Abstract
BACKGROUND: Depression is a leading mental health concern in adolescents. The maturation of the hypothalamic pituitary adrenal (HPA) axis during adolescence coincides with higher basal cortisol, and elevations in evening cortisol have been associated with depressive symptoms. Autism spectrum disorder (ASD) is differentiated by challenges with socioemotional reciprocity. Research in autistic youth has shown earlier and higher rates of depressive symptoms and elevated evening cortisol. The extent to which cortisol profiles may be linked to depressive symptoms in ASD has not been explored. METHODS: Participants included 244 youth, 140 autistic and 104 neurotypical, aged 10-16 over four years. The Children's Depression Inventory (CDI) 2nd Edition Total T-score assessed depressive symptoms. Salivary morning and evening cortisol collected over three days in the home were log transformed and averaged. We fit a mixed effects model for CDI Total scores with log-transformed cortisol (fit with natural cubic splines) as the main variable of interest, adjusting for diagnosis (ASD or TD), nonlinear age (fit with natural cubic splines), sex, and use of psychotropic medication. We also allowed for diagnosis-by-cortisol and sex-by-cortisol interactions. RESULTS: There was a main effect of morning cortisol on the CDI total score (p = 0.028, robust effect size index (RESI) = 0.17), but no main effect for evening cortisol (p = 0.421, RESI=0.00). There was a diagnosis-by-evening cortisol interaction (p = 0.001, RESI = 0.25), but no diagnosis-by-morning cortisol interaction; the ASD CDI scores were flat across evening cortisol values (p = 0.824), however increasing evening cortisol in the interval 0.11-0.80 nmol/L was associated with increasing CDI in the TD group (p = 0.005). We observed a significant morning cortisol by age interaction (p = 0.029, RESI = 0.18). CONCLUSIONS: Results replicate previous findings in autism showing higher depressive symptoms, but do not show a clear association with cortisol levels. Elevations in evening cortisol were associated with higher depressive symptoms in neurotypical youth; a link previously found in non-autistic adolescents and adults.