Analysis of Health-Related Quality of Life in Elderly Patients with Stroke Complicated by Hypertension in China Using the EQ-5D-3L Scale

Figures & data

Figure 1 Flowchart for inclusion of research subjects and analytical methods. MSE, Mean Square Error; R², R-squared.

Table 1 The EQ-5D-3L Scale Based on the Health Utility Value Point System of Chinese Residents

Variable	Define	Coefficients
C	At least one dimension is at level 2, 3	0.067
N3	At least one dimension is at level 3	0.016
Mobility
MO1	The dimension of action ability is at level 1	0.000
MO2	The dimension of action ability is at level 2	0.101
MO3	The dimension of action ability is at level 3	0.275
Self-Care
SC1	Self-Care dimension at level 1	0.000
SC2	Self-Care dimension at level 2	0.103
SC3	Self-Care dimension at level 3	0.239
Usual Activities
UA1	Daily activity dimension at level 1	0.000
UA2	Daily activity dimension at level 2	0.086
UA3	Daily activity dimension at level 3	0.217
Pain/Discomfort
PD1	Pain/Discomfort dimension at level 1	0.000
PD2	Pain/Discomfort dimension at level 2	0.110
PD3	Pain/Discomfort dimension at level 3	0.232
Anxiety/Depression
AD1	Anxiety/Depression dimension at level 1	0.000
AD2	Anxiety/Depression dimension at level 2	0.074
AD3	Anxiety/Depression dimension at level 3	0.172

Table 2 Basic Information, Health Utility Values and Univariate Analysis in Hypertensive Stroke Patients(n=301)

Features	N (%)	Health Utility value [M (P25, P75)]	P value
Gender			0.909
Male	169(56.15)	0.458(0.186, 0.745)
Female	132(43.85)	0.419(0.186, 0.745)
State of residence			0.650
Live alone	35(11.63)	0.360(0.186, 0.646)
Living with family	266(88.37)	0.451(0.186, 0.745)
Annual income			<0.001
Less than 50, 000	22(7.31)	0.747(0.513, 0.859)
50, 000–100, 000	218(72.43)	0.463(0.220, 0.745)
Larger than 100, 000	61(20.27)	0.186(0.186, 0.360)
Medical insurance			0.219
Yes	273(90.70)	0.427(0.186, 0.745)
None	28(9.30)	0.480(0.186, 0.859)
Blood pressure control			0.836
Control to normal	235(78.07)	0.455(0.186, 0.748)
Uncontrolled	66(21.93)	0.396(0.258, 0.745)
Physical activity level			<0.001
Lower	216(71.76)	0.360(0.186, 0.651)
Medium	52(17.28)	0.641(0.423, 0.859)
Highly	33(10.96)	0.611(0.412, 0.745)
Social events			<0.001
Yes	149(49.50)	0.510(0.238, 0.758)
None	152(50.50)	0.360(0.112, 0.662)
Length of sleep at night			<0.001
Not enough	131(43.52)	0.455(0.203, 0.745)
Normal	78(25.91)	0.643(0.412, 0.859)
Excessive length	92(30.56)	0.233(0.186, 0.462)
Smoking			0.003
Never smoked	38(12.62)	0.326(0.186, 0.482)
Still smoking	147(48.84)	0.412(0.186, 0.745)
Have quit smoking	116(38.54)	0.517(0.186, 0.839)
Naps			0.824
Yes	188(62.46)	0.458(0.186, 0.745)
None	113(37.54)	0.360(0.186, 0.749)
Age			<0.001
<55	39(12.96)	0.642(0.386, 0.754)
55~64	79(26.25)	0.604(0.281, 0.766)
64~74	125(41.53)	0.422(0.186, 0.648)
≥74	58(19.27)	0.205(0.129, 0.461)
Marital status			0.156
Unmarried	64(21.26)	0.360(0.186, 0.617)
Married	237(78.74)	0.455(0.186, 0.746)
Level of education			0.004
No education	70(23.26)	0.360(0.175, 0.548)
Primary and below	137(45.51)	0.422(0.186, 0.672)
Middle school	86(28.57)	0.615(0.194, 0.758)
College and above	8(2.66)	0.690(0.502, 0.859)
Urban and Rural residence			0.011
Rural	156(51.83)	0.360(0.186, 0.637)
Urban	145(48.17)	0.510(0.186, 0.758)
Area of residence			0.449
Northern China	48(15.95)	0.396(0.186, 0.619)
North-eastern	27(8.97)	0.427(0.156, 0.745)
Eastern China	81(26.91)	0.458(0.186, 0.758)
Central China	57(18.94)	0.510(0.223, 0.745)
South China	26(8.64)	0.438(0.186, 0.834)
Southwestern	43(14.29)	0.424(0.203, 0.745)
Northwestern	19(6.31)	0.326(0.083, 0.535)
Drinking			0.001
Non-alcoholic	227(75.42)	0.360(0.186, 0.671)
Drinking	74(24.58)	0.626(0.332, 0.758)
BMI			<0.001
Normal	81(26.91)	0.604(0.326, 0.773)
Overweight	109(36.21)	0.513(0.274, 0.758)
Obese	89(29.57)	0.223(0.186, 0.465)
Underweight	22(7.31)	0.186(0.179, 0.447)
Fat accumulation in the abdomen			0.040
Yes	222(73.75)	0.412(0.186, 0.745)
None	79(26.25)	0.482(0.273, 0.758)

Abbreviations: [M (P₂₅, P₇₅)], median and interquartile spacing; BMI, Body Mass index.

Table 3 Tobit Regression Modelling of Factors Influencing HRQoL in Hypertensive Stroke Patients(n=301)

Independent Variable	β	SE	Z value	P value	95% CI
Annual income (Reference value = less than 50,000)
50,000–100,000	−0.162	0.065	−2.476	0.013	(−0.289, −0.034)
Larger than 100,000	−0.179	0.083	−2.163	0.031	(−0.342, −0.017)
Physical activity level (reference value = low)
Medium	0.118	0.045	2.589	0.010	(0.029, 0.207)
Highly	0.161	0.056	2.865	0.004	(0.051, 0.272)
Fat accumulation in the abdomen (reference value = yes)
None	0.036	0.048	0.756	0.450	(−0.058, 0.130)
Social events (reference value = yes)
None	−0.141	0.035	−4.078	<0.001	(−0.209, −0.073)
Length of sleep at night (reference value = normal)
Not enough	−0.114	0.041	−2.781	0.005	(−0.194, - 0.034)
Excessive length	−0.046	0.058	−0.797	0.426	(−0.159, 0.067)
Smoking (reference value = still smoking)
Never smoked	−0.058	0.052	−1.105	0.269	(−0.161, 0.045)
Have quit smoking	0.140	0.036	3.924	<0.001	(0.070, 0.210)
Drinking (reference value = no alcohol)
Drinking	0.057	0.039	1.441	0.150	(−0.020, 0.134)
Age (reference <55 years)
55~64	−0.054	0.056	−0.960	0.337	(−0.165, 0.056)
65~74	−0.104	0.052	−1.977	0.048	(−0.206, −0.001)
>74	−0.093	0.063	−1.479	0.139	(−0.216, −0.030)
Level of education (reference value = not educated)
Primary school and below	−0.053	0.044	−1.205	0.228	(−0.140, 0.033)
Secondary Schools	−0.021	0.051	−0.406	0.685	(−0.121, 0.079)
Tertiary and above	0.092	0.111	0.832	0.406	(−0.125, 0.309)
Urban and Rural residence (reference value = Rural)
Urban	0.041	0.034	1.187	0.235	(−0.027, 0.108)
BMI (reference value = Normal)
Overweight	−0.025	0.050	−0.495	0.621	(−0.122, 0.073)
Obesity	−0.209	0.054	−3.892	<0.001	(−0.314, −0.104)
Underweight	−0.189	0.074	−2.572	0.010	(−0.334, −0.045)

Table 4 Comparison of Model Performance

	Training set		Test Set
	MSE	R2	MSE	R2
Tobit regression	0.0753	0.3004	0.0691	0.2926
Random Forest1	0.0402	0.6262	0.0710	0.2726
Random Forest2	0.0376	0.6502	0.0759	0.2224
XGBoost1	0.0519	0.5182	0.0653	0.3313
XGBoost2	0.0459	0.5731	0.0663	0.3210

Notes: The number of people in the training set is 218, The number of people in the test set is 83. Random Forest1 and XGBoost1: Include in the model those characteristics that were statistically significant in the one-way analysis of variance. Random Forest2and XGBoost2: Include all characteristics in the model.

Figure 2 Importance ranking of Random Forest outcome features(n=218).

Figure 3 Chart representing importance ranking of features obtained from result of XGBoost classifier(n=218).

Figure 4 Contribution of each characteristic to the predicted value for a single sample.

Figure 5 Ranking of feature importance(n=218).

Figure 6 BMI Contribution Map(n=218).