Self-reported sleep duration and prediction of proteinuria: a retrospective cohort study

Background: Although multiple studies have shown that sleep duration is a predictor of cardiovascular diseases and mortality, few studies have reported an association between sleep duration and chronic kidney disease.

Study design: Retrospective cohort study.

Setting & participants: 6,834 employees of Osaka University aged 20-65 years who visited Osaka University Healthcare Center for their mandatory annual health examinations between April 2006 and March 2010 and did not have estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m(2), proteinuria, or treatment for self-reported kidney disease.

Predictor: Self-reported questionnaires about life style, including sleep duration, and blood and urine testing at the first examinations during the study period. An association between sleep duration and outcome was assessed using multivariate Poisson regression models adjusting for clinically relevant factors.

Outcome: Time to the development of proteinuria defined as 1+ or higher by dipstick test.

Results: Self-reported baseline sleep duration was 6.0 ± 0.9 hours, which reflected the mean sleep duration during a median of 2.5 (25th-75th percentile, 1.4-3.9) years of the observational period. Development of proteinuria was observed in 550 employees (8.0%). A multivariate Poisson regression model clarified that shorter sleep duration, especially 5 or fewer hours, was associated with the development of proteinuria in a stepwise fashion (vs 7 hours; incidence rate ratios of 1.07 [95% CI, 0.87-1.33; P = 0.5], 1.28 [95% CI, 1.00-1.62; P = 0.05], and 1.72 [95% CI, 1.16-2.53; P = 0.007] for 6, 5, and ≤4 hours, respectively), along with younger age, heavier current smoking, trace urinary protein by dipstick test, higher eGFR, higher serum hemoglobin A(1c) level, and current treatment for heart disease. A stepwise association between shorter sleep duration and the development of proteinuria also was verified in 4,061 employees who did not work the night shift.

Limitations: Self-reported sleep duration might be biased. Results in a single center should be confirmed in the larger cohort including different occupations.

Conclusion: Short sleep duration, especially 5 or fewer hours, was a predictor of proteinuria.