Inflammatory markers in late pregnancy in association with postpartum depression-A nested case-control study
Emma Bränn1, Fotios Papadopoulos2, Emma Fransson3
1Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden.
Psychoneuroendocrinology
|March 13, 2017
Summary
Inflammation markers in late pregnancy plasma may indicate postpartum depression risk. While no single marker is definitive, specific proteins like STAM-BP and a summary inflammation score show potential for future predictive models.
Area of Science:
- Immunology
- Perinatal Psychiatry
- Molecular Biology
Background:
- Immune system adaptations during pregnancy are implicated in perinatal depression.
- Identifying predictive markers for postpartum depression (PPD) is crucial for early intervention.
Purpose of the Study:
- To investigate if late pregnancy plasma inflammation markers can predict postpartum depressive symptoms.
- To explore novel inflammation-associated markers and their potential role in PPD pathophysiology.
Main Methods:
- Analyzed plasma samples from 291 pregnant women (63 with PPD, 228 controls) using a 92-marker inflammation panel (multiplex proximity extension assay).
- Employed logistic regression, LASSO, and Elastic net analyses to identify predictive markers.
- Validated protein-level findings for STAM-BP and ST1A1 using gene methylation data from an external population.
Main Results:
- Forty inflammation markers were found at lower levels in late pregnancy among women who later developed PPD.
- STAM-BP (AMSH), AXIN-1, ADA, ST1A1, and IL-10 remained significant after Bonferroni correction.
- A summary inflammation variable was the second-best predictor of PPD, after personal history of depression.
Conclusions:
- Exploratory study reveals differences in late pregnancy inflammation markers between women with and without postpartum depressive symptoms.
- Individual markers are not sufficient for PPD risk assessment, but STAM-BP and the summary inflammation variable show promise for future combined predictive models.


