Urinary Potassium Excretion and Progression of CKD

Abstract

Background and objectives: Data on whether low or high urinary potassium excretion is associated with poor kidney outcome have been conflicting. The aim of this study was to clarify the association between urinary potassium excretion and CKD progression.

Design, setting, participants, & measurements: We investigated the relationship between lower urinary potassium excretion and CKD progression and compared three urinary potassium indices among 1821 patients from the Korean Cohort Study for Outcome in Patients with CKD. Urinary potassium excretion was determined using spot urinary potassium-to-creatinine ratio, spot urinary potassium concentration, and 24-hour urinary potassium excretion. Patients were categorized into four groups according to quartiles of each urinary potassium excretion metric. The study end point was a composite of a ≥50% decrease in eGFR from baseline values and ESKD.

Results: During 5326 person-years of follow-up, the primary outcome occurred in 392 (22%) patients. In a multivariable cause-specific hazard model, lower urinary potassium-to-creatinine ratio was associated with higher risk of CKD progression (adjusted hazard ratio, 1.47; 95% confidence interval, 1.01 to 2.12) comparing the lowest quartile with the highest quartile. Sensitivity analyses with other potassium metrics also showed consistent results in 855 patients who completed 24-hour urinary collections: adjusted hazard ratios comparing the lowest quartile with the highest quartile were 3.05 (95% confidence interval, 1.54 to 6.04) for 24-hour urinary potassium excretion, 1.95 (95% confidence interval, 1.05 to 3.62) for spot urinary potassium-to-creatinine ratio, and 3.79 (95% confidence interval, 1.51 to 9.51) for spot urinary potassium concentration.

Conclusions: Low urinary potassium excretion is associated with progression of CKD.

Keywords: Confidence Intervals; Disease Progression; Follow-Up Studies; Kidney Failure, Chronic; Potassium; Proportional Hazards Models; Renal Insufficiency, Chronic; Urinary potassium excretion; chronic kidney disease; creatinine; glomerular filtration rate; kidney.

Graphical abstract

Figure 1.

Cumulative incidence of CKD progression according to baseline spot urinary potassium-to-creatinine ratio. Adverse kidney events occurred more as urinary spot urinary potassium-to-creatinine ratio was lower (P<0.001 for the Gray test). Q, quartile.

Figure 2.

The relationships between three urinary potassium metrics and CKD progression. Cubic spline analyses showed that there were inverse associations of spot urinary potassium-to-creatinine ratio (A), spot urinary potassium concentration (B), and 24-hour urinary potassium excretion (C) with CKD progression. All curves represent multivariable-adjusted hazard ratios. Hazard ratios were adjusted for age, sex, smoking status, body mass index, comorbid disease (diabetes and cardiovascular disease), mean arterial pressure, eGFR, C-reactive protein level, intact parathyroid hormone level, serum albumin level, serum calcium level, spot urinary Na-to-Cr ratio (24-hour urinary sodium excretion for 24-hour urinary potassium excretion), spot urinary protein-to-creatinine ratio (24-hour urinary protein for 24-hour urinary potassium excretion), LDL cholesterol level, and renin-angiotensin system blocker, statin, and diuretics usage. The histograms in A–C represent the frequency distribution of spot urinary potassium-to-creatinine ratio, spot urinary potassium, and 24-hour urinary potassium, respectively. Median value of each urinary potassium metric was set as a reference.

Figure 3.

Subgroup associations of urinary potassium-to-creatinine ratio with CKD progression. The significant association between urinary potassium excretion and CKD progression was seen in most subgroups. Cause-specific hazard ratios (HRs; 95% confidence intervals [95% CIs]) per 1-SD lower spot urinary potassium-to-creatinine ratio (millimoles per gram) were on log scale. HRs were adjusted for age, sex, smoking status, body mass index, comorbid disease (diabetes and cardiovascular disease), mean arterial pressure, eGFR, high-sensitivity C-reactive protein level, intact parathyroid hormone level, serum albumin level, serum calcium level, spot urinary Na-to-Cr ratio, spot urinary protein-to-creatinine ratio, LDL cholesterol level, and renin-angiotensin system blocker, statin, and diuretics usage. UPCR, urinary protein-to-creatinine ratio.