Table 4

Novel Marker Combinations Used to Predict CAC Progression

Combination Novel Marker ModelOR per 1 SD (95% CI)c StatisticAICLR Test p Value
Base model0.71102,524.65
Model 10.71642,518.03<0.01
 Fibrinogen1.06 (1.05–1.07)
 sICAM1.11 (0.81–1.52)
 Factor VIIIc1.05 (0.89–1.25)
 CIMT1.06 (1.02–1.10)
Model 20.71602,517.55<0.01
 LDLpn0.91 (0.76–1.08)
 CRP1.06 (0.96–1.18)
 Fibrinogen0.98 (0.87–1.11)
 Urine albumin1.11 (0.95–1.31)
 sICAM1.17 (0.96–1.43)
Model 3 (backward stepwise selection)0.71582,513.01
 sICAM1.18 (0.97–1.43)
 Factor VIIIc1.09 (0.99–1.20)
 Age1.08 (1.06–1.09)
 Sex1.53 (1.17–2.00)
 Race/ethnicity0.87 (0.80–0.95)
 DBP1.02 (1.01–1.03)
 Antihypertensive medication use1.76 (1.39–2.23)
 Total cholesterol1.01 (1.00–1.01)
 HDL0.99 (0.98–0.99)
 Smoking1.42 (0.98–2.06)
 BMI1.03 (1.01–1.05)
 Physical activity1.00 (1.00–1.00)
 Family history of heart attack1.45 (1.19–1.78)
sICAM1.17 (0.96–1.43)0.71502,515.21<0.01
CRP alone1.07 (0.97–1.18)0.7112,524.660.15
Ln baseline CAC score (for those with baseline CAC >0, n =790)1.59 (1.43–1.77)0.739

AIC = Akaike information criterion; CI = confidence interval; Ln = natural log; LR = likelihood ratio; other abbreviations as in Tables 1, 2, and 3.

  • Base model includes age, sex, race/ethnicity, SBP, DBP, antihypertensive medication use, current smoking, total and HDL cholesterol, family history of heart attack, BMI, and physical activity. †The rest of models are adjusted for the base model.