BIRTH-WEIGHT AND INFANT-MORTALITY IN THE MEXICAN ORIGIN AND ANGLO POPULATIONS

Objective. The current near parity that exists between Mexican origin and Anglo infant mortality rates is perplexing because the former popu lation is characterized by a high risk profile, including higher ferti lity, younger age at first childbearing, less adequate prenatal care, and lower socioeconomic status. The purpose of this analysis is to det ermine whether this ''epidemiologic paradox'' can be accounted for by a more favorable Mexican origin birth weight distribution, as some rec ent research has suggested. Methods. Standard decomposition techniques are applied over a 30-year period to determine whether birth weight d istributions provide an explanation of how a socially disadvantaged mi nority came to have infant mortality rates so nearly equal to those of the majority population. Results. While improvement in birth weight c omposition appears to be responsible for one-fifth to one-fourth of th e decline in Anglo neonatal and postneonatal mortality rates, respecti vely, none of the decline in Mexican origin rates is explained in this manner. However, at specific weights at which the vast majority of al l births occur, the huge Mexican origin disadvantage so evident in 195 3 either disappeared entirely or was reduced to no more than two death s per 1,000 in 1983. Conclusions. Thus, the results indicate that ethn ic differentials in birth weight distributions provide neither a resol ution of the epidemiologic paradox as it exists currently nor an expla nation of why Mexican origin and Anglo infant mortality converged. In contrast, changes in weight-specific risk constitute by far the most i mportant explanation in regard to both neonatal and postneonatal morta lity.