Nomogram model based on clinical factors and autonomic nervous system activity for predicting residual renal function decline in patients undergoing peritoneal dialysis

Abstract

Background: Several heart rate variability (HRV) parameters were reported to be associated with residual renal function (RRF) in patients undergoing continuous ambulatory peritoneal dialysis (CAPD). However, it is unclear whether using HRV or other autonomic nervous system (ANS) activity indexes can predict RRF decline in CAPD patients.

Methods: Patients undergoing CAPD in 2022 from the First Affiliated Hospital of Nanjing Medical University were enrolled in this study. Their clinical characteristics, 5-min HRV parameters and average voltage of 5-min skin sympathetic nerve activity (aSKNA) were collected. According to the 12-month glomerular filtration rate (GFR) decline rate compared with the upper quartile, these patients were categorized into two groups: RRF decline (RRF-D) group and RRF stable (RRF-S) group. Clinical factors and ANS activity indexes for predicting 1-year RRF decline were analyzed using logistic regression, and a nomogram model was further established. The relationships between volume load related indexes and aSKNA were displayed by Spearman's correlation graphs.

Results: Ninety-eight patients (53 women, average age of 46.7 ± 13.0 years old) with a median dialysis vintage of 24.5 months were enrolled in this study. Seventy-three patients were categorized into the RRF-S group and 25 patients into the RRF-D group. Compared with RRF-S group, patients in the RRF-D group had higher systolic blood pressure (BP; p = 0.019), higher GFR (p = 0.016), higher serum phosphorous level (p = 0.030), lower total Kt/V (p = 0.001), and lower levels of hemoglobin (p = 0.007) and albumin (p = 0.010). The RRF-D group generally exhibited lower HRV parameters and aSKNA compared with the RRF-S group. A nomogram model included clinical factors (sex, systolic BP, hemoglobin, GFR, and total Kt/V) and aSKNA showed the largest AUC of 0.940 (95% CI: 0.890-0.990) for predicting 1-year RRF decline.

Conclusion: The nomogram model included clinical factors (sex, systolic BP, hemoglobin, GFR and total Kt/V) and ANS activity index (aSKNA) might be a promising tool for predicting 1-year RRF decline in CAPD patients.

Keywords: chronic kidney disease; heart rate variability; nomogram model; peritoneal dialysis; skin sympathetic nerve activity.