Circadian Rhythm of Salivary Cortisol in Obese Adolescents With and Without Apnea: A Pilot Study

BACKGROUND AND OBJECTIVE: Obstructive sleep apnea (OSA) and obesity are associated with stress system activation involving the hypothalamic-pituitary-adrenal (HPA) axis in adults, but these effects in childhood and adolescence remain unclear. We examined diurnal salivary cortisol as a measurement of the HPA axis function in obese adolescents with and without OSA and the relationships between cortisol levels, body weight, and parameters of polysomnography (PSG). METHODS: After PSG, saliva samples were collected from obese participants (with and without OSA) and lean participants four times over a 24-h period, namely, at 7:00 h (m-sCort), 13:00 h (a-sCort), 19:00 h (e-sCort), and 23:00 h (n-sCort). An enzyme-linked immunosorbent assay (ELISA) was used to measure salivary cortisol levels. The mean values of cortisol levels and fixed-time point diurnal cortisol slope (DCS) were calculated and compared among the three study groups. Correlations between parameters were analyzed using Spearman's correlation coefficients. RESULTS: Obese OSA participants had significantly higher e-sCort and n-sCort levels than both obese non-OSA participants and lean controls. However, m-sCort and a-sCort in these patients had a pronounced upward trend. M-sCort was significantly correlated with both the lowest oxygen saturation (SpO(2)) and time with SpO(2) <90%. Moreover, in the obese OSA group, DCS was significantly flatter than in the other two groups. The a-sCort in obese non-OSA participants was significantly higher than that in the lean control group and, surprisingly, was positively correlated with the apnea/hypopnea index. Additionally, m-sCort was related to body weight. CONCLUSION: This study provided further evidence for alterations in diurnal cortisol production in obese adolescents, which may indicate a chronically stressed HPA axis. However, there were significant differences in salivary cortisol parameters between participants with and without OSA. Furthermore, patients with OSA had more associations between time-point cortisol levels and OSA-related indices. Nonetheless, this research is a pilot study, and further investigations are necessary.