Estimation of cardiovascular disease from polysomnographic parameters in sleep-disordered breathing

TURHAN M., Bostanci A., Bozkurt S.

EUROPEAN ARCHIVES OF OTO-RHINO-LARYNGOLOGY, vol.273, no.12, pp.4585-4593, 2016 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 273 Issue: 12
  • Publication Date: 2016
  • Doi Number: 10.1007/s00405-016-4176-1
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.4585-4593
  • Keywords: Bayesian network, Obstructive sleep apnea, Polysomnography, Cardiovascular disease, AMERICAN-HEART-ASSOCIATION, APNEA-HYPOPNEA, MORTALITY, PRESSURE, EVENTS, UPDATE, COHORT, INDEX, RISK
  • Akdeniz University Affiliated: Yes


We aimed to illustrate the causal relationships between cardiovascular diseases (CVDs) and various polysomnographic variables, and to develop a CVD estimation model from these variables in a population referred for assessment of possible sleep-disordered breathing (SDB). Clinical and polysomnographic data of 1162 consecutive patients with suspected SDB whose comorbidity status was known, were reviewed, retrospectively. Variable selection was performed in two steps using univariate analysis and tenfold cross validation information gain analysis. The resulting set of variables with an average merit value (m) of > 0.005 was considered to be causal factors contributing to the CVDs, and used in Bayesian network models for providing estimations. Of the 1162 patients, 234 had CVDs (20.1 %). In total, 28 parameters were evaluated for variable selection. Of those, 19 were found to be associated with CVDs. Age was the most effective attribute in estimating CVD (m = 0.051), followed by total sleep time with oxygen saturation < 90 % (m = 0.021). Some other important variables were apnea-hypopnea index during non-rapid eye movement (m = 0.018), lowest oxygen saturation (m = 0.018), body mass index (m = 0.016), total apnea duration (m = 0.014), mean apnea duration (m = 0.014), longest apnea duration (m = 0.013), and severity of SDB (m = 0.012). The modeling process resulted in a final model, with 76.9 % sensitivity, 96.2 % specificity, and 92.6 % negative predictive value, consisting of all selected variables. The study provides evidence that the estimation of CVDs from polysomnographic parameters is possible with high predictive performance using Bayesian network analysis.