Heterogeneities in Sleep Duration and Quality Among U.S. Immigrants From Different Racial and Ethnic Backgrounds

Abstract

Objectives:

Sleep plays an essential role in well-being. Although U.S. immigrants are considerably growing, few studies have examined sleep in this diverse population, particularly those from Asian backgrounds. It is also unclear how sleep differs by the length of residence (LOR) across immigrant groups. In this study, we examined the relationships among race/ethnicity, LOR, and sleep using a nationally representative cohort of U.S. immigrants.

Methods:

We analyzed data from the 2013–2018 National Health Interview Survey. The sample (N = 27,761; 14% ≥ 65 years old) included foreign-born adults from the following racial/ethnic backgrounds: non-Hispanic (NH) White, NH Black, Asian (Chinese, Filipino, Asian Indian), and Hispanic/Latino. LOR was categorized as < 5, 5–9, 10–14, and ≥ 15 years. Sleep was assessed with self-reported sleep duration (normal, short, and long) and poor sleep quality (trouble falling asleep, trouble staying asleep, and waking up unrested).

Results:

Filipino and Hispanic/Latino immigrants reported the highest prevalence of short (41.8%) and long (7.0%) sleep, respectively. NH White immigrants had the highest prevalence rate across all three poor sleep quality measures (range 17.7–41.5%). LOR ≥ 15 years was significantly associated with worse sleep, and it moderated White-Asian differences in sleep quality. Immigrants from different racial/ethnic groups showed variations in sleep patterns as they resided longer in the U.S.

Conclusions:

Immigrants exhibited substantial heterogeneities in sleep. Future research should investigate the contributing factors to the variations in their sleep patterns, both between groups and within the same group of immigrants, in order to inform tailored interventions.

INTRODUCTION

Immigrants make up a considerable portion of the U.S. population. The number of foreign-born individuals has grown more than quadrupled in size since 1965, now accounting for 13.7% of the total population.¹ Notably by 2055, Asians are projected to surpass Hispanics and become the largest immigrant group in the U.S.² Migration itself can be a challenging experience, which is further complicated by acculturation – a dynamic process that involves an interplay between the practices, values, and identifications of both the heritage culture and the new, receiving culture.³ Consequently, immigrants face multiple stressors after emigrating to a new country, such as language barriers, discrimination, economic instability, and limited access to health care, all of which can adversely influence their health.^4,5

Chronic exposure to abnormal sleep duration and poor sleep quality is strongly associated with morbidities and all-cause mortality.⁶ Substantial epidemiological evidence has pointed to racial/ethnic disparities in sleep.^7,8 However, sleep research focusing on immigrant populations, especially those from Asian countries of origin, remains scarce.^9,10 Albeit limited, available evidence suggests suboptimal sleep health among Asian adults in comparison to their non-Hispanic (NH) White counterparts.¹¹ One study of community-dwelling adults reported that Chinese immigrants slept nearly one hour less than NH White adults.¹² Nevertheless, such findings are often limited by the lack of generalizability due to the use of convenience sampling. More importantly, past studies have predominantly considered Asians as a monolithic group. According to the U.S. standards for classifying race and ethnicity, over 40 countries fall within the Asian category.¹³ Aggregating individuals from different ethnicities and cultural backgrounds under the same umbrella can mask the diverse sleep patterns within Asian subgroups.¹⁴ Thus, investigating sleep health among various Asian ethnic subgroups and comparing their sleep with other major racial/ethnic groups is essential. This approach promises a more nuanced understanding of sleep health for immigrant populations.

While newly arrived immigrants tend to have better health outcomes than native residents, this phenomenon, known as the “healthy immigrant effect”, tends to deteriorate as their length of residence (LOR) increases.¹⁵ LOR has been commonly used as a proxy measure of acculturation.¹⁰ A meta-analysis showed that immigrants with longer LOR in North America had a higher risk of smoking, alcohol and drug use, and physical inactivity.¹⁶ Emerging evidence has suggested a similar link between extended LOR and sleep disadvantages for immigrants.¹⁰ Given the diverse migration experiences and the unique cultural challenges encountered by different racial/ethnic groups, it is likely that the relationship between LOR and sleep may differ across the immigrant groups. To date, no data have conclusively supported this hypothesis. This area of inquiry could shed light on the influence of LOR on sleep among immigrant populations as they acculturate over time.

In this study, we examined sleep duration and quality using a nationally representative sample of NH White, NH Black, Asian, and Hispanic/Latino immigrant adults residing in the U.S. Within the Asian cohort, we specifically focused on subgroups of Asian immigrants from China, India, and Philippines, which represent the top three largest countries of Asian origin in the U.S.¹ This research sought to address three questions:

1. Do sleep measures vary among immigrant adults from different racial/ethnic backgrounds?

2. Is there an association between LOR and sleep measures among immigrant adults?

3. Does LOR moderate the association between race/ethnicity and sleep measures among immigrant adults?

METHODS

Study Design and Dataset

We conducted a cross-sectional analysis of data from the 2013–2018 National Health Interview Survey (NHIS).¹⁷ The NHIS is a survey that collects a range of health topics through personal household interviews by the National Center for Health Statistics. By incorporating stratification, clustering, and oversampling of specific population subgroups, the NHIS ensures the representative of the civilian non-institutionalized population in the U.S. One sample adult and one sample child (if present) are randomly selected from each NHIS family to answer questionnaires.

Sample

Our sample consisted of immigrants aged 18 years or older, who were born either outside the U.S. or in the U.S. territories.

Independent Variables

Race/Ethnicity

Our study sample included adults from four major racial/ethnic backgrounds (NH White, NH Black, Asian, and Hispanic/Latino). Within the Asian group, we included individuals of Chinese, Filipino, and Asian Indian descent. We defined Hispanic/Latino as adults who are of Hispanic or Latino origin and may be of any race or combination of races. We excluded adults who identified themselves under the “Other Race” category or who were part of the “Other Asian” group due to limited representation. The final race/ethnicity variable consisted of six categories—NH White, NH Black, Chinese, Filipino, Asian Indian, and Hispanic/Latino.

Length of Residence

Foreign-born adults were asked about the number of years they spent in the U.S. with five choices: < 1, 1–4, 5–9, 10–14, and ≥ 15 years. We consolidated these choices to create three different categorizations of U.S. residence – two categories of LOR (< 10, ≥ 10 years), three categories of LOR (< 5, 5–9, ≥ 10 years), and four categories of LOR (< 5, 5–9, 10–14, ≥ 15 years). Sensitivity analyses were performed to identify the most appropriate categorization for our study.

Outcome Variables

Our outcomes of interest included one measure of sleep duration and three measures of sleep quality. Sleep duration was assessed using the response to the question, “On average, how many hours of sleep do you get in a 24-hour period?” The responses were recorded as integers. We categorized the responses as short (< 7 hours), normal (7–8 hours), and long (> 8 hours).¹⁸ The participants also reported how often in the past week they experienced trouble falling asleep, trouble staying asleep, and waking up feeling rested. Responses were recorded on a scale from 1 (no trouble/never felt rested this past week) to 7 (seven or more times). To ensure consistency, we reverse-coded the last question to reflect “waking up feeling unrested”. Subsequently, we dichotomized the responses for each of the poor sleep quality measures into categories: “yes” for those who experienced the symptom 3–7 times a week, and “no” for those who experienced the symptom ≤ 2 times a week.

Covariates

The analyses were controlled for sociodemographic information, health status, and health behaviors, all of which were coded as categorical variables. Sociodemographic information included: age (18–34, 35–64, ≥ 65 years old), sex (male or female), education (high school diploma/General Educational Development [GED] or less, some college/associate degree, bachelor’s degree or above), employment status (employed or unemployed), poverty status operationalized by the ratio of family income to the federal poverty threshold (≤ 1.38, 1.39–2.00, 2.01–4.00, ≥ 4.01), insurance status (insured or uninsured), marital status (married or not married), number of children under 5 years old (yes or no). Health status included: feeling sad in the past 30 days (yes or no), self-reported health status (good/excellent or fair/poor), and history of chronic diseases (yes or no). The respondents were considered to have a history of chronic diseases if they reported one or more of the following conditions: heart attack, stroke, coronary heart disease, heart conditions/diseases, cancer, chronic liver conditions, arthritis, chronic obstructive pulmonary disease, or arthritis/gout/lupus/fibromyalgia. Health behaviors included: smoking status (current smoker or lifetime abstainer/former smoker), drinking status (current drinker or lifetime abstainer/former drinker), use of sleep medications in the past week (yes or no), and body mass index (BMI). Considering the different criteria of BMI for Asian populations, we used ≥23 kg/m² as a cutoff for Asian adults and ≥25 kg/m² as a cutoff for other racial/ethnic groups to determine their overweight/obese status.¹⁹

Statistical Analysis

This study was exempted by the university’s Institutional Review Board due to the use of publicly available, secondary data without identifiable information. The NHIS data were accessed using the Integrated Public Use Microdata Service.²⁰ Considering the complex, multi-stage sampling design, we applied the survey-specific sampling weight and strata for all analyses. Descriptive statistics were employed to summarize the sample characteristics with weighted proportions. We performed Rao-Scott X² analyses for comparison of sociodemographic and health status among NH White, NH Black, Chinese, Filipino, Asian Indian, and Hispanic/Latino immigrants. Each sleep variable was examined across all racial/ethnic groups using weighted frequencies (percentages). We conducted generalized linear models with Poisson distribution to obtain prevalence ratios (PRs) and 95% confidence intervals (CIs) for each sleep measure. We chose Poisson regression over logistic regression in our analysis due to concerns that logistic modeling may overestimate prevalence ratios, especially when the prevalence of the outcome exceeds 10%.^21,22 The baseline model (Model 0) included age and sex variables, and Model 1 expanded on Model 0 by including race/ethnicity and the covariates. Model 2 was further built upon Model 1 by including LOR as an additional predictor. We ran the regression models with different categorizations of LOR separately for the sensitivity analysis. Once the final LOR categories were selected, we added the interaction term, race/ethnicity × LOR, to Model 3. Further, we assessed the relationship between LOR and sleep measures stratified by race/ethnicity. All statistical analyses were carried out using R 4.3.2.²³ The significance level was set at an alpha level of 0.05.

RESULTS

Table 1 presents data about sociodemographic and health status across different racial/ethnic groups. The total sample included 27,761 adults. This represented over 234 million adult immigrants living in the U.S., translating to 21% NH White, 9% NH Black, 18% Asian (32% Chinese, 26% Filipino, 42% Asian Indian), and 52% Hispanic/Latino immigrants. The weighted prevalence of the sample showed that 14% were adults aged 65 years or older, and over half were female. A large proportion resided in the U.S. for at least 15 years, ranging from 49% to 74%. About 9% of the sample used sleep medications in the past week. We observed significant differences for all variables across the six racial/ethnic groups (p-values < 0.001).

Table 1.

Sociodemographic and Health Status of the Sample

Variables	Total (N = 27,761)	NH White (N = 5,387)	NH Black (N = 2,420)	Asian (N = 4,940)	Chinese (N = 1,670)	Filipino (N = 1,350)	Asian Indian (N = 1,920)	Hispanic/Latino (N = 15,014)	p
Weighted sample size (%)	234,179,685 (100%)	48,206,666 (21%)	21,721,845 (9%)	42,620,181 (18%)	13,515,494 (32%)	11,084,359 (26%)	18,020,328 (42%)	121,630,993 (52%)
Age									<0.001
18-34 years old	28	23	31	28	29	17	33	29
35-64 years old	58	55	59	58	55	64	57	59
≥ 65 years old	14	23	10	14	16	20	9	12
Sex									<0.001
Male	49	48	51	46	45	41	51	50
Female	51	52	49	54	55	59	49	50
Length of residence in the U.S.									<0.001
< 5 years	10	10	14	16	16	7	20	7
5-9 years	10	8	13	14	13	12	16	9
10-14 years	13	9	14	14	14	12	16	15
≥ 15 years	67	74	59	57	57	69	49	69
Education									<0.001
High school/GED or below	48	26	37	20	27	18	16	68
Some college/Associate degree	20	25	29	16	14	28	10	18
Bachelor’s degree or above	32	48	34	64	58	53	74	13
Poverty status									<0.001
≤ 1.38	29	15	28	15	22	10	13	39
1.39-2.00	14	9	13	9	11	11	6	19
2.01-4.00	28	27	31	25	22	33	22	28
≥ 4.01	29	48	27	52	45	47	60	14
Employed	65	61	71	67	62	67	71	65	<0.001
Insured	76	90	82	92	92	91	94	64	<0.001
Married	63	66	51	73	67	69	80	60	<0.001
Have children ≤ 5 years old	16	11	16	14	10	11	20	18	<0.001
Feel sad in the past 30 days	27	27	26	25	27	26	22	29	<0.001
Fair/poor health status	13	11	9	8	10	10	6	16	<0.001
Overweight/obese	68	59	67	64	46	70	74	73	<0.001
History of chronic diseases	21	31	17	18	16	27	14	20	<0.001
Current drinker	55	68	46	48	48	54	44	53	<0.001
Current smoker	9	13	6	6	6	9	5	9	<0.001
Use of sleep medications in the past week	9	13	8	5	6	8	3	9	<0.001

Data are presented as percentages.

Abbreviations: GED = General Education Diploma; NH = non-Hispanic

The unadjusted prevalence rates of sleep duration and quality for all racial/ethnic groups are presented in Figure 1. Of the total sample, 36.5% reported either short or long sleep duration, 15.6% experienced difficulties falling asleep, 17.1% struggled to stay asleep, and 40.2% woke up feeling unrested. Filipinos had the highest percentage of adults sleeping < 7 hours (41.8%), whereas Hispanics/Latinos had the highest percentage of adults sleeping > 8 hours (7.0%). The prevalence rates of poor sleep quality remained the highest across all three measures among NH Whites, who were closely followed by Hispanic/Latino and Filipino immigrants. Conversely, Chinese and Asian Indian immigrants had relatively lower prevalence rates of experiencing poor sleep quality symptoms than other racial/ethnic groups.

Figure 1. Prevalence Rate of Sleep Duration and Quality Among U.S. Immigrants and Different Racial/Ethnic Groups.

Filipino immigrants exhibited the highest prevalence of sleeping < 7 hours, while Hispanic/Latino immigrants showed the highest prevalence of sleeping > 8 hours. The highest prevalence rates were observed among NH White immigrants for difficulties in falling asleep, staying asleep, and waking up feeling unrested. The data for the Asian group represent the average rates of sleep duration and quality across Chinese, Filipino, and Asian Indian immigrants. Abbreviation: NH = non-Hispanic.

As displayed in Table 2, age and sex were significant predictors of sleep duration and sleep quality in Model 0. Model 1 incorporated race/ethnicity as an additional predictor and adjusted for the covariates. As a result, age and sex remained significantly related to the sleep variables. We observed racial/ethnic differences in short sleep, trouble falling asleep, trouble staying asleep, and waking up unrested. NH Black, Filipino, and Hispanic/Latino immigrants were more likely to report short sleep duration in comparison to NH White immigrants. Notably, Asian Indian immigrants consistently had lower prevalence rates across all three poor sleep quality measures.

Table 2.

Association of Age, Sex, Race/Ethnicity, with Sleep

Study variables	< 7 hours		> 8 hours		Trouble Falling Asleep		Trouble Staying Asleep		Waking Up Unrested
	Model 0	Model 1	Model 0	Model 1	Model 0	Model 1	Model 0	Model 1	Model 0	Model 1
	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)	PR (95% CI)
Age (Ref: 18-34 years old)
35-64 years old	1.04*** (1.03, 1.05)	1.03*** (1.01, 1.04)	0.94*** (0.92, 0.96)	0.93*** (0.91, 0.95)	1.16** (1.06, 1.26)	0.90* (0.82, 0.99)	1.47*** (1.34, 1.61)	1.21*** (1.09, 1.34)	1.01 (0.96, 1.05)	0.97 (0.93, 1.01)
≥ 65 years old	1.05*** (1.03, 1.07)	1.02 (1.00, 1.04)	1.12*** (1.09, 1.15)	1.01 (0.97, 1.04)	1.39*** (1.26, 1.54)	0.71*** (0.63, 0.80)	2.04*** (1.84, 1.30)	1.14* (1.00, 1.30)	0.82*** (0.77, 0.87)	0.75*** (0.70, 0.80)
Sex (Ref: Male)
Female	1.02* (1.00, 1.03)	1.01 (1.00, 1.02)	1.04*** (1.02, 1.06)	1.01 (0.99, 1.03)	1.54*** (1.43, 1.66)	1.31*** (1.21, 1.41)	1.50*** (1.39, 1.61)	1.30*** (1.20, 1.40)	1.23*** (1.18, 1.28)	1.18*** (1.14, 1.24)
Race/Ethnicity (Ref: NH White)
NH Black		1.10*** (1.08, 1.13)		1.03 (0.99, 1.07)		0.89 (0.77, 1.03)		0.80** (0.70, 0.93)		0.98 (0.90, 1.07)
Chinese		0.99 (0.96, 1.01)		0.98 (0.94, 1.02)		0.61*** (0.50, 0.75)		0.63*** (0.52, 0.75)		1.02 (0.93, 1.11)
Filipino		1.11*** (1.08, 1.14)		0.99 (0.94, 1.04)		1.06 (0.89, 1.25)		0.81* (0.69, 0.96)		0.96 (0.87, 1.06)
Asian Indian		0.99 (0.97, 1.01)		0.99 (0.96, 1.03)		0.74** (0.61, 0.91)		0.57*** (0.48, 0.69)		0.77*** (0.70, 0.85)
Hispanic/Latino		1.02* (1.00, 1.04)		1.00 (0.97, 1.03)		1.00 (0.91, 1.10)		0.89** (0.82, 0.97)		1.04 (0.98, 1.09)

Note: Model 0 included age and sex. Model 1 was built upon Model 0 by adding race/ethnicity as a predictor; this model was additionally adjusted for education, employment status, poverty status, insurance status, marital status, have children < 5 years old, feel sad in the past 30 days, fair/poor health status, history of chronic diseases, smoking status, drinking status, and use of sleep medications in the past week.

Abbreviations: CI = confidence interval; NH = non-Hispanic; PR = prevalence ratio

Asterisks indicate level of statistical significance:

^*p < 0.05

^**p ≤ 0.01

^***p ≤ 0.001

After including LOR in Model 2, we found that increased LOR was consistently related to suboptimal sleep. Yet, the relationship yielded slightly different results when varying categorizations of LOR were used. When we divided LOR into 4 categories, LOR ≥ 15 years was significantly related to short sleep duration and all three measures of poor sleep quality, whereas LOR 10–14 years was associated with waking up unrested (PR: 1.15, 95% CI: 1.05–1.25) in comparison to LOR < 5 years. Based on these results of the sensitivity analysis, we adopted four categories of LOR for the subsequent analyses (Table 3).

Table 3.

Sensitivity Analysis of the Relationship Between Length of Residence and Sleep

Categories of LOR	< 7 hours			> 8 hours			Trouble Falling Asleep			Trouble Staying Asleep			Waking Up Unrested
Two categories (Ref: < 10 years)	PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI
LOR ≥10 years	1.03 **	1.01	1.04	0.99	0.96	1.01	1.31 ***	1.18	1.45	1.44 ***	1.29	1.60	1.16 ***	1.10	1.22
Three categories (Ref: < 5 years)	PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI
LOR 5–9 years	1.03 *	1.01	1.06	0.99	0.95	1.03	1.06	0.88	1.28	1.02	0.84	1.22	1.06	0.96	1.16
LOR ≥10 years	1.05 ***	1.02	1.07	0.98	0.95	1.01	1.35 ***	1.16	1.57	1.45 ***	1.24	1.69	1.19 ***	1.11	1.29
Four categories (Ref: < 5 years)	PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI
LOR 5–9 years	1.03 *	1.01	1.06	0.99	0.95	1.03	1.06	0.88	1.28	1.02	0.85	1.23	1.06	0.96	1.16
LOR 10–14 years	1.02	0.99	1.05	1.00	0.96	1.03	1.16	0.97	1.39	1.12	0.92	1.36	1.15 **	1.05	1.25
LOR ≥ 15 years	1.05 ***	1.03	1.08	0.98	0.94	1.01	1.41 ***	1.21	1.65	1.56 ***	1.34	1.82	1.21 ***	1.12	1.30

Note: Analyses were adjusted for age, sex, race/ethnicity, education, employment status, poverty status, insurance status, marital status, have children < 5 years old, feel sad in the past 30 days, fair/poor health status, history of chronic diseases, smoking status, drinking status, and use of sleep medications in the past week.

Abbreviations: CI = confidence interval; LOR = length of residence; NH = non-Hispanic; PR = prevalence ratio

Asterisks indicate level of statistical significance:

^*p < 0.05

^**p ≤ 0.01

^***p ≤ 0.001

Table 4 presents data on the moderating role of LOR on the relationship between race/ethnicity and sleep outcomes. The interaction between Chinese ethnicity and LOR ≥ 15 years was significantly related to short sleep duration (PR: 1.08, 95% CI: 1.02–1.14), trouble staying asleep (PR: 2.88, 95% CI: 1.31–6.33), and wake up unrested (PR: 1.46, 95% CI: 1.10–1.94). Similarly, the interaction between Asian Indian ethnicity and LOR ≥ 15 years was significantly related to wake up unrested (PR: 1.45, 95% CI: 1.10–1.92).

Table 4.

Length of Residence Moderates Racial/Ethnic Differences in Sleep

Interaction Term	Short (< 7 hours)			Long (> 8 hours)			Trouble Falling Asleep			Trouble Staying Asleep			Wake Up Unrested
	PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI		PR	95% CI
NH Black x LOR (5-9)	0.91	0.81	1.02	1.13	0.97	1.32	1.06	0.48	2.33	0.93	0.42	2.08	0.86	0.60	1.25
NH Black x LOR (10-14)	0.96	0.87	1.07	1.05	0.91	1.22	0.97	0.48	1.96	0.78	0.36	1.67	1.26	0.91	1.74
NH Black x LOR (≥15)	0.95	0.87	1.04	1.07	0.95	1.20	1.07	0.61	1.90	1.19	0.66	2.15	1.25	0.96	1.63
Chinese x LOR (5-9)	1.00	0.91	1.09	1.07	0.92	1.26	1.29	0.56	2.96	2.53	0.96	6.69	1.29	0.90	1.84
Chinese x LOR (10-14)	1.08	0.99	1.17	1.05	0.93	1.18	0.64	0.25	1.65	2.06	0.80	5.29	0.85	0.58	1.24
Chinese x LOR (≥15)	1.08 **	1.02	1.14	0.97	0.87	1.08	1.60	0.91	2.81	2.88 **	1.31	6.33	1.46 **	1.10	1.94
Filipino x LOR (5-9)	0.98	0.86	1.11	0.99	0.79	1.24	0.81	0.35	1.85	0.68	0.28	1.62	1.34	0.81	2.21
Filipino x LOR (10-14)	1.03	0.91	1.17	0.92	0.72	1.17	0.44	0.18	1.08	0.57	0.23	1.42	1.11	0.65	1.91
Filipino x LOR (≥15)	1.06	0.96	1.18	0.96	0.78	1.17	0.96	0.52	1.78	0.88	0.46	1.67	1.37	0.91	2.06
Asian Indian x LOR (5-9)	0.99	0.90	1.08	1.09	0.97	1.24	1.07	0.47	2.44	1.30	0.55	3.09	1.32	0.92	1.90
Asian Indian x LOR (10-14)	1.04	0.95	1.13	1.06	0.93	1.21	1.22	0.55	2.71	1.28	0.58	2.79	1.39	0.97	1.99
Asian Indian x LOR (≥15)	1.01	0.94	1.07	0.98	0.90	1.08	1.46	0.86	2.48	1.54	0.85	2.82	1.45 **	1.10	1.92
Hispanic/Latino x LOR (5-9)	0.99	0.92	1.06	1.05	0.93	1.17	0.94	0.60	1.47	0.83	0.52	1.32	0.95	0.74	1.21
Hispanic/Latino x LOR (10-14)	1.03	0.97	1.10	1.01	0.92	1.12	0.72	0.46	1.13	0.98	0.61	1.58	0.97	0.75	1.24
Hispanic/Latino x LOR (≥15)	1.00	0.95	1.05	0.97	0.89	1.05	0.90	0.65	1.25	0.98	0.70	1.38	0.99	0.83	1.18

Analyses were adjusted for age, sex, race/ethnicity, education, employment status, poverty status, insurance status, marital status, have children < 5 years old, feel sad in the past 30 days, fair/poor health status, history of chronic diseases, smoking status, drinking status, use of sleep medications in the past week, and length of residence.

Reference: NH White x LOR < 5 years.

Abbreviations: CI = confidence interval; LOR = length of residence; NH = non-Hispanic; PR = prevalence ratio

Asterisks indicate level of statistical significance:

^*p < 0.05

^**p ≤ 0.01

^***p ≤ 0.001

Subsequently, we examined sleep duration and quality stratified by LOR categories among immigrants (Figure 2 and Table S1). In general, immigrants who lived in the U.S. ≥ 15 experienced shorter sleep duration and/or poorer sleep quality in comparison to the newly arrived (LOR < 5 years). Long sleep duration did not significantly differ by LOR categories among immigrant groups. The relationship between LOR and sleep for each racial/ethnic group is discussed as follows. NH White: NH White immigrants who resided in the U.S. for ≥ 15 years were more likely to report short sleep, trouble falling asleep, and trouble staying asleep in comparison to their counterparts with LOR < 5 years (p-values < 0.05). NH Black: NH Black immigrants who resided in the U.S. for 10–14 years and ≥ 15 years faced a notably higher risk of waking up unrested than those with LOR < 5 years (p-values < 0.05). Chinese: Chinese immigrants with LOR 10–14 and LOR ≥15 years were more likely to report short sleep (p-values < 0.05). A distinct V-shaped pattern emerged in sleep quality measures for this group. When compared to recent arrivals, individuals with LOR 5–9 years exhibited higher rates of trouble staying asleep and waking up unrested. This trend dipped for those with 10–14 years of residence but surged again for those who resided in the U.S. for 15 years or longer. Filipino: While Filipino immigrants exhibited sleep patterns similar to Chinese immigrants, the prevalence of sleep duration and quality did not significantly differ by LOR categories. Asian Indian: Asian Indian immigrants who lived in the U.S. for 15 years or longer were more likely to report trouble falling asleep and staying asleep. In comparison to those with LOR < 5 years, the prevalence of waking up unrested for this group increased steadily with prolonged residency (p-values < 0.05). Hispanic/Latino: Hispanic/Latino immigrants with LOR 5–9 years were more likely to complain short sleep in comparison to those with LOR < 5 years. In addition, Hispanics/Latinos with LOR ≥ 15 were more likely to report short sleep duration, having trouble falling asleep, and having trouble staying asleep (p-values < 0.05).

Figure 2. Relationship Between Length of Residence and Sleep by Racial/Ethnic Groups.

In comparison to the newly arrived, immigrants from different racial/ethnic backgrounds demonstrated various patterns of sleep duration and quality as their length of residence in the U.S. increased. The asterisks indicate that there is a statistical difference in sleep measures when immigrants live in the U.S. for 5–9, 10–14, or ≥15 years in comparison to their counterparts who live in the U.S. for < 5 years. Reference: Length of residence < 5 years. Abbreviation: NH = non-Hispanic.

DISCUSSION

Our study assessed self-reported sleep quantity and sleep quality among U.S. immigrant adults from different racial/ethnic backgrounds, including three Asian ethnic subgroups. The prevalence of sleep duration and poor sleep quality varied among immigrant groups. Our findings indicate a positive relationship between LOR and worsening sleep health. Further, we discovered that LOR moderated White-Asian differences in poor sleep quality. Except for Filipinos, all immigrant groups experienced higher rates of short sleep and/or poor sleep quality after residing in the U.S. for 15 years or longer, compared to their counterparts who were newly arrived. However, different racial/ethnic groups exhibited varying patterns in sleep duration and quality as their LOR increased.

Our study builds upon the literature by highlighting that U.S. immigrant adults face notable sleep challenges. Consistent with previous findings, Filipino adults in the U.S. are particularly susceptible to inadequate sleep.²⁴ Wang and colleagues²⁵ reported that Filipinos had 0.55 times lower odds of reporting sufficient sleep in comparison to NH Whites, and foreign-born status was significantly associated with short sleep in this population. Evidence has also indicated that Hispanic/Latino individuals were more likely to report long sleep duration.²⁶ Surprisingly, our study found that NH Whites had the highest prevalence rate of poor sleep quality within the cohort, highlighting the need for more focused research on sleep issues among White immigrants. Though several studies indicate that immigrants often experience better sleep outcomes than their US-born counterparts²⁷, findings are mixed regarding nativity status and sleep. Some research has indicated poorer sleep among immigrant populations in comparison to the native-born. For example, Jackson and colleagues²⁸ found a higher prevalence of insufficient sleep among Black immigrant workers than US-born Black workers. Another study reported that Hispanic immigrants from the Dominican Republic, Cuba, and Central/South America were more prone to short sleep compared to White adults born in the U.S.²⁹ However, it is important to acknowledge that the selection of subjective versus objective sleep measurements can significantly affect the study results. Often, minoritized populations, especially Black and Asian immigrants, tend to endure adverse sleep outcomes in silence.^12,14 Research using actigraphs has found shorter sleep and poorer sleep quality among Chinese immigrants compared to NH White adults, but no marked differences were observed in self-reported sleep outcomes between the two groups.¹² The discrepancy implicates a potential underreporting of sleep problems in minority communities.¹¹ Since our sleep data were derived from self-reporting, the actual rates of poor sleep among immigrant populations could be higher than our data suggested. It is thus crucial to integrate both subjective and objective measurements to accurately identify sleep problems for immigrants.

Length of residence has been recognized as a vital risk stratification tool for monitoring and predicting health changes in immigrant populations.¹⁶ Our study reaffirmed the evidence of sleep disadvantages among immigrants who live in the U.S. for extended durations.^30,31 In the overall sample, LOR ≥ 15 years had a strong, consistent association with short sleep and poor quality. The results implicate that immigrants might begin to show a more evident sleep deterioration after they live in the U.S. for 15 years or longer, as opposed to the previously presumed 10 years.³² Our study additionally identified the moderating role of LOR ≥ 15 years in exacerbating NH White-Chinese and NH White-Asian Indian disparities in sleep quality. Sleep health for immigrants is influenced by an interplay of factors. The initial better sleep due to selective migration and healthy practices from one’s home country can be eroded by acculturation-related stressors over time. Increased LOR in the host country is strongly linked to acculturative stress, and this relationship has been observed across various immigrant groups.^33–35 For example, a study involving 400 Asian immigrants found that a one-unit increase in acculturative stress correlated with a decrease of 0.08 hour in sleep; moreover, those with higher levels of acculturative stress were 1.18 times more likely to experience sleep disturbances. Indeed, the mental and emotional challenges of adjusting to a new cultural environment may disrupt immigrants’ sleep-wake cycles, leading to poor sleep health. Beyond acculturation, cumulative exposure to social inequalities may further contribute to worsening sleep in long-term immigrants.¹⁶ Factors such as low income, lack of insurance, language barriers, suboptimal housing environment, restricted access to healthcare resources, discrimination, and systemic racism are among the numerous social determinants that directly or indirectly impair sleep to varying extents.³⁶

Our study revealed a nuanced relationship between LOR and sleep, which differed both in patterns as well as magnitudes across immigrant groups. In particular, the research shed light on the non-linear relationship between LOR and sleep in some immigrant groups. A particularly notable discovery is the distinct V-shaped pattern for sleep quality among Chinese immigrants. The findings implicate that assimilating into the host country may not be a straightforward progression but rather an intricate process.³⁷ Additionally, even though Chinese and Asian Indians overall had better sleep outcomes than other minority groups, their sleep health appeared to decline more sharply upon establishing a long-term residency in the U.S. The reasons for this are unclear; however, we speculate that resilience may play a potential role. Resilience-based coping has been shown to mediate the impact of psychological well-being on sleep quality.³⁸ It is plausible that immigrants who are less resilient to stress and challenges faced during their migration journey may experience more sleep problems. Equipping immigrants with robust coping skills could be an approach to assist them in managing stressors and maintaining optimal sleep health. Another attributable factor might be cultural identity. In Lee’s study, Chinese and Korean immigrants who strongly identified with their Asian heritage reported greater risks of sleep disturbances than those who considered themselves as “mostly Asian” or “bicultural/western”.³⁵ The researchers believed that individuals who rooted deeply in their original culture might perceive acculturative stress as a great threat, consequently resulting in poor sleep outcomes.³⁵ As such, the cultural conflicts could potentially explain why some Asian ethnic groups in our study exhibited vulnerability to poor sleep as LOR increased.

Improving sleep health is essential for immigrants. Our study points toward several new avenues for further exploration. More studies should be encouraged to elucidate the biological, psychological, and sociocultural factors contributing to the differences in sleep among immigrant groups. Our study underscores the importance of employing disaggregated data for Asian ethnic subgroups. Given the sleep disparities observed in the Asian cohort, we encourage researchers to investigate why Filipino immigrants are susceptible to short sleep, what mechanisms underlie the unique V-shaped sleep patterns among Chinese immigrants, as well as how cultural beliefs as well as practices shape Asian immigrants’ sleep health.³⁹ It is of paramount importance to perform routine sleep assessments for immigrants using culturally and ethnically sensitive tools. The tailored evaluation holds promise for better identification of sleep problems and treatments for immigrant communities.

Limitations

The strengths of our study include analyzing a racially and ethnically diverse sample of U.S. immigrants, desegregating sleep data for Asian ethnic groups, and examining sleep patterns across different LOR categories. Despite the strengths, our study is subject to several limitations. First, the cross-sectional design limited us from assessing the longitudinal relationship between LOR and sleep. Second, as previously discussed, the study results solely relied on self-reported data, which might be prone to recall bias or potential underreporting of sleep problems. We were also unable to examine other dimensions of sleep, such as daytime sleepiness and sleep disorders. Third, the limited study variables precluded us from assessing the influence of other acculturation proxies (e.g., acculturative stress) on sleep. Lastly, considering the scope of this study, we only focused on subgroups of Asian immigrants and did not differentiate between subgroups among NH Black and Hispanic immigrants.

CONCLUSIONS

Taken together, our study demonstrated considerable heterogeneities in sleep duration and quality among immigrants from diverse racial/ethnic backgrounds. Robust relationships were observed between race/ethnicity, LOR, and sleep outcomes. Given the essential role of sleep in overall health and the significant sleep health disparities faced by immigrant populations, our study represents a call to action to enhance sleep assessment and clinically indicated treatment of sleep symptoms for immigrant populations. Future research is essential to identify specific drivers that contribute to suboptimal sleep health, both within and among different immigrant subgroups.