1. HISPANIC/LATINO PARADOX IN ADULT MORTALITY: AN UPDATE Kyriakos S. Markides, Ph.D. University of Texas Pan American February 26, 2010

2. AN EPIDEMIOLOGIC PARADOX Hispanics (except Cuban Americans) are socioeconomically disadvantaged, but have favorable overall mortality Hispanics (except Cuban Americans) are socioeconomically disadvantaged, but have favorable overall mortality Markides and Coreil (1986) Risk factor profiles High rates of DIABETES High rates of DIABETES High rates of OBESITY High rates of OBESITY Similar rates of hypertension, cholesterol Similar rates of hypertension, cholesterol High SMOKING rates among men, lower among women (fewer cigarettes). Cuban American males smoke the most High SMOKING rates among men, lower among women (fewer cigarettes). Cuban American males smoke the most High ALCOHOL (binge) drinking rates among men, low among women. Alcohol consumption in women increases with acculturation High ALCOHOL (binge) drinking rates among men, low among women. Alcohol consumption in women increases with acculturation Low rates of physical ACTIVITY Low rates of physical ACTIVITY Strong families Strong families Migration selection Migration selection

3. AGING, MIGRATION AND MORTALITY: CURRENT STATUS OF RESEARCH ON THE HISPANIC MORTALITY PARADOX Data based on Vital Statistics show the greatest mortality advantage compared to Non-Hispanic Whites for all Hispanics combined. The advantage is greatest among older people. Data based on Vital Statistics show the greatest mortality advantage compared to Non-Hispanic Whites for all Hispanics combined. The advantage is greatest among older people. National Community Surveys linked to the National Death Index show a narrowing of the advantage and one study suggests that the Mexican origin mortality advantage (Palloni & Arias, 2004) can be attributed to selective return migration of less healthy immigrants to Mexico. National Community Surveys linked to the National Death Index show a narrowing of the advantage and one study suggests that the Mexican origin mortality advantage (Palloni & Arias, 2004) can be attributed to selective return migration of less healthy immigrants to Mexico. The Medicare – NUDIMENT data show a much lower advantage of Hispanic elders than the Vital Statistics Method. The Medicare – NUDIMENT data show a much lower advantage of Hispanic elders than the Vital Statistics Method. Markides & Eschbach, J. Gerontology: Social Sciences (2005) Social Sciences (2005)

4. CONCLUSION Markides & Eschbach, 2005 The majority of the evidence continued to support a mortality advantage at a minimum for Mexican Americans. Greatest advantage is in old age. The majority of the evidence continued to support a mortality advantage at a minimum for Mexican Americans. Greatest advantage is in old age. Self-reports of health status in old age do not support an advantage. Self-reports of health status in old age do not support an advantage. Suggested that older Mexican Americans live longer with more disability. Data not yet available on trends in disability. Suggested that older Mexican Americans live longer with more disability. Data not yet available on trends in disability. Greatest challenge was Palloni & Arias’ (2004) suggestion of a “salmon bias”. Greatest challenge was Palloni & Arias’ (2004) suggestion of a “salmon bias”.

5. RECENT EVIDENCE OF A SALMON BIAS Turra & Elo (2008) used the Medicare- NUDIMENT data to examine the existence of a salmon bias. Turra & Elo (2008) used the Medicare- NUDIMENT data to examine the existence of a salmon bias. Data supported a salmon bias: foreign-born social security beneficiaries living abroad had higher mortality rates than foreign-born beneficiaries living in the U.S.. Too small to explain mortality advantage. Data supported a salmon bias: foreign-born social security beneficiaries living abroad had higher mortality rates than foreign-born beneficiaries living in the U.S.. Too small to explain mortality advantage. Effect of salmon bias on death rates is partially offset by the high mortality of Hispanic emigrants returning to the U.S. Effect of salmon bias on death rates is partially offset by the high mortality of Hispanic emigrants returning to the U.S.

6. A DIFFERENT TEST OF THE SALMON BIAS Hummer Colleagues examined infant mortality rates among Hispanics by nativity and in comparison to non-Hispanic whites. Hummer Colleagues examined infant mortality rates among Hispanics by nativity and in comparison to non-Hispanic whites. They found that first hour, first day and first week mortality rates among infants born in the U.SA. To Mexican immigrant women are about 10% lower than those of infants of U.S. born non-Hispanic white women. They found that first hour, first day and first week mortality rates among infants born in the U.SA. To Mexican immigrant women are about 10% lower than those of infants of U.S. born non-Hispanic white women. It is unlikely that such favorable rats are the result of out-migration of Mexican origin women and infants It is unlikely that such favorable rats are the result of out-migration of Mexican origin women and infants

7. EVIDENCE FROM MHAS (Wong and Colleagues) While there is considerable return migration back to Mexico, MHAS data show that the vast majority of return migrants are younger While there is considerable return migration back to Mexico, MHAS data show that the vast majority of return migrants are younger Very few older people return to Mexico because their children live in the U.S. Very few older people return to Mexico because their children live in the U.S.

8. THE SES GRADIENT Turra & Goldman (2007) National Health Interview (NIH) 1989-94 linked the National Death Index (NDI) through 1997. Turra & Goldman (2007) National Health Interview (NIH) 1989-94 linked the National Death Index (NDI) through 1997. Differences in mortality by education were smaller for Hispanic groups than for non- Hispanic whites. Differences in mortality by education were smaller for Hispanic groups than for non- Hispanic whites. Mortality advantage was greater among foreign- born than the native-born especially in old age. Mortality advantage was greater among foreign- born than the native-born especially in old age. “The Hispanic mortality advantage pertains primarily to persons of lower SES” “The Hispanic mortality advantage pertains primarily to persons of lower SES”

9. MORTALITY AT YOUNGER AGES (Eschbach and Colleagues, 2007) Study used Texas and California vital registration data from 1999-2001 linked to 2000 census population data. Study used Texas and California vital registration data from 1999-2001 linked to 2000 census population data. Focus on ages 15 to 44 where little attention had been paid and where consistent advantages have not always been found. Focus on ages 15 to 44 where little attention had been paid and where consistent advantages have not always been found. Mortality advantage confined to foreign-born Hispanics. Advantages primarily attributed to social and behavioral causes (substance abuse, HIV, suicide) Mortality advantage confined to foreign-born Hispanics. Advantages primarily attributed to social and behavioral causes (substance abuse, HIV, suicide)

10. ASCERTAINMENT OF HISPANIC ETHNICITY ON CALIFORNIA DEATH CERTIFICATES (Eschbach and Colleagues, 2006) Under ascertainment was lower among foreign born Hispanics. After appropriate adjustments foreign-born Hispanics exhibited a significant mortality advantage of 25% to 30% compared to non-Hispanic Whites. No advantage among native-born Hispanics. Under ascertainment was lower among foreign born Hispanics. After appropriate adjustments foreign-born Hispanics exhibited a significant mortality advantage of 25% to 30% compared to non-Hispanic Whites. No advantage among native-born Hispanics.

11. “IS THE HISPANIC MORTALITY ADVANTAGE A DATA ARTIFACT” (Arias, Eschbach, Schauman et al (in press)) National Longitudinal Mortality Study data showed that Hispanic origin reporting on death certificates is quite good. National Longitudinal Mortality Study data showed that Hispanic origin reporting on death certificates is quite good. Correction for death certificate misclassification had only a small effect on death rates with the Hispanic age-adjusted death rate increasing from 79 to 83 percent of the non-Hispanic white rate. Foreign- born Hispanics had higher rates of correct classification than U.S.-born Hispanics giving added credence to the immigrant advantage in mortality compared to the native-born. Correction for death certificate misclassification had only a small effect on death rates with the Hispanic age-adjusted death rate increasing from 79 to 83 percent of the non-Hispanic white rate. Foreign- born Hispanics had higher rates of correct classification than U.S.-born Hispanics giving added credence to the immigrant advantage in mortality compared to the native-born.

12. OVERALL IMMIGRANT ADVANTAGE (Singh & Hyatt, 2006) Immigrant mortality advantage not confined to Hispanics. There appears to be an overall immigrant advantage which may have increased in recent years. Immigrant advantage was evident for cardiovascular diseases, major cancers, diabetes, respiratory diseases, suicide, and unintentional injuries. These trends due to growing heterogeneity of immigrant population, continuing advantages in behavioral characteristics, and migration selectivity. Immigrant mortality advantage not confined to Hispanics. There appears to be an overall immigrant advantage which may have increased in recent years. Immigrant advantage was evident for cardiovascular diseases, major cancers, diabetes, respiratory diseases, suicide, and unintentional injuries. These trends due to growing heterogeneity of immigrant population, continuing advantages in behavioral characteristics, and migration selectivity. Asian/Pacific Islanders had the highest life expectancy followed by Hispanics and non-Hispanic Whites. For each ethnic origin, there was an immigrant advantage except for Asian/Pacific Islanders which likely reflects compositional differences between the native-born and immigrants (Markides & Colleagues, 2007) Asian/Pacific Islanders had the highest life expectancy followed by Hispanics and non-Hispanic Whites. For each ethnic origin, there was an immigrant advantage except for Asian/Pacific Islanders which likely reflects compositional differences between the native-born and immigrants (Markides & Colleagues, 2007)

13. IMMIGRANT ENCLAVES Osypuk, Diez Roux, Hadley & Kandula (2009) used data from the Multi-Ethnic Study of Atherosclerosis in four U.S. cities (New York, Los Angeles, St Paul and Chicago). They found that high neighborhood immigrant concentration was associated with lower consumption of high fat foods among Hispanics and Chinese but also less walk ability, fewer recreational exercise resources, worse safety, and other negative characteristics. Osypuk, Diez Roux, Hadley & Kandula (2009) used data from the Multi-Ethnic Study of Atherosclerosis in four U.S. cities (New York, Los Angeles, St Paul and Chicago). They found that high neighborhood immigrant concentration was associated with lower consumption of high fat foods among Hispanics and Chinese but also less walk ability, fewer recreational exercise resources, worse safety, and other negative characteristics.

14. HISPANIC PARADOX IN BIOLOGICAL PROFILES (Crimmins & Colleagues, 2007) Use NHANES (1999-2002) to compare blood pressure, metabolic, and inflammatory risk profiles. Use NHANES (1999-2002) to compare blood pressure, metabolic, and inflammatory risk profiles. After controls for SES differences U.S. born Mexican Americans had higher biological risk scores than did non-Hispanic Whites and foreign-born Mexican Americans. There were no differences between foreign-born Mexican Americans and non-Hispanic Whites leading some to the notion of migration selectivity and superior health behaviors of immigrants. After controls for SES differences U.S. born Mexican Americans had higher biological risk scores than did non-Hispanic Whites and foreign-born Mexican Americans. There were no differences between foreign-born Mexican Americans and non-Hispanic Whites leading some to the notion of migration selectivity and superior health behaviors of immigrants.

15. A TEST OF THE ”HEALTHY MIGRANT EFFECT” USING DATA FROM THE MEXICAN FAMILY LIFE SURVEY (Rubalcava, Teruel, Thomas et al, 2008) 6446 respondents aged 15-29 in 2002. Some moved to the U.S. during the next three years. 6446 respondents aged 15-29 in 2002. Some moved to the U.S. during the next three years. Health significantly predicted migration among females and rural males. However associations were weak with considerable variation in the estimates between males and females and between urban and rural dwellers. Health significantly predicted migration among females and rural males. However associations were weak with considerable variation in the estimates between males and females and between urban and rural dwellers. Note: Small numbers of migrants might be a problem: 113 rural men; 87 urban men; 90 rural women, and 52 urban women Note: Small numbers of migrants might be a problem: 113 rural men; 87 urban men; 90 rural women, and 52 urban women

16. CONCLUSIONS/NEXT STEPS Migration selection still most viable explanation Migration selection still most viable explanation A Healthy migrant effect is found in most immigrant groups to the U.S., but also to Canada and Austria. A Healthy migrant effect is found in most immigrant groups to the U.S., but also to Canada and Austria. Salmon bias exist but has minor effects. Salmon bias exist but has minor effects. Immigrant communities need search for mechanisms. Immigrant communities need search for mechanisms. Low class has different meanings for Mexican Immigrants than for non-Hispanic Whites and African Americans. Low class has different meanings for Mexican Immigrants than for non-Hispanic Whites and African Americans. Increasing rates of obesity and diabetes including in old age. Hispanic EPESE data suggest increases in the prevalence of diabetes, disability, and cognitive impairment in very old Mexican Americans. Increasing rates of obesity and diabetes including in old age. Hispanic EPESE data suggest increases in the prevalence of diabetes, disability, and cognitive impairment in very old Mexican Americans. Need closer examination of changes with time in the U.S. in health behaviors, diet, and obesity. Need closer examination of changes with time in the U.S. in health behaviors, diet, and obesity.