Intestinal microbiome and maternal mental health: preventing parental stress and enhancing resilience in mothers

Maternal psychiatric disorders are increasing in developed countries, with significant implications for children’s mental health and development. Preventing maternal mental illness requires identifying both risk and resilience factors. The intestinal microbiota has emerged as a key player influencing the central nervous system via autonomic, neuropeptide, hormonal, and immune pathways and is modifiable by diet and lifestyle. Prior studies link depression and anxiety with increased inflammatory bacteria and decreased short-chain fatty acid (SCFA)-producing bacteria, notably butyrate producers with beneficial intestinal effects. Physical health is also closely tied to mental health, with fatigue, insomnia, and other somatic symptoms correlating with depression, anxiety, and stress. Indices such as the Multidimensional Physical Scale (MDPS), muscle mass (SMI), and strength (handgrip) can indicate mental health risks. However, comprehensive evaluations in young mothers are scarce. Parenting stress is a key risk factor for psychiatric disorders and child maltreatment; yet, its relationship with gut microbiota in nonclinical postpartum mothers remains understudied. Given multiple influences on maternal stress (child factors, environment, physical health), the Parenting Stress Index (PSI) is appropriate. This study investigated, in nonclinical Japanese mothers, the associations among gut microbiota, physical/physiological conditions (body composition/strength, autonomic function, oxytocin), parenting stress risk, and psychological resilience. The authors hypothesized that high parenting stress would associate with lower microbiota diversity and fewer SCFA producers, poorer physical conditions, and that in early postpartum primiparous mothers, gut microbiota and autonomic function (particularly vagal activity) would relate to resilience; oxytocin associations were explored without directional prediction.

Existing literature indicates the gut microbiota influences brain function via autonomic, endocrine, and immune pathways and is alterable by diet, lifestyle, and exercise. Systematic reviews have reported that patients with depression and anxiety show increased inflammatory taxa and decreased SCFA-producing bacteria, especially butyrate producers with beneficial mucosal and anti-carcinogenic effects. Physical health and mental illness are intertwined, with somatic symptoms (fatigue, insomnia) correlating with depression and anxiety; MDPS has been used to predict mental illness risk in women. Muscle mass and strength relate to psychiatric risk. Prior research on maternal stress and microbiota has focused on pregnancy using perceived stress scales; relationships between parenting stress and microbiota in nonclinical postpartum mothers are largely unexplored. Vagal activity may be a biomarker of resilience and mediates gut–brain signaling; oxytocin relates to parenting and stress responses but shows individual variability and mixed effects on cortisol and emotional processing. Postpartum physical recovery occurs within months but is not well quantified in relation to mental health.

Design: Two cross-sectional studies in nonclinical Japanese mothers. Study 1: Participants and procedures: 474 postpartum women (children aged 0–4 years) recruited via nursery schools nationwide (Jan–Feb 2021). Exclusions: current medical treatment (physical/psychiatric) (n=55), antibiotics within 3 months (n=69), psychiatric/hormone/fertility medications (n=3), incomplete PSI (n=8), yielding N=339 (mean age 34.66±4.81). Measures: Parenting stress (Japanese PSI; total and parent/child subscales), physical condition (MDPS: physical activity, somatic disorders, hormone activity, microvascular disorders, meteoropathy-related indices), sleep time and quality, dietary/lifestyle habits (Mykinso Pro), socioeconomic data. Microbiome: Stool collected with guanidine thiocyanate kits; DNA extracted via automated system. 16S rRNA V1–V2 sequencing (Illumina MiSeq 250 bp PE). Bioinformatics: QIIME2 (2020.8), DADA2; taxonomy via naïve Bayes classifier trained on arts-SILVA (SILVA 138-based). Diversity: alpha (Shannon) and beta (unweighted/weighted UniFrac). Prevalent genera defined as present in ≥50% of participants with ≥0.01% relative abundance. Statistics: Mann–Whitney U-tests for group comparisons (high-risk vs healthy) on physical indices, sleep, diet, alpha diversity, prevalent genera; FDR-adjusted q-values. Beta diversity compared by PERMANOVA. ANCOVA post hoc controlling for age and education where relevant. Study 2: Participants and procedures: 27 primiparous mothers 3–6 months postpartum (mean age 33.63±4.18) assessed Sep 2019–Mar 2020 at Kyoto University Baby Lab. All breastfeeding (55.56% exclusive), 81.48% vaginal delivery; 40.74% antimicrobials during delivery. Two visits: Visit 1—physical assessments and questionnaires (RS25, J-RS, CES-D). Physical measures: body composition (InBody 770: BMI, body fat %, SMI, ECW/TBW), dominant handgrip strength (digital dynamometer), two-step test (locomotive function), 5 m normal and maximum gait speed. Visit 2—saliva oxytocin (ELISA; Enzo ADI-900-153A-0001), resting ECG (3 min, 1000 Hz; cardiac vagal index [CVI], cardiac sympathetic index [CSI] via Lorenz plot analysis), PSI, demographics. Stool collected at home within 2–3 days after Visit 2. Microbiome: 16S rRNA V1–V2 via QIIME2 (2019.4), DADA2, taxonomy via Greengenes; shotgun metagenomics on DNBSEQ-G400, libraries via KAPA Hyper Prep, read mapping via Kraken2 against custom database (hg19, RefSeq bacteria/fungi, NCBI-nt viruses). Statistics: Exploratory correlations to select variables (p<0.05 or r≥0.30). Regression models relating relative abundance (shotgun-identified taxa) to psychological (RS25, J-RS, PSI), physical (handgrip, gait speeds, two-step), and physiological (CVI, CSI, oxytocin) variables. Significance at p<0.05 or R²≥0.13 (moderate effect); no multiple-comparison correction due to small sample, with emphasis on effect sizes. Visualizations and analyses in R (versions 3.5.1–4.2.1) with qiime2R, microbiome, tidyMicro, ggplot2; q-values via Bioconductor.

Study 1 (N=339): - Prevalence: 65 mothers (19.17%) exceeded PSI cutoffs and were classified as high parenting stress risk; 274 were healthy controls. - Physical condition: High-risk group had significantly poorer subjective sleep quality and higher (worse) MDPS scores across all five subscales (p<0.01, q<0.01). No significant difference in sleep duration. - Microbiota diversity: Alpha diversity (Shannon) was lower in the high-risk group (p=0.026, q=0.019). Beta diversity differed significantly between groups (PERMANOVA): unweighted UniFrac F=1.94, p=0.007; weighted UniFrac F=2.29, p=0.003. - Taxa differences: Significant group differences in relative abundance for Odoribacter, Alistipes, Erysipelatoclostridium, Lachnospira, Monoglobus, Phascolarctobacterium, Veillonella, Sutterella, and Escherichia–Shigella. Patterns suggest altered SCFA producers (e.g., Lachnospira, Veillonella, Alistipes, Phascolarctobacterium), fewer immunity/antibacterial-related taxa (Odoribacter, Sutterella), and more inflammatory/pathogenic taxa (Escherichia–Shigella). Results were robust to dietary habit comparisons and ANCOVA controlling for age and education. Study 2 (N=27 primiparous, 3–6 months postpartum): - Physical status vs references: 40.74% had SMI <5.7 kg/m² (sarcopenia criterion); 85.19% had handgrip below age-referenced values; 96.30% below reference on two-step test; 70.37% below normal gait speed reference; 92.59% below maximum gait speed reference, indicating reduced muscle mass and motor function. - Autonomic function and resilience: Shannon alpha diversity positively correlated with vagal activity (CVI) (r=0.59, p=0.003, 95% CI 0.005–0.71). CVI positively correlated with psychological resilience (J-RS) (r=0.42, p=0.047, 95% CI 0.23–0.80). - Shotgun metagenomics associations (partial regressions): • Blautia SC05B48 positively associated with resilience (J-RS) (R²=0.15, p≈0.06) and two-step test (R²=0.18, p=0.04). • Clostridium SY8519 positively associated with resilience (J-RS) (R²=0.20, p=0.03) and negatively with oxytocin (R²=0.35, p=0.01). • Collinsella aerofaciens positively associated with resilience (J-RS) (R²=0.26, p=0.01) and negatively with maximum gait speed (R²=0.27, p=0.01). • Eggerthella lenta negatively associated with resilience (RS25) (R²=0.26, p=0.01); positively with handgrip strength (R²=0.31, p=0.01) and PSI-infant (child aspect) (R²=0.16, p=0.05). • Faecalibacterium prausnitzii positively associated with sympathetic activity (CSI) (R²=0.22, p=0.03).

The studies demonstrate that even among nonclinical mothers, higher parenting stress risk relates to poorer physical condition and altered gut microbiota, characterized by reduced alpha diversity, distinct beta diversity, disrupted SCFA-producing taxa, decreased immunity/antibacterial-related genera, and increased inflammatory/pathogenic bacteria. These microbial signatures support gut dysbiosis and possible intestinal inflammation in high-stress mothers. In early postpartum primiparous mothers, greater gut microbial diversity associated with stronger vagal activity, and vagal activity linked to higher psychological resilience, suggesting autonomic function as a physiological pathway connecting gut microbiota with resilience. Specific taxa, notably butyrate-producing Blautia and Clostridium, were positively associated with resilience and certain physical functions, whereas Eggerthella lenta related to higher parenting stress and lower resilience, aligning with prior depressive disorder findings. The negative association between Clostridium SY8519 and oxytocin suggests complex interactions with stress-system homeostasis via HPA axis modulation. Findings underscore the potential for microbiota-targeted and autonomic-focused strategies to enhance resilience and mitigate parenting stress, with consideration for diet and ethnic microbiome differences.

This work links maternal parenting stress risk and psychological resilience to gut microbiota diversity and composition, autonomic nervous system function, and physical condition in nonclinical Japanese mothers. Study 1 shows that approximately 19% of mothers have high parenting stress risk, associated with poorer physical health and gut microbiota dysbiosis. Study 2 highlights that in early postpartum primiparous mothers, gut diversity correlates with vagal activity and resilience, and specific taxa (Blautia SC05B48, Clostridium SY8519, Collinsella aerofaciens, Eggerthella lenta) relate to resilience, oxytocin, and physical function. These findings inform conceptual models of resilience and suggest personalized interventions—such as diet (e.g., traditional Japanese foods influencing Blautia, isoflavones/equol metabolism), probiotics, lifestyle, and strategies to strengthen autonomic function—may enhance resilience before mothers become vulnerable. Future research should use larger, longitudinal, and cross-cultural samples, incorporate clinical high-stress groups, and include covariates (diet, body composition, exercise) to clarify causal pathways and optimize tailored support considering ethnic microbiome differences.

Study 2 had a small sample size (N=27) due to COVID-19 constraints, limiting statistical power and precluding inclusion of covariates; no multiple-comparison correction was applied, increasing potential type I error, though moderate-to-large effect sizes were emphasized and nonsignificant results reported. Both studies were cross-sectional, preventing causal inference and assessment of intraindividual change from prepregnancy to postpartum. Physical function recovery was not quantified relative to baseline; many Study 2 participants had reduced muscle mass and motor function versus references. Dietary habits did not differ between Study 1 groups, yet taxa patterns suggest possible diet-related effects requiring further investigation. Generalizability may be limited to Japanese mothers; ethnic differences in microbiota warrant international comparisons. Timing differences (Study 2 data pre-pandemic) and potential unmeasured confounders (diet, body composition, exercise) remain considerations.