Frailty Assessment in Polymyalgia Rheumatica: Comparative Performance and Determinants of Clinical-Research Frailty...

Background/Objectives: Frailty is increasingly recognized as an important condition in polymyalgia rheumatica (PMR), although different instruments may assess different dimensions of vulnerability. This study aimed to compare the determinants and clinical correlates of frailty measured by the Comprehensive Rheumatologic Assessment of Frailty (CRAF) and the Italian Kihon Checklist (I-KCL). Methods: Fifty-two patients with PMR were evaluated. The discriminative performance of CRAF and I-KCL was assessed against the Cardiovascular Health Study (CHS) frailty phenotype using ROC analysis. Multivariable linear regression identified the independent determinants of each frailty measure. Results: The distribution of frailty status varied across the three assessment tools. The CRAF score identified a higher proportion of non-frail patients (44.2%) and a lower proportion of frail individuals (19.2%) compared to both the I-KLC and CHS criteria. Discriminative accuracy of CRAF and I-KCL scores against the CHS/Fried phenotype reference standard was excellent for both instruments: CRAF AUC = 0.96 (95% CI 0.888-1, bootstrap, 2000 stratified resamples) and I-KCL AUC = 0.899 (95% CI 0.704-1). The difference between CRAF and I-KCL discriminative accuracy was therefore not statistically significant (z = 0.941, p = 0.3467). In multivariable linear regression models adjusted for age, disease duration, PMR-AS(CRP), comorbidity count, HAQ, and fatigue, PMR-AS(CRP) (standardized β = 0.51, p < 0.001) and comorbidity count (standardized β = 0.36, p < 0.001) were independently associated with the CRAF score (adjusted R² = 0.787, n = 51), while PMR-AS(CRP) (standardized β = 0.40, p < 0.001) and comorbidity count (standardized β = 0.46, p < 0.001) were independently associated with the I-KCL score (adjusted R² = 0.675, n = 51). A further finding was that functional impairment, measured by the HAQ, was independently associated with the CRAF score (standardized β = 0.19, p = 0.018) but not with the I-KCL score (standardized β = 0.06, p = 0.512). Conclusions: Although CRAF and I-KCL share these core determinants, they capture different dimensions of frailty: CRAF is more strongly influenced by functional impairment, whereas I-KCL appears less affected by disability. These instruments should therefore be considered complementary rather than interchangeable.