Abstract

Background: Patients with chronic kidney disease (CKD) are at an increased risk for blood electrolyte imbalances and compromised cardiac and renal regulatory functions. Our objective is to characterize serum potassium trajectories in patients with CKD. Methods: Data from the Chronic Renal Insufficiency Cohort Study were used to analyze 5,625 participants with CKD. Serum potassium measurements were collected annually. Hyperkalemia was defined as >5.0 mmol/L, hypokalemia as <3.5 mmol/L, and normokalemia as 3.5 - 5.0 mmol/L. Trajectory analysis using a semiparametric modeling strategy was conducted, incorporating hierarchical modeling and latent class models. Group-based trajectory models were used to fit longitudinal data and estimate probabilities for multiple trajectories. Models were tested using a censored normal model with various numbers of trajectory groups and functional forms (e.g., linear, quadratic, cubic). Maximum likelihood was used to estimate model parameters and model fit was assessed by comparing Bayesian Information Criterion. Posterior predicted probabilities for each individual were estimated for each trajectory group and participants were assigned the highest posterior predicted probability. Results: Participants were 56% male, 43% Black, and had a median age of 67.0 years. Overall mean serum potassium was 4.3 mmol/L over 12 annual visits with a median follow-up of 9 years. Based on model fit and statistical significance, five distinct trajectory groups were identified (Figure). Four trajectories (one hyperkalemia and three normokalemia) were stable over time. One trajectory started in the normokalemia range and increased into hyperkalemia. Conclusions: Five trajectories of serum potassium were identified among CKD patients. Characterizing serum potassium trajectories over time can help to understand who may be at increased risk for abnormal potassium and related complications and potentially inform the development of interventions to maintain normokalemia in patients with CKD.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call