The body mass index change is associated with death or hemodialysis transfer in Japanese patients initiating peritoneal dialysis

Figures & data

Figure 1. Flow chart of the procedure of study participant recruitment.

Table 1. Baseline characteristics of the study population and the groups divided according to body mass index change (T1: BMI change: < −4.13%, T2: BMI change: −4.13–0.67%, and T3: BMI change: >0.67%).

Variables	Total (n = 122)	T1 (n = 41)	T2 (n = 40)	T3 (n = 41)	p-Value
Age (year)	61.1 ± 12.1	60.8 ± 14.0	59.8 ± 10.7	62.8 ± 11.5	0.53
Sex (% male)	90 (74%)	31 (76%)	30 (75%)	29 (71%)	0.9
APD/CAPD (at the commencement of PD)	0/122 (0/100%)	0/41 (0/100%)	0/40 (0/100%)	0/41 (0/100%)	1
SPIED/SMAP	111/11 (91/9%)	39/2 (95/5%)	34/6 (85/15%)	38/3 (93/7%)	0.27
Double/triple cuffs (Swan neck silicon PD catheter)	101/21 (83/17%)	31/10 (76/24%)	36/4 (90/10%)	34/7 (83/17%)	0.23
Diabetes mellitus	41 (34%)	14 (34%)	13 (33%)	14 (34%)	1
Cerebrocardiovascular disease	32 (26%)	14 (34%)	11 (28%)	7 (17%)	0.22
Antihypertensive drugs
ACEi/ARB	84 (69%)	33 (80%)	27 (68%)	24 (59%)	0.09
CCB	107 (88%)	34 (83%)	35 (88%)	38 (93%)	0.4
MRB	3 (2%)	1 (2%)	1 (3%)	1 (2%)	1
β-blocker	28 (23%)	10 (24%)	11 (28%)	7 (17%)	0.55
α-blocker	12 (10%)	6 (15%)	4 (10%)	2 (5%)	0.33
Loop diuretics	55 (45%)	20 (49%)	21 (53%)	14 (34%)	0.22
Thiazide diuretics	6 (5%)	2 (5%)	3 (8%)	1 (2%)	0.53
Charlson comorbidity index	3.0 (2.0–4.0)	3.0 (2.0–5.0)	3.0 (2.0–4.0)	3.0 (2.0–4.0)	0.72
Systolic blood pressure (mmHg)	140.1 ± 21.8	146.7 ± 24.9	138.4 ± 16.9	135.0 ± 21.7*	0.04
Diastolic blood pressure (mmHg)	81.8 ± 16.9	86.4 ± 22.1	81.9 ± 10.2	77.0 ± 15.0*	0.04
Mean blood pressure (mmHg)	101.2 ± 17.2	106.5 ± 21.8	100.7 ± 10.9	96.3 ± 15.7*	0.02
Body weight (kg)	65.5 ± 15.3	71.3 ± 16.0	62.3 ± 13.1*	62.8 ± 15.2*	0.01
BMI (kg/m^2)	23.8 (21.3–26.8)	25.9 (23.7–29.2)	23.0 (21.0–24.9)**	22.7 (20.9–26.4)**	<0.001
IVCD (cm)	1.7 ± 0.4	1.7 ± 0.4	1.6 ± 0.4	1.7 ± 0.3	0.58
Albumin (mg/dL)	3.5 (3.1–3.8)	3.4 (3.1–3.8)	3.6 (3.4–3.9)	3.6 (3.1–3.8)	0.23
Urea (mg/dL)	78.1 (62.8–89.4)	79.2 (60.9–94.5)	73.4 (60.1–85.9)	80.8 (63.2–89.0)	0.72
Creatinine (mg/dL)	8.6 (7.6–9.9)	8.8 (8.0–10.0)	8.6 (7.7–9.9)	8.3 (7.5–9.4)	0.35
eGFR (mL/min/1.73m^2)	5.4 ± 1.4	5.3 ± 1.7	5.4 ± 1.1	5.4 ± 1.3	0.88
Hemoglobin (g/dL)	9.6 ± 1.3	9.8 ± 1.3	9.6 ± 1.2	9.4 ± 1.5	0.48
Calcium(mg/dL)	8.3 ± 1.1	8.3 ± 1.2	8.4 ± 0.9	8.3 ± 1.0	0.77
Phosphorus (mg/dL)	5.9 (5.1–6.8)	6.0 (5.3–6.9)	5.8 (5.1–6.7)	5.4 (4.8–6.6)	0.16
PTH (pmol/L)	352 (222–500)	298 (201–496)	344 (201–496)	402 (269–521)	0.43
CRP (mg/L)	0.08 (0.03–0.18)	0.10 (0.03–0.17)	0.04 (0.02–0.10)	0.10 (0.05–0.33)^#	0.01
BNP (pg/mL)	120.5 (50.9–289.5)	166.1 (92.4–413.4)	74.2 (42.5–166.9)*	132.8 (43.4–236.0)	0.02
GNRI	97.3 ± 11.1	99.8 ± 10.9	96.9 ± 10.2	95.3 ± 11.8	0.18
D/P4 (n = 118)	0.59 ± 0.13	0.57 ± 0.14	0.57 ± 0.12	0.63 ± 0.14	0.07
6-month percentage change in BMI	−2.14 (−5.57–1.84)	−8.80 (−11.90 to −5.57)	−2.14 (−3.12 to −0.42)***	3.52 (1.93–5.40)^***###	<0.001

*p < 0.05; **p < 0.01; ***p < 0.001 versus T1. ^#p < 0.05; ^##p < 0.01; ^###p < 0.001 versus T2. p-Values were estimated by Fisher’s exact test for categorical variables and the one-way analysis of variance and the Kruskal–Wallis test for normally and non-normally distributed continuous variables, respectively.

Abbreviations: SD: standard deviation; SPIED: short-term PD induction and education; SMAP: stepwise initiation of PD using the Moncrief–Popovich technique; ACEi/ARB: angiotensin converting enzyme inhibitor/angiotensin II receptor blocker; CCB: calcium channel blocker; MRB: mineralocorticoid receptor blocker; BMI: body mass index; IVCD: inferior vena cava diameter; eGFR: estimated glomerular filtration rate; PTH: parathyroid hormone; CRP: C-reactive protein; BNP: brain natriuretic peptide; GNRI: geriatric nutritional risk index; D/P4: dialysate-to-plasma ratio of creatinine at 4 h.

Table 2. Causes of HD transfer among tertile categories.

Cause of HD transfer	Total (n = 122)	T1 (n = 41)	T2 (n = 40)	T3 (n = 41)	p-Value
Peritonitis	29 (23.8%)	11 (26.8%)	7 (17.5%)	11 (26.8%)	0.16
HF due to volume overload	8 (6.6%)	6 (14.6%)	0 (0%)*	2 (4.9%)	0.04
Uremic solute retention	4 (3.3%)	0 (0%)	3 (7.5%)	1 (2.4%)	0.12
Frailty	6 (4.9%)	3 (7.3%)	2 (5.0%)	1 (2.4%)	0.87
Major abdominal surgery	4 (3.3%)	1 (2.4%)	3 (7.5%)	0 (0%)	0.22
Tunnel infection	2 (1.6%)	1 (2.4%)	1 (2.5%)	0 (0%)	1
Catheter malfunction	1 (0.8%)	0 (0%)	1 (2.5%)	0 (0%)	0.60
Cerebral hemorrhage	1 (0.8%)	0 (0%)	1 (2.5%)	0 (0%)	0.60

*p < 0.05 versus T1. p-Values were estimated by Fisher’s exact test.

Abbreviations: HD: hemodialysis; HF: heart failure.

Figure 2. Kaplan–Meier curves (A) and Cumulative incidence curves (B) for death or hemodialysis transfer in the T1, T2, and T3 groups.

Table 3. Association between BMI change and death or HD transfer in patients on PD using standard Cox regression model and Fine and Gray sub-distribution hazards model.

HD transfer or all-cause mortality	T1 vs. T2 (ref)			T3 vs. T2 (ref)			T1 vs. T3 (ref)
	HR (95% CI)	p-Value		HR (95% CI)	p-Value		HR (95% CI)	p-Value
Cause-specific hazards
Model 1	2.48 (1.41–4.37)	0.002		1.18 (0.66–2.11)	0.57		2.10 (1.18–3.74)	0.01
Model 2	2.53 (1.42–4.50)	0.002		1.18 (0.66–2.12)	0.58		2.14 (1.18–3.87)	0.01
Model 3	2.16 (1.19–3.94)	0.01		0.86 (0.46–1.62)	0.64		2.52 (1.36–4.66)	<0.001
Subdistribution hazards
Model 1	1.99 (1.20–3.30)	0.008		1.10 (0.65–1.84)	0.73		1.82 (1.01–3.26)	<0.05
Model 2	2.02 (1.20–3.40)	0.008		1.09 (0.65–1.83)	0.74		1.85 (1.01–3.38)	<0.05
Model 3	1.70 (1.02–2.83)	0.04		0.78 (0.43–1.40)	0.41		2.18 (1.16–4.08)	0.02

Model 1 was a minimally adjusted model with age, sex, Charlson comorbidity index, estimated glomerular filtration rate, geriatric nutritional risk index, and the categories of BMI change. Model 2 was adjusted for the same variables as model 1 in addition to the use of the mean blood pressure and angiotensin converting enzyme inhibitor/angiotensin II receptor blocker. Model 3 was adjusted for the same variables as model 2 in addition to the logarithmic brain natriuretic peptide and logarithmic C-reactive protein.

Abbreviations: BMI: body mass index; HD: hemodialysis; PD: peritoneal dialysis; ref: reference; HR: hazard ratio; CI: confidence interval.

Figure 3. Restricted cubic spline curves showing death or transfer to hemodialysis for the unadjusted (A) and multivariable-adjusted model 1 (B), model 2 (C), and model 3 (D). In splines, solid lines represent the hazard ratio and dashed lines represent the 95% confidence interval.

Table 4. Association between the BMI change and peritonitis or heart failure in patients on PD using the standard Cox regression model.

	Peritonitis									Heart failure-associated hospitalization
	T1 vs. T2 (ref)			T3 vs. T2 (ref)			T1 vs. T3 (ref)			T1 vs. T2 (ref)			T3 vs. T2 (ref)			T1 vs. T3 (ref)
	HR (95% CI)	P-value		HR (95% CI)	p-Value		HR (95% CI)	p-Value		HR (95% CI)	p-Value		HR (95% CI)	p-Value		HR (95% CI)	p-Value
Model 1	1.89 (0.98–3.62)	0.06		1.51 (0.81–2.82)	0.20		1.25 (0.67–2.35)	0.49		1.89 (0.95–3.77)	0.07		1.29 (0.66–2.49)	0.46		1.46 (0.75–2.85)	0.27
Model 2	1.85 (0.95–3.61)	0.07		1.51 (0.81–2.83)	0.20		1.22 (0.63–2.35)	0.55		1.76 (0.89–3.48)	0.10		1.43 (0.73–2.80)	0.30		1.24 (0.61–2.52)	0.56
Model 3	1.79 (0.89–3.60)	0.10		1.48 (0.77–2.85)	0.24		1.21 (0.62–2.37)	0.58		1.72 (0.85–3.51)	0.13		1.50 (0.73–3.06)	0.27		1.15 (0.56–2.38)	0.7

Model 1 was a minimally adjusted model with age, sex, Charlson comorbidity index, estimated glomerular filtration rate, geriatric nutritional risk index, and the categories of BMI change. Model 2 was adjusted for the same variables as model 1 in addition to the use of the mean blood pressure and angiotensin converting enzyme inhibitor/angiotensin II receptor blocker. Model 3 was adjusted for the same variables as model 2 in addition to the logarithmic brain natriuretic peptide and logarithmic C-reactive protein.

Abbreviations: BMI: body mass index; PD: peritoneal dialysis; ref: reference; HR: hazard ratio; CI: confidence interval.