Fatigue in spondyloarthritis: a marker of disease activity. A cross-sectional study of 266 patients
K. Chauffier, S. Paternotte, V. Burki, A. Durnez, M. Elhaï, E. Koumakis, M. Meyer, J. Payet, I. Fabreguet, F. Roure, M. Dougados, L. Gossec
2013 Vol.31, N°6
PI 0864, PF 0870
Free to view
(click on article PDF icon to read the article)
PMID: 24144262 [PubMed]
Accepted : 17/04/2013
In Press: 11/10/2013
Fatigue is an important aspect of spondyloarthritis (SpA). However the influencing factors of fatigue in SpA are unclear. The objective of this study was to explore if fatigue is related to disease activity or to patient characteristics.
This was a retroelective observational study (Cochin COSPA study) in one tertiary-referral centre. The primary outcome was fatigue, evaluated on a 0–100mm Visual Analogue Scale (VAS). The covariates were demographic characteristics, disease subtype (axial vs. peripheral) and disease-related factors, e.g. Bath Disease Activity Index (BASDAI), patient global assessment (VAS), Bath Functional Index (BASFI). To explain fatigue, univariate then multivariate logistic regressions were conducted (with fatigue analysed as above or below 50 mm), as well as multiple linear regressions with the different covariates.
Two hundred and sixty-six SpA patients were analysed. Sixty-one percent were male; mean age and disease duration were 44.5±13.5 years and 16.8±11.7 years, respectively. Mean VAS fatigue was 49.3±32.7mm; 49.6% of patients had fatigue VAS>50mm. Logistic regression showed high fatigue was associated with disease: BASDAI and BASFI (p<0.0001), as well as female gender (p=0.025) and aerobic exercise (p=0.005), but there was no difference in the subtypes of SpA. In multivariate analysis, the single factor explaining fatigue was patient global assessment (p<0.001 and odds ratio =1.35). By linear regression, demographic variables explained 2.8% of the variance, whereas disease characteristics and activity explained 44.6%.
Fatigue levels were high in SpA patients whatever the subtype and appeared more strongly related to the disease than to patient-related variables, thus confirming its usefulness as an outcome measure.