Triglyceride-glucose index and nocturnal oxygenation impairment in patients with obstructive sleep apnea

In obstructive sleep apnea (OSA), respiratory events differ not only in frequency but also in the depth and duration of oxygen desaturation. The triglyceride-glucose (TyG) index is a routinely available marker related to insulin resistance, but its relationship with conventional nocturnal oxygenation-related parameters in OSA remains incompletely clarified. Objective: We examined the association of the TyG index with hypoxemia-related sleep parameters and explored whether these associations were influenced by monitoring modality and AHI-defined OSA severity.

This single-center retrospective cross-sectional study screened 154 consecutive symptomatic adult inpatients from 2021 to 2025; after exclusion of 11 patients, 143 were included in the retrospective dataset. Diagnostic sleep monitoring was performed using in-laboratory polysomnography or Nox T3 portable monitoring. Laboratory measurements were obtained within 3 days before or after sleep monitoring. Severe nocturnal oxygen desaturation was defined as lowest SpO₂ <80%, and severe AHI-defined OSA was defined as AHI ≥30 events/h.

TyG was available in 116 patients, and 101 had complete TyG, lowest SpO₂, and AHI data. TyG was higher in patients with lowest SpO₂ <80% than in those with lowest SpO₂ ≥80% [9.252 (8.872, 9.577) vs. 8.780 (8.574, 8.963), P<0.001], and higher in patients with AHI ≥30 events/h than in those with AHI <30 events/h [9.242 (8.869, 9.624) vs. 8.823 (8.571, 9.121), P<0.001]. TyG correlated positively with AHI and CT90 and negatively with mean and lowest SpO₂. The AUC was 0.760 for lowest SpO₂ <80% and 0.715 for AHI ≥30 events/h. In models adjusted for age, sex, BMI, and monitoring modality, TyG remained associated with lowest SpO₂ <80% and AHI ≥30 events/h. However, the association with lowest SpO₂ <80% was attenuated after additional adjustment for AHI.

Higher TyG was associated with worse nocturnal oxygenation-related parameters and greater AHI-defined OSA severity, but these exploratory findings may partly reflect overall OSA severity. TyG may provide a simple metabolic clue for risk stratification, but it cannot replace formal sleep monitoring.