Menstrual cycle length identifies perinatal risk information beyond routine metabolic antenatal screening: a retrospective cohort study of 17,792 deliveries.

BACKGROUND: Preterm birth (PTB) and low birth weight (LBW) are leading causes of neonatal mortality and long-term morbidity. In China, the high prevalence of gestational diabetes mellitus (GDM) and thyroid dysfunction (TF) creates a metabolically complex obstetric environment in which identifying risk markers that add value beyond existing screening is a priority. Menstrual cycle length is routinely available, yet whether it provides risk information non-redundant with GDM and TF has not been rigorously examined. METHODS: Retrospective cohort of 17,792 consecutive singleton deliveries at a tertiary centre in Wuhan, China (2020-2024). Cycle length was recorded at intrapartum admission: short (<25d), normal (25-35d; reference), or long (>35d). Primary outcomes were PTB (<37 weeks) and LBW (<2,500 g). Three nested logistic models were fitted; stability across models was used to assess whether GDM and TF acted as strong confounders. Restricted cubic splines, additive interaction (RERI), population-attributable fractions (PAF), and five sensitivity analyses were used, with Benjamini-Hochberg FDR correction. RESULTS: Long cycle (>35d; n = 1,810, 10.2%) was independently associated with PTB (aOR 1.36, 95% CI 1.12-1.64; FDR p = 0.006) and LBW (aOR 1.46, 1.18-1.78; FDR p = 0.002) after full adjustment for GDM and TF. Estimates were stable across progressive adjustment (PTB 1.34→1.36; LBW 1.42→1.46), indicating GDM and TF were not strong confounders. RERI for long cycle × GDM/TF crossed zero for both outcomes, excluding meaningful synergism. RCS showed a monotonic increase from ~30 days. PAF: PTB 3.1% (0.9-5.4%); LBW 3.9% (1.2-6.4%). The birth weight reduction (β = -33 g) attenuated at term (β = -8 g, NS), but long cycle was associated with higher odds of small-for-gestational-age (aOR 1.39, 1.11-1.73), including at term, indicating both earlier delivery and impaired fetal growth. Period length showed no association. CONCLUSIONS: Long menstrual cycle length was associated with PTB and LBW, not materially attenuated by adjustment for GDM and TF, consistent with prognostic information non-redundant with metabolic screening; however, stable estimates cannot exclude residual confounding or establish an independent pathway. Because cycle length reflects the pre-pregnancy ovulatory interval, any clinical value would lie in collecting menstrual history before conception or in early pregnancy. These findings are hypothesis-generating and require prospective validation in community-based cohorts.