Functional translation of exercise responses from graded exercise testing to exercise training Carl Foster, Ph.D., Kathy Lemberger, MS., Nancy N. Thompson, MS., Sheila M. Sennett, MS., John Hare, B.S., Michael L. Pollock, Ph.D, Albert E. Pels, III, Ph.D., and Donald H. Schmidt, M.D. MiZwaukee, W&c. Establishment of target heart rate is the most widely employed method for prescribing the relative inten- sity of exercise training in prevention/rehabilitation programs.’ The use of target heart rate has been shown to be effective in defining minimum criteria of exercise intensity necessary to promote the long- term physiologic adaptation commonly referred to as the training effect. 2 Training above the target heart rate has also been shown to relate to the incidence of cardiovascular emergencies in rehabili- tation programs.3 Although the target heart rate is a useful marker of the relative exercise intensity there is little infor- mation available regarding the absolute exercise intensity (pace of ambulation, power output on cycle ergometer) necessary to achieve a prescribed target heart rate. The translation of exercise responses from the incremental non steady-state exercise employed during graded exercise testing (GXT) to the steady-state submaximal exercise usually recom- mended for training is currently made primarily on an empiric basis. Several investigators have shown that because of limited time for accommodation during GXT stages and because of cardiovascular drift during exercise training, the absolute workload required to elicit target heart rate during training is less than during GXT.4,5 However, the specific degree to which the training workload needs to be reduced to achieve the desired heart rate has not been established. The relationship between absolute workload and heart rate is biased by fitness level during both GXT From Human Performance Laboratory, Cardiovascular Disease Section, Department of Medicine, University of Wisconsin Medical School (Mil- waukee Clinical Campus), Mount Sinai Medical Center. Received for publication May 8, 1986; accepted June 16, 1986. Reprint requests: Carl Foster, Ph.D.. Human Performance Laboratory, Mount Sinai Medical Center, 950 N. 12th St., Milwaukee, WI 53233. and exercise training. Fitter individuals must ambu- late at a faster pace to attain target heart rate. They also achieve target heart rate later during a standard GXT protocol. With this bias in mind, we were interested in determining whether the heart rate responses during GXT could be used to provide guidelines for absolute exercise intensity. Accord- ingly, the purpose of the present investigation was to evaluate the relationship between the time (or pow- er output) during GXT when a given target heart rate was observed and the pace of level ground ambulation (or power output on the cycle ergome- ter) during exercise trainng required to achieve the same heart rate. METHODOLOGY This investigation was conducted with two inde- pendent groups. The treadmill group were involved with the translation of exercise responses from GXT on the treadmill to level ground ambulation. The cycle group were involved with the translation of exercise responses from GXT on a cycle ergometer to exercise training using cycle ergometry. Subjects. All 345 subjects in the investigation were participants at our institution in one of the follow- ing: (1) a cardiac rehabilitation program, (2) a preventive exercise program, or (3) experimental studies of exercise training. On the basis of pilot studies, well-trained individuals were not included. Within each group, 90% were randomly selected and were designated the validation group. Exercise responses from the validation group were used to define the relationship between GXT and training. Data from the remaining 10% of subjects (the cross validation group) were used to test the accuracy of the derived equations. Descriptive physical charac- teristics and exercise response for the subjects are presented in Table I. All subjects provided informed consent prior to participation. 1309