Introduction

There is a large body of evidence that recent immigrants are healthier than persons born in the USA [1]. This phenomenon, also known as the healthy immigrant effect, describes a pattern whereby first generation immigrants to North America have significantly less all-cause and disease-specific mortality [2, 3] and lower rates of chronic disease [4, 5] than their U.S.-born counterparts of the same race or ethnicity. As time in the USA increases, the health advantages of immigrants decrease significantly in what has been termed the Paradox of Assimilation [6]. Explanations for the observed initial health advantage of immigrants and the decline in health that comes with time in the USA include: selective immigration of healthier people [7, 8]; acculturation with negative health behaviors including changes in diet and substance use [9, 10]; and undiagnosed or undetected disease at time of immigration [11, 12].

Since race and ethnicity play an important role in shaping health outcomes in the U.S. population, much of the existing research compares the health of immigrants to U.S.-born populations, using the broad categories that constitute the racial hierarchy in the USA. However, the degree of immigrant pre-migration exposure to these (or any) racial hierarchies varies based on immigrants’ country of origin. For example, immigrants from countries such as Canada, France, and England have had exposure to racial structures that are comparable to the USA; whereas, immigrants from sub-Saharan Africa, South America, and Southeast Asia may not be as familiar with a heterogeneous, white majority system of racial stratification [13], although systems that advantage more light-skinned persons are still in place. As foreign-born persons enter a specific form of racialized space in the USA where, historically, opportunity has been provided based on whiteness and where racial health disparities are persistent [14–17], immigrants may begin to be forced into U.S. racial hierarchical patterning or may distinguish themselves from native born populations to avoid racial marginalization.

Immigrants in the USA may share similar experiences of migrating, but their immigrant status plays a smaller role than their race and ethnicity in determining socio-economic outcomes in the USA [18]. Hispanic/Latino immigrants and immigrants from racial minority groups are significantly more disadvantaged, while non-Hispanic White immigrants quickly become like U.S.-born Whites—occupying higher positions within the social hierarchy [19, 20]. Hispanic immigrants, even though they constitute a racially heterogeneous group, are racialized as one group of individuals with many races but who are linked closely by ethnicity [21]. In general, Hispanics and Latinos in the USA are perceived to occupy a distinct and sometimes intermediate position between non-Hispanic Whites and non-Hispanic Blacks. And, although most light-skinned Latinos, classify themselves as White, they are often excluded from the privilege that non-Latino/Hispanic Whites experience [22, 23]. Race and ethnicity complicate the immigrant experience such that it no longer makes sense to talk about a monolithic immigrant population. Immigrants experience racial and ethnic differences in opportunities and constraints that might have implications for health over time.

The main goal of this study is to explore the effects of U.S. race-based and ethnicity stratifications on the health trajectories of immigrants over time. Health trajectories are associated with subjective perceptions of self-rated health (SRH). Identifying racial and ethnic patterns in SRH among immigrants is relevant if there is genuine interest in eliminating health disparities in the USA and in guaranteeing overall population health and wellbeing. Cross-sectional, nationally representative survey data are used to address three specific objectives: (1) Examine racial and ethnic differences in self-rated health among immigrants; (2) Assess the degree to which differences in education, income, employment status, and access to insurance explain racial and ethnic differences in self-rated health among immigrants; and (3) Determine the effect of time in the USA on racial and ethnic patterning of self-rated health among immigrants. Since more time in the USA reflects longer exposure to the U.S. racial terrain where persons who belong to racial minority groups and who are of Hispanic ethnicity are exposed to circumstances such as socio-economic disadvantage and discrimination that impact health [24–26], we hypothesize that longer time in the USA should be associated with poorer SRH among immigrants who belong to minority groups compared to non-Hispanic White immigrants. Assessing racial and ethnic patterns in SRH among immigrants is an important step towards predicting future patterns of disparities in the general population, especially as immigration continues to be the dominant factor in population growth in the USA.

Materials and Methods

Data

Data were obtained from the Integrated Health Interview Series (IHIS). The IHIS is an online resource of harmonized data from the National Health Interview Survey (NHIS) from 1969 to present [27]. The NHIS is a cross-sectional annual survey that uses a multi-stage probability sample design to collect health data on a nationally representative non-institutionalized U.S. sample. Data from 2012 were used for analyses. The analytic sample consisted of 16,288 persons who were not born in the USA or in a U.S. territory. Distinction of immigrants based on reason for immigration is not provided in dataset, but we assumed that the sample included short-term visitors such as students and temporary workers, permanent residents, asylees, refugees, undocumented non-citizens, and naturalized citizens.

Measures

Main Variables

The dependent variable is self-rated or self-reported health. In the NHIS, respondents are asked whether they would rate their health in general as excellent, very good, good, fair, or poor. Responses were dichotomized into poor health (fair or poor) and good health (excellent, very good, or good), as is commonly done in health research [28–30]. The key independent variable is race/ethnicity. Respondents selected a racial category and indicated whether they were of Hispanic ethnicity. Since Hispanics/Latinos do not conceptualize racial categories exclusively in terms of skin color, and are racialized based on ethnicity [21], a combined race/ethnicity variable was created with ethnicity assigned first. The variable consists of the following categories: non-Hispanic White, non-Hispanic Black, Hispanic, and Asian. Non-Hispanic persons who identified as American Indian/Alaskan Native and multiple and other race were excluded because of their limited sample size among the foreign-born population.

Socio-economic Indicators

Socio-economic indicators included educational attainment (less than high school, completed high school, some college, college degree, and graduate degree), employment status (employed/unemployed or not in the labor force), annual household income in relation to percentage of the Federal Poverty Line (FPL) (less than 100 % FPL, 100–199 % FPL, 200–299 % FPL, 300–399 % of FPL 400 % or more of FPL), and health insurance status (insured/uninsured).

Time in the USA

In the NHIS, length of stay in the USA is divided into five categories (<1, <5, <10, <15, or ≥15 years). The first two categories are combined in our analyses to increase cell sizes.

Other Covariates

Other variables controlled for in this study include age (continuous), gender (male/female), marital status (single/never married, divorced/separated/widowed, married), U.S. citizenship (yes/no), and language in which the survey was completed (English only, Spanish/Spanish, and English). These factors are associated with health status and might confound the relationships between race, length of stay in the USA and SRH among immigrants [31–36].

Analyses

Bivariate differences between groups on socio-demographic factors and health status were assessed using two sample tests of proportions with non-Hispanic Whites as the reference. Logistic regressions were computed to assess the association between race/ethnicity and poor SRH. The unadjusted model was fitted first. Next, demographic variables and socio-economic indicators were included successively to assess if any of these factors intervened in the relationship between race/ethnicity and SRH (Aim 2). Then, U.S. citizenship status, language of survey, and length of stay in the USA were adjusted. An interaction between race/ethnicity and length of stay was tested to assess whether the association between race/ethnicity and SRH varied by length of stay (Aim 3). In all analyses, sampling weights were used to account for differential selection probability [24]. Standard errors were adjusted for the complex sampling design. We used the standard alpha level of 0.05 to assess significance of relationships.

Results

Table 1 describes characteristics of the immigrant population by race. Non-Hispanic Black immigrants make up the smallest percentage of foreign-born persons (8.1 %) and Hispanics the largest accounting for almost one half of the immigrant population (45.8 %). There are important differences in SRH. Compared to non-Hispanic White immigrants (13.1 %), a significantly smaller proportion of non-Hispanic Black immigrants (8.8 %) and Asian immigrants (10.9 %) have poor SRH, while a greater proportion of Hispanic immigrants have poor SRH (14.7 %).

Table 1 Descriptive statistics of U.S. immigrants by race
Full size table

There are also important differences in demographic and socio-economic characteristics. Hispanic immigrants are less likely than non-Hispanic White immigrants to be women. Non-Hispanic White immigrants are also significantly older than each of the other immigrant groups. Black and Hispanic immigrants are more likely to be socio-economically disadvantaged than their non-Hispanic White counterparts. For example, about four times as many Hispanic immigrants have less than a high school education compared to non-Hispanic White immigrants. Both Black and Hispanic immigrants are less likely to have a college degree, more likely to be unemployed or living in poverty, and are more likely to be uninsured compared to non-Hispanic White immigrants. Relative to Whites, the only area where Asian immigrants show greater disadvantage is likelihood of living in poverty. The majority of immigrants who took part or all of the survey in Spanish are Hispanic immigrants. Finally, White immigrants have spent longer time in the U.S., and a greater proportion of them are naturalized U.S. citizens compared to the other groups. Although approximately two thirds of Hispanic immigrants are not naturalized U.S. citizens, more than one half of foreign-born Hispanics have lived in the USA for at least 15 years.

Results from logistic regressionsFootnote 1 are presented in Table 2. The unadjusted model shows that non-Hispanic Black immigrants and Asian immigrants had lower odds of rating their health as poor compared to non-Hispanic White immigrants (OR = 0.69, CI = 0.54–0.87 and OR = 0.82 CI = 0.70–0.96, respectively). However, Hispanic immigrants had higher odds of poor SRH than their non-Hispanic White counterparts (OR = 1.25, CI = 1.09–1.43). Once socio-demographic characteristics were controlled in Model 2, the odds of poor SRH became similar for Asians and non-Hispanic Whites, and for Hispanics and non-Hispanic Whites. Age accounted for the White-Asian difference in SRH as the odds of poor SRH for Asians became similar to non-Hispanic Whites with the addition age to the model. Controlling for education and family income accounted for the White-Hispanic difference in SRH. However, holding demographic and socio-economic indicators constant across immigrant groups did not eliminate the Black-White difference. Non-Hispanic Blacks still fared better and had significantly lower odds of poor SRH compared to non-Hispanic Whites (OR = 0.76, CI = 0.58–0.98).

Table 2 Logistic regressions of poor SRH on race, demographics, socio-economic characteristics, health insurance, U.S. citizenship, and length of stay in the USA
Full size table

In the third model, after language, U.S. citizenship and length of stay were controlled; Blacks and Whites had similar odds of poor SRH. Persons who completed the interview survey in Spanish only, or in a combination of English and Spanish had slightly higher odds of poor SRH than respondents who completed the survey entirely in English (OR = 1.11, CI = 1.01–1.49). While citizenship status did not have a significant effect on SRH; immigrants who had lived in the USA for 15 years or longer had higher odds of poor SRH than immigrants who had lived in the USA for less than 5 years (OR = 1.58, CI = 1.23–2.02). Although controlling for length of time in the USA reduced the odds of reporting poor SRH for Blacks compared to Whites to non-significance in Model 3, the size of the effect (OR = 0.81, CI = 0.70–1.08) is not substantially different than that found in Model 2 (OR = 0.76, CI = 0.58–0.98). Finally, we assessed whether the association between race/ethnicity and SRH varied by length of stay by interacting race/ethnicity and length of stay; results were not significant (not tabled).

Discussion

The first aim of this study was to describe racial differences in SRH among immigrants. Before adjusting for employment, education, and income, Hispanic immigrants appeared to have worse SRH than non-Hispanic White immigrants, while non-Hispanic Black and Asian immigrants appeared to have better SRH compared to White immigrants. However, after including demographic and socio-economic factors in the model (addressing Aim 2), we find that lower levels of education and family income explained poorer SRH among Hispanic immigrants and that younger age explained better SRH among Asian immigrants. Socio-economic status did not explain better SRH among non-Hispanic Black immigrants compared to their White counterparts; indeed, given their greater disadvantage, we would expect that they would have worse SRH than non-Hispanic Whites. The Black advantage in SRH was reduced to non-significance when time in the USA was added to the model (Aim 3). However, the interaction of time in country with Black race was not significant, indicating that the SRH of Black immigrants did not decrease significantly with time in the USA. Taken together, these findings suggest that although the SRH of all immigrants decreases with time in country, there are no racial/ethnic differences in this trend.

For all immigrants, having lived in the USA for 15 years or longer was associated with higher odds of poor self-rated health—a finding consistent with the body of research suggesting that more time in the USA is associated with poorer health. Although time in the USA is not synonymous with acculturation, prior research suggests that longer time in the USA facilitates the adoption of U.S. norms and ways of life and the loss of cultural aspects that may have protective effects on health [37, 38]. Acculturation may therefore explain greater odds of poor SRH among immigrants who have resided in the USA for 15 years or longer.

The Hispanic paradox—findings of similar or better health outcomes among Hispanics compared to Whites despite the lower socio-economic status of Hispanics in the USA—has been consistently documented [39–41]. However, SRH is an exception to this paradox because SRH for Latinos has been constantly poorer than for Whites [36, 42, 43]. Evidence from this study supports previous findings that SRH is an exception to the Hispanic paradox. Hispanic immigrants were more socio-economically disadvantaged and still had poorer SRH than Whites in the unadjusted model. However, controlling for income and education differences eliminated this disparity. Therefore, poor socio-economic conditions have strong negative effects on the SRH of Hispanic immigrants.

Findings of similar SRH among non-Hispanic Black compared to White immigrants are contrary to general observations based on research that focuses on the U.S.-born population where the greatest heath disparities are between Blacks and Whites with Blacks having poor SRH, worse health outcomes, and experiencing excess morbidity and mortality [16, 44, 45]. Researchers have argued that racial differences in socio-economic status (SES) are fundamentally responsible for the racial disparities in health that are observed in the U.S. population [46, 47]. We demonstrate that despite the fact that non-Hispanic Black immigrants face greater socio-economic disadvantage than their White counterparts, they have similar SRH.

One possible explanation for the finding of no difference in SRH between non-Hispanic Black and non-Hispanic White immigrants may be a function of health selection. The majority (89.5 %) of Black immigrants in our sample are from the African continent, Central America, and South America. Studies show that immigration to the USA from these regions, compared to other predominantly White regions, is significantly driven by poverty, war, economic depression, political instability, repressive governments, and unemployment caused by the breakdown of the post-colonial economy [48–51]. Relative to White immigrants, Black immigrants might need to be healthier to initiate and complete the migration process. Contrary to this hypothesis, however, is previous work suggesting that health selection of migrants from Africa and from Western European countries and other countries where immigrants are more likely to be White tends to be similar [37, 52].

A second potential explanation is that there might be differences in the reference for social comparison of health among Black and White immigrants, and these may influence ratings of subjective assessments of health [53]. Generally, immigrants are more likely to compare their achievements to persons left behind in sending countries than to persons of similar or other races in the USA [54–58]. At the same level of objective health, Blacks who have migrated from more disadvantaged countries than Whites may be more likely to rate their health positively. More research is needed to explore how the choice of reference for rating current overall health might change with time in the USA.

Limitations

This study should be considered in light of important limitations. First, it was impossible to obtain means of entry and reason for immigration to the USA. These are measures that would have enriched the analyses given that there is a lot of variation among immigrants of the same race and ethnicity. Second, percentage of life in the USA would have been a better estimate for the level of integration in the U.S. Society. The lack of age at immigration or a continuous variable for time in the USA is a major limitation of the public-use data for this study. Third, this paper only considers one measure of health status—self-rated health. Whether these racial and ethnic differences are consistent with other measures of health remains to be determined. Finally, immigrants were grouped together based on racial and ethnic identification. Other research suggests that there may be important differences in health status of persons who identify of the same race depending on their country of origin [13].

Implication for Health Disparities

Findings from this study have implications for health disparities research and for addressing gaps in health outcomes in the USA. Much of the health disparities literature group persons into broad racial/ethnic categories, such as White, Black, or Hispanic, and does not consider immigrant status. Our research adds to the growing body of literature that emphasizes that simply examining differences among race/ethnic categories is not enough. Indeed, the Black-White health disparities that have received so much policy attention do not play out when we examine self-assessed health among immigrants. We find that race is not associated with SRH among immigrants. The reasons why non-Hispanic Black immigrants have similar SRH to non-Hispanic White immigrants, despite greater social disadvantage, warrant further attention.