Perinatal depression is one of the most common medical complications during pregnancy and postpartum period, affecting 10% to 20% of pregnant individuals, with higher rates among Black and Latina women who are also less likely to be diagnosed and treated. Machine learning (ML) models based on electronic medical records (EMRs) have effectively predicted postpartum depression in middle-class White women but have rarely included sufficient proportions of racial/ethnic minorities, which has contributed to biases in ML models. Our goal is to determine whether ML models could predict depression in early pregnancy in racial/ethnic minority women by leveraging EMR data. We extracted EMRs from a large U.S. urban hospital serving mostly low-income Black and Hispanic women (n = 5875). Depressive symptom severity was assessed using the Patient Health Questionnaire-9 self-report questionnaire. We investigated multiple ML classifiers using Shapley additive explanations for model interpretation and determined prediction bias with 4 metrics: disparate impact, equal opportunity difference, and equalized odds (standard deviations of true positives and false positives). Although the best-performing ML model's (elastic net) performance was low (area under the receiver operating characteristic curve = 0.61), we identified known perinatal depression risk factors such as unplanned pregnancy and being single and underexplored factors such as self-reported pain, lower prenatal vitamin intake, asthma, carrying a male fetus, and lower platelet levels. Despite the sample comprising mostly low-income minority women (54% Black, 27% Latina), the model performed worse for these communities (area under the receiver operating characteristic curve: 57% Black, 59% Latina women vs. 64% White women). EMR-based ML models could moderately predict early pregnancy depression but exhibited biased performance against low-income minority women.

Higher stress during pregnancy associates with negative outcomes and elevated inflammation. The gut microbiota, reflecting environment and social interactions, alongside host immune responses have the potential to better understand perceived stress and identify when stress is excessive in pregnancy. Two U.S. cohorts of 84 pregnant individuals, composed of urban women of color and suburban white women, completed the Perceived Stress Scale-10 (PSS-10) and provided fecal and blood samples at two time points. Confirmatory Factor Analysis assessed the robustness of a two-factor PSS-10 model (Emotional Distress/ED and Self-Efficacy/SE). Gut microbiota composition was measured by 16 S rRNA amplicon sequencing and the immune system activity was assessed with a panel of 21 T-cell related cytokines and chemokines. ED levels were higher in the suburban compared to the urban cohort, but levels of SE were similar. ED and SE levels were associated with distinct taxonomical signatures and the gut microbiota data improved the prediction of SE levels compared with models based on socio-demographic characteristics alone. Integration of self-reported symptoms, microbial and immune information revealed a possible mediation effect ofBacteroides uniformisbetween the immune system (through CXCL11) and SE. The study identified links between distinct taxonomical and immunological signatures with perceived stress. The data are congruent with a model where gut microbiome and immune factors, both impacting and reflecting factors such as close social relationships and dietary fiber, may modulate neural plasticity resulting in increased SE during pregnancy. The predictive value of these peripheral markers merit further study.

Early-onset colorectal cancer (EOCRC) is defined as a diagnosis of colorectal cancer (CRC) in individuals younger than 50 years of age. While overall CRC rates in the United States (US) decreased between 2001 and 2018, EOCRC rates have increased. This research project aims to evaluate the feasibility and acceptability of Time-Restricted Eating (TRE), Mindfulness, or TRE combined with Mindfulness among young to middle-aged adults at risk of EOCRC. Forty-eight participants will be randomly assigned to one of four groups: TRE, Mindfulness, TRE and Mindfulness, or Control. Data on feasibility, adherence, and acceptability will be collected. Measures assessed at baseline and post-intervention will include body weight, body composition, dietary intake, physical activity, sleep behavior, circulating biomarkers, hair cortisol, and the gut microbiome. The effects of the intervention on the following will be examined: (1) acceptability and feasibility; (2) body weight, body composition, and adherence to TRE; (3) circulating metabolic, inflammation, and oxidative stress biomarkers; (4) intestinal inflammation; and (5) the gut microbiome. TRE, combined with Mindfulness, holds promise for stress reduction and weight management among individuals at risk of EOCRC. The results of this pilot study will inform the design and development of larger trials aimed at preventing risk factors associated with EOCRC.

Pregnancy alters many physiological systems, including the maternal gut microbiota. Diet is a key regulator of this system and can alter the host immune system to promote inflammation. Multiple perinatal disorders have been associated with inflammation, maternal metabolic alterations, and gut microbial dysbiosis, including gestational diabetes mellitus, pre-eclampsia, preterm birth, and mood disorders. However, the effects of high-inflammatory diets on the gut microbiota during pregnancy have yet to be fully explored. We aimed to address this gap using a system-based approach to characterize associations among dietary inflammatory potential, a measure of diet quality, and the gut microbiome during pregnancy. Forty-seven pregnant persons were recruited prior to 16 weeks of gestation. Participants completed a food frequency questionnaire (FFQ) and provided fecal samples. Dietary inflammatory potential was assessed using the Dietary Inflammatory Index (DII) from the FFQ data. Fecal samples were analyzed using 16S rRNA amplicon sequencing. Differential taxon abundances with respect to the DII score were identified, and the microbial metabolic potential was predicted using PICRUSt2. Inflammatory diets were associated with decreased vitamin and mineral intake and a dysbiotic gut microbiota structure and predicted metabolism. Gut microbial compositional differences revealed a decrease in short-chain fatty acid producers such as Faecalibacterium, and an increase in predicted vitamin B12 synthesis, methylglyoxal detoxification, galactose metabolism, and multidrug efflux systems in pregnant individuals with increased DII scores. Dietary inflammatory potential was associated with a reduction in the consumption of vitamins and minerals and predicted gut microbiota metabolic dysregulation.