Nine subjects, mean age 34.3 years (S.D. 13.9 years; 5 males) completed the protocol. The study had local ethical committee approval. All subjects were judged healthy on the basis of a medical questionnaire. Eight infrared markers were attached to each lower limb, inside of the left and outside of the right, over the main anatomical points. The subject walked on a treadmill, at a normal pace with markers in view of a CODA mpx 30 motion analysis system positioned to the walker’s right. The cognitive task was to respond while walking to a verbal command presented via radio headphones. On the command ‘Red’ the subject was to respond with ‘Yes’ and on ‘Green’ with ‘No’. The Red/Green commands were generated from a sound bank of pre-recorded, digitised words and selected randomly by the control program. Gait was recorded over 20 s epochs with one or other command presented once during each epoch. The timing of a command was triggered via CODA from a marker on the right heel as it elevated in the stride cycle. Forty such epochs overall were recorded for each subject. In the first 20 epochs the subject was required to listen to the commands but not to respond; in the remaining 20 epochs the subject had to respond to the commands.
The difference between the mean and standard error of the amplitude and timing of gait parameters during control and perturbed walking were used to determine significance. Even in a group of healthy and active subjects, on comparing step sequences with and without a verbal response we still found in three of the nine subjects significant perturbations in amplitude and timing of gait parameters, which were synchronized to the command delivery (P < 0.05; unpaired t test).
Although our results cannot directly be linked to the elderly we suggest that a method of this kind, which potentially provides a measure of gait instability under dual tasking pressure, might provide a useful method of risk assessment in vulnerable groups.
We thank the Tana Trust and the Dunhill Medical Trust for their support.