Hair Cortisol Concentration and Perceived Chronic Stress in Low-Income Urban Pregnant and Postpartum Black Women

Evaluating Bias-Mitigated Predictive Models of Perinatal Mood and Anxiety Disorders

Importance

Machine learning for augmented screening of perinatal mood and anxiety disorders (PMADs) requires thorough consideration of clinical biases embedded in electronic health records (EHRs) and rigorous evaluations of model performance.

Objective

To mitigate bias in predictive models of PMADs trained on commonly available EHRs.

Design, Setting, and Participants

This diagnostic study collected data as part of a quality improvement initiative from 2020 to 2023 at Cedars-Sinai Medical Center in Los Angeles, California. The study inclusion criteria were birthing patients aged 14 to 59 years with live birth records and admission to the postpartum unit or the maternal-fetal care unit after delivery.

Exposure

Patient-reported race and ethnicity (7 levels) obtained through EHRs.

Main Outcomes and Measures

Logistic regression, random forest, and extreme gradient boosting models were trained to predict 2 binary outcomes: moderate to high-risk (positive) screen assessed using the 9-item Patient Health Questionnaire (PHQ-9), and the Edinburgh Postnatal Depression Scale (EPDS). Each model was fitted with or without reweighing data during preprocessing and evaluated through repeated K-fold cross validation. In every iteration, each model was evaluated on its area under the receiver operating curve (AUROC) and on 2 fairness metrics: demographic parity (DP), and difference in false negatives between races and ethnicities (relative to non-Hispanic White patients).

Results

Among 19 430 patients in this study, 1402 (7%) identified as African American or Black, 2371 (12%) as Asian American and Pacific Islander; 1842 (10%) as Hispanic White, 10 942 (56.3%) as non-Hispanic White, 606 (3%) as multiple races, 2146 (11%) as other (not further specified), and 121 (<1%) did not provide this information. The mean (SD) age was 34.1 (4.9) years, and all patients identified as female. Racial and ethnic minority patients were significantly more likely than non-Hispanic White patients to screen positive on both the PHQ-9 (odds ratio, 1.47 [95% CI, 1.23-1.77]) and the EPDS (odds ratio, 1.38 [95% CI, 1.20-1.57]). Mean AUROCs ranged from 0.610 to 0.635 without reweighing (baseline), and from 0.602 to 0.622 with reweighing. Baseline models predicted significantly greater prevalence of postpartum depression for patients who were not non-Hispanic White relative to those who were (mean DP, 0.238 [95% CI, 0.231-0.244]; P < .001) and displayed significantly lower false-negative rates (mean difference, -0.184 [95% CI, -0.195 to -0.174]; P < .001). Reweighing significantly reduced differences in DP (mean DP with reweighing, 0.022 [95% CI, 0.017-0.026]; P < .001) and false-negative rates (mean difference with reweighing, 0.018 [95% CI, 0.008-0.028]; P < .001) between racial and ethnic groups.

Conclusions and Relevance

In this diagnostic study of predictive models of postpartum depression, clinical prediction models trained to predict psychometric screening results from commonly available EHRs achieved modest performance and were less likely to widen existing health disparities in PMAD diagnosis and potentially treatment. These findings suggest that is critical for researchers and physicians to consider their model design (eg, desired target and predictor variables) and evaluate model bias to minimize health disparities.

A Pilot Study Exploring the Feasibility of Virtual Written Exposure Therapy with Underserved Black Perinatal Women

In the USA, Black pregnant women are at the highest risk for maternal morbidity. They also experience the highest rates of trauma exposure and posttraumatic stress disorder (PTSD). PTSD takes a toll on Black women's mental and physical health, placing them at risk for maternal morbidity. It increases several mental health symptoms such as suicidality, anxiety, re-living the trauma, and numbness. These mental health conditions adversely affect health behaviors, including those essential for maternal health, such as attending prenatal and postpartum appointments. Furthermore, untreated PTSD is associated with higher blood pressure, increasing Black mothers' risk of pre-eclampsia. For a variety of reasons including cultural mistrust, stigma, transportation, time constraints, and access to care, PTSD is frequently underassessed and undertreated among Black perinatal women. Written exposure therapy (WET) is a state-of-the-art brief treatment for PTSD. In this study, we explored the initial feasibility of the virtual delivery of WET to reduce PTSD symptoms among Black perinatal women. Results found the virtual delivery of WET to be feasible. Symptom reduction for PTSD in participants was 50-100% during follow-up, suggesting potential effectiveness of the intervention. Implications for virtual delivery of WET in reducing risk for Black maternal morbidity are discussed.

Age and Hair Cortisol Levels as Predictors of SARS-CoV-2 Infection

Chronic psychosocial stress is known to adversely impact immune function. During the SARS-CoV-2 pandemic, occupational stress among workers in healthcare was at an unprecedented level due to risks of infection and work demands. We performed a nested case-control study to investigate the associations between chronic stress and the risks of contracting SARS-CoV-2. We collected 3 cm of hair from employees at an academic medical center who tested positive for SARS-CoV-2 (N = 49) and controls who tested negative (N = 49), matched for age, race, and sex. The diagnosis of SARS-CoV-2 was based on polymerase chain reaction or antibody tests. As a proxy for chronic stress, we segmented hair into 1 cm sections each representing one month and measured cortisol levels using a cortisol enzyme-linked immunosorbent assay. For cases, we used cortisol concentrations measured in hair segments from the month prior to a positive SARS-CoV-2 test, and for controls, we used time-matched hair segments. We fitted conditional logistic regression models adjusted for sex, age, race, body mass index, and healthcare worker status, and stratified models by older vs. younger age (cutoff = 41 years). African Americans had higher hair cortisol levels relative to participants of other races and ethnicities. In adjusted models, higher hair cortisol concentrations were associated with an increased odds of infection with SARS-CoV-2 (OR = 1.84; CI: 1.10-3.07) among older, but not younger, participants. The results suggest that psychosocial stress may be a risk factor for SARS-CoV-2 infection; stress management may be an important part of a comprehensive approach to protect against SARS-CoV-2 infection.

Association of Increasing the Minimum Wage in the US With Experiences of Maternal Stressful Life Events

Importance

Exposure to stressful life events (SLEs) before and during pregnancy is associated with adverse health for pregnant people and their children. Minimum wage policies have the potential to reduce exposure to SLEs among socioeconomically disadvantaged pregnant people.

Objective

To examine the association of increasing the minimum wage with experience of maternal SLEs.

Design, Setting, and Participants

This repeated cross-sectional study included 199 308 individuals who gave birth between January 1, 2004, and December 31, 2015, in 39 states that participated in at least 2 years of the Pregnancy Risk Assessment Monitoring Survey between 2004 and 2015. Statistical analysis was performed from September 1, 2022, to January 6, 2023.

Exposure

The mean minimum wage in the 2 years prior to the month and year of delivery in an individual's state of residence.

Main Outcomes and Measures

The main outcomes were number of financial, partner-related, traumatic, and total SLEs in the 12 months before delivery. Individual-level covariates included age, race and ethnicity, marital status, parity, educational level, and birth month. State-level covariates included unemployment, gross state product, uninsurance, poverty, state income supports, political affiliation of governor, and Medicaid eligibility levels. A 2-way fixed-effects analysis was conducted, adjusting for individual and state-level covariates and state-specific time trends.

Results

Of the 199 308 women (mean [SD] age at delivery, 25.7 [6.1] years) in the study, 1.4% were American Indian or Alaska Native, 2.5% were Asian or Pacific Islander, 27.2% were Hispanic, 17.6% were non-Hispanic Black, and 48.8% were non-Hispanic White. A $1 increase in the minimum wage was associated with a reduction in total SLEs (-0.060; 95% CI, -0.095 to -0.024), financial SLEs (-0.032; 95% CI, -0.056 to -0.007), and partner-related SLEs (-0.019; 95% CI, -0.036 to -0.003). When stratifying by race and ethnicity, minimum wage increases were associated with larger reductions in total SLEs for Hispanic women (-0.125; 95% CI, -0.242 to -0.009).

Conclusions and Relevance

In this repeated cross-sectional study of women with a high school education or less across 39 states, an increase in the state-level minimum wage was associated with reductions in experiences of maternal SLEs. Findings support the potential of increasing the minimum wage as a policy for improving maternal well-being among socioeconomically disadvantaged pregnant people. These findings have relevance for current policy debates regarding the minimum wage as a tool for improving population health.

Perinatal mood and anxiety disorders: biomarker discovery using plasma proteomics

BACKGROUND

Perinatal mood and anxiety disorders encompass a range of mental health disorders that occur during pregnancy and up to 1 year postpartum, affecting approximately 20% of women. Traditional risk factors, such as a history of depression and pregnancy complications including preeclampsia, are known. Their predictive utility, however, is not specific or sensitive enough to inform clinical decision-making or prevention strategies for perinatal mood and anxiety disorders. Better diagnostic and prognostic models are needed for early identification and referral to treatment.

OBJECTIVE

This study aimed to determine if a panel of novel third-trimester plasma protein biomarkers in pregnant women can be used to identify those who have a high predisposed risk for perinatal mood and anxiety disorders within 3 months postpartum.

STUDY DESIGN

We studied 52 women (n=34 with a risk for perinatal mood and anxiety disorders and n=18 controls) among whom mental health screening was conducted at 2 time points, namely in the third trimester and again at 3 months postdelivery. An elevated perinatal mood and anxiety disorder risk was identified by screening individuals with above-validated cutoffs for depression (Edinburgh Postnatal Depression Scale ≥12), anxiety (Overall Anxiety Severity and Impairment Scale ≥7), and/or posttraumatic stress disorder (Impact of Events Scale >26) at both time points. Plasma samples collected in the third trimester were screened using the aptamer-based SomaLogic SomaScan proteomic assay technology to evaluate perinatal mood and anxiety disorder-associated changes in the expression of 1305 protein analytes. Ingenuity Pathway Analysis was conducted to highlight pathophysiological relationships between perinatal mood and anxiety disorder-specific proteins found to be significantly up- or down-regulated in all subjects with perinatal mood and anxiety disorder and in those with perinatal mood and anxiety disorders and no preeclampsia.

RESULTS

From a panel of 53 significant perinatal mood and anxiety disorder-associated proteins, a unique 20-protein signature differentiated perinatal mood and anxiety disorder cases from controls in a principal component analysis (P<.05). This protein signature included NCAM1, NRCAM, and NTRK3 that converge around neuronal signaling pathways regulating axonal guidance, astrocyte differentiation, and maintenance of GABAergic neurons. Interestingly, when we restricted the analysis to subjects without preeclampsia, a 30-protein signature differentiated perinatal mood and anxiety disorder cases from all controls without overlap on the principal component analysis (P<.001). In the nonpreeclamptic perinatal mood and anxiety disorder group, we observed increased expression of proteins, such as CXCL11, CXCL6, MIC-B, and B2MG, which regulate leucocyte migration, inflammation, and immune function.

CONCLUSION

Participants with perinatal mood and anxiety disorders had a unique and distinct plasma protein signature that regulated a variety of neuronal signaling and proinflammatory pathways. Additional validation studies with larger sample sizes are needed to determine whether some of these molecules can be used in conjunction with traditional risk factors for the early detection of perinatal mood and anxiety disorders.

An evaluation of the feasibility and acceptability of sister circles as an anxiety intervention for pregnant Black women

BACKGROUND

Black women and their infants face heightened health risks during pregnancy, labor, and delivery that, for many, cost them their lives. Such health risks during this critical period are linked to increased rates of anxiety among Black pregnant and postpartum women. Black women also endure racism when seeking support from mental health and healthcare providers, which further contributes to pregnancy and postpartum-related anxiety. Evidence on sister circles has demonstrated that this indigenous form of healing may provide Black pregnant and postpartum women with the support and skills needed to cope with stressors associated with mental and physical health concerns and racism.

OBJECTIVE

Our study aimed to evaluate the feasibility and acceptability of the Birthing Beautiful Babies Sisters Offering Support, a sister circle cognitive behavioral therapy-based stress and anxiety intervention for Black pregnant and postpartum women.

DESIGN

Our study employed a mixed-methods framework.

METHODS

Descriptive analyses, paired samples t-test, and open and selective coding were conducted. We utilized the following measures: Penn State Worry Questionnaire-Abbreviated, Kessler Psychological Distress Scale, Pregnancy-Related Anxiety Scale, focus booklet, and focus group discussion.

RESULTS

Participants reported experiencing moderate levels of worry and distress and low levels of pregnancy-related anxiety prior to the start of the Birthing Beautiful Babies Sisters Offering Support intervention. Eighty women participated in Birthing Beautiful Babies Sisters Offering Support. Eleven focus groups were conducted to assess their experience. Participants reported a perceived increase in their knowledge about panic attacks and stress and ability to manage stressors effectively. They found that all completed activities contributed to their development and application of skills. They reported they enjoyed Birthing Beautiful Babies Sisters Offering Support because of the supportive environment, openness, and emphasis on sisterhood. Participants provided helpful feedback about the structure and flow of the intervention.

CONCLUSION

Limitations of the present study and future directions are discussed.

Maternal stress and hair cortisol among pregnant women following hurricane Florence

Natural disasters represent major stressors, resulting in psychological distress and physiological responses such as increased cortisol. During pregnancy, this impacts not only maternal well-being, but also fetal development. In 2018, Hurricane Florence caused extensive damage across the eastern United States. Studies indicated that compared to married pregnant women, unmarried pregnant women had higher risk of distress. Here we assess hair cortisol among a subsample of participants, and variations based on marital status.

METHODS

We analyzed multiple stress measures among 37 participants who were pregnant during Hurricane Florence. We used questionnaires modeled on previous studies to assess hardship associated with the hurricane, psychological distress, sociodemographic characteristics, social support, and food security. We analyzed cortisol concentrations in proximal and distal hair sections, representing stress around the time of the disaster (distal) and 3-4 months following the disaster (proximal). We used linear regression to test relationships between hair cortisol and self-report stress measures, and variations based on marital status.

RESULTS

Self-report measures of distress and hardship were similar among married and unmarried participants. Mean cortisol levels in distal and proximal sections were higher among unmarried participants. Controlling for confounding variables, hardship was not associated with hair cortisol. Distress predicted cortisol in distal sections (β = .482, p = .018), with a trend for proximal sections (β = .368, p = .055). Marital status was a significant predictor of distal (β = .388, p = .027) and proximal (β = .333, p = .047) hair cortisol, explaining 8.6%-11.7% of unique variance.

CONCLUSIONS

Preexisting and intersecting risk factors likely place unmarried pregnant individuals at risk of stress during and following a disaster.

COVID-19 and Systemic Racism Pandemics Impact Daily Life for Immigrant/Refugees, Minneapolis, USA

Background:

The pandemics of COVID-19 and systemic racism have a deleterious impact on the daily life experiences and health for populations of color. The experiences are compounded for immigrant/refugee communities that may have other barriers such as English language literacy or trauma. Cumulative stress due to everyday racism is harmful for health.

Objectives:

This study describes the impact of day-to-day lived experiences of Karen, Somali, and Latinx communities during the COVID-19 pandemic and aftermath of the police murder of George Floyd in the Minneapolis/St. Paul metro area.

Methods:

In-depth interviews were conducted over three weeks in September and October 2020 to understand the daily life experiences of Karen, Somali and Latinx adults drawn from community contacts during the COVID-19 pandemic and the aftermath of the police murder of George Floyd. Interviewers were bilingual and from the communities they interviewed. Nine questions were asked, ranging from their knowledge of COVID-19, prevention practices, experiences during shelter-in-place, and the perceptions of the police murder of George Floyd. Qualitative analysis included transcript review, coding facilitated by Atlas.ti Cloud software, summaries, and validation by interviewers.

Findings:

Thirty-two adults were interviewed (Latinx = 10, Karen = 10, Somali = 12). One-third were in person per participant request and complying with COVID-19 precautions, and the remainder were remote. The average age recorded was 37 years (range 20–66 years), 43.8% males and 56.3% females. Respondents reported experiences of discrimination and systemic racism while engaging in daily life activities, including accessing foods and common goods, school, work, transportation, and healthcare, all of which were exacerbated by COVID-19 and the police murder of George Floyd.

Conclusions:

Immigrant/refugee communities of color in Minneapolis/St. Paul face daily experiences of racism that were compounded by the events of 2020. Discrimination and systemic racism contribute to the persistent health inequities among populations of color.

Father involvement in infancy predicts behavior and response to chronic stress in middle childhood in a low-income Latinx sample

Fathers' involvement in early childhood is important for children's physical, emotional, and cognitive development, particularly in low-income families. However, little is known about the longitudinal relations between early father involvement and children's later physiological responses to chronic stress and behaviors impacted by stress in the context of poverty. These issues are particularly important among Latinx immigrant families who face significant psychosocial and poverty-related risk. In the current study, we examined the relationship between father involvement in infancy and physiological chronic stress in the middle childhood period, as measured through hair cortisol concentration (HCC), and several behavioral measures (attention problems, working memory) in a Latinx immigrant sample with low income. Father involvement in infancy predicted children's later HCC, and working memory in second to third grade. Father involvement also moderated the effect of HCC on working memory, such that increased HCC predicted better working memory when fathers were not involved. These findings suggest that the fathers' involvement in infancy has lasting impacts on health and behavior and that associations between physiological and behavioral measures of stress may be moderated by differences in early father involvement.