Skip to main content

Table 2 Regression coefficients and OR (95% CI) of best models based on backwards regression analysis for the progression of carotid plaque over 3 years in RA patients

From: Radial BMD and serum CTX-I can predict the progression of carotid plaque in rheumatoid arthritis: a 3-year prospective cohort study

 

Model 1: Clinical factors (n = 252)

Model 2: CTX-I (n = 244)

Model 3: BMD radius (n = 223)

Model 4: CTX-I+ BMD radius (n = 216)

Baseline parametersa

β coefficient (SE)

OR (95% CI)

β coefficient (SE)

OR (95% CI)

β coefficient (SE)

OR (95% CI)

β coefficient (SE)

OR (95% CI)

Age

0.036 (0.014)

1.037 (1.010–1.065)**

0.040 (0.014)

1.041 (1.012–1.070)**

–

–

–

–

Hypercholesterolemia

0.896 (0.381)

2.450 (1.162–5.164)*

1.024 (0.393)

2.784 (1.290–6.010)**

0.963 (0.390)

2.621 (1.221–5.624)*

1.099 (0.404)

3.002 (1.360–6.626)**

Smoking: ever and current

0.611 (0.212)

1.843 (1.217–2.791)**

0.649 (0.216)

1.913 (1.252–2.924)*

0.903 (0.261)

2.467 (1.480–4.111)**

0.988 (0.272)

2.687 (1.577–4.579)***

Presence of carotid plaque

0.992 (0.288)

2.697 (1.534–4.739)**

1.002 (0.295)

2.723 (1.527–4.853)**

1.291 (0.282)

3.637 (2.095–6.316)***

1.353 (0.291)

3.869 (2.190–6.838)***

Serum CTX-I (pg/mL)

NA

NA

1.539 (0.667)

4.662 (1.262–17.219)

NA

NA

1.718 (0.731)

5.574 (1.331–23.341)*

BMD radius (g/cm2)

NA

NA

NA

NA

–3.057 (1.193)

0.047 (0.005–0.488)*

–3.667 (1.250)

0.026 (0.002–0.296)**

R2 (%)

24.3

27.3

25.6

29.7

  1. aExcluded variables which has statistical significance of p < 0.10 for the progression of carotid plaque in univariate analysis were as follows: sex, waist-hip ratio, the presence of hypertension, serum LDL-cholesterol level, serum uric acid level, mHAQ, RF, and ESR
  2. *p < 0.05; **p < 0.01; ***p < 0.001. R2: Nagelkerke R square
  3. Abbreviations: β partial regression coefficient; SE standard error; OR odds ratio; CI confidence interval; CTX-I C-terminal telopeptides type I collagen; BMD bone mineral density; NA not-available; R2 coefficient of determination