4.1 Baseline characteristics
In the final analysis, the mean age of the 4766 participants was 51.88 years, with males accounting for 49.03% and female 50.97%, of that, a total of 2351 participants (49.33%) suffered from periodontitis and 2415(50.67%)without periodontitis. Among tertiles, there were significant differences in the distribution of gender, race, education, annual household income, smoking status, alcohol consumption, BMI, diabetes, hypertension and congestive heart failure (all p < 0.05). Compared to the other two groups, participants in tertile 3 displayed a propensity for lower education level, lower income, larger waist circumference, higher BMI and a higher prevalence of diabetes, hypertension, and congestive heart failure. detailed information is presented in Table 1.
Table 1. Weighted baseline characteristics of study participants according to tertile groups of CMI.
Variable
|
Overall
(N=4766)
|
T1
(N=1589)
|
T2
(N=1588)
|
T3
(N=1589)
|
p value
|
Age (years)
|
51.88 ± 14.14
|
51.22 ± 14.32
|
52.60 ± 14.47
|
51.82 ± 13.59
|
0.021
|
Gender (%)
|
|
|
|
|
<0.001
|
Male
|
49.0
|
37.5
|
50.1
|
59.5
|
|
Female
|
51.0
|
62.5
|
49.9
|
40.5
|
|
Race (%)
|
|
|
|
|
<0.001
|
Mexican American
|
14.3
|
9.0
|
15.1
|
18.9
|
|
Non-Hispanic White
|
44.0
|
43.9
|
41.6
|
46.3
|
|
Non-Hispanic Black
|
19.4
|
25.7
|
20.6
|
12.0
|
|
Other
|
22.3
|
21.5
|
22.7
|
22.8
|
|
Education (%)
|
|
|
|
|
<0.001
|
<High school
|
23.7
|
17.9
|
24.6.
|
28.7
|
|
≥High school
|
76.3
|
82.1
|
75.4
|
71..3
|
|
Marital status (%)
|
|
|
|
|
0.238
|
Married/Living with Partner
|
66.5
|
65.3
|
66.6
|
67.5
|
|
Widowed/Divorced/ Separated
|
21.9
|
22.8
|
21.2
|
21.8
|
|
Never married
|
10.9
|
11.4
|
11.6
|
9.6
|
|
Not recorded
|
0.7
|
0.5
|
0.6
|
1.0
|
|
Annual household income (%)
|
|
|
|
|
0.01
|
< $20,000
|
20.5
|
17.8
|
20.2
|
23.5
|
|
≥ $20,000
|
75.5
|
77.5
|
76.1
|
72.9
|
|
Not recorded
|
4.0
|
4.7
|
3.7
|
3.6
|
|
Smoked at least 100 cigarettes in life (%)
|
|
|
|
|
<0.001
|
Yes
|
43.4
|
39.2
|
41.8
|
49.1
|
|
No
|
56.6
|
60.8
|
58.2
|
50.9
|
|
Had at least 12 alcohol drinks/one year? (%)
|
|
|
|
|
<0.001
|
Yes
|
68.4
|
68.0
|
66.9
|
70.4
|
|
No
|
24.5
|
22.8
|
25.6
|
25.0
|
|
Not recorded
|
7.1
|
9.1
|
7.5
|
4.6
|
|
BMI (kg/m2) (%)
|
|
|
|
|
<0.001
|
Normal(<25)
|
27.0
|
49.5
|
22.2
|
9.4
|
|
Overweight (25-30)
|
34.9
|
32.5
|
38.5
|
33.5
|
|
Obese(≥30)
|
38.1
|
18.0
|
39.2
|
57.1
|
|
Diabetes (%)
|
|
|
|
|
<0.001
|
Yes
|
11.7
|
5.4
|
11.9
|
17.9
|
|
No
|
85.5
|
92.3
|
85.5
|
78.7
|
|
Borderline
|
2.8
|
2.3
|
2.6
|
3.4
|
|
Hypertension (%)
|
|
|
0.004
|
|
<0.001
|
Yes
|
37.5
|
29.4
|
41.8
|
49.1
|
|
No
|
62.5
|
70.6
|
58.2
|
50.9
|
|
CHD (%)
|
|
|
|
|
0.078
|
Yes
|
2.9
|
2.1
|
3.0
|
3.5
|
|
No
|
97.1
|
97.9
|
97.0
|
96.5
|
|
Angina (%)
|
|
|
|
|
0.017
|
Yes
|
1.8
|
1.0
|
2.0
|
2.3
|
|
No
|
98.2
|
99.0
|
98.0
|
97.7
|
|
CHF (%)
|
|
|
|
|
<0.001
|
Yes
|
2.0
|
1.1
|
1.8
|
3.0
|
|
No
|
98.0
|
98.9
|
98.2
|
97.0
|
|
Stroke (%)
|
|
|
|
|
0.904
|
Yes
|
2.7
|
2.9
|
2.6
|
2.7
|
|
No
|
97.3
|
97.1
|
97.4
|
97.3
|
|
Abbreviations: BMI: body mass index. CHD: coronary heart disease. CHF: congestive heart failure.
4.2 Relationship between CMI and periodontitis
The multivariable logistic regression analysis between CMI and periodontitis were displayed in Table 2. A positive association between CMI and periodontitis was found in all three models. In the fully adjusted model, participants in the highest quantile had a 33% higher risk of periodontitis compared to those in the lowest quantile of CMI [1.332(1.115, 1.591), p =0.00161], Furthermore, the smoothed curve fitting supported a positive linear relationship as shown in Fig.2. The linear relationship between the two was reinforced by two-piecewise regression (p for log-likelihood ratio > 0.05, the linear regression providing a better fit), as indicated in Table 3.
Table 2. Association between CMI and periodontitis in multivariable logistic regression
|
|
OR (95%CI), p-value
|
|
Exposure
|
Crude model
|
Model I
|
Model II
|
CMI
|
1.227 (1.132, 1.330) <0.00001
|
1.228 (1.126, 1.339) <0.00001
|
1.158 (1.052, 1.273) 0.00259
|
CMI(tertiles)
|
|
|
|
T1 Group
|
1.0
|
1.0
|
1.0
|
T2 Group
|
1.357 (1.180, 1.560) 0.00002
|
1.184 (1.018, 1.337) 0.02824
|
1.199(1.019, 1.410) 0.02858
|
T3 Group
|
1.573(1.367,1.809) <0.00001
|
1.405 (1.205, 1.639) 0.00001
|
1.332(1.115, 1.591) 0.00161
|
Abbreviations: CI: confidence interval;OR: odds ratio;CMI: Cardiometabolic index;T: Tertile
The crude model adjusts for: None.
Model Ⅰ adjusts for gender; age; race.
Model II adjusted for gender, age, race, marital status, education level, annual family income level, alcohol consumption, smoking status, BMI, diabetes, hypertension, coronary heart disease, congestive heart failure, angina and stroke
Model Ⅱ adjust for: gender; age; race; marital status; education level; Annual family income level; alcohol consumption; smoking status; BMI; diabetes; hypertension; coronary heart disease; congestive heart failure; angina and stroke;
Figure.2 The association between CMI and periodontitis: smooth curve fitted by a generalized additive model, a solid red line denoted the fitted curve and the two blue lines represented 95% confidence intervals. Model adjusted for gender, age, race, marital status, education level, annual family income level, alcohol consumption, smoking status, BMI, diabetes, hypertension, coronary heart disease, congestive heart failure, angina and stroke.
Table 3. Association between CMI and periodontitis using two-piecewise linear regression
Outcome
|
Periodontitis
OR (95%CI)
|
P value
|
Model 1 (linear regression)
|
1.158 (1.052, 1.273)
|
0.0026
|
Model 2 (segmented regression)
|
|
|
Turning point(K)a
|
0.407
|
|
CMI < K effect
|
2.220 (0.969, 5.085)
|
0.0592
|
CMI > K effect
|
1.126 (1.018, 1.246)
|
0.0207
|
P for log-likelihood ratio
|
0.121
|
|
4.3 Results of subgroup analysis
Subgroup analyses and interaction tests stratified by age, sex, race, smoke, alcohol consumption, BMI, hypertension and diabetes were performed to assess potential differences in the association between CMI and periodontitis. The logistic regression analysis results for the subgroups indicated that CMI and periodontitis remained positive correlation (Fig 3). However, the correlation was more pronounced in female participants (1.20 [1.02, 1.41], p < 0.05) than in male participants (1.14 [1.01, 1.28], p < 0.05). Furthermore, CMI was positively associated with periodontitis in individuals between the ages of 30 and 42, non-Hispanic White, overweight, no-diabetes, no smoke, had no hypertension, drinking or no drinking (p <0.05). However, none of the interaction factors exhibited statistical significance (p interaction > 0.05).