[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Previous Article in Journal
Bayesian Evidence Synthesis to Infer Unobserved Population Dynamics: An Application to International Migration Into the United States, 2000–2019
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Populations
Volume 1
Issue 1
10.3390/populations1010004
populations-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Does Genetic Predisposition Explain the “Immigrant Health Paradox”? Evidence for Non-Hispanic White Older Adults in the U.S.

by
Zoya Gubernskaya
1,* and
Dalton Conley
2
1
Department of Sociology, University at Albany, Albany, NY 12222, USA
2
Department of Sociology, Princeton University, Princeton, NJ 08544, USA
*
Author to whom correspondence should be addressed.
Populations 2025, 1(1), 4; https://doi.org/10.3390/populations1010004
Submission received: 18 October 2024 / Revised: 30 December 2024 / Accepted: 15 January 2025 / Published: 29 January 2025
Browse Figures
Figure 1
<p>Unstandardized coefficients from the OLS regression models predicting polygenic indices (PGIs) for selected traits.</p> "> Figure A1
<p>Relative Risk Ratio coefficients for being foreign-born (vs. U.S.-born) from the multinominal logistic regression models predicting selected traits. Model 1 includes age, age squared, sex, region, and year of survey; Model 2: Model 1 + PGI and 10 principal component variables; Model 3: Model 2 + education (in years).</p> ">
Versions Notes

Abstract

:
This study uses data from the 2006–2012 Health and Retirement Study (HRS) genetic sample (N = 11,667) to explore the “immigrant health paradox” from a novel perspective by examining the nativity differences in genetic predisposition to health-related outcomes. Polygenic indices (PGIs) were used to evaluate whether older non-Hispanic white foreign-born individuals have genotypes that predispose them to better health profiles compared to their U.S.-born counterparts. The results show an immigrant advantage with respect to genetic predisposition to cognitive function, BMI, and smoking frequency. There are no significant differences in genetic predisposition to height, smoking initiation, and depression. Including respective PGIs in multinomial regression models partially explains an immigrant advantage with respect to cognitive function and obesity. The findings are consistent with the “healthy immigrant effect” or selective migration of individuals with a favorable genetic predisposition to health as one of the explanations of the immigrant health paradox.

1. Introduction

With 281 million people currently living outside of their respective countries of birth, immigration has become a global phenomenon [1]. Immigrants’ health is of particular interest to researchers and policymakers [2]. Previous research has produced compelling evidence of an “immigrant health paradox” with respect to mortality: the foreign-born typically enjoy longer life expectancy than the native-born population despite their relatively disadvantaged socioeconomic status (e.g., [3,4]). This pattern of health disparities was first documented for Hispanics in the U.S. [5], but similar findings have also been reported for non-Hispanic white and Black foreign-born individuals in the U.S. (e.g., [6,7,8,9]) and immigrants in other countries (e.g., [10,11,12,13]).
Even though the “immigrant health paradox” with respect to mortality is well documented, it is difficult to resolve between multiple causal stories. First, missing place-of-birth information, misclassified race/ethnicity, and misreported age could bias downward mortality estimates among immigrants (e.g., [14]). Second, selective out-migration of less healthy immigrants, or “salmon bias,” leads to overly optimistic estimates of health and mortality of the foreign-born in host countries (e.g., [15,16,17]). Third, the “healthy immigrant effect” posits positive health selection of incoming migrants: the foreign-born would be healthier than an average person in both sending and receiving countries if healthier people are more likely to migrate [18,19,20,21,22]. Migration to the U.S. is not easy due to geographic separation from other continents and restrictive immigration laws, limiting it to those who are willing and able to overcome these barriers. In addition, since the late 19th century, legal permanent migrants have been subject to medical screening [23], but relatively few were denied entry in the early 20th century [24], and medically based rejections are even rarer today. Finally, migrant culture—healthier diets, social support from family and kin, and risk-avoidant behavior—is protective of health and may help counterbalance negative influences, such as poor working conditions, lack of access to health care, and discrimination (e.g., [25,26]). The above explanations are not mutually exclusive, and each has received some empirical support, but none explains the immigrant health advantage entirely.
The findings on health outcomes other than mortality are more mixed, especially among older foreign-born individuals. For example, in the U.S., older immigrants report worse self-rated health [27,28] and have higher rates of disability than their native-born counterparts [29] but have lower rates of obesity, smoking, and alcohol consumption (e.g., [25,30]). The inconsistent results could be due to an imperfect relationship between mortality and morbidity and/or additional methodological challenges in measuring health, such as reporting biases and changes in health over the life course. Nevertheless, with respect to some health-related indicators, such as BMI and current smoking, immigrants’ advantage is well documented [26,29,31,32,33] and, similar to the immigrant health paradox in mortality, is subject to multiple causal interpretations.
In this area of research, genetic data provide a unique opportunity to test whether immigrants tend to have better health endowments compared to non-immigrants, most likely due to the “healthy immigrant effect” or positive selective migration. Genetic factors predict longevity, physical and mental health (e.g., [34,35,36]), health risk behaviors such as smoking and obesity (e.g., [37,38,39,40]), and social determinants of health such as education (e.g., [41,42,43,44,45]). Methodologically, genotypes, which are fixed at conception, have unique advantages over survey data: they do not vary over time, are not affected by the environment, and are not subject to recall or other biases. Thus, genetic data can be used to test whether the nativity differences in health are explained, at least partially, by differences in genetic predisposition to health and/or to the social determinants of health.
We use one of the few nationally representative datasets containing demographic, health, and genetic information, the Health and Retirement Study (HRS), to explore whether there is an immigrant health advantage with respect to a genetic predisposition to height, BMI, smoking, cognitive function, and depressive symptoms among older non-Hispanic white adults in the U.S. We focus on these outcomes for three reasons: (1) their well-documented association with physical health and mortality (e.g., [46,47]); (2) the high predictive validity of genetic measures and polygenic indices (PGIs) (e.g., [48]); (3) the availability of both a PGI and a matching observed phenotype in the data. We restrict our sample to non-Hispanic white adults due to the limitations of PGIs explained in the following sections.
Height is inversely associated with overall mortality, lung and heart disease mortality, and some cancers [49,50,51,52]. As twin and adoption studies show, stature is highly heritable [53], but it is also often considered a proxy for adverse in utero and childhood conditions that can suppress genetic endowment. At the individual level, given the survival, adverse pre- and post-natal conditions such as malnutrition and infections lead to a shorter stature and elevated mortality risk in adulthood and later life [46,54]. Previous research found that migrants are, on average, taller than non-migrants in their home countries [9,26]. However, at least in the case of Mexicans in the U.S., immigrants are not necessarily taller than their U.S.-born counterparts [20].
Body mass index (BMI) is often used as a proxy for health behaviors such as diet and physical activity, and high BMI is a well-documented risk factor for multiple health conditions such as heart disease, stroke, sleep apnea, and diabetes. The heritability of BMI estimated in twins and adoption studies, although not as high as for height, is also substantial [53]. Immigrants typically have lower average BMI compared to their U.S.-born counterparts [26], but some studies find that immigrant advantage in BMI tends to deteriorate with increased duration of stay in the host country [25]. Others argue that nativity differences may also depend on the birth cohort and period of migration, reflecting increasing obesity rates in immigrant origin countries [29].
Smoking is one of the most important health risk behaviors with well-documented genetic roots [39,40,53]. Research specifically points to smoking as one of the strongest predictors of mortality, and a possible explanation of the immigrant health advantage in mortality [9,30,32,33]. Immigrants are less likely to smoke than U.S.-born individuals or non-immigrants in their home countries [9,26,30,31]. However, the immigrant advantage is typically larger in comparison to non-migrants in their home countries, which is not surprising given that the smoking rates are higher in many immigrant origin countries than in the U.S., especially among males [31,55].
There are often mixed findings regarding immigrants’ health (dis)advantage with respect to cognitive function, in part because results can be sensitive to sample selection (e.g., by race and ethnicity, gender, and cohort), measurement, and statistical models’ specification. Several studies find that foreign-born individuals have higher rates of cognitive impairments than their native-born counterparts [56,57,58,59], while others report no differences or find an immigrant advantage once other covariates—especially education—are taken into account [59,60,61]. Previous studies link immigrant cognitive health (dis)advantage to differential selective migration [62] and bilingualism [63,64] and conclude that the inconsistent findings can reflect large within-group heterogeneity.
Research on mental health in general and depressive symptoms specifically among immigrant populations is also often mixed (e.g., [65]). Some studies find a higher prevalence of depression among foreign-born individuals, especially at older ages (e.g., [66,67]). However, other studies find that the prevalence of major depression and other mood and anxiety disorders is higher among U.S.-born individuals compared to their foreign-born counterparts [68].
Recent studies show that most complex human traits are highly polygenic, e.g., affected by multiple variants in many genes [69]. These discoveries led to the development of measures of individual genetic predisposition—polygenic indices (PGIs), also known as polygenic risk scores. PGIs are summary measures of genetic predisposition to a disease or trait (phenotype). The PGI for a given person is a weighted sum of the number of alleles associated with a phenotype. The phenotype-specific alleles are identified, and respective weights are calculated and reported in the previously published genome-wide association studies (GWASs) conducted in large independent samples. PGIs typically have better predictive accuracy than any single candidate gene and therefore are increasingly widely used in social science genomics research [48,69,70]. Another advantage of PGIs is that after they are standardized using the sample mean and standard deviation, they can be treated as continuous variables and easily incorporated into conventional statistical models.
PGIs also have important limitations. First, because a single mutation typically has a small effect on a trait, very large samples are needed to detect these effects in GWASs. Additionally, most genotyping techniques measure only common variants rather than the entire genome and rely on imputation methods (of other common variants) to correct for missing data. As a result, even the most recent GWASs and PGIs capture only a fraction of the putative genetic effects on the outcomes of interest. Second, because populations differ in terms of the presence and frequency of genetic variants, the PGIs are sensitive to confounding by genetic ancestry. Differences in the underlying genetic structure of the groups may be correlated with certain traits yet reflect environmental differences. Despite the recent increase in the availability of genetic data on non-white populations, the currently available large GWASs were conducted predominantly on individuals of European descent. Consequently, the PGIs derived from these GWASs are the most predictive for non-Hispanic whites and perform poorly in samples of African Americans or admixed populations, such as Hispanics [48]. Because of this last limitation, our analyses focus on non-Hispanic whites.
Based on the previous research on immigrant health advantage and theoretical argument about positive immigrant selectivity, we ask the following research question:
R1: Do immigrants have genotypes that, on average, predispose them to better health-related outcomes?
We expect the following:
Hypothesis 1: 
Compared to US-born older adults, foreign-born individuals will have higher mean PGIs for health-promoting characteristics (height and cognitive function) and lower mean PGIs for health-risk characteristics (BMI, smoking initiation and frequency, and depressive symptoms).
If Hypothesis 1 is supported for certain PGIs, then we ask the next logical question:
R2: Does accounting for genetic predisposition to certain health risk characteristics help explain the nativity differences with respect to these outcomes?
Hypothesis 2: 
Accounting for related PGIs will reduce the size of the nativity differences (i.e., between immigrants and natives) in observed health-related outcomes.
If Hypothesis 1 is not supported for certain PGIs, then accounting for genetic predisposition should not significantly change the size of the nativity differences in these health outcomes.

2. Materials and Methods

We use data from the 2006–2012 Health and Retirement Study (HRS) genetic sample [71]. The HRS is a longitudinal population-representative study of the U.S. population aged 50 and over [72]. The data have been collected every two years since 1992 and include multiple indicators of health, along with many demographic, socio-economic, and family characteristics. In 2006, HRS initiated an enhanced face-to-face interview (EFTF) that included, among other things, the collection of blood and saliva samples [73] that were subsequently genotyped. The most recent release of public-use data in November 2020 included polygenic indices (PGIs) for various phenotypes created by the staff researchers [74]. This dataset was merged with the main 2006–2012 HRS data [75] to obtain demographic and health-related measures.
The current research focuses on older adults of European ancestry. The HRS staff performed principal component analysis (PCA) of the entire sample to identify genetic ancestry sub-populations and outliers. The final European American sample included all self-reported non-Hispanic whites that had PC loadings within ± one standard deviation of the mean for eigenvectors 1 and 2 in the analysis of all unrelated study subjects [74]. Once the ancestry-specific analysis sample was identified (n = 12,090), PCA was run again within the sample to create eigenvectors to use as covariates in the statistical models to adjust for possible population stratification. From this sample, 7 respondents were excluded because of the missing information about their place of birth. After removing respondents younger than 50 as they are not a target population for the HRS, the final sample consisted of 11,667 individuals, of which 414 (3.55%) were foreign-born. The effective sample sizes in some models are slightly smaller because of the missing data on smoking (0.48%), BMI (0.32%), cognitive score (0.21%), and depressive symptoms (0.21%).
Individual genetic predisposition is determined by PGIs. The PGIs selected for this study were constructed by the HRS researchers (see [74] for detailed methodology) based on the recently published GWASs for height and BMI [76], smoking [77], general cognition [78], and major depression [79].
The foreign-born dummy distinguishes between those who were born in the U.S. (reference) and those who were born outside of the U.S. We control for basic demographic and technical characteristics such as age (in years), sex, and region of residence, as well as age at first interview in the HRS panel and year of the survey.
The health-related dependent variables are height, BMI, smoking, cognitive score, and depressive symptoms. Height is measured in meters. BMI is calculated as weight in kilograms divided by height in meters squared. We also use the categorical measure of BMI: normal or low weight (BMI < 25), overweight (BMI 25–29.9), and obese (BMI ≥ 30). In the HRS, those who reported smoking at least 100 cigarettes during their lifetime were coded as “ever smoked”. Those who reported smoking at the time of the interview were coded as “current smokers”. We distinguish between current smokers, former smokers (reference), and those who never smoked. The cognitive score is based on the 20-word list recall scale: 8 or fewer words, 9–12 words (reference), and 13 or more words (which roughly corresponds to comparing the lower and upper quartiles to the middle 50% of the distribution). Unfortunately, other measures of cognitive function available in the HRS were not asked in all years, which resulted in a significant sample reduction. Depressive symptoms are a sum of the Center for Epidemiologic Studies Depression Scale (CES-D 8) items, and we treat it as a continuous and a categorical variable: 0 symptoms (reference); 1–2 symptoms, and 3+ symptoms. All measures used in the analyses were taken from the wave in which a respondent participated in the in-person interview, either 2006, 2008, 2010, or 2012.
First, we describe the nativity differences in the main independent variables based on the sample statistics. To test Hypothesis 1, we run OLS regressions on each PGI and report statistically significant coefficients for the foreign-born indicator using conventional p-value thresholds. To account for possible unobservable differences in genetic ancestry, we include the ten principal components from the sample-specific PCA analysis as controls.
To test Hypothesis 2, we construct a series of multinomial logistic regressions to model the relationship between PGIs, nativity, and health outcomes. We focus on the categorical dependent variables because the nativity differences in many of our measures are at the extremes of the distribution. Model 1 is the baseline and includes a foreign-born indicator, age, age squared, sex, region of residence, age at first interview in the HRS panel, and interview year. Model 2 adds PGIs for respective traits along with the principal component variables. We use the KHB method [80] to compare the coefficients across the nested multinomial regression models. Model 3 adds educational attainment, which is one of the strongest predictors of health, to assess how education affects nativity differences when the genetic predisposition is taken into account.

3. Results

3.1. Descriptive Statistics

The descriptive health profile of the older foreign-born individuals is presented in Table 1. The immigrants in the sample are slightly older (69.53 vs. 67.32) and slightly shorter than their U.S.-born counterparts (1.68 vs. 1.69 m). Consistent with the previous research, they also have lower mean BMI (26.88 vs. 28.05), primarily because they are less likely to be obese (23.4% vs. 30.9%). The older immigrants are less likely to smoke currently (9.2% vs. 13.7%) but are more likely to be former smokers (51.8% vs. 43.3%) than the U.S.-born individuals. There are no significant differences in the mean number of years of education (13.6 vs. 13.26). However, the mean cognitive score is lower for immigrants than for natives (9.67 vs. 10.21), which is primarily due to the higher proportion of those with low cognitive scores among the foreign-born individuals (24.4% vs. 19.1%). Nativity differences in the likelihood of lifetime smoking and the mean number of depressive symptoms are not statistically significant. Compared to their U.S.-born counterparts, older foreign-born individuals are more likely to reside in the Northeast and the West and less likely to reside in the Midwest or the South.

3.2. Nativity Differences in Genetic Predisposition

Figure 1 presents the foreign-born coefficients from the OLS regression models predicting PGIs (the models are presented in Appendix A Table A1). As the PGI variables are standardized, the coefficients can be interpreted as the difference between the reference and the observed category in standard deviations. Consistent with Hypothesis 1, foreign-born individuals have higher mean cognitive function PGIs, lower mean smoking frequency PGIs, and lower mean BMI PGIs. The nativity differences in PGIs for height, smoking initiation, and depression are not statistically significant.

3.3. Genetic Predisposition and Nativity Differences in Health-Related Outcomes

Next, we explore how accounting for genetic predisposition affects nativity differences in phenotypical characteristics. We only present results for the outcomes for which Hypothesis 1 was supported: cognitive function, BMI, and smoking. (The results for the other measures are presented in Appendix A Table A3).
Table 2 presents the coefficients from the multinomial regression models predicting cognitive function. The predicted Relative Risk Ratios for the foreign-born vs. U.S.-born individuals are also presented in Appendix A Figure A1. Compared to their U.S.-born counterparts, the foreign-born individuals are more likely to exhibit low rather than average cognitive function, although these differences are not significant (Model 1). Accounting for genetic predisposition to higher cognition (Model 2) increases the size of the foreign-born coefficient with respect to low cognition and reduces it with respect to high cognition, although the KHB test is statistically significant only for the latter. Thus, Hypothesis 2 is partially supported for the cognitive score. Model 3 shows that education is strongly predictive of cognitive function in later life, but genetic predisposition remains statistically significant even after accounting for educational attainment.
When it comes to BMI (Table 3), the immigrant health advantage is statistically significant only with respect to obesity (Model 1). Even though the BMI PGIs added in Model 2 is highly predictive of being both overweight and obese, and accounting for genetic predisposition reduces the size of the nativity coefficients for both categories, the difference is marginally statistically significant only for obesity (vs. BMI < 25). Education is strongly negatively associated with being overweight and obese, but it explains neither the nativity differences nor the effect of the genetic predisposition on BMI (Model 3).
The results for smoking presented in Table 4 show that compared to U.S.-born older adults, their foreign-born counterparts are both significantly less likely to never smoke and to smoke currently than to be former smokers (Model 1). Genetic factors work to increase the immigrant disadvantage with respect to never smoking and reduce the immigrant advantage with respect to current smoking (Model 2), but these effects are relatively small and not statistically significant. Thus, Hypothesis 2 is not supported for smoking.

3.4. Robustness Checks

One of the major concerns when assessing differences in genetic predisposition to a trait between groups is that these differences may be driven by ancestral heterogeneity—that is, differences in the underlying population genetic structure of the groups that may be correlated with certain traits yet reflect environmental differences. The common approach that we followed is to restrict the analyses to a relatively homogeneous population (e.g., European ancestry) and to include principal component variables that capture the overall genetic structure of the population in statistical models. Because of the history of migration to the U.S., many non-Hispanic white U.S.-born and foreign-born older adults share common European ancestry. The top countries of origin of the non-Hispanic foreign-born individuals in the HRS are Canada, Germany/Austria, Great Britain, and Italy. Unfortunately, the HRS does not have a question about the ethnic ancestry of U.S.-born individuals, but the top first reported ancestry categories for the U.S.-born non-Hispanic whites aged 50 and over in the 2008 American Community Survey are German, English, Irish, and Italian (authors’ calculations). Given that the HRS is representative of the U.S. population, there are many similarities in ethnic ancestry of the foreign-born and U.S.-born older non-Hispanic white adults. Possible genetic outliers with respect to ancestry were identified by the PCA analyses and removed from the sample.
To confirm our findings about the nativity differences in PGIs, we also used propensity score matching (Appendix A Table A2). We match foreign- and US-born individuals on the first five principal component variables using one-to-one (upper panel) and one-to-six (lower panel) nearest-neighbor matching algorithms. The coefficients are average treatment effects (ATEs) of being foreign-born on genetic predisposition to a trait. The strongest and most statistically significant effects are observed for cognitive function and smoking frequency. All other coefficients are not statistically significant using standard p-value thresholds.

4. Discussion

This research offers a novel examination of the “immigrant health paradox” by exploring nativity differences in genetic predisposition to health outcomes among older non-Hispanic white adults in the 2006–2012 Health and Retirement Study genetic sample. The results generally support the “healthy immigrant effect” or the idea that immigrants are positively selected on health. We found a statistically significant immigrant advantage in polygenic indices for cognitive function, BMI, and smoking frequency. These results are consistent with many studies that find immigrant health advantage in actual BMI and current smoking [9,26,30,31,32,33]. The differences in PGIs for height, smoking initiation, and depression are not statistically significant, which is also not surprising given that the findings about the nativity differences in these outcomes are more mixed. Noteworthy is that none of the examined PGIs revealed a significant immigrant disadvantage.
We also found that genetic factors partially explain nativity differences in cognitive function and BMI. Taking into account genetic predisposition for high cognition explains the foreign-born advantage (albeit a small one) with respect to high vs. average cognitive score. Genetic predisposition also partially explains the immigrant health advantage with respect to obesity. However, the genetic influences on the immigrant health advantage in smoking are relatively small and not statistically significant. As we did not find any significant nativity differences in genetic predisposition to height and depression, genetic factors seem to play a minor role in explaining the nativity differences in these characteristics (which are also often inconsistent).
The results have important implications. First, accounting for genetic predisposition could potentially explain, at least partially, “unexplained” nativity differences in observed health outcomes if genetic data were available in other datasets. Second, even though genetic predisposition is strongly associated with observed traits and health behaviors, genetic factors explain only a small proportion of the nativity differences in health-related outcomes. Nor do they explain away the effects of social factors, such as education, which remained highly statistically significant in models with controls for genetic predisposition. If anything, accounting for genetic predisposition should produce a more accurate estimate of environmental factors. Finally, when it comes to possible explanations of the immigrant health paradox, our findings support the “healthy immigrant effect” or selective migration of individuals with a favorable genetic predisposition to health. The true test of the “healthy immigrant effect” would require comparing genotypes of immigrants and non-immigrants in origin countries, and we are not aware of any large datasets that would allow such a test (but see [41] for the case of New Zealand out-migration). But because genotypes are fixed at conception, unaffected by the environment, and do not change over time, selective migration is the most likely explanation for the nativity differences in genetic predisposition we documented in this study.
This research has several limitations. First, the sample size of the foreign-born individuals in the HRS genetic sample is relatively small, which means that the current research is adequately powered to detect only relatively large effects. Second, we relied on the traditional PGI estimates that may be confounded by parental genetic nurture and by other environmental effects that are associated with the PGIs even when controlling for principal components of the genetic data. The ideal PGI is constructed from within family genome-wide association studies, where the variation in genotype is entirely random and conditional on parental genotypes and thus directly causal. However, such analyses are scarce due to data limitations. While the traditional approach is not ideal, it is still the case that this PGI has the desired attribute of being fixed at conception and unchangeable, and thus whatever environmental confounders that it may pick up are pre-migration and are thus still part of the selective migration that results in the “healthy immigrant effect”. Third, there is significant heterogeneity within the older immigrant population by reasons for migration, birth cohort, age and period of arrival, and other characteristics that are linked to selective migration. Unfortunately, we could not explore these differences due to the small sample size of the foreign-born individuals. Finally, the analyses are focused only on non-Hispanic white older adults in the U.S. and should not be extrapolated to all immigrants in the U.S. or immigrants elsewhere. Younger cohorts of immigrants in the U.S. or immigrant populations in other countries may be different from the HRS respondents in terms of their genetic predisposition, genetic ancestry (countries of origin), and health outcomes. Also, many immigrants in the U.S. are not white. Despite the recent increase in the availability of genetic data on non-white populations, the currently available PGIs remain the most predictive for respondents of European ancestry. Although this is clearly a limitation that should be addressed in the future once better data become available, non-Hispanic whites currently constitute about 85% of U.S.-born and about one-third of foreign-born adults aged 50 and over in the U.S.
These concerns aside, the results from this research provide the first empirical evidence in support of the “immigrant health paradox” with respect to genetic predisposition. Future studies should address the limitations of the current research and use other data sources to assess the immigrant (dis)advantage in genetic risk profiles and to test whether nativity differences in health outcomes can be explained by genetic factors.

Author Contributions

Conceptualization, Z.G. and D.C.; methodology, Z.G. and D.C.; formal analysis, Z.G.; data curation, Z.G.; writing—original draft preparation, Z.G.; writing—review and editing, Z.G. and D.C.; visualization, Z.G.; funding acquisition, Z.G. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Russell Sage Foundation grant “Genetic Influences on Immigrant Health in Midlife and Older Age” (#G-1810-08943). Z.G. benefited from the training course at the University of Michigan (NIA R25AG053227), the RSF 2019 Summer Institute in Social Genomics (#G-1902-12533), and ongoing support of the Center for Demographic and Social Analysis (CSDA) and Center for Aging and Policy Studies (CAPS, P30AG066583). D.C.: Funding from NJ ACTS: A Platform for Translational Science in New Jersey (NCATS 1UM1TR004789-01).

Institutional Review Board Statement

Ethical review and approval were waived for this study, due to this work uses secondary analyses of publicly available datasets, in which the data are stripped of identifying information. The authors did not use any sensitive or restricted data, and the data access required a simple registration to ensure responsible use.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

We are grateful to Seulki Kim for excellent research assistance and to the anonymous reviewers for their insightful comments and helpful suggestions.

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix A

Table A1. Unstandardized coefficients from the OLS regression models predicting polygenic indices (PGIs) for selected traits.
Table A1. Unstandardized coefficients from the OLS regression models predicting polygenic indices (PGIs) for selected traits.
VariablesHeightBMIEver SmokedCigarettes per DayCognitive FunctionMajor Depression
(Ref. US-born)
Foreign-born0.047−0.078 †−0.039−0.109 *0.134 **0.026
(0.046)(0.045)(0.046)(0.049)(0.045)(0.046)
10 PCsyesyesyesyesyesyes
Constant−0.0030.000−0.001−0.002−0.001−0.002
(0.009)(0.008)(0.009)(0.009)(0.009)(0.009)
N11,66711,66711,66711,66711,66711,667
R-squared0.1790.1920.1430.0410.1840.152
** p < 0.01, * p < 0.05, † p < 0.1. Standard errors in parentheses.
Table A2. Average treatment effects of nativity on PGIs from the propensity score matching (on the first 5 principal components) models: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
Table A2. Average treatment effects of nativity on PGIs from the propensity score matching (on the first 5 principal components) models: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
VariablesHeightBMIEver SmokedCigarettes per DayCognitive FunctionMajor Depression
Panel A: One-to-one matching
(Ref. US-born)
Foreign-born0.023−0.082−0.077−0.166 **0.198 ***−0.011
(0.070)(0.061)(0.055)(0.064)(0.060)(0.051)
Panel B: One-to-six matching
(Ref. US-born)
Foreign-born0.028−0.042−0.056−0.113 *0.193 ***−0.008
(0.054)(0.053)(0.053)(0.055)(0.053)(0.045)
N11,66711,66711,66711,66711,66711,667
N1 (foreign-born)414414414414414414
N0 (US-born)11,25311,25311,25311,25311,25311,253
*** p < 0.001, ** p < 0.01, * p < 0.05. Standard errors in parentheses.
Table A3. Unstandardized coefficients from the OLS regression models predicting health-related outcomes: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
Table A3. Unstandardized coefficients from the OLS regression models predicting health-related outcomes: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
VariablesHeightCognitive ScoreBMIDepressive
Symptoms
(Ref. U.S.-born)
Foreign-born−0.007 *−0.396 **−0.638 **−0.068
(0.003)(0.146)(0.231)(0.093)
Age0.002 ***0.222 ***0.369 ***−0.157 ***
(0.001)(0.034)(0.057)(0.021)
Age squared−0.000 ***−0.002 ***−0.003 ***0.001 ***
(0.000)(0.000)(0.000)(0.000)
(Ref. Male)
Female−0.152 ***1.117 ***−0.594 ***0.296 ***
(0.001)(0.053)(0.098)(0.033)
(Ref. Northeast)
Midwest−0.001−0.1030.547 ***−0.051
(0.002)(0.084)(0.164)(0.054)
South0.004 *−0.165 *−0.1230.099
(0.002)(0.080)(0.154)(0.052)
West0.005 **0.047−0.424 *0.111
(0.002)(0.091)(0.168)(0.059)
PGI (respective)0.031 ***0.384 ***1.792 ***0.165 ***
(0.001)(0.030)(0.056)(0.019)
10 PCsYesYesYesYes
Education (yrs)-0.319 ***−0.142 ***−0.120 ***
(0.011)(0.021)(0.007)
Constant1.735 ***3.267 *21.089 ***8.043 ***
(0.023)(1.280)(2.049)(0.725)
N11,66411,63911,62711,638
R-squared0.6590.2940.1410.052
*** p < 0.001, ** p < 0.01, * p < 0.05. Robust standard errors in parentheses.
Populations 01 00004 g0a1
Figure A1. Relative Risk Ratio coefficients for being foreign-born (vs. U.S.-born) from the multinominal logistic regression models predicting selected traits. Model 1 includes age, age squared, sex, region, and year of survey; Model 2: Model 1 + PGI and 10 principal component variables; Model 3: Model 2 + education (in years).
Figure A1. Relative Risk Ratio coefficients for being foreign-born (vs. U.S.-born) from the multinominal logistic regression models predicting selected traits. Model 1 includes age, age squared, sex, region, and year of survey; Model 2: Model 1 + PGI and 10 principal component variables; Model 3: Model 2 + education (in years).
Populations 01 00004 g0a1

References

  1. McAuliffe, M.; Triandafyllidou, A. (Eds.) World Migration Report 2022; International Organization for Migration (IOM): Geneva, Switzerland, 2022. [Google Scholar]
  2. Castañeda, H.; Holmes, S.M.; Madrigal, D.S.; Young, M.-E.D.; Beyeler, N.; Quesada, J. Immigration as a Social Determinant of Health. Annu. Rev. Public Health 2015, 36, 375–392. [Google Scholar] [CrossRef] [PubMed]
  3. Borrell, L.N.; Lancet, E.A. Race/Ethnicity and All-Cause Mortality in US Adults: Revisiting the Hispanic Paradox. Am. J. Public Health 2012, 102, 836–843. [Google Scholar] [CrossRef] [PubMed]
  4. Cunningham, S.A.; Ruben, J.D.; Narayan, K.M.V. Health of Foreign-Born People in the United States: A Review. Health Place 2008, 14, 623–635. [Google Scholar] [CrossRef] [PubMed]
  5. Markides, K.S.; Coreil, J. The Health of Hispanics in the Southwestern United States: An Epidemiologic Paradox. Public Health Rep. 1986, 101, 253. [Google Scholar] [PubMed]
  6. Hamilton, T.G.; Robert, A. Hummer. Immigration and the Health of U.S. Black Adults: Does Country of Origin Matter? Soc. Sci. Med. 2011, 73, 1551–1560. [Google Scholar] [CrossRef]
  7. Mehta, N.K.; Elo, I.T.; Engelman, M.; Lauderdale, D.S.; Kestenbaum, B.M. Life Expectancy among US-Born and Foreign-Born Older Adults in the United States: Estimates from Linked Social Security and Medicare Data. Demography 2016, 53, 1109–1134. [Google Scholar] [CrossRef]
  8. Palarino, J.V. The immigrant health advantage: An examination of African-origin Black immigrants in the United States. Popul. Res. Policy Rev. 2021, 40, 5–929. [Google Scholar] [CrossRef]
  9. Riosmena, F.; Kuhn, R.; Jochem, W.C. Explaining the Immigrant Health Advantage: Self-Selection and Protection in Health-Related Factors among Five Major National-Origin Immigrant Groups in the United States. Demography 2017, 54, 175–200. [Google Scholar] [CrossRef]
  10. Ichou, M.; Wallace, M. The healthy immigrant effect: The role of educational selectivity in the good health of migrants. Demogr. Res. 2019, 40, 61–94. [Google Scholar] [CrossRef]
  11. Shor, E.; Roelfs, D.; Vang, Z.M. The ‘Hispanic Mortality Paradox’ Revisited: Meta-Analysis and Meta-Regression of Life-Course Differentials in Latin American and Caribbean Immigrants’ Mortality. Soc. Sci. Med. 2017, 186, 20–33. [Google Scholar] [CrossRef] [PubMed]
  12. Vang, Z.M.; Sigouin, J.; Flenon, A.; Gagnon, A. Are Immigrants Healthier than Native-Born Canadians? A Systematic Review of the Healthy Immigrant Effect in Canada. Ethn. Health 2017, 22, 209–241. [Google Scholar] [CrossRef] [PubMed]
  13. Wallace, M.; Kulu, H. Low Immigrant Mortality in England and Wales: A Data Artefact? Soc. Sci. Med. 2014, 120, 100–109. [Google Scholar] [CrossRef] [PubMed]
  14. Acevedo-Garcia, D.; Bates, L.M. Latino Health Paradoxes: Empirical Evidence, Explanations, Future Research, and Implications. Lat. Os United States Chang. Face Am. 2008, 101–113. [Google Scholar]
  15. Arenas, E.; Goldman, N.; Pebley, A.R.; Teruel, G. Return Migration to Mexico: Does Health Matter? Demography 2015, 52, 1853–1868. [Google Scholar] [CrossRef]
  16. Palloni, A.; Arias, E. Paradox Lost: Explaining the Hispanic Adult Mortality Advantage. Demography 2004, 41, 385–415. [Google Scholar] [CrossRef] [PubMed]
  17. Turra, C.M.; Elo, I.T. The Impact of Salmon Bias on the Hispanic Mortality Advantage: New Evidence from Social Security Data. Popul. Res. Policy Rev. 2008, 27, 515–530. [Google Scholar] [CrossRef] [PubMed]
  18. Akresh, I.R.; Frank, R. Health Selection among New Immigrants. Am. J. Public Health 2008, 98, 2058–2064. [Google Scholar] [CrossRef] [PubMed]
  19. Bostean, G. Does Selective Migration Explain the Hispanic Paradox? A Comparative Analysis of Mexicans in the US and Mexico. J. Immigr. Minor. Health 2013, 15, 624–635. [Google Scholar] [CrossRef]
  20. Crimmins, E.M.; Soldo, B.J.; Kim, J.K.; Alley, D.E. Using Anthropometric Indicators for Mexicans in the United States and Mexico to Understand the Selection of Migrants and the ‘Hispanic Paradox’. Soc. Biol. 2005, 52, 164–177. [Google Scholar] [CrossRef]
  21. Jasso, G.; Massey, D.S.; Rosenzweig, M.R.; Smith, J.P. Immigrant health: Selectivity and acculturation. In IFS Working Papers No. 04/23; Institute for Fiscal Studies (IFS): London, UK, 2004. [Google Scholar]
  22. Feliciano, C. Educational Selectivity in US Immigration: How Do Immigrants Compare to Those Left Behind? Demography 2005, 42, 131–152. [Google Scholar] [CrossRef] [PubMed]
  23. MacPherson, D.W.; Gushulak, B.D. Health screening in immigrants, refugees, and international adoptees. In The Travel and Tropical Medicine Manual, 5th ed.; Elsevier: Amsterdam, The Netherlands, 2016; pp. 260–270. [Google Scholar]
  24. Fairchild, A.L. Science at the Borders: Immigrant Medical Inspection and the Shaping of the Modern Industrial Labor Force; JHU Press: Baltimore, MD, USA, 2003. [Google Scholar]
  25. Abraído-Lanza, A.F.; Chao, M.T.; Flórez, K.R. Do Healthy Behaviors Decline with Greater Acculturation?: Implications for the Latino Mortality Paradox. Soc. Sci. Med. 2005, 61, 1243–1255. [Google Scholar] [CrossRef]
  26. Riosmena, F.; Wong, R.; Palloni, A. Migration Selection, Protection, and Acculturation in Health: A Binational Perspective on Older Adults. Demography 2013, 50, 1039–1064. [Google Scholar] [CrossRef]
  27. Angel, J.L.; Buckley, C.J.; Sakamoto, A. Duration or Disadvantage? Exploring Nativity, Ethnicity, and Health in Midlife. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 2001, 56, S275–S284. [Google Scholar] [CrossRef] [PubMed]
  28. Wakabayashi, C. Effects of Immigration and Age on Health of Older People in the United States. J. Appl. Gerontol. 2010, 29, 697–719. [Google Scholar] [CrossRef]
  29. Hayward, M.D.; Hummer, R.A.; Chiu, C.-T.; González-González, C.; Wong, R. Does the Hispanic Paradox in US Adult Mortality Extend to Disability? Popul. Res. Policy Rev. 2014, 33, 81–96. [Google Scholar] [CrossRef] [PubMed]
  30. Blue, L.; Fenelon, A. Explaining Low Mortality among US Immigrants Relative to Native-Born Americans: The Role of Smoking. Int. J. Epidemiol. 2011, 40, 786–793. [Google Scholar] [CrossRef] [PubMed]
  31. Bosdriesz, J.R.; Lichthart, N.; Witvliet, M.I.; Busschers, W.B.; Stronks, K.; Kunst, A.E. Smoking prevalence among migrants in the US compared to the US-born and the population in countries of origin. PLoS ONE 2013, 8, e58654. [Google Scholar] [CrossRef] [PubMed]
  32. Fleischer, N.L.; Ro, A.; Bostean, G. Smoking selectivity among Mexican immigrants to the United States using binational data, 1999–2012. Prev. Med. 2017, 97, 26–32. [Google Scholar] [CrossRef]
  33. Lariscy, J.T.; Hummer, R.A.; Hayward, M.D. Hispanic Older Adult Mortality in the United States: New Estimates and an Assessment of Factors Shaping the Hispanic Paradox. Demography 2015, 52, 1–14. [Google Scholar] [CrossRef] [PubMed]
  34. Ehret, G.B.; Munroe, P.B.; Rice, K.M.; Bochud, M.; Johnson, A.D.; Chasman, D.I.; Smith, A.V.; Tobin, M.D.; Verwoert, G.C.; Hwang, S.J.; et al. Genetic Variants in Novel Pathways Influence Blood Pressure and Cardiovascular Disease Risk. Nature 2011, 478, 103–109. [Google Scholar] [CrossRef]
  35. Hjelmborg, J.v.; Iachine, I.; Skytthe, A.; Vaupel, J.W.; McGue, M.; Koskenvuo, M.; Kaprio, J.; Pedersen, N.L.; Christensen, K. Genetic Influence on Human Lifespan and Longevity. Hum. Genet. 2006, 119, 312–321. [Google Scholar] [CrossRef] [PubMed]
  36. Sullivan, P.F.; Neale, M.C.; Kendler, K.S. Genetic Epidemiology of Major Depression: Review and Meta-Analysis. Am. J. Psychiatry 2000, 157, 1552–1562. [Google Scholar] [CrossRef] [PubMed]
  37. Belsky, D.W.; Moffitt, T.E.; Sugden, K.; Williams, B.; Houts, R.; McCarthy, J.; Caspi, A. Development and Evaluation of a Genetic Risk Score for Obesity. Biodemogr. Soc. Biol. 2013, 59, 85–100. [Google Scholar] [CrossRef]
  38. Boardman, J.D.; Blalock, C.L.; Pampel, F.C. Trends in the Genetic Influences on Smoking. J. Health Soc. Behav. 2010, 51, 108–123. [Google Scholar] [CrossRef]
  39. Domingue, B.W.; Conley, D.; Fletcher, J.; Boardman, J.D. Cohort Effects in the Genetic Influence on Smoking. Behav. Genet. 2016, 46, 31–42. [Google Scholar] [CrossRef] [PubMed]
  40. Furberg, H.; Kim, Y.; Dackor, J.; Boerwinkle, E.; Franceschini, N.; Ardissino, D.; Bernardinelli, L.; Mannucci, P.M.; Mauri, F.; Merlini, P.A.; et al. Genome-Wide Meta-Analyses Identify Multiple Loci Associated with Smoking Behavior. Nat. Genet. 2010, 42, 441–447. [Google Scholar] [CrossRef]
  41. Belsky, D.W.; Moffitt, T.E.; Corcoran, D.L.; Domingue, B.; Harrington, H.; Hogan, S.; Houts, R.; Ramrakha, S.; Sugden, K.; Williams, B.S. The Genetics of Success: How Single-Nucleotide Polymorphisms Associated with Educational Attainment Relate to Life-Course Development. Psychol. Sci. 2016, 27, 957–972. [Google Scholar] [CrossRef] [PubMed]
  42. Branigan, A.R.; McCallum, K.J.; Freese, J. Variation in the Heritability of Educational Attainment: An International Meta-Analysis. Soc. Forces 2013, 92, 109–140. [Google Scholar] [CrossRef]
  43. Domingue, B.W.; Belsky, D.W.; Conley, D.; Harris, K.M.; Boardman, J.D. Polygenic Influence on Educational Attainment: New Evidence from the National Longitudinal Study of Adolescent to Adult Health. AERA Open 2015, 1, 2332858415599972. [Google Scholar] [CrossRef] [PubMed]
  44. Lee, J.J.; Wedow, R.; Okbay, A.; Kong, E.; Maghzian, O.; Zacher, M.; Nguyen-Viet, T.A.; Bowers, P.; Sidorenko, J.; Linnér, R.K.; et al. Gene Discovery and Polygenic Prediction from a Genome-Wide Association Study of Educational Attainment in 1.1 Million Individuals. Nat. Genet. 2018, 50, 1112–1121. [Google Scholar] [CrossRef]
  45. Okbay, A.; Beauchamp, J.P.; Fontana, M.A.; Lee, J.J.; Pers, T.H.; Rietveld, C.A.; Turley, P.; Chen, G.B.; Emilsson, V.; Meddens, S.F.W.; et al. Genome-Wide Association Study Identifies 74 Loci Associated with Educational Attainment. Nature 2016, 533, 539–542. [Google Scholar] [CrossRef] [PubMed]
  46. Crimmins, E.M.; Finch, C.E. Infection, Inflammation, Height, and Longevity. Proc. Natl. Acad. Sci. USA 2006, 103, 498–503. [Google Scholar] [CrossRef] [PubMed]
  47. Marmot, M. Social Determinants of Health Inequalities. Lancet 2005, 365, 1099–1104. [Google Scholar] [CrossRef]
  48. Dudbridge, F. Power and Predictive Accuracy of Polygenic Risk Scores. PLoS Genet. 2013, 9, e1003348. [Google Scholar] [CrossRef]
  49. Barker, D.J.; Osmond, C.; Golding, J. Height and mortality in the counties of England and Wales. Ann. Hum. Biol. 1990, 17, 1–6. [Google Scholar] [CrossRef] [PubMed]
  50. Jousilahti, P.; Tuomilehto, J.; Vartiainen, E.; Eriksson, J.; Puska, P. Relation of adult height to cause-specific and total mortality: A prospective follow-up study of 31, 199 middle-aged men and women in Finland. Am. J. Epidemiol. 2000, 151, 1112–1120. [Google Scholar] [CrossRef]
  51. Leon, D.A.; Smith, G.D.; Shipley, M.; Strachan, D. Adult height and mortality in London: Early life, socioeconomic confounding, or shrinkage? J. Epidemiol. Community Health 1995, 49, 5–9. [Google Scholar] [CrossRef] [PubMed]
  52. Perkins, J.M.; Subramanian, S.V.; Davey Smith, G.; Özaltin, E. Adult height, nutrition, and population health. Nutr. Rev. 2016, 74, 149–165. [Google Scholar] [CrossRef] [PubMed]
  53. Polderman, T.J.; Benyamin, B.; De Leeuw, C.A.; Sullivan, P.F.; Van Bochoven, A.; Visscher, P.M.; Posthuma, D. Meta-analysis of the heritability of human traits based on fifty years of twin studies. Nat. Genet. 2015, 47, 7. [Google Scholar] [CrossRef] [PubMed]
  54. Bozzoli, C.; Deaton, A.; Quintana-Domeque, C. Adult height and childhood disease. Demography 2009, 46, 4–669. [Google Scholar]
  55. Ng, M.; Freeman, M.K.; Fleming, T.D.; Robinson, M.; Dwyer-Lindgren, L.; Thomson, B.; Lopez, A.D. Smoking prevalence and cigarette consumption in 187 countries, 1980–2012. J. Am. Med. Assoc. 2014, 311, 2–192. [Google Scholar] [CrossRef]
  56. Franco, Y.; Choi, E.Y. The relationship between immigrant status and undiagnosed dementia: The role of limited English proficiency. J. Immigr. Minor. Health 2020, 22, 914–922. [Google Scholar] [CrossRef] [PubMed]
  57. Liew, H.P. Cognitive disparities between US- and foreign-born individuals. J. Public Health 2021, 29, 1121–1134. [Google Scholar] [CrossRef]
  58. Moon, H.; Badana, A.N.; Hwang, S.Y.; Sears, J.S.; Haley, W.E. Dementia prevalence in older adults: Variation by race/ethnicity and immigrant status. Am. J. Geriatr. Psychiatry 2019, 27, 241–250. [Google Scholar]
  59. Xu, H.; Zhang, Y.; Wu, B. Association between migration and cognitive status among middle-aged and older adults: A systematic review. BMC Geriatr. 2017, 17, 184. [Google Scholar] [CrossRef] [PubMed]
  60. Hill, T.D.; Angel, J.L.; Balistreri, K.S.; Herrera, A.P. Immigrant status and cognitive functioning in late-life: An examination of gender variations in the healthy immigrant effect. Soc. Sci. Med. 2012, 75, 12–2084. [Google Scholar] [CrossRef] [PubMed]
  61. Garcia, M.A.; Ortiz, K.; Arévalo, S.P.; Diminich, E.D.; Briceño, E.; Vega, I.E.; Tarraf, W. Age of migration and cognitive function among older Latinos in the United States. J. Alzheimer’s Dis. 2020, 76, 1493–1511. [Google Scholar] [CrossRef] [PubMed]
  62. Casanova, M.; Aguila, E. Gender differences in cognitive function among older Mexican immigrants. J. Econ. Ageing 2020, 16, 100226. [Google Scholar] [CrossRef] [PubMed]
  63. Jang, Y.; Choi, E.Y.; Wu, B.; Dong, X.; Kim, M.T. Linguistic adaptation and cognitive function in older Chinese and Korean immigrants in the United States: A cross-sectional study. J. Aging Health 2022, 34, 951–960. [Google Scholar] [CrossRef] [PubMed]
  64. Padilla, C.; Mendez, M.F.; Jimenez, E.E.; Teng, E. Bilingualism in older Mexican-American immigrants is associated with higher scores on cognitive screening. BMC Geriatr. 2016, 16, 189. [Google Scholar] [CrossRef] [PubMed]
  65. Bas-Sarmiento, P.; Saucedo-Moreno, M.J.; Fernández-Gutiérrez, M.; Poza-Méndez, M. Mental health in immigrants versus native population: A systematic review of the literature. Arch. Psychiatr. Nurs. 2017, 31, 111–121. [Google Scholar] [CrossRef] [PubMed]
  66. González, H.M.; Tarraf, W.; Whitfield, K.E.; Vega, W.A. The epidemiology of major depression and ethnicity in the United States. J. Psychiatr. Res. 2010, 44, 15–1051. [Google Scholar] [CrossRef] [PubMed]
  67. Ladin, K.; Reinhold, S. Mental Health of Aging Immigrants and Native-Born Men Across 11 European Countries. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 2013, 68, 298–309. [Google Scholar] [CrossRef] [PubMed]
  68. Grant, B.F.; Stinson, F.S.; Hasin, D.S.; Dawson, D.A.; Chou, S.P.; Anderson, K. Immigration and Lifetime Prevalence of DSM-IV Psychiatric Disorders Among Mexican Americans and Non-Hispanic Whites in the United States: Results From the National Epidemiologic Survey on Alcohol and Related Conditions. Arch. Gen. Psychiatry 2004, 61, 1226–1233. [Google Scholar] [CrossRef] [PubMed]
  69. Conley, D.; Fletcher, J. The Genome Factor: What the Social Genomics Revolution Reveals About Ourselves, Our History, and the Future; Princeton University Press: Princeton, NJ, USA, 2017. [Google Scholar] [CrossRef]
  70. Belsky, D.W.; Israel, S. Integrating Genetics and Social Science: Genetic Risk Scores. Biodemogr. Soc. Biol. 2014, 60, 137–155. [Google Scholar] [CrossRef] [PubMed]
  71. Health and Retirement Study, (HRS Polygenic Scores–Release 3) Public Use Dataset. Produced and Distributed by the University of Michigan with Funding from the National Institute on Aging (Grant Number NIA U01AG009740). Ann Arbor, MI, 2020. Available online: https://hrs.isr.umich.edu/news/data-announcements/hrs-polygenic-scores-2006-2012-genetic-data-release-3 (accessed on 18 October 2024).
  72. Servais, M.A. Overview of HRS Public Data Files for Cross-Sectional and Longitudinal Analysis. 2010. Available online: http://hrsonline.isr.umich.edu/sitedocs/dmgt/OverviewofHRSPublicData.pdf (accessed on 1 August 2020).
  73. Crimmins, E.M.; Faul, J.D.; Kim, J.K.; David, R.W. Documentation of Biomarkers in the 2010 and 2012 Health and Retirement Study; University of Michigan Institute for Social Research Survey Research Center: Ann Arbor, MI, USA, 2015. [Google Scholar]
  74. Ware, E.B.; Gard, A.M.; Faul, J.D.; Schmitz, L.L. HRS Polygenic Scores—Release; University of Michigan Institute for Social Research Survey Research Center: Ann Arbor, MI, USA, 2020. [Google Scholar]
  75. Health and Retirement Study, (RAND HRS Longitudinal File 2016) Public Use Dataset. Produced and Distributed by the University of Michigan with Funding from the National Institute on Aging (Grant Number NIA U01AG009740). Ann Arbor, MI, 2018. Available online: https://hrs.isr.umich.edu/news/data-announcements/rand-hrs-2016-data-now-available (accessed on 18 October 2024).
  76. Yengo, L.; Sidorenko, J.; Kemper, K.E.; Zheng, Z.; Wood, A.R.; Weedon, M.N.; Frayling, T.M.; Hirschhorn, J.; Yang, J.; Visscher, P.M. Meta-Analysis of Genome-Wide Association Studies for Height and Body Mass Index in 700000 Individuals of European Ancestry. Hum. Mol. Genet. 2018, 27, 3641–3649. [Google Scholar] [CrossRef] [PubMed]
  77. Liu, M.; Jiang, Y.; Wedow, R.; Li, Y.; Brazel, D.M.; Chen, F.; Datta, G.; Davila-Velderrain, J.; McGuire, D.; Tian, C. Association Studies of up to 1.2 Million Individuals Yield New Insights into the Genetic Etiology of Tobacco and Alcohol Use. Nat. Genet. 2019, 51, 237–244. [Google Scholar] [CrossRef] [PubMed]
  78. Davies, G.; Lam, M.; Harris, S.E.; Trampush, J.W.; Luciano, M.; Hill, W.D.; Hagenaars, S.P.; Ritchie, S.J.; Marioni, R.E.; Fawns-Ritchie, C.; et al. Study of 300,486 Individuals Identifies 148 Independent Genetic Loci Influencing General Cognitive Function. Nat. Commun. 2018, 9, 2098. [Google Scholar] [CrossRef] [PubMed]
  79. Wray, N.R.; Ripke, S.; Mattheisen, M.; Trzaskowski, M.; Byrne, E.M.; Abdellaoui, A.; Adams, M.J.; Agerbo, E.; Air, T.M.; Andlauer, T.M.F.; et al. Genome-Wide Association Analyses Identify 44 Risk Variants and Refine the Genetic Architecture of Major Depression. Nat. Genet. 2018, 50, 668–681. [Google Scholar] [CrossRef] [PubMed]
  80. Kohler, U.; Karlson, K.B.; Holm, A. Comparing Coefficients of Nested Nonlinear Probability Models. Stata J. 2011, 11, 420–438. [Google Scholar] [CrossRef]
Populations 01 00004 g001
Figure 1. Unstandardized coefficients from the OLS regression models predicting polygenic indices (PGIs) for selected traits.
Figure 1. Unstandardized coefficients from the OLS regression models predicting polygenic indices (PGIs) for selected traits.
Populations 01 00004 g001
Table 1. Descriptive statistics: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
Table 1. Descriptive statistics: Non-Hispanic whites aged 50+, 2006–2012 HRS genetic sample.
VariablesNative-BornForeign-Born
Mean (SD)/ProportionMin/MaxMean (SD)/ProportionMin/MaxSign.
Age67.32 (10.62)50/10169.53 (10.98)50/96***
Female0.562 0.582
Education (years)13.26 (2.50)0/1713.36 (3.06)2/17
Region
  Northeast0.156 0.276 ***
  Midwest0.289 0.189 ***
  South0.371 0.235 ***
  West0.183 0.300 ***
Height (meters)1.69 (0.102)1.22/2.251.68 (0.097)1.42/1.93**
BMI28.05 (5.78)13.3/64.926.88 (4.80)17.9/46.0***
  Normal or low (BMI < 25)0.309 0.365 *
  Overweight (BMI 25–29.9)0.382 0.401
  Obese (BMI 30+)0.309 0.234 **
Ever smoked0.570 0.610
  Never smoked0.430 0.390
  Former smoker0.433 0.518 ***
  Current smoker0.137 0.092 **
Cognitive score (word recall)10.21 (3.30)0/209.67 (3.55)0/19**
  Cognitive score < 80.191 0.244 **
  Cognitive score 8–120.571 0.524
  Cognitive score 13+0.238 0.232
Depression (CESD-8)1.29 (1.85)0/81.26 (1.87)0/8
  No depressive symptoms0.491 0.505
  1–2 depressive symptoms0.323 0.321
  3+ depressive symptoms0.187 0.174
N11,253 414
*** p < 0.001, ** p < 0.01, * p < 0.05 based on two-tailed t-tests.
Table 2. Log odds coefficients from the multinominal logistic regression models predicting cognitive score categories (word recall 20).
Table 2. Log odds coefficients from the multinominal logistic regression models predicting cognitive score categories (word recall 20).
VariablesModel 1Model 2Model 3
(Ref. 8–12)(Ref. 8–12)(Ref. 8–12)
<813+<813+<813+
(Ref. U.S.-born)
Foreign-born0.2430.0880.265 *0.039 a0.227−0.030
(0.133)(0.128)(0.133)(0.132)(0.140)(0.136)
Age−0.094 **0.145 ***−0.099 **0.155 ***−0.118 **0.156 ***
(0.035)(0.041)(0.035)(0.042)(0.036)(0.043)
Age squared0.001 ***−0.001 ***0.001 ***−0.002 ***0.001 ***−0.001 ***
(0.000)(0.000)(0.000)(0.000)(0.000)(0.000)
(Ref. Male)
Female−0.425 ***0.578 ***−0.444 ***0.613 ***−0.512 ***0.730 ***
(0.053)(0.049)(0.053)(0.050)(0.054)(0.051)
(Ref. Northeast)
Midwest0.066−0.175 *0.067−0.203 **0.071−0.143
(0.081)(0.073)(0.086)(0.077)(0.088)(0.079)
South0.168 *−0.1100.176 *−0.1210.187 *−0.089
(0.078)(0.070)(0.081)(0.073)(0.084)(0.074)
West−0.1400.079−0.1150.0250.011−0.044
(0.092)(0.077)(0.096)(0.080)(0.098)(0.082)
Cognitive function PGIs −0.187 ***0.390 ***−0.117 ***0.288 ***
(0.030)(0.027)(0.031)(0.028)
10 PCs YesYesYesYes
Education (yrs) −0.155 ***0.208 ***
(0.011)(0.012)
Constant−0.760−3.657 *−0.700−3.960 **1.874−6.727 ***
(1.362)(1.461)(1.363)(1.487)(1.392)(1.525)
N11,63911,63911,639
Pseudo R20.0980.1130.143
Chi2170019062249
Df224446
Log likelihood−10,287−10,114−9774
*** p < 0.001, ** p < 0.01, * p < 0.05. Robust standard errors in parentheses. Models also include controls for the year of interview and age at the first interview in the HRS. a Coefficient is different from Model 1, p < 0.05.
Table 3. Log odds coefficients from the multinominal logistic regression models predicting BMI categories.
Table 3. Log odds coefficients from the multinominal logistic regression models predicting BMI categories.
VariablesModel 1Model 2Model 3
(Ref. BMI < 25)(Ref. BMI < 25)(Ref. BMI < 25)
BMI 25–29.9BMI 30+BMI 25–29.9BMI 30+BMI 25–29.9BMI 30+
(Ref. U.S.-born)
Foreign-born−0.040−0.296 *−0.024−0.241 b−0.026−0.240
(0.118)(0.138)(0.120)(0.143)(0.120)(0.143)
Age0.102 ***0.280* **0.096 ***0.274 ***0.093 ***0.267 ***
(0.026)(0.032)(0.027)(0.034)(0.027)(0.034)
Age squared−0.001 ***−0.002 ***−0.001 ***−0.002 ***−0.001 ***−0.002 ***
(0.000)(0.000)(0.000)(0.000)(0.000)(0.000)
(Ref. Male)
Female−0.684 ***−0.383 ***−0.708***−0.439 ***−0.719 ***−0.463 ***
(0.046)(0.050)(0.047)(0.052)(0.047)(0.052)
(Ref. Northeast)
Midwest0.1040.275 ***0.1400.281***0.1340.268 **
(0.072)(0.076)(0.076)(0.083)(0.076)(0.083)
South0.0100.0200.028−0.0070.024−0.019
(0.068)(0.073)(0.071)(0.079)(0.071)(0.079)
West−0.075−0.171 *−0.042−0.161−0.027−0.122
(0.076)(0.083)(0.080)(0.089)(0.080)(0.089)
BMI PGIs 0.358 ***0.859 ***0.355 ***0.847 ***
(0.027)(0.030)(0.027)(0.030)
10 PCs YesYesYesYes
Education (yrs) −0.026 **−0.069 ***
(0.010)(0.011)
Constant−2.229 *−7.831 ***−1.959 *−7.609 ***−1.474−6.329 ***
(0.915)(1.089)(0.932)(1.140)(0.946)(1.160)
N11,63911,63911,639
Pseudo R20.0300.0680.070
Chi2693.514321458
Df204244
Log likelihood−12,343−11,860−11,838
*** p < 0.001, ** p < 0.01, * p < 0.05. Robust standard errors in parentheses. Models also include controls for the year of the interview. b Coefficient is different from Model 1, p < 0.1.
Table 4. Log odds coefficients from the multinominal logistic regression models predicting smoking status.
Table 4. Log odds coefficients from the multinominal logistic regression models predicting smoking status.
VariablesModel 1Model 2Model 3
(Ref. Former)(Ref. Former)(Ref. Former)
NeverCurrentNeverCurrentNeverCurrent
(Ref. U.S.-born)
Foreign-born−0.243 *−0.455 *−0.259 *−0.440 *−0.261 *−0.433 *
(0.110)(0.185)(0.111)(0.187)(0.112)(0.188)
Age−0.120 ***0.090−0.105 ***0.080−0.104 ***0.081
(0.027)(0.049)(0.027)(0.049)(0.027)(0.050)
Age squared0.001**−0.001 **0.001 **−0.001 **0.001 **−0.001 **
(0.000)(0.000)(0.000)(0.000)(0.000)(0.000)
(Ref. Male)
Female0.793 ***0.351 ***0.837 ***0.322 ***0.858 ***0.294 ***
(0.042)(0.061)(0.042)(0.061)(0.043)(0.062)
(Ref. Northeast)
Midwest0.281 ***0.1560.248 ***0.1080.259 ***0.080
(0.064)(0.093)(0.068)(0.099)(0.068)(0.100)
South0.136 *0.198 *0.184 **0.1320.186 **0.100
(0.062)(0.089)(0.065)(0.093)(0.065)(0.094)
West0.093−0.259 *0.096−0.288 **0.063−0.205
(0.070)(0.105)(0.073)(0.110)(0.073)(0.111)
Smoking initiation PGIs −0.336 ***0.229 ***−0.324 ***0.192 ***
(0.023)(0.034)(0.023)(0.035)
Smoking frequency PGIs −0.0170.127 ***−0.0070.098 **
(0.021)(0.031)(0.021)(0.031)
10 PCs YesYesYesYes
Education (yrs) 0.054 ***−0.148 ***
(0.009)(0.013)
Constant3.040 **−2.3362.487 *−2.0481.669−0.082
(1.013)(1.753)(1.023)(1.765)(1.035)(1.799)
N11,58811,58811,588
Pseudo R20.0430.0640.075
Chi2849.312741422
Df224648
Log likelihood−11,044−10,801−10,677
*** p < 0.001, ** p < 0.01, * p < 0.05. Robust standard errors in parentheses. Models also include controls for the year of interview and age at the first interview in the HRS.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Gubernskaya, Z.; Conley, D. Does Genetic Predisposition Explain the “Immigrant Health Paradox”? Evidence for Non-Hispanic White Older Adults in the U.S. Populations 2025, 1, 4. https://doi.org/10.3390/populations1010004

AMA Style

Gubernskaya Z, Conley D. Does Genetic Predisposition Explain the “Immigrant Health Paradox”? Evidence for Non-Hispanic White Older Adults in the U.S. Populations. 2025; 1(1):4. https://doi.org/10.3390/populations1010004

Chicago/Turabian Style

Gubernskaya, Zoya, and Dalton Conley. 2025. "Does Genetic Predisposition Explain the “Immigrant Health Paradox”? Evidence for Non-Hispanic White Older Adults in the U.S." Populations 1, no. 1: 4. https://doi.org/10.3390/populations1010004

APA Style

Gubernskaya, Z., & Conley, D. (2025). Does Genetic Predisposition Explain the “Immigrant Health Paradox”? Evidence for Non-Hispanic White Older Adults in the U.S. Populations, 1(1), 4. https://doi.org/10.3390/populations1010004

Article Metrics

Back to TopTop