Reliability of graded exercise testing after traumatic brain injury: Submaximal and peak responses

Objective: To assess the reliability of the submaximal and peak responses to a treadmill graded exercise test in individuals recovering from traumatic brain injury. Design: A total of 15 individuals (11 men, 4 women; 28.5 ± 9.2 yrs) with moderate to severe traumatic brain injury admitted into a postacute residential treatment center 10 ± 7 mos after injury performed two treadmill graded exercise tests separated by 4-8 days. Heart rate, oxygen consumption (in milliliters per minute per kilogram), minute ventilation (in liters per minute), and respiratory exchange ratio (carbon dioxide output/oxygen consumption) were continuously monitored at 1-min intervals. Submaximal and peak values were analyzed for absolute level of agreement using the intraclass correlation coefficient. Results: The submaximal intraclass correlation coefficient values for all variables between minutes 3 and 7 ranged from 0.80 to 0.93. Submaximal intraclass correlation coefficients before and in the subsequent minutes were less than optimal for heart rate and minute ventilation. Lower intensity workloads elicited slightly better agreement than higher intensity workloads. The peak response intraclass correlation coefficients ranged from 0.77 (heart rate) to 0.92 (oxygen consumption). Conclusion: Despite the presence of other co-morbidities, subjects recovering from traumatic brain injury provided consistent individual responses, allowing for reliable assessments of cardiorespiratory conditioning programs. Assessments of ambulatory efficiency, endurance, and aerobic fitness adaptations as a result of treatment for persons with traumatic brain injury should include analyses of submaximal responses to graded exercise.