Association of acculturation and country of origin with self-reported hypertension and diabetes in a heterogeneous Hispanic population

Hispanics are the fastest growing minority group in the U.S., comprising 15% of the population [1]. Compared with non-Hispanic whites, Hispanics have significantly higher rates of diabetes and hypertension [2‐8]. These differences account for the disparity in cardiovascular-related morbidity and age adjusted mortality between Hispanics and non-Hispanic whites [9, 10].

While disease prevalence has been shown to vary by Hispanic subgroups [11‐14], few studies have examined the differential effect of acculturation across Hispanic subgroups defined by country of origin. Acculturation is defined as a multidimensional process through which foreigners adopt the customs of a host country [13]. While a standard acculturation metric is not fully agreed upon[15], language remains an important acculturation metric since more than three-quarters of the U.S. Hispanic population speak a language other than English at home [1]. Other important components of acculturation include: nativity (U.S.-born or foreign-born), years of residence in the United States and citizenship status [15]. Although Hispanics have been traditionally studied as a homogenous group [8], understanding determinants of chronic diseases among distinct subgroups remains salient in planning public health interventions.

Since the Hispanic population in California is two times greater than that of the remainder of the U.S. (36.2% versus 15.1%) [16], we examined the hypothesis that diabetes and hypertension prevalence among U.S. Hispanics varies by country of origin and by the degree of acculturation using the California Health Interview Survey (CHIS).

We examined the adult participants in the 2001, 2003, 2005, and 2007 CHIS, the largest health survey conducted in any state. CHIS is a population-based random-digit dial telephone survey of California’s population conducted every other year since 2001. CHIS employs a multi-stage sample design and includes cellphone-only households. In addition, interviews are conducted in English, Spanish, and three Asian languages. Using CHIS allows sampling of non-Carribean Hispanics in larger numbers than what is possible through national datasets. A detailed description of the history, design, data collection, and available data for CHIS is available at the website [17].

There were 33,633 self-identified Hispanics and 126,448 NHWs. There were 13,645 U.S.-born Hispanics (Table 1) and 19,988 foreign-born Hispanics (Table 2). Total self-identified Hispanic subtypes included: Mexican (n = 26,091), Central American (n = 3,092), South American (n = 921), 2 + Hispanic (n = 2,245), and Other Hispanic (n = 1,284). Overall, regardless of nativity status, Hispanics were slightly younger than NHWs. Since our hypertension and diabetes findings did not change in separate sensitivity analyses stratifying for age greater than 45 years, our data is not presented stratified by age.

Using a large state sample of Hispanics over several years, our study has several important findings on the association of acculturation and chronic disease prevalence. First, self-reported prevalence of hypertension and diabetes varied by country of origin. Second, stratifying by nativity status demonstrated significant differences in disease reporting by country of origin. Third, among all foreign-born subgroups, only Mexicans reported lower odds of hypertension after adjustment for socioeconomic and acculturation factors. Fourth, acculturation, as measured by years of residence in the U.S. and citizenship status, was an important predictor of hypertension and diabetes. This suggests that acculturation may differentially impact Hispanic subgroups. Our findings highlight the importance of disaggregation of Hispanics by country of origin and acculturation factors whenever possible.

Nativity had important differential effects. Compared with NHW, we found that U.S.-born Mexicans, Central Americans South Americans and foreign-born Mexicans had lower odds of reporting hypertension. The rate of hypertension increased after adjustment for socioeconomic status and acculturation factors in all groups except for South Americans. There was no difference in the reported odds of hypertension compared to NHWs in fully adjusted models for the other Hispanic subgroups. As people spend more time in the U.S., lifestyle changes and a consequent increase in BMI may significantly lead to a poorer cardiovascular profile [19‐21]. Our study also found that U.S.-born Hispanics had higher BMIs compared to foreign-born Hispanics, regardless of country of origin. However, changes in clinical risk factors might differ in each Hispanic subgroup. Prior studies have shown that despite similar cardiac risk factors, Hispanic subgroups have varying degrees of subclinical CVD, including higher coronary artery calcium (CAC) scores and inflammation as measured by C-reactive protein [12, 22‐24]. This may be attributed to the genetic diversity in Hispanic subgroups [25]. The low prevalence of hypertension among Mexicans, despite acculturation, might be explained by the geographic proximity to their native land, which may play a stronger role in preservation of family ties and traditions despite length of residency in the U.S. and English language acquisition [26].

The “Hispanic Paradox” refers to the epidemiological finding that foreign-born Hispanics – largely Mexican – often fare better than their white counterparts on morbidity and mortality outcomes, despite lower levels of income, education, and worse health care access [27‐29]. This difference has been attributed to a healthy migrant effect, healthier behaviors, and/or cultural traditions. The healthy migrant effect posits that healthier persons are more likely to migrate thus producing increased longevity and health in the emigrant population. The acculturation hypothesis states that Hispanic cultural orientation results in healthier behaviors that result in better health outcomes and is thus protective against the effects of lower socioeconomic status in the U.S. [29]. Prior studies of the healthy migrant effect have focused primarily on Mexicans with few studies focusing on Hispanic subgroups [30].

Similar to prior studies, we also found lower odds of reported hypertension among foreign-born Mexicans even after adjustment for acculturation [31]. However, we found no significant difference in the odds of reported hypertension compared with NHWs after adjustment for socioeconomic and acculturation factors among all other foreign-born Hispanic subgroups. Our findings suggest that acculturation may differentially impact Hispanic subgroups and highlights the importance of disaggregation of Hispanics by country of origin and acculturation factors.

In agreement with existing literature, we found that reported diabetes rates varied with country of origin [6, 7, 32]. In our study, Mexicans, regardless of nativity status, have higher odds of reporting diabetes compared to NHWs, even after adjusting for BMI. In contrast, U.S.-born Central and South Americans had lower odds of reporting diabetes compared to NHWs. In addition, U.S.-born Other Hispanic and foreign-born Central Americans had higher odds of reporting diabetes. These findings might be attributed to genetic and behavioral differences, which make different subgroups more susceptible to diabetes and the metabolic syndrome and not hypertension. We found an increasing trend in the odd ratios for reporting diabetes as the years of residence in the U.S. increased (Table 4). This is consistent with prior studies that have shown an inverse relationship between increasing acculturation and a less healthy diet and increased rates of obesity [5, 8, 21, 30, 33]. Our findings show a complex relationship between greater acculturation and diabetes, likely because of the distinct Hispanic subgroups and numerous acculturation metrics available in our study.

We found higher odds of reporting hypertension among foreign-born Other Hispanics and higher rates of diabetes among U.S.-born Other Hispanics compared with NHWs. These differences may be due to a higher socioeconomic and acculturation status for this subgroup. Compared to other Hispanic subgroups, the Other Hispanics had high rates of college education, higher incomes, private insurance, and higher unadjusted rates of hypertension and diabetes. Among the foreign-born Hispanics, Other Hispanics were the most acculturated with 82% living in the U.S. greater than fifteen years and most speaking English well. The self-identification as “Other Hispanic” may reflect people who do not fit a particular pre-established Hispanic subgroup (i.e., European Hispanics or Brazilian) or people who identify as biracial or bi-ethnic (i.e., participant with only one Hispanic parent). The self-identity of Hispanics is complex and is influenced by the country of origin and nativity status [16]. Future studies should include bi-racial or bi-ethnic Hispanics as they constitute a poorly studied subgroup.

In conclusion, we found significant differences in rates of self-reported hypertension and diabetes by Hispanic subgroups, nativity, and acculturation. Our findings underscore the importance of exploring within-group differences and heterogeneity among Hispanics in the U.S., differences that may account for the mixed and paradoxical associations between acculturation and various health indicators. Finally, better standardized data collection strategies that incorporate Hispanic subgroups and acculturation factors will allow for improved tracking of health status among Hispanics, which may better inform public health interventions.