Abstract
Background
Prior studies report less favorable sleep characteristics among non-Whites as compared with non-Hispanic Whites. However, few population-based studies have used objective measures of sleep duration, especially in more than two racial/ethnic groups. We tested whether objectively-estimated sleep duration and self-reported sleep quality varied by race and whether differences were at least partially accounted for by variability in clinical, psychological and behavioral covariates.
Methods
Adults aged 35-64 years who self- identified as White, Black, Asian or Hispanic were randomly sampled from Chicago, IL and surrounding suburbs. Our analytic sample included adults who had an apnea hypopnea index <15 after one night of screening and who completed 7 nights of wrist actigraphy for determination of sleep duration, sleep percentage, minutes of wake after sleep onset and sleep fragmentation (n=495). Daytime sleepiness was estimated using the Epworth Sleepiness Scale (ESS) and sleep quality was estimated from the Pittsburgh Sleep Quality Index (PSQI).
Results
Following statistical adjustment for age, gender, education, work schedule (i.e, day vs. night shift) smoking status, depressive symptoms, BMI, hypertension and diabetes, sleep duration (minutes) was significantly (all p<0.01) shorter in Black (mean=399.5), Hispanic (mean=411.7) and Asian (mean=409.6) participants than White participants (mean=447.4). All remaining sleep characteristics were significantly less favorable among Black participants as compared with White participants. Asian participants also reported significantly more daytime sleepiness than White participants.
Conclusions
Differences in sleep characteristics by race/ethnicity are apparent in a sample of adults with a low probability of sleep apnea and following adjustment for known confounders.
Keywords: sleep duration, disparities, epidemiologic studies
1. Introduction
There are notable and persistent disparities in the prevalence of major cardiovascular and metabolic disorders by race and ethnicity. Adverse health behaviors such as poor diet and physical inactivity account for a substantial proportion of these disparities [1]. Given research describing the contribution of sleep duration and quality to the development of obesity, hypertension, diabetes, cardiovascular disease and mortality [2-7], it is equally plausible that differences in sleep characteristics by race/ethnicity contribute to disparities in cardiovascular disease and metabolic disorders.
Variability in short sleep duration and poor quality sleep by sociodemographic characteristics have been observed in population research studies. Most often, non-White adults and adults from lower socioeconomic status groups report less favorable sleep characteristics including a higher prevalence of short and long sleep and poorer quality sleep than White adults [8-16]. Limitations of prior research on sociodemographic variability in sleep include reliance on self-reported vs. objectively-determined sleep duration, limited racial/ethnic variability within studies (i.e., most studies compare 2 groups) and potential confounding by the prevalence of sleep disorders such as obstructive sleep apnea which are independently associated with shorter sleep and poorer quality sleep [17].
Thus, the objective of our study was to describe and compare objectively measured sleep duration and quality via actigraphy and self-reported sleep quality and sleepiness in a population-based sample of White, Black, Hispanic and Asian adults who have a low probability of sleep apnea. An additional advantage over the previous studies is the ability to statistically adjust for potential behavioral and clinical confounders of the association between race/ethnicity and sleep characteristics. Our a priori hypothesis was that we would observe less favorable sleep characteristics in non-Whites vs. Whites, but that these differences are at least partially accounted for by differences in clinical and behavioral covariates across groups.
2. Methods
2.1 Study Participants and Design
We carried out our analyses in the Chicago Area Sleep Study (CASS), a cross-sectional population based epidemiologic study. Men and women ages 35-64 years who were living in Chicago, IL or the surrounding suburbs were identified via commercial telephone listings and contacted by mail and telephone. CASS staff administered a telephone screening to determine the likelihood of sleep apnea based on the Berlin Sleep Questionnaire [18], a modified STOP-BANG [19] (modified to use self-reported neck circumference for men) and body mass index (BMI). Adults were invited to participate if they met each of the following criteria: BMI <35 kg/m2; Berlin score < 3 (women) or <2 (men); and, a STOP-BANG <2 affirmative responses for women or <3 affirmative responses for men.
Eligible potential participants were invited to attend two clinical examinations approximately 1 week apart. Informed consent was obtained from all participants and all protocols were approved by the Northwestern University Institutional Review Board. Women were scheduled to attend their first examination during the mid-follicular phase of their menstrual cycle. At the first examination, staff explained the procedures for wearing the Apnealink Plus ® apnea screening device and the wrist actigraph and gave participants a set of questionnaires to complete prior to the next examination that was scheduled to take place a minimum of 8 days later and a maximum of 14 days later.
2.2 Measurements
2.2.2. Race/ethnicity
A primary goal of the study was to include equal representation of adults from four race/ethnic groups. Because race/ethnicity is primarily a social and cultural construct, we relied on self-report to determine race/ethnicity. Potential study participants were recruited from geographic areas in Chicago, IL and surrounding suburbs that had a high proportion of the targeted racial/ethnic groups. The commercial telephone listings included an indicator of race/ethnicity and telephone recruiters asked participants to confirm their race/ethnicity. Additionally, when participants attended the clinical examination, they were also asked to complete sociodemographic questionnaires indicating their race (Black, White, Asian) and whether or not they were of Hispanic ethnicity. Very few Hispanic participants reported both their ethnicity and their race and so we had substantial missing data for the question about race among Hispanics. We addressed this by classifying all participants indicating Hispanic ethnicity as Hispanic for our study. Where there was disagreement between self-reports of race/ethnicity, we classified participants based on the race/ethnicity that they reported on the sociodemographic questionnaires. Based on prior research reporting disparities in cardiovascular and metabolic disorders in Asian ancestral groups, we attempted to reduce variability by targeting Asians who reported Chinese ethnicity. However, we were unable to identify a sufficient number of eligible Chinese participants so we broadened our inclusion criteria to other East Asians (e.g., Korean, Japanese and Vietnamese) because South and Southeast Asians (e.g., Filipinos and South Asian Indians) have a markedly different cardiovascular and metabolic profile [20].
2.2.3 Sleep Characteristics
To restrict the sample to participants who had a low likelihood of obstructive sleep apnea, we asked participants at the first visit to wear the Apnealink Plus® apnea screening device for one night. We made post hoc exclusions for analysis to those participants whose AHI was <15 based on a minimum of 4 hours of wear time and using a combination of measurements including the nasal cannula, a chest belt to detect respiratory effort and a pulse oximeter to measure oxygen saturation. We repeated the analyses in the subset of 361 participants with AHI ≤ 5 for our sensitivity analysis. There is a high sensitivity (91%) and specificity (95%) between the ApneaLink Plus ® and laboratory polysomnography [21].
At the first examination, participants took home the wrist-worn Actiwatch™ 2 device (Phillips Respironics, Bend, OR). They were asked to wear the watch continuously for 7 days and nights until the second clinical examination and to keep a daily sleep log to record when they went to sleep and awoke each day and any times that they napped during the preceding 24- hour interval. The Actiware software program (version 5) and its built-in algorithm was used to analyze the actigraphy data. Automated scoring was not used because it has not been validated. Instead a member of the research team identified bed time and wake time using the sleep logs and the event markers. A proportion of the records (10%) were independently scored by KK and comparisons were made to ensure data quality. Time in bed was determined based on the Actiwatch™ device marker, which the participants were asked to press when they went to bed to sleep and when they woke. If participants did not use the marker, study staff estimated bed times and wake times based on self-reports recorded on the sleep log. Sleep duration was determined using the device software algorithms that quantified the amount of movement during time in bed. Minutes of wake after sleep onset (WASO) were calculated. The percentage of time the participant was asleep during the sleep period (sleep onset to sleep end) was calculated (sleep percentage). Sleep fragmentation is an index of restlessness during the sleep period expressed as a percentage. It is calculated by summing two percentages: (1) the percentage of the sleep period spent moving (an epoch with >2 activity counts is considered moving) and (2) the percentage of the number of immobile phases (consecutive epochs with no movement) that are only 1 minute long or less. Average values for each of the sleep characteristics were calculated for the 7 days. Participants were asked to complete the Pittsburgh Sleep Quality Index (PSQI) [22]. Scores range from 0 to 21 with higher scores indicating a worse sleep quality. Daytime sleepiness was measured using 8- item Epworth Sleepiness Scale;[23] higher scores (range 0-24) indicate greater sleepiness. Both the PSQI and ESS were treated as continuous variables in analyses.
2.2.4 Covariates
Participants were asked to fast for a minimum of 12 hours prior to the second clinic examination and to bring all prescription medications and over the counter supplements that they were currently taking. All clinical measurements (i.e., phlebotomy, blood pressure, anthropometry) were collected between 7:30am and 11:00am. Blood was drawn from participants in the seated position into citrate vacutainer tubes, centrifuged at 3,000 rpm at 4 °C for 20 minutes, and stored at -70 °C. Fasting glucose was determined from plasma using spectrophotometry. Whole blood was assayed for determination of hemoglobin A1c using an immunoturbidimetric assay. Diabetes status was determined if fasting glucose ≥ 126 mg/dL, or hemoglobin A1c ≥ 6.5% or if participants reported taking diabetes control medications [24]. Blood pressure was measured using an Omron automated cuff from participants in a seated position after five minutes of rest. Three measurements were collected and the final two were averaged. Hypertension was defined if participants had systolic blood pressure ≥ 140, diastolic blood pressure ≥ 90 or self-reported using antihypertensive medications. Height and weight were measured in light examination clothes and no shoes. Body mass index (BMI) was calculated as weight in kilograms divided by the height in meters2.
Age, gender and race/ethnicity were queried. We captured years of education completed as our measure of socioeconomic status. Work status (i.e., full-time, part-time, unemployed) and schedule (i.e., regular day shift, night shift or swing shift) were queried as both an additional measure of socioeconomic status and an indicator of regular sleep patterns. Questionnaires were used to ascertain smoking status, which were categorized into current, former and never use. Physical activity was assessed using the Global Physical Activity Questionnaire (GPAQ), a self-reported surveillance instrument that captures the frequency and duration of engagement in three domains of activity (i.e., work, leisure, transportation) and sedentary behavior [25]. Depressive symptoms were assessed using the Centers for Epidemiologic Studies Depression Scale (CES-D) [26].
2.3 Analysis Sample
Among the 630 who had valid actigraphy data to determine sleep duration and maintenance, 595 completed the clinical examination. We excluded 14 participants who did not have valid apnea-hypopnea index (AHI) values based on the multi-channel Apnealink® Plus (ResMed Germany Inc) and 52 participants with AHI ≥15. We additionally excluded 33 participants who did not report race/ethnicity or were using sleep medications or hypnotic antidepressants. After exclusions, there were 496 participants available for analysis.
2.4 Statistical Methods
Characteristics of the study sample are presented as means and standard deviations for continuous variables or proportions for categorical variables in the total sample and stratified by race/ethnicity. Sleep characteristics are described by race/ethnicity according to their mean and standard deviation, median and interquartile range, minimum and maximum. We used analysis of variance (ANOVA) to calculate adjusted means and 95% confidence intervals for each of the sleep characteristics comparing Black, Asian and Hispanic participants to White participants. Prior to modeling, we evaluated the presence of effect modification by gender and education level (centered at a mean=0 and standard deviation=1) by including multiplicative interaction terms in each model. Given the large number of interaction tests, we maintained conservative criteria for determining interaction at p<0.05. Models were adjusted for age, gender, education, work schedule, smoking status, depressive symptoms, BMI, hypertension and diabetes. Statistical significance was determined as p<0.05. All analyses were carried out using SAS version 9.2 (SAS Institute, Cary, NC).
3. Results
The distribution of participant characteristics is reported in Table 1. A roughly equal distribution of participants self-reported Black (n=155, 31%), Asian (n=109, 22%), Hispanic (n=103, 21%) and White (n=129, 26%) race/ethnicity. The majority (≥ 60%) of participants were female within each racial/ethnic group except for Whites (50% female). There was considerable variability in other sociodemographic characteristics; namely, White and Asian participants had the highest levels of education and the largest proportions were employed during the day. Cardiovascular risk factor profiles were also more favorable among White and Asian participants—fewer were current smokers, depressive symptom scores and BMI were lower and they had a lower prevalence of overweight, obesity, diabetes and hypertension than Black and Hispanic participants. Two exceptions were that both Black and Hispanic participants reported more minutes of physical activity per week than Whites and Asians.
Table 1. Distribution of Sample Characteristics by Race/Ethnicity.
| Total | Black | Asian | Hispanic | White | |
|---|---|---|---|---|---|
| N | 496 | 155 | 109 | 103 | 129 |
| Age, years | 47.8 (8.2) | 49.1 (8.3) | 48.9 (7.6) | 46.6 (8.1) | 46.2 (8.4) |
| Sex, % female | 60% | 66% | 60% | 66% | 50% |
| Education, years | 15.6 (3.9) | 14.7 (3.2) | 18.3 (2.8) | 12.2 (4.4) | 17.1 (2.5) |
| Work schedule, % | |||||
| Retired/Unemployed | 31% | 45% | 19% | 36% | 19% |
| Employed- Day shift | 51% | 36% | 70% | 42% | 59% |
| Employed- Other shift | 19% | 18% | 11% | 22% | 22% |
| Smoking status, % | |||||
| Current | 19% | 32% | 5% | 20% | 16% |
| Former | 20% | 16% | 18% | 18% | 28% |
| Never | 61% | 52% | 77% | 61% | 57% |
| Physical activity, min/week | 753.9 (1027.4) | 828.5 (1038.2) | 463.9 (621.3) | 1160.5 (1559.3) | 627.3 (669.7) |
| Depressive symptoms, CES-D score | 11.5 (9.6) | 12.6 (9.0) | 10.4 (9.8) | 12.0 (8.9) | 10.5 (10.5) |
| Body mass index, kg/m2 | 26.3 (4.5) | 28.2 (4.5) | 23.0 (3.1) | 28.5 (4.3) | 25.1 (3.6) |
| Weight status, % | |||||
| Normal weight | 44.4 | 27.7 | 80.7 | 18.5 | 54.3 |
| Overweight | 32.3 | 31.6 | 16.5 | 47.6 | 34.1 |
| Obese | 23.4 | 40.7 | 2.8 | 34.0 | 11.6 |
| Glucose, mg/dL | 91.8 (17.4) | 91.3 (16.3) | 91.7 (9.9) | 96.2 (25.3) | 89.2 (15.4) |
| Hemoglobin A1c, % | 5.7 (0.6) | 5.8 (0.6) | 5.7 (0.4) | 5.8 (0.8) | 5.5 (0.5) |
| Diabetes, % | 5.4 | 6.5 | 2.8 | 11.7 | 1.6 |
| Systolic blood pressure, mmHg | 115.3 (14.3) | 121.9 (14.6) | 112.3 (14.0) | 112.2 (13.2) | 112.6 (12.5) |
| Diastolic blood pressure, mmHg | 71.6 (10.4) | 76.6 (10.5) | 69.5 (9.7) | 69.1 (9.5) | 69.2 (9.6) |
| Hypertension, % | 16.9 | 35.5 | 8.3 | 13.6 | 4.7 |
Sleep characteristics were roughly normally distributed within race/ethnic groups, but reflect variability by race/ethnicity (Table 2). Black participants had the shortest sleep duration, the lowest sleep percentage and greater minutes of wake after sleep onset and sleep fragmentation. Self-reported sleep quality (PSQI) was worse and daytime sleepiness (ESS) was among the highest in Black participants. For each sleep characteristic, Hispanic and Asian participants had values that fell between those of Black (least favorable) and White (most favorable).
Table 2. Distribution of Sleep Characteristics by Race/Ethnicity.
| Mean (SD) | Median (Interquartile Range) | Minimum | Maximum | |
|---|---|---|---|---|
| Sleep Duration (minutes) | ||||
| Black | 409.3 (67.3) | 410.6 (91.3) | 116.5 | 589.6 |
| Asian | 413.0 (50.9) | 416.5 (61.1) | 201.7 | 555.5 |
| Hispanic | 417.0 (79.5) | 433.7 (107.3) | 60.1 | 592.7 |
| White | 444.7 (52.7) | 449.6 (59.7) | 200.3 | 660.2 |
| Sleep Percentage (%) | ||||
| Black | 87.6 (5.4) | 88.2 (7.3) | 68.9 | 97.4 |
| Asian | 90.7 (4.8) | 91.7 (5.2) | 60.7 | 96.0 |
| Hispanic | 89.7 (4.4) | 90.5 (5.6) | 71.5 | 96.8 |
| White | 91.5 (3.6) | 92.5 (3.9) | 72.4 | 96.9 |
| Wake After Sleep Onset (minutes) | ||||
| Black | 52.5 (27.1) | 47.4 (31.4) | 9.3 | 176.3 |
| Asian | 38.6 (18.4) | 34.9 (21.7) | 12.4 | 140.8 |
| Hispanic | 43.4 (19.6) | 40.4 (21.7) | 11.1 | 103.9 |
| White | 37.8 (16.7) | 33.1 (17.7) | 15.3 | 121.2 |
| Sleep Fragmentation (%) | ||||
| Black | 23.1 (7.7) | 22.9 (10.5) | 5.3 | 47.5 |
| Asian | 18.3 (8.4) | 15.8 (9.2) | 5.6 | 60.1 |
| Hispanic | 20.3 (7.7) | 18.5 (9.6) | 7.6 | 48.0 |
| White | 18.3 (7.2) | 16.5 (8.6) | 6.9 | 45.1 |
| Pittsburgh Sleep Quality Index | ||||
| Black | 6.8 (3.3) | 7.0 (3.0) | 0 | 18 |
| Asian | 5.4 (2.7) | 5.0 (4.0) | 0 | 16 |
| Hispanic | 5.8 (3.3) | 5.0 (4.0) | 0 | 16 |
| White | 5.0 (3.0) | 4.0 (3.0) | 1 | 15 |
| Epworth Sleepiness Scale | ||||
| Black | 7.2 (4.0) | 6.0 (5.0) | 0 | 19 |
| Asian | 7.6 (4.0) | 8.0 (5.0) | 0 | 20 |
| Hispanic | 6.9 (4.3) | 6.0 (5.0) | 0 | 22 |
| White | 5.9 (3.9) | 5.0 (5.0) | 0 | 18 |
Prior to modeling, we tested and confirmed the absence of statistical interaction by gender for the relationships of race with sleep duration (F = 1.2, p=0.31), sleep percentage (F= 1.32, p= 0.26), WASO (F=1.28, p=0.28), fragmentation (F=0.87, p=0.45), PSQI (F=0.67, p=0.57) and ESS (F=1.78, p=0.15). We pooled all analyses by gender. There was no evidence of statistical interaction by education level for sleep duration (F=2.21, p=0.09), sleep percentage (F=1.78, p=0.15) or fragmentation (F=1.14, p=0.33). We did find a significant interaction of race with WASO (F=2.81, p=0.04) and daytime sleepiness (F=3.6, p=0.01). Among Black and Hispanic participants, there was an inverse association of education with WASO, whereas there was no association among Asian or White participants. Education was positively associated with daytime sleepiness among Black participants but there was no association or an inverse association in the other race/ethnic groups. When we modeled the associations adjusted for sociodemographic characteristics, health behaviors, and cardiovascular disease risk factors, we included an interaction term for education in the models with WASO and daytime sleepiness as outcomes (Table 3). Black, Asian and Hispanic participants each had significantly (p<0.01) shorter average sleep duration than White participants. However, only Black participants had significantly lower sleep percentage, more minutes of WASO, greater sleep fragmentation and poorer self-reported sleep quality (PSQI) as compared with Whites; differences in those same measures comparing Hispanic and Asians with Whites were not statistically significant. Asians did report significantly (p<0.01) higher daytime sleepiness (ESS) than Whites.
Table 3. Adjusted means (95% confidence intervals) association of race with sleep characteristics.
| Black | Asian | Hispanic | White (Referent) | |
|---|---|---|---|---|
| Sleep Duration, minutes | 399.5** (388.3, 410.7) | 409.6** (394.7, 424.5) | 411.7** (397.4, 426.1) | 447.4 (435.7, 459.1) |
| Sleep Percentage, % | 87.9** (87.1, 88.8) | 89.7 (88.6, 90.8) | 89.5 (88.5, 90.6) | 90.9 (90.0, 91.8) |
| Wake After Sleep Onset, minutes† | 50.2** (46.4, 54.0) | 43.0 (38.0, 48.1) | 43.5 (38.6, 48.4) | 41.2 (37.2, 45.1) |
| Sleep Fragmentation | 22.9** (21.5, 24.3) | 20.0 (18.2, 21.8) | 20.9 (19.1, 22.7) | 19.5 (18.0, 20.9) |
| Pittsburgh Sleep Quality Index | 6.6** (6.0, 7.2) | 5.6 (4.8, 6.4) | 6.0 (5.3, 6.8) | 5.2 (4.6, 5.9) |
| Epworth Sleepiness Scale† | 7.1* (6.4, 7.8) | 7.4** (6.5, 8.4) | 6.9 (6.0, 7.9) | 5.8 (5.1, 6.6) |
Adjusted for age, gender, education, work schedule, smoking status, depressive symptoms, BMI, hypertension and diabetes
Interaction term between education and race included in the model
p<0.05
p<0.01
Restricting the sample to 403 day shift workers and unemployed/retired workers did not substantially change these associations (data not shown). Similarly, restricting to the subset of participants with AHI≤5 did not change the results (data not shown).
4. Discussion
We observed patterns of objectively determined sleep duration and self-reported sleep quality that were consistent with our primary hypothesis that non-Whites in our sample (i.e., Black, Asian and Hispanics) would have shorter sleep duration, worse subjective sleep quality and more daytime sleepiness than Whites. However, contrary to our secondary hypothesis, accounting for differences in social, psychological, behavioral and clinical characteristics across groups did not attenuate these differences. Findings from our population-based observational study describe racial/ethnic differences in sleep duration and quality that are not attributable to common covariates that we measured in our study.
Our findings are consistent with prior reports that are based on both self-reported measures of sleep and sleep symptoms by race/ethnicity and socioeconomic status. In one of the largest studies of 32,749 adults from the National Health Interview Survey (NHIS), Hale et al [8] reported that Black participants were 41% more likely to self-report being short sleepers and 62% more likely to report being long-sleepers than White participants; Hispanics and “others” were 26-35% more likely to report being short sleepers. In a 2013 report from the National Health and Nutrition Examination Survey (NHANES), race/ethnic minorities were more likely to report adverse sleep symptoms [27]. For example, black participants were more likely to report longer sleep latency than white participants and Hispanics/Latinos reported snoring more often than white participants. Grandner et al [27] indicated that the reporting varied by how questions were worded, which makes an argument for objective assessment of sleep. The findings were similar to other studies that relied on self-report [9, 10, 12, 14, 16, 27, 28].
Despite relatively low correlations between self-reported and objective measures of sleep duration as described in the Coronary Artery Risk Development in Young Adults (CARDIA) study (r=0.47) [29], findings are similar when sleep is objectively determined. The CARDIA study captured sleep using wrist actigraphy over 6 nights in 669 participants and sleep duration ranged from a high of 6.71 hours (SD=0.89) per night in White women to a low of 5.10 hours (SD=1.30) per night in Black men [11]. Disparities in CARDIA were also apparent by self-reported income (which in the US is closely correlated with race/ethnicity), whereby household income was positively associated with sleep duration and sleep efficiency and inversely associated with latency and time in bed. Although the absolute difference between groups amounts to a little over an hour and a half, meta-analyses suggest that crossing the threshold into “short” sleep is associated with significantly increased risks for hypertension, diabetes and mortality [2, 3, 7].
One strength of our study over prior studies that used actigraphy to capture sleep is that most prior studies compare only two racial/ethnic groups—most commonly Black and White adults. By contrast, we additionally included Asians and Hispanics whose sleep has been less frequently characterized. It was not unexpected that Black participants in our study had less favorable sleep characteristics than Whites according to each objective and self-reported measure. However, there is ample evidence that rates of metabolic disorders are higher in Hispanics than Whites and lower in adults of Asian ancestry as compared with Whites, yet we saw few differences in sleep characteristics other than duration and sleepiness (Asians only) as compared with Whites. Our findings contradict those in Study of Women Across the Nation and NHIS that describe variability in sleep characteristics across these groups [8, 9, 15].
We observed unexpectedly short sleep duration and high reports of daytime sleepiness in our Asian participants as compared with Whites. Prior studies have observed similar or even higher rates of sleep apnea among Asians despite lower BMI [30], which may be attributable to differences in craniofacial structure resulting in a smaller airway [31]. However, we attempted to exclude participants who had apnea using questionnaires to identify participants who had a high likelihood of apnea and using apnea screening equipment. Thus, the shorter sleep duration and greater daytime sleepiness that we observe in Asians as compared to Whites may not be due to apnea and therefore may be attributable to other unknown, and thus unmeasured, factors. Furthermore, the previously identified correlates of short sleep were not observed in Asians. For example, Asians had more education, lower BMI and fewer cardiovascular risk factors at baseline. The only notable difference was that physical activity levels were lower in Asians than for other racial/ethnic groups. Additional research on the contribution of cultural factors, attitudes towards sleep and family and household environment is warranted to explore the racial/ethnic differences in sleep.
Given the high rate of metabolic disorders in the Hispanic population overall and in this sample, it was unexpected that most sleep characteristics—with the exception of duration, were similar between Whites and Hispanics. A prior study in the Study of Women Across the Nation (SWAN) identified greater reporting of sleep complaints among US Born Latinas and Asians and among immigrant Latinas and Japanese (but not Chinese) with greater English language acculturation [15]. We did not capture details regarding acculturation in our cohort and so may have missed an important source of variability in perceptions of sleep among Hispanics and Asians. However, the findings reported in SWAN relied on self-reported complaints, which may be more susceptible to bias arising from social or cultural factors, whereas we captured objective measures of sleep.
A primary advance of our study over previous research is that we attempted to exclude participants who have moderate to severe obstructive sleep apnea (OSA) using screening equipment in addition to questionnaires. We additionally confirmed our findings in the subset of adults with a very low likelihood of apnea (apnea hypopnea index <5). Prior research has consistently demonstrated that sleep apnea is associated with adverse metabolic and cardiovascular disease risk. However, OSA is estimated to affect 2-4% of the population [32], which means that the very large remaining proportion of the population expresses variability in sleep duration and quality that is attributable to psychological, social, behavioral and environmental factors.
It is possible that genetic factors could contribute to differences in sleep duration. However, the majority of research on sleep genetics is related to sleep disorders [33] or genetically “short sleepers” who do not have a biological need for more than 5 hours of sleep per night [34]. There is very little research on genes in relation to the range of regular sleep duration. Further, if genes were the source of racial/ethnic variability in sleep duration and efficiency then those genes would be differentially distributed by race/ethnicity in the population. Few studies have a large enough sample size of non-Whites to explore the plausibility of the genetic hypothesis. Additionally, focusing attention on an immutable characteristic such as genes does not permit an opportunity for intervention to modify sleep behaviors.
While additional strengths of our study include population-based sampling required to describe the distribution of sleep in the general population, 7 days of actigraphy and standardized data collection in a research clinic setting, our findings should be interpreted in light of some limitations. The primary goal of our study was to study sleep characteristics in adults who were free from obstructive sleep apnea and so we excluded adults who had a high likelihood of apnea based on symptom screeners (i.e., Berlin and STOP-BANG). Those symptom screeners indicate that adults who have common cardiovascular disease risk factors such as hypertension are at high risk. Given that rates of hypertension are much higher in Black and Hispanic adults as compared with White adults [35], our sample distribution by race and ethnicity may have been biased towards a healthier sample of Black and Hispanic study participants and may not be representative of Black and Hispanic adults in the US. However, even with this limitation, we still observed shorter sleep duration among Black and Hispanic participants as compared with White participants. Although our sample size is relatively large compared with other studies that included objective determination of sleep, we did not have sufficient power to include more potential confounders in our statistical models. Similarly, we were underpowered to evaluate interaction in our sample. We did observe interactions by education level for two outcomes, WASO and daytime sleepiness, but those primarily indicated an isolated effect in Black and Hispanic adults who were also more likely to have lower levels of education. Future studies with larger sample sizes should investigate whether education exerts an independent effect on these sleep characteristics across race/ethnic groups. Despite findings from other studies indicating the role of psychological stressors such as perceived discrimination on sleep [36], we did not include a comprehensive battery of scales to assess discrimination, perceived stress, anxiety or other psychological factors which could vary by race/ethnicity and interfere with sleep. We did assess depressive symptoms and despite differences in the presence of symptoms by race/ethnicity, statistical adjustment for depressive symptoms did not change the patterns that we observed. Finally, we only included a single measure of socioeconomic status, years of education completed. Prior studies have consistently demonstrated that numerous markers of socioeconomic status including income, occupational classification and hours worked were all associated with sleep duration [9, 12, 37, 38]. It is possible that the racial/ethnic differences that we observed are attributable to residual confounding by socioeconomic status.
In summary, we observed shorter objectively-determined sleep duration in non-Whites as compared with Whites in our population-based sample of adults who had a low likelihood of sleep apnea that persisted following statistical adjustment for known correlates of short sleep and poor quality sleep. Future studies with larger sample sizes and more comprehensive measurements of psychological and social characteristics may be able to extend beyond our findings by investigating factors that contribute to variability in sleep by race/ethnicity. These observed patterns in sleep differences remain important given the potential for racial/ethnic differences sleep characteristics to contribute to disparities in the onset of cardiovascular and metabolic disease.
Supplementary Material
Highlights.
Largest epidemiologic study to include objective determination of sleep in a multi-ethnic population of Black, Asian, Hispanic and White participants
Adverse sleep characteristics observed in Black as compared with White participants
Despite similar cardiovascular risk profiles, Asian participants reported more daytime sleepiness than Whites
Acknowledgments
The study was funded by the National Heart, Lung and Blood Institute/National Institutes of Health grant R01HL092140.
Footnotes
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 citable 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.Shay CM, Ning H, Allen NB, Carnethon MR, Chiuve SE, Greenlund KJ, et al. Status of cardiovascular health in US adults: prevalence estimates from the National Health and Nutrition Examination Surveys (NHANES) 2003-2008. Circulation. 2012;125:45–56. doi: 10.1161/CIRCULATIONAHA.111.035733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cappuccio FP, Cooper D, D'Elia L, Strazzullo P, Miller MA. Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies. Eur Heart J. 2011;32:1484–92. doi: 10.1093/eurheartj/ehr007. [DOI] [PubMed] [Google Scholar]
- 3.Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care. 2010;33:414–20. doi: 10.2337/dc09-1124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Stranges S, Cappuccio FP, Kandala NB, Miller MA, Taggart FM, Kumari M, et al. Cross-sectional versus prospective associations of sleep duration with changes in relative weight and body fat distribution: the Whitehall II Study. Am J Epidemiol. 2008;167:321–9. doi: 10.1093/aje/kwm302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stranges S, Dorn JM, Cappuccio FP, Donahue RP, Rafalson LB, Hovey KM, et al. A population-based study of reduced sleep duration and hypertension: the strongest association may be in premenopausal women. J Hypertens. 2010;28:896–902. doi: 10.1097/HJH.0b013e328335d076. [DOI] [PubMed] [Google Scholar]
- 6.Knutson KL. Sleep duration and cardiometabolic risk: A review of the epidemiologic evidence. Best Practice & Research Clinical Endocrinology & Metabolism. 2010;24:731–43. doi: 10.1016/j.beem.2010.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies. Sleep. 2010;33:585–92. doi: 10.1093/sleep/33.5.585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hale L, Do DP. Racial differences in self-reports of sleep duration in a population-based study. Sleep. 2007;30:1096–103. doi: 10.1093/sleep/30.9.1096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Krueger PM, Friedman EM. Sleep duration in the United States: a cross-sectional population-based study. Am J Epidemiol. 2009;169:1052–63. doi: 10.1093/aje/kwp023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hall M, Bromberger J, Matthews K. Socioeconomic status as a correlate of sleep in African-American and Caucasian women. Ann N Y Acad Sci. 1999;896:427–30. doi: 10.1111/j.1749-6632.1999.tb08161.x. [DOI] [PubMed] [Google Scholar]
- 11.Lauderdale DS, Knutson KL, Yan LL, Rathouz PJ, Hulley SB, Sidney S, et al. Objectively measured sleep characteristics among early-middle-aged adults: the CARDIA study. Am J Epidemiol. 2006;164:5–16. doi: 10.1093/aje/kwj199. [DOI] [PubMed] [Google Scholar]
- 12.Stamatakis KA, Kaplan GA, Roberts RE. Short sleep duration across income, education, and race/ethnic groups: population prevalence and growing disparities during 34 years of follow-up. Ann Epidemiol. 2007;17:948–55. doi: 10.1016/j.annepidem.2007.07.096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jean-Louis G, Kripke DF, Ancoli-Israel S, Klauber MR, Sepulveda RS. Sleep duration, illumination, and activity patterns in a population sample: effects of gender and ethnicity. Biological Psychiatry. 2000;47:921–7. doi: 10.1016/s0006-3223(99)00169-9. [DOI] [PubMed] [Google Scholar]
- 14.Whinnery J, Jackson N, Rattanaumpawan P, Grandner MA. Short and long sleep duration associated with race/ethnicity, sociodemographics, and socioeconomic position. Sleep. 2014;37:601–11. doi: 10.5665/sleep.3508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hale L, Troxel WM, Kravitz HM, Hall MH, Matthews KA. Acculturation and sleep among a multiethnic sample of women: the Study of Women's Health Across the Nation (SWAN) Sleep. 2014;37:309–17. doi: 10.5665/sleep.3404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Piccolo RS, Yang M, Bliwise DL, Yaggi HK, Araujo AB. Racial and socioeconomic disparities in sleep and chronic disease: results of a longitudinal investigation. Ethnicity & disease. 2013;23:499–507. [PMC free article] [PubMed] [Google Scholar]
- 17.Rosenzweig I, Glasser M, Polsek D, Leschziner GD, Williams SC, Morrell MJ. Sleep apnoea and the brain: a complex relationship. Lancet Respir Med. 2015;3:404–14. doi: 10.1016/S2213-2600(15)00090-9. [DOI] [PubMed] [Google Scholar]
- 18.Chung F, Yegneswaran B, Liao P, Chung SA, Vairavanathan S, Islam S, et al. Validation of the Berlin questionnaire and American Society of Anesthesiologists checklist as screening tools for obstructive sleep apnea in surgical patients. Anesthesiology. 2008;108:822–30. doi: 10.1097/ALN.0b013e31816d91b5. [DOI] [PubMed] [Google Scholar]
- 19.Chung F, Yegneswaran B, Liao P, Chung SA, Vairavanathan S, Islam S, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108:812–21. doi: 10.1097/ALN.0b013e31816d83e4. [DOI] [PubMed] [Google Scholar]
- 20.Palaniappan LP, Araneta MR, Assimes TL, Barrett-Connor EL, Carnethon MR, Criqui MH, et al. Call to action: cardiovascular disease in Asian Americans: a science advisory from the American Heart Association. Circulation. 2010;122:1242–52. doi: 10.1161/CIR.0b013e3181f22af4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Erman MK, Stewart D, Einhorn D, Gordon N, Casal E. Validation of the ApneaLink for the screening of sleep apnea: a novel and simple single-channel recording device. Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine. 2007;3:387–92. [PMC free article] [PubMed] [Google Scholar]
- 22.Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213. doi: 10.1016/0165-1781(89)90047-4. [DOI] [PubMed] [Google Scholar]
- 23.Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14:540–5. doi: 10.1093/sleep/14.6.540. [DOI] [PubMed] [Google Scholar]
- 24.Standards of Medical Care in Diabetes—2010. Diabetes Care. 2010;33:S11–S61. doi: 10.2337/dc10-S011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Armstrong T, Bull F. Journal of Public Health. 2006. Development of the World Health Organization Global Physical Activity Questionnaire (GPAQ) [DOI] [Google Scholar]
- 26.Radloff LS. The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977;1:385–401. [Google Scholar]
- 27.Grandner MA, Petrov ME, Rattanaumpawan P, Jackson N, Platt A, Patel NP. Sleep symptoms, race/ethnicity, and socioeconomic position. J Clin Sleep Med. 2013;9:897–905. doi: 10.5664/jcsm.2990. A-D. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Grandner MA, Patel NP, Jean-Louis G, Jackson N, Gehrman PR, Perlis ML, et al. Sleep-related behaviors and beliefs associated with race/ethnicity in women. J Natl Med Assoc. 2013;105:4–15. doi: 10.1016/s0027-9684(15)30080-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lauderdale DS, Knutson KL, Yan LL, Liu K, Rathouz PJ. Self-reported and measured sleep duration: how similar are they? Epidemiology. 2008;19:838–45. doi: 10.1097/EDE.0b013e318187a7b0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Mirrakhimov AE, Sooronbaev T, Mirrakhimov EM. Prevalence of obstructive sleep apnea in Asian adults: a systematic review of the literature. BMC pulmonary medicine. 2013;13:10. doi: 10.1186/1471-2466-13-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lam B, Ip MSM, Tench E, Ryan CF. Craniofacial profile in Asian and white subjects with obstructive sleep apnoea. Thorax. 2005;60:504–10. doi: 10.1136/thx.2004.031591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lee W, Nagubadi S, Kryger MH, Mokhlesi B. Epidemiology of Obstructive Sleep Apnea: a Population-based Perspective. Expert review of respiratory medicine. 2008;2:349–64. doi: 10.1586/17476348.2.3.349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sehgal A, Mignot E. Genetics of sleep and sleep disorders. Cell. 2011;146:194–207. doi: 10.1016/j.cell.2011.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Pellegrino R, Kavakli IH, Goel N, Cardinale CJ, Dinges DF, Kuna ST, et al. A Novel BHLHE41 Variant is Associated with Short Sleep and Resistance to Sleep Deprivation in Humans. Sleep. 2014;37:1327–36. doi: 10.5665/sleep.3924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Egan BM, Zhao Y, Axon R. US trends in prevalence, awareness, treatment, and control of hypertension, 1988-2008. JAMA. 2010;303:2043–50. doi: 10.1001/jama.2010.650. [DOI] [PubMed] [Google Scholar]
- 36.Grandner MA, Hale L, Jackson N, Patel NP, Gooneratne NS, Troxel WM. Perceived racial discrimination as an independent predictor of sleep disturbance and daytime fatigue. Behav Sleep Med. 2012;10:235–49. doi: 10.1080/15402002.2012.654548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Basner M, Fomberstein KM, Razavi FM, Banks S, William JH, Rosa RR, et al. American time use survey: sleep time and its relationship to waking activities. Sleep. 2007;30:1085–95. doi: 10.1093/sleep/30.9.1085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Patel NP, Grandner MA, Xie D, Branas CC, Gooneratne N. “Sleep disparity” in the population: poor sleep quality is strongly associated with poverty and ethnicity. BMC Public Health. 2010;10:475. doi: 10.1186/1471-2458-10-475. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.