Stress Response Patterns Among African American Youth Living in Public Housing Communities: A Latent Class Analysis

Background & Purpose:

African American youth are vulnerable to chronic stress exposure given their overrepresentation in urban neighborhoods characterized by concentrated poverty, violence, and other socio-ecological stressors. Chronic stress exposure may lead to atypical patterns of cortisol reactivity for youth including chronically elevated or blunted cortisol levels, which may be associated with negative health consequences. However, there have been inconsistent findings regarding stress-induced variability of cortisol responses among African American youth. These inconsistencies have important implications for interventions targeting this population, and thus motivated our exploration. Our study 1) investigates heterogeneity in cortisol response among African American youth, and 2) examines the association between biological stress response patterns of cortisol and psychological and socio-ecological measures.

Methods:

Data are from 123 African American youth ages 13 to 21 living in public housing and low-income apartments in a Northeastern U.S. city. Salivary cortisol samples were collected three times: 08:00am, 08:30am, and 09:00am. Cortisol levels were categorized into three groups: up to the median, between median and 75th percentile, and above 75th percentile. We used latent class analysis (LCA) to identify discrete groups in which members had similar patterns of cortisol levels. Mann-Whitney tests were performed to examine the bivariate associations between psychological (e.g., anxiety, depressive symptoms) and socio-ecological (e.g., community violence exposure, perceived neighborhood risks, exposure to delinquent peers) correlates and class membership. Finally, a logistic regression was utilized to assess the relative effects of the correlates on class membership.

Results:

Sample mean cortisol levels across three time points was 0.26 μg/dL (SD = 0.15). Results of LCA indicated that a 2-class model of cortisol reactivity was the best. For class 1 youth, mean cortisol at waking was 0.396 μg/dL (SD = 0.035), 0.483 μg/dL (SD = 0.033) at time 2, and 0.439 μg/dL (SD = 0.033) at time 3. This group was indicative of a normative waking stress response. For class 2 youth, mean cortisol at waking was 0.198 μg/dL (SD = 0.012), 0.210 μg/dL (SD = 0.010) at time 2, and 0.179 μg/dL (SD = 0.010) at time 3. This group was indicative of a blunted stress response. The psychological and socio-ecological factors were not significantly associated with class membership in both bivariate and multivariable models.

Discussions: Findings suggest that African American youth display distinct physiological stress reactivity patterns. A subset of youth exhibit stress response patterns consistent with hypocortisolism, which may be consequential for youth health and development. The non-significant results suggest that physiological measures such as salivary cortisol may be suitable for assessing stress responses that are not captured by subjective psychosocial measures. Interventions to assist youth in coping with stress, while transforming the upstream factors that give rise to adverse community conditions, are needed to advance racial, political, and social justice.