Immune changes in pregnancy: associations with pre-existing conditions and obstetrical complications at the 20th gestational week-a prospective cohort study

Abstract

Background: Pregnancy is a complex biological process and serious complications can arise when the delicate balance between the maternal and semi-allogeneic fetal immune systems is disrupted or challenged. Gestational diabetes mellitus (GDM), pre-eclampsia, preterm birth, and low birth weight pose serious threats to maternal and fetal health. Identification of early biomarkers through an in-depth understanding of molecular mechanisms is critical for early intervention.

Methods: We analyzed the associations between 47 proteins involved in inflammation, chemotaxis, angiogenesis, and immune system regulation, maternal and neonatal health outcomes, and the baseline characteristics and pre-existing conditions of the mother in a prospective cohort of 1049 pregnant women around the 20th gestational week. We used Bayesian linear regression models to examine the impact of risk factors on biomarker levels and Bayesian cause-specific parametric proportional hazards models to analyze the effect of biomarkers on maternal and neonatal outcomes. We evaluated the predictive value of baseline characteristics and 47 proteins using machine-learning models and identified the most predictive biomarkers using Shapley additive explanation scores.

Results: Associations were identified between specific inflammatory markers and several conditions, including maternal age and pre-pregnancy body mass index, chronic diseases, complications from prior pregnancies, and COVID-19 exposure. Smoking during pregnancy affected GM-CSF and 9 other biomarkers. Distinct biomarker patterns were observed for different ethnicities. Within obstetric complications, IL-6 inversely correlated with pre-eclampsia risk, while birth weight to gestational age ratio was linked to markers including VEGF and PlGF. GDM was associated with IL-1RA, IL-17D, and eotaxin-3. Severe postpartum hemorrhage correlated with CRP, IL-13, and proteins of the IL-17 family. Predictive modeling yielded area under the receiver operating characteristic curve values of 0.708 and 0.672 for GDM and pre-eclampsia, respectively. Significant predictive biomarkers for GDM included IL-1RA and eotaxin-3, while pre-eclampsia prediction yielded the highest predictions when including MIP-1β, IL-1RA, and IL-12p70.

Conclusions: Our study provides novel insights into the interplay between preexisting conditions and immune dysregulation in pregnancy. These findings contribute to our understanding of the pathophysiology of obstetric complications and the identification of novel biomarkers for early intervention(s) to improve maternal and fetal health.

Keywords: COVID-19; Gestational diabetes; Immunology; Pre-eclampsia; Pregnancy; Risk prediction; Women’s health.

Fig. 1

a Flowchart for inclusion. b Timepoints for data collection. Data from the electronic health record was obtained throughout the pregnancy and postpartum

Fig. 2

Heatmap of phenotypes across participants using hierarchical clustering based on MSD biomarkers. Phenotypes are divided into maternal characteristics, pregnancy-related outcomes for the current pregnancy (observed before or at inclusion), fetal characteristics, pregnancy-related outcomes for prior pregnancies, prior diseases to the current pregnancy, and phenotypes observed after inclusion. ASA, acetylsalicylic acid; BMI, body mass index; C-section, cesarean section; DM, diabetes mellitus; GA, gestational age; GDM, gestational diabetes mellitus; IBD, inflammatory bowel diseases; IUI, intrauterine insemination; IVF, in vitro fertilization; MoM, multiples of medians; PCOS, polycystic ovary syndrome; PE, pre-eclampsia; PPH, postpartum hemorrhage; SGA, small for gestational age; UTI, urinary tract infection

Fig. 3

Previous pregnancies and pre-existing conditions effects on inflammatory markers. Effect size is increase or decrease in standard deviations. Only associations where the 95% Bayesian credible interval does not include zero are shown. ASA, acetylsalicylic acid; BMI, body mass index; GA, gestational age; GDM, gestational diabetes mellitus; IUI, intrauterine insemination; IVF, in vitro fertilization; MOM, multiples of medians; PCOS, polycystic ovary syndrome; PE, pre-eclampsia; PPH, postpartum hemorrhage

Fig. 4

Associations between markers and later obstetrical outcomes. The effect size is the log-hazard rate for all outcomes, except birth weight-gestational age ratio (BWGA). The effect of biomarkers on BWGA is a change in percentile from the 50th percentile. Only associations where the 95% Bayesian credible interval does not include zero are shown. BMI, body mass index; GA, gestational age; IUI, intrauterine insemination; IVF, in vitro fertilization

Fig. 5

Summarized SHAP values for each feature category (red = clinical measure, blue = MSD biomarker) across all outcomes investigated in the prognostic model. PPH, postpartum hemorrhage

Fig. 6

SHAP values for all prognostic features for gestational diabetes mellitus (GDM) and pre-eclampsia, respectively. The top ten features with the highest median value for each outcome are indicated with light colors. Only IL-1RA and CRP were part of both top ten lists. Features are colored based on the feature category as clinical measure (red) or MSD biomarker (blue). BMI, pre-pregnancy body mass index; MoM, multiple of medians; SHAP, Shapley additive explanation