were more likely to report unintentional daytime sleep than sheltered and unsheltered individuals. Although the reasons for this association are not clear, perhaps unstably housed individuals were more likely to have access to environments that supported sleep during the day. In support of this hypothesis, a greater proportion of unstably housed individuals reported optimal sleep duration and fewer reported low sleep duration compared to sheltered and unsheltered individuals (this difference did not reach statistical significance). More research is needed to determine how housing status may be related to various dimensions of sleep.
Optimal sleep duration (7–9 hours) was associated with good to excellent self-rated health and fewer days of poor physical health relative to those who reported shorter sleep duration. Previous research suggests that both under- and over-sleeping are associated with all-cause mortality (Cappuccio et al., 2010). Additionally, reporting of 1 or more days of inadequate sleep was associated with having more days of poor mental and physical health. Results reflected a dose-response association of inadequate sleep with increased days of poor physical and mental health, such that the odds of having three or more days of poor mental health increased with the total days of inadequate sleep. These findings are consistent with the findings of Geiger et al. (2012), which showed that the odds of poor self-rated health increased with more days of inadequate sleep in a nationally representative sample. Finally, unintentional daytime sleep was associated with more days of poor mental and physical health. Previous research has indicated that sleep problems can be both a cause of and a key feature of mental health disorders (Anderson & Bradley, 2013). The current study results extend these findings to PEH and provide further evidence that inadequate sleep may adversely influence both physical and mental health.
Importantly, findings suggested that sleep was generally poor among PEH. Results showed that 59% of PEH reported less than seven hours of sleep per night, with only 18.2% reported having no days of insufficient sleep/rest during the past 30 days. This was nearly twice the rate of inadequate sleep duration reported among domiciled adults (35% and 30.7%, respectively) on a survey of more than 444,000 Americans (Liu et al., 2014; McKnight et al, 2009). Study findings are consistent with other research showing that PEH experience more sleep problems (Chang et al., 2015; Léger et al., 2017; Gonzalez & Tyminski, 2020). Overall, findings suggest the need to create safe, quiet, and comfortable sleeping spaces for PEH to promote better physical and mental health. Healthcare practitioners could play a role in identifying at-risk individuals by inquiring about sleep in their contacts with PEH. Given the probable bi-directional relationship between sleep and health, service providers for PEH might consider combining comfortable and safe sleeping spaces with physical and mental health treatments to optimize health. Shelter intake workers may also work with PEH to identify barriers to accessing shelter at night, and assist with obtaining permanent housing.
Limitations
Study results should be interpreted cautiously in light of several study limitations. First, information regarding housing status beyond the previous evening was not available, which limited our ability to classify and detect sleep differences by typical housing status. Future studies should assess housing status over a longer period of time (e.g., over the past 30-days) in order to more accurately characterize individual housing situations. Second, since sleep characteristics were self-reported, sleep problems might have been under- or overestimated by participants. If feasible, future work should incorporate objective sleep assessment technologies such as actigraphy to enhance the accuracy and specificity of sleep data collection. In addition, the cross-sectional designed limited our ability to characterize the direction of the relationships between sleep problems and health outcomes; hence, the causal pathways between sleep problems and health issues could not be established. Finally, the findings from this study may be less generalizable to PEH in other states, regions, or countries, where sleep may be affected differently by local policies and access to services, as well as differing weather and climate patterns.
Conclusion
Study findings address an important gap within the literature by shedding light on the relationship between sleep problems and health among PEH. Sleep was strongly related to physical and mental health in this vulnerable population. Study results suggest the need to increase attention and shift resources towards creating comfortable, quiet, and safe sleeping spaces for PEH. Poor physical and mental health are commonplace among PEH, and interventions to promote better sleep have the potential to improve health in this vulnerable population.
AUTHOR NOTE
The authors have no conflicts of interest to disclose. This research was supported by the Oklahoma Tobacco Settlement Endowment Trust (TSET) grant R23-02, NCI Support Grant P30CA225520 awarded to the Stephenson Cancer Center, and NIMHD Grant K01MD015295 (to ACA).
Footnotes
CRediT STATEMENT
Nhu Huynh: Conceptualization, Formal Analysis, Writing – Original Draft Darla E. Kendzor: Supervision, Conceptualization, Writing – Review and Editing, Funding Acquisition Chaelin K. Ra: Writing – Review and Editing Summer G. Frank-Pearce: Writing – Review and Editing, Validation Adam C. Alexander: Writing – Review and Editing Michael S. Businelle: Writing – Review and Editing David R. Samson: Supervision, Conceptualization, Writing – Review and Editing
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Alexander AC, Waring JJC, Olurotimi O, Kurien J, Noble B, Businelle MS, Ra CK, Ehlke SJ, Boozary LK, Cohn AM, & Kendzor DE (2022). The relations between discrimination, stressful life events, and substance use among adults experiencing homelessness. Stress and Health, 38(1), 79–89. 10.1002/smi.3073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Anderson KN, & Bradley AJ (2013). Sleep disturbance in mental health problems and neurodegenerative disease. Nature and Science of Sleep, 5(default), 61–75. 10.2147/NSS.S34842 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Boozary LK, Frank-Pearce SG, Alexander AC, Sifat MS, Kurien J, Waring JJC, Ehlke S, Businelle MS, Ahluwalia JS, Kendzor DE (2022). Tobacco use characteristics, treatment preferences, and motivation to quit among adults accessing a day shelter in Oklahoma City. Drug and Alcohol Dependence Reports, 5, 100117. DOI: 10.1016/j.dadr.2022.100117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.BRFSS, 2021. Survey Questions, pg. 17, Core Section 2, Healthy Days <https://www.cdc.gov/brfss/questionnaires/index.htm>. Accessed Sep 21, 2022.
- 5.Cappuccio FP, D’Ellia L, Strazzullo P, & Miller MA (2010). Sleep Duration and All-Cause Mortality: A Systematic Review and Meta-Analysis of Prospective Studies. Sleep (New York, N.Y.), 33(5), 585–592. 10.1093/sleep/33.5.585 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Center for Disease Control and Prevention. 1 in 3 Adults don’t get enough sleep. (2016). https://www.cdc.gov/media/releases/2016/p0215-enough-sleep.html. Accessed Sep 9, 2022.
- 7.Centers for Disease Control and Prevention (CDC). (2017). Surveillance - Sleep and Sleep Disorder. Available at: https://www.cdc.gov/sleep/surveillance.html#brfss. Accessed Sep 9, 2022.
- 8.Chang H-L, Fisher FD, Reitzel LR, Kendzor DE, Nguyen MAH, & Businelle MS (2015). Subjective Sleep Inadequacy and Self-rated Health among Homeless Adults. American Journal of Health Behavior, 39(1), 14–21. 10.5993/AJHB.39.1.2 [DOI] [PubMed] [Google Scholar]
- 9.Chaput J-P, Dutil C, Featherstone R, Ross R, Giangregorio L, Saunders TJ, Janssen I, Poitras VJ, Kho ME, Ross-White A, & Carrier J (2020). Sleep duration and health in adults: an overview of systematic reviews. Applied Physiology, Nutrition, and Metabolism, 45(10 (Suppl. 2)), S218–S231. 10.1139/apnm-2020-0034 [DOI] [PubMed] [Google Scholar]
- 10.Corning Megan Elizabeth, “Sleep Disturbance in the Homeless Population: The Relationship between Homelessness, Sleep and Health” (2010). Honors Scholar Theses. 128. https://opencommons.uconn.edu/srhonors_theses/128 [Google Scholar]
- 11.Ehlke SJ, Cohn AM, Boozary LK, Alexander AC, Waring JJC, Businelle MS, & Kendzor DE (2022). Discrimination, Substance Use, and Mental Health among Sexual and Gender Minority Adults Accessing Day Shelter Services. Substance Use & Misuse, 57(8), 1237–1247. 10.1080/10826084.2022.2076874 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Geiger SD, Sabanayagam C, & Shankar A (2012). The Relationship between Insufficient Sleep and Self-Rated Health in a Nationally Representative Sample. Journal of Environmental and Public Health, 2012, 518263–518268. 10.1155/2012/518263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gonzalez A, & Tyminski Q (2020). Sleep deprivation in an American homeless population. Sleep Health, 6(4), 489–494. 10.1016/j.sleh.2020.01.002 [DOI] [PubMed] [Google Scholar]
- 14.Itani O, Jike M, Watanabe N, & Kaneita Y (2017). Short sleep duration and health outcomes: A systematic review, meta-analysis and meta-regression. Sleep Medicine, 32, 246–256. 10.1016/j.sleep.2016.08.006 [DOI] [PubMed] [Google Scholar]
- 15.Kakizaki M, Kuriyama S, Nakaya N, Sone T, Nagai M, Sugawara Y, Hozawa A, Fukudo S, & Tsuji I (2013). Long sleep duration and cause-specific mortality according to physical function and self-rated health: the Ohsaki Cohort Study. Journal of Sleep Research, 22(2), 209–216. 10.1111/j.1365-2869.2012.01053.x [DOI] [PubMed] [Google Scholar]
- 16.Karkoulias K, Lykouras D, Sampsonas F, Karaivazoglou K, Sargianou M, Drakatos P, … & Assimakopoulos K. (2013). The impact of obstructive sleep apnea syndrome severity on physical performance and mental health. The use of SF-36 questionnaire in sleep apnea. Eur Rev Med Pharmacol Sci, 17(4), 531–6. [PubMed] [Google Scholar]
- 17.Léger D, Beck F, & Richard JB (2017). Sleep Loss in the Homeless—An Additional Factor of Precariousness: Survey in a Group of Homeless People. JAMA Internal Medicine, 177(2), 278–279. 10.1001/jamainternmed.2016.7827 [DOI] [PubMed] [Google Scholar]
- 18.Liu Y, Wheaton AG, Chapman DP, Cunningham TJ, Lu H, & Croft JB (2016). Prevalence of Healthy Sleep Duration among Adults — United States, 2014. MMWR. Morbidity and Mortality Weekly Report, 65(6), 137–141. 10.15585/mmwr.mm6506a1 [DOI] [PubMed] [Google Scholar]
- 19.McConnell CF, Bretz KM, & Dwyer WO (2003). Falling Asleep at the Wheel: A Close Look at 1,269 Fatal and Serious Injury-Producing Crashes. Behavioral Sleep Medicine, 1(3), 171–183. 10.1207/S15402010BSM0103_4 [DOI] [PubMed] [Google Scholar]
- 20.McKnight-Eily L., Liu L, Perry G., Presley-Cantrell L., Strine T., Lu H, & Croft J. (2009). Perceived Insufficient Rest or Sleep Among Adults — United States, 2008. MMWR. Morbidity and Mortality Weekly Report, 58(42), 1175–1179. [PubMed] [Google Scholar]
- 21.R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/. [Google Scholar]
- 22.Ra CK, Hébert ET, Alexander A, Kendzor DE, Suchting R, & Businelle MS (2021). Unsheltered homeless and unstably housed adults have higher levels of stress and more health risk factors than sheltered homeless adults. Journal of Social Distress and Homeless, 32(1), 42–50. 10.1080/10530789.2021.1961990 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Reitzel LR, Short NA, Schmidt NB, Garey L, Zvolensky MJ, Moisiuc A, Reddick C, Kendzor DE, & Businelle MS (2017). Distress Tolerance Links Sleep Problems with Stress and Health in Homeless. American Journal of Health Behavior, 41(6), 760–774. 10.5993/AJHB.41.6.10 [DOI] [PubMed] [Google Scholar]
- 24.Sifat MS, Ehlke SJ, Ogunsanya M, Frank-Pearce SG, Boozary LK, Alexander AC, Businelle MS, & Kendzor DE (2023). Greater Discrimination Frequency and Lower Distress Tolerance Are Associated with Mental Health Problems Among Racially Privileged and Minoritized Adults Accessing an Urban Day Shelter. Journal of Racial and Ethnic Health Disparities. 10.1007/s40615-022-01496-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Strine TW, & Chapman DP (2005). Associations of frequent sleep insufficiency with health-related quality of life and health behaviors. Sleep Medicine, 6(1), 23–27. 10.1016/j.sleep.2004.06.003 [DOI] [PubMed] [Google Scholar]
- 26.Sullivan JH, Warkentin M, & Wallace L (2021). So many ways for assessing outliers: What really works and does it matter? Journal of Business Research, 132, 530–543. 10.1016/j.jbusres.2021.03.066 [DOI] [Google Scholar]
- 27.US Department of Housing and Urban Development (HUD). (2020). The 2017 annual homeless assessment report to Congress, part 1: Point-in-time estimates of homelessness. Accessed Sep 21, 2022. [Google Scholar]