This study is an attempt to quantitatively combine the effects of the following factors to allow the prediction of a given person’s metabolic energy expenditure :​ (1) the magnitude of external and inertial forces on the arm(s), (2) the position of the arm(s)B and (3) the average speed at which a particular arm movement is performsdo The empirical development is based on the concept that these factors can be characterized by a single dimension, which is the resultant torques at the shoulder(s), elbow(s)e or wrist(s).

Measured metabolic energy expenditure rates of four subjects performing static arm tasks are reportedo These disclose that the rate of metabolic energy expenditure by a person for a given level of torque at the shoulder, elbow, or wrist is also dependent on the following;​ (1) the torque direction (ioeo clockwise or counterclockwise rotation at the shoulder), (2) the angle at the elbow and shoulder, (3) the torque at adjacent articulations, and (4) whether one or both arms are active * The effects of each of these factors are combined into a whole-arm metabolic prediction model for static activities0 Validation of the whole-arm prediction model (accomplished by having the subjects hold various amounts of weight in their hands in different positions with respect to the body), is presented. The validation disclosed that if each subject's metabolic energy expenditure rates for given shoulder, elbow, and wrist torques were ascertained first and used in the model, the resulting predictions accounted for between 85 and 97 percent of the variance in the measured metabolic energy expenditure rates during whole-arm activities.

An attempt is made to extend the metabolic prediction model for static arm tasks to also apply to dynamic arm activities * One subject's metabolic energy expenditure rates during elbow flexions at different torques and angular velocities are reported * A wholearm prediction model for dynamic tasks is developed from these results * Validation of the dynamic model, which involved having the subject perform weight lifting activities, revealed that the metabolic predictions were significantly lower than the measured energy expenditure rates during the slow lifts * A hypothesized rationale for this result is discussed»