Anemia, defined as a hemoglobin concentration below a given threshold, remains one of the most common nutritional problems on the planet, and pregnant women and adolescent girls bear a disproportionate share of the burden. The condition is far from benign. Iron deficiency, the leading cause of anemia, impairs oxygen delivery to tissues, including the brain, and has long been suspected of influencing mood, cognition, and energy levels. At the same time, perinatal depression affects a substantial proportion of women worldwide, with prevalence estimates particularly high in low- and middle-income countries, and it carries consequences that ripple outward to infant development, mother-child bonding, and even birth outcomes such as preterm delivery and low birth weight. Whether the two conditions are causally connected, however, has been surprisingly difficult to pin down.

To address that uncertainty, the research team conducted a systematic review and dose-response meta-analysis, searching eight bibliographic databases for peer-reviewed studies published from 1980 onward that reported on the relationship between maternal hemoglobin status or anemia and maternal depression during pregnancy or up to one year postpartum. The screening process began with 718 titles and abstracts, from which 42 studies ultimately met the eligibility criteria for inclusion. Rather than pooling results with conventional fixed or random-effects techniques alone, the investigators used a flexible Bayesian meta-regression approach, a statistical framework capable of modeling nonlinear relationships between hemoglobin concentration and depression risk while accommodating the considerable variation in study designs, populations, and measurement methods.

The dose-response curves generated by the models are the study’s most technically interesting output. In longitudinal studies, the analysis of peripartum depression drew on 21 observations from 10 studies, while the postnatal depression analysis used 19 observations from 9 studies. Across these curves, the risk of depressive symptoms was minimized at a mean hemoglobin concentration of approximately 128 grams per liter and rose steadily as concentrations dropped. This pattern is biologically plausible. Hemoglobin is the body’s principal oxygen carrier, and the neural circuits that regulate mood are metabolically demanding. Iron, which sits at the center of every hemoglobin molecule, also serves as a cofactor in the synthesis of dopamine and serotonin, neurotransmitters intimately involved in mood regulation. Prior research, including work by Corwin and colleagues showing that low hemoglobin is a risk factor for postpartum depression, and studies linking plasma ferritin measured shortly after delivery to depressive symptoms, has hinted at these mechanisms.

Pregnancy complicates the picture in an important way. During gestation, plasma volume expands substantially, diluting red blood cells and producing a physiological hemodilution that lowers hemoglobin concentrations even in iron-replete women. One recent large-scale cohort study even proposed gestational hemodilution itself as a putative risk factor for postpartum depression. This means that interpreting a low hemoglobin reading during pregnancy requires care: it may reflect normal dilution, true iron deficiency, or both. The World Health Organization has recently updated its guidance on hemoglobin cutoffs for defining anemia, including adjustments for elevation and other factors, and the new analysis aligns itself with this evolving landscape, underscoring how definitional choices can shift estimates of global anemia burden and, by extension, the size of the population potentially at risk.

Yet the authors are emphatic that their headline finding comes with a major caveat. Under the Burden of Proof framework, which assigns a star rating from one to five to reflect the strength of evidence for a risk-outcome relationship, the hemoglobin-depression link received only one star out of five. That rating stems from substantial between-study heterogeneity: the included studies varied widely in how they measured hemoglobin, when during pregnancy or the postpartum period they measured it, how they defined and assessed depression, and which confounding factors they controlled for. Many relied on screening instruments such as the Edinburgh Postnatal Depression Scale rather than clinical diagnosis, and few accounted for the full range of socioeconomic, obstetric, and psychological factors that influence both anemia and mood. The one-star rating does not mean the association is absent; it means the evidence base is too weak to support strong causal claims.

This is precisely what the Burden of Proof framework, first described by Zheng and colleagues in Nature Medicine in 2022, was designed to expose. Traditional meta-analyses often report a pooled effect size without systematically grading how much confidence readers should place in it. The framework borrows ideas from GRADE methodology but pushes further, applying conservative risk-function modeling that penalizes sparse, heterogeneous, or biased data. The result is a more honest accounting: a risk curve can be estimated, but its evidentiary weight is made explicit. For low hemoglobin and peripartum depression, the message is that a plausible and potentially modifiable risk factor has been identified, but the field needs far better studies before iron supplementation or transfusion policies can be justified on mental health grounds alone.

The practical implications are nonetheless significant. Anemia during pregnancy is already a target of global health policy, with routine iron and folic acid supplementation recommended in antenatal care guidelines. If future research confirms that correcting low hemoglobin also reduces the risk of perinatal depression, the case for aggressive screening and treatment would strengthen considerably, particularly in settings where both anemia and perinatal mental disorders are highly prevalent. Maternal depression is associated with impaired early interactions, compromised parenting and safety practices, and adverse effects on child cognitive and emotional development, so even a modest causal contribution from anemia would translate into a meaningful population-level burden. Conversely, if the association turns out to be largely confounded, by poverty, chronic stress, or poor overall nutrition, interventions targeting hemoglobin alone would miss the mark.

The research team, which includes investigators from the Institute for Health Metrics and Evaluation, the University of Queensland, and the Queensland Centre for Mental Health Research, and which received support from the Gates Foundation, is explicit about what comes next. High-quality longitudinal studies with robust, repeated measurement of hemoglobin and validated assessment of peripartum depression are needed to strengthen the evidence base. Such studies should ideally track iron status across trimesters, distinguish iron deficiency from other causes of low hemoglobin, and use diagnostic interviews rather than screening questionnaires alone. In the meantime, the study offers clinicians and policymakers a carefully quantified signal: the lowest observed risk of peripartum depression occurs near a hemoglobin concentration of 128 grams per liter, and the evidence linking lower levels to higher risk, while weak by the strictest standards, is consistent enough to warrant attention. For a condition as common as anemia and an outcome as consequential as maternal depression, that combination of plausibility and uncertainty is exactly the kind of finding that should galvanize the next generation of research.

Subject of Research: The association between low maternal hemoglobin concentration and peripartum depressive symptoms assessed through a systematic review and Burden of Proof meta-analysis

Article Title: The effect of low hemoglobin on maternal peripartum depressive symptoms: a Burden of Proof framework

Article References: The effect of low hemoglobin on maternal peripartum depressive symptoms: a Burden of Proof framework. (n.d.). https://doi.org/10.1038/s44220-026-00716-3

Image Credits: AI Generated

DOI: 10.1038/s44220-026-00716-3

Keywords: anemia, hemoglobin, peripartum depression, postpartum depression, pregnancy, Burden of Proof, meta-analysis, iron deficiency, maternal mental health, Bayesian meta-regression, global health, risk factors