Prevalence and correlates of sleep apnea among US Veterans with chronic kidney disease

The prevalence and correlates of sleep apnea (SA) among Veterans with chronic kidney disease (CKD), a population at high risk of both SA and CKD, are unknown. We performed a cross-sectional analysis of 248 Veterans (18-89 years) selected only for presence of moderate to severe CKD. All participants underwent full, unattended polysomnography, measurement of renal function and a sleepiness questionnaire. Logistic regression with backward selection was used to identify predictors of prevalent SA (apnea-hypopnea index [AHI, >= 15 events/hr] and prevalent nocturnal hypoxia [NH, % of total sleep time spent at <90% oxygen saturation]). The mean age of our cohort was 73.2 +/- 9.6 years, 95% were male, 78% were Caucasian and the mean body mass index (BMI) was 30.3 +/- 4.8 kg/m(2). The prevalence of SA was 39%. There was no difference in daytime sleepiness among those with and without SA. In the final model, older age, higher BMI and diabetes mellitus (DM) were associated with higher odds of SA, after controlling for age, BMI, race and sex. Higher BMI, DM, unemployed/retired status, current smoking and higher serum bicarbonate level were associated with prevalent NH. To sum, SA was common among Veterans with moderate to severe CKD. Although some traditional risk factors for SA were associated with SA in this population, sleepiness did not correlate with SA. Further study is needed to validate our findings and understand how best to address the high burden of SA among Veterans with CKD.