Postpartum psychosis is associated with elevated neuromelanin-MRI signal in the midbrain

Postpartum psychosis (PP) occurs in about 1-3 per 1,000 births with symptoms including delusions, hallucinations, and mood episodes. PP is considered a medical emergency, as it can pose substantial risks for both parent and baby.

PP is difficult to detect and predict, in part because we do not understand the biological factors contributing to the illness well. Neuroimaging techniques, such as magnetic resonance imaging (MRI), have been used to study the role of emotion and memory systems, gray matter, and myelin in patients with PP.

Previous studies have shown that dopamine transmission abnormalities—like irregular release, uptake, or receptor signaling of dopamine in the brain—may be linked to psychotic disorders, but this association has not been investigated in the context of PP specifically. Neuromelanin (NM) is a dark pigment found in the brain that is a by-product of dopamine metabolism. Thus, researchers can measure the NM-MRI signal as a proxy for dopamine levels in the brain.

To better understand the role of dopamine level changes in PP, this study compared NM-MRI signal intensity in 30 women who had experienced PP in the past 10 years and a control group of 24 women who had given birth but did not experience PP. The study focused on differences in signal intensity in the substantia nigra (SN) and ventral tegmentum area (VTA), dopamine-rich areas of the midbrain, between the two groups. Additionally, the study examined resting-state functional connectivity specifically within the SN. The women with histories of PP were recruited from the larger Massachusetts General Hospital Postpartum Psychosis Project (MGHP3), and the control group was recruited through flyers, social media, and the MGH website.

Participants completed two study visits, one virtual visit of clinical assessments followed by an in-person MRI scan at the Athinoula A. Martinos Center for Biomedical Imaging in Boston, MA. Participants were paid $250 for completing the study and reimbursed for associated travel costs. Additional details about the methods and data analyses used in the study can be found in the full paper.

The PP and control groups had similar demographic characteristics. However, the PP group had significantly higher levels of subclinical psychotic symptoms, mania, emotion reactivity, and general anxiety, and lower levels of overall wellbeing than the control group.

Key findings:

• Having a history of PP was associated with increased NM-MRI signal intensity in the SN and the VTA. 

• Among the women with a history of PP, NM-MRI signal intensity was associated with levels of subclinical psychotic symptoms like unusual thoughts, suspicious/persecutory ideas, and perceptual abnormalities (hallucinations and distortions).

• The PP group had decreased SN connectivity to subcortical brain regions, compared to the control group. These decreases were associated with levels of SN NM-MRI signal and subclinical psychotic symptoms.

What does this mean?

• Changes in midbrain NM levels may be detected long after recovery from PP and may continue at lower levels after remission in the form of subclinical psychotic symptoms.

• It may be important to monitor subclinical psychotic symptoms in women with a history of PP.

• An increase in midbrain NM may impact the coordinated function of brain regions associated with the SN in women with a history of PP.

A limitation of the study is that findings may be related to pre-existing psychiatric conditions instead of the PP episode specifically. In addition, we cannot infer a causal relationship between MRI findings and subclinical psychotic symptoms in this cross-sectional study.

Thus, longitudinal studies are needed to better understand the relationship between NM-MRI signal intensity and psychotic symptoms over time (before pregnancy, during pregnancy, postpartum, and after postpartum). 

Citation: McKenna, F., Vinke, L.N., Williams, M. et al. Postpartum psychosis is associated with elevated neuromelanin-MRI signal in the midbrain. Mol Psychiatry 31, 3524–3532 (2026). https://doi.org/10.1038/s41380-026-03476-9

Written by: Peri Barest and Lily Cork, Clinical Research Coordinators