Premium
Determinants of generalized fatigue in individuals with symptomatic knee osteoarthritis: The MOST Study
Author(s) -
Fawole Henrietta O.,
Riskowski Jody L.,
Dell'Isola Andrea,
Steultjens Martijn P.,
Nevitt Michael C.,
Torner James C.,
Lewis Cora E.,
Felson David T.,
Chastin Sebastien F. M.
Publication year - 2020
Publication title -
international journal of rheumatic diseases
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.795
H-Index - 41
eISSN - 1756-185X
pISSN - 1756-1841
DOI - 10.1111/1756-185x.13797
Subject(s) - medicine , osteoarthritis , confidence interval , cohort , cohort study , physical therapy , population , visual analogue scale , linear regression , pathology , alternative medicine , environmental health , machine learning , computer science
Aim The aim of the study was to identify sociodemographic, disease‐related, physical and mental health‐related determinants of fatigue at 2‐year follow‐up in individuals with symptomatic knee osteoarthritis (OA). Methods A longitudinal analysis of participants with symptomatic knee OA from the Multicenter Osteoarthritis Study (MOST) was conducted to identify predictors of fatigue at 2‐year follow‐up. Participants self‐reported fatigue at baseline for the first time in the MOST cohort and at follow‐up using a 0‐10 visual analog scale. At baseline, questionnaires on sociodemographics, disease‐related symptoms, physical and mental health factors were completed. Data were analyzed using linear regressions with a backwards elimination approach. Results Of the 2330 individuals in the MOST cohort at baseline, 576 had symptomatic knee OA and of these, 449 with complete fatigue values at baseline and follow‐up were included in this analysis. Minimally important fatigue change (ie, worsening [≥1.13], no change [<0.82 or <1.13] and improvement [≥−0.82]) from baseline to follow‐up were unequal within the population (34.5%, 26.9%, 38.5%; χ 2 [2, N = 449] = 9.32, P = .009). The multiple linear regression showed that baseline fatigue (unstandardized coefficient [ Β ] = 0.435; 95% confidence interval [CI] 0.348‐0.523, P < .001), slow gait speed ( Β = −1.124; 95% CI −1.962 to −0.285, P = .009), depressive symptoms ( Β = 0.049; 95% CI 0.024‐0.075, P < .001) and higher numbers of comorbidities ( Β = 0.242; 95% CI 0.045‐0.439, P = .016) were significant predictors of greater fatigue at follow‐up. Conclusion Fatigue is strongly associated with physical‐ and mental‐related health factors. Individualized treatments that include combined psychological and physical function rehabilitation might be modalities for fatigue management.