|Home | About | Journals | Submit | Contact Us | Français|
An impressive body of public health knowledge on health care disparities among Latinos has been produced. However, inconclusive and conflicting results on predictors of health care disparities remain.
We examined the theoretical assumptions and methodological limitations of acculturation research in understanding Latino health care disparities, the evidence for socioeconomic position as a predictor of health care disparities, and the effectiveness of cultural competency practice. Persistent use of culture-driven acculturation models decenters social determinants of health as key factors in health disparities and diminishes the effectiveness of cultural competency practice. Social and economic determinants are more important predictors than is culture in understanding health care disparities.
Improvements in the material conditions of low-income Latinos can effectively reduce health care disparities.
Health disparities, and the association between disparities and social determinants, are not new concepts. W. E. B. DuBois noted racial and ethnic disparities more than a century ago in his analysis of 1890 census data in Philadelphia, Pennsylvania.1 DuBois observed that health disparities reflected a vast set of problems having a common center that must be studied according to some general plan.2 Since 1906, considerable interest in understanding the causes of these differences has prompted attention to health care disparities and their association with race, ethnicity, gender, socioeconomic position (SEP),3 and culture.4 Acknowledgment is widespread that population health is more a function of good public health measures and socioeconomic conditions than biomedical advances. Public health has focused on population health since the 19th century but has not consistently addressed social and economic inequality and inequity. Scientific assumptions that omit consideration of structural and economic factors in the study of Latino health disparities must be challenged.5–10
Tentative associations between acculturation research and health care disparities, as the empirical foundation for cultural competency practice, require further interrogation. In our consideration of this issue, we asked three questions: (1) How does acculturation research elucidate predictive factors that contribute to health care disparities? (2) How does the body of work on social determinants of health and structural inequality inform acculturation research? (3) Does cultural competency practice contribute to decreasing disparities among Latino groups?
Acculturation research has been a central lens in the study of Latinos in all disciplines for the past 40 years and has assumed a central role in the scientific discourse on Latinos and health disparities in public health. The epistemology of the construct of acculturation provides insights into its origins and meaning. The acculturation construct was developed in the early 20th century by sociologists known as melting pot theorists.11 The framework was designed as an irreversible, three stage model for acculturation that consisted of contact, accommodation, and assimilation. Multiple revisions of the acculturation construct have expanded its scope, but the underlying assumptions and focus on culture have not changed significantly.12–16
Measurement scales of acculturation often assume that an individual must move out of the culture of origin domain into the host culture domain (e.g., Bicultural Involvement Questionnaire, National Health and Nutrition Examination Survey acculturation questionnaire).16–18 A bidimensional model that separates maintenance of culture of origin from adoption of host culture has also been developed (e.g., Acculturation Rating Scale for Mexican Americans-II,19 Bidimensional Acculturation Scale for Hispanics20).
The acculturation construct is rooted in the study of immigrant populations that recently arrived in the United States and assessment of the interaction of new immigrant groups with the host culture. Flawed assumptions are evident in the theory of acculturation (e.g., individuals born outside the United States must abandon their culture of origin to be acculturated, and individuals are free to choose to become an integral part of American society). Both the National Health and Nutrition Examination Survey questionnaire and the Bidimensional Acculturation Scale for Hispanics derive acculturation scores mostly from language-based questions (an indirect measure of the cultural process) rather than more representative measures of culture change (e.g., changes in values, attitudes, and maintenance of traditions).20–23
Measurement scales capture a static and proxy view of acculturation. Multiple unmeasured factors are associated with acculturation processes, including educational level, family structure, religious beliefs, power relationships between majority and minority groups, and social stigma.12 Social science and public health definitions of acculturation23–31 are based on the initial assumptions of culture-driven models derived from earlier work and are operationalized through either individual proxy indicators of a unidirectional process (e.g., English language proficiency, self-reported ethnic identity) or a composite of these indicators (Table 1).
Acculturation studies have steadily increased in public health over the past 40 years. A Medline search with the keyword “socioeconomic” yielded more than 75 420 articles published since 2000; a search of the same period with the keyword “acculturation” produced 3485 articles. With the keywords “Hispanic” and “socioeconomic,” a search produced 2089 articles. Although SEP measures are more common than acculturation measures in the health literature, a search for “Hispanic” and “acculturation” yielded 729 articles; a search for “socioeconomic,” “Hispanic,” and “acculturation” together identified only 177 articles (24% of Hispanic acculturation articles).
Latinos (especially those of Mexican origin) are the most studied group in research on culture or ethnic groups, acculturation, and immigrant status.32–36 Although acculturation research acknowledges the role of socioeconomic factors,6,37 it emphasizes culture as a determinant of health disparities. Acculturation studies as a whole produce a pan-ethnic view of the Latino population as a foreign, homogeneous cultural group; differences between Latino subgroups and by nativity are ignored. This body of work has been plagued by multiple conceptual and methodological issues, leading to conflicting and inconclusive results. For example, a recent review found that US-born and English-speaking Latinos experience more physical and mental health problems than do more recent immigrants.33,34 Other studies show that Latino immigrant groups may be at greater risk of adverse health behaviors.38–40
In their study of immigrants, Carrasquillo et al. concluded, “Among Latinos, health insurance, rather than language, residency, income, or ethnicity, has been found to be the most important determinant of access.”41(p922) Evidence that SEP is a central variable in understanding Latino health disparities includes recent research that contests the Latino health paradox and shows that most health indicators (particularly health care access and quality) are worsening for Latinos and Mexicans.42 Other studies conclude that access to health care is a key indicator of SEP and is most associated with foreign place of birth and citizenship status for Mexican-origin and other Latino subgroups.43,44
Although Abraido-Lanza et al. critiqued the omission of structural factors in acculturation research, they argued for a continued “consideration of the intersection of large scale social forces and culture.”44(p1345) If acculturation is heralded as an independent factor in explaining health care disparities, structural inequality and the role of SEP as it intersects with culture will remain invisible in acculturation research. Latinos no longer constitute one culture, and ethnic differences within and between groups by SEP, race, and geographic segregation provide a compelling argument for a broad social determinants framework in public health research and practice to reduce well-documented factors associated with racial/ethnic health disparities.
Current critiques suggest a rethinking of the limitations of acculturation as an organizing and theorizing principle in the study of Latino health care disparities. Scholars have provided critical observations on what acculturation research measures.43–49 The principal critiques are that the central concepts remain simple, ambiguous, and inconsistent and that models of acculturation are implicit or poorly specified, lack clear definitions, insufficiently conceptualize acculturation, often lead to inadequate generalizations regarding Latinos, and have not significantly advanced our understanding of health care disparities.
Viruell-Fuentes presented cogent arguments regarding the assumptions of acculturation research: (1) culture is located within an individual, and cultural traits are inherent to members of a particular group, and (2) the onus of culture is placed on the individual at the expense of addressing the structural contexts that reproduce social and economic inequities. Attention is diverted from examining the historical, political, and economic contexts of migration and how these contexts impinge on immigrant health.45 Hunt et al. agreed:
Acculturation has become a popular variable in research on health disparities among certain ethnic minorities, in the absence of serious reflection about its central concepts and assumptions. The explicit concern is that acculturation as a variable in health research may be based more on ethnic stereotyping than on objective representations of cultural difference.47(p973)
The authors concluded,
In the absence of a clear definition and an appropriate historical and socio-economic context, the concept of acculturation has come to function as an ideologically convenient black box, wherein problems of unequal access to health posed by more material barriers, such as insurance, transportation, education, and language, are pushed from the foreground, and ethnic culture is made culpable for health inequalities.47(p982)
These critiques highlight major methodological flaws of previous research that have contributed to inconclusive or conflicting evidence and direct our attention to parallel scientific findings regarding the relationship between Latino health disparities and SEP.
SEP, a dimension of social stratification, encompasses a variety of terms, including socioeconomic status, social class, and social status. In general, better health is “more closely associated with social advantage than with social disadvantage,”50(p47) although there is variation in the magnitude of the association with particular diseases. A recently released federal disparities strategic framework characterizes the literature as “replete with examples of the associations between socioeconomic status (SES) and morbidity/mortality and the significant implications of SES for health.”51
No single indicator of SEP is best for all circumstances, because each emphasizes particular aspects of social stratification. Education is commonly thought to capture knowledge-related assets of individuals, and in part, it reflects circumstances of early life and country of educational experience, which are associated with health. As Galobardes et al. note, “the same indicator of SEP may not capture equally the socioeconomic distribution in different ethnic groups,”50(p67) and may have different meanings. Several theoretical models can be used to conceptualize how SEP influences disease risk.50 The Critical Period, Barker, and Fetal Origins Hypothesis models hold that exposure during a critical period may have lasting effects that result in greater risk of disease. Alternative models suggest that risks accumulate over the life course or can cluster.
Regardless of socioeconomic indicator, Latinos have lower socioeconomic status than do non-Latino Whites.52–54 For instance, 39.4% of Latinos but only 10.6% of non-Latino Whites aged 25 years or older do not have a high school degree (Table 2).52 More than twice as many Latinos (20.7%) as non-Latino Whites (9.0%)52 are living below the federal poverty level. Total Latino per capita income is less than half that of non-Latino Whites ($15 502 and $31138, respectively).54 Latino households have less wealth than do White households, regardless of income.53 Latino households are more likely to receive food stamps (12.6%) than are non-Latino White households (5.2%), and Latinos aged 16 years and older in the labor force are more likely than are non-Latino Whites to be unemployed (7.3% and 5.2%, respectively).52 Latinos are three times as likely as non-Latino Whites to lack health insurance (32.1% and 10.4%, respectively).54
Important within-group and between-group differences exist among foreign-born and US-born Latinos in demographic, language, and access indicators. Six out of every 10 Latinos are US born, but Latino subgroups differ (e.g., 38.9% of Cubans, 7.1% of Central Americans, and 6.8% of South Americans are foreign born).55 Approximately 70% of Mexican-origin persons living in the United States are US citizens,56 and 40.1% of Mexico-born persons immigrated to the US more than 10 years ago.
Studies have demonstrated that SEP has a stronger effect than acculturation on health care use among Latinos. Almost two decades ago, data from the Hispanic Health and Nutrition Examination Survey showed that use of preventive services among Mexicans, Cubans, and Puerto Ricans was “predicted more strongly by access to care than by acculturation.”57(p11) More recently, Morales et al. noted “the weak associations between access to care and acculturation, survey language, and citizenship status” among HIV-infected Latino patients once education and insurance status were taken into account.58(p1120) Data from the National Health Interview Survey showed access factors and previous screening were more strongly associated with current use of Papanicolaou tests and mammograms than were language and ethnic factors.59 Suarez also found no association between acculturation and the use of Papanicolaou tests and mammograms in El Paso, Texas, once income, insurance, and education were controlled.60 Similarly, Palmer et al. found no significant effect for acculturation in predicting mammography screening among Latino women living in farmworker communities.61
By contrast, O’Malley et al. found that Latino women in New York City who were more acculturated had significantly higher odds of recently receiving a mammogram than did less acculturated women, once demographic, socioeconomic, and health system characteristics were controlled.62 The authors noted, “If these factors are not controlled for, acculturation may simply act as a proxy for socioeconomic status.”62(p223)
Variations in how acculturation is measured may partly explain these conflicting findings, but acculturation measures, including limited English language proficiency, are strongly associated with education and health insurance coverage.63 Dubard and Gizlice recently found that Spanish-speaking Latinos are more likely to be uninsured, not to have a personal doctor, and to receive less preventive care than English-speaking Latinos after adjustment for demographic and socioeconomic factors.64 Language can therefore be viewed as both an access factor and a communication barrier between health care provider and patient.61
The 1985 government task force report on Black and minority health acknowledged that racial/ethnic minority groups exhibited unique health care patterns and that research and interventions were needed to provide more responsive care to these underrepresented groups.65 Acculturation theorizing formed the basis for acceptance of cultural competency practice to redress unmet needs of Latinos as a cultural group, with limited attention paid to addressing economic disadvantage. Cultural competence is a
set of congruent behaviors, attitudes, and policies that come together in a system, agency or profession that enables that system, agency or profession to work effectively in cross-cultural situations.66(piv)
This practice identifies systems, agencies, or professions as potential points of intervention. For example, cultural competency education of physicians tends to emphasize culture over social determinants of health. The American Association of Medical Colleges (AAMC) has identified definitions of race, ethnicity, and culture as an integral part of the education and training of medical students and physicians in cultural competency and health disparities.67 However, the AAMC did not provide these definitions. Even when medical schools use the AAMC’s Tool for Assessing Cultural Competence Training to guide the development of cultural competency and health disparities curriculum for undergraduate and graduate medical education, wide variation exists in curricula. In a proposed revision of this instrument, only 1 of the 42 learning objectives specifically addresses social determinants; 8 address culture, and 3 address interpretation.68 Lie et al. identified the need to “critically appraise literature on health disparities” as an underaddressed curriculum content area.68(p5)
Carrasquillo and Lee-Rey69 based their argument for diversification of the medical classroom on the finding that approximately half of all underrepresented minority graduates plan to care for under-served populations, whereas fewer than 20% of other new doctors had such plans.70 In addition,
less than half of all students in these anonymous surveys responded that access to care was a major problem, and only 42% responded that everyone is entitled to adequate health care.69(p1204)
Like acculturation research, cultural competency practice has emphasized a panethnic culture, with less attention to SEP, within-group differences, and the importance of social determinants of health as causal factors in health disparities.
Reports and studies on cultural competency practice have identified challenges to its implementation: (1) lack of agreement on the terms, definitions, and core approaches; (2) limited research on impact and effectiveness; and (3) absence of a funding source considered sufficient to implement new initiatives.71–84 These initiatives have not yielded significant improvements in reducing health disparities for Latinos. Cultural competency practice can be most effective when (1) resources are available to provide high-quality and appropriate care, (2) providers know the population they are serving and respond to their patients’ unmet needs appropriately for their language and literacy, and (3) providers are committed to providing comprehensive and high-quality care. Resource-poor cultural competency programs represent an incomplete solution to decreasing health care disparities for low-income Latinos and are no substitute for well-funded, high-quality public service programs that are inclusive of ethical public health features of responsiveness to constituencies served.
An impressive body of evidence has been amassed on racial and ethnic disparities in health care: where they exist, who is affected, and what are their consequences.85–94 Persistent use of individual or culture-driven models in public health with Latino subgroups ignores the effect of residence in low-resource communities, low SEP, the social construction of marked cultural identities, and institutional patterns of unequal treatment, all of which contribute to health disparities. Expanding and shifting the boundaries of public health research to include social determinants of health, institutional barriers, and structural inequality will enhance our understanding of health disparities among Latinos by subgroup and will promote the effective implementation of policy solutions that are consistent with the findings.
The American Public Health Association has proposed several policy solutions, including raising income to protect health; providing universal health care; focusing on prevention; increasing the proportion of underrepresented US racial/ethnic minorities in the health professions; increasing the number of community health workers and expanding their roles; increasing reporting, monitoring, and tracking of state, hospital, and local data; and using community-based participatory approaches for research and interventions.95 Latino health research will be strengthened by expanding its theoretical lens and designing improved methods for capturing factors that are most powerful in predicting inequity in access to goods and benefits among different racial and ethnic groups in the United States.
O. Carter-Pokras received funding from the National Heart, Lung, and Blood Institute (grant 1 K07 HL079255).
The authors gratefully acknowledge the exceptionally thoughtful and insightful comments from reviewers, assistance with literature reviews and article preparation by Laura A. Logie, PhD, and research assistance by Sylvette La Touche-Howard.
Reprints can be ordered at http://www.ajph.org by clicking the “Reprints/Eprints” link.
ContributorsR. E. Zambrana originated the article, searched and reviewed the literature, and wrote and revised each draft. O. Carter-Pokras contributed to the review of the literature and to writing and reviewing drafts of the article.
An earlier version of this paper was presented by R. E. Zambrana at the Society for Public Health Education on November 2, 2007, in Arlington, VA.
Ruth E. Zambrana, Department of Women’s Studies and the Consortium on Race, Gender and Ethnicity, University of Maryland, College Park.
Olivia Carter-Pokras, Department of Epidemiology and Biostatistics, School of Public Health, University of Maryland, College Park.