Childhood stress and birth timing among African American women: Cortisol as biological mediator

Abstract

Preterm birth (PTB) occurs among 1:11U.S. white women and 1:7.5 African American women and is a significant driver of racial disparities in infant mortality. Maternal stress is the most common clinical phenotype underlying spontaneous PTB. Specific patterns of stress and biological mediators driving PTB remain unclear. We examined the effect of childhood stress on birth timing among African American women and evaluated maternal cortisol elevation as a biological mediator. A prospective observational design was employed, with a single study visit at 28-32 weeks gestation and medical record review. The Stress and Adversity Inventory was administered, which provides a comprehensive estimate of childhood stress, stress in adulthood, and five core characteristic subscales (interpersonal loss, physical danger, humiliation, entrapment, role disruption). Venipuncture was performed between 11:00am and 4:00pm and plasma cortisol quantified by ELISA. Analyses controlled for stress in adulthood. Among a final sample of 89, cumulative childhood stress predicted birth timing (p=0.01). The association was driven by stress related to interpersonal loss and physical danger, with support for maternal cortisol as a biological mediator (ab=0.02, 95% CI [0.001, 0.045]; ab=0.02, 95% CI [0.001, 0.043], respectively). Results were similar, overall, in sub-group analyses among spontaneously laboring women (n=53); however, role disruption arose as an additional predictor, as mediated by cortisol elevations (ab=0.03, 95% CI [0.005, 0.074]). Of note, cortisol was no longer supported as a mediator linking physical danger to birth timing after adjusting for sleep quality and hours awake prior to venipuncture (ab=0.02, 95% CI [-0.0001, 0.046]). We provide preliminary evidence that, independent of stress in adulthood, childhood stress of specific core characteristics may shape birth timing, with cortisol elevation as a biological mediator. Further investigation is warranted and may bolster the development of biologically-informed screening tools for the prediction and targeted prevention of stress-related PTB.

Keywords: African Americans; Cohort studies; Cortisol; Pregnancy; Premature birth; Psychological; Stress.

Figure 1

Associations among stress z-scores and maternal cortisol level. Greater cumulative childhood stress (r_s = .27), childhood interpersonal loss-related stress (r_s = .22), childhood physical danger-related stress (r_s = .22), and childhood role disruption-related stress (r_s = .37) were each associated with greater maternal cortisol level per Spearman rank correlation. Note: The interpersonal loss, physical danger, humiliation, entrapment, and role disruption subscale scores are each specific to childhood. *p < .05

Figure 2

Associations among stress z-scores and birth timing. Greater cumulative childhood stress (r_s = −.24) and childhood interpersonal loss-related stress (r_s = −.23) were each associated with earlier birth timing per Spearman rank correlation. Physical danger (r_s = −.20, p = .06) and role disruption (r_s = −.18, p = .09) showed marginal but non-statistically significant associations. Note: The interpersonal loss, physical danger, humiliation, entrapment, and role disruption subscale scores are each specific to childhood. Birth timing is untransformed. *p < .05