Ni Gusti Ayu Nanditha, Dong Keun Rhee, Corey J Teply, Heidi A Tandiono, Ihunaya Okorie, Heather J Taylor, Bridget L Stollfus, Huong T Chu, Taylor S Noyes, Kyle Humphrey, Jackie Markt-Maloney, Aishe Memetova, Eunice Chung, Ashley A Harris, Kia C Fuller, Maegan A Dirac, Zulfiqar A Bhutta, William M Gardner, Ksenia I Ershova, Domenico Girelli, Khaled M Musallam, Parminder S Suchdev, Susan A McLaughlin, Reed J D Sorensen, Aleksandr Y Aravkin, Peng Zheng, Nora M Gilbertson, Christopher J L Murray, Simon I Hay, Theresa A McHugh, Nandita Perumal, Nicholas J Kassebaum
This study evaluated dose-response relationships between hemoglobin concentration in pregnancy and maternal and neonatal outcomes based on comprehensive systematic reviews using the Burden of Proof meta-analytic framework. Risk estimates accounting for uncertainty and quantitative star ratings were generated to assess evidence strength. Our analysis revealed that even small deviations above or below hemoglobin levels with the lowest predicted risk - particularly on the low end (anemia)-were associated with non-linear increased risks of all-cause maternal and neonatal mortality, postpartum hemorrhage, maternal sepsis, preterm birth, low birthweight and large-for-gestational-age. Across outcomes with at least moderate evidence strength (>2 stars), maternal hemoglobin levels associated with the lowest risk ranged from 109-135 g/L, challenging the appropriateness of fixed anemia thresholds. Even modest population-level shifts in hemoglobin distributions could meaningfully affect the burden of adverse outcomes, supporting more refined, trimester- and outcome-specific thresholds to better identify at-risk individuals and guide clinical interventions.