Table 3.
Predictors of Carotid Artery Intima-Media Thickness (CIMT) in All Subjects
| Covariate | All Subjects |
|||
|---|---|---|---|---|
| Univariate Models |
Multivariate Models |
|||
| Parameter Estimate (95% CI) | P | Parameter Estimate (95% CI) | P | |
| Baseline CIMT (per µm) | ||||
| Baseline sCD14 (per 1 µg/mL) | −6.90 (−15.97 to 2.17) | .130 | … | … |
| Baseline LPS (per 100 pg/mL) | −0.84 (−2.86 to 1.18) | .410 | … | … |
| Age (per 10 y) | 39.15 (12.04…66.266) | .006 | … | … |
| Female sex | −81.30 (−157.16 to −5.44) | .036 | … | … |
| Nonwhite vs white | −59.56 (−106.10 to −13.02) | .013 | −65.39 (−111.10 to −19.89) | .006 |
| BMI (per kg/m2) | 9.53 (3.66–15.37) | .002 | 10.01 (4.28–15.74) | .001 |
| Waist-hip ratio (per 0.1 unit) | 26.65 (−1.21 to 54.51) | .060 | … | … |
| Insulin (per 10 mU/L) | 30.68 (−1.68 to 63.03) | .063 | … | … |
| Yearly rate of change in CIMT (per 1 µm/y) | ||||
| Baseline sCD14 (per 1 µg/mL) | 1.35 (.14–2.53) | .029 | 1.62 (.44–2.79) | .007 |
| Baseline LPS (per 100 pg/mL) | 0.36 (.11–.62) | .005 | 0.42 (.17–.68) | .001 |
| HDL cholesterol (reference, ≥35 mg/dL) | −5.05 (−10.70 to .61) | .080 | … | … |
| Body mass index (per kg/m2) | −0.88 (−1.77 to −.02) | .055 | … | … |
| Yearly rate of change in sCD14 (per 1 µg/mL/y) | −4.53 (−8.71 to −.36) | .034 | … | … |
Abbreviations: BMI, body mass index; CI, confidence interval; HDL, high-density lipoprotein; LPS, lipopolysaccharide; sCD14, soluble CD14.
P < .10 for all predictors, except baseline sCD14 and LPS. The reference group used for determination of the parameter estimate was clarified as needed. In univariate analysis, the baseline covariates considered were age, sex, race, fasting lipid measurements (total cholesterol, low-density lipoprotein cholesterol, HDL cholesterol, triglycerides, and non-HDL cholesterol), use of lipid-lowering drugs, fasting glucose, BMI, waist circumference, waist-hip ratio, insulin, high-sensitivity C-reactive protein, homocysteine, sCD14, and LPS. Only covariates significant (P < .05) in univariate analysis were considered in the multivariate models; the final multivariate models in this table include only the covariates remaining significant.