Table 3.

Factors associated with baseline glucocorticoid use, using multivariable logistic regression.

VariablesBaseline Use, OR (95% CI), N = 17,514p
Age, yr
  < 50Reference
  50–651.05 (0.95–1.17)0.35
  65–800.88 (0.78–0.99)0.03
  ≥ 800.93 (0.80–1.07)0.31
Female0.83 (0.74–0.91)< 0.001
Race
  WhiteReference
  Nonwhite1.22 (1.12–1.33)< 0.001
  Unknown0.81 (0.73–0.90)< 0.001
Smoking1.16 (1.09–1.24)<0.001
BMI, kg/m20.98 (0.98–0.99)< 0.001
Ln(CRP), mg/dl1.13 (1.10–1.16)< 0.001
Depression1.08 (1.01–1.16)0.04
Anxiety1.09 (1.00–1.18)0.05
Diabetes0.77 (0.71–0.84)< 0.001
Cancer0.90 (0.80–1.02)0.09
ILD1.28 (1.08–1.53)0.005
Asthma/COPD1.18 (1.09–1.28)< 0.001
Charlson score, per 1 point1.11 (1.08–1.14)< 0.001
HCQ use1.38 (1.26–1.50)< 0.001
Prior conventional DMARD0.85 (0.79–0.92)< 0.001
Year 2011–2014 vs 2006–20101.05 (0.98–1.12)0.16
  • Tested but excluded with p > 0.2: anti-CCP antibody positive, posttraumatic stress disorder, congestive heart failure, chronic kidney disease, myocardial infarction, concomitant leflunomide/sulfasalazine. BMI: body mass index; HCQ: hydroxychloroquine; ln(CRP): natural logarithm of C-reactive protein; COPD: chronic obstructive pulmonary disease; DMARD: disease-modifying antirheumatic drug; ILD: interstitial lung disease; CCP: cyclic citrullinated peptide.