Growing epidemiological evidence supports a link between sleep duration and obesity in children and adolescents. 1 Furthermore, a recent study demonstrated an association between short sleep duration and insulin resistance. 2 Identification of other conditions, such as sleep deprivation, which could be additionally implicated in the development of both obesity and its associated cardiovascular risk factors, is urgently needed in the treatment and prevention of obesity. Impaired sleep could certainly be one of these conditions as it influences both energy intake and expenditure. Shorter sleep duration results in tiredness that may hamper physical activity, and alters metabolic hormones to increase appetite and affect food selection. Additionally, extra time awake provides increased opportunity for food intake. 1 We previously reported an independent association between sleepdisordered breathing and both glucose intolerance and dyslipidaemia, but in that paper we did not include information on sleep duration. 3 We therefore retrospectively analysed the interaction between parentally-reported sleep duration and metabolic dysregulation in a sample of overweight and obese children and adolescents attending a weight-loss program at our paediatric obesity clinic. The inclusion criteria and details of the various measurements were described previously. 3 We used data on body fat distribution (body mass index (BMI) z score, waist circumference and waist-hip ratio), sleepdisordered breathing severity (respiratory disturbance index, oxygen desaturation index,, SaO2. and SaO2nadir) and metabolic dysregulation (fasting glucose, insulin, C-peptide, cholesterol, HDL-cholesterol and triglycerides; area under the curve for glucose, insulin and C-peptide during oral glucose tolerance test; and HOMA-index reflecting insulin resistance). Sleep duration was assessed by questionnaire and subjects were divided according to sleep duration based on published normative data and our own unpublished data. 4 Cut-off values were 10 h of sleep for 6–10-year-old children, 9. 5 h for 10–12-year-old children and 9 h for subjects older than 12. Table 1 presents the subject characteristics according to sleep duration. The difference in waist circumference did not persist after correction for age (p= 0.2), and the difference in area under the curve for C-peptide did not persist after controlling for waist circumference (p= 0. 097). On the other hand, the observed difference in C-peptide did remain after correcting for both age and waist circumference (p= 0.02). Table 2 presents correlation coefficients between sleep duration and certain obesity-related factors. The correlation between sleep duration and waist circumference became borderline significant after controlling for age (p= 0.049). The other coefficients became insignificant after additionally controlling for waist circumference.

In summary, our study partly supports the hypothesis of an interaction between shorter sleep duration and obesity (waist circumference) and early metabolic abnormalities (C-peptide). Our analyses also show that future studies on this subject should take into account the confounding effects of age and abdominal obesity. In view of these