Table 6.
Multivariate linear regression analysis of log transformed RHI on the 7 individual endothelial related drugs (n = 868)a.
|
Model 1b Expβ (95% CI) |
p-value |
Model 2c Expβ (95% CI) |
p-value | |
|---|---|---|---|---|
| NSAIDS | 0.98 (0.93, 1.02) | 0.366 | 0.98 (0.93, 1.03) | 0.424 |
| Methotrexate | 0.97 (0.93, 1.02) | 0.295 | 0.98 (0.93, 1.03) | 0.346 |
| Leflunomide | 0.96 (0.88, 1.05) | 0.367 | 0.97 (0.89, 1.05) | 0.404 |
| Hydroxychloroquine | 1.06 (0.99, 1.12) | 0.086 | 1.07 (1.00, 1.14) | 0.044 |
| Sulfasalazine | 0.88 (0.78, 0.98) | 0.020 | 0.89 (0.80, 0.99) | 0.032 |
| TNF inhibitor | 0.99 (0.94, 1.04) | 0.574 | 1.00 (0.94, 1.05) | 0.784 |
| Non-TNF inhibitors | 0.93 (0.84, 1.03) | 0.167 | 0.91 (0.83, 1.01) | 0.084 |
a
Using multiple imputaion (MI) with n = 598 complete observations and n = 270 non-complete observations across covariates in models 1 and 2. MI was perfromed using chained equations and the KNN predictive mean-matching algorithm.
b
Model 1: All 7 drugs and adjusted for age and gender.
c
Model 2: All 7 drugs and adjusted for age, gender, BMI, diabetes, HDL-cholesterol, LDL-cholesterol, family history of CHD, smoking status (Never, Light, Moderate, Severe, or Former), duration of RA, DAS28 score, steroid dose (mg/day), Hypertension, and CRP.