Are hypoxic burden and respiratory event duration superior to the apnea–hypopnea index in predicting disease severity and clinical outcomes in obstructive sleep apnea?

Obstructive Sleep Apnea Syndrome (OSAS) is a common disorder characterized by recurrent apnea and hypopnea episodes, with a prevalence that increases with obesity and aging. According to recent epidemiological data, approximately 1 billion people worldwide are affected by OSAS. Patients often present with symptoms such as snoring, witnessed apnea, excessive daytime sleepiness, morning headaches, non-restorative sleep, fatigue, and tiredness. These symptoms vary due to heterogeneity related to factors such as age, obesity, sex, and race. Although the Apnea-Hypopnea Index (AHI), measured through polysomnography (PSG), is currently the standard parameter used in clinical practice to diagnose and grade OSAS severity, it does not reflect the extent and duration of nocturnal oxygen desaturation. Additionally, its limited role in predicting diseases associated with OSAS, such as cerebrovascular disease, hypertension, coronary artery disease, heart failure, chronic kidney disease, and type 2 diabetes mellitus, has led to investigations into new diagnostic parameters. Hypoxic burden (HB) and respiratory event duration are among these novel markers. This study aimed to investigate the importance of HB and respiratory event duration in predicting OSAS severity and outcomes. In the study, patients who were admitted to the Sleep Disorders Diagnosis and Treatment Center of the Department of Chest Diseases at Tokat Gaziosmanpaşa University Research and Training Hospital with a preliminary diagnosis of sleep-disordered breathing between 01.02.2024 and 01.10.2024 and underwent PSG were evaluated using test records, pre-test sleep questionnaires, and PSG reports. A total of 721 patients were included, consisting of 497 (69%) males and 224 (31%) females, with a mean age of 47.64±11.44 years and a mean BMI of 31.88±5.82. Based on AHI, the cases were divided into three groups: mild OSAS (n=200, 28%), moderate OSAS (n=225, 31%), and severe OSAS (n=296, 41%). Parameters of respiratory event duration including average apnea duration, average hypopnea duration, total apnea-hypopnea duration, and all HB parameters (HB-TST, HB-REM, HB-NREM, HB-supine, HB-non-supine) showed significant differences between the groups. Regarding symptoms and comorbidities, symptoms such as snoring, witnessed ix apnea, EDS, and insomnia were found to be similar across all three groups classified by AHI. Among those with witnessed apnea, HB-TST, HB-NREM, and HB-supine values were significantly higher (p<0.05). Comparison of comorbidities revealed significant differences in the prevalence of HT, CAD, CVD, and DM between the groups (p<0.05). Correlation analyses showed that all HB parameters were strongly associated with traditional PSG parameters such as AHI (REM, NREM, supine, lateral), average SpO2, minimum SpO2, ODI, and T90. Among respiratory event duration parameters, total apnea-hypopnea duration showed a positive correlation with AHI, AHI-NREM, AHI-lateral, and ODI, and a negative correlation with minimum and average SpO2. ROC analyses were performed to determine cutoff values of HB parameters and respiratory event duration for predicting the presence of CVD. The cutoff value for total apnea-hypopnea duration was found to be 41.2 minutes, and for HB-TST it was 39.4. Multivariate logistic regression analyses were conducted to identify independent predictors for cardiac and metabolic diseases. Witnessed apnea, age, BMI, total apnea-hypopnea duration, AHI, ESS, and T90 were identified as independent predictors for CVD. In conclusion, considering the heterogeneity of OSAS in terms of risk factors, pathophysiological mechanisms, clinical presentation, and outcomes, defining such a broad-spectrum disease with a single parameter would oversimplify its complexity. Among the newly developed diagnostic indicators, HB appears to be one of the most promising. Further research is needed to determine the normal and abnormal levels of HB and to evaluate its role in monitoring treatment response. We believe the data obtained from our study will provide valuable insights for future research in this field.