Estimating MET values using the ratio of heart rate for persons with paraplegia.
Metabolic equivalent (MET) values are levels assigned to a physical activity that measures its intensity and energy expenditure. The compendium of physical activity (CPA) is a listing of more than 600 physical activities and their corresponding MET value. However, because these values were obtained from the general population, they may not be applicable to people with disabilities. Furthermore, the list does not include any disability-specific activities.
It is thought that heart-rate variability should be a good predictor of a MET level for any given physical activity. The error rate in using heart rate as a predictor can be minimized by using individual calibration and cross-validation. This study sought to determine if the heart-rate ratio is an accurate predictor of MET values, to examine how individual calibration can reduce error, and to explore the use of cross-validation for prediction.
A total of 27 participants with paraplegia participated in this study, 12 of whom had a complete lesion and the remaining 15 had an incomplete lesion. The level of the spinal cord injury ranged from T6 to L4, and the age of injury ranged from 6-45 years.
Resting metabolic rate (RMR) was measured for all participants. The participants then rested quietly by lying down for 30 minutes; then, data was collected for an additional 30 minutes for RMR calculations.
Individuals also participated in 10 various physical activities, such as working on a computer, vacuuming, pushing a wheelchair- on a tile floor, on a side walk, and up and down a ramp. Four of the ten activities were considered low-intensity. Each activity lasted 10 minutes, and oxygen consumption during activity was measured. Heart rate was also measured during rest and during activities. Body composition was also measured as could have an effect on RMR.
The correlation coefficient between heart rate and MET level was .77. The correlation between observed and predicted MET values was also high, r = .78.
The study showed that the lower-intensity physical activities showed a lower relationship (r = .53) between expected and predicted energy expenditure than the higher-intensity activities (r = .93). There were also higher error rates in the lower-intensity activities measured with both the group calibration and individual calibration when compared to the higher-intensity activities. However, using individual calibration did reduce the prediction error significantly. There was also a strong relationship between heart rate and the observed MET values. Heart-rate ratio was a good predictor of estimated MET values.