3.1 Baseline characteristics
Overall, 10200 aged adults (4809 males and 5391 females) were included in the analysis, with an average age of 67.43 ± 9.55 years. The quartiles for WWI in males were as follows: Q1 (≤10.36), Q2 (10.36-10.80), Q3 (>10.80-11.26), and Q4 (>11.26), and the quartiles were as follows: Q1 (≤10.85), Q2 (10.85-11.42), Q3 (11.42-12.00), and Q4 (>12.00) for females. Baseline characteristics, including age, alcohol consumption, smoking status, history of hypertension, dyslipidemia and diabetes, blood glucose, BMI, UA, total cholesterol, HDL, and LDL, were significantly different among the four subgroups (Table 1). During the analysis of sex differences, age, education, smoking status, residential history, history of chronic diseases, hypertension status, diabetes status, BMI, UA levels, total cholesterol levels, HDL levels and LDL levels were significantly different (P <0.05); only hyperlipidemia and GLU levels were not significantly different (P> 0.05) (Table 1).
Table 1. Baseline characteristics of the study participants according to sex
Characteristics
|
|
sex
|
t/x2
|
p
|
Female (n =5391)
|
Male (n =4809)
|
No. of subjects
|
|
5391
|
4809
|
|
|
age
|
mean ± SD
|
66.86±9.83
|
68.08±9.17
|
-6.307
|
0.000**
|
education level, n (%)
|
junior middle school
|
4116(85.59)
|
5029(93.29)
|
162.774
|
0.000**
|
|
High school
|
598(12.44)
|
318(5.90)
|
|
College or above
|
95(1.98)
|
44(0.82)
|
place of residence, n (%)
|
Urban
|
4556(84.51)
|
3922(81.56)
|
18.092
|
0.001**
|
|
Rural
|
829(15.38)
|
885(18.40)
|
|
|
Drinking, n (%)
|
yes
|
671(12.45)
|
2792(58.06)
|
2358.018
|
0.001**
|
Smoking, n (%)
|
yes
|
302(5.60)
|
2894(60.18)
|
3561.31
|
0.001**
|
Chronic disease history, n (%)
|
yes
|
1255(23.28)
|
922(19.17)
|
29.754
|
0.001**
|
Hypertension, n (%)
|
yes
|
286(5.31)
|
205(4.26)
|
6.326
|
0.042*
|
Diabetes, n (%)
|
yes
|
286(5.31)
|
205(4.26)
|
6.326
|
0.042*
|
Dyslipidemia, n (%)
|
yes
|
377(6.99)
|
308(6.40)
|
2.21
|
0.331
|
BMI
|
mean ±SD
|
23.76±3.73
|
22.76±3.26
|
14.427
|
0.000**
|
UA
|
M (IQR)
|
3.860(3.3,4.6)
|
4.79(4.1,5.7)
|
-34.414
|
0.000**
|
Total cholesterol
|
M (IQR)
|
194.46(170.5,220.8)
|
185.18(162.8,208.4)
|
-10.831
|
0.000**
|
HDL
|
M (IQR)
|
50.64(41.8,60.7)
|
48.71(39.4,59.5)
|
-5.015
|
0.000**
|
LDL
|
M (IQR)
|
117.53(95.5,141.9)
|
110.57(90.5,131.4)
|
-9.502
|
0.000**
|
Blood glucose (mg/dl)
|
M (IQR)
|
101.88(94.3,112.3)
|
102.78(94.1,113.8)
|
-1.475
|
0.143
|
Mean ± SD for continuous variables: The p value was calculated by the weighted linear regression model. (%)
For categorical variables, the p value was calculated by the weighted chi-square test, * P<0.05 ** P<0.001.
Abbreviations: BMI: Body mass index; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol;
TC: Total cholesterol; UA: Uric acid.
3.2 Survival curve analysis
For men without other confounding factors, the log-rank test (P=0.830) revealed no significant difference in survival rates between WWI quartiles. However, for women without confounding factors, the log-rank test (P=0.045) indicated a statistically significant difference in survival rates between WWI quartiles. . Additionally, the second quartile had the highest survival rate, suggesting that the Cox proportional hazards regression model could be used to analyze the relationship between the WWI and CKD (Fig. 2).
3.3 The relationship between the WWI and CKD via the Cox proportional hazards model
The prevalence of CKD in each group was 5.86%, 4.89%, 6.16%, and 6.83%, respectively, in the female subgroup. According to the adjusted age model generated with female data, the HR (95% CI) of the highest WWI quartile was 1.57 (1.12-2.20) compared to that of the second WWI quartile. After adjusting for age, smoking status, drinking status, BMI, smoking status, drinking status and chronic disease status (dyslipidemia, diabetes, hypertension), the HR (95% CI) in the lowest quartile was 1.42 (1.01-2.00), and that in the highest quartile was 1.50 (1.05-2.13) compared to that in the second WWI quartile (Table 2). After adjusting for age, smoking status, alcohol consumption, BMI, smoking status, alcohol consumption, chronic disease status (dyslipidemia, diabetes, hypertension) and UA, total cholesterol, HDL and blood glucose levels, the HR (95% CI) in the lowest quartile was 1.56 (1.04 - 2.34), that in the third WWI quartile was 1.59 (1.08 - 2.34), and that in the highest quartile was 1.53 (1.01 - 2.31) compared to that in the second WWI quartile (Table 2). In the male subgroup, the prevalence of CKD in each group was 8.80%, 8.51%, 7.85%, and 8.37%; after adjusting for various variables, the prevalence of CKD in each group did not significantly differ (P>0.05). Based on nonrestrictive cubic splines, we found no nonlinear relationship between the WWI and CKD, regardless of sex (P>0.05) (Supplementary Fig. 1).
Table 2. Association between the WWI and CKD in male and female subjects
Gender
|
WWI
|
Rate/1000 p_years
|
case,(n%)
|
CKD
|
model 1
|
|
model 2
|
|
model 3
|
HR (95% CI)
|
P value
|
|
HR (95% CI)
|
P value
|
|
HR (95% CI)
|
P value
|
|
Total
|
|
403(8.38)
|
|
|
|
|
|
|
|
|
Male
|
Q2
|
13.75
|
102(8.51)
|
Ref
|
|
|
Ref
|
|
|
Ref
|
|
|
Q1
|
14.04
|
105(8.80)
|
0.99 (0.74 - 1.31)
|
0.922
|
|
1.01 (0.75 - 1.37)
|
0.947
|
|
1.19 (0.82 - 1.71)
|
0.356
|
|
Q3
|
12.72
|
95(7.85)
|
0.99 (0.75 - 1.31)
|
0.944
|
|
1.11 (0.82 - 1.52)
|
0.497
|
|
1.10 (0.78 - 1.54)
|
0.592
|
|
Q4
|
13.92
|
101(8.37)
|
0.92 (0.69 - 1.22)
|
0.574
|
|
1.05 (0.79 - 1.40)
|
0.742
|
|
1.08 (0.76 - 1.54)
|
0.682
|
|
Total
|
|
320(5.94)
|
|
|
|
|
|
|
|
|
Female
|
Q2
|
7.51
|
66(4.89)
|
Ref
|
|
|
Ref
|
|
|
Ref
|
|
|
Q1
|
9.22
|
78(5.86)
|
1.29 (0.92 - 1.80)
|
0.138
|
|
1.42 (1.01 - 2.00)
|
0.047
|
|
1.56 (1.04 - 2.34)
|
0.031
|
|
Q3
|
9.63
|
84(6.16)
|
1.30 (0.93 - 1.81)
|
0.119
|
|
1.31 (0.93 - 1.85)
|
0.116
|
|
1.59 (1.08 - 2.34)
|
0.023
|
|
Q4
|
11.09
|
92(6.83)
|
1.57 (1.12 - 2.20)
|
0.011
|
|
1.50 (1.05 - 2.13)
|
0.025
|
|
1.53 (1.01 - 2.31)
|
0.044
|
Note: Male: Q1 (≤10.36), Q2 (10.36-10.80), Q3 (10.80-11.26), and Q4 (>11.26); Female: Q1 (≤10.85), Q2 (10.85-11.42), Q3 (11.42-12.00), and Q4 (>12.00).
Model 1 was adjusted for age.
Model 2 was further adjusted for BMI, smoking status, drinking status, and chronic diseases (dyslipidemia, diabetes, hypertension) based on Model 1.
Model 3 was further adjusted for UA, total cholesterol, HDL, and blood glucose based on Model 2.
Abbreviations: CKD, chronic kidney disease; WWI: weight-adjusted waist circumference index; BMI: body mass index; HDL-C:
High-density lipoprotein cholesterol; UA: Uric acid.
3.4 The relationship between the WWI and CKD according to subgroup
In further analysis, we divided the WWI into two categories, with the median as the cutoff, to analyze the relationship between the WWI and CKD in different subgroups. In the subgroup analysis, significant associations of a higher WWI with the risk of CKD were observed in the hypertension male subgroup. We also observed a significant interaction effect between the WWI and hypertension on CKD incidence. However, no significant association between the WWI and CKD was found in the other subgroups of male participants (Fig. 3). In females, significant associations between a higher WWI and the risk of CKD were not detected in any of the hypertension subgroups; however, significant associations between a higher WWI and the risk of CKD were detected. We also found a significant relationship between a greater WWI and the risk of CKD in the normal subgroup; nonetheless, we observed a significant relationship between a greater WWI and the risk of CKD. Nevertheless, no significant relationship between the WWI and CKD was found in the other subgroups of female participants (Fig. 4).
3.5 Sensitivity analysis
To test the robustness of our findings, we conducted a sensitivity analysis on the common causes of CKD, including diabetes and hypertension. Consequently, the male WWI still had no statistically significant effect on the occurrence of CKD after excluding diabetic patients from the study. Among females, on the other hand, Q3 and Q4 had a positive impact on the occurrence of CKD, unlike Q2; this was in line with the results of not excluding diabetes patients. After excluding hypertensive patients, there was no statistically significant effect of male sex on the WWI on CKD occurrence. However, compared with Q2, Q3 and Q4 had a positive effect on the occurrence of CKD in females; this finding was consistent with the results of not excluding hypertensive patients. This shows the robustness of our findings.
Table 3. Sensitivity analysis of the common causes of CKD
|
Variables
|
Female
|
|
Male
|
|
HR (95%CI)
|
P
|
|
HR (95%CI)
|
P
|
|
WWI
|
|
|
|
|
|
Exclude people with diabetes
|
Q2
|
Ref
|
|
|
|
|
|
Q1
|
1.39 (0.91 - 2.10)
|
0.125
|
|
1.01(0.72,1.42)
|
0.951
|
|
Q3
|
1.46 (1.08 - 2.18)
|
0.048
|
|
0.92(0.65,1.30)
|
0.633
|
|
Q4
|
1.48 (1.09 - 2.23)
|
0.046
|
|
1.04(0.74,1.46)
|
0.826
|
Exclude people with hypertension
|
Q2
|
|
|
|
|
|
|
Q1
|
1.37(0.86,2.19)
|
0.183
|
|
.880(0.61,1.27)
|
0.512
|
|
Q3
|
1.76(1.12,2.74)
|
0.014
|
|
.860(0.60,1.25)
|
0.441
|
|
Q4
|
1.75(1.08,2.83)
|
0.023
|
|
.730(0.49,1.09)
|
0.130
|
A sensitivity analysis of the common causes of CKD, including diabetes and hypertension, was performed to test the robustness of the results.
Abbreviations: CKD, chronic kidney disease; WWI: weight-adjusted waist circumference index.