Anemia of aging and obstructive sleep apnea

World Health Organization defined anemia of aging (AOA) when men and women greater than 65 years, respectively, have unexplained hemoglobin (Hgb) less than 13 and 12 g/dl. Recent evidence suggests that this is likely a chronic inflammatory process involving interleukins (IL) 6, 12, and C-reactive protein. Among elderly with obstructive sleep apnea (OSA), hypoxic stimulation of erythropoiesis may obscure AOA. Treatment of OSA may paradoxically restore AOA. We sought to identify OSA and AOA coexistence and OSA treatment AOA interaction.

Records of 101 successive patients older than 65 years and with OSA who were treated with continuous positive airway pressure were analyzed retrospectively. Differences among pre/post-treatment of OSA hemograms were assessed using paired two-tailed Student's t test. Hemogram changes were compared to apnea-hypopnea index (AHI), respiratory effort related arousals (RERA), and duration patients slept with oxyhemoglobin saturation <89% (hypoxic time (HT)) and were assessed for correlative significance using Pearson coefficient correlation.

Eighty-two of one hundred one patients (27 men, 55 women; average age 71 years) had charted all the study data variables cited above. Mean pre-treatment and mean 1-year post-treatment of OSA Hgb/hematocrit (Hct) for men and women, respectively, were 13 g/dl/40.7% to 12.7 g/dl/39.1% and 12.1 g/dl/38.1% to 11.9 g/dl/37.6%. Hct changed significantly among both men and women (p < 0.05). Among 56% and 30% of the 82 patient study cohort, 1-year post-treatment of OSA, Hct declined (mean 4.8%) and increased (mean 3.7%), respectively; both changes were statistically significant (p < 0.01). These changes did not correlate significantly with AHI, RERA, or HT.

Among the entire cohort for both men and women, we did not see AOA before OSA treatment and we did see AOA 1 year after OSA treatment. However, post-treatment of OSA Hct distributed bimodally, with significant increases and declines of Hct. While these Hct changes did not correlate significantly with selected sleep-breathing variables, we remain intrigued by a possible AOA-OSA interaction. AOA and OSA share common inflammatory processes. We believe OSA inflammatory processes interact with OSA hypoxia-induced erythropoiesis. The balance of these sets of processes determines the effect of OSA and OSA treatment on AOA.