Association by logistic regression of the three oxygenation variables with the four individual types of pain. Estimation of the change in odds ratio (95% CIs) reflected the effect of arterial desaturation (i.e., a reduction in average nocturnal Sao₂, a reduction in the minimum nocturnal Sao₂, and an increase in the % of total sleep time spent at an Sao₂ <90%) on pain. The odds ratio for each of the three oxygenation variables was estimated thrice: (1) unadjusted for any confounders (model 1), (2) with an adjustment for age, sex, race, body mass index, continuous airway pressure use during polysomnography, and analgesic/antiinflammatory medicine use (model 2), and (3) with an additional adjustment for sleep fragmentation (SF) and systemic inflammation (model 3). A decrease in the minimum nocturnal Sao₂ significantly increased the odds for “morning headache” (P = 0.009), “headache disrupting sleep” (P = 0.002), and “chest pain while in bed” (P = 0.004) even after adjusting for clinical confounders, SF, and systemic inflammation. A decrease in average nocturnal Sao₂ significantly increased “headache disrupting sleep” (P = 0.008) even after adjusting for clinical confounders, SF, and systemic inflammation. Sao₂ = arterial oxyhemoglobin saturation.