Table 4 

Linear and binary regression between classical CVD risk and endothelial function in RA patients from the crosssectional study 

Microvascular function 
Macrovascular function 



Endotheliumdependent 
Endothelium independent 
Endothelium dependent 
Endothelium independent 



CVD risk factors 

Insulin 
β = 0.13, P = 0.23 
β = 0.04, P = 0.73 
β = 0.04, P = 0.71 
β = 0.25, R^{2 }= 0.065^{a} 
HOMA 
β = 0.09, P = 0.41 
β = 0.01, P = 0.95 
β = 0.06, P = 0.61 
β = 0.23, R^{2 }= 0.052^{a} 
QUICKI 
β = 0.07, P = 0.53 
β = 0.02, P = 0.88 
β = 0.11, P = 0.30 
β = 0.30, R^{2 }= 0.095^{b} 
High cholesterol 
OR = 1.01 (0.991.00) 
OR = 1.00 (0.991.00) 
OR = 1.08 (0.991.17) 
OR = 1.09 (1.021.17)^{a} 
Hypertension 
OR = 1.00 (1.001.00) 
OR = 1.00 (0.991.00) 
OR = 1.08 (0.991.17) 
OR = 1.07 (1.011.14)^{a} 
Cigarette smoking 
OR = 1.00 (1.001.00) 
OR = 1.00 (1.001.00) 
OR = 1.04 (0.971.13) 
OR = 1.04 (0.991.10) 
Global CVD risk 

FRS 
β = 0.26, R^{2 }= 0.068^{b} 
β = 0.18, P = 0.09 
β = 0.12, P = 0.26 
β = 0.32, R^{2 }= 0.101^{b} 
TC score 
β = 0.28, R^{2 }= 0.077^{a} 
β = 0.17, P = 0.18 
β = 0.23, P = 0.08 
β = 0.14, P = 0.30 
TC/HDL score 
β = 0.26, R^{2 }= 0.068^{a} 
β = 0.17, P = 0.19 
β = 0.25, P = 0.05 
β = 0.17, P = 0.18 
Reynolds Risk Score 
β = 0.30, R^{2 }= 0.088^{a} 
β = 0.12, P = 0.37 
β = 0.27, R^{2 }= 0.072^{a} 
β = 0.51, R^{2 }= 0.259^{b} 
QRISK 2 
β = 0.33, R^{2 }= 0.109^{b} 
β = 0.19, P = 0.08 
β = 0.22, R^{2 }= 0.048^{a} 
β = 0.51, R^{2 }= 0.255^{b} 


CVD risk factors and global CVD risk scores were entered as independent variables, and endothelial functions, as dependent variables in the linear and binary regression analysis. Odds ratio (OR) with 95% confidence interval is presented for all binary regression analysis. CVD, cardiovascular disease; FRS, Framingham risk score; HDL, highdensity lipoprotein; HOMA, homeostasis model assessment; QUICKI, quantitative insulin sensitivity check index; SCORE, systematic coronary risk evaluation; TC, total cholesterol. ^{a}P < 0.05; ^{b}P < 0.01. R^{2 }value is shown for all significant associations only. 

Sandoo et al. Arthritis Research & Therapy 2012 14:R117 doi:10.1186/ar3847 