Serum Potassium and Risk of Death or Kidney Replacement Therapy in Older People With CKD Stages 4-5: Eight-Year Follow-up

Rationale & ObjectiveHypokalemia may accelerate kidney function decline. Both hypo- and hyperkalemia can cause sudden cardiac death. However, little is known about the relationship between serum potassium and death or the occurrence of kidney failure requiring replacement therapy (KRT). We investigated this relationship in older people with chronic kidney disease (CKD) stage 4-5.Study DesignProspective observational cohort study.Setting & ParticipantsWe followed 1,714 patients (≥65 years old) from the European Quality (EQUAL) study for 8 years from their first estimated glomerular filtration rate (eGFR) < 20 mL/min/1.73 m2 measurement.ExposureSerum potassium was measured every 3 to 6 months and categorized as ≤3.5, >3.5-≤4.0, >4.0-≤4.5, >4.5-≤5.0 (reference), >5.0-≤5.5, >5.5-≤6.0, and >6.0 mmol/L.OutcomeThe combined outcome death before KRT or start of KRT.Analytical ApproachThe association between categorical and continuous time-varying potassium and death or KRT start was examined using Cox proportional hazards and restricted cubic spline analyses, adjusted for age, sex, diabetes, cardiovascular disease, renin-angiotensin-aldosterone system (RAAS) inhibition, eGFR, and subjective global assessment (SGA).ResultsAt baseline, 66% of participants were men, 42% had diabetes, 47% cardiovascular disease, and 54% used RAAS inhibitors. Their mean age was 76 ± 7 (SD) years, mean eGFR was 17 ± 5 (SD) mL/min/1.73 m2, and mean SGA was 6.0 ± 1.0 (SD). Over 8 years, 414 (24%) died before starting KRT, and 595 (35%) started KRT. Adjusted hazard ratios for death or KRT according to the potassium categories were 1.6 (95% CI, 1.1-2.3), 1.4 (95% CI, 1.1-1.7), 1.1 (95% CI, 1.0-1.4), 1 (reference), 1.1 (95% CI, 0.9-1.4), 1.8 (95% CI, 1.4-2.3), and 2.2 (95% CI, 1.5-3.3). Hazard ratios were lowest at a potassium of about 4.9 mmol/L.LimitationsShorter intervals between potassium measurements would have allowed for more precise estimations.ConclusionsWe observed a U-shaped relationship between serum potassium and death or KRT start among patients with incident CKD 4-5, with a nadir risk at a potassium level of 4.9 mmol/L. These findings underscore the potential importance of preventing both high and low potassium in patients with CKD 4-5.Plain-Language SummaryAbnormal potassium blood levels may increase the risk of death or kidney function decline, especially in older people with chronic kidney disease (CKD). We studied 1,714 patients aged ≥65 years with advanced CKD from the European Quality (EQUAL) study and followed them for 8 years. We found that both low and high levels of potassium were associated with an increased risk of death or start of kidney replacement therapy, with the lowest risk observed at a potassium level of 4.9 mmol/L. In patients with CKD, the focus is often on preventing high blood potassium. However, this relatively high optimum potassium level stresses the potential importance of also preventing low potassium levels in older patients with advanced CKD. Hypokalemia may accelerate kidney function decline. Both hypo- and hyperkalemia can cause sudden cardiac death. However, little is known about the relationship between serum potassium and death or the occurrence of kidney failure requiring replacement therapy (KRT). We investigated this relationship in older people with chronic kidney disease (CKD) stage 4-5. Prospective observational cohort study. We followed 1,714 patients (≥65 years old) from the European Quality (EQUAL) study for 8 years from their first estimated glomerular filtration rate (eGFR) < 20 mL/min/1.73 m2 measurement. Serum potassium was measured every 3 to 6 months and categorized as ≤3.5, >3.5-≤4.0, >4.0-≤4.5, >4.5-≤5.0 (reference), >5.0-≤5.5, >5.5-≤6.0, and >6.0 mmol/L. The combined outcome death before KRT or start of KRT. The association between categorical and continuous time-varying potassium and death or KRT start was examined using Cox proportional hazards and restricted cubic spline analyses, adjusted for age, sex, diabetes, cardiovascular disease, renin-angiotensin-aldosterone system (RAAS) inhibition, eGFR, and subjective global assessment (SGA). At baseline, 66% of participants were men, 42% had diabetes, 47% cardiovascular disease, and 54% used RAAS inhibitors. Their mean age was 76 ± 7 (SD) years, mean eGFR was 17 ± 5 (SD) mL/min/1.73 m2, and mean SGA was 6.0 ± 1.0 (SD). Over 8 years, 414 (24%) died before starting KRT, and 595 (35%) started KRT. Adjusted hazard ratios for death or KRT according to the potassium categories were 1.6 (95% CI, 1.1-2.3), 1.4 (95% CI, 1.1-1.7), 1.1 (95% CI, 1.0-1.4), 1 (reference), 1.1 (95% CI, 0.9-1.4), 1.8 (95% CI, 1.4-2.3), and 2.2 (95% CI, 1.5-3.3). Hazard ratios were lowest at a potassium of about 4.9 mmol/L. Shorter intervals between potassium measurements would have allowed for more precise estimations. We observed a U-shaped relationship between serum potassium and death or KRT start among patients with incident CKD 4-5, with a nadir risk at a potassium level of 4.9 mmol/L. These findings underscore the potential importance of preventing both high and low potassium in patients with CKD 4-5.