Impact of sleep apnea severity and nocturnal hypoxemia on the triglyceride-rich lipoproteins particles: the ELSA-Brasil study.

Intermittent hypoxia, a hallmark of obstructive sleep apnea (OSA), promotes impaired triglyceride-rich lipoprotein (TRL) clearance in mice models, but data in humans are unclear. We evaluated the association of total TRL-p, its subclasses and TRL mean size with OSA severity and nocturnal hypoxemia in participants with sleep study and lipid profile assessed by nuclear magnetic resonance spectroscopy. OSA was classified as none, mild, moderate, and severe. Sleep time under SpO2 <90% (TSA90) was categorized into tertiles and as 0%, >0-1%, 1%-10%, and ≥10%. Non-linear associations were tested using restricted cubic spline (RCS) regression. Models were adjusted for age, sex, race, hypertension, diabetes, smoking, low-density lipoprotein, and waist circumference. Thus, 2,397 participants (49 ± 8 years, 57% women) were included (33.5% moderate/severe OSA). After full adjustment, large/very large TRL-p were associated with mild (P = 0.016) and moderate (P = 0.032) but not with severe OSA. TRL mean size was associated with TSA90 tertiles 2-3, driven by large/very large TRL-p, irrespective of sex. RCS analysis confirmed a non-linear dose-response between TSA90 and large/very large TRL-p (P = 0.021) and TRL mean size (P = 0.011), characterized by a rise-and-plateau pattern with maximal effect at TSA90 ∼3% and progressive attenuation thereafter. Sensitivity analyses considering insulin resistance, lipid-lowering and anti-hypertensive drugs have shown similar results. Summary, large/very large TRL-p are independently associated with mild-moderate OSA and show a non-linear, rise-and-plateau dose-response with nocturnal hypoxemia, saturating at hypoxic burdens around 3% of sleep time-suggesting that even minimal hypoxic burden engages atherogenic TRL pathways and challenging AHI-based risk stratification.