Translational Psychiatry (Aug 2021)
Adverse childhood experiences, daytime salivary cortisol, and depressive symptoms in early adulthood: a longitudinal genetically informed twin study
Abstract
Abstract Dysregulated hypothalamic–pituitary–adrenal (HPA)-axis function might underlie the relationship between adverse childhood experiences (ACEs) and depression. However, limited research has examined the possible mediating role of the HPA-axis among young people using longitudinal data. Moreover, it remains unclear whether genetic influences could contribute to these associations. Participants were 290 children from the Twins Early Development Study. ACEs were assessed from age 3–11 years. We calculated a cumulative risk score and also derived different ACEs clusters using factor analysis and latent class analysis. HPA-axis activity was indexed by daytime salivary cortisol at age 11. Depressive symptoms were ascertained at age 21. Genetic liability to altered cortisol levels and elevated depressive symptoms was measured using a twin-based method. We performed causal mediation analysis with mixed-effects regression models. The results showed that ACEs cumulative exposure (b = −0.20, p = 0.03), bullying (b = −0.61, p = 0.01), and emotional abuse (b = −0.84, p = 0.02) were associated with lower cortisol levels at age 11. Among participants exposed to multiple ACEs, lower cortisol was related to higher depressive symptoms at age 21 (b = −0.56, p = 0.05). Lower cortisol levels mediated around 10–20% of the total associations of ACEs cumulative exposure, bullying, and dysfunctional parenting/emotional abuse with higher depressive symptoms. Genetic factors contributed to these associations, but the mediation effects of cortisol in the associations of ACEs cumulative exposure (b = 0.16 [0.02–0.34]) and bullying (b = 0.18 [0.01–0.43]) remained when genetic confounding was accounted for. In conclusion, ACEs were linked to elevated depressive symptoms in early adulthood partly through lower cortisol levels in early adolescence, and these relationships were independent of genetic confounding.