Demographic, clinical, and biomarker predictors of apnea-hypopnea index from polysomnography in Midwestern sleep clinic patients.

Background: To examine if selected demographic (age, gender), clinical (diabetes, coronary artery disease, hyperlipidemia, myocardial infarction, stroke, lung disease, smoking history, alcohol intake), and biomarker [blood pressure (BP), heart rate, body mass index (BMI), neck circumference, Mallampati score] variables are predictors of apnea-hypopnea index (AHI) from polysomnography (PSG).

Methods: This cross-sectional study recruited a sample of adults (N = 170) who were being evaluated for OSA. Participants completed self-reported demographic and clinical questionnaires, and then completed PSG (n = 142). Multi-collinearity was assessed. Confounding factors, correlations, and potential interactions were explored.

Results: The final regression model was performed on 130 participants; 61 (46.9%) had an AHI ≥ 15. Systolic and diastolic BPs were highly correlated. Interactions were tested between gender and other variables (high cholesterol, BMI, neck circumference, systolic BP) and between systolic BP and other variables (high cholesterol, BMI, neck circumference, and lung disease). No interactions occurred between gender or systolic BP and other variables, meaning that the effects of the variables on AHI levels from PSG did not vary depending on gender or systolic BP. BMI, systolic BP, and absence of lung disease were predictors for AHI levels ≥ 15 from PSG.

Conclusions: BMI and systolic BP were significant predictors of OSA in this study. The absence of lung disease as a significant predictor was unique and may be due to the small number of participants who self-reported lung disease. To our knowledge, this is the first study to report this combination of variables to predict AHI levels ≥ 15 from PSG.