Dyskalemia in people at increased risk for heart failure: findings from the heart ‘OMics’ in AGEing (HOMAGE) trial

Aims: In people at risk of heart failure (HF) enrolled in the Heart ‘OMics’ in AGEing (HOMAGE) trial, spironolactone reduced circulating markers of collagen synthesis, natriuretic peptides, and blood pressure and improved cardiac structure and function. In the present report, we explored factors associated with dyskalaemia. Methods and results: The HOMAGE trial was an open-label study comparing spironolactone (up to 50 mg/day) versus standard care in people at risk for HF. After randomization, serum potassium was assessed at 1 and 9 months and was defined as low when ≤3.5 mmol/L (hypokalaemia) and high when ≥5.5 mmol/L (hyperkalaemia). Multivariable logistic regression models were constructed to identify clinical predictors of dyskalaemia. A total of 513 participants (median age 74 years, 75% men, median estimated glomerular filtration rate 71 mL/min/1.73 m2) had serum potassium available and were included in this analysis. At randomization, 88 had potassium < 4.0 mmol/L, 367 had potassium 4.0–5.0 mmol/L, and 58 had potassium > 5.0 mmol/L. During follow-up, on at least one occasion, a serum potassium < 3.5 mmol/L was observed in 6 (1.2%) and <4.0 mmol/L in 46 (9%) participants, while a potassium > 5.0 mmol/L was observed in 38 (8%) and >5.5 mmol/L in 5 (1.0%) participants. The median (percentile25−75) increase in serum potassium with spironolactone during the study was 0.23 (0.16; 0.29) mmol/L. Because of the low incidence of dyskalaemia, for regression analysis, hypokalaemia and hyperkalaemia thresholds were set at <4.0 and >5.0 mmol/L, respectively. The occurrence of a serum potassium > 5.0 mmol/L during follow-up was positively associated with the presence of diabetes mellitus {odds ratio [OR]: 1.21 [95% confidence interval (CI) 2.14; 3.79]} and randomization to spironolactone (OR: 2.83 [95% CI 1.49; 5.37]). Conversely, the occurrence of a potassium concentration < 4.0 mmol/L was positively associated with the use of thiazides (OR: 2.39 [95% CI 1.32; 4.34]), blood urea concentration (OR: 2.15 [95% CI 1.34; 3.39] per 10 mg/dL), and history of hypertension (OR: 2.32 [95% CI 1.02; 5.29]) and negatively associated with randomization to spironolactone (OR: 0.30 [95% CI 0.18; 0.52]). Conclusions: In people at risk for developing HF and with relatively normal renal function, spironolactone reduced the risk of hypokalaemia and, at the doses used, was not associated with the occurrence of clinically meaningful hyperkalaemia