Effect of clinical factors on trajectory of functional performance in patients undergoing hemodialysis

Figures & data

Figure 1. Participants’ flow diagram.

Figure 2. Three-dimensional plot using the Karnofsky Performance Status scores from 2015 to 2017 based on K-means clustering algorithm. High KPS cluster, navy dots; low KPS cluster, maroon dots.

Table 1. Baseline characteristics of hemodialysis patients categorized by KPS cluster (derived from KPS scores between 2015 and 2016) (N = 82).

Variables	High KPS cluster (n = 56)	Low KPS cluster (n = 26)	p
Age (years)	61 ± 11	67 ± 7	0.008
Gender (male, %)	25 (44.64%)	10 (38.46%)	0.598
Dialysis vintage (years)	11.0 ± 6.90	10.0 ± 5.96	0.516
Diabetes	10 (17.86%)	9 (34.62%)	0.158
Antihypertensive drugs	15 (26.78%)	8 (30.77%)	0.709
Primary kidney disease			0.547
Chronic glomerulonephritis	14 (25.00%)	4 (15.38%)
Hypertensive nephropathy	25 (44.64%)	16 (61.54%)
Chronic interstitial nephritis	1 (1.79%)	—
Cause unknown	16 (28.57%)	6 (23.08%)
KPS scale
2015	90.54 ± 10.52	86.92 ± 13.79	0.195
2016	92.50 ± 8.37	80.38 ± 13.11	<0.001
2017	93.39 ± 5.81	65.38 ± 13.92	<0.001
Laboratory data
Hb (g/dL)	10.71 ± 1.31	10.52 ± 0.94	0.51
Albumin (g/dL)	3.99 ± 0.28	3.77 ± 0.31	0.002
BUN (mg/dL)	61.29 ± 0.28	63.73 ± 15.65	0.545
Cr (mg/dL)	10.08 ± 2.45	9.64 ± 1.77	0.413
Ca (mg/dL)	9.34 ± 1.05	9.60 ± 0.86	0.272
P (mg/dL)	6.66 ± 11.83	4.82 ± 1.43	0.433
K (meq/L)	4.58 ± 0.74	6.17 ± 9.98	0.236
Kt/V	1.83 ± 0.40	1.72 ± 0.23	0.214
iPTH^a (pg/ml)	205.05 (68.80–433.35)	293.7 (108.80–664.90)	0.259
PCS(ug/ml)	27.09 ± 20.30	30.59 ± 27.05	0.516
IS (ug/ml)	43.53 ± 23.50	40.09 ± 17.05	0.507
Beta-2-microglobulin^a (ug/L)	25918.80 (20300.00–30199.60)	28240.70 (20241.80–35610.00)	0.213

BUN: blood urea nitrogen; Cr: creatinine; Ca: calcium; Hb: hemoglobin; iPTH: intact parathyroid hormone; IS: indoxyl sulfate; K: potassium; KPS: Karnofsky performance status; P: phosphate; PCS: p-cresyl sulfate.

^aMedian, interquartile range.

p-value was estimated using two-sample t-test, chi-squared test or Fisher's exact test appropriately.

Table 2. Logistic regression for KPS cluster (derived from KPS scores between 2015 and 2017) (N = 82)

Variables	Univariate			Full-adjusted model			Stepwise selection (p < 0.2) model
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
Dependent variables (reference: high KPS cluster)
Age	1.07	1.02–1.13	0.012	1.15	1.04–1.28	0.008	1.12	1.03–1.23	0.011
Gender (male, %)	0.78	0.30–2.00	0.599	0.15	0.02–0.98	0.048	0.20	0.04–0.96	0.045
Dialysis vintage	0.98	0.91–1.05	0.511	0.92	0.80–1.06	0.265	0.89	0.79–1.01	0.066
Diabetes	2.44	0.84–7.02	0.099	1.91	0.34–10.83	0.464	—
Hb	0.87	0.58–1.30	0.505	0.77	0.35–1.70	0.513	—
Albumin	0.06	0.01–0.44	0.005	0.07	<0.01–1.45	0.086	0.02	<0.01–0.40	0.009
BUN	1.01	0.98–1.04	0.540	1.09	1.01–1.16	0.022	1.08	1.02–1.16	0.015
Cr	0.91	0.72–1.14	0.410	0.79	0.50–1.26	0.323	0.69	0.45–1.07	0.096
Ca	1.33	0.80–2.23	0.270	2.01	0.65–6.19	0.226	3.24	1.20–8.73	0.020
P	0.86	0.60–1.24	0.420	0.51	0.23–1.10	0.084	0.54	0.28–1.04	0.065
K	1.05	0.94–1.18	0.358	1.08	0.94–1.24	0.297	1.11	0.97–1.27	0.148
Kt/V	0.41	0.10–1.67	0.213	0.22	0.01–3.95	0.302	—
iPTH	>1.00	<1.00–>1.00	0.260	>1.00	<1.00–>1.00	0.339	—
PCS	1.01	0.99–1.03	0.513	1.01	0.98–1.05	0.479	—
IS	>0.99	0.97–1.01	0.502	0.97	0.92–1.01	0.160	—
Beta-2-microglobulin	>1.00	<1.00–>1.00	0.212	>1.00	>1.00–<1.01	0.047	>1.00	>1.00–<1.01	0.031

Data availability statement

The raw data used to support the findings of this study are available from the corresponding author upon reasonable request.