The rise and fall of excess male infant mortality

doi:10.1073/pnas.0800221105

The rise and fall of excess male infant mortality

*Davis School of Gerontology,
^‡College of Letter Arts and Sciences, University of Southern California, 3715 McClintock Avenue, Los Angeles, CA 90089-0191

Communicated by Samuel H. Preston, University of Pennsylvania, Philadelphia, PA, January 16, 2008 (received for review July 28, 2007)

Abstract

The male disadvantage in infant mortality underwent a surprising rise and fall in the 20th century. Our analysis of 15 developed countries shows that, as infant mortality declined over two centuries, the excess male mortality increased from 10% in 1751 to >30% by approximately 1970. Remarkably, since 1970, the male disadvantage in most countries fell back to lower levels. The worsening male disadvantage from 1751 until 1970 may be due to differential changes in cause-specific infant mortality by sex. Declines in infant mortality from infections and the shift of deaths to perinatal conditions favored females. The reduction in male excess infant mortality after 1970 can be attributed to improved obstetric practices and neonatal care. The additional male infants who survived because of better conditions were more likely to be premature or have low birth weight, which could have implications for their health in later life. This analysis provides evidence of marked changes in the sex ratio of mortality at an age when behavioral differences should be minimal.

Footnotes

^†To whom correspondence should be addressed. E-mail: crimmin{at}usc.edu
Author contributions: G.L.D., E.M.C., S.V., and C.E.F. designed research, performed research, analyzed data, and wrote the paper.
The authors declare no conflict of interest.
This article contains supporting information online at www.pnas.org/cgi/content/full/0800221105/DC1.
↵ § We also estimated the same models with country-specific fixed effects to account for the nonindependence of observations in each country. In each model, interpretations were unchanged, although slope coefficients were attenuated, and R ² values were higher.