Obstructive sleep apnea (OSA) may contribute to metabolic and inflammatory deregulation but previous studies failed to consider sleep duration, sleep fragmentation, insomnia, and daytime sleepiness as potential confounders. Consecutive non-diabetic middle-aged participants from the ELSA-Brasil cohortwere invited to perform a clinical evaluation, home sleep study for 1 night, and wrist actigraphy for 7 days. OSA was defined by an apnea-hypopnea index ≥ 15 events/h. Participants were stratified according to the presence of OSA measuring the following markers: fasting glucose, glucose tolerance test, homeostatic model assessment of insulin resistance (HOMA-IR) index, fasting insulin, insulin after 2 h of glucose load, glycated hemoglobin, total cholesterol and their fractions, triglycerides, C-reactive protein, TNF-alpha, interleukin-6, interleukin-10, leptin, adiponectin, E-selectin, ADMA, MCP-1, TGF, apolipoprotein B, fibrinogen, and lipoprotein(a). Differences between groups were identified by chi-square test and ANOVA. We studied 708 participants (mean age: 46 ± 5 years, men: 44%, BMI 26.1 ± 4.1 kg/m2). Compared to no OSA, participants withOSA presented higher levels while fasting and after 2 h glucoseloadof insulin, HOMA-IR, cholesterol, triglycerides, and C-reactive protein (all p < 0.001). After linear regression analysis adjusting for traditional risk factors plus sleep duration, fragmentation, insomnia, and daytime sleepiness, OSA was negatively associated with adiponectin (β = - 0.271 CI 95% - 0.456 - 0.085) and positively associated with cholesterol (β = 9.707 CI 95% 2.737 16.678). Sex-stratification revealed that these associations were significant for men but not women. In non-diabetic middle-age adults, men with OSA presented with lower adiponectin and higher cholesterol levels independently of sleep duration, sleepfragmentation, insomnia, and daytime sleepiness.