HISPANIC PREGNANCY OUTCOMES - ADDITIONAL EVIDENCE

Objective. Our objectives are to examine differentials in pregnancy ou tcomes across Hispanic groups and to address the question of whether, in addition to Mexican Americans, other Hispanic populations are chara cterized by an ''epidemiologic paradox,'' that is, a combination of a high-risk sociodemographic profile and favorable pregnancy outcomes. M ethods. Based on a national data set that contains several million bir ths and more risk factors than heretofore available, we employ logisti c regression to estimate ethnic-specific models as well as models in w hich ethnicity is included as a predictor. Results. Except among Puert o Ricans, rates of adverse pregnancy outcomes among Hispanics are rath er similar to Angle rates. The adjusted odds of prematurity and low bi rth weight, however, are significantly higher than the Angle risk for all Hispanic groups, while the odds of Hispanic infant mortality are s ignificantly lower. Maternal smoking,low weight gain, and low educatio n significantly increase the risk of adverse outcomes, as do both inad equate and ''adequate plus'' prenatal care. Conclusions. The ''paradox '' is reversed in the case of birth outcomes, but remains in the case of infant mortality. The similarity in the direction of effects of ris k factors regardless of ethnicity implies that interventions that impr ove pregnancy outcomes for one group will also benefit others.