A comparison of Hispanic and Anglo compromised birth outcomes and cause-specific infant mortality in the United States, 1989-1991

Objective. Recent research has documented the consistency of the epidemiolo gic paradox. For all major Hispanic groups, despite high-risk profiles, adj usted infant-mortality rates are similar to, or more favorable than, those for non-Hispanic whites. Few studies have examined the infant cause-of-deat h structure for Hispanics. Methods. Using the Linked Birth and Infant Death Data files from NCHS for 1989-91, this paper employs a relatively new appr oach to cause-of-death classification, where deaths are categorized by a ty pology that focuses on causes most likely to be affected by similar prevent ion strategies. In addition, we utilize a refinement of the conventional me asure of compromised birth outcome, which includes a component of maturity not typically included in social science research. We explore the distribut ion of infant deaths by these new classifications, as well as the distribut ion by cause and outcome, for four major Hispanic groups and non-Hispanic w hites (Anglos). Results. Most Hispanic groups have quire similar distributi ons of infant death by cause category, and, unlike the situation among adul ts, the distributions are very similar to those observed among Anglos. The cause-by-outcome patterns extend partly to the four leading causes of infan t death, with puerto Ricans being the most likely to show disparate pattern s. Conclusions. The most important findings from multinomial logistic regre ssion models explain (and reduce) observed differentials, and strongly sugg est that, should Hispanics experience risk profiles identical to those of A nglos, cause-specific differences would be reduced sharply, and even disapp ear, for all Hispanic groups.