Miriam E Tucker
September 22, 2017
Elevated potassium is common and associated with poor outcomes in a real-world population of patients with diabetes, new research finds.
Results from the Danish population-based cohort study were presented September 15 here at the European Association for the Study of Diabetes (EASD) 2017 Annual Meeting by Reimar W Thomsen, MD, PhD, of Aarhus University, Copenhagen, Denmark.
"Hyperkalemia is quite frequent in diabetes patients. There are clearly identifiable risk factors for clinicians to look for. We also observed that hyperkalemia is associated with quite severe clinical outcomes, including cardiac outcomes, acute hospitalizations, and [intensive care unit] admissions. So, it may be a greater problem in diabetes than previously thought," Dr Thomsen told Medscape Medical News.
Asked to comment on the findings, session comoderator Sten Madsbad, professor of medicine at the University of Copenhagen and chief physician in the department of endocrinology, Hvidovre University Hospital, Denmark, said, "It's an interesting paper ... We have to realize that patients with high potassium are high-risk patients."
Because the use of potassium-sparing diuretics was identified in the study as one of several risk factors, Dr Madsbad also advised, "Maybe we should treat the hyperkalemia with potassium inhibition or change the drug we're using ... or at least, stop the spironolactone."
One in Six With Diabetes Have Hyperkalemia; CKD and HF Up Risk
Dr Thomsen and colleagues note that data are scarce on hyperkalemia in patients with diabetes.
They used linked individual data from hospital, prescription, and laboratory databases covering the population of Northern Denmark, which included complete information for about 1.8 million inhabitants, or approximately 30% of the country's population during 2000–2012. The study subjects were 68,601 individuals with a first-time prescription for a glucose-lowering agent.
Overall, about 16% developed hyperkalemia, defined as elevated blood potassium > 5.0 mmol/L, based on a first blood test in primary care or the hospital, over a mean 3.4-year follow-up.
Chronic kidney disease (CKD) was present in 26% at the time diabetes medication was started, and the risk for developing hyperkalemia within the first year increased with worsening kidney function, from 3% among the 74% with no CKD to 7% among the 16% with stage 3A kidney disease (eGFR 45–59), to 13% of the 7% with stage 3B disease, 22% for the 3% with stage 4, and 22% to 23% of the 1.2% with stage 5 CKD (eGFR < 15/on dialysis).
Hyperkalemia recurrence was also common, with 43.6% experiencing a second event within 0.65 years. Of those, 56.6% had a third event, and of those, 64.6% had a fourth.
"So, previous hyperkalemia predicts future hyperkalemia, and the time between these events is typically short," Dr Thomsen commented.
In a second study phase, the investigators compared 11,208 patients with a first hyperkalemia episode with 11,127 diabetes patients without hyperkalemia, matched for gender, age, and diabetes duration.
Factors Significantly Predicting Hyperkalemia in Patients With Diabetes
"One out of six patients with diabetes develops hyperkalemia; those with concomitant kidney disease and heart failure are at particularly increased risk," the researchers observe.
Hyperkalemia Portends Poorer Outcomes
Dr Thomsen and colleagues also compared outcomes 6 months prior to the hyperkalemia event to those 6 months after among those with hyperkalemia.
Risk ratios before/after were 1.18 for outpatient contact, 1.67 for acute hospitalization, 1.78 for hospitalization for any cardiac diagnosis, and 6.27 for hospitalization with ICU admission.
They then further compared these with controls with diabetes who didn't have elevated potassium, matched for index date and time since diabetes diagnosis.
By comparison with the controls, the risk for hospitalization was more than double (2.16-fold) in those with hyperkalemia vs those without.
For hospitalization for any cardiac diagnosis, the increased risk was nearly fivefold higher (4.73), for hospitalization for cardiac arrest it was increased 6.59-fold, and for ICU admission, almost fivefold (4.77).
"Hyperkalemia is clearly associated with severe clinical outcomes, including acute hospitalizations, and is a clear predictor of death in patients with diabetes. It's difficult to prove causality, but there is clearly something ... going on after the hyperkalemia event in these patients," Dr Thomsen concluded.
The study was funded by a grant to Aarhus University from AstraZeneca, which is developing a drug to treat hyperkalemia. Dr Thomsen has no further relevant financial relationships, and Dr Madsen has no conflicts of interest.