For decades, clinical guidance on hemoglobin in pregnancy has focused almost exclusively on anemia — the lower bound. This large Chinese cohort study reframes the conversation by demonstrating that elevated hemoglobin carries risks comparable to deficiency, suggesting that current upper-limit thresholds may be insufficiently protective for both mother and newborn.
Drawing on over 120,000 hemoglobin measurements from nearly 42,000 singleton pregnancies at a tertiary hospital in China (2018–2023), investigators modeled associations between second- and third-trimester hemoglobin and six outcomes: gestational diabetes mellitus (GDM), preeclampsia, postpartum hemorrhage (PPH), preterm birth, small-for-gestational-age (SGA) birth, and neonatal asphyxia. Using Poisson regression with restricted cubic splines and generalized estimating equations, the team found consistent U-shaped relationships between hemoglobin and most adverse outcomes — preeclampsia, PPH, preterm birth, SGA, and neonatal asphyxia — across both trimesters. GDM showed a near-linear positive association with rising hemoglobin. Optimal hemoglobin windows were identified at approximately 10.6–11.8 g/dL in the second trimester and 10.7–12.1 g/dL in the third trimester, derived from a low-risk subgroup approach cross-validated with a combined risk curve method.
The U-shaped pattern is biologically plausible: high hemoglobin often reflects hemoconcentration from reduced plasma volume expansion — a known marker of inadequate placental perfusion and a precursor to preeclampsia and fetal growth restriction. Conversely, severe anemia compromises oxygen delivery, increasing hemorrhage risk and fetal distress. What makes this study notable is its scale and the narrow optimal bands identified, which sit somewhat below the WHO's conventional anemia cutoff of 11.0 g/dL — a finding that could challenge standard clinical thresholds if replicated in diverse populations. Key limitations include its retrospective, single-center design in a Chinese tertiary hospital, limiting generalizability to lower-resource or ethnically distinct settings. The analysis is observational, precluding causal inference, and unmeasured confounders such as iron supplementation regimens, nutritional status, and altitude could influence results. Still, the dataset's size and methodological rigor make this an incrementally important contribution that warrants prospective validation across varied obstetric populations.