Abstract

OBJECTIVES--Although periarticular osteoporosis is a well-recognised phenomenon in rheumatoid arthritis (RA), there is considerable controversy over whether RA is associated with more generalised osteoporosis. The aetiology of this bone loss is probably multifactorial, including both life-style risk factors and disease-related determinants. Population-based studies on bone mineral density (BMD) in RA have not previously been conducted, and the purpose of the present cross-sectional population-based study was to determine whether patients with RA are at an increased risk of having osteoporosis. Furthermore, the determinants of BMD in RA patients were investigated. METHODS--BMD at the spine and femoral neck was measured in 143 women with RA. The control group consisted of 1611 women with no disease or taking any drugs known to affect bone metabolism. The study population was a random stratified sample from the Kuopio Osteoporosis Study, which included all perimenopausal women aged 47-56 years residing in Kuopio Province, Eastern Finland in 1989 (n = 14,220). The mean age of the patients at the time of densitometry was 53.7 years. RESULTS--The mean (SD) spinal and femoral neck BMD was significantly lower in patients with RA compared with controls [spine: 1.067 (0.161) v 1.129 (0.157) g/cm2, p < 0.001; femoral neck: 0.851 (0.136) v 0.932 (0.123) g/cm2, p < 0.001]. Analysis of variance showed that at the spine the difference was significant only in patients having corticosteroid treatment, whereas at the femoral neck patients with non-steroid treatment also had significantly lower BMD. When confounding factors were corrected, no significant difference could be found between non-steroid and corticosteroid treated patients with RA, suggesting that the independent effect of corticosteroids on BMD is only minimal. Multiple regression analysis found age, weight and functional grade to be significant predictors of spinal BMD (R2 = 0.403, p < 0.001). In the femoral neck weight, cumulative corticosteroid dose and functional grade were significant predictors of BMD (R2 = 0.410, p < 0.001). CONCLUSIONS--RA is associated with generalised osteoporosis. The physical impairment and body weight are the major determinants of both spinal and femoral bone mass in RA patients. The cumulative corticosteroid dose was also a significant determinant of femoral neck BMD. However, the independent effect of corticosteroids is questionable because the use of corticosteroids may be an indicator of more severe disease.