Potential for primary prevention of low birth weight in pregnancies with and without gestational diabetes mellitus: a prospective cohort study in Central China

BACKGROUND: Low birth weight (LBW) remains a critical global health challenge, yet the differential contributions of risk factors in pregnancies with and without gestational diabetes mellitus (GDM) are poorly characterized. This study aimed to identify modifiable first-trimester risk factors for LBW in GDM and non-GDM subpopulations within a Chinese context, and quantify the preventable LBW burden by population attributable fraction (PAF) estimation adjusted for risk factor interdependence. METHODS: This prospective cohort study enrolled 34,031 pregnant women from Central China (2013-2019), stratified by GDM (n = 5414) and non-GDM (n = 28,617) groups. Early-pregnancy exposures included advanced maternal age, passive smoking, alcohol intake, no folic acid used, low education, and pre-pregnancy underweight. Multivariable logistic regression and principal component analysis-adjusted PAFs were used to quantify preventable LBW burdens while accounting for risk factor overlap. RESULTS: Five risk factors were identified in both GDM (advanced maternal age, passive smoking, underweight, no folic acid used, and low education) and non-GDM (advanced maternal age, alcohol intake, underweight, no folic acid used, and low education) pregnancies. It was found that 18.4% and 33.4% of LBW cases could be theoretically prevented through these five identified risk factors in both GDM and non-GDM cohorts, respectively. Among four shared contributors, low education emerged as the primary modifiable risk factor (adjusted PAF: 9.3% in GDM; 20.1% in non-GDM), followed by underweight (adjusted PAF: 3.7% in GDM; 5.8% in non-GDM), advanced maternal age (adjusted PAF: 3.3% in GDM; 4.8% in non-GDM), no folic acid used (adjusted PAF: 0.9% in GDM; 2.2% in non-GDM). Passive smoking and alcohol intake are specific risk factors for pregnant women with GDM and non-GDM, respectively, with adjusted PAF of 1.1% and 0.5%. CONCLUSION: This study significantly contributes to the advancement of precision public health approaches by elucidating modifiable risk profiles associated with LBW in both GDM and non-GDM populations. These findings highlight population-specific heterogeneity in preventable disease burdens, revealing that a significant proportion of LBW cases could be reduced through targeted antenatal interventions.