Total Hypoxic Burden as a Supplementary Metric in Mild Obstructive Sleep Apnea

Objectives: The apnea–hypopnea index (AHI) does not capture desaturation depth or duration. We examined whether total hypoxic burden (THB) describes hypoxemic variation beyond event frequency and whether a THB threshold identifies a higher-risk subgroup in mild obstructive sleep apnea (OSA; 5≤AHI <15).Methods: We retrospectively analyzed 924 patients undergoing overnight polysomnography at a single tertiary center. THB was the cumulative desaturation area. Within mild OSA, the ≥25%·min/h threshold was the subgroup mean plus one standard deviation. Hierarchical logistic models were adjusted for age, sex, body mass index (BMI), and AHI. Incremental value was assessed by likelihood ratio tests and DeLong comparisons of areas under the curve (AUCs).Results: Mean age was 47.96±13.31 years; 73.6% were male. AHI and THB correlated strongly (r=0.848). Within mild OSA (n=230), THB varied 20.9-fold across a 3.0-fold AHI range (R2=0.389). THB ≥25 (n=483) was crudely associated with hypertension, diabetes mellitus, arrhythmia, and stroke/cerebrovascular accident (CVA), but none persisted after adjustment for age, sex, and BMI (hypertension odds ratio, 1.23; 95% confidence interval, 0.85 to 1.77). For the first three, adding THB ≥25 to adjusted AHI models negligibly changed AUCs (ΔAUC≤0.004); stroke/CVA modeling was unstable. Within mild OSA, no excess comorbidity was detected among 35 patients with THB ≥25.Conclusions: THB quantifies hypoxemic heterogeneity within AHI-defined categories, but its association with comorbidities was not independent of age, sex, and BMI; the ≥25%·min/h threshold remains exploratory.