Examining the relationship of sociodemographic factors, neighborhood cohesion and abnormal sleep duration among U.S. foreign-born subpopulations in the National Health Interview Survey

Background Limited studies have examined the relationship of neighborhood cohesion and sleep duration between U.S. foreign-born Hispanics/Latinos and non-Hispanics/Latinos. Methods We conducted a multinomial logistic regression using the 2013-2018 National Health Interview Survey on U.S. foreign-born adults ≥18 (N = 27,253). The outcome variable, sleep duration, was categorized as short sleep (≤6 hours), normal sleep (7 to 8 hours), and long sleep (≥9 hours). Neighborhood cohesion was categorized using tertiles (low, medium, high) from self-reported Likert scores. Our model included sociodemographic factors (i.e., age, marital status), socioeconomic status (i.e., education, employment status), health risk behaviors (i.e., body mass index, smoking status, alcohol drinking status), ethnic identity (i.e., Mexican, Puerto Rican, Cuban, Dominican, Central or South American, other/multiple Hispanic/Latino, and non-Hispanic/Latino), and acculturation factors (i.e., years lived in the U.S.; the language of interview). Results Participants reporting low and medium neighborhood cohesion compared to high neighborhood cohesion had 45% (95% confidence interval [CI]:1.33-1.58) and 15% (95%CI:1.05-1.26) increased odds of short sleep (≤6 hours), compared to normal average sleep. Mexican participants had decreased odds of experiencing short sleep (adjusted odds ratio [AOR] = 0.82, 95%CI:0.73-0.92), while Puerto Ricans had increased odds of experiencing short sleep (AOR = 1.25, 95%CI:1.03-1.51) compared to non-Hispanics/Latinos. Conclusion Neighborhood cohesion was associated with increased odds of short sleep duration. Social determinants, acculturation, and behavioral risk factors in the context of neighborhood cohesion are critical to understand U.S. foreign-born Hispanic/Latino sleep duration, as these factors may negatively synergize to exacerbate risk, worsening mental and physical health outcomes.

with abnormal sleep duration include cardiovascular disease, diabetes mellitus, inflammation, and obesity [1][2][3]. Abnormal sleep has also been associated with increased mortality [4]. And while about 35% of the U.S. population suffers from short sleep [5], minority groups are reported to have a higher prevalence of short sleep when compared to non-Hispanic Whites [6,7] that may further widen the health disparities gap.
The U.S. Hispanic/Latino population will reach approximately 111 million by 2060 [8], making them the fastest growing minority group in the country. Health disparities among this group are prevalent in the U.S., with studies having reported that Hispanic's/ Latino's general and mental health are associated with overall negative outcomes when compared to that of non-Hispanic Whites [9]. Several complex and interconnected determinants of health at individual, interpersonal, community and societal levels, coupled with the process of immigration and acculturation contribute to negative health outcomes that may synergize to worsen physical and mental health among U.S. foreignborn [10,11]. A systematic review that explored the relationship between sleep and racial/ethnic disparities in cardiovascular disease found that minorities, in comparison to non-Hispanic Whites, were more likely to experience sleep deficiency that potentially increases the probability of developing or dying from this disease [12]. A possible explanation was that racial/ethnic minorities were more likely to reside in low-income/ underserved neighborhoods or communities with poor cohesion where environmental stressors and hazards are common. These exposures ultimately increase their risk of cardiovascular disease and other adverse health outcomes [12]. Other factors including occupational stressors, access to treatment, and psychosocial stressors (e.g., perceived discrimination), are also known to contribute to the increased risk of cardiovascular disease among minorities [12].
Using the Behavioral Risk Factor Surveillance System (BRFSS) data, Grandner et al. [13], examined the social and behavioral factors of short sleep duration and found that socioeconomic and sociodemographic factors (e.g., income, education, employment, and sex) were associated with insufficient sleep [13]. Similarly, longer-thanaverage sleep or long sleep has also been associated with socioeconomic status (SES), obesity, and mental health disorder [14]. Using a nationally representative sample, Grandner et al. [15], further analyzed sleep complaints in approximately 160,000 adults in the U.S. and found that lower SES was associated with an increased prevalence of sleep complaints [15]. Moreover, ecological factors like neighborhood cohesion have also been associated with sleep [16].
Neighborhood cohesion is defined as the connectivity and social relationship among members of a community based on norms, values, and cultures that impact a neighborhood [17]. This relationship is not just based on ethnic homogeneity or SES, rather, a spectrum of ecological factors. High neighborhood cohesion has been associated with superior health outcomes. For instance, one study that explored neighborhood cohesion and sleep outcomes found that those reporting low neighborhood cohesion were more likely to have shorter sleep duration [16].
Hispanics/Latinos represent the largest immigrant group in the U.S. Subsequently, their health constitutes a significant proportion of the U.S. public health burden, as they represent more than 18% of the population [18]. Existing evidence shows that sleep duration among U.S. Hispanics/Latinos varies by socioeconomic, sociodemographic, place of residence, and acculturation (i.e., adaption to the host culture) factors [19]. Despite the importance of neighborhood environment, acculturation, and their effect on sleep duration, few studies have investigated the relationship between abnormal sleep and neighborhood cohesion among diverse, foreign-born Hispanic/Latino subpopulations in the U.S.
The purpose of this study was to examine sleep duration and neighborhood cohesion among U.S. foreignborn subpopulations, specifically Hispanics/Latinos compared to non-Hispanics/Latinos. We hypothesized that individuals living in a low cohesion neighborhood would experience abnormal sleep durations, that is short and long sleep. We also comprehensively examined the relationship of abnormal sleep and SES, health risk behaviors, and acculturation. While previous studies have examined the associations between neighborhood cohesion and short sleep duration among Hispanics/ Latinos, our study not only included short sleep but long sleep duration as well. We analyzed complex survey data from the 2013-2018 National Health Interview Survey (NHIS) to fill a major gap in the current literature on sleep duration, neighborhood cohesion, Hispanic/Latino ethnicities, and health disparities.

Study participants
Data from the 2013-2018 NHIS were analyzed (see https:// www. cdc. gov/ nchs/ nhis/ index. htm). The NHIS is a cross-sectional household interview survey that targets the civilian noninstitutionalized population of the U.S. These analyses were restricted to foreign-born adults, i.e., those who were born outside of the 50 U.S. states and Washington, DC. Respondents with missing values for any of the variables of interest were excluded (i.e., 4,789), resulting in a final sample size of 27,253.

Outcome variable Sleep duration
Sleep duration was categorized using the following question: "On average, how many hours of sleep do you get in a 24-hour period? Enter hours of sleep in whole numbers, rounding 30 minutes (0.5 hour) or more up to the next whole hour and dropping 29 or fewer minutes. " Responses were categorized as ≤6 hours for short sleep, 7 to 8 hours for normal average sleep, and ≥9 hours for long sleep [20], where normal average sleep was used as the reference group.

Predictor variable Neighborhood cohesion
Neighborhood cohesion was based on NHIS participants' agreement with the following four statements: "(1) this is a close-knit neighborhood, (2) there are people you can count on in this neighborhood, (3) people in this neighborhood can be trusted, and (4) people in this neighborhood help each other out. Values were assigned to the responses as follows: definitely agree = 1, somewhat agree = 2, somewhat disagree = 3, definitely disagree = 4. " The sum of the four values was subtracted from 17 to create a neighborhood cohesion score ranging from 1 to 13. Tertiles were created based on the weighted distribution of the neighborhood cohesion score among foreign-born adults, and categorized as low, medium, and high. Low and medium cohesion were compared to high neighborhood cohesion.

Statistical analysis
Multinomial logistic regression models were used to examine the association between neighborhood cohesion and sleep duration. Sleep duration was the response variable, where normal average sleep (7 to 8 hours) was the reference category. Model

Sample characteristics
As seen in Table 1, the largest subpopulation among the sample was foreign-born non-Hispanic/Latino (50.4%), followed by Mexican (27.0%), Central or South American (11.6%), Puerto Rican (4.3% ), Cuban (3.6%), Dominican (2.4%), and lastly other/multiple Hispanic/Latino (0.7%). The majority of the participants (63.2%) reported a daily average sleep duration of 7 to 8 hours, while 30.5% reported that they slept less than or equal to 6 hours and 6.4% slept more than or equal to 9 hours. Regarding neighborhood cohesion, the largest proportion was found in the medium tertile of cohesion (37.7%), followed by high neighborhood cohesion (34.1%), and lastly low neighborhood cohesion (28.2%). About half of the participants reported being current drinkers (54.7%) and 75.0% reported that they have never smoked. When examining the number of years lived in the U.S., 81.0% reported that they have lived in the U.S. for 10 years or more. Regarding the language of the interview, the majority (71.0%) had the interview in English. See Table 1 for more details.

Multinomial logistic regression Short sleep duration
As indicated in Model 1, we found that factors such as neighborhood cohesion, sex, marital status, education, and employment were associated with reporting short sleep. In Model 2, health risk behaviors including BMI and alcohol drinking status were associated with short sleep duration. Model 3, the most comprehensive model, was selected for interpretation.

Short sleep duration comprehensive model
Compared to males, females had higher odds of reporting short sleep duration (adjusted odd ratio [AOR] = 1.13, 95% confidence interval [CI]: 1.04-1.22). See Table 2. Additionally, participants with low neighborhood cohesion and medium neighborhood cohesion, relative to those reporting high neighborhood cohesion, had higher Ethnic identities such as Hispanic/Latino ethnicity were found to have an association with sleep duration.  Table 2 for more details. Table 3 shows the relationship between neighborhood cohesion and long sleep duration. All the sociodemographic factors are shown to be associated with long sleep duration (see Model 1). We observed a statistically significant association between participants age, marital status, having a bachelor's degree or higher, and being employed with long sleep duration. However, no statistically significant association was found between neighborhood cohesion and long sleep duration among U.S. foreign-born subpopulations aged ≥18 years using the NHIS data from 2013 to 2018. In Model 2 after adding health behaviors, we found that individuals who reported having underweight BMI (<18.50), obese BMI (≥30.00), and former drinkers had an increased odds of experiencing long sleep duration. Again, Model 3 was selected for interpretation.

Long sleep duration comprehensive model
In our comprehensive analysis (i.e., Model 3), we assessed acculturation factors while controlling for sociodemographic and health behavior factors (see Table 3). Our results showed that females had increased    Table 3 for more details.

Discussion
To our knowledge, this study is amongst the first to examine the relationship between neighborhood cohesion and sleep duration among U.S. foreign-born (i.e., Hispanics/Latinos and non-Hispanics/Latinos) that accounted for SES, health risk behaviors, and acculturation. We found that low and medium neighborhood cohesion were strongly associated with short sleep duration among U.S. foreign-born adults. Furthermore, our study also found that the odds of abnormal sleep duration varied among foreign-born Hispanic/Latino subpopulations. Specifically, we found that Mexicans were 18% less likely to report short sleep, while Puerto Ricans had 25% increased odds of short sleep duration. These Hispanic/ Latino subpopulation differences could be explained not only by neighborhood cohesion, but by SES, behavioral risk factors, and acculturation factors in context of neighborhood environments, which have been reported to affected sleep duration among U.S. foreign-born adults.
Using NHIS data on 13,537 participants from 2010-2015, Murillo et al. [22], assessed the relationship between normal versus short sleep duration and neighborhood cohesion by Hispanic/Latino subgroups. The authors reported significant associations between medium and high neighborhood cohesion and normal sleep for all Hispanic/Latino subgroups, but the strongest association was among Cubans/Cuban Americans. These associations were accounted for by assessing sociodemographic and socioeconomic factors (i.e., sex/gender, education, employment status, marital status) in the model. Dudley [22], they only accounted for sleep outcomes (i.e., sleep duration hours, sleep fragmentation index, sleep maintenance efficiency) in their model. On the other hand, using the NHIS, Garcia et al. [24], reported that except for U.S.-born Cubans, all Hispanic/Latino subgroups had higher odds of short sleep duration compared to non-Hispanic Whites in the U.S. Our study did not find any relationship between Cuban, Dominican, Central or South American, and other/ multiple Hispanic/Latino ethnicity and sleep duration among foreign-born adults in the U.S., when compared to foreign-born non-Hispanics/Latinos. Garcia et al. [24] accounted for demographic characteristics, acculturation measures, socioeconomic characteristics, and behavioral health characteristics (e.g., age, sex, marital status, the language of interview, education status, smoking status) in their model. Sociodemographic factors in context of the neighborhood were found to have a significant relationship to neighborhood cohesion and sleep duration. We must also acknowledge that individual level factors also play a role in sleep duration and neighborhood dynamics. For instance, sex/gender has been reported to affect both short and long sleep duration [25][26][27][28]. Earlier research has suggested that poor neighborhood conditions were positively associated with higher relative risk of reporting very short sleep, long sleep, as well as napping in both men and women [28]. A systematic review that explored sleep characteristics across the lifespan in individuals from the Netherlands, United Kingdom, and the U.S., found that women were more likely to experience short sleep or slightly less efficient sleep than men [27]. A review that focused on sleep and women's health associated menstruation, pregnancy (i.e., prenatal & postnatal), and menopause as contributors to abnormal sleep among women due to the hormonal changes that occur within these physiological events [29]. In contrast, a study that analyzed the American Time Use Survey, a nationally representative sample from 2003 to 2007 (N = 56,149), and examined gender differences in time for sleep, [26] revealed women were more likely to sleep better than men, although the difference was small. Our study revealed that U.S. foreign-born women were 13% and 24% more likely than men to experience both short and long sleep duration, respectively. While abnormal sleep in female participants was more prevalent in our study, research assessing sleep patterns in males and females was limited, specifically long sleep. Further research on factors contributing to abnormal sleep as well as its variation among males and females is necessary.
Regarding socioeconomic factors, our study found that employed individuals had 14% higher odds of short sleep and 45% lower odds of long sleep. Similarly, individuals with some college or associate's degree were more likely to report short sleep (18%) while those with bachelor's degree or more were less likely to experience long sleep duration (38%) in comparison to those that had a Lifetime abstainer (≤12 drinks in life); Former drinker (no drinks past year); Current drinker (1+ drinks past year) Other ref.
--less than a high school education. We observed a nonstatistically significant association between attaining a bachelor's degree or higher and short sleep duration in U.S. foreign-born adults. Although it has been reported that individuals with higher educational attainment have an association with lower rates of chronic diseases compared to those with less education levels, higher educational attainment does not translate to improved sleep outcomes in our sample. However, prior studies on sleep duration found that having higher education was associated with better sleep outcomes [15,28]. Using the Jackson Heart Study on 5,301 African American participants, Johnson et al. [30], investigated the association between individual-level socioeconomic position, which was measured by income and education, and neighborhood characteristics with sleep duration and sleep quality. The authors reported an association between lower individual socioeconomic position and sleep duration/quality, that is, individuals with lower socioeconomic position were associated with increased odds of long sleep and poorer sleep quality. The authors also found that higher neighborhood problems were associated with poor sleep duration/quality [30]. More research, especially prospective studies, is necessary to provide a better understanding of the impact of socioenvironmental factors on sleep. Acculturation factors such as years lived in the U.S. are essential when analyzing sleep duration among U.S. foreign-born subpopulations. The present study indicated that those that have lived in the U.S. for 10 years or more have 16% higher odds of short sleep duration. A potential cause is the adoption of U.S. health risk behaviors that have conflicting effects. An example would be an increase of high caloric foods that lack nutritional value resulting in an increase in BMI. We found that those that were overweight or obese had higher odds of short sleep. Additionally, those that were obese and underweight had increased odds of long sleep duration. Individuals that have adopted or are adapting to the American diet are more likely to experience abnormal sleep duration [31][32][33]. A study that explored short sleep duration, BMI, and appetite regulatory hormones (e.g., leptin and ghrelin) found that individuals that were considered short sleepers had an elevated ghrelin level and reduced leptin level, which was associated with increased BMI [34,35]. Another study that examined the association between sleep duration and dietary consumption of fruits/vegetables reported that among adults in the United Kingdom, those that consumed lower amounts of fruits/vegetables were more likely to experience abnormal sleep [36].
While the language of interview was associated with short sleep, that is, English and/or Spanish compared to other language, this may be indicative of the diverse linguistic spectrum among U.S. foreign-born Hispanics/ Latinos. Furthermore, this may be the result of a diverse mix from both highly acculturated and non-acculturated foreign-born adults in the U.S. While we cannot make global assumptions about acculturation using language of interview, we did use it as a proxy. The assumptions regarding acculturation were limited in our study, as such future studies must go beyond language used for interview to capture language usage in specific everyday situations.
Of note was that long sleep duration was not found to have a significant relationship with neighborhood cohesion for U.S. foreign-born adults. Prior studies have found that long sleep duration was associated with cardiovascular disease, obesity, and diabetes, [2] but our study did not find association between neighborhood cohesion and long sleep duration. To our knowledge, while this study was not the first to explore neighborhood cohesion and abnormal sleep duration, it is amongst the first to explore the concept of long sleep among the U.S. foreignborn and its association with neighborhood cohesion.
Our study, however, was not without limitations. First, the NHIS data are cross-sectional, which does not allow us to assess changes in short and long sleep duration over time or establish and understand the causal link between neighborhood cohesion and abnormal sleep duration. Second, the number of acculturation variables in the dataset was limited. Third, other language was used as the reference for "language of survey" due to non-Hispanic/Latino being the largest group. Consequently, the purpose of our study was not only to examine Hispanic/Latino subgroups in the U.S. but also foreignborn non-Hispanic immigrant groups. Fourth, the sleep duration measures were self-reported which could have been biased. Despite these challenges, our findings are consistent with previous studies and contribute to the body of knowledge regarding the effect of neighborhood cohesion and abnormal sleep among immigrants and Hispanic/Latino subpopulations. Overall, our results demonstrate that there was an association with low and medium neighborhood cohesion and short sleep duration among U.S. foreign-born.

Conclusion
The dynamic changes and growth in U.S. immigrant populations call for innovative efforts to ensure their health and well-being. Immigrant sleep duration has only recently been explored and needs to be a priority in the public health system in the U.S. Sleep duration has many components that will affect its outcome, such as health risk behaviors, SES, acculturation factors, ethnic identity, and neighborhood cohesion. This study found that sleep duration varies among U.S. foreign-born Hispanics/Latinos when compared to foreign-born non-Hispanics/Latinos.