Toward Personalized Exercise Medicine: A Cautionary Tale

Purpose This study aimed to examine individual exercise response rates across a range of cardiometabolic variables, cardiorespiratory fitness, and body composition in adults. Methods A retrospective analysis of data from three randomized controlled trials was used in this study. Participants include those who completed the given trial (control, n = 87; intervention, n = 251). Anthropometric (weight, body mass index, waist circumference), cardiorespiratory fitness (V?O-2peak), MRI-measured total adipose tissue (AT), abdominal subcutaneous AT, and visceral AT and common cardiometabolic variables were assessed pre- and postintervention using standard methodologies. The technical error (TE), which includes both the day-to-day variability and instrument error, was calculated using pre- and postintervention data from the time-matched control group. Results On average, all anthropometric, MRI, and V?O-2peak variables improved significantly after intervention compared with the control group (P < 0.05). With the exception of glucose disposal rate (37%), after intervention less than 13% of participants improved cardiometabolic outcome measures beyond the day-to-day variability of measurement. In other words, the individual response for 63%-96% of participants fell within the uncertain range (2 TE). Similarly, for absolute V?O-2peak (L center dot min(-1)), only 45% of participants improved beyond 2 TE. By comparison, for MRI-derived variables, the majority of participants (77%, 58%, and 51% for total AT, abdominal subcutaneous AT, and visceral AT, respectively) improved beyond 2 TE. The observed reductions beyond 2 TE for WC and body weight were 53% and 63%, respectively. Conclusions The findings suggest extreme caution when inferring that the cardiometabolic and cardiorespiratory fitness response for a given individual is attributable to the exercise dose prescribed.

Automated summary

This study used data from three randomized controlled trials to examine individual exercise response rates in cardiometabolic variables, cardiorespiratory fitness, and body composition. The findings suggest that the majority of participants' response falls within the uncertain range, except for MRI-derived variables which showed improvement beyond the uncertainty range.