No abstract

Common elements governing the responses to a standard programme of exercise ' training ' have been investigated by the etatietical technique of principal component analysis. Convergence of the data wes checked by a pilot trial on a desk calculator and three definitive computer analyses were then carried out. The first and second computer analyses were based on a wide range of messurernents of pulse rate, respiratory minute volume, and personality. Six components accounted for 70 per cent of the variance of the data. These were tentatively identified aa (i) cardiac response to exercise, (ii) resting state, (iii) ventiletory changes with training, (iv) and (v) influence of personality on resting state and exercise response, and (vi) body size. Components (i) and (v) were related to the subjects' initial physical fitness. For the third computer ~~~l y S i S several moasurements shown to be redundant were excluded and specific metabolic measurements were included. Six components then described 77 per cent of the variance. The initial metabolic cost of exercise was represented in components (i) and (ii) and changes with training in component (iv). Fitness wag correleted most closely with the initial oxygen cost of exercise, end was also more closely related to pulse then to ventilatory measurements, and to results on Day 1 than to changes during training. Selection and weighting of parameters to yield en optimum objective assessment of physical fitneas rn discussed.