Article Title



Anthony Campitelli1, Cody Diehl1, Megan Jones1, Josh Gills2, Ray Urbina1, Kelsey Bryk3, Jordan Glenn3 and Michelle Gray1

1University of Arkansas, Fayetteville, AR

2Rutgers University, New Brunswick, NJ

3Neurotrack Technologies, Inc.

Functional impairment (FI) is responsible for a massive increase in socioeconomic costs and reduction in quality of life among older adults. Impaired walking ability (WA) is an important component of FI. To better predict FI, researchers have focused on investigating the role of manifest biomotor variables in their development of predictive models. This, however, has come at the expense of measuring latent physiological constructs which may predict FI better than manifest measures. One such latent variable is whole-body muscle quality (MQ) which takes into account how well the muscles in the body can produce force and power but cannot be measured directly. Purpose: The purpose of this investigation was to produce a structural equation model to measure MQ and determine its ability to predict WA. Methods: Adults age 45-75 years (n = 231) were evaluated in this study. All participants completed a 5-times sit-to-stand task (5xSTS), a maximal handgrip dynamometry test (HG), sit-to-stand maximal muscular power task (STSP), a 6-minute walking task (6MinW), a 4-meter fast walking task (4MW), and a 10-meter habitual walking task (10MW). Confirmatory factor analysis (CFA) was used to produce a measurement model wherein MQ was dimensionally reduced from 5xSTS, HG, and STSP, and WA from 6MinW, 4MW, and 10MW. If a good model was produced, a structural equation model was fit which regressed MQ on WA and included age and sex as potential exogenous contributing variables. Results: The CFA measurement model for MQ and WA met all model assumptions and demonstrated good model fit. After removing sex from the subsequent structural model, the resulting model demonstrated adequate absolute fit (χ2(13) = 21.08, p = .07, χ2/df = 1.62, and SRMR = 0.065) and excellent incremental fit (CFI = .98 and TLI = .97). The final structural model demonstrated that MQ is a good predictor of WA (β = .66), displaying better predictive ability than age (β = .41). This model also illustrates how MQ mediates the effect of age on WA. Conclusion: MQ is a good predictor of WA. Latent biomotor measures should not be discounted in the assessment and prediction of FI.

This document is currently not available here.