Table 5.
Association between the DII and CRP a.
| Subjects | Dietary Inflammatory Beta Estimates (95% CI) | |||
|---|---|---|---|---|
| Tertile 1 | Tertile 2 | Tertile 3 | Continuous | |
| Whole sample | ||||
| Model 1 b | 1.00(Ref.) | −0.004(−0.064,0.055) | −0.032(−0.101,0.036) | −0.012(−0.033,0.009) |
| Model 2 c | 1.00(Ref.) | 0.035(−0.018,0.089) | 0.040(−0.024,0.103) | 0.012(−0.008,0.031) |
| Subjects with MetS | ||||
| Model 1 b | 1.00(Ref.) | 0.072(−0.010,0.154) | 0.110(0.009,0.211) | 0.033(0.003,0.062) |
| Model 2 c | 1.00(Ref.) | 0.086(0.004,0.167) * | 0.145(0.045,0.245) * | 0.040(0.010,0.069) * |
| Subjects without MetS | ||||
| Model 1 b | 1.00(Ref.) | −0.021(−0.097,0.056) | −0.048(−0.134,0.038) | −0.018(−0.045,0.009) |
| Model 2 c | 1.00(Ref.) | 0.021(−0.050,0.091) | 0.001(−0.080,0.082) | 0.002(−0.024,0.027) |
DII: dietary inflammatory index, CRP: C-reactive protein, MetS: metabolic syndrome. a CRP was analyzed as a continuous variable with linear regression; b Model 1 was used to obtain the crude beta estimates; c Model 2 was adjusted for age, gender, city, education level, family monthly expenditure on food, smoking status, and BMI. * Beta estimates (95% CI) of adjusted model that were significant.