Abstract
Job discrimination, a social stressor, may lead to sleep health disparities among workers; yet, limited research has examined this relationship and specific sources of job discrimination. We used a US sample of working women (n = 26,085), participants in the Sister Study (2008–2016), to examine the associations of perceived job discrimination due to sex, race, age, health conditions, and/or sexual orientation with sleep health. Cross-sectionally, linear or logistic regression models revealed that each source of job discrimination was independently associated with different sleep problems after controlling for other sources of job discrimination. Longitudinally, among participants without short sleep (<7 hours/night) at time 1 (2012–2014), age-specific job discrimination was associated with 21% increased odds of new-onset short sleep (odds ratio = 1.21, 95% confidence interval: 1.03, 1.43) at time 2 (2014–2016). Among those without insomnia symptoms at time 1, race-specific job discrimination was associated with 37% increased odds of new-onset insomnia symptoms (odds ratio = 1.37, 95% confidence interval: 1.07, 1.75) at time 2. Sex- and health-specific job discrimination also predicted new-onset sleepiness. There were dose-response relationships such that a greater number of sources of job discrimination (≥3) was associated with greater odds of prevalent and incident sleep problems. Perceived job discrimination may contribute to working women’s poor sleep health over time, raising concerns about sleep health disparities emanating from the workplace.
Keywords: health disparities, job discrimination, sleep health, worker health
Abbreviations
- CI
confidence interval
- OR
odds ratio
- SE
standard error
Perceived job discrimination, a social stressor, may be linked to health disparities in working populations (1). Women appear to be more vulnerable than men to experiencing discrimination at work, as evidenced by the gender gaps in wages and promotions that remain prevalent in the US workplace (2, 3). Perceived job discrimination may impact female workers’ sleep health; however, limited research has examined this relationship (4, 5).
Although not specific to perceived job discrimination, there is evidence for adverse associations of perceived discrimination with health outcomes. A meta-analysis showed that discrimination, prejudice, racism, sexism, and unfair treatment were each negatively associated with health across psychological and physical domains (1). Greater perceptions of inequality (or unequal treatment by others) have been associated with greater levels of allostatic load due to “wear and tear” resulting from sustained levels of stress (6, 7). Experiencing more discrimination in any form has also been associated with greater visceral fat in middle-aged women after adjustment for risk factors like depressive symptoms (8).
Relatively little research regarding perceived job discrimination and sleep has been conducted. In the Chicago Community Adult Health Study cohort, greater exposure to everyday racial and nonracial discrimination, major experiences of discrimination attributed to race/ethnicity, and workplace harassment and incivility were associated with shorter sleep and more sleep difficulties (5). A recent study using a population-based sample of middle-aged workers identified longitudinal associations of perceived unfairness about work with insomnia symptoms mediated by negative work-to-nonwork spillover over a 20-year time period (4). Given that stress appears to strongly influence sleep (9, 10), ever experiencing job discrimination may be associated with poor sleep health.
Working women may perceive different types of job discrimination due to varying social identities. Women’s sociodemographic characteristics may influence their perceptions of discrimination at work (11–14). Previous studies have found high prevalences of perceived job discrimination due to workers’ race (15, 16), weight (17–19), and age (20, 21). However, less is known about how different types of perceived job discrimination (due to race, sex, age, health conditions, and/or sexual orientation) contribute to sleep health in working women.
The prevalence of these types of job discrimination and their associations with sleep health may be more pronounced for relevant minority groups (e.g., racial/ethnic minorities, sexual minorities). The minority stress model suggests that minority groups may face conflicts with dominant values in the social environment and experience resultant unique stress (22). Older age and an unfavorable health condition may also produce inequalities related to discrimination. Thus, racial/ethnic minority women, older women, women with poorer health, or sexual minority women may be more likely to experience perceived job discrimination, and perceived job discrimination due to marginal status may degrade their sleep health.
Moreover, there may be a cumulative health burden associated with multiple exposures to job discrimination due to multiple social identities. According to the cumulative disadvantages perspective (23), persons with multiple minority statuses, such as older racial minorities or older racial and sexual minorities, may be at greater risk for job discrimination and poor sleep health. Examining various forms of perceived job discrimination in working women and their cumulative associations with sleep health can inform how disparities emanating from the workplace are associated with sleep health among those who are at particularly greater risk.
In this study, we examined cross-sectional and longitudinal associations of various types of perceived job discrimination (i.e., due to sex, race, age, health conditions, and/or sexual orientation) with multiple dimensions of sleep health, including sleep duration, insomnia symptoms, and daytime sleepiness, in working women. These 3 dimensions of sleep health capture both the quantity and quality of habitual sleep health frequently used in previous large-scale studies (4, 5, 13, 24). We hypothesized that, cross-sectionally, perceived job discrimination would be associated with having any of the following: shorter sleep duration, greater odds of insomnia symptoms, and greater odds of excessive daytime sleepiness. We explored whether the cross-sectional associations of each type of job discrimination with sleep health would be more apparent in relevant marginalized groups (i.e., racial minorities, older women, those who rated their health as good/fair/poor, and sexual minorities). Longitudinally, we hypothesized that perceived job discrimination would be associated with new onset of a sleep health problem among those who did not previously have one. We also evaluated dose-response relationships between perceived job discrimination and sleep health, hypothesizing that persons who experienced more (or multiple types of) job discrimination would exhibit poorer sleep health, cross-sectionally and longitudinally.
METHODS
Participants and procedure
Participants came from a sample of eligible women from the Sister Study (sisterstudy.niehs.nih.gov). The Sister Study, a prospective cohort study of women aged 35–74 years residing in the United States, was designed to explore genetic and environmental determinants of breast cancer (25, 26). The Sister Study enrolled 50,884 participants between August 2003 and July 2009. Recruitment materials and questionnaires were provided in both English and Spanish. Survey materials were first developed in English and then translated into Spanish by bilingual research staff. Translated material was retranslated independently into English to make sure the intended meaning was presented correctly across the 2 languages.
In baseline interviews (2003–2009), participants provided informed consent and extensive information about their sociodemographic characteristics, health behaviors, and health outcomes. A comprehensive follow-up questionnaire was administered every 2–3 years for detailed updates on lifestyle and health. Our main exposure—perceived job discrimination—was assessed during the first follow-up (2008–2012), and information on sleep characteristics—our outcomes of interest—was collected at the second (2012–2014) and third (2014–2016) follow-ups.
To be eligible for this study, participants had to be employed and could not have missing data regarding perceived job discrimination and sleep measures. Among 50,884 women who were included at baseline, 32,845 reported at the first follow-up (the baseline for this investigation) that they were currently employed. Therefore, 18,039 women who reported being retired (n = 9,972; 55%), unemployed (n = 764; 4%), a homemaker (n = 5,722; 32%), or something else (n = 1,581; 9%) were excluded. Of the 32,845 working women, 32,803 worked more than 8 hours/week; and of these women, 3,588 participants (10.9%) did not provide responses for at least 1 of the 5 perceived job discrimination measures. Of the remaining 29,215 participants, 3,130 (10.7%) did not provide responses to any of the questions on sleep characteristics at the second follow-up (hereafter called time 1). Compared with respondents, nonrespondents (those who did not provide responses to either job discrimination or sleep measures) were younger, were less educated, worked more hours per week, were in poorer self-rated health, and were more likely to be racial/ethnic minorities, foreign-born, widowed/divorced/separated, and raised in poor or low-income households. Thus, our final analytical sample for cross-sectional analyses was 26,085 working women. Among these eligible participants, 23,377 women who provided any sleep data at the third follow-up (hereafter called time 2) were included in longitudinal analyses.
The study was approved by the institutional review board of the National Institute of Environmental Health Sciences, and written informed consent was provided by each participant.
Measures
Perceived job discrimination.
At time 1, a series of questions were posed to women who had ever held a full-time or part-time job other than homemaking for 1 year or longer. The questions read, “Have you ever been treated unfairly in job hiring, promotion, or firing due to… 1) your race or ethnicity, 2) your sex, 3) your age, 4) an illness or medical condition, and/or 5) your sexual orientation?”. Participants could select multiple sources of job discrimination. Responses to each of the questions were coded as “yes” or “no.” We used the 5 binary variables as indicators for exposure to each source of job discrimination. We also created a summary score across the 5 variables to examine the cumulative burden associated with multiple exposures to job discrimination (1, 2, or ≥3 exposures vs. no exposure).
Sleep characteristics.
Three sleep dimensions/characteristics were assessed consistently at time 1 and time 2. Participants reported their habitual sleep duration (“In the past year, how many hours of sleep per night on average did you typically get?”; 1–24 hours), insomnia symptoms (“Do you have difficulty falling asleep or staying asleep on a regular basis?”; yes/no), and feeling excessive sleepiness (“Do you ever feel excessively sleepy during the day, even after getting your usual sleep?”; yes/no). Sleep duration was analyzed as a continuous outcome (1–24 hours) and as a binary outcome to detect those who had insufficient sleep duration. Using age-appropriate criteria for sufficient sleep duration (27), we dichotomized sleep duration into short sleep (<7 hours/night) versus longer sleep (≥7 hours/night).
Covariates and potential modifiers.
We considered sociodemographic characteristics that may be associated with sleep health (4, 28–30). Those included the participant’s age (years), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic/Latina, or other), nativity (foreign-born or US-born), educational attainment (ranging from 1 (no formal schooling) to 10 (doctoral degree)), marital status (legally married/living as married; widowed, divorced, or separated; or never married), sexual orientation (sexual minority (including homosexual/bisexual/asexual) or heterosexual), childhood socioeconomic status (“your family’s income level during the majority of your time growing up”; well-off, middle-income, low-income, or poor), body mass index (weight (kg)/height (m)2), depression (10-item Center for Epidemiologic Studies Depression Scale (31) score ≥10), and weekly work hours. All sociodemographic covariate data came from the baseline interview; weekly work hours were calculated considering multiple job-related variables at time 1. Continuous covariates (i.e., age, educational attainment, body mass index, and weekly work hours) were centered at the sample mean. We also used subgroups defined by age, race/ethnicity, and sexual orientation to examine whether the associations between perceived job discrimination (each type) and sleep health were more apparent in older people (≥53.4 years; dichotomized at median age), racial/ethnic minorities, and sexual minorities. We additionally used self-rated general health (“In the past 24 months, would you say your health has generally been... excellent, very good, good, fair, or poor?”) to evaluate the association between health-specific job discrimination and sleep health (optimal health (excellent/very good) vs. suboptimal health (good/fair/poor)).
Statistical analysis
We used descriptive statistics to examine the prevalence of perceived job discrimination and sleep health problems. To test our hypotheses, we used linear regression (for continuous sleep duration) and logistic regression (for binary indicators of short sleep, insomnia symptoms, and excessive sleepiness). In cross-sectional models, we also included all 5 types of job discrimination simultaneously in one model to examine independent associations of each source of job discrimination with sleep health after taking into account shared variance with other sources. To explore whether the relationship was more apparent in marginalized groups, we included the interaction between each source of job discrimination and a relevant subgroup—for example, race-specific job discrimination × racial/ethnic group. To test the longitudinal hypothesis, we determined whether perceived job discrimination was associated with the onset of each sleep health problem at time 2 among persons who did not have it at time 1. In all models, we adjusted for covariates. We used SAS 9.4 (SAS Institute, Inc., Cary, North Carolina) for the analyses. Statistical significance was defined as a 2-sided P value less than 0.05 or a 95% confidence interval that excluded 1. There were cases with incomplete data for some covariates (1%–4% missing data). Because less than 5% missingness is unlikely to contribute to differences in results/inferences (32), we treated those cases as missing at random (33).
RESULTS
Study population characteristics
Table 1 shows the sociodemographic characteristics of our study sample (n = 26,805) by experience of job discrimination (ever experienced discrimination from any source (25.48%) vs. never experienced discrimination (74.52%)). Among those who had ever experienced job discrimination and provided valid responses for each type, the prevalence by specific type was as follows: 17% for sex (4,308/26,051), 10% for age (2,499/26,019), 5% for race/ethnicity (1,205/26,036), 4% for health conditions (1,015/26,046), and 1% for sexual orientation (317/26,042). The correlation between types of job discrimination was low (see Web Table 1, available at https://academic.oup.com/aje), indicating that each type had its own, unique variance.
Table 1.
Sociodemographic Characteristics (%) of Working Women According to Their Experience of Job Discrimination, Sister Study, 2008–2012
| Experiencing Any Type of Job Discrimination | |||
|---|---|---|---|
| Sociodemographic Characteristic |
Total Sample (n = 26,085) |
Ever
a
(n = 6,646; 25.48%) |
Never (n = 19,439; 74.52%) |
| Race/ethnicity | |||
| White | 87.33 | 81.13 | 89.45 |
| Black | 6.82 | 13.32 | 4.60 |
| Hispanic/Latina | 3.42 | 2.78 | 3.64 |
| Other | 2.43 | 2.77 | 2.31 |
| Nativity | |||
| Foreign-born | 3.72 | 2.90 | 3.99 |
| US-born | 96.28 | 97.10 | 96.01 |
| Age categoryb | 53.28 (7.60) | 54.05 (7.26) | 53.01 (7.69) |
| Younger (<53.4 years) | 49.97 | 45.28 | 51.57 |
| Older (≥53.4 years) | 50.03 | 54.72 | 48.43 |
| Educational attainmentb,c | 7.34 (1.68) | 7.65 (1.59) | 7.24 (1.69) |
| ≥4 years of college | 55.80 | 62.96 | 53.35 |
| <4 years of college | 44.20 | 37.04 | 46.65 |
| Weekly work hoursb | 38.04 (13.14) | 39.66 (13.76) | 37.48 (12.88) |
| Marital status | |||
| Married/living as married | 75.15 | 67.91 | 77.63 |
| Widowed/divorced/separated | 18.83 | 24.43 | 16.91 |
| Never married | 6.02 | 7.66 | 5.46 |
| Sexual orientation | |||
| Heterosexual | 97.22 | 95.42 | 97.83 |
| Homosexual/bisexual/asexual | 2.78 | 4.58 | 2.17 |
| Childhood socioeconomic status | |||
| Well-off | 6.43 | 6.29 | 6.48 |
| Middle-income | 63.01 | 59.06 | 64.36 |
| Low-income | 24.13 | 25.76 | 23.57 |
| Poor | 6.43 | 8.88 | 5.59 |
| Body mass indexb,d | 27.51 (6.10) | 28.42 (6.45) | 27.20 (5.94) |
| Self-rated health condition | |||
| Excellent/very good | 71.50 | 64.49 | 73.90 |
| Good/fair/poor | 28.50 | 35.51 | 26.10 |
a Among women who reported that they had ever experienced job discrimination, 17% of the events (n = 4,308) were due to sex, 5% (n = 1,205) were due to race, 10% (n = 2,499) were due to age, 4% (n = 1,015) were due to health condition, and 1% (n = 317) were due to sexual orientation.
b Values are expressed as mean (standard deviation).
c Codes for educational attainment ranged from 1 (no formal schooling) to 10 (doctoral degree); 7 indicates an associate’s or technical degree (including nursing), and 8 indicates a bachelor’s degree.
d Weight (kg)/height (m)2.
Table 2 shows the prevalence of each type of job discrimination by relevant subgroup. The prevalence was higher in marginalized groups; race-specific job discrimination was higher in racial minorities, age-specific job discrimination was higher in older women, health-specific job discrimination was higher among women who rated their health as good/fair/poor (vs. excellent/very good), and sexual orientation-specific job discrimination was higher in sexual minorities. At time 1, mean sleep duration was 6 hours and 52 minutes per night; 64% of participants met the age-appropriate criterion for sufficient sleep (≥7 hours/night), and 36% had short sleep (<7 hours/night). Less than half of the sample reported insomnia symptoms (46%) or excessive daytime sleepiness (40%).
Table 2.
Prevalence (%) of Different Types of Perceived Job Discrimination Among Working Women, by Relevant Subgroup, Sister Study, 2008–2012
| Type and Experience of Perceived Job Discrimination | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Sex-Specific | Race-Specific | Age-Specific | Health-Specific | Sexual Orientation-Specific | ||||||
| Subgroup | Ever | Never | Ever | Never | Ever | Never | Ever | Never | Ever | Never |
| Sex (female) | 16.54 | 83.46 | 4.63 | 95.37 | 9.60 | 90.40 | 3.90 | 96.10 | 1.22 | 98.78 |
| Race/ethnicity | ||||||||||
| White | 16.34 | 83.66 | 1.45 | 98.55 | 9.31 | 90.69 | 3.79 | 96.21 | 1.10 | 98.90 |
| Black | 20.82 | 79.18 | 43.42 | 56.58 | 13.45 | 86.55 | 5.46 | 94.54 | 2.65 | 97.35 |
| Hispanic/Latina | 10.69 | 89.31 | 8.35 | 91.65 | 7.76 | 92.24 | 2.37 | 97.63 | 1.24 | 98.76 |
| Other | 19.91 | 80.09 | 5.70 | 94.30 | 11.87 | 88.13 | 5.54 | 94.46 | 1.27 | 98.73 |
| Age categorya | ||||||||||
| Younger (<53.4 years) | 15.85 | 84.15 | 4.58 | 95.42 | 6.25 | 93.75 | 4.09 | 95.91 | 1.27 | 98.73 |
| Older (≥53.4 years) | 17.22 | 82.78 | 4.68 | 95.32 | 12.95 | 87.05 | 3.70 | 96.30 | 1.17 | 98.83 |
| Self-rated health conditionb | ||||||||||
| Excellent/very good | 15.65 | 84.35 | 3.83 | 96.17 | 8.32 | 91.68 | 2.43 | 97.57 | 1.05 | 98.95 |
| Good/fair/poor | 18.74 | 81.26 | 6.56 | 93.44 | 12.70 | 87.30 | 7.55 | 92.45 | 1.64 | 98.36 |
| Sexual orientation | ||||||||||
| Heterosexual | 16.12 | 83.88 | 4.61 | 95.39 | 9.52 | 90.48 | 3.82 | 96.18 | 0.86 | 99.14 |
| Sexual minorityc | 30.89 | 69.11 | 5.39 | 94.61 | 12.72 | 87.28 | 6.77 | 93.23 | 13.75 | 86.25 |
a Age was dichotomized using the median value (median age, 53.4 years).
b General health condition was assessed by means of 1 item asking, “In the past 24 months, would you say your health has generally been: 1 = excellent, 2 = very good, 3 = good, 4 = fair, or 5 = poor?”.
c Homosexual, bisexual, or asexual.
Cross-sectional associations between types of job discrimination and sleep health problems
In model 1 (Table 3), ever experiencing any job discrimination (regardless of type) was associated with higher odds of insomnia symptoms and excessive sleepiness. In model 2, which included all types of job discrimination, race-specific job discrimination was significantly and independently associated with shorter sleep duration and higher odds of having short sleep duration (<7 hours/night). Sex-specific and health-specific job discrimination were associated with higher odds of having insomnia symptoms and excessive daytime sleepiness. Age-specific job discrimination was associated with higher odds of excessive daytime sleepiness. Sexual orientation-specific job discrimination was associated with higher odds of short sleep duration. These associations were found after controlling for all covariates.
Table 3.
Independent Cross-Sectional Associations of Different Types of Perceived Job Discrimination With Sleep Health Among Working Women, Sister Study, 2008–2014
| Sleep Health Characteristic | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
|
Sleep Duration, hours/night |
Short Sleep (  7 Hours/Night vs.  7 Hours/Night)
|
Insomnia Symptoms (Yes vs. No) |
Excessive Sleepiness (Yes vs. No) |
||||||||
| Parameter | β (SE) | P Value | OR | 95% CI | P Value | OR | 95% CI | P Value | OR | 95% CI | P Value |
| Model 1a | |||||||||||
| Intercept | 6.97 (0.02) | <0.001 | |||||||||
| Any JD | −0.01 (0.02) | 0.233 | 1.03 | 0.97, 1.10 | 0.304 | 1.19 | 1.12, 1.27 | <0.001 | 1.33 | 1.25, 1.41 | <0.001 |
| Model 2a | |||||||||||
| Intercept | 6.96 (0.02) | <0.001 | |||||||||
| Race-specific JD | −0.13 (0.40) | 0.001 | 1.17 | 1.01, 1.35 | 0.037 | 1.06 | 0.91, 1.22 | 0.457 | 1.09 | 0.94, 1.27 | 0.233 |
| Sex-specific JD | −0.01 (0.02) | 0.511 | 0.96 | 0.89, 1.03 | 0.248 | 1.20 | 1.11, 1.29 | <0.001 | 1.24 | 1.15, 1.33 | <0.001 |
| Age-specific JD | −0.02 (0.03) | 0.420 | 1.03 | 0.94, 1.13 | 0.531 | 1.03 | 0.94, 1.13 | 0.542 | 1.06 | 0.97, 1.17 | 0.219 |
| Health-specific JD | 0.03 (0.04) | 0.462 | 1.09 | 0.95, 1.25 | 0.237 | 1.17 | 1.02, 1.35 | 0.028 | 1.30 | 1.13, 1.49 | <0.001 |
| Sexual orientation-specific JD | −0.10 (0.07) | 0.120 | 1.40 | 1.09, 1.79 | 0.008 | 1.12 | 0.88, 1.44 | 0.356 | 1.16 | 0.91, 1.49 | 0.238 |
Abbreviations: CI, confidence interval; JD, job discrimination; OR, odds ratio; SE, standard error.
a Models adjusted for race/ethnicity, foreign birth, age, educational level, weekly work hours, marital status, sexual orientation, childhood socioeconomic status, body mass index, depression, and self-rated health.
We then examined cross-sectional associations of the different types of perceived job discrimination with sleep health by subgroup. Results for significant interactions are presented in Figure 1. There was a significant interaction between race-specific job discrimination and Hispanic race/ethnicity (vs. non-Hispanic White) in the association with insomnia symptoms (β = 0.81 (standard error (SE), 0.29); P < 0.01). Hispanics/Latinas who perceived job discrimination due to their race had 2.2 times’ higher odds of insomnia symptoms (odds ratio (OR) = 2.15, 95% confidence interval (CI): 1.28, 3.62) than their counterparts who never perceived any job discrimination due to race; this association was not found in non-Hispanic White women. There were interactive associations of age-specific job discrimination and age with the odds of short sleep (β = –0.20 (SE, 0.09); P < 0.05) and insomnia symptoms (β = –0.20 (SE, 0.09); P < 0.05). Associations of age-specific job discrimination with higher odds of short sleep (OR = 1.20, 95% CI: 1.03, 1.39) and insomnia symptoms (OR = 1.29, 95% CI: 1.11, 1.50) were found in younger women (age <53.4 years) but not in older women. These differences by age were not found when we tested the association using a different age cutoff (age <65 years).
Figure 1.
Cross-sectional associations of different types of perceived job discrimination (JD) (ever experienced vs. never experienced) with sleep health in marginalized sociodemographic groups, Sister Study, 2008–2014. A) Odds ratio for insomnia symptoms associated with race-specific JD. Results were significant for Hispanics (odds ratio (OR) = 2.15, P < 0.01). B) Odds ratio for short sleep (<7 hours/night) associated with age-specific JD. Results were significant for younger women (age <53.4 years) but not older women (age ≥53.4 years) (OR = 1.20, P < 0.05). C) Odds ratio for insomnia symptoms associated with age-specific JD. Results were significant for younger women but not older women (OR = 1.29, P < 0.01). D) Odds ratio for short sleep associated with health-specific JD. Results were significant for women who rated their health as good/fair/poor but not for those with excellent/very good health (OR = 1.29, P < 0.01). All models adjusted for covariates. Bars, 95% confidence intervals.
There was also a significant interaction between health-specific job discrimination and self-rated health associated with sleep duration (β = 0.19 (SE, 0.07); P < 0.05) and the odds of short sleep (β = –0.30 (SE, 0.14); P < 0.05). Among women who rated their health as good/fair/poor, perceived job discrimination due to one’s health condition was associated with 1.3 times’ higher odds of short sleep (OR = 1.29, 95% CI: 1.05, 1.51) in comparison with their counterparts who had never perceived job discrimination due to health; this was not found in women with excellent/very good health. The association of the interaction between sexual orientation-specific job discrimination and participants’ sexual orientation with excessive sleepiness reached statistical significance (β = –0.64 (SE, 0.28); P < 0.05); however, the negative association with job discrimination due to sexual orientation was found in heterosexual women (OR = 1.65, 95% CI: 1.24, 2.20), not in sexual minorities. All subgroup patterns (regardless of the significance of interactions) are shown in Web Figures 1−4.
Longitudinal associations between types of job discrimination and sleep health problems
We further examined whether ever experiencing job discrimination was associated with the new onset of a sleep health problem at time 2 among women who did not have the sleep health problem at time 1 (Figure 2). Age-specific job discrimination uniquely predicted the new onset of short sleep at time 2 (OR = 1.21, 95% CI: 1.03, 1.43). For the new onset of insomnia symptoms, race-specific job discrimination was an independent predictor after controlling for other sources (OR = 1.37, 95% CI: 1.07, 1.75). For the new onset of excessive daytime sleepiness, sex-specific (OR = 1.18, 95% CI: 1.04, 1.33) and health-specific (OR = 1.35, 95% CI: 1.07, 1.72) job discrimination were independent predictors.
Figure 2.
Longitudinal associations of different types of perceived job discrimination with the new onset of sleep health problems at time 2 (2014–2016) among working women in the Sister Study who did not have the problem at time 1 (2012–2014). A) Odds ratio (OR) for short sleep (<7 hours/night) (14,945 women did not have short sleep at time 1; 14,584 observations (97.6%) were used in the analysis due to missing responses for some variables). B) OR for insomnia symptoms (12,567 women did not have insomnia symptoms at time 1; 12,275 observations (97.7%) were used due to missing responses). C) OR for excessive daytime sleepiness (14,164 women did not report excessive daytime sleepiness at time 1; 13,833 observations (97.7%) were used due to missing responses). All models adjusted for covariates. Bars, 95% confidence intervals (CIs).
Dose-response relationships between job discrimination and sleep health problems
Cross-sectionally, working women who had ever experienced job discrimination for 3 or more reasons exhibited greater odds of having short sleep, insomnia symptoms, and excessive daytime sleepiness than those who had experienced job discrimination for 2 or 1 reasons and those who had never experienced job discrimination (Figure 3). In particular, there were dose-response relationships such that the odds of insomnia symptoms and excessive daytime sleepiness increased as the number of types of job discrimination increased. Longitudinally, women who had ever experienced job discrimination for 3 or more reasons also exhibited greater odds of developing a new sleep health problem than those who had experienced job discrimination for 2 or 1 reasons and those who had never experienced job discrimination (Figure 4).
Figure 3.
Cross-sectional dose-response relationship between the sum of different types of perceived job discrimination and sleep health problems among working women in the Sister Study, 2008–2014. A) Odds ratio (OR) for short sleep (<7 hours/night) (25,608 observations were used in the model due to missing responses for some variables). B) OR for insomnia symptoms (25,608 observations were used due to missing responses). C) OR for excessive daytime sleepiness (25,669 observations were used due to missing responses). All models adjusted for covariates. Bars, 95% confidence intervals (CIs).
Figure 4.
Longitudinal dose-response relationship between the sum of different types of perceived job discrimination and the new onset of sleep health problems at time 2 (2014–2016) among working women in the Sister Study who did not have the problem at time 1 (2012–2014). A) Odds ratio (OR) for short sleep (<7 hours/night) (14,945 women did not have short sleep at time 1; 14,674 observations were used in the analysis due to missing responses for some variables). B) OR for insomnia symptoms (12,567 women did not have sleep difficulty at time 1; 12,357 observations were used due to missing responses). C) OR for excessive daytime sleepiness (14,164 women did not report excessive daytime sleepiness at time 1; 13,908 observations were due to missing responses). All models adjusted for covariates. Bars, 95% confidence intervals (CIs).
Supplementary analyses
Although there were only a few women who completed the questionnaires in Spanish (1.17%), we conducted sensitivity analyses adjusting for language of survey administration (Spanish vs. not). All cross-sectional and longitudinal results were unaltered.
We further examined whether perceived job discrimination was associated with long sleep duration (>9 hours/night), although very few women in our sample had long sleep durations both at time 1 (0.99%) and at time 2 (1.13%). In cross-sectional models, experiencing health-specific (OR = 3.21, 95% CI: 2.14, 4.80), any type (OR = 1.37, 95% CI: 1.05, 1.80), and 2 types (OR = 1.57, 95% CI: 1.02, 2.42) of job discrimination were associated with greater odds of long sleep. Longitudinally, among participants who did not report long sleep at time 1, health-specific job discrimination (OR = 2.55, 95% CI: 1.54, 4.24) was associated with greater odds of new onset of long sleep.
In addition, we examined whether perceived job discrimination was associated with symptom severity at time 2 among women who had insomnia symptoms or daytime sleepiness at time 1. For those who reported insomnia symptoms at time 1, job discrimination did not predict the number of nights with insomnia symptoms at time 2. However, for those who reported daytime sleepiness at time 1, age-specific job discrimination marginally predicted more frequent symptoms of daytime sleepiness at time 2 (β = 0.07 (SE, 0.04); P = 0.059); experiencing 3 or more types of job discrimination also marginally predicted more frequent symptoms of daytime sleepiness at time 2 (β = 0.14 (SE, 0.08); P = 0.066).
DISCUSSION
Using a large sample of midlife working women in the United States, this study specified sources of perceived job discrimination (due to race, sex, age, health conditions, and/or sexual orientation) and used 3 different indicators of sleep health (sleep duration, insomnia symptoms, and excessive daytime sleepiness) to examine associations between perceived job discrimination and sleep health. Overall, findings from this study support the notion that perceived job discrimination is an important social stressor that may degrade workers’ sleep health (1, 6). This study contributes to the epidemiologic literature by demonstrating poor sleep health in working women who reported ever experiencing job discrimination (overall and across various types) due to their sociodemographic characteristics.
Investigators in previous studies have reported a high prevalence of perceived discrimination at work due to race (15, 16), weight (17–19), and age (20). However, few studies have examined the prevalence of and health associations with perceived job discrimination in working women and specific sources of it. In our large sample of working women in midlife, nearly one-third reported that they had ever experienced job discrimination due to their sociodemographic characteristics. Not surprisingly, the prevalence of perceived job discrimination was higher in marginalized groups, such as racial minorities, older women, women with suboptimal self-rated health, and sexual minorities. The results are consistent with the minority stress model (22). More research is needed on this topic, including more detailed information such as perceived severity of job discrimination (rather than experienced vs. not) and specific job contexts that induce discrimination (e.g., hiring, promotion, salary, and/or task distribution).
Each source of job discrimination was independently associated with poor sleep health across different dimensions, after taking into account associations of other sources of job discrimination. Moreover, each source of job discrimination independently contributed to the development of a new sleep problem longitudinally, and the magnitude differed by source of job discrimination and sleep dimension. Thus, perceived job discrimination was generally associated with poor sleep health concurrently and over time; however, its sensitivity to multiple aspects of sleep health differed by each source of discrimination that related to participants’ social and health status. The link between race-specific job discrimination and short sleep duration is consistent with previous research showing that racial minorities (who are more likely to report race-specific job discrimination) report short sleep but fewer sleep complaints than Whites (24, 34). The associations of sex- and health-specific job discrimination with insomnia symptoms and excessive sleepiness are also in line with the observation that women and persons those suffering from poor health tend to report more sleep difficulties (35). Our results show the importance of examining multiple aspects of sleep health (36) in the same sample, as well as multiple sources of job discrimination (15–21).
As expected, the associations between perceived job discrimination and sleep problems were more apparent in racial/ethnic minorities and those with poorer self-rated health. Hispanic/Latina participants had particularly higher odds of insomnia symptoms as a function of race-specific job discrimination compared with non-Hispanic Whites. In previous research, Kaufmann et al. (37) reported that, compared with non-Hispanic Whites, Hispanics are at a higher risk for insomnia and prone to worsening insomnia severity with age. Our finding additionally shows that Hispanic/Latina working women are susceptible to having insomnia symptoms if they perceive job discrimination due to their race. Moreover, women who rated their health as not optimal (poor/fair/good) had higher odds of shorter sleep duration. Given that the Sister Study recruited women who had a biological sister with breast cancer (26, 38), our sample might have been more sensitive to participants’ experiencing any negative acts associated with their health and medical conditions (even though they did not have breast cancer themselves). Inconsistent with our expectation, the associations of age-specific job discrimination with higher odds of short sleep and insomnia symptoms were pronounced for the younger age group including middle-aged women (ages 35–53.4 years), but not for the older age group. Working women in midlife may be particularly vulnerable to perceiving discrimination at work and may attribute it to their age (21).
This study further demonstrates that exposure to job discrimination for multiple reasons has a dose-response negative relationship with sleep health in working women. Women who experienced job discrimination for 3 or more reasons exhibited poorer sleep health concurrently and over time, compared with those who experienced job discrimination for 2 or 1 reasons and those who never experienced job discrimination. The results show cumulative disadvantages (23) in the sleep health of women who have been exposed to more types of discrimination.
This study had several limitations that may provide useful directions for future research. First, the design of this study was observational; thus, we could not determine whether there was a causal relationship between perceived job discrimination and sleep health. Second, our measure of perceived job discrimination was assessed at only 1 time point, at which participants were asked whether they had ever experienced job discrimination. Future research including a refined measure (including information on the severity and length of exposure) with more time points could examine longitudinal changes in the experiences and extent of perceived job discrimination and how those changes are associated with changes in sleep health. Third, despite the Sister Study’s having oversampled racial/ethnic minorities, there was underrepresentation of these groups. Moreover, our analytical sample appeared to be more socioeconomically advantaged than nonrespondents, and without data on those who were potentially more disadvantaged, we probably underestimated the associations of perceived job discrimination with sleep health. Lastly, we used self-reported sleep measures. Future research may benefit from incorporating objective measures of sleep, such as actigraphic sleep data (39).
On the basis of our finding that perceived job discrimination was associated with poor sleep health, employers could provide workplace-based intervention programs to decrease the incidence of discrimination at work as well as to enhance employees’ perceptions regarding work situations. Because managers are often involved in making decisions for promotions and evaluations of employees’ performance, it is important to reduce conscious and unconscious biases among managers toward minority employee groups to mitigate their detrimental influence (40). Given that workers’ sleep problems can significantly reduce work productivity (41–44), employers may consider providing educational sessions about “discrimination and fairness” to reduce unfair treatment and negative perceptions of the work environment among their employees, which are associated with various sleep problems.
In conclusion, we found that perceived job discrimination is an important social stressor that is associated with working women’s poor sleep health over time. Sleep health degraded by the experience of discrimination at work may affect work productivity, decision-making, and perception of stressors by workers, and more research is warranted (9, 42, 45, 46). To avoid creating a vicious cycle between poor sleep and perceived job discrimination, researchers and organizations should make more efforts to decrease all forms of discrimination in the workplace.
Supplementary Material
ACKNOWLEDGMENTS
Author affiliations: School of Aging Studies, College of Community and Behavioral Sciences, University of South Florida, Tampa, Florida (Soomi Lee); Department of Biobehavioral Health, College of Health and Human Development, Pennsylvania State University, University Park, Pennsylvania (Anne-Marie Chang, Orfeu M. Buxton); College of Nursing, Pennsylvania State University, University Park, Pennsylvania (Anne-Marie Chang); Division of Sleep Medicine, Harvard Medical School, Boston, Massachusetts (Orfeu M. Buxton); Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, Boston, Massachusetts (Orfeu M. Buxton); Division of Sleep and Circadian Disorders, Departments of Medicine and Neurology, Brigham and Women’s Hospital, Boston, Massachusetts (Orfeu M. Buxton); Intramural Program, National Institute on Minority Health and Health Disparities, Bethesda, Maryland (Chandra L. Jackson); and Epidemiology Branch, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina (Chandra L. Jackson).
This work was funded in part by the Intramural Research Program of the National Institutes of Health, National Institute of Environmental Health Sciences (grant Z1A ES103325-01 to C.L.J.).
We thank Dr. Aimee D’Aloisio, Jenna Waggoner, and Dr. Ross Ulmer for their help with data and variable construction.
Outside the scope of the current work, O.M.B. received subcontract grants to Pennsylvania State University from Proactive Life, Inc. (formerly Mobile Sleep Technologies LLC), doing business as Sleep Space (New York, New York) (National Science Foundation/Small Business Technology Transfer Program grant 1622766 and National Institutes of Health/National Institute on Aging Small Business Innovation Research Program grants R43AG056250 and R44 AG056250), honoraria/travel support for lectures from Boston University (Boston, Massachusetts), Boston College (Chestnut Hill, Massachusetts), Tufts University School of Dental Medicine (Boston, Massachusetts), New York University (New York, New York), and Allstate Corporation (Northbrook, Illinois), and an honorarium from the National Sleep Foundation for his role as the Editor-in-Chief of Sleep Health (https://www.sleephealthjournal.org/). None of the authors have any financial conflicts of interest relevant to the current study.
REFERENCES
- 1. Pascoe EA, Richman LS. Perceived discrimination and health: a meta-analytic review. Psychol Bull. 2009;135(4):531–554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Puhl RM, Andreyeva T, Brownell KD. Perceptions of weight discrimination: prevalence and comparison to race and gender discrimination in America. Int J Obes (Lond). 2008;32(6):992–1000. [DOI] [PubMed] [Google Scholar]
- 3. Ortiz SY, Roscigno VJ. Discrimination, women, and work: processes and variations by race and class. Sociol Q. 2009;50(2):336–359. [Google Scholar]
- 4. Lee S, Mogle JA, Jackson CL, et al. What’s not fair about work keeps me up: perceived unfairness about work impairs sleep through negative work-to-family spillover. Soc Sci Res. 2019;81:23–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Slopen N, Williams DR. Discrimination, other psychosocial stressors, and self-reported sleep duration and difficulties. Sleep. 2014;37(1):147–156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Schwartz JA. Long-term physical health consequences of perceived inequality: results from a twin comparison design. Soc Sci Med. 2017;187:184–192. [DOI] [PubMed] [Google Scholar]
- 7. Geronimus AT, Hicken M, Keene D, et al. “Weathering” and age patterns of allostatic load scores among Blacks and Whites in the United States. Am J Public Health. 2006;96(5):826–833. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Lewis TT, Kravitz HM, Janssen I, et al. Self-reported experiences of discrimination and visceral fat in middle-aged African-American and Caucasian women. Am J Epidemiol. 2011;173(11):1223–1231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Lee S, Crain TL, McHale SM, et al. Daily antecedents and consequences of nightly sleep. J Sleep Res. 2017;26(4):498–509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Benham G. Sleep: an important factor in stress-health models. Stress Heal. 2010;26(3):204–214. [Google Scholar]
- 11. Huffman ML, Cohen PN. Racial wage inequality: job segregation and devaluation across U.S. labor markets. Am J Sociol. 2004;109(4):902–936. [Google Scholar]
- 12. Chung-Bridges K, Muntaner C, Fleming LE, et al. Occupational segregation as a determinant of US worker health. Am J Ind Med. 2008;51(8):555–567. [DOI] [PubMed] [Google Scholar]
- 13. Hale L. Inadequate sleep duration as a public health and social justice problem: can we truly trade off our daily activities for more sleep? Sleep. 2014;37(12):1879–1880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Bobo LD, Fox C. Race, racism, and discrimination: bridging problems, methods, and theory in social psychological research. Soc Psychol Q. 2006;66(4):319–332. [Google Scholar]
- 15. Meyer JD. Race-based job discrimination, disparities in job control, and their joint effects on health. Am J Ind Med. 2014;57(5):587–595. [DOI] [PubMed] [Google Scholar]
- 16. Shippee TP, Schafer MH, Ferraro KF. Beyond the barriers: racial discrimination and use of complementary and alternative medicine among Black Americans. Soc Sci Med. 2012;74(8):1155–1162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Carr D, Friedman MA. Is obesity stigmatizing? Body weight, perceived discrimination, and psychological well-being in the United States. J Health Soc Behav. 2005;46(3):244–259. [DOI] [PubMed] [Google Scholar]
- 18. Robinson E, Sutin A, Daly M. Perceived weight discrimination mediates the prospective relation between obesity and depressive symptoms in U.S. and U.K. adults. Health Psychol. 2017;36(2):112–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Sutin AR, Stephan Y, Terracciano A. Weight discrimination and risk of mortality. Psychol Sci. 2015;26(11):1803–1811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Chou RJ-A, Choi NG. Prevalence and correlates of perceived workplace discrimination among older workers in the United States of America. Ageing Soc. 2011;31(06):1051–1070. [Google Scholar]
- 21. Petit P. The effects of age and family constraints on gender hiring discrimination: a field experiment in the French financial sector. Labour Econ. 2007;14(3):371–391. [Google Scholar]
- 22. Meyer IH. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychol Bull. 2003;129(5):674–697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Dannefer D. Systemic and reflexive: foundations of cumulative dis/advantage and life-course processes [published online ahead of print October 6, 2018] J Gerontol Ser B Psychol Sci Soc Sci. (doi: 10.1093/geronb/gby118). [DOI] [PubMed] [Google Scholar]
- 24. Jackson CL, Redline S, Kawachi I, et al. Racial disparities in short sleep duration by occupation and industry. Am J Epidemiol. 2013;178(9):1442–1451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Sandler DP, Hodgson ME, Deming-Halverson SL, et al. The Sister Study cohort: baseline methods and participant characteristics. Environ Health Perspect. 2017;125(12):127003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Chan SH, Van Hee VC, Bergen S, et al. Long-term air pollution exposure and blood pressure in the Sister Study. Environ Health Perspect. 2015;123(10):951–958. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Consensus Conference Panel, Watson NF, Badr MS, et al. Joint consensus statement of the American Academy of Sleep Medicine and Sleep Research Society on the recommended amount of sleep for a healthy adult: methodology and discussion. Sleep. 2015;38(8):1161–1183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Buxton OM, Lee S, Beverly C, et al. Work-family conflict and employee sleep: evidence from IT workers in the Work, Family and Health Study. Sleep. 2016;39(10):1871–1882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Berkman LF, Liu SY, Hammer L, et al. Work-family conflict, cardiometabolic risk, and sleep duration in nursing employees. J Occup Health Psychol. 2015;20(4):420–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lee S, Hale L, Chang AM, et al. Longitudinal associations of childhood bedtime and sleep routines with adolescent body mass index. Sleep. 2019;42(1):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Irwin M, Artin KH, Oxman MN. Screening for depression in the older adult: criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Arch Intern Med. 1999;159(15):1701–1704. [DOI] [PubMed] [Google Scholar]
- 32. Marshall A, Altman DG, Holder RL. Comparison of imputation methods for handling missing covariate data when fitting a Cox proportional hazards model: a resampling study. BMC Med Res Methodol. 2010;10(112):Article 110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Bryk AS, Raudenbush SW. Hierarchical Linear Models: Applications and Data Analysis. Newbury Park, CA: Sage Publications; 1992. [Google Scholar]
- 34. Grandner MA, Patel NP, Gehrman PR, et al. Who gets the best sleep? Ethnic and socioeconomic factors related to sleep complaints. Sleep Med. 2010;11(5):470–478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Knutson KL. Sociodemographic and cultural determinants of sleep deficiency: implications for cardiometabolic disease risk. Soc Sci Med. 2013;79:7–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Buysse DJ. Sleep health: can we define it? Does it matter? Sleep. 2014;37(1):9–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Kaufmann CN, Mojtabai R, Hock RS, et al. Racial/ethnic differences in insomnia trajectories among U.S. older adults. Am J Geriatr Psychiatry. 2016;24(7):575–584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Xu Z, Bolick SCE, Deroo LA, et al. Epigenome-wide association study of breast cancer using prospectively collected Sister Study samples. J Natl Cancer Inst. 2013;105(10):694–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Jackson CL, Patel SR, Jackson WB 2nd, et al. Agreement between self-reported and objectively measured sleep duration among White, Black, Hispanic, and Chinese adults in the United States: Multi-Ethnic Study of Atherosclerosis. Sleep. 2018;41(6):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Courtois M, Herman G. Managers are potential levers of discrimination within organizations: the role of group identification, assimilation endorsement and power. J Appl Soc Psychol. 2015;45(8):471–487. [Google Scholar]
- 41. Kessler RC, Berglund PA, Coulouvrat C, et al. Insomnia and the performance of US workers: results from the America Insomnia Survey. Sleep. 2011;34(9):1161–1171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Rosekind MR, Gregory KB, Mallis MM, et al. The cost of poor sleep: workplace productivity loss and associated costs. J Occup Environ Med. 2010;52(1):91–98. [DOI] [PubMed] [Google Scholar]
- 43. Lawson KM, Lee S. Better previous night sleep is associated with less next day work-to-family conflict mediated by higher work performance among female nursing home workers. Sleep Health. 2018;4(5):485–491. [DOI] [PubMed] [Google Scholar]
- 44. Lee S, Buxton OM, Andel R, et al. Bidirectional associations of sleep with cognitive interference in employees’ work days. Sleep Health. 2019;5(3):298–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Swanson LM, Arnedt JT, Rosekind MR, et al. Sleep disorders and work performance: findings from the 2008 National Sleep Foundation Sleep in America poll. J Sleep Res. 2011;20(3):487–494. [DOI] [PubMed] [Google Scholar]
- 46. Barnes CM, Schaubroeck J, Huth M, et al. Lack of sleep and unethical conduct. Organ Behav Hum Decis Process. 2011;115(2):169–180. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.