(June 2013) Despite advances in health care and increases in income over the past 50 years, significant gaps in life expectancy and health by race and ethnicity persist among older Americans. This newsletter highlights recent work by National Institute on Aging (NIA)-supported researchers and others who examined life expectancy and health trends among older blacks and Hispanics. By 2030, the U.S. elderly population is expected to become more racially and ethnically diverse than it is today. Understanding their differences in health and addressing disparities are critically important to improving the nation’s overall health and well-being.

Life Expectancy

Disparities in life expectancy between older minorities and their non-Hispanic white counterparts are well documented. But the nature and extent of these differences may surprise some people. According to the National Center for Health Statistics (NCHS), life expectancy for blacks—whether at birth, at age 65, or at age 75—tends to be lower than for non-Hispanic whites. For example, a 65-year-old African American male in 2010 was expected to live another 15.9 years, compared with 17.7 years for a white man at that age (see table). A black woman age 65 is expected to live another 19.3 years, a full year less than a white woman of that age.

Yet while older blacks have lower life expectancies than older whites, Hispanics actually are expected to live longer— again, both at birth and at older ages. At age 65, for example, Latino males can expect to live, on average, an additional 18.8 years, while Latina women have a life expectancy of an additional 22 years—almost two years more than 65-year-old white females. Similar differences can be found at age 75 for both men and women.

Recent research suggests that life expectancy among racial and ethnic groups may not be converging as quickly as some expected. Montez and colleagues (2011) examined life expectancy trends, documenting the widening gap between lower- and higher-educated adults between 1986 and 2006, focusing on blacks and non-Hispanic whites. They argued that this expanding disparity by education level does not mean that policy initiatives to reduce socioeconomic differences in health failed. Rather, they suggest other factors played a stronger role, particularly differences in income and tobacco use, but also neighborhood living environments, access to medical care, job fulfillment, occupational hazards, psychosocial resources, and the availability and sophistication of health-related information.

Olshansky and colleagues (2012) also examined life expectancy trends and found that the overall gap in life expectancy at birth shrank between blacks and whites but increased between Hispanics and non-Hispanic whites between 1990 and 2008. Contributing to these shifts were declines in life expectancy among whites without high school degrees. When the researchers explored the impact of race and education together, they found stark differences:

In 2008, white men and women with 16 or more years of education had life expectancies at birth that were more than a decade longer than for black men and women with less than 12 years of education. Even among males with 16 or more years of education, the gap between white and black men was about five years. They called the disparities in life expectancy by race/ethnicity and education levels “alarming” and the role of education “profound.” The researchers argued that when it comes to life expectancy, there are at least “two Americas” divided along lines of race/ethnicity and educational attainment. They urge policymakers to improve and expand education opportunities for people of all ages, including lifelong learning programs for older adults.

Black-White Differences Explored: A body of research suggests that socioeconomic status (education and income levels) accounts for much—but not all—of the racial disparity in life expectancy. To quantify the influences, Geruso (2012) explored what black life expectancy would be if blacks shared the same socioeconomic characteristics (income, education, employment, and occupation) as the white population as a whole. His analysis suggests that socioeconomic differences explain 70 percent of the life expectancy gap between black and white females and 80 percent of the gap for males. Probing further, he found that differences in education and income levels are the most important explanatory factors.

Blacks tend to have lower life expectancy than whites until about age 85, after which blacks tend to outlive their white peers (a phenomenon called racial crossover). Jackson and colleagues (2011) suggest that discrimination and chronic stress may lead to increased mortality among blacks beginning at midlife, leaving those who reach older ages hardier than whites of similar ages. Compared to whites, U.S. blacks experience more hardship and trauma early in life, as well as stressors related to their disadvantaged social status (discrimination, residential segregation). Among blacks, these stressors build up over the years, contributing to earlier onset of disease (particularly cardiovascular disease and Type 2 diabetes) and leading to lower life expectancy throughout early old age.

Living in a disadvantaged neighborhood also may limit life expectancy. Focusing on mortality before and after age 80, Yao and Robert (2011) examined national data on blacks and whites from the Americans’ Changing Lives study and census tract information. They found that older adults who lived in more disadvantaged neighborhoods (mainly measured by high levels of poverty and low levels of education) had a higher mortality rate, even when individual income was taken into account. Compared to their white peers, black older adults have higher risks of dying before age 80, reflecting the impact of neighborhood and individual socioeconomic status. They argue that both individual and neighborhood socioeconomic disadvantages must be addressed to reduce the mortality disparities evident in old age.

Poor quality health care in predominantly black communities also may contribute to differences in life expectancy differences (Bach et al. 2004; Skinner et al. 2005). Researchers have found that doctors and hospitals that treat a disproportionate share of black patients tend to be less qualified and to have access to fewer resources than those that treat mainly white patients. (Only limited evidence suggests individual health providers treat black and white patients differently.) Analysis by Deaton and Lubotsky (2009) supports the explanation of differing quality of care by location. They examined data for U.S. cities and found a link between mortality and income inequality that virtually disappeared once they took into account the proportion of black residents. This finding suggests that whites who live in cities with large black populations also receive inferior care, affecting mortality rates.

Causes of Death: Despite differences in life expectancy, the leading causes of death among Americans ages 65 and over are very similar across the major racial and ethnic groups. For example, heart disease and cancer were the two leading causes of death in 2010 among older whites, blacks, and Hispanics, according to NCHS data. Stroke was the third leading cause of death for older blacks and Hispanics, while chronic respiratory diseases ranked third among older whites. (Stroke was fourth.) Further down the list, there were some differences. For example, diabetes was the fourth leading cause of death among both African American and Hispanic elderly, while Alzheimer’s disease was the fifth leading cause of death among older whites (both men and women). And the sixth leading cause of death among older whites, accidents, did not rank among the top 10 causes for older blacks. There were not many gender differences within each racial group: Slightly more older black males died from cancer in 2010 than from heart disease, while Alzheimer’s disease was the fifth leading cause of death among older black women. (It ranked only ninth among older black men.)


Most researchers agree that socioeconomic status, access to health care, and health behaviors interact in complex ways to create racial and ethnic health disparities. While there is no consensus on which factors matter the most, there is some general agreement on the main causes. These include education, income, health-risk behavior, psychosocial factors (especially stress), access to and quality of health care, culture, genetic factors, and environmental and occupational risks.

Deaton (2011) explored the dynamics that produce racial differences in overall health, highlighting the consequences of living in a racist society, the relatively poorer quality of hospitals that predominantly serve African Americans, and environmental factors (as measured by relative proximity to sites judged to be toxic or hazardous). Deaton also speculated that three other factors could account for the poorer health status of some older minority groups: income disparities caused by unhealthy behaviors; a propensity toward more physically demanding jobs, which causes great stresses on the body; and increased income inequality, by which persons with wealth often are less likely to support initiatives to improve health conditions of people who are less financially secure.

Diez Roux and colleagues (2009) examined the role of accumulated stress, exploring the biological mechanisms that underlie racial and ethnic health differences among older people. After adjusting for income, education, and a variety of behavioral factors (including smoking and physical activity), they found that the “biological aging” process occurs faster for blacks and Hispanics than for whites. They studied telomere length, a genetic measure of stress on a person’s body. While they found marked differences in telomere length among adults, the length did not differ among black, white, and Hispanic newborns, suggesting that racial differences in telomere length emerge and increase with age. Similarly, Geronimus and colleagues (2010) focused on black and white middle-age women (ages 49 to 55) and found that such factors as health behaviors (smoking, for example) and perceived stress (resulting from such conditions as economic hardship and discrimination) led black women to age biologically faster than white women, even when controlling for socioeconomic status.

Using data from the Americans’ Changing Lives study, Yao and Robert (2008) found that racial and socioeconomic differences in self-reported health persist into old age and that among older African American adults, self-reported health declines more rapidly over time than that of older white adults. While individual and neighborhood socioeconomic factors account for much of the initial difference, they do not explain the more rapid decline in self-reported health among older blacks.

Pereira and colleagues (2011) probed black-white differences further. Using data from the National Health Measurement Study, they report that among persons ages 65 to 89, white men and women had higher health-related quality of life scores than black men and women, respectively. Moreover, elderly black men and women were far more likely to report poor health, but only older black women were more likely to report more than one chronic medical condition. The fact that black men are less likely to seek routine medical care (and therefore are less likely to have health conditions diagnosed) may contribute to the latter result.

Cancer Trends: The CDC’s National Program of Cancer Registries reported cancer-incidence data for all states except Wisconsin in 2009. Data showed that the incidence of cancer (adjusted for age) was higher for older whites than it was for minorities in 2009. For example, the age-adjusted cancer rate for elderly African Americans was 3 percent lower than that for whites, while the incidence rate for older Latinos was 24 percent lower. The overall pattern was true for both men and women, although the age-adjusted cancer incidence rate for older black males was slightly higher (6 percent) than it was for white males.

Depression: Steffens and colleagues (2009) examined depression among persons ages 70 and over by race, drawing on a nationally representative sample that included people with cognitive impairment (a milder condition than dementia). They found that the prevalence of depression was about equal in men and women, but it was significantly higher (up to three times) among older whites and Latinos than among older African Americans. Other studies that do not include cognitively impaired people tend to find higher rates of depression among African Americans than whites. George and Lynch (2003) examined black and white adults ages 65 and older and found a relationship between stress levels (measured by instances of death and serious illness among family members and close friends) and increases in depressive symptoms. The older African Americans studied experienced more stress and more depressive symptoms than their white counterparts. Additionally, a study by Spence, Adkins, and Dupre (2011) that examined black and white women ages 52 to 81 found consistently higher levels of depressive symptoms among black women. Their analysis suggests that differences in physical health and socioeconomic status explain the racial gap.

Cognitive Functioning: Researchers increasingly view low formal education levels as a risk factor for developing Alzheimer’s disease, and there is some evidence that educational quality also may play a role (see Mehta et al. 2009 for summary). Using data from the Aging, Demographics, and Memory Study, Mehta and colleagues (2009) found a link between self-reported “below average” school performance and increased odds of developing Alzheimer’s disease among whites, blacks, and Hispanics. They argue that both quantity and quality of education may be key to building the cognitive capacity (known as cognitive reserve) that adults may draw on if they begin to suffer mental decline. “Not only is ‘staying in school’ important,” they write, “but increasing school performance is equally important” … as a “possible protective mechanism to stave off the onset of Alzheimer’s.”

Studies on the prevalence of dementia and cognitive impairment have found a great deal of variability in rates, sometimes finding black rates similar to white rates and at other times finding black rates higher than white rates (see Potter et al. 2009 for summary). While differences between blacks and whites may reflect differences in education levels and socioeconomic status, Potter and colleagues (2009) suggest that the varying prevalence rates documented in studies may reflect cultural differences as well. Using data from the Aging, Demographics, and Memory Study, they compared direct tests of cognitive function, reports by family members, and clinical diagnoses. Both blacks and whites perceived that their older family members were having cognitive difficulties, and their reports predicted a clinical diagnosis of dementia. However, whites’ reports also tended to predict the emergence of cognitive impairment in their older family members, but blacks’ reports did not. This suggests that family members’ predictions of early cognitive changes may be influenced by cultural differences in perception and reporting, including that whites may be more inclined to report early signs of impairment, that blacks may be less inclined to report these changes, or that both patterns exist.

Another recent study found that most of the differences in cognitive function scores among whites, blacks, and Hispanics could be explained by differences in both educational attainment and current physical activity levels (Masel, Faji, and Peck 2010). Masel and colleagues used data from the Health and Retirement Study to examine adults ages 51 to 61 and found that higher levels of education and leisuretime physical activity (but not housework) were associated with higher cognitive test scores. The researchers suggest that programs promoting cognitive training and physical activity be examined to assess their impact on reducing the onset of cognitive decline.

Disability: Disabilities tend to be more prevalent among elderly in minority groups than among their non-Hispanic white counterparts. Preliminary estimates from the National Health and Aging Trends Study (NHATS) find that blacks and Hispanics are more likely than non-Hispanic whites to be among the 10 percent of Medicare enrollees ages 65 and older who became disabled in some way before their 65th birthday (Spillman 2012). A study examining disability trends based on National Health Interview Survey data from 1982 to 2002 showed that the difference between the disability rates of minorities and non-Hispanic whites narrowed over the 20-year period, from a 10 percentage point difference to about a 5-point gap in 2002 (Schoeni, Freedman, and Martin 2009). The same study found that differences in disability levels by socioeconomic status and education widened during the period.

Liang and colleagues (2010) examined differences in how disabilities emerge and worsen during the aging process. Using data from the Health and Retirement Study (among other data), they measured levels of physical functioning based on the need for assistance with personal activities (such as bathing, dressing, eating, walking, or getting in and out of bed) and with household tasks (preparing hot meals, grocery shopping, managing money, taking medication, or using a telephone). They found that older blacks and Latinos tended to experience more rapid functional declines than non-Hispanic whites as they aged. When they controlled for factors such as socioeconomic status, baseline health, and marital status, white-black differences persisted while white-Latino differences were all but eliminated. Their findings suggest that other factors—such as neighborhood characteristics, lifestyle, and discrimination—may play a role in older blacks’ level of disability and contribute to the pace of their physical decline as they age.

Disability at older ages is closely linked to lifelong health. In an analysis based on the Health and Retirement Study, Hayward and Montez (2012) found that both poor health and low socioeconomic status during childhood each increase the likelihood that older blacks and non-Hispanic whites would have a disabling condition later in adulthood. But they also found that higher levels of educational attainment appeared to offset the impact of low childhood socioeconomic status, thereby decreasing the likelihood of having a disability in later life. The researchers suggest that these findings reflect the powerful role education plays in enabling individuals to overcome childhood disadvantages. They argue that education policy should be part of public policy strategies to limit disability, improve health, and reduce disparities. Additionally, a study of older African Americans not living in nursing homes found that having symptoms of major depression or two or more medical conditions significantly increased the odds of having a disability (Thorpe et al. 2011). The researchers suggest that chronic disease prevention efforts should target African Americans in midlife in order to slow or stop the progression of mobility problems that worsen in old age.

Freedman and colleagues (2009) found racial and ethnic disparities in the use of assistive technology and devices (walkers, scooters) and environmental modifications (grab bars, bath seats, or ramps) that allow people to function independently. Use of assistive technology is more widespread among non-Hispanic whites than among other groups. Minorities and those with lower socioeconomic levels were more likely to rely on someone else for help. The researchers suggest that public information campaigns and more widespread access could increase the use of assistive technology and devices by blacks and Hispanics, narrowing this gap, promoting independence, and improving quality of life.

End-of-Life Health Care Decisions: Compared to whites, blacks are more likely to distrust the health care system and hospice, to request aggressive care at the end of life, and to express discomfort discussing death (Carr 2011; Johnson, Kuchibhatla, and Tulsky 2008). The Health and Retirement Study asked respondents to make treatment choices for a hypothetical person, describing a woman in her 80s with cancer and asking what decisions respondents would make given different probabilities of survival and out-of-pocket costs (Chao, Pagan, and Soldo 2008). Older adults opted for treatment when costs were covered by Medicare, even if chances for survival were low. Blacks were more likely to choose to continue treatment regardless of the likelihood of survival or cost. Hispanics were more likely than whites to recommend opting out of costly end-of-life care, particularly if it would deplete savings and impoverish a surviving spouse.

Blacks and Hispanics are significantly less likely to complete advance directives (writing living wills, selecting a durable power of attorney for health care) than whites, even when socioeconomic differences are taken into account (Carr 2011). Compared to whites, Hispanics are less likely to believe that illness is a burden on their families and more likely to prefer family-centered decisionmaking (including informal discussions) rather than formal legal plans. Blacks and Hispanics’ limited use of advance directives has been linked in part to the religious belief that God alone determines the timing and nature of death. Carr explains that “legal documents specifying one’s medical treatment preferences may be deemed irrelevant, undesirable, or as intruding on God’s plan.” Patients who have an advance directive are more likely to receive the level of care they prefer at death; end-of-life Medicare spending was lower among patients with advance directives, particularly in areas of the country with high Medicare expenditures (Nicholas et al. 2011). These racial and ethnic differences will have growing implications for Medicare spending as minorities make up an increasing share of the older population.

Kelvin Pollard is a senior demographer at the Population Reference Bureau; Paola Scommegna is a senior/writer editor at PRB.


Please refer to PDF of report.