Fig. 1

Histograms showing median erosion scores in patients with Clinically Suspect Arthralgia comparing anti-citrullinated protein antibodies (ACPA)-positive and ACPA-negative patients (a), patients positive or negative for local subclinical joint inflammation (b), ACPA positivity and negativity in relation to the concomitant presence of magnetic resonance imaging (MRI)-detected subclinical inflammation (c), or rheumatoid factor (d). Median erosion scores with the upper limit of the interquartile range (75th percentile): **p < 0.01; *p < 0.05; NS, non-significant. The following comparisons have been made: ACPA+ vs. ACPA− (a) (p = 0.006) and MRI+ vs. MRI− (b) (p < 0.001). Next, ACPA+MRI− vs. ACPA–MRI– patients (c) (p = 0.68), ACPA+MRI+ vs. ACPA−MRI– patients (c) (p < 0.001) and finally ACPA+MRI− vs. ACPA+MRI+ (c) (p = 0.016). ACPA+ rheumatoid factor (RF)− patients vs. ACPA−RF− patients (d) (p = 0.30) and ACPA+RF+ patients vs. ACPA−RF− patients (d) (p = 0.006)