Abstract
BACKGROUND: Hypertrophic tonsillar tissue in children with obstructive sleep apnea (OSA) has enhanced expression of glucocorticoid receptors, which may reflect low endogenous cortisol levels. We have evaluated the effect of the interaction between tonsillar hypertrophy and OSA severity on morning serum cortisol levels. METHODS: Children with and without snoring underwent polysomnography, tonsillar size grading, and measurement of morning serum cortisol. RESULTS: Seventy children (2-13 years old) were recruited: 30 with moderate-to-severe OSA (apnea-hypopnea index [AHI] > 5 episodes/h), 26 with mild OSA (AHI > 1 and ≤ 5), and 14 controls (no snoring; AHI ≤ 1). Tonsillar hypertrophy was present in 56.7%, 53.8%, and 42.9% of participants in each group, respectively. Application of a general linear model demonstrated a significant effect of the interaction between severity of OSA and tonsillar hypertrophy on cortisol levels (P = 0.04), after adjustment for obesity, gender, and age. Among children with tonsillar hypertrophy, subjects with moderate-to-severe OSA (n = 17; AHI 14.7 ± 10.6), mild OSA (n = 14; AHI 2.3 ± 1.2), and control participants (n = 6; AHI 0.7 ± 0.2) were significantly different regarding cortisol levels (P = 0.02). Subjects with moderate-to-severe OSA had lower cortisol (16.9 ± 8.7 mcg/dL) than those with mild OSA (23.3 ± 4.2; P = 0.01) and those without OSA (controls) (23.6 ± 5.3 mcg/dL; P = 0.04). In contrast, children with normal-size tonsils and moderate-to-severe OSA, mild OSA, and controls did not differ in cortisol levels. CONCLUSIONS: Children with moderate-to-severe obstructive sleep apnea and the phenotype of hypertrophic tonsils have reduced morning serum cortisol levels and potentially decreased glucocorticoid inhibitory effects on tonsillar growth.