Abstract
Background and Aims Sleep disturbances are common in patients on dialysis, but little is known about their prevalence and determinants in earlier chronic kidney disease (CKD) stages. We assessed the prevalence of insomnia, excessive daytime sleepiness (EDS), and sleep apnea, and studied their relation with kidney function, before and after accounting for various other potential determinants in patients with moderate or severe CKD. Method This study included 2657 nephrology outpatients with non-dialysis dependent CKD (mean age, 67 ± 12 years; 66% men; mean eGFR, 33 ± 12 mL/min/1.73 m2) from the French CKD-REIN (CKD-Renal Epidemiology and Information Network) cohort study who completed the Basic Nordic Sleep Questionnaire at baseline. Insomnia was defined as having badly or rather badly been sleeping, or having had difficulties falling asleep or awakened too early without being able to fall asleep again at least 3 times per week, over the past 3 months. And EDS, as feeling excessively sleepy during daytime at least 3 times per week over the past 3 months. Sleep apnea with or without continuous positive airway pressure use were identified from medical records. For each sleep disturbance, we assessed the crude prevalence by sex and by estimated glomerular filtration rate (eGFR) categories. We estimated adjusted odds ratios (OR) with 95% confidence interval (95% CI) associated with eGFR decrease (per 10 mL/min/1.73 m2) and urinary albumin-to-creatinine ratio (ACR) categories (A1-A3), as well as with age, sex, education (years), smoking, alcohol consumption, physical activity (estimated by the Global Physical Activity Questionnaire), performance score in Instrumental Activities of Daily Living [IADL], obesity, diabetes, cardiovascular disease, depression, patient-reported cramps, itching and stress or worry about CKD, and relevant drug use. Results Overall prevalence of insomnia, EDS, and sleep apnea were 23%, 15%, and 14%, respectively, with differences between women and men for insomnia, 30% vs 19%, EDS, 17% vs 13%, and sleep apnea, 10% vs 17%, respectively (all P-values <.001). Crude prevalence of EDS significantly increased with decreasing eGFR category (without interaction with sex), but not those for insomnia and sleep apnea (Figure). ACR categories were not associated with either sleep disturbances. In multivariable analyses, factors significantly associated with higher prevalence of insomnia included female sex, smoking, low physical activity, depression, and thyroid hormone use. There were also strong dose-response relations between insomnia prevalence and to what extend patients reported being bothered by cramps or itching, or stressed/worried about CKD. The crude relation of eGFR with EDS tended to weaken (ORs [95% CI], 1.10 [0.99-1.22] per 10 mL/min decrease, p-value 0.09) in multivariable analysis, which showed higher ORs associated with past smoking, depression, lower IADL score, insomnia, and betablocker use, as well as dose-response relations with itching, and stress/worry about CKD. Factors significantly associated with increased prevalence of sleep apnea included male sex, alcohol consumption, lower IADL score, obesity, depression, and cardiovascular disease. Conclusion This analysis of baseline data from the CKD-REIN cohort shows that, in patients with non-dialysis dependent CKD, EDS alone appeared to be related to kidney function level. The study, however, suggests the potential role of modifiable CKD-related symptoms, cramps and itching, and of being stressed or worried about CKD, in addition to that well-documented for behavioral and clinical risk factors, and drugs in the risk of insomnia and EDS.