Do two measures of frailty identify the same people? An age‐gender comparison. Issue 3 (18th August 2019)

• Record Type: Journal Article
• Authors: Gordon, Susan; Grimmer, Karen; Baker, Nicky
• Is Part Of: Journal of evaluation in clinical practice. Volume 26:Issue 3(2020)
• Page Start: 879
• Page End: 888
• Publication Date: 2019-08-18
• Subjects: aged -- Clinical Frailty Scale -- correlation -- frailty -- Fried frailty phenotype -- middle age; Clinical medicine -- Periodicals; 616.005
• DOI: 10.1111/jep.13265
• Languages: English
• ISSNs: 1356-1294

Abstract

Abstract: Rationale, aims, and objectives: Most frailty assessments have been developed for people aged over 65 years. However, there is growing evidence that frailty is detectable in younger people. This paper tests the hypothesis that the Fried frailty phenotype and the CFS categories identify the same people in age‐gender subgroups in community‐dwelling 40 to 75‐year‐olds. Method: Participants were recruited via comprehensive community‐sampling strategies. They self‐reported frailty using the Clinical Frailty Scale (CFS), and frailty was also estimated using the Fried phenotype (self‐reported unintended weight loss, exhaustion and low regular exercise; observed slow gait speed and poor grip strength). CFS and Fried scores were compared overall, and for age‐gender subgroups (40‐49 years, 50‐59 years, 60‐69 years, and 70‐75 years). Spearman rho and differences in mean integer Fried scores were calculated across CFS categories using ANOVA. Correlations were determined between Fried categories of not‐frail, pre‐frail, and frail and ranked CFS categories, using ranked scores (tau‐c) and Cochran‐Mantel‐Haenszel (C‐M‐H) tests. Results: Of 656 participants (67% female; mean age 59.9 years, SD 10.6), Fried phenotype classified 59.2% not frail, 39.0% pre‐frail, and 1.8% frail, with no gender or age differences. CFS data were missing for 25 participants, with N = 631 reporting categories of very well (24.6%), well (44.6%), managing well (21.9%), vulnerable (6.3%), mildly frail (0.5%), and moderately frail (0.2%). Overall, the mean Fried frailty scores increased incrementally and significantly across ranked CFS categories ( P < .01), with weak linear correlation (rho = 0.09). There were variable correlations in age‐gender groups, with the best correlation found for women aged 50 years or older, and men aged 60 to 69 years. Conclusion: Frailty assessments using the two assessments became more consistent, as age increased. Pre‐frailty was identified by both assessments in all age‐gender groups. The validity of self‐reported CFS, and of pre‐frailty criteria relevant to people younger than 65 years, needs investigation.