Examining the Complex Relation Between Sleep Problems and Chronic Kidney Disease

Abstract

Sleep-related problems are rapidly growing public-health concerns that often result in adverse short-term consequences (e.g., reduced quality of life) and long-term health effects (e.g., cardiovascular disease). However, the relation between sleep problems and chronic kidney disease (CKD) has been under-investigated. To better understand this relation, we addressed three aims. The first aim was to examine temporal trends in the prevalence of 5 self-reported sleep problems in U.S. adults and their associations with CKD and all-cause mortality, using data from 5 National Health and Nutrition Examination Surveys (2005-14). The prevalence of trouble sleeping and diagnosed sleep disorder increased over the decade, while nocturia (urinating ≥2 times/night), inadequate sleep (<7 hours/night), and excessive sleep (>9 hours/night) remained stable. All sleep problems, except inadequate sleep, were more common among adults with CKD than without CKD, especially for excessive sleep and nocturia, which were positively associated with all-cause mortality. Second, we conducted a large retrospective cross-sectional study of obstructive sleep apnea (OSA) in U.S. veterans who sought care in Veteran Administration (VA) facilities in fiscal year (FY) 2018 to better understand the population burden of OSA and its relation with CKD and its risk factors. Using data from 6.2 million veterans for FY2014-18, we estimated OSA point prevalence at the last visit to a VHA facility in FY2018 (index time T) and period prevalence of OSA (excluding prevalent cases at T) going back to the start of FY2014. Period prevalence for each duration was estimated by taking into account left censoring of veterans followed back for different durations. At time T, OSA point prevalence was 24.9% in veterans with CKD and 15.2% in those without CKD. The overall 60-month period prevalence was 11.6% and was positively associated with CKD, obesity, being male, having hypertension, or diabetes, and inversely associated with age ≥65. Lastly, we conducted a large retrospective cohort study of U.S. veterans to test the hypothesis that CKD is a mediator in the causal pathway linking race/ethnicity with OSA incidence. Four statistical methods of mediation analysis with different advantages and limitations were used: informal difference method, 4-way decomposition, flexible mediation analysis, and dynamic path analysis. Blacks and Hispanics had higher incidence rates than did non-Hispanic Whites. The percentages of the total race/ethnicity effects mediated by CKD were small and similar using all 4 mediation methods; e.g., using flexible mediation analysis, the percentage of the Black/White effect on OSA incidence mediated by CKD was 5.8%. However, when CKD and its 3 risk factors were treated jointly as mediators in flexible mediation analysis, the percentage mediated increased to 30.3%. The high prevalence of sleep problems including OSA among persons with CKD, their associations with mortality, and the mediated effect by CKD on racial disparity in OSA incidence suggest their potential importance to clinical practice. Future work could address the feasibility of early identification, objective characterization and management of sleep problems among patients with CKD, and studying the effects of proactive practices including control of mediators (CKD) on disease progression and other outcomes.