Abstract
OBJECTIVE: To identify factors associated with the loss of cortical diaphyseal bone over time in patients with rheumatoid arthritis (RA). METHODS: We measured the combined cortical thickness (CCT) of the second metacarpal bone from digitized hand radiographs in an RA cohort. We estimated the rate of loss of CCT, and tested the effect of factors that could accelerate the rate. RESULTS: We studied 649 patients, who had 2990 hand radiographs. The median interval between the first and last followup radiograph was 2 years (range 0 to 23 yrs). The mean CCT at baseline was 4.04 mm (standard deviation 1.18). CCT loss was greatest during the earliest observation stages, following a square-root function of time at a rate of 0.393 mm/year(1/2) (95% CI 0.360, 0.423). Patients with a mean erythrocyte sedimentation rate (ESR) < 30 mm/h lost CCT at a rate of 0.303 mm/year(1/2) (95% CI 0.247, 0.358); those with a mean ESR > 30 and < or = 60 mm/h lost CCT at 0.395 mm/year(1/2) (95% CI 0.345, 0.446); and those with ESR > 60 mm/h lost CCT at 0.554 mm/year(1/2) (95% CI 0.480, 0.628). Patients who received a cumulative dose of glucocorticoids > or = 11.7 g lost CCT at 0.659 mm/year(1/2) (95% CI 0.577, 0.742), compared to 0.361 mm/year(1/2) (0.323, 0.401) in patients who did not receive glucocorticoids. In a multivariable model, the effect of the ESR and cumulative glucocorticoids was independent of age at RA onset, RA duration, sex, ethnic group, body mass index, presence of rheumatoid nodules, rheumatoid factor, and the HLA-DRB1 shared epitope. CONCLUSION: Early rapid loss of cortical diaphyseal bone occurs in patients with RA, followed by gradual slowing. Systemic inflammation and glucocorticoids seem to accelerate bone loss independently of other risk factors. Cortical diaphyseal bone may be an important target of the disease process in RA.