LAUSR

OBJECTIVE Anemia is a common problem in rheumatoid arthritis (RA), associated with radiographic progression and disability. We explored the association of hemoglobin with a comprehensive set of variables in RA patients. METHODS We included RA outpatients in the routine setting. For each patient we performed measurements (clinical measures, blood tests including serology, markers of acute phase response and iron metabolism, including hepcidin, and circulating hematopoietic precursor levels) at baseline and 12 weeks thereafter, and analyzed their changes in patients with a treatment adaptation at baseline. We performed principal component analysis (PCA) to identify thematic groups hemoglobin was related to. Then we constructed multivariable linear models to assess the contribution of individual variables to the variability of hemoglobin. RESULTS In total, 88 patients were included (age: 58 ± 12; disease duration: 9.3 ± 9.6 years). Cross-sectionally (at baseline and week 12) hemoglobin levels were tied to iron metabolism and hematopoiesis, but not to clinical activity, based on thematic groups extracted from the PCA. In contrast, longitudinal changes in hemoglobin levels were closely linked to changes in clinical activity. Conversely, hepcidin reflected iron metabolism cross-sectionally, but changes in acute phase response longitudinally. In multivariable analysis variability components of hemoglobin were explainable by ferritin, ESR, evaluator global assessment (EGA), and iron levels, while components of hemoglobin changes were explained by changes in EGA mostly. Hepcidin was not independently associated with hemoglobin. CONCLUSION Besides its dependence on body iron status, changes in hemoglobin levels are strongly tied to disease activity, possibly revealing more about disease activity than other laboratory markers.