Invited commentary by Dr. Timothy Yau.
As nephrologists who regularly manage the complications of chronic kidney disease (CKD), our visceral reaction to high potassium intake is often negative. Although we also treat interesting hypokalemic conditions, more common encounters occur with elevated serum potassium levels in patients with diminished glomerular filtration rate (GFR), and potassium is considered the enemy.
However, higher dietary potassium intake in the general population is associated with lower blood pressure and a reduction in cardiovascular events and subsequent mortality. This has been demonstrated in several studies, including a 2014 report by Kieneker et al., and although the mechanisms for this influence remain debated, there is an hypothesis: higher potassium intake lowers blood pressure by inducing natriuresis and decreasing vascular resistance. There may also be a role for high potassium intake in reducing left ventricular modeling through mechanisms not depedent on blood pressure. However, experience tells us that interventions that are significant in the general population do not always apply to those with kidney disease. This is a harsh lesson that nephrologists have learned through countless trials, including the notable AURORA lipid-lowering trial.
Two trials have previously evaluated the relationship between potassium intake and kidney outcomes, unfortunately with conflicting results. The first based its conclusion on a single spot urine potassium measurement and included patients who were predominantly receiving renin-angiotensin system inhibition therapy (angiotensin II receptor blockers, angiotensin converting enzyme [ACE] inhibitors, or both) or placebo. It demonstrated a 30% lower risk of GFR loss in individuals with a high (>2.7 g/d) and moderate (1.7 to 2.7 g/d) urine potassium excretion, as compared with low (<1.7 g/d) excretion. The second was based off CRIC (the Chronic Renal Insufficiency Cohort) and came to the opposite conclusion: urine potassium excretion was associated with more rapid GFR loss.
A recent study published in AJKD was based on the MDRD (Modified Diet in Renal Disease) Study cohort, and divided patients into 4 quartiles based on 24-hour urine potassium excretion: low (mean, 1.41 g/day), mid-low (2.01 g/day), mid-high (2.54 g/day), and high (3.6 g/day) groups. An important assumption that Leonberg-Yoo et al recognize is that 24-hour urine potassium excretion is a potentially flawed surrogate for potassium intake, as gastrointestinal secretion of potassium increases with declining GFR (the mean GFR for these patients was 32.6 mL/min). Nonetheless, after robust multivariate analysis, the authors ascertained that higher 24-hour urine potassium excretion led to a reduction in all-cause mortality (17% reduction death rate per 1 standard deviation higher urine potassium excretion). Although there were trends towards significance in increased kidney failure events (dialysis or transplantation) in lower urine potassium excretion, they were not significant in adjusted analyses.
What do we make of these findings? One could argue that higher 24-hour urine potassium excretion implies the capability of the reduced nephron mass to adequately achieve homeostasis and subsequently maintain normal serum potassium levels (serum levels for this cohort were not established in the paper). Muddling the picture in this group is the concomitant use of ACE inhibitors (36.1%) and diuretics (40.2%), which would either enhance or impair urine potassium excretion. Additionally, as with all studies based off the MDRD Study cohort, we must remember that the cause of CKD in this group is not representative of CKD in the US general population. The majority of patients enrolled in the MDRD Study had polycystic kidney disease or glomerulonephritis, and only a small portion (5.2%) were diabetic. If practicing clinicians were to apply the findings from this trial to a generalized CKD population (predominantly those with diabetic nephropathy or hypertensive nephrosclerosis treated with ACE inhibitors) by recommending a higher potassium intake, they potentially are putting their patients at risk for hyperkalemia.
This study suggests that high urine potassium excretion is associated with a lower mortality and is one of the few that shows similar findings when applied to both the general and CKD population. However, more prospective and generalizable data are needed before high potassium diets can be recommended to patients with reduced GFR for this questionable benefit.
Timothy Yau, MD
Assistant Professor of Medicine
Washington University School of Medicine
St. Louis, Missouri