Table 5.

Effect of antibody trajectory and level on erosions and DAS remission.

Univariate AnalysesMultiple Logistic Regression
Erosions OR (95% CI)DAS Remission OR (95% CI)Erosionsa OR (95% CI)DAS Remissionb OR (95% CI)
Followup 12 months
  Antibody trajectory
    ACPA
      Always negative (n = 110)1111
      Conversion (n = 13)1.37 (0.26, 7.17)1.21 (0.38, 3.83)1.83 (0.28, 11.85)1.22 (0.33, 4.56)
      Always positive (n = 212)2.81 (1.42, 5.56)*1.21 (0.76, 1.91)3.86 (1.68, 8.92)*1.11 (0.65, 1.91)
      Reversion (n = 7)3.64 (0.55, 23.97)2.59 (0.48, 13.92)3.24 (0.4, 26.2)1.53 (0.26, 8.94)
    RF
      Always negative (n = 165)1111
      Conversion (n = 21)1.8 (0.4, 8.13)0.49 (0.19, 1.22)1.22 (0.16, 9.23)0.55 (0.2, 1.48)
      Always positive (249)1.15 (0.64, 2.07)0.59 (0.39, 0.87)*0.58 (0.26, 1.3)0.59 (0.37, 0.92)
      Reversion (n = 85)1.09 (0.52, 2.29)0.73 (0.43, 1.24)0.74 (0.28, 2.01)0.77 (0.43, 1.28)
  Antibody level
    ACPA
      Negative (n = 123)1111
      Low positive (n = 26)2.03 (0.63, 6.53)2.29 (0.93, 5.65)3.14 (0.86, 11.5)2.06 (0.76, 5.58)
      Moderate/high positive (n = 193)2.84 (1.48, 5.46)1.12 (0.71, 1.75)3.4 (1.61, 7.2)1.02 (0.6, 1.73)
    RF
      Negative (n = 186)1111
      Low positive (n = 122)0.75 (0.39, 1.46)0.58 (0.37, 0.92)0.78 (0.37, 1.66)0.56 (0.33, 0.95)
      Moderate/high positive (n = 212)1.37 (0.77, 2.46)0.73 (0.49, 1.09)1.25 (0.64, 2.44)0.75 (0.48, 1.17)
Erosions OR (95% CI)DAS Remission OR (95% CI)Erosionsc OR (95% CI)DAS Remissiond OR (95% CI)
Followup 24 months
  Antibody trajectory
    ACPA
      Always negative (n = 83)1111
      Conversion (n = 13)1.76 (0.43, 6.98)0.66 (0.2, 2.12)3.39 (0.72, 16.05)0.47 (0.14, 1.59)
      Always positive (n = 166)1.22 (0.58, 2.54)1.25 (0.73, 2.14)1.24 (0.53, 2.92)1.17 (0.66, 2.09)
      Reversion (n = 15)1.23 (0.21, 7.12)1.15 (0.38, 3.52)2.24 (0.34, 14.88)0.98 (0.3, 3.24)
    RF
      Always negative (n = 115)1111
      Conversion (n = 21)1.82 (0.54, 6.2)0.89 (0.35, 2.28)3.25 (0.86, 12.22)0.77 (0.29, 2.08)
      Always positive (n = 173)1.45 (0.71, 2.96)1.28 (0.68, 2.39)1.87 (0.85, 4.14)0.74 (0.44, 1.23)
      Reversion (n = 67)1.38 (0.29, 3.24)0.83 (0.52, 1.34)1.78 (0.71, 4.44)1.18 (0.61, 2.28)
  Antibody level
    ACPA
      Negative (n = 123)1111
      Low positive (n = 26)1.15 (0.28, 4.81)1.51 (0.49, 4.68)1.43 (0.32, 6.38)1.12 (0.33, 3.8)
      Moderate/high positive (n = 193)1.64 (0.79, 3.43)0.91 (0.53, 1.59)1.98 (0.88, 4.44)0.76 (0.41, 1.41)
    RF
      Negative (n = 186)1111
      Low positive (n = 122)0.88 (0.38, 2.06)0.64 (0.36, 1.39)1.14 (0.46, 2.81)0.59 (0.32, 1.1)
      Moderate/high positive (n = 212)1.83 (0.89, 3.77)0.81 (0.49, 1.32)2.37 (1.08, 5.22)0.71 (0.42, 1.2)
  • Multiple logistic regression model included the following variables associated with the outcome in univariate analyses with p < 0.1:

  • a baseline erosions, CRP (baseline ACPA positivity for analyses of RF trajectory and levels);

  • b age, sex, smoking, baseline DAS28, HAQ, SJC;

  • c baseline erosions, DAS28;

  • d age, baseline DAS28, HAQ. ACPA: anticitrullinated protein antibodies; RF: rheumatoid factor; CRP: C-reactive protein; DAS28: Disease Activity Score-28; HAQ: Health Assessment Questionnaire 28 score; SJC: swollen joint count. Bonferroni correction for multiple comparisons was formed.

  • * p < 0.0083;

  • p < 0.017.