Table 1. Multivariate logistic regression analysis for predicting radiographic progression using conventional risk factors plus urinary protein markers.

Model	Odds ratio (95% CI)	P-value
Model 1 (Basic model+ORM1, per 1-s.d. increase)	1.27 (0.90–1.79)	0.166
Model 2 (Basic model+ORM2, per 1-s.d. increase)	1.25 (1.01–1.54)	0.044
Model 3 (Basic model+GSN, per 1-s.d. increase)	0.79 (0.59–1.05)	0.101
Model 4 (Basic model+CD14, per 1-s.d. increase)	1.00 (0.99–1.00)	0.501

Abbreviations: CI, confidence interval; GSN, gelsolin; ORM, orosomucoid (α1-acid glycoprotein).

Basic model included age, disease duration, C-reactive protein, anti-cyclic citrullinated peptide antibody positivity and glomerular filtration rate at baseline.

Logistic regression was used to obtain basic model-adjusted odds ratios for the presence of radiographic progression.