Adversity in early life and pregnancy are immunologically distinct from total life adversity: macrophage-associated phenotypes in women exposed to interpersonal violence

The study addresses whether adversity during two sensitive immunological periods—early life (birth to age 5) and pregnancy—has distinct and lasting associations with immune phenotypes compared to adversity accumulated across the total lifespan. Prior work shows adversity is linked to cardiometabolic, psychiatric, neurodevelopmental disorders and premature mortality. Sensitive periods feature heightened CNS and immune plasticity, enabling environmental embedding via epigenetic mechanisms. Pregnancy represents a period of profound maternal immune remodeling involving both pro-inflammatory and immunosuppressive states to support fetal tolerance. Despite known links between adversity and inflammation, the specific macrophage-associated phenotypes tied to developmental timing of adversity remain unclear. The authors hypothesized that higher adversity during early life and during pregnancy would be associated with shifts toward a pro-inflammatory macrophage gene expression phenotype (higher M1/M2), and that these associations would be distinct from those related to total lifetime adversity.

Evidence indicates early-life adversity induces epigenetic changes that heighten inflammatory responses, influence T-cell repertoires, and contribute to neuroinflammation. The innate immune system can acquire trained immunity—an epigenetically mediated memory that augments responses to subsequent challenges. Pregnancy involves dynamic immune adaptations balancing inflammatory and tolerogenic processes at the maternal-fetal interface; excessive M1-like macrophage polarization is linked to adverse obstetric outcomes, whereas M2-like profiles support tolerance. Stress and low socioeconomic status have been associated with heightened inflammatory gene expression during pregnancy and with altered immune profiles in maternal and fetal tissues. Traditional single biomarkers (e.g., CRP) may lack sensitivity/specificity, motivating composite and cell-type-specific transcriptional phenotypes. Endotoxin tolerance (ET) represents a macrophage-associated immunosuppressive program observed clinically (e.g., sepsis) and may reflect cumulative stress exposure over time.

Design: Cross-sectional baseline analysis within the CPP-HEALTH study (2013–2015) of mothers seeking child-parent psychotherapy following child exposure to interpersonal trauma. Participants: 53 low-income, ethnically diverse biological mothers of children aged 2–6 years (serum cytokine cohort); a convenience subsample of 42 provided monocyte gene expression data. Inclusion: biological mothers, English/Spanish fluency, child exposed to a traumatic event. Exclusion: non-biological mothers, current substance abuse, child ward of the state, current pregnancy, developmental disorder, psychosis, chronic medical conditions in mother or child. Participants consented; IRB-approved. Adversity assessment: Structured interviews assessed early life adversity (ELA; <5 years), pregnancy adversity (PA; adversity during a specific pregnancy), and total lifetime adversity (TLA). ELA and PA were treated as sensitive-period exposures; TLA captured cumulative exposure. Biospecimen collection: Morning fasting blood draws; serum stored at −80°C. Serum biomarkers: Pro-inflammatory cytokines/markers measured included TNF-α, IL-1β, IL-6, and CRP (high-sensitivity assay). Anti-inflammatory cytokines included IL-1RA, IL-4, and IL-10. Multiplex chemiluminescent assay platforms used (vendor details provided). Pro-inflammatory (PRO) and anti-inflammatory (ANTI) aggregate scores were computed by averaging normalized component markers (PRO: CRP, TNF-α, IL-1β; ANTI: IL-1RA, IL-10, IL-6). Monocyte RNA-seq: CD14+ monocytes isolated via positive selection from PBMCs, stabilized in RNAlater. RNA extracted, cDNA prepared with Illumina TruSeq Stranded protocol, sequenced on Illumina HiSeq 4000. Reads aligned to human transcriptome (HISAT2) and expression quantified/normalized (StringTie). Primary immune phenotype (M1/M2): A priori macrophage-associated gene sets reflecting M1-like (pro-inflammatory) and M2-like (immunoregulatory; focused on M2b) expression were aggregated. An M1 score and an M2 score were computed and combined as an M1/M2 ratio index of relative pro-inflammatory imbalance. Comparator phenotype (Endotoxin tolerance, ET): Clinically-validated macrophage immunosuppressive gene expression signature representing a tolerant/desensitized state following repeated endotoxin stimulation. Lower ET scores indicate prolonged inflammatory responses (reduced tolerance). ET was evaluated to test specificity of associations versus M1/M2. Statistical analysis: Variables assessed for normality; cytokines transformed as needed (e.g., log). Multivariable linear regression models tested associations between each adversity construct (ELA, PA, TLA) and immune outcomes (M1/M2, ET, PRO, ANTI). Covariates: Model 1 adjusted for age, BMI, antidepressant use, and Hispanic ethnicity; Model 2 added socioeconomic status indicators (poverty per census definition), US-born status, and education. Additional models adjusted for PTSD and depressive symptom severity (PSS, CESD-R). Analyses also conducted on non-normalized data; exploratory gene-level analyses identified contributors within the M1/M2 phenotype (e.g., TLR2). Power analysis indicated 80% power to detect moderate effects (d=0.40) at α=0.04.

- Sample: Mean maternal age ~32 years; predominantly Hispanic (72%); 66% living in poverty; mean CRP 2.50 mg/L (SEM 0.42; range 0.29–11.60), indicating low-grade chronic inflammation. - Primary outcome (M1/M2): Adjusted analyses showed higher early life adversity (ELA) and higher pregnancy adversity (PA) were each associated with a higher M1/M2 macrophage gene expression phenotype, indicating a relative pro-inflammatory skew (ELA β = 0.337, p = 0.029; PA β = 0.332, p = 0.031). In multivariate models including both ELA and PA plus mental health covariates, ELA remained significant; PA showed a trend when controlling for ELA (p = 0.052). - Total lifetime adversity (TLA): The abstract reports TLA was associated with lower M1/M2 (β = -0.341, p = 0.031). In results focused on specificity, TLA was significantly associated with a more immunosuppressed endotoxin tolerance (ET) phenotype (lower ET scores implying prolonged inflammatory responses), and this TLA–ET association remained significant when adjusting for ELA, PA, PTSD, and depression (e.g., multivariate model β ≈ -0.374 to -0.442; p = 0.026–0.031). - Serum cytokines: Neither ELA nor PA were significantly associated with aggregated serum pro-inflammatory cytokines (PRO) after adjustment. TLA showed limited associations with PRO only when adjusting for socioeconomic factors. Higher PA was associated with lower anti-inflammatory cytokines (ANTI) in unadjusted analyses, but not after adjustment. - Exploratory gene-level finding: Within the M1/M2 signature, TLR2 showed the highest contribution to the association pattern.

Findings support the hypothesis that adversity during immunologically sensitive periods—early life and pregnancy—is linked to a specific macrophage-associated transcriptional phenotype characterized by a higher M1/M2 ratio, suggesting a durable pro-inflammatory bias potentially reflecting trained immunity. These associations persisted after controlling for sociodemographic, medical, and psychological factors, indicating specificity to timing of adversity rather than generalized distress. In contrast, total lifetime adversity related more strongly to an endotoxin tolerance phenotype, consistent with a potential weathering effect where repeated exposure diminishes tolerance mechanisms, leading to prolonged or dysregulated inflammatory responses. The absence of robust associations with conventional serum cytokines underscores the value of composite, cell-type-specific transcriptional biomarkers over single-protein measures. Together, results highlight distinct immunologic imprints of adversity depending on developmental timing, with implications for identifying high-risk individuals and tailoring interventions, particularly among low-income women exposed to interpersonal violence who face elevated chronic disease risks.

The study identifies a macrophage-associated M1/M2 gene expression phenotype as a specific biomarker of adversity during early life and during pregnancy. Greater adversity in these sensitive periods was associated with higher M1/M2 (pro-inflammatory skew) in circulating monocytes, independent of sociodemographic and mental health covariates. In contrast, cumulative lifetime adversity was associated with an endotoxin tolerance immunosuppressive phenotype, suggesting distinct biological embedding pathways by timing. These results support the concept that sensitive-period adversity may instill trained immunity-like biases, whereas repeated lifetime adversity may erode tolerance mechanisms, potentially contributing to chronic disease risk. Future research should use longitudinal designs, functional validation of phenotypes, single-cell and surface marker profiling, and examine generalizability across populations and pregnancies to inform precision prevention and intervention strategies.

- Generalizability: Sample comprised low-income, ethnically diverse women seeking treatment for child trauma with high violence exposure; findings may not generalize to other populations. - Pregnancy adversity measurement: Adversity was assessed for one specific pregnancy; adversity during other pregnancies was not captured. - Cross-sectional design: Limits causal inference and temporal ordering; longitudinal validation is needed. - Phenotype simplification: M1/M2 transcriptomic index may oversimplify dynamic in vivo macrophage polarization; functional assays and cell-surface phenotyping were not performed. - Sample size: Modest, particularly for the RNA-seq subsample (n=42), which may limit power for some associations and subgroup analyses. - Potential unmeasured confounding: Breastfeeding status could not be included due to small n; other biological or environmental factors may influence immune phenotypes.