Predicting Preterm Birth from the Unpredictable

Abstract

Preterm birth has many equivocal risk factors in the medical literature. This research attempts to unify the risk factors for preterm birth using a mid-level evolutionary biology theory termed Life History (LH) Theory. Many of the risk factors for preterm birth overlap considerably with indicators of a latent LH factor. This study reports a measurement model in which a latent construct of maternal LH speed was constructed using many of the common risk factors for preterm birth as manifest indicators. The Maternal LH Factor loads in the predicted directions on to the risk factors of preterm birth and other theoretically specified manifest indicators. Three samples were collected: (1) a clinical sample, (2) a Facebook sample, and (3) a Reddit sample. A structural model was tested that included the LH variables and sociodemographic variables of interest as predictors of preterm birth. The structural model, a sequential canonical cascade model, indicates that developing in a faster family accelerates maternal LH speed and reduces maternal educational attainment. These events then reduce maternal income, which increases self-reported perceived stress and ultimately leads to the birth of a preterm infant.