A Comparison of Estimation Models of Physical Fitness Age for Elderly People using the Japan Fitness Test

Abstract

This study examines the validity for estimating physical fitness age (PFA) using the Japan Fitness Test through the application of single and multiple regression analyses and principal component analysis. The participants were 484 community-dwelling Japanese elderly men and women. A Physical fitness test consisting of 6 performance tests, namely grip strength, sitting trunk flexion, foot balance with opened eye, 10-m hurdle walk, sit-ups, and 6-min walk were used. Participants were divided into two groups for modeling PFA equations (n=322) and for cross-validation of them (n=162). The equation models for estimating the PFA were a single regression model with a physical fitness test score, a multiple regression model with the 6 performance tests, and the first principal component model obtained from the 6 performance tests. The validity of three PFA models was compared by concurrent validity with chronological age (CHA), cross-validity using shrinkage, and degree of agreement between aging effect for physical fitness and the correlation coefficients of PFA and CHA. Results of statistical analyses reveal that the three PFA models satisfy concurrent validity and cross-validity, but the single regression PFA model detected a greater degree of agreement for aging than the other models.