Preterm birth etiological pathways: a Bayesian networks and mediation analysis approach.

Area of Science:

• Obstetrics and Gynecology
• Reproductive Health
• Epidemiology

Background:

• Spontaneous preterm birth (PTB) remains a significant global health concern.
• Understanding the etiological pathways of PTB is crucial for developing targeted interventions.
• Sociodemographic factors and infections are implicated in PTB risk.

Purpose of the Study:

• To determine the mediating effect of spontaneous preterm birth predictors.
• To identify etiological pathways contributing to spontaneous preterm birth.
• To suggest potential intervention strategies based on identified pathways.

Main Methods:

• A case-control study was conducted with multiparous women in Tucumán, Argentina (2005-2010).
• Sociodemographic, health, and obstetric data were collected from women delivering at term and preterm.
• Penalized logistic regression and Bayesian networks were used to identify predictive factors and mediation pathways.

Main Results:

• Three significant PTB pathways were identified: maternal age <20 mediated by few prenatal visits, first-trimester bleeding mediated by second-trimester bleeding, and UTI mediated by second-trimester bleeding.
• Few prenatal visits (<5) and vaginal bleeding were identified as major predictors.
• Mediation effects were sensitive to confounders influencing mediator-outcome and exposure-mediator variables.

Conclusions:

• Identified PTB pathways suggest etiological links to social disparities and genitourinary tract infections.
• Few prenatal visits and vaginal bleeding play a mediating role in PTB risk.
• Findings underscore the importance of addressing social determinants and infection management in PTB prevention.