Hispanic Baby Boomers: Health Inequities Likely to Persist in Old Age

Abstract

Purpose:

As the Baby-Boom generation enters the ranks of the elderly adults over the next 4 decades, the United States will witness an unprecedented growth in racial/ethnic diversity among the older adult population. Hispanics will comprise 20% of the next generation of older adults, representing the largest minority population aged 65 years and older, with those of Mexican-origin comprising the majority of Hispanics. Little is known about the health status of this population.

Data/Methods:

Data are for Baby Boomers born between 1946 and 1964 (ages 43–61) in the 2007 California Health Interview Survey. Logistic regression estimates the odds of diabetes, hypertension, obesity, fair/poor self-rated health (SRH), and functional difficulties among U.S.-born non-Hispanic Whites (NHW), U.S.-born Mexicans, naturalized Mexican immigrants, and noncitizen Mexican immigrants.

Results:

The Mexican-origin populations are disadvantaged relative to NHW for all socioeconomic status (SES) and several health outcomes. The Mexican origin disadvantage in health attenuates when controlling for SES and demographics, but the disadvantage remains for diabetes, obesity, and fair/poor SRH.

Implications:

Baby Boomers of Mexican origin do not share the advantages of health, income, and educational attainment enjoyed by U.S.-born NHW. As this cohort moves into old age, the cumulative disadvantage of existing disparities are likely to result in continued or worse health disparities. Reductions in federal entitlement programs for the elderly adults that delay eligibility, scale back programs and services, or increase costs to consumers may exacerbate those inequities.

As the Baby-Boom generation enters the ranks of the elderly adults over the next four decades, they will represent the largest and most diverse older adult population in American history. Among the almost 80 person million individuals projected to be age 65+ by 2050, about 42% are expected to be members of an ethnic minority population, up from 20% in 2010 (U.S. Census Bureau, 2010). The fastest growing population among those aged 65 years and over is Hispanics. Currently, Hispanics represent 7% of the U.S. older adult population or 2.9 million individuals; by 2050, they are expected to represent nearly 20% of the population aged 65 years and over or 17.5 million individuals. This is more than a sixfold increase in the number of elderly Hispanics over the next four decades (U.S. Census Bureau, 2010). Despite the fact that they will represent the largest ethnic minority population among the next generation of older persons, relatively little is known about the health status of the Hispanic Baby-Boom population. The entire Baby-Boom population is often described in general terms as well educated, healthy, and economically secure (Congressional Budget Office, 1999; Frey, 2010). Baby Boomers as a group have benefited from improvements in education (U.S. Department of Education, 1998), pensions, and improvements in the quality of and access to health care (Lubitz, Greenberg, Gorina, Wartzman, & Gibson, 2001). There is also evidence of decreases in disability and improvements in late-life functioning among the middle-aged and older adult population in the aggregate (Crimmins, 2004). As a result, much of the public discourse around the future of Social Security and Medicare, as well as other aging programs and services, is based on the image of the Baby-Boom generation as healthy, wealthy, and wise (Wallace & Villa, 2009). Although this image may fit the Baby-Boom population in aggregate, it is unlikely to represent the situation of the Hispanic Baby-Boom population.

The empirical evidence regarding the health of older Hispanics finds a mixed picture of health advantages and disadvantages. Hispanics have better mortality outcomes, but worse morbidity and disability rates overall than non-Hispanic Whites (NHW; Markides & Wallace, 2007). Hispanics often have low incomes, are poorly educated and housed, and work in dangerous and difficult jobs, which are risk factors for early mortality in the general population. Yet Hispanics have better mortality outcomes than more privileged NHW, which has been described as a “Hispanic paradox.” This advantage is concentrated among immigrant adults and dissipates with years in the United States (Markides & Eschbach, 2005). Explanations for the paradox include the selective migration of healthier persons to start with, protective cultural factors after migration such as diet and smoking patterns as well as family dynamics and resulting social support, and the selective return migration of unhealthy migrants, that is, a salmon bias (Palloni & Arias, 2004). In spite of the mortality advantage found among the Hispanic population, it is important to recognize that the population does experience disparities in many other health indicators, similar to other minority and low-income populations, including morbidity, poor self-rated health (SRH), and functional difficulties (Markides & Eschbach, 2005).

Over the past two decades, new data have become available through national datasets that include oversamples of Hispanics, for example, the Health and Retirement Survey, as well as regional datasets that include large samples of Hispanics such as the Hispanic Established Population for the Epidemiological Study of the Elderly, the San Antonio Heart Study, the San Luis Valley Heart Study, and the Sacramento Area Latino Study on Aging. Evidence from these studies shows that the Hispanic population, primarily those identified as Mexican American, is living longer and has experienced gains in life expectancy that outpace NHW and other minority populations (Hayward, Warner, Crimmins, & Hidajat, 2007). At the same time, however, compared with NHW, Hispanics have a higher prevalence and mortality for certain chronic conditions including diabetes (Angel & Angel, 2006; Beard, Al Ghatrif, Samper-Ternent, Gerst, & Markides, 2009), infectious and parasitic diseases (Vega & Amaro, 1994), and obesity (Angel & Angel, 2006). They have higher rates of hospitalization for uncontrolled diabetes and diabetes-related complications (Beard et al., 2009; Agency for Healthcare Research and Quality, 2011), increased prevalence of hypertension (Ghatrif et al., 2011), worse SRH (Himes, 2005), and greater difficulty with functioning (Markides & Eschbach, 2005; Palloni & Arias, 2004). At all ages, Hispanics are less likely to have insurance coverage or a usual source of care (U.S. Census Bureau, 2008). According to Angel and Angel (2006), the health profile of older Mexican Americans is further compromised by the population’s lower than average levels of income. Williams (2005) concludes that the relatively low socioeconomic status (SES) profiles of Hispanic immigrants and native-born Hispanics, and the ongoing challenges that Hispanics face with educational and occupational mobility (Camarillo & Bonilla, 2001), suggest that the health status of Hispanics is likely to be worse than the U.S. average in the future.

Robert et al. (2009) also maintain that aggregate trends showing health status improvements among future cohorts of older adults should be viewed with caution because of the substantial health disparities that are tied to socioeconomic inequities. There is evidence that health disparities associated with SES are increasing in the United States (Mensah, Mokdad, Ford, Greenlund, & Croft, 2005) at the same time that income inequality is rising (U.S. Congressional Budget Office, 2011). The increase in minority populations among the elderly population in the future may result in an increase in health disparities given the cumulative disadvantage associated with a lifetime experience of poor SES, blocked social and educational opportunities, and discrimination (Dannefer, 2003; Ferraro & Shippee, 2009; Williams, 2005). In fact, racial/ethnic and socioeconomic inequities in health have risen to the point of becoming a major public health policy issue. Several federal agencies, including the Institute of Medicine, the Department of Health and Human Services, the National Institutes of Health overall, and the National Institute of Aging specifically have included in their goals and strategic plans an emphasis on identifying and reducing health disparities (Heckler, 1985; Smedley, Stith, & Nelson, 2002), improving our understanding of health disparities and eliminating health inequalities among older adults (Sood & Stahl, 2011), and achieving health equity (U.S. Department of Health and Human Services, 2011). Reaching these goals requires empirical investigation of the health status and correlates of poor health found among the population.

Whether or not Hispanic Baby Boomers will experience the health disparities of their predecessors in old age depends in part on the current level of health and SES found among the population. The available evidence finds that at all ages of adulthood, Hispanics have higher poverty rates, lower average incomes, and fewer accumulated assets than their NHW counterparts (Fry, Kochhar, Passel, & Suro, 2005; Taylor, Fry, & Kochhar, 2011). Hispanics are also more likely to experience longer spells of unemployment/underemployment, to not have a usual source of health care, and to live in segregated underresourced neighborhoods and attend under resourced poor performing schools (Hispanics in Philanthropy, 2011). The handful of studies that examine the health and SES of the Hispanic Baby-Boom population support these findings. The studies on health that were primarily conducted on preretirement Hispanic Boomers aged 50–61 years find evidence of increased levels of obesity and diabetes (Wray, Alwin, McCammon, Manning, & Best, 2006) and poorer functional health (Liang, Xu, Bennett, Ye, & Quiñoes, 2009). Data on the SES of Hispanic Baby Boomers (ages 35–53 in 2000) find that Hispanic Baby Boomers compared to NHW had lower levels of education, home ownership, lower investment income, and higher rates of material hardship and poverty; but there was significant variation by citizenship status with noncitizens experiencing the most SES disparity (Gassoumis, Wilber, Baker, & Torres-Gil, 2010). There is also evidence that Hispanic Baby Boomers have been disproportionately negatively affected by the economic downturn, especially with regards to the accumulation of wealth and assets (Pew Hispanic Center, 2009) and financial hardship (AARP, 2010). Therefore, it seems that the general or aggregate profile of the Baby-Boom population as “healthy, wealthy, and wise” is not in keeping with the experience of Hispanics.

The current analysis extends the work in this area by examining the health status and socioeconomic correlates of poor health among the Mexican origin and NHW Baby-Boom populations disaggregated by citizenship status. Our purpose is to investigate the likelihood of health inequities persisting between Mexican-origin populations and NHW in the coming generation of older adults. We first document current health and SES disparities between U.S.-born NHW, U.S.-born Mexicans, naturalized Mexicans, and Mexican-born noncitizens among the Baby-Boom generation. Next, we examine the extent to which inequities in several health outcomes are the result of race/ethnicity/citizenship, SES, and demographic differences between the populations. Examination of the experience of those of Mexican origin specifically is warranted given the heterogeneity of the Hispanic population with regards to some measures of health and SES, migration history, the proportion of immigrants in the population, and cultural practices (Angel & Angel, 2003). Furthermore, disaggregating the population by nativity is of particular importance because the immigrant generation has better mortality and better health for some health measures when compared to U.S.-born Mexicans, although their advantage declines with length of stay in the United States (Vega & Amaro, 1994; Williams, 2005). Lastly, the sustained and continued level of population migration from Mexico plays a significant factor in the growth of the Hispanic population in the United States (Angel & Angel, 2003).

The health and welfare of Mexican-origin Baby Boomers as reviewed earlier raises two theoretical issues: the Hispanic paradox and cumulative disadvantage. These frameworks are typically discussed separately because they come to different conclusions about the impact of SES throughout the life course. Cumulative disadvantage predicts that when a group experiences patterns of disadvantage, it can accumulate over the life course to create inequalities in old age (Dannefer, 2003). Applied to health disparities, a cumulative disadvantage perspective is concerned with the ways in which social inequality creates the accumulation of risk for low-income and minority populations that affect health status through a variety of mechanisms. Sources of inequality that can accumulate over the life course and therefore affect health in old age include childhood poverty, substandard education, institutional- and individual-level discrimination, and delayed health care (Schafer & Ferraro, 2011; Wallace, in press; Williams, 2005), predicting poor health outcomes for older Hispanics. But the Hispanic immigrant paradox notes that older Hispanics have lower mortality for most major conditions, contrary to the prediction of the cumulative disadvantage framework. The “cumulative inequality” framework is able to integrate the paradox with a cumulative disadvantage perspective by incorporating modifiers into the life course of disadvantage, including human agency, resource mobilization, and perceived trajectories that are independent of the larger contextual disadvantages (Ferraro & Shippee, 2009). The following analysis builds on the cumulative inequality framework to understand both health inequities experienced by Baby-Boom generation Mexican-origin Americans, as well as the divergences from the expected pattern demonstrated by immigrants. We expect that Mexican-origin Baby Boomers will have worse health than NHW. Among the Mexican-origin population, we expect that the Mexican–U.S.-born population will have worse health than naturalized and noncitizen Mexicans and that noncitizen Mexicans will have better health than either naturalized or U.S.-born Mexicans.

Methods

Data and Sample

Data are from the California Health Interview Survey (CHIS), conducted by the University of California, Los Angeles (UCLA), Center for Health Policy Research in collaboration with the California Department of Public Health, the Department of Health Care Services, and the Public Health Institute. (California Health Interview Survey, 2008; California Health Interview Survey, 2009) CHIS uses a multistage sampling design in conjunction with a random-digit-dial telephone survey of both landlines and cell phones to interview a random sample of households throughout California. Interviews are conducted in English, Spanish, and four Asian languages. The response rate for the screener was 35.5%, and among those screened, the completion of interviews for the adult sample was 52.8% for the landline sample and 52% for the cell phone–only sample. These rates are consistent with other telephone surveys, and the quality assurance studies by CHIS indicate that the weighted data are representative of California’s noninstitutionalized population (California Health Interview Survey, 2008). A methodology report provides more detailed information regarding the sample selection procedure (Cervantes & Brick, 2009).

The public-use dataset contains information for 51,048 adults 18 years and older. Our analysis is restricted to respondents who were between the ages of 43 and 61 years (i.e., “Baby Boomers”) at the time of the 2007 survey and who are of Mexican origin (U.S. born, naturalized, and noncitizens) or U.S.-born NHW, leaving a sample of 14,723. Because CHIS replaces missing values in data files through imputation for nearly every variable, there were no missing data and the analytic sample is the same (King, Rauch, & Edwards, 2008). California is uniquely suited to an empirical analysis of the health and SES of the Hispanic population. The state has both the largest overall Hispanic population, as well as the largest Mexican-origin population in the United States (Pew Hispanic Center, 2011) and the state’s demographic projections mirror national trends. Specifically, the California population is aging at a similar rate with 20% of the population expected to be age 65+ by 2050. Moreover, the diversity of the California population (all ages) and among those 65+ is increasing and will continue to increase over the next four decades. As with the U.S. population, the greatest increase will occur among the Hispanic population who are expected to represent 38% of the older adult population of California by 2050 (State of California, Department of Finance, 2007).

Measures

Health Outcomes.—

Six health-related dependent variables are examined: diabetes, obesity, high blood pressure, heart disease, condition that limits physical activity (CLPA), and self-rated general health (SRH). Diabetes, obesity, and high blood pressure are dichotomous variables that indicate whether or not the participant reports ever being told by a doctor that they had the specific disease/condition. Obesity is a body mass index of 30 or higher. CLPA was also a dichotomous variable based on a question that asked whether the respondents have a condition that substantially limits one or more basic physical activities such as walking, climbing stairs, reaching, lifting, or carrying. SRH was based on the standard SRH status question where respondents were asked to rate their general health, dichotomized as excellent/very good/good health versus fair/poor health.

Demographic Variables.—

The main independent variable is one that combines race, ethnicity, and citizenship status. The variable created allows for comparison of the four groups of interest to the study based on race, ethnicity, and citizenship status: U.S.-born NHW, U.S.-born Mexicans, naturalized Mexicans, and noncitizen Mexicans. We refer to this variable as Race/Ethnicity/Citizenship status. Gender, age, and marital status are included as control variables. Age is a continuous variable for ages 43–61 in 2007. Marital status was dichotomized as married or not married.

Socioeconomic Status Variables.—

Several variables indicate SES. In addition to the standard SES measure of poverty, education, and homeownership, we examine other measures that have been found to be related to health disparities, including perceived neighborhood safety (Subramanian, Kubzansky, Berkman, Fay, & Kawachi, 2006), resource availability (Cagney, Browning, & Wen, 2005), and health insurance status (Institute of Medicine, 2001). Poverty is a four-category measure using federal poverty line (FPL): 300% FPL and over, 200%–299% FPL, 100%–199% FPL, and 0%–99% FPL. Education reflects the highest level of education completed, some graduate school/graduate degree, some college/college degree, high school graduate, and less than a high school diploma. Home ownership includes whether the respondents owned or rented their home or had some other arrangement. Neighborhood safety is based on an item that asked respondents if they feel safe in their neighborhood all of the time, most of the time, some of the time, or none of the time. Two items were selected to indicate community-level resource availability as part of SES; both involve dichotomous measures of parks and open spaces. The first measures availability and asked respondents whether there is a park, playground, or open space within walking distance of their home. The second measures utilization and asked whether the respondent went to a park, playground, or public open space in the previous 30 days. The last SES variable measures access to health care based on whether or not the respondent had health insurance at the time of the survey.

Data Analysis

Because CHIS uses a complex sample design, it is necessary to adjust the standard errors. Replicate weights were used in all the analyses using the svy commands in STATA MP 11. We first examine bivariate differences in measures of health and SES by race/ethnicity/citizenship, using chi-square tests and analysis of variance (ANOVA) to examine statistical significance as appropriate. We next controlled for SES and demographic differences between the populations using logistic regression to estimate the net effects of race/ethnicity/citizenship for a series of health outcomes.

Results

Table 1 presents the bivariate results by race/ethnicity/citizenship for the Baby-Boom generation. The mean age and percent female is statistically equivalent across NHW, U.S.-born Mexicans, naturalized Mexican immigrants, and noncitizen Mexican immigrants. The percent not married or living with a partner is also similar across groups except for the lower rate among naturalized Mexican immigrants (16%), which is driven mostly by their lower rates of divorce (data not shown).

Table 1.

Socioeconomic Status and Health Status by Race/Ethnicity/Citizenship, Ages 43–61, California, 2007

	Non-Hispanic Whites (U.S. born)	Mexican (U.S. born)	Mexican (naturalized)	Mexican (noncitizen)
Unweighted sample size	12,469	1,082	587	585
Weighted population estimate	4,529,295	662,118	506,611	578,087
Age (mean)	52	51	50	49
Female (%)	51	48	49	46
Not married** (%)	24	27	16	23
Poverty***
300%+ FPL	78	56	29	6
200%–299% FPL	10	17	16	15
100%–199% FPL	7	14	35	41
0%–99% FPL	5	13	20	38
Education***
Some grade school/graduate degree (%)	19	6	2	1
Some college/college degree (%)	54	40	17	9
High school graduate (%)	22	37	24	15
Less than high school (%)	5	17	57	75
Home ownership***
Own (%)	80	70	76	49
Rent (%)	18	28	22	48
Other (%)	3	2	2	3
Feel neighborhood safe***
All of the time (%)	68	64	64	73
Most of the time (%)	29	30	24	14
Some of the time (%)	2	5	12	10
None of the time (%)	1	1	0	3
No park availability (%)	16	18	21	14
No park use past month (%)**	30	35	38	39
Currently uninsured (%)***	8	9	23	47
Diabetes (%)***	7	15	17	21
Obese (body mass index ≥ 30)***	24	43	36	28
High blood pressure**	29	35	30	22
Condition limits physical activity*	19	23	24	18
Fair/poor self-rated general health***	13	24	42	46

Notes: FPL = federal poverty line.

*p < .05. **p < .01. ***p < .001.

Bivariate Relationships Between Race/Ethnicity/Citizenship and SES

Chi-square and ANOVA tests indicated significant differences by race/ethnicity/citizenship for all the SES variables, except for park availability (Table 1).

The pattern is consistent across the majority of the SES variables with the exception of home ownership and neighborhood safety. U.S-born NHW report the best SES followed by U.S.-born Mexicans, naturalized Mexicans, and lastly noncitizen Mexicans. The most compelling differences were found for poverty and education. Of the NHW group, 5% lived below the poverty line compared to 13% of the U.S.-born Mexican group, 20% of the naturalized Mexican group, and 38% of the noncitizen Mexican group. The differences in education by race/ethnicity/citizenship are similar. Of the U.S.-born NHW group, 5% had less than a high school education compared to 17% of U.S.-born Mexicans, 57% of naturalized Mexicans, and 75% of Mexican noncitizens. Conversely, whereas 19% of NHW had some graduate education or a graduate degree, 6% of U.S.-born Mexicans, 2% of naturalized Mexicans, and 1% of Mexican noncitizens had some graduate education or a graduate degree. Home ownership was remarkably similar; 80% of the U.S.-born NHW group owned their own home, 70% of the U.S.-born Mexican group, and 76% of the naturalized Mexican group. The outlier was only 49% of the noncitizen Mexican group reported owning their home.

The differences in perceptions of neighborhood safety are small but statistically significant. Among U.S.-born NHW, 68% felt their neighborhood was safe all of the time, compared to 64% of U.S.-born Mexicans and naturalized Mexicans and 73% of non-citizen Mexicans. Although park availability was not significantly different, park utilization had small but statistically significant differences. Differences were also statistically significant but small for marital status. Lastly, about 8% of the U.S.-born NHW group and 9% of the U.S.-born Mexican group were not currently insured, compared to 23% and 47% of naturalized Mexicans and noncitizen Mexicans, respectively.

Bivariate Relationships Between Race/Ethnicity/Citizenship and Health Outcomes

All the health outcomes are statistically different across the groups at the bivariate level (Table 1). Although U.S.-born NHW have the best health across the majority of health measures, the pattern of worse health among Mexican-origin populations varies. For diabetes, the rate was the highest for noncitizen immigrant Mexicans (21%) and decreased with citizenship (17%) and U.S. nativity (15%). The trend was the opposite for obesity among the Mexican-origin populations, with immigrants having the lowest rate (28%) and U.S.-born Mexicans the highest (43%). High blood pressure follows the same trend. Although there were differences in CLPA, those differences are comparatively small. Noncitizen Mexicans had a CLPA rate similar to that of U.S.-born NHW (18% vs. 19%), whereas naturalized Mexicans and U.S.-born Mexican-origin persons had a similar rate (23% vs. 24%). SRH fair and poor health exhibits the largest differences, from a high of 46% among non-citizen Mexicans to 24% for U.S.-born Mexicans.

Multivariate Analysis of Health Outcomes by Race/Ethnicity/Citizenship and SES

A series of logistic regressions tested whether race/ethnicity/citizenship is associated with each of the health outcomes and whether the relationship between these variables and the outcomes remain once SES and other demographic controls are included in the model.

The unadjusted logistic regressions mirror the simple bivariate analysis, showing U.S.-born Mexicans have higher disease rates for all conditions, followed by naturalized Mexican immigrants, and Mexican noncitizens having the best pattern after U.S.-born NHW (Table 2, Model 1). Diabetes and fair/poor SRH show the strongest and most consistent disadvantage compared to U.S.-born NHW (the reference group). Mexican noncitizens have lower rates of high blood pressure than U.S.-born NHW, whereas U.S.-born Mexicans have higher rates. There is no difference between Mexican noncitizens and U.S.-born NHW for obesity and conditions that limit activities, but both are higher for naturalized Mexicans and U.S.-born Mexicans.

Table 2.

Logistic Regression Predicting Chronic Conditions, U.S.-Born Non-Hispanic Whites, and Mexican Americans Ages 43–61, California, 2007

	Diabetes	High blood pressure	Obesity	Condition limits activity	Fair/poor self-reported health
	OR	OR	OR	OR	OR
Model 1
Race/ethnicity/citizenship
Non-Hispanic White U.S. born (reference)
Mexican U.S. born	2.22***	1.29**	2.40***	1.32*	2.05***
Mexican naturalized	2.49***	1.05	1.78***	1.40*	4.59***
Mexican noncitizen	3.32***	0.70**	1.22	0.96	5.42***
Model 2
Race/ethnicity/citizenship
Non-Hispanic White U.S. born (reference)
Mexican U.S. born	1.86***	1.16	2.07***	0.8	1.28
Mexican naturalized	1.61*	0.91	1.35	0.61**	1.67***
Mexican noncitizen	1.93**	0.61**	0.86	0.28***	1.32
Female	0.77*	0.81***	0.83***	1.19*	0.99
Age	1.09***	1.08***	1.01.	1.06***	1.05***
Poverty
300% + FPL (reference)
200%–299% FPL	1.65*	1.25*	1.21	1.45***	2.05***
100%–199% FPL	2.55***	1.45***	1.10	2.91***	3.16***
0%–99% FPL	2.33***	1.25	1.16	4.73***	4.03***
Education
Some graduate or higher (reference)
Some college/college degree	1.79***	1.22*	1.56***	1.44***	1.50***
HS diploma	1.65***	1.71***	1.77***	1.93***	2.16***
Less than HS	2.25***	1.44*	2.07***	2.30***	4.11***
Homeownership
Own (reference)
Rent	0.99	1.12	1.04	1.34***	1.19.
Other arrangement	1.24	1.22	1.37	1.68*	1.79*
Not married	1.18	1.10	1.06	1.06	1.28*
Park not available	0.92	0.96	1.10	1.01	1.17
Park not utilized	1.36*	1.27***	1.41***	1.38***	1.43***
Neighborhood feels safe
All of the time (reference)
Most of the time	0.92	1.07	1.06	1.29***	1.20
Some of the time	0.94	1.40*	1.08	1.80***	2.04***
None of the time	1.40	2.62**	1.51	1.82	2.78***
Uninsured	0.71	0.67***	0.91	0.80	0.95

Notes: OR = odds ratio; FPL = federal poverty line; HS = high school.

*p < .05. **p < .01. ***p < .001.

Adjusting for population differences in demographics and SES attenuates the differences in health status unevenly (Table 2, Model 2). Diabetes and obesity remain almost twice as likely for U.S.-born Mexicans as U.S.-born NHW. The other three conditions—high blood pressure, conditions that limit activities, and SRH—no longer show differences between U.S.-born Mexicans and NHW after controlling for other variables. Diabetes remains higher for noncitizen Mexicans and naturalized Mexicans than U.S.-born NHW. Conditions that limit activities reverse direction, from being higher or no different for all Mexican-origin groups than U.S.-born NHW, to being lower or no different. And self-reported fair/poor health remains higher only for naturalized Mexicans.

As expected, higher poverty and low education levels are associated with worse health across the majority of health indicators and have the strongest effects of any of the variables in the model (see Table 2, Model 2). For example, being poor or near poor compared to living at 300% FPL or above greatly increase the odds of having diabetes, high blood pressure, a condition that limits activity, and fair/poor SRH. The only condition not associated with poverty among these Baby Boomers is obesity. Lower education is similarly associated with higher rates of health conditions, with most conditions increasing as education decreases. The other consistent predictor of worse health was not utilizing a park. Those that did not utilize a park for exercise/leisure reported worse health across all health outcomes compared to those who utilized a park in the past 30 days. Those that perceive their neighborhood as unsafe were nearly three times more likely to report having high blood pressure and fair/poor health. The other measures of SES, homeownership and being uninsured, had minimal impact on the health outcomes in this analysis.

Discussion

The cumulative inequality framework predicts that inequalities experienced at young ages, and throughout the life course, will have a negative impact on the health and well-being of the Mexican-origin population as they age (Ferraro & Shippee, 2009). The Mexican-origin Baby-Boom generation has experienced low levels of education in youth and continued low levels of income in middle-age presage poor outcomes in older ages. Differences between U.S.-born Mexicans and the U.S.-born NHW in California become a gulf when comparing noncitizen Mexican immigrant Baby Boomers with U.S.-born NHW. The income and education gap for immigrants can be explained in part by the segmented labor market that creates a demand for low-skilled immigrant labor to fill low-wage jobs. The lower educational level of U.S.-born Mexicans compared to U.S.-born NHW is related to worse educational opportunities in childhood and barriers to advanced education (Rodriguez, 2004; Wallace, in press). Both education and location in segmented job markets occur early in life but become persistent sources of inequality throughout the life course. It is likely that Mexican-origin Baby Boomers will continue to have lower income and education levels when they become older adults, both of which will continue to have an impact and result in poorer health status. The lower educational attainment, income, and health insurance experienced by Mexican immigrant Boomers in California will almost surely persist into old age. The strong effects of education and income on the chronic conditions examined means that this coming generation of Hispanic elders will continue to suffer high rates of chronic and disabling conditions as a result of the life-long effects of lower SES.

In the unadjusted model that only includes race/ethnicity/citizenship status, U.S.-born NHW have better health across all outcomes. The unadjusted model finds U.S.-born Mexican Baby Boomers with the most consistent disparity in health status across outcomes and noncitizen Mexican immigrants with the best pattern among the Mexican subgroups, consistent with the healthy immigrant effect. The pattern changes when demographic and SES controls are added, with diabetes being the only condition that remains higher for all three categories of Mexican-origin Baby Boomers compared to their U.S.-born White peers. Because diabetes is a risk factor for heart disease, other circulatory problems, eye disease, and kidney problems, especially when poorly managed (Beard et al., 2009), the already high rates of diabetes pose a particular challenge for the Hispanic Baby-Boom generation as it ages and develops cumulative diabetes-related problems.

Controlling for SES reduces the relative odds of all five health outcomes for all the Mexican-origin groups relative to U.S.-born NHW. Immigrants have the largest effects, with naturalized Mexicans changing from higher rates than U.S.-born NHW for four health outcomes in the unadjusted model to only two in the adjusted model, condition that limits activity went from higher to lower. The higher rates among U.S.-born than immigrant Mexican Baby Boomers may be due to the U.S. born not benefiting from the “health capital” that immigrants often bring, while experiencing a lifetime of discrimination and stress associated with being Hispanic (Guimarães, 2007). The literature on the Hispanic health paradox also identifies that some immigrants return to Mexico when they become ill, which is unlikely among U.S.-born Hispanics, and culturally linked healthy behaviors among immigrants that decline with acculturation (Palloni & Arias, 2004). The multiple, potential health advantages of immigrants largely disappear by the second generation, allowing the disadvantages of lower SES to have their full effect. Naturalization is associated with time in the United States (Passel, 2007), which likely explains much of the smaller advantages of the naturalized Mexicans in our analysis (Markides & Eschbach, 2005). This also implies that the relative health advantages exhibited by noncitizen Mexican immigrants are likely to decline as they age in the United States and become elderly Baby Boomers. This is consistent with the cumulative inequality framework that would identify culturally linked health behaviors as moderators of inequality-driven health problems, together with other human and social network resources that immigrants bring that are not present to the same extent among U.S.-born Boomers of Mexican origin.

Although health insurance is associated with access to health care and better health status among adults (McWilliams, 2009), the net effect of having no health insurance is either not significant or is protective in our sample for the target conditions. The most likely explanation is that those without health insurance primarily seek episodic care for acute conditions and are therefore less likely to know that they have a chronic condition. For example, being uninsured and Hispanic are both associated with undiagnosed diabetes nationally (Danaei, Friedman, Oza, Murray, & Ezzati, 2009). Rather than being at lower risk, having lower rates of diagnosis potentially means that those without health insurance will wait to seek care until their condition is more advanced and has a greater harm on their health (McWilliams, 2009). In this case, being uninsured over a lifetime can lead to undiagnosed chronic diseases in middle age and more severe complications from those conditions in older ages. Thus, the lack of association between current health insurance and current health status may presage worse health at older ages for Hispanics because they have exceptionally high rates of uninsurance.

Baby Boomers are only now entering the ages of eligibility for Medicare and Social Security. This analysis of Mexican-origin Baby Boomers reveals that federal entitlement programs will continue to be an important safety net for this population. Their high poverty rates, low education rates, and poor health suggest that this is not a population that has adequate economic resources going into old age, nor will they all be healthy enough to work beyond the current retirement age in order to supplement their income. Further, the low educational levels suggest that Mexican-origin Boomers will continue to be found in low-paying jobs that often lack retiree benefits. Proposals to raise the eligibility age for Medicare will extend the years of uninsurance for Mexican-origin Boomers, risking additional years of undetected and untreated chronic conditions. This low-income population also does not generally have the resources to pay for better coverage under voucher plans for Medicare and will not fare well due to their low education and incomes under proposals for Social Security privatization or raised retirement ages.

Conclusion

Our analysis documents health and socioeconomic disparities between U.S.-born NHW and Mexican-origin populations in the Baby-Boom cohort of Californians. The most consistent disparities that are not mediated by SES are among U.S.-born Boomers of Mexican origin. Much, but not all, of the health disparities are driven by SES differences in the populations, especially education and income. Immigrants do better than U.S.-born Mexicans and in some cases better than U.S.-born NHW, but there is evidence that this immigrant advantage declines over time. Baby Boomers have already finished their formal education and are established in occupations that are not likely to lead to significant upward changes in wealth before retirement. This means that the SES disparities that contribute to health disparities are unlikely to change without intervention by social policies at the same time as the immigrant health advantages decline. Following the cumulative inequality framework (Ferraro & Shippee, 2009), we can expect that current patterns of disparities between the Mexican-origin and U.S.-born NHW population will continue to worsen as the Boomer generation moves into old age. Although this analysis is just for Mexican-origin Baby Boomers in California, the state is a microcosm of patterns that are likely to be found elsewhere in the country.

In the current economic context, it is unlikely that there will be new initiatives to reduce poverty among Baby Boomers, now or in their older adult years. And in the current immigration policy context, it is unlikely that there will be new initiatives to improve the pay or working conditions of immigrants. This means that many Mexican American and Mexican immigrant Boomers will enter their later years of life with two strikes against them—the health risks accumulated over a lifetime of low SES, compounded by the additional risks created by being of Mexican origin. Weakening the safety net of income and health programs for seniors would be a third strike, for the current cohort of Mexican-origin Boomers is significantly less likely than U.S.-born NHW to have the means to pay extra for health insurance and other basic expenses when they need it the most. Policy discussions regarding changing Social Security and Medicare benefits or eligibility must take into account the diversity of the Baby-Boom generation and the potential risk these changes may pose for exacerbating the SES and health disparities that already exist.

Funding

The National Institute of Mental Health (grant R21MH080624), supported M.P.A. work on the manuscript. S.P.W. received support from the UCLA Resource Centers for Minority Aging Research Center for Health Improvement of Minority Elderly under National Institutes of Health/National Institute on Aging (grant P30-AG021684), as well as the National Institute of Environmental Health Sciences (NIEHS) (grant RC1ES018121) and the Centers for Disease Control (CDC) (grant U58DP002952).