SUN-183 DIURNAL RHYTHMS OF SERUM PHOSPHATE IN HEMODIALYSIS PATIENTS

Abstract

Hyperphosphatemia is a common complication of chronic kidney disease that associates with increased risks for all-cause and cardiovascular mortality. Phosphate binder therapies are an efficacious strategy to reduce serum phosphate levels. In clinical practice, the decision on binder therapy initiation and dosage is typically determined from the pre-dialysis blood phosphate level. The healthy population exhibits diurnal rhythms in blood phosphate regulation with fluctuations of more than 1 mg/dL throughout the day (Vervloet et al. 2017). Whether such circadian rhythms persist in dialysis patients and whether the fluctuation pattern is different to that of healthy individuals is only partially understood. We profiled the diurnal rhythms of serum phosphate levels in a population of chronic hemodialysis (HD) patients. We used data from adult in-center HD patients (age ≥18 years) with HD vintage >90 days who were treated at a large dialysis organization in the United States in 2018. We computed the mean pre-dialysis blood phosphate levels at the time of the start of each HD treatment in 2-hour periods through 24 hours for the entire population, as well as stratified by phosphate binder use. Time of HD initiation was treated as time-varying exposure, allowing patients to switch time groups if they had pre-dialysis blood draws in multiple time periods of the day. The mean phosphate levels based on the time of the day with 95% confidence limits were plotted and we used a nonparametric smoothing spline model to fit data and calculate the trajectories of phosphate levels throughout the day. A total of 158,156 adult HD patients were included in the analysis. Patients were: mean age 62.8±14.3 years, 57.1% male, 58.2% white race, 15.8% Hispanic ethnicity, 41.9% with diabetes, 19.6% with congestive heart failure, and HD vintage 1509±1288 days. Mean pre-dialysis albumin was 3.8±0.4 g/dL and phosphate was 5.4±1.5 mg/dL for the population. Overall, mean phosphate levels were found to vary remarkably by 0.9 mg/dL throughout the day, with nadir levels from 0900 to 1100 hours (5.3±1.6 mg/dL), followed by a rise to a mid-day peak from 1700 to 1900 hours (5.8±1.8 mg/dL), and the highest levels had a peak time from 2300 to 0100 hours (6.2±2.1 mg/dL) (Figure 1a). Similar fluctuations and phosphate levels were observed among patients using a phosphate binder (n=143,364), however, patients not using a phosphate binder (n=14,792) exhibited absolute values of mean phosphate levels that were consistently about 1 mg/dL lower at every timepoint (Figure 1b). All patients treated with HD from 2300 through 0300 hours (n=315) were taking a phosphate binder. Our results suggest phosphate levels have a diurnal variation of about 1 mg/dL in patients on chronic hemodialysis, characterized by a nadir in the morning, a mid-day peak and a larger nocturnal peak. Further adjusted analyses are needed to confirm these results that account for patient characteristics, phosphate binder dose and duration of use.