Key points are not available for this paper at this time.
Background Early lactation failure among mothers of moderate preterm infants poses a persistent public health concern. This study identified key predictors and developed a risk stratification tool. Methods We conducted a prospective cohort study of 3,210 mother-infant dyads (32.0–34.9 weeks gestation) at a tertiary hospital in China (February 2022–April 2025). Early lactation failure was defined as absence of direct breastfeeding with documented latch or provision of 5 ml cumulative expressed breast milk within 72 h postpartum. Data included psychosocial assessments breastfeeding self-efficacy scale-short form (BSES-SF), Edinburgh postnatal depression scale (EPDS), family support, obstetric factors, neonatal characteristics, and early care variables. Missing data was imputed using multiple chained equations (50 datasets). Multilevel logistic regression with ward-level random intercepts identified predictors, with bootstrap validation (1,000 resamples) assessing performance. Results Among 3,210 mother-infant dyads enrolled, 716 (22.3%) experienced early lactation failure within 72 h postpartum. Key predictors included higher BSES-SF scores adjusted odds ratio (aOR) = 0.96 per point, 95% confidence interval (CI): 0.94–0.98, higher EPDS scores (aOR = 1.08, 95% CI: 1.04–1.12), cesarean delivery (aOR = 1.42, 95% CI: 1.15–1.75), neonatal respiratory support (aOR = 1.28, 95% CI: 1.05–1.56), and shorter kangaroo care duration (aOR = 0.92 per 10 min, 95% CI: 0.87–0.97). A five-factor risk tool stratified mothers into low-risk (13.0% failure rate), moderate-risk (20.5%), and high-risk (31.0%) groups, with strong discrimination ( C -statistic = 0.704; calibration slope = 0.952; Hosmer-Lemeshow p = 0.267). Population-attributable risks were highest for cesarean delivery (20.4%), low self-efficacy (18.9%), and depression (18.2%), with numbers needed to treat (NNT) ranging from 11 to 18. Conclusion This study confirms the multifactorial basis of early lactation failure, highlighting maternal psychosocial factors as key predictors. The validated risk tool enables identification of high-risk dyads for targeted nutritional support interventions.
Hong et al. (Tue,) studied this question.