Joint Effect of Paternal and Maternal Age on the Prevalence of Low Birth Weight in the United States

OBJECTIVE: Prior studies of the effect of father's age on the risk of low birth weight (LBW, infant weight <2500 g) have been inconclusive. In addition, previous investigators have not conducted a detailed analysis to determine whether paternal age (PA) interacts with maternal age (MA) on the outcome of LBW. We aimed to quantify the joint effect of PA and MA on the prevalence of LBW. METHODS: Cross-sectional prevalence data from the National Survey of Family Growth 2017-2019 (female pregnancy file) were analyzed using SAS 9.4 software. The exposure of interest was the PA group. Six PA groups were created based on the mother's and father's age at the time of conception. Weighted frequencies and weighted percent estimates of selected characteristics were calculated. Odds ratios (ORs) and 95% confidence intervals (CIs) for the binary outcome of LBW were calculated from logistic regression models. A causal diagram (ie, directed acyclic graph) identified maternal race-ethnicity as a variable that needed to be controlled for to reduce confounding bias when estimating the association under study. Preterm birth, according to our directed acyclic graph, was a collider on some pathways between parental age and LBW (and an intermediate on other pathways) and hence was not controlled for. All of the analyses accounted for the complex survey sample design. CONCLUSIONS: We did not detect a joint effect of maternal and PA on the prevalence of LBW in a national US sample.