ABSTRACT
Mexican-American infants have surprisingly low mortality rates, given their high-risk demographic characteristics. One explanation for this well-known paradox is the beneficial influence of a traditional Mexican cultural orientation. However, many studies have focused on individual, rather than contextual, markers of acculturation to explain the reasons for this paradox. This study incorporated community-level data into the analysis to further elucidate the Mexican paradox. Data from the National Linked Birth and Infant Death files for 1995-1997 were used to stratify infants born in counties of Arizona, California, New Mexico, and Texas into tertiles based on the proportion of Mexican births in each county. We calculated mortality rates for infants in each tertile. Logistic regression, with generalized estimating equations, was used to calculate odds ratios comparing infant mortality in low and medium concentration counties to high concentration counties. Odds ratios were adjusted for maternal age, education, parity, marital status, and maternal nativity status. Among Mexican-American infants, mortality rates ranged from 4.3 in counties with high proportions of Mexican births to 5.5 in counties with low proportions of Mexican births. However, this association was limited to US-born mothers, whose rates ranged from 4.4 in high concentration counties to 7.0 in low concentration counties (adjusted OR, 1.56 [1.35-1.81]); a substantial proportion of that difference was due to lower birth-weight specific mortality among normal birth-weight infants (1.9 vs. 3.1 deaths/1,000 live births). Among infants with Mexico-born mothers, there was no association between community context and mortality (OR, 1.01). Residence in counties with high proportions of Mexican births had a positive influence on birth outcomes among women of Mexican origin born in the United States. Exposure to the Mexican culture may reinforce healthy behaviors that Mexican-American women may lose through acculturation.