More information about text formats
We read with interest the paper by Verhulst et al 1 and there are a few areas that warrant further discussion.
On the accuracy of sleep polysomonography study, the accuracy of the thoracoabdominal strain gauge used in their study requires elaboration. Although strain gauge is one of the qualitative methods to measure thoracoabdominal circumference, it is not very sensitive to detect shallow...
On the accuracy of sleep polysomonography study, the accuracy of the thoracoabdominal strain gauge used in their study requires elaboration. Although strain gauge is one of the qualitative methods to measure thoracoabdominal circumference, it is not very sensitive to detect shallow breathing as seen in cases of patients with myopathy2. Hence, use of strain gauge may result in mislabeling of obstructive apnea as central apnea4.A better alternative would be respiratory inductive plethysmograph (RIP) that measures the change of the cross-sectional area of rib cage and abdomen. It is more sensitive and accurate in detecting lung volume change.2 According to the American Academy of Sleep Medicine, a quantitative measurement of ventilatory effort such as esophageal manometry, calibrated respiratory inductance plethysmography or diaphragmatic / intercostals EMG) should be performed for classifying obstructive, central or mixed hypopnea.3 The authors should also provide and justify the upper limit of normal for central apnea / hypopnea in their study. The absence of end-tidal CO2 monitor in the PSG set-up also decreased the sensitivity of diagnosing obstructive sleep-disordered breathing, i.e. obstructive sleep hypoventilation defined as end-tidal CO2 more than 45 mm Hg for more than 60% of total sleep time or more than 55mmHg for more than 8% of total sleep time.4 Verhulst et al. should also report the arousal index and wake after sleep onset (WASO) so as to allow comparison between normal and the apneic group.
On the analysis of data, this study evaluated the risk factors for obstructive sleep apnea and central apnea in a clinical sample of children referred to a pediatric obesity clinic. Verhulst et al concluded that “none of the anthropometric variables was a significant predictor for the presence of mild obstructive sleep apnea” although this conclusion was not supported by any quantitative analysis. Also the presentation of the results in Verhulst et al’s study did not conform to the standard suggested by Moss et al,5 e.g. inclusion of odds ratio and confidence intervals for the final model. The readers cannot judge for themselves if the conclusion was a result of insufficient statistical power. Some very important results were not presented, for example, the prevalence of tonsils hypertrophy, data on waist-hip ratio despite the fact that these results were used in their partially-disclosed logistic regression results.
We have conducted a systematic review of the literature and found that the prevalence of obstructive sleep apnea syndrome in obese children ranged from 13% to 36%.6 We also demonstrated a dose-response relationship between obstructive sleep apnea and obesity.7 We suspected that Verhulst et al’s failure to corroborate our findings was due to a type II error and misdiagnosis of obstructive sleep apnea as central apnea.