Measuring physical strain during ambulation with acceleration

Purpose: To study the feasibility of ambulatory accelerometry in the evaluation of physical strain in walking at different speeds and different levels of economy. Methods: Twelve able-bodied subjects performed a walking test on a treadmill with increasing walking speed. After a 6-wk period, these measurements were repeated, with an additional test of perturbed walking. Motility (the intensity of body segment movements, measured with accelerometry), heart rate, and oxygen uptake were simultaneously recorded. Feasibility was determined by comparing motility with percentage heart rate reserve (%HRR), using oxygen uptake (VO₂) as reference. The relation with oxygen uptake, the sensitivity to change when walking speed increased and gait was perturbed, and intrasubject and intersubject variability were studied. Results: Walking with increasing walking speed resulted in a higher pooled r² value (0.91) and a lower residual standard deviation (1.44 mL· kg^-1·min^-1) for the motility-VO₂ relation than for the %HRR-VO₂ relation (0.84 and 1.92 mL·kg^-1·min^-1, respectively). Furthermore, the motility-VO₂ relation for this part of the protocol showed lowest interindividual differences. The sensitivity to changes due to an increase in walking speed, and the test-retest reliability were the highest for motility. Walking with a brace resulted in a lower pooled r² value for the motility-VO₂ relation than for the %HRR-VO₂ relation (0.31 and 0.67, respectively); the sensitivity to changes due to walking with a brace was lower for motility than for %HRR. Conclusion: Compared with heart rate, accelerometry allows accurate measurement with little error in any able-bodied individual, without the need for individual calibration. The response of motility to changes and differences in economy in patients requires further study.