Objectives: To measure the change in body composition in a pre-menopausal female systemic lupus erythematosus (SLE) population over 3 yr, and to identify predictors of change in body composition including the effects of disease-, corticosteroid (CS)- and patient-related variables.
Methods: All 55 pre-menopausal females with SLE who participated in a cross-sectional study of body composition in 1994 were invited to undergo interview, examination, medical record review, and body composition assessment by dual-energy X-ray absorptiometry (DXA).
Results: Twenty-eight subjects participated with a mean (S.E.M.) age of 34.4 (1.6) yr, duration of SLE of 6.8 (0.8) yr and mean (range) time to follow-up of 3.2 (2.9-3.4) yr. Seventeen subjects were exposed to CS during the study period with a mean (range) daily dose of prednisolone of 12.0 (2.8-22.9) mg. There was a significant increase in body mass index (BMI) (24.53+/-0.83 vs 25.37+/-1.04, P = 0.03) and fat-free mass (41.04+/-0.83 vs 41.53+/-0.92, P = 0.05) over the 3 yr period. Univariate analysis revealed that change in fat-free mass was significantly associated with change in total body bone mineral density (BMD) (P = 0.03). Stepwise multiple linear regression analysis revealed a significant independent association of disease activity with increases in both BMI (r2 = 0.41, P = 0.006) and fat mass (r2 = 0.39, P = 0.007), and of exercise and Modified Health Assessment Questionnaire with an increase in fat-free mass (r2 = 0.51, P = 0.007). Age at SLE diagnosis and smoking were significant independent predictors for loss of total body BMD, while CS duration was predictive of an increase in total body BMD (r2 = 0.80, P < 0.0001).
Conclusion: In this SLE population, disease activity was predictive of deleterious changes in body composition, including increases in BMI and fat mass. Patient-related variables were also important predictors of body composition change with exercise independently predicting an increase in fat-free mass, and smoking predictive of loss of total body BMD. In contrast, CS-related variables were not found to have harmful effects on body composition. Change in fat-free mass, and not fat mass, was predictive of change in total body BMD.