Haemoglobin Concentrations and Maternal and Neonatal Outcomes: Identifying Optimal Haemoglobin Ranges

Both low and high haemoglobin (Hb) concentrations associate with increased risks of adverse pregnancy outcomes, with the optimal Hb concentrations remaining elusive.

This study investigated the associations between Hb concentrations and maternal/neonatal outcomes and determined optimal Hb ranges during the second and third trimesters.

Retrospectively, data from 41,956 singleton live births delivered at Fuyang Women and Children's Hospital, a tertiary hospital in China, from 2018 to 2023, were analysed. Poisson regression with restricted cubic splines was employed to associate Hb concentrations with maternal (gestational diabetes mellitus [GDM], preeclampsia, and postpartum haemorrhage [PPH]) and neonatal outcomes (preterm birth [PTB], small for gestational age [SGA] birth, and neonatal asphyxia), applying generalised estimating equations to account for repeated measures. Optimal Hb ranges were determined using interquartile ranges from a low-risk subgroup and thresholds identified via the combined risk curve method, incorporating PPH and neonatal outcomes.

A total of 120,263 Hb measurements were analysed. U-shaped relationships emerged between Hb concentration and preeclampsia, PPH, PTB, SGA and neonatal asphyxia, whereas that with GDM was nearly linear. These patterns were consistent in both the second and third trimesters. The optimal ranges of 10.6-11.8 and 10.7-12.1 g/dL were identified by the low-risk subgroup approach for the second and third trimesters, respectively, whereas those by the combined risk curve method yielded slightly broader ranges of 9.6-11.7 g/dL and 9.9-12.0 g/dL, respectively.

The findings of this study reveal U-shaped relationships between maternal Hb concentrations and maternal/neonatal outcomes, with 10.5-12.5 g/dL as the optimal Hb concentration during the second and third trimesters, considering combined risks and clinical feasibility. Future multicentre studies are warranted to validate these ranges in broader populations.