Body composition and bone mineral density in rheumatoid arthritis : influence of inflammation and treatment with glucocorticoids and TNF-blocking agents

Abstract

Rheumatoid arthritis (RA) is a chronic inflammatory disease. Besides symptoms from the joints, changes in body composition and reduced bone mineral are common. Increased fat mass (FM) and loss of muscle mass contribute to increased morbidity and mortality. The aim of this thesis was to study the impact of inflammation and medical treatment on body composition and bone mineral density (BMD) in patients with RA.

The work is based on two cross-sectional studies in established RA and two prospective, randomized studies in early RA, analysing the effects of glucocorticoids (GCs) and anti-TNF therapy, respectively. The patients were assessed clinically and by assays, such as markers of bone remodelling, insulin-like growth factor-1 (IGF-1), apolipoproteins, leptin and adiponectin. Dual X-ray energy absorptiometry (DXA) was performed and fat free mass index (FFMI, kg/m2) and fat mass index (FMI, kg/m2) were calculated.

Low fat free mass (FFM), below the 10th percentile of a reference material, was frequent. The highest frequency, 50%, was found in RA inpatients and was associated with high disease activity, physical disability and low bioavailable IGF-1. The proportion of outpatients with established disease that had low FFM was 38% and in patients with early RA 19%. Between 34 and 45% of the patients had high FM, above the 90th percentile of the reference population, and 80% had FM% corresponding to overweight or obesity. The frequency of osteoporosis was 26-28% in established RA and 9% in early RA. In established RA, patients with long-term GC therapy had higher FM than those without, whereas BMD in lumbar spine and femoral neck did not differ between patients treated with GC and those not. Further, there was no significant difference between the treatment groups in the markers of bone formation.

In contrast, in the prospective study in early RA, GC treatment was associated with a rapid decrease in the marker of bone synthesis during the first 3 months in contrast to patients not treated with GC. Also the markers of bone resorption decreased and it seemed that GC treatment could counteract the negative impact of inflammation on bone tissue, as BMD in femoral neck was preserved after 2 years treatment. However, GC treatment could not prevent bone loss in spine in postmenopausal women, where BMD decreased significantly more than in those not treated with GC. Treatment with TNF antagonists was associated with a significant increase in FM after 2 years, an increase that was not found in patients treated with a combination of disease modifying anti-rheumatic drugs (DMARDs), despite similar reduction of disease activity. The increase of FM seemed thus to be a specific effect of anti-TNF blockade and was not associated with an atherogenic lipid profile. Further, levels of leptin and adiponectin increased, of which adiponectin normally is associated with improved insulin sensitivity and endothelial function. This may at least partially explain the reduced frequency of CVD found when disease activity is reduced in RA, also when anti-TNF therapy is used.

In conclusion, a large proportion of RA patients have changes in body composition, which contribute to increased morbidity and mortality. GC treatment is associated with increased FM and can counteract the negative impact of inflammation on bone but need special attention to postmenopausal women. Anti-TNF therapy also increases FM but at the same time increases adiponectin, which have favourable effects on CVD risk factors.