Rethinking Psychosocial and Energetic Influences on Cortisol, Testosterone, and Hormonal Coupling among Jordanian, Syrian, and Indigenous Qom Adolescents

Abstract

Extreme forms of stress derived from displacement, poverty, and armed conflict exposures may confer lasting changes to the neuroendocrine system and may impact the pace of development for children and adolescents. Evolutionary theorists suggest that early life stressors may accelerate or suppress pubertal development due to high psychosocial stress, nutritional disadvantage and shortened life expectancies. A key mechanism by which puberty may be regulated is through the relationships between the hypothalamic pituitary adrenal (HPA) and gonadal axes (HPG) that in adulthood may suppress one another, but in puberty may upregulate in concordance to meet developmental demands of puberty (e.g., growth, responses to stress). ‘Hormonal coupling’ describes the relationship between two or more hormones over time, to quantify the relationship between the HPA (cortisol) and HPG (testosterone). We predicted cortisol and testosterone would not couple at pre-pubertal ages, positively couple at pubertal ages, and de-couple post-puberty. While testosterone reliably increases in puberty, past research on pubertal and age-related changes in cortisol are inconsistent. In addition, studies of hormonal coupling have rarely focused on diverse populations outside of the United States wherein socioecologies and childhood exposures may include more significant forms of marginalization and extreme exposures to conflict. In U.S. teens, the connection between cortisol and testosterone weakens in the presence of early life stress. However, it's unclear how stress signals related to insecurity, trauma, and mortality risks may affect hormone coupling in teens affected by war and socioeconomic marginalization and how hormonal trajectories might vary in disparate contexts. In this dissertation, I leverage collaborations and data among Indigenous Qom adolescents in Argentina who have been affected by socioeconomic marginalization and among Jordanian refugee and Syrian non-refugee adolescents living in Jordan to advance understanding of how cortisol varies with pubertal timing and linear growth, and whether hormonal coupling varies between these two populations and in response to trauma, insecurity, distress, mental health, and resilience among the cohort in Jordan. In my first chapter, I explored relationships between age, linear growth, adiposity, C-peptide (proxy for insulin), and cortisol across puberty, and tested whether higher cortisol levels are associated with earlier ages at menarche and peak height velocity. I used longitudinal data (n = 777 urine samples) from Qom females ages 7-14 (n = 46) and tested my pre-registered analysis using Bayesian longitudinal mixed effects models and joint modeling. I found limited evidence supporting the overarching hypothesis that HPA upregulation is associated with pubertal maturation or timing but that cortisol may be more clearly related to differences in relative linear growth at early-mid puberty, as measured by height-for-age Z-scores. In my second chapter, I draw from two populations: Indigenous Qom females living in Argentina and male and female adolescents living in Jordan ages 10-19 (n =769). Qom participants had >1 year of quarterly morning urine samples assayed with commercially available kits and in-house assays. Jordanian and Syrian participants had dried blood spots, which were analyzed for free cortisol and testosterone using multiple reaction monitoring mass spectrometry. I used Bayesian hierarchical models with age, log cortisol, and an age-specific log cortisol slope to estimate testosterone. In both studies, I found small positive associations between cortisol and age on testosterone, consistent with the notion of hormonal coupling. For Qom females, the age-specific effect of cortisol on testosterone increased in magnitude at ages 12-14, possibly suggesting greater post-menarcheal changes among Qom females whose median age at menarche is 11.5. Among Jordanian/Syrian adolescents, the age-specific effect of cortisol on testosterone was somewhat greater for females vs. males at earlier ages, potentially indicative of pubertal timing differences and relatively later pubertal timing. Overall, I found evidence of positive hormonal coupling and no evidence of de-coupling. In the third chapter, with data from Jordanian and Syrian participants, I focus specifically on differences between Jordanian and Syrians who have very different life circumstances as non-refugee host teens and refugee teens displaced into Jordan. I focus on the potential effects of extreme stressors such as lifetime trauma, human insecurity, human distress, mental health, resilience, and experiencing an 8-week psychosocial intervention on cortisol-testosterone coupling at baseline (n=769 participants) and 12 months later (n=225 participants). Overall I found that when disaggregating by refugee status, Syrians as compared to Jordanians had relatively higher cortisol and changes in cortisol across age groups and Syrian males had slightly higher testosterone compared to Jordanians. I found evidence of positive coupling across participants, with mixed findings for the effects of trauma, insecurity, distress, mental health and resilience. I found scant evidence of de-coupling. Syrian females with higher trauma, insecurity, and distress showed robust hormonal coupling but less of an age-related trend, implying that there may be early life programming of HPA regulation affecting overall hormone levels. Further, some of the evidence in Chapter Three implies that hormonal coupling and de-coupling are context-dependent rather than a universal developmental pattern. Overall, my dissertation research propels an understanding of energetic and psychosocial influences on the neuroendocrine system forward and helps to bridge research on hormonal coupling to global populations with diverse socioecologies.