Apnea-hypopnea index-oxygen desaturation index (AHI-ODI) discordance in obstructive sleep apnea: association with symptom burden and cardiometabolic comorbidity: A retrospective cross-sectional study.

BACKGROUND

The apnea-hypopnea index (AHI) counts respiratory events but not the depth or duration of accompanying desaturation. We examined AHI-ODI discordance (AHI minus oxygen desaturation index, ODI) and its associations with daytime symptom burden and prevalent cardiometabolic comorbidity.

METHODS

Single-center retrospective cross-sectional study of 200 adults with OSA (January 2023-January 2025). Five indices (AHI, ODI, mean SpO2, T90 and AHI-ODI discordance) were examined; excessive daytime sleepiness (EDS; Epworth Sleepiness Scale ≥ 11) was the primary outcome. Multivariable logistic regression adjusted for age, sex, BMI and an AHI-ODI × sex interaction.

RESULTS

The cohort was 77.5% male, median age 57 years (48.8-64.2). For EDS, AHI-ODI discordance showed the numerically highest AUC (0.639, p < 0.001), followed by T90 (0.627), and mean SpO2 and ODI (0.609 each); AHI was not significant (0.546, p = 0.264). No pairwise difference was significant. AHI-ODI discordance remained independently associated with EDS after adjustment (OR 1.04 per event/h, 95% CI 1.01-1.07, p = 0.008), with no significant sex interaction. Among secondary outcomes, T90 showed the highest AUCs for diabetes mellitus (0.644) and hypertension (0.615), and mean SpO2 for heart failure (0.608). No index discriminated against previous stroke or atrial fibrillation.

CONCLUSIONS

AHI-ODI discordance was independently, but only modestly, associated with EDS, whereas oxygenation indices were related to prevalent cardiometabolic comorbidity. The discrimination was insufficient, and all reported cut-offs are exploratory. AHI-ODI discordance is not a measure of sleep apnea-specific hypoxic burden.