Inflated Denominators, Selection In Utero, and the Black Male Neonatal Death Paradox.

Abstract

Background: Epidemiologists speculate that comparatively high rates of fetal death among males conceived by non-Hispanic Black (NHB) women in the United States (USA) could explain the unexpectedly low neonatal death rate among extremely preterm (ePTB) NHB males. Consistent with this 'selection in utero' argument, conception cohorts exhibiting high sex ratios (M:F) of NHB stillbirths reportedly exhibit greater NHB advantages in ePTB male neonatal death rates. Sceptics, however, attribute this association to an artefact that spuriously inflates the denominators of neonatal death rates in highly stressed populations.

Objective: To determine whether the positive association over conception cohorts between the NHB male neonatal death advantage and the sex ratio of NHB stillbirths survives correction for inflated denominators.

Methods: We retrieved vital statistics for NHB and non-Hispanic white (NHW) singleton ePTB infants born in the USA from 1995 through 2018. We aggregated these data into 282 monthly conception cohorts. We avoided the inflated denominator problem by substituting a 'NHB share of burden' variable for the difference between NHB and NHW neonatal death rates. We specify this variable as the NHB proportion of neonatal deaths among NHB and NHW ePTB males born from each conception cohort. We determined, using Box-Jenkins methods, whether cohorts exhibiting unusually high sex ratios of NHB stillbirths also exhibited unusually low NHB shares of the burden of ePTB male neonatal death.

Results: Consistent with the selection in utero argument, the NHB share of neonatal deaths among ePTB males fell 7% below expected among the cohorts exhibiting unusually high sex ratios of NH Black stillbirths.

Conclusions: Stillbirth affects the composition of birth cohorts by selecting against less fit males in conception cohorts. Although clinical manifestations of this bias remain largely unexplored, they likely include the Black male neonatal death paradox.