P0274 - The Effects of Testosterone, Estradiol, Dehydroepiandrosterone, and Cortisol on Anxiety and Depression Symptoms in Ecuadorian Adolescents

Background: Adrenal and sex hormone dysregulation have been independently associated with increased depression and anxiety. Cortisol, an adrenal hormone often used as a marker of stress, can modify the production of sex hormones and may modify the hormone-mood associations. This study evaluated the associations and interplay between sex and adrenal hormones with depression and anxiety among adolescents.

Methods: We examined 545 adolescents in rural Pedro Moncayo, Ecuador in 2016 (ages 11-17y, 50.4% female). Depression and anxiety symptoms were assessed using the Children’s Depression Inventory 2nd Edition (CDI-2) and the Multidimensional Anxiety Scale for Children 2nd Edition (MASC-2), respectively. Testosterone, cortisol, dehydroepiandrosterone (DHEA), and estradiol were measured in saliva collected upon awakening. Estradiol was only measured in boys. Linear regression was used to estimate associations; effect modification by cortisol was assessed. We calculated odds ratios (OR) of elevated depression or anxiety scores (clinical cutoff scores ≥60) comparing participants with low hormone concentrations (≤10th percentile of the cohort, using gender- and age-specific cutoffs) and elevated hormone concentrations (≥90th percentile of the cohort) to those with normal hormone concentrations (10th-90th percentile) using logistic regression. Models were adjusted for age, BMI-for-age z-score, parental education, household income, hemoglobin concentration, awakening time, saliva collection time, and sexual maturation.

Results: Mean (SD) depression and anxiety scores were 53.8 (8.95) and 58.5 (9.37) for females, and 52.1 (9.60) and 56.8 (9.80) for males, respectively. In boys, elevated estradiol concentrations were associated with elevated depression scores (OR=4.75 [95%CI: 1.95, 11.56]) and elevated anxiety scores (OR=2.43 [1.01, 5.84]) compared to boys with normal levels (Figure 1). In pooled analyses (boys and girls), elevated testosterone concentrations (OR=1.79 [0.99, 3.24]) and elevated cortisol concentrations (OR=1.77 [0.99, 3.16]) were also marginally associated with elevated anxiety scores (Figure 1). Low and elevated DHEA levels showed no association in either stratified or pooled analysis. In linear regression models, only estradiol was independently associated with depression (difference per 10% increase (β= 0.45 [0.15, 0.75]) and anxiety scores (β= 0.42 [0.13, 0.72]) (Table 1). Higher cortisol levels strengthened the estradiol-depression association in boys (β for upper median of cortisol [βumc]= 0.54 [0.12, 0.96]), and the depression associations with testosterone (βumc = -0.19 [-0.35, -0.03]) and DHEA (βumc = -0.12 [-0.22, -0.01]) in girls (Table 1).

Conclusion: We observed that elevated levels of testosterone, cortisol and estradiol (≥90th percentile) were associated with altered mood of adolescents. Cortisol concentration was a considerable effect modifier to the associations of most hormones with depression and anxiety. Given the effect modification seen between adrenal and gonadal hormones on depression and anxiety symptoms, our findings confirm the need to further characterize the influence of these relationships on mood. Replication of our findings in longitudinal analyses are warranted.