Relationship Between Salivary Cortisol and Serum Testosterone in Boys with Autism

Background:  Data are limited regarding the hormonal profile of children with autism spectrum disorders. Adrenal hormones such as cortisol are increased in conditions of stress and in turn mediate the physiological response to stress. Some studies have reported that children with ASD have a lower diurnal variation of cortisol compared with controls. Data are conflicting regarding levels of gonadal hormones such as testosterone in ASD, with some (but not all) studies reporting higher testosterone levels in this condition. Data are also conflicting regarding levels of thyroid hormones in ASD.

Objectives:  Our objective was to examine levels of cortisol, testosterone, TSH and free thyroxine in peripubertal boys with ASD versus controls, and particularly to determine the diurnal variation in salivary cortisol and its relationship with serum testosterone and free T4 levels.

Methods:   In 18 peripubertal boys (mean age 10.5±0.4 years) with ASD and 19 age matched controls (11.2±0.3 years) (p=0.23) 8-14 years old, we measured  morning levels of testosterone, IGF-1 and free T4, and  8 AM and 11 PM salivary cortisol .

Results:  The morning levels of serum testosterone, IGF-1, free T4 and TSH, and 8 AM salivary cortisol did not differ in children with ASD compared with controls. However, the 11 PM salivary cortisol was higher (1.44±0.37 vs. 0.47±0.16 nmol/L, p=0.004), and the ratio of AM/PM salivary cortisol lower in boys with ASD compared with controls [reported as median (25^th quartile- 75^thquartile) using a non-parametric test (Wilcoxon rank sum test); ASD: 1.14 (0.60-3.81); Controls: 3.36 (1.96-6.33), p=0.03)]. In controls, testosterone levels correlated positively with IGF-1 (rho=0.59, p=0.008) and 11 PM cortisol (rho=0.60, p=0.01), and inversely with free T4 (rho= -0.48, p=0.04). In boys with ASD, associations of testosterone with IGF-1 (rho=0.68, p=0.003) and with free T4 levels (rho= -0.60, p=0.01) were preserved, however, the association of testosterone with 11 PM cortisol was lost (rho=0.14, p=0.60). There were no associations of testosterone, IGF-1 or free T4 with the AM/PM cortisol ratio in either group. 

Conclusions:  This is the first study to describe a loss in the relationship of PM salivary cortisol with AM serum testosterone in peripubertal boys with ASD compared to controls. . Associations of testosterone levels with other hormones such as IGF-1 and free T4 are preserved in boys with ASD. Further studies are necessary to better understand the implications of these findings.