Unraveling a Public Health Enigma: Why Do Immigrants Experience Superior Perinatal Health Outcomes?

RESEARCH IN THE SOCIOLOGY OF HEALTH CARE HEALTH CARE DELIVERY SYSTEM CHANGES: NEW ROLES FOR PROVIDERS, INSURERS, AND PATIENTS

Editor: JENNIE JACOBS KRONENFELD School of Health Administration and Policy Arizona State University

VOLUME 13 • 1996 (PART B)

@ Greenwich, Connecticut

JAI PRESS INC. London! England

LINDA H. AIKEN and MARGERY MULLIN

336

Pereda, c., and M. Cifuentes. 1991. Preliminary Results: Combined Study.on Parmers ~elati~n; and Risk ofHIV Infection; Knowledge. Attitudes. Beliefs. and Practices on AIDS m Chile s Capital City and TWo City Ports 1990. Wo~ld Healt~ Organization, G~obal Programme on AIDS, Social and Behavioural Unit. Santiago, Chile; School of Pubhc Health, Faculty of Medicine, University of Chile. P'tt D 1993 "Brief comments." Health Transition Review 3: 215-216. ~~ro' M.B.S., M.G. Suarez, J.M. Uribe, PA. Cerda, H.V. Es~oz, and R.L. Arroya:e. 1989. Q ' "Marcadores de Infeccion por VIH en Prostitutas: Analisls de Factores Preventlvos y de Riesgo." Revista Medica de Chile 117: 624-628. "Parti6 distribuci6n gratuita del AZT." La Epoca (December 2), p. 13. Ramos, C.G. 1992 . 6 16 19 Rodriguez, V.M. 1991. "Afectadisimos por el SIDA." ~PSI40 :. - . . "Research on AIDS Interventions m Developmg Countnes; State of the Art. 1990 D h . Scopper". Social Science and Medicine 30: 1265-1272. . S 'th H L 1993 "On the Limited Utility of KABP-Style Survey Data 10 the ml , Epid~miolo~y of AIDS, With Reference to the AIDS Epidemic in Chile." Health TfGlnsitiorL Review 3: 1-16. . Stone, V.E., 1.S. Weissman, and P.D. Clea~. 199~. "Satisfact~on With Anlbullat'[lry Persons With AIDS: Predictors of Patient Ratmgs of QualIty. Medicine 10: 239-245. 'k M 1990 "SIDA en Chile-Las Culpas de Silencio. APS1338: 4-13. VodanOVI,. . . ' . al SOI~io..Ge:ogJrap,hlc • •·.•. Wallace, R. 1991. "Traveling Waves of HIV Infectton on a Low DImenSIOn Network." Social Science and Medicine 32: 847-852. . . . 1 P 1994 "Forecasting the Effect of Health Reform on U.S. PhySICIan Wemer, .. . " J, I ,r th American Requirement: Evidence From HMO Staffing Patterns. ouma OJ e Association 272(3): 222-230. World Health Organization. 1994. Global Programme on AIDS: The Current Global of the HIV/ AIDS Pandemic. . Wykoff, R.F., C.W. Heath, S.L. Hollis, S.T. Leonard, C.B. QUlller, 1.L. Jones, ~. and R.L. Parker. 1988. "Contact Tracing to Identify Human ImmunodefiCiency a Rural Community." Journal ofthe Ameri,;an Medic~l Association 259: 3563i5c;6",6,'mble. Wykoff, R.F., l,L. Jones, S.T. Longshore, S.L. HolliS, c.B. QUlller, H. Dow~a: and W,.. 1991. "Notification of the Sex and Needle-Sharing Partners of IndlVldu,aJs w~~h Immunodeficiency Virus in Rural South Carolina: 30-month Expenence. Se):ually} Transmitted Diseases 18: 217-22.

UNRAVELING A PUBLIC HEALTH ···········.··ENIGMA • .• >

WHY DO IMMIGRANTS EXPERIENCE SUPERIOR PERINATAL HEALTH OUTCOMES?

Ruben G. Rumbaut and John R. Weeks

ABSTRACT Recent research, has pointed to an apparent public health enigma among new immigrants to the United States: high-risk groups, particularly low-income immigrants from Mexico and Southeast Asia, show unexpectedly favorable perinatal outcomes and seem to be "superior health achievers." This study attempts to unravel the reasons for this paradox by examining an in-depth data set drawn from a Comprehensive Perinatal Program (CPP) in San Diego County providing prenatal care services to low-income pregnant women, most of whom were immigrants from Mexico and various Asian countries, The CPP data set consists of nearly 500 independent variables per case (including most of those listed in the research literature as likely biomedical and sociocultural detenninants 'of pregnancy outcomes) for a sample of both foreign-born and U.S.-born women, :matched to infant health outcome measures collected from hospital records for

Research in the Sociology of Health Care, Volume 13D, pages 337·391. opyright@ 1996 by JAI Press Inc. Il rights of reproduction in any form reserved. BN: 0-7623·0049·3

337

338

RUBEN G. RUMBAUT and JOHN R. WEEKS

every baby delivered by CPP mothers during 1989-1991. The analysis focuses on the identification of maternal risk factors that best explain observed ethnic andj or nativity differences in pregnancy outcomes (such as birth weight, diagnoses at birth, complications, length of hospitalization of the baby). Asians and Hispanics (mostly foreign-born) had superior outcomes relative to Anglos and African Americans (mostly U.S.-born), and within ethnic-racial groups outcomes were generally better for immigrants than for natives. For immigrants, outcomes seem to worsen as the general assimilation or "Americanization" of the mother increases. The comparative socioeconomic advantages of U.S.-born mothers-in education, employment, income, English literacy and proficiencyappear to be overwhelmed by biomedical, nutritional, and psychosocial disadvantages. However, when other risk factors were controlled in regression analyses, nativity and ethnicity washed out of the equations. Ten independent variables-a set of both biomedical and sociocultural risk factors-remained in the multivariate analysis as significant predictors of infant health outcomes. Four qualitative case histories of CPP mothers are drawn from psychosocial reports to shed further light on the quantitative findings derived from the statistical analysis. Some limitations and implications of these results for health policy and future research are considered.

INTRODUCTION The National Commission to Prevent Infant Mortality (NCPIM), establ:ishe:di< by Congress in 1987, has sought to make the health of mothers and chi.ldre:n/ a national priority and to focus attention on what one commission me:mlJetii called "a national disgrace" (Bradley 1989) and another analyst posed "capital crime" (Boone 1989). Over the past three decades the infant mClrtalitYi ranking of the United States has slipped from sixth to 24th place "1'T1;l'm,aHII"/ nations of the world (Population Reference Bureau 1991). More than American babies die each year, most of them needlessly, incurringan en4Drnl011ls i i if not incalculable human and societal cost. The NCPIM was born in the of the federal Task Force on Black and Minority Health, which pVlllL"U"U/ a "national paradox of phenomenal scientific achievement and improvement in overall health status, while at the same time, peJrsis:tel1lt significant health inequities exist for minority Americans" (U.S. DHHS p. 2). Infant mortality was identified by the task force as one of six of death which together account for more than 80 percent of the mortality observed among ethnic minority groups in excess of that in the white majority population (cr., Becerra, Hogue, Atrash, and Perez 1991; Eberstein 1991; Johnson 1987; Langlois, Ritter, and Casey 1991; Mangold and Powell-Griner 1991; Markides 1983; Rogers 1989; Samuels 1986; Scott 1990). Coincidentally, over the same period during which the United States infaIlt mortality ranking was sliding internationally, immigration to the United States

Unraveling a Public Health Enigma: Immigrants' Pregnancy Outcomes

339

was growing sharply-notably after the passage of the 1965 immigration law which eliminated racist national-origin quotas, the end of the Bracero program in 1964 which led to increased illegal immigration from Mexico, and the resettlement of large flows of Cold War refugees, especially after the end of the Vietnam War in 1975. This new and rapidly accelerating immigration to the United States is extraordinary both in its size and its diversity (Portes and Rumbaut 1990). The 1990 U.s. Census counted 19.8 million immigrants, an all-time high. Annual arrivals during the past decade have rivaled those of the historic peak years of the pre-World War I period. At that time 90 percent of the immigrants came from Europe and most settled in the northeastern and midwestern regions of the country. Today nearly 90 percent come from developing countries in Asia, Latin America, and the Caribbean, with the largest numbers-fully a third of the national total-concentrating in California. In fact, by 1990,22 percent of aU Californians were foreign-born, and their U.S.-born children nearly doubled that proportion (Rumbaut 1994). Heated public debate about the costs and benefits of immigration has intensified precisely in such areas of dense concentration, as illustrated most dramatically in California in 1994 by the passage of Proposition 187 (the initiative that would deny health care, public education, and other social services to illegal immigrants and their children). Yet lost in the clamor is the sheer complexity and irony of the phenomenom; immigrants today include at Once the most and the least educated and skilled groups in American society, as well as those with the lowest and the highest rates of poverty, welfare dependency, and fertility_ Given current trends, the size of the immigrant population may well be eclipsed and its composition further diversified during the I990s. The Immigration Act of 1990 increased worldwide legal immigration limits by about 40 percent, to 700,000 per year; of these, employment-based visas (reserved largely for professionals) nearly tripled to 140,000, and familySponsored admissions also expanded to about half a million annually. Ironically, since the end of the Cold War in 1989 and of the USSR in 1991 refugee admissions have also increased, adding well over 100,000 to the regula; annual flows. And the illegal immigrant population has not only grown but diversified. EXcluding the nearly 3 million formerly undocumented immigrants "\Vhose status was legalized under the Immigration Reform and Control Act 0[1986 (over 2 million were Mexican nationals, followed by Salvadorans and (juatemalans), the. Immigration and Naturalization Service estimated that by the end of 1992 the illegal resident population totaled about 3.2 million, and was growing at a rate of perhaps 300,000 annually, with approximately half [the total located in California (U.S. INS 1992). As a result of this sharply increased immigration, combined with the higher rtility offoreign-born women, for the last few decades the Hispanic and Asian opulations have been growing at a more rapid rate than African Americans-

340

RUBEN G. RUMBAUT and JOHN R. WEEKS

a phenomenon that is redefining the American ethnic mosaic. Within the Hispanic and Asian groups, respectively, Mexicans and Southeast Asians have had the highest levels of both immigration and natural increase. Mexico remains by far the leading source of both legal and undocumented immigration to the United States. Mexicans alone accounted for a huge 22 percent (4.3 million) of the total foreign-born population counted by the 1990 Census, and 26 percent of all immigrants arriving since 1970. The Southeast Asians (Vietnamese, Cambodians, and Laotians) predominate among refugee admissions-over one million have resettled in the United States since 1975and they also have exhibited the highest fertility rates among ethnic groups .' ." in the United States (Rumbaut 1995; Rumbaut and Weeks 1986; Weeks, •.. .•. . . Rumbaut, Brindis, Korenbrot, and Minkler 1989). In just over a decade, the . •. .••.. . Indochinese had grown to form the third largest Asian-origin population in. the country, behind the Chinese and F i l i p i n o s . / > Although today's immigrants and refugees are drawn from some of the. poorest countries of the Third World, it is not the case that the United States .•.•. . . • . . is thereby importing poor or surplus populations. International migration requires resources and resourcefulness, and the poorest of the poor are least/> apt to be in a position to migrate (Portes and Rumbaut 1990). In fact, the» United States is importing some of the best and brightest from all over thei\ world. This "brain drain" (or "brain gain" for the United States) is partlya>/ function of scarce opportunities in source countries and partly the result of •. professional occupational preferences built into U.S. immigration laws. For> example, among immigrant adults in the United States in 1990, over 60 percent of those from India and Taiwan had college degrees, as did between a third to a half of those from Iran, Hong Kong, the Philippines, Japan, South Korea, and China, and about 50 percent of all immigrants from Africa (Rumbaut 1994). Those figures are well above the corresponding norms (20%) for nativeborn Americans. Such positively selected high-socioeconomic-status immigrants are likely to constitute a low-risk population with respect to health" and their superior socioeconomic resources should be reflected in better than average health profiles. Yet, by contrast, only 3.5 percent of adult Mexican immigrants in 1990 had college educations-the lowest proportion of any ethnic group in the country as did only 4 percent or 5 percent of Southeast Asian refugees from Laos and Cambodia and largely undocumented immigrants from EI Salvador and Guatemala. These lower socioeconomic status immigrants combine high fertility rates with high poverty rates (higher than for African Americansil1 the United States) and face formidable barriers in accessing health care services (Rumbaut, Chavez, Moser, Pickwell, and Wishik 1988). They may therefore be expected to constitute a high-risk population exhibiting worse than average health outcomes. But recent research indicates that this, in fact, is not the case. Contrary to expectations, Mexican immigrants and Southeast Asian refugees

Unraveling a Public Health Enigma: Immigrants' Pregnancy Outcomes

341

appear to have lower than average death rates, posing what some investigators have c~lled a "contemporary public health enigma" (Williams, Chen, Binkin, and C~mgman 1986~ and an "epidemiological paradox" (Collins and Shay 1994; MarkIdes and Korell 1986; Scribner 1991) or what is more specifically an infant health paradox.

AN EPIDEMIOLOGICAL PARADOX? i\A study of ~erin~tal outcomes in California (Williams et al. 1986) analyzed data for Califorma for 1981 from the state's matched birth-death cohort file //for .four ,groups: non-Hispanic whites, blacks, U.S.-born Hispanics (Who, in •.. .•. .•.•. . . Callforma, are mostly of Mexican descent), and Mexican immigrants. In terms >iof maternal risk factors, the Mexican-born women had less education more children, shorter birth spacing, and a later start to prenatal care than ~ny of /< the other three groups. Yet, in terms of perinatal outcomes, the Mexican-born i.women had :he lowest percentage of low birthweight babies, the lowest