Adiposity in relation to age as predictor of severity of sleep apnea in children with snoring.

Correlation between obesity and obstructive sleep apnea has been documented in both adults and children. This investigation evaluated importance of body mass index (BMI) in relation to age as predictor of severity of obstructive sleep-disordered breathing (SDB). Children with habitual snoring referred for polysomnography were recruited. BMI Z score (> or =1.036 vs <1.036, i.e. at risk for overweight or overweight vs normal) was assessed as predictor of severity of SDB (apnea-hypopnea index [AHI] >five vs < or =five episodes per hour) at different ages (< or =6 vs >6 years). Two hundered eighty-four participants were recruited: 75 young children (4.6 +/- 1 years) with high BMI (1.9 +/- 0.7); 95 young subjects (4.5 +/- 1.1 years) with low BMI (-0.2 +/- 1.3); 55 older children (9.2 +/- 1.8 years) with high BMI (1.8 +/- 0.5); and 59 older participants (9.7 +/- 2.2 years) with low BMI (-0.2 +/- 1.1). Odds ratios for AHI >5 in young/high BMI children, young/low BMI subjects, and older/high BMI subjects relative to older/low BMI participants were: 6.5 (95% confidence interval 2.1-19.9), 7.3 (2.4-22) and 2 (0.6-7.3), respectively. Large tonsil size was associated with young age (odds ratio 1.97; 1.2-3.3). Among children with habitual snoring, adiposity does not predict severity of obstructive SDB in early childhood probably due to the prominent role of adenotonsillar hypertrophy. However, it may have a more important contribution to severity of SDB in older children.

Nocturnal change of circulating intercellular adhesion molecule 1 levels in children with snoring.

Hypoxia promotes adherence of leukocytes to endothelial cells by inducing expression of adhesion molecules like intercellular adhesion molecule 1 (ICAM-1). Increased serum levels of circulating ICAM-1 (cICAM-1) have been reported in adults with sleep apnea and associated hypoxemia. This investigation assessed the hypothesis that the overnight change of cICAM-1 levels in children with snoring is correlated with the severity of obstructive sleep-disordered breathing. Evening and morning serum levels of cICAM-1 were measured in children with snoring referred for polysomnography. Twenty-five children with an apnea-hypopnea index greater than or equal to 5 episodes/h (5.5 +/- 1.8 years), 30 subjects with an index less than 5 and greater than 1 (6.3 +/- 2 years), and 19 children with an index less than or equal to 1 (7.1 +/- 3 years) were recruited. Overnight change in cICAM-1 (log-transformed ratio of morning-to-evening levels) was similar in subjects with an apnea-hypopnea index greater than or equal to 5 episodes/h compared to those with an index less than 5 and greater than 1 or to children with an index less than or equal to 1 (-0.001 +/- 0.08 vs -0.03 +/- 0.09 vs -0.06 +/- 0.1; p > 0.05). When multiple regression analysis was applied, apnea-hypopnea index, respiratory arousal index, and oxygen saturation of hemoglobin nadir were not significant predictors of overnight change in cICAM-1 levels. Thus, in children with snoring, overnight change in cICAM-1 levels is not related to severity of obstructive sleep-disordered breathing.

Obstructive sleep-disordered breathing and plasma levels of vascular endothelial growth factor in children.

Vascular endothelial growth factor (VEGF) may be one of the pathophysiologic links in the association between obstructive sleep apnea-hypopnea and cardiovascular disease. Morning serum VEGF levels are increased in children with obstructive sleep apnea. However, release of VEGF by platelets and leukocytes during blood clotting may affect its concentration in serum. In the present study, VEGF levels were measured in children with and without habitual snoring using plasma specimens. Evening and morning plasma VEGF concentrations were determined in: (i) 20 children with habitual snoring and apnea-hypopnea index (AHI)5 episodes/h (median age 5; range 1.9-13 years); (ii) 55 children with snoring and AHI<5 episodes/h (median age 6; 2-13 years); and (iii) 25 controls without snoring (median age 6.5; 3-13 years). No differences were identified between the three study groups regarding evening [median 2.5 (range 2.5-174.5) versus 22.5 (2.5-159.4) versus 26.8 (2.5-108) pg/mL; P>0.05] and morning VEGF levels [median 7.7 (range 2.5-120.5) versus 25.1 (2.5-198.4) versus 48.4 (2.5-147.7) pg/mL; P>0.05]. AHI and % sleep time with oxygen saturation of hemoglobin less than 90% were not significant predictors of log-transformed morning VEGF concentrations (P>0.05). In summary, both evening and morning plasma VEGF levels were similar in children with obstructive sleep-disordered breathing of variable severity and in controls without snoring. VEGF may not play an important pathophysiologic role in all cases of obstructive sleep-disordered breathing in childhood.

Obstructive sleep-disordered breathing and fasting insulin levels in nonobese children.

A positive correlation of severity of sleep-disordered breathing with morning fasting insulin levels, which is independent of obesity, was reported in adults and obese children. We hypothesized that both severity of sleep-disordered breathing and relative body mass index predict fasting insulin and homeostasis model assessment (HOMA) index values in nonobese children with habitual snoring. One hundred and ten subjects with habitual snoring (median age, 6 years; range, 2-13 years) underwent polysomnography and measurement of morning fasting insulin and glucose levels. The HOMA index was calculated. Thirty children had an apnea-hypopnea index (AHI) >/= 5 episodes/hr (median, 7.8 episodes/hr; range, 5-42.3 episodes/hr), and 80 subjects had an AHI < 5 episodes/hr (median, 1.9 episodes/hr; range, 0.2-4.9 episodes/hr). Insulin and HOMA index values were similar in children with AHI >/= 5 episodes/hr (median insulin, 4.9 mU/l; range, 1.66-19.9 mU/l; and median HOMA, 1; range, 0.36-4.95) and in subjects with AHI < 5 episodes/hr (median insulin, 5.8 mU/l; range, 0.74-41.1 mU/l; and median HOMA, 1.3; range, 0.13-9.72) (P > 0.05). No significant correlations were identified between insulin or HOMA index values and any polysomnography indices (P > 0.05). When multiple linear regression was carried out, relative body mass index was a significant predictor of log-transformed insulin levels or HOMA index values, but AHI and percentage of sleep time with saturation <95% were not. In conclusion, contrary to findings in adults and in obese children, severity of sleep-disordered breathing is not a significant predictor of fasting insulin or HOMA index values in nonobese children with habitual snoring.