Incidence of Diabetes and Cardiovascular Disease in Mexican Americans

BACKGROUND:

The San Antonio Heart Study 1 was a population-based survey of diabetes and cardiovascular risk factors in Mexican Americans and non-Hispanic whites conducted from 1979 to 1982. Participants included all men and non-pregnant women ages 24 to 64, who resided in households randomly sampled from three socio-culturally distinct neighborhoods of San Antonio. The low-income barrio residents were almost exclusively traditional Mexican-Americans. The middle-income neighborhood was ethnically balanced with equal numbers of Mexican Americans and Anglos whereas the upper income suburb was approximately 90 percent Anglo. Risk factors measured included obesity, glucose intolerance, hypertriglyceridemia, low levels of high density lipoprotein cholesterol, and blood pressure. The total number of people who were examined medically was 2,386, divided as follows: 1,288 Mexican Americans; 929 Anglos; and 169 other Hispanics such as Cubans. The San Antonio Heart Study 1 was designed to test the hypothesis that as Mexican Americans became progressively more affluent and accultured to mainstream United States culture, they would gradually lose their diabetic pattern of cardiovascular risk factors of obesity, diabetes, and hypertriglyceridemia.

Mexican Americans are the second largest minority group in the United States. Despite this, as recently as 1977, little was known about their health status. In recent years this situation has begun to change with publication of results from several major studies including the San Antonio Heart Study, the Starr County Study, and most recently the Hispanic Health and Nutrition Examination Survey (HHANES).

DESIGN NARRATIVE:

Beginning in 1987, each participant in the prospective San Antonio Heart Study 1 was recalled as close as possible to the eighth anniversary of his or her baseline examination and re-examined. Subjects had fasting lipid and lipoprotein determinations, a glucose tolerance test to diagnose incident cases of non-insulin dependent diabetes mellitus according to the National Diabetes Data Group criteria, blood pressure measurements, a resting electrocardiogram, and completed a questionnaire to elicit information on possible cardiovascular endpoints. Medical records were obtained to document the endpoints. Subjects with a history of peripheral vascular disease or diabetes had ankle/arm blood pressure ratios determined before and after exercise. Deaths were ascertained in the course of routine follow-up supplemented by search of state health department and National Death Index records. Death certificates were requested and for those certificates indicating diabetes or cardiovascular disease as the cause of death, hospital or physician records were requested and reviewed. Standardized criteria were used to validate cause of death. Endpoints included diabetes, sudden cardiac death, non-fatal (including silent) myocardial infarction, angina pectoris, cerebrovascular and peripheral vascular disease. The study was renewed in 1992 to continue follow-up.

The study was renewed in 1998 to perform a 15 to 24 year mortality follow-up of the cohort. Time dependencies in the ethnic mortality ratio were examined as well as other covariates such as biochemical parameters, hemodynamic variables, and anthropometric variables such as socioeconomic status, health care access and utilization and migrant status. A sample of 200 Mexican-American and 200 non-Hispanic white death certificates were verified by medical record review. The purpose of the review was to determine whether Mexican Americans, deemed to have died of cardiovascular causes based on medical record review, were more or less likely to have cardiovascular codes (ICDA 390-459) listed on their death certificates than non-Hispanic whites deemed to have died of these causes. The review aided in the interpretation of data relating to possible ethnic differences in cause-specific mortality based on death certificate coding.

The study should help to explain whether the "Hispanic paradox" is an artifact of underascertainment of deaths in this group. The "Hispanic paradox" refers to the concept that all-cause and cardiovascular (CV) mortality are widely thought to be lower in Hispanics, including Mexican Americans, than in non-Hispanic whites in the United States. Preliminary mortality data from this study suggest that all-cause mortality is actually higher in Mexican Americans than in non-Hispanic whites in San Antonio.