Background: This study sought to better understand the psychometric properties of the SARC-F, by examining the baseline and training-related relationships between the five SARC-F items and objective measures of muscle function. Each of the five items of the SARC-F are scored from 0 to 2, with total score of four or more indicative of likely sarcopenia.
Methods: This manuscript describes a sub-study of a larger step-wedge, randomised controlled 24-week progressive resistance and balance training (PRBT) program trial for Australian community dwelling older adults accessing government supported aged care. Muscle function was assessed using handgrip strength, isometric knee extension, 5-time repeated chair stand and walking speed over 4 m. Associations within and between SARC-F categories and muscle function were assessed using multiple correspondence analysis (MCA) and multinomial regression, respectively.
Results: Significant associations were identified at baseline between SARC-F total score and measures of lower-body muscle function (r = - 0.62 to 0.57; p ≤ 0.002) in 245 older adults. MCA analysis indicated the first three dimensions of the SARC-F data explained 48.5% of the cumulative variance. The initial dimension represented overall sarcopenia diagnosis, Dimension 2 the ability to displace the body vertically, and Dimension 3 walking ability and falls status. The majority of the 168 older adults who completed the PRBT program reported no change in their SARC-F diagnosis or individual item scores (56.5-79.2%). However, significant associations were obtained between training-related changes in SARC-F total and item scores and changes in walking speed and chair stand test performance (r = - 0.30 to 0.33; p < 0.001 and relative risk ratio = 0.40-2.24; p < 0.05, respectively). MCA analysis of the change score data indicated that the first two dimensions explained 32.2% of the cumulative variance, with these dimensions representing whether a change occurred and the direction of change, respectively.
Discussion: The results advance our comprehension of the psychometric properties on the SARC-F, particularly its potential use in assessing changes in muscle function. Older adults' perception of their baseline and training-related changes in their function, as self-reported by the SARC-F, closely matched objectively measured muscle function tests. This is important as there may be a lack of concordance between self-reported and clinician-measured assessments of older adults' muscle function. However, the SARC-F has a relative lack of sensitivity to detecting training-related changes, even over a period of 24 weeks.
Conclusions: Results of this study may provide clinicians and researchers a greater understanding of how they may use the SARC-F and its potential limitations. Future studies may wish to further examine the SARC-F's sensitivity of change, perhaps by adding a few additional items or an additional category of performance to each item.