Abstract
OBJECTIVE: Frailty occurs prematurely in rheumatoid arthritis (RA) and is associated with poor health outcomes. We compared the performance of four frailty instruments, including a pragmatic alternative measure using chair sit-to-stand (STS), and evaluated their abilities to predict poor health outcomes.
METHODS: Frailty was measured at baseline using four instruments: the Fried Frailty Phenotype with STS (Fried-STS), the Fried Frailty Phenotype with hand grip strength (Fried-HGS), the Veterans Affairs Frailty Index (VA-FI), and the FRAIL Scale. Outcomes collected at the one-year follow-up included category of falls (none, one, more than one), category of days hospitalized (none, one to three, more than three), and a composite outcome of fall, hospitalization, or death. Ordinal logistic or logistic regression models, adjusted for age and sex, explored the association of frailty and each outcome.
RESULTS: A total of 143 participants were included (mean ± SD age 64.5 ± 11.7 years, 73% male, and 69% White). Categorization as frail differed by instrument: Fried-STS, 17%; Fried-HGS, 15%; VA-FI, 36%; and FRAIL Scale, 20%. There was poor agreement between frailty instruments (k = 0.07-0.31) except for the Fried-STS and Fried-HGS (k = 0.62). Frailty by the Fried-STS, Fried-HGS, and FRAIL Scale was associated with falls (adjusted odds ratios [aORs] 3.83-9.54, P < 0.05). Frailty by the VA-FI was associated with days hospitalized (aOR 5.21, P = 0.017). Frailty by the Fried-STS, VA-FI, and FRAIL Scale was associated with higher odds of the composite measure of incident fall, hospitalization, or death (aORs 2.93-7.25, P < 0.05).
CONCLUSION: Each frailty measure predicted adverse health outcomes, with phenotypic and patient-reported measures predicting falls and the deficit accumulation model predicting hospitalization days. Being frail by the Fried-HGS did not predict poor outcomes as well as the other frailty instruments, including the Fried-STS.