Association of Hemoglobin to Red Blood Cell Distribution Width-Standard Deviation (RDW-SD) Ratio and 3-Month Readmission in Elderly Chinese Patients with Heart Failure: A Retrospective Cohort Study

Figures & data

Table 1 The Baseline Characteristics of Participants

Hb/RDW-SD	Q1 (0.44 to ≤2.01)	Q2 (2.01 to ≤2.42)	Q3 (2.42 to ≤2.79)	Q4 (2.79 to ≤4.14)	P-value
Participants	448	446	448	448
Age, years, n (%)					<0.001
≥80	220 (49.11%)	207 (46.41%)	166 (37.05%)	139 (31.03%)
70–79	178 (39.73%)	178 (39.91%)	190 (42.41%)	158 (35.27%)
60–69	50 (11.16%)	61 (13.68%)	92 (20.54%)	151 (33.71%)
Gender, n (%)					<0.001
Female	274 (61.16%)	295 (66.14%)	273 (60.94%)	225 (50.22%)
Male	174 (38.84%)	151 (33.86%)	175 (39.06%)	223 (49.78%)
Occupation, n (%)					0.693
Urban Resident	376 (85.26%)	378 (86.30%)	375 (85.03%)	372 (83.41%)
Farmer	45 (10.20%)	40 (9.13%)	38 (8.62%)	45 (10.09%)
Others	20 (4.54%)	20 (4.57%)	28 (6.35%)	29 (6.50%)
BMI, kg/m^2	21.25 ± 4.12	20.75 ± 3.47	21.18 ± 3.66	21.85 ± 4.09	<0.001
SBP, mmHg	132.92 ± 25.07	132.74 ± 22.72	131.75 ± 22.66	131.06 ± 24.88	0.616
DBP, mmHg	74.59 ± 13.32	76.87 ± 14.05	77.58 ± 13.91	77.53 ± 14.67	0.004
Creatinine, µmol/L	122.70 (72.80–183.90)	88.90 (67.75–121.95)	82.15 (63.20–108.20)	78.30 (63.40–103.60)	<0.001
Urea, mmol/L	11.01 (6.88–16.03)	8.25 (5.84–11.34)	7.39 (5.74–10.11)	7.20 (5.52–9.77)	<0.001
GFR, mL/min/1.73 m^2	40.34 (25.14–70.00)	56.83 (41.02–81.18)	68.44 (48.64–92.39)	73.20 (54.11–93.40)	<0.001
RDW (%)	16.79 ± 2.58	14.81 ± 1.31	14.27 ± 1.08	13.83 ± 1.16	<0.001
RDW-SD, fL	54.02 ± 8.03	49.23 ± 4.46	47.28 ± 3.94	45.07 ± 3.79	<0.001
Hb, g/L	84.28 ± 18.19	109.80 ± 10.44	122.94 ± 10.89	137.63 ± 13.62	<0.001
Albumin, g/L	33.99 ± 4.73	36.06 ± 4.31	37.40 ± 4.43	38.41 ± 4.45	<0.001
Cystatin, mg/L	1.59 (1.23–2.26)	1.55 (1.20–2.15)	1.51 (1.21–2.11)	1.54 (1.22–2.28)	0.310
MCV, fL	91.77 ± 8.61	91.95 ± 8.80	92.66 ± 8.37	91.89 ± 9.25	0.430
MCH, pg	29.81 ± 3.38	29.84 ± 3.52	30.21 ± 3.14	29.88 ± 3.57	0.254
MCHC, g/L	324.29 ± 13.57	323.94 ± 15.42	325.82 ± 13.69	324.71 ± 13.31	0.212
Potassium, mmol/L	4.09 ± 0.75	4.00 ± 0.70	3.94 ± 0.66	3.87 ± 0.60	<0.001
Sodium, mmol/L	138.21 ± 4.65	138.42 ± 4.95	138.59 ± 4.97	138.20 ± 5.06	0.594
BNP, pg/mL	840.81 (341.78–1956.94)	853.22 (391.23–1944.24)	780.23 (308.85–1574.08)	620.11 (216.44–1527.81)	0.001
Total bilirubin, µmol/L	15.00 (9.35–25.55)	18.10 (12.50–29.13)	18.50 (12.80–26.87)	19.00 (13.40–27.40)	0.002
MSCKD, n (%)					<0.001
No	291 (64.96%)	332 (74.44%)	353 (78.97%)	377 (84.34%)
Yes	157 (35.04%)	114 (25.56%)	94 (21.03%)	70 (15.66%)
Type of HF, n (%)					0.003
Both	352 (78.57%)	339 (76.01%)	336 (75.00%)	302 (67.41%)
Left	84 (18.75%)	95 (21.30%)	106 (23.66%)	133 (29.69%)
Right	12 (2.68%)	12 (2.69%)	6 (1.34%)	13 (2.90%)
Diabetes, n (%)					0.125
No	328 (73.21%)	351 (78.70%)	336 (75.00%)	353 (78.79%)
Yes	120 (26.79%)	95 (21.30%)	112 (25.00%)	95 (21.21%)
NYHA class, n (%)					0.122
III	223 (49.78%)	230 (51.57%)	230 (51.34%)	239 (53.35%)
IV	161 (35.94%)	132 (29.60%)	141 (31.47%)	123 (27.46%)
II	64 (14.29%)	84 (18.83%)	77 (17.19%)	86 (19.20%)
Readmission within 3-month					0.012
No	321 (71.65%)	324 (72.65%)	330 (73.66%)	360 (80.36%)
Yes	127 (28.35%)	122 (27.35%)	118 (26.34%)	88 (19.64%)
LVEF (%)	53.13 ± 12.51	50.00 ± 12.10	50.31 ± 13.79	50.62 ± 13.28	0.194
Presence of LVEF, n (%)					0.192
No	323 (72.10%)	304 (68.16%)	305 (68.08%)	293 (65.40%)
Yes	125 (27.90%)	142 (31.84%)	143 (31.92%)	155 (34.60%)

Notes: Results in table: Mean±SD; Median (Q1-Q3); n (%).

Abbreviations: Q, quartile; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; GFR, glomerular filtration rate; Hb, hemoglobin; RDW, red blood cell distribution width; RDW-SD, standard deviation of red blood cell distribution width; MCV, mean corpuscular volume; MCH, mean hemoglobin volume; MCHC, mean hemoglobin concentration; BNP, B-type natriuretic peptide; MSCKD, moderate to severe chronic kidney disease with glomerular filtration rate < 60 mL/min; HF, heart failure; NYHA, New York Heart Association; LVEF, left ventricular ejection fraction.

Table 2 The Results of Univariate Analysis

	Statistics	OR (95% CI)	P-value
Occupation
Urban Resident	1520 (84.869%)	Ref
Farmer	173 (9.659%)	0.481 (0.310, 0.746)	0.00109
Others	98 (5.472%)	1.028 (0.647, 1.634)	0.90534
MSCKD
No	1372 (75.634%)	Ref
Yes	442 (24.366%)	1.471 (1.160, 1.864)	0.00144
Age
≥80	741 (40.804%)	Ref
70–79	712 (39.207%)	1.048 (0.830, 1.324)	0.69107
60–69	363 (19.989%)	0.715 (0.527, 0.969)	0.03070
SBP	132.02 ± 23.86	0.988 (0.983, 0.992)	<0.00001
DBP	76.66 ± 14.01	0.988 (0.981, 0.996)	0.00326
Type of HF
Both	1345 (74.064%)	Ref
Left	427 (23.513%)	0.575 (0.437, 0.757)	0.00008
Right	44 (2.423%)	0.673 (0.320, 1.413)	0.29491
Diabetes
No	1391 (76.597%)	Ref
Yes	425 (23.403%)	1.344 (1.055, 1.712)	0.01657
Creatinine	109.99 ± 76.31	1.002 (1.001, 1.003)	0.00494
Urea	9.56 ± 5.26	1.045 (1.025, 1.065)	<0.00001
GFR	65.58 ± 33.53	0.993 (0.989, 0.996)	0.00001
Hb	113.62 ± 23.91	0.995 (0.991, 1.000)	0.04262
RDW	14.92 ± 2.00	1.076 (1.023, 1.132)	0.00455
RDW-SD	48.90 ± 6.28	1.032 (1.016, 1.049)	0.00012
Hb/RDW-SD	2.37 ± 0.58	0.728 (0.608, 0.872)	0.00058
Potassium	3.98 ± 0.68	1.289 (1.108, 1.499)	0.00098
Sodium	138.36 ± 4.90	0.952 (0.932, 0.972)	<0.00001
BNP	1283.96 ± 1361.78	1.000 (1.000, 1.000)	0.00102
Total bilirubin	21.77 ± 14.60	1.010 (1.003, 1.017)	0.00541
NYHA class
II	933 (51.377%)	Ref
IV	562 (30.947%)	1.435 (1.137, 1.811)	0.00239
I	321 (17.676%)	0.679 (0.492, 0.939)	0.01906

Abbreviations: CI, Confidence; Ref, Reference.

Table 3 Relationship Between Hb/RDW-SD and HF 3-Month Readmission in Different Models

Variable	Crude Model (OR, 95% CI, P)	Model I (OR, 95% CI, P)	Model II (OR, 95% CI, P)
Hb/RDW-SD	0.73 (0.61, 0.87) 0.0006	0.74 (0.62, 0.89) 0.0017	0.70 (0.55, 0.89) 0.0031
Hb/RDW-SD (quartile)
Q1	Ref	Ref	Ref
Q2	0.95 (0.71, 1.28) 0.7403	0.96 (0.72, 1.29) 0.8070	1.02 (0.73, 1.43) 0.9088
Q3	0.90 (0.67, 1.21) 0.5000	0.92 (0.69, 1.24) 0.5932	0.88 (0.62, 1.26) 0.4939
Q4	0.62 (0.45, 0.84) 0.0024	0.64 (0.47, 0.88) 0.0065	0.65 (0.44, 0.96) 0.0309
P for trend	0.0033	0.0097	0.0226

Notes: Crude model: we did not adjust other covariates. Model I: we adjusted age, gender. Model II: we adjusted age, gender, occupation, MSCKD, SBP, DBP, type of HF, diabetes, creatinine, urea, GFR, albumin, potassium, sodium, BNP, total bilirubin, NYHA class, presence of LVEF.

Table 4 Threshold Effect Analysis of Hb/RDW-SD on HF Readmission Within 3-Month Using Piece-Wise Linear Regression

	HF Readmission Within 3-Month (OR, 95% CI) P-value
Fitting model by standard linear regression	0.70 (0.55, 0.89)	0.0031
Fitting model by two-piecewise linear regression
Inflection points of Hb/RDW-SD	1.78, 2.17
< 1.78	0.38 (0.17, 0.87)	0.0209
1.78–2.17	1.02 (0.16, 6.34)	0.9815
> 2.17	0.44 (0.27, 0.71)	0.0007
P for log likelihood ratio test	0.002

Notes: We adjusted age, gender, occupation, MSCKD, SBP, DBP, type of HF, diabetes, creatinine, urea, GFR, albumin, potassium, sodium, BNP, total bilirubin, NYHA class, presence of LVEF.

Figure 1 The non-linear relationship between Hb/RDW-SD and HF 3-month readmission. A non-linear relationship between them was detected after adjusting for age, gender, occupation, MSCKD, SBP, DBP, type of HF, diabetes, creatinine, urea, GFR, albumin, potassium, sodium, BNP, total bilirubin, NYHA class, presence of LVEF.

Figure 2 The association between Hb/RDW-SD and the risk of HF readmission within 3-month in various subgroups.

Notes: Above model adjusted for age, gender, occupation, MSCKD, SBP, DBP, type of HF, diabetes, creatinine, urea, GFR, albumin, potassium, sodium, BNP, total bilirubin, NYHA class, presence of LVEF. In each case, the model is not adjusted for the stratification variable.