Table 2

Multivariable Cox and logistic regression analyses to evaluate the association between anti-RNA polymerase III antibodies and development of cancersa
Variable Frequency ( n ) Hazard/Odds ratio 95% CI P-value
2A: Multivariable Cox regression analysis
   Age (per 10 years) 1,740 valid 1.46 1.28 to 1.67 <0.001
   Male vs. female 299 vs. 1,441 0.83 0.50 to 1.39 0.487
   RNAP+ vs. RNAP- 268 vs. 1,472 2.55 1.75 to 3.74 <0.001
2B: Multivariable Cox regression analysis
   Age (per 10 years) 1,694 valid 1.43 1.21 to 1.70 <0.001
   Male vs. female 294 vs. 1,401 0.93 0.53 to 1.80 0.931
   RNAP+ vs. RNAP- 250 vs. 1,445 2.10 1.27 to 3.48 0.004
2C: Logistic regression analysis
   Age (per 10 years) 1,617 valid 1.63 1.27 to 2.09 <0.001
   Male vs. female 265 vs. 1,352 0.73 0.28 to 1.90 0.514
   RNAP+ vs. RNAP- 249 vs. 1,368 5.83 3.11 to 10.92 <0.001

aCI, Confidence interval; RNAP+, positivity for anti-RNA polymerase III antibodies; RNAP-, anticentromere antibodies, anti-Scl-70, antinuclear antibody-negative and other antibodies. These analyses include all events (row 2A) with total frequency N = 2,177, 437 cases dropped, 128 events, 1,612 censored; events post-systemic sclerosis onset (row 2B) with total frequency N = 2,177, 483 cases dropped, 83 events, 1,611 censored; and for events within 36 months of systemic sclerosis onset (row 2C) with total frequency N = 1,623, 6 missing cases, 42 events.

Moinzadeh et al.

Moinzadeh et al. Arthritis Research & Therapy 2014 16:R53   doi:10.1186/ar4486

Open Data