Cachexia measured by bioelectrical impedance vector analysis and risk of infection in women with rheumatoid arthritis

Rheumatoid arthritis (RA) patients have a higher frequency of infections than the healthy population. The reason has yet to be explained but involves several factors, of which body composition and rheumatoid cachexia are often overlooked. This study aimed to evaluate whether patients with cachexia, measured by bioelectrical impedance vector analysis, are at an increased risk of developing infections compared with patients without cachexia. A secondary analysis of 186 women with RA enrolled in a randomized trial (ClinicalTrials.gov ID: NCT02900898, September 14, 2016) was completed. Medical records and phone calls were used to record infectious events diagnosed and treated during follow-up. Hazard ratios were calculated using Cox proportional hazard regression analysis, and a predictive model of infection was created. After 36 months of follow-up, 62 patients (26.7% non-cachectic and 44.3% cachectic, p < 0.01) developed at least one infectious event. The most common site of was the urinary tract, followed by the lungs and respiratory tract. The presence of cachexia (HR 1.90, 95% CI 1.15-3.13) and the use of glucocorticoids (HR 1.77, 95% CI 1.01-3.09) were associated with infection in univariate and multivariate models. Body mass index (BMI), smoking, and methotrexate use were not associated with a higher frequency of infections. The presence of cachexia and the use of glucocorticoids were identified as predictors of infections in a cohort of female RA patients. More extensive measurements of body composition should be performed beyond BMI in RA patients to better understand its impact and to prevent additional comorbidities and complications. Key Points • The presence of cachexia measured by bioelectrical impedance vector analysis was associated with infectious events in women with rheumatoid arthritis, whereas body mass index did not show an association. • Glucocorticoids were the only drug associated with a higher frequency of infection. None of the disease-modifying antirheumatic drugs, including methotrexate, showed an association.