Figure 1
Performance decrements in the cognitive task due to resource limitations. A continuous motor task (in our example: walking) continuously draws on limited resources throughout the entire movement cycle (blue rectangle). If an intermittent cognitive task is triggered by a stimulus at a given moment in that cycle, both tasks have to share the limited amount of processing resources (vertical axis). For simplicity, we restrict this illustration here to the case that motor processes claim a constant amount of resources and that only the cognitive task is affected by this limitation. Yet, we will loosen these restraints later. Optimal processing of the cognitive task would require a given processing capacity (amount × duration), here represented by the area A + B. However, due to the overall limitation and the parallel demand of the motor task, only the amount A is currently available. If now, the task needs to be as fast as under typical single-task (ST) conditions, quality of task achievement will drop due to the loss of B. Alternatively, the duration of task processing might be prolonged, allowing temporally extended use of resources (duration of C) to compensate for the shortfall of B. Typically, this temporal delay is taken as measure for the amount of task interference in the latter case.
Figure 2
Fixed cycle-linked alteration of processing requirements of the motor task. The motor tasks requires different amounts of processing resources at different moments in the cycle. The cognitive task is presented at different temporal locations in that cycle (stimulus onset at 0%, resp. 60%). Dependent on the amount of unmet requirements (B), it takes longer (larger horizontal extension of compensatory process supply C in the rightward [60%] location) to complete the cognitive task. Differences in response delays are thus informative about changes in processing requirement of the motor task across the motor cycle.
Figure 3
Triggered interruption of motor task processing. An unexpectedly occurring cognitive secondary task interrupts processing of the motor task. The shift of resource allocation cannot be effected immediately. Dependent on the amount of unmet requirements (B), accumulated during that shifting period, it takes longer to complete the cognitive task. Response delays (horizontal extension of C) inform about the quickness of the shifting in resource allocation. Whether motor performance is also affected depends on whether the interrupted motor processes can easily be shifted to a different location (E).
Figure 4
Scheduled temporal interleaving of processes. The secondary task has to be processed at a predictable temporal position in the motor cycle. In order to free sufficient capacity (A) to execute the cognitive task with performance loss, part of the motor processing is shifted in time. This could either be before (D) or after (E) the period of cognitive task processing.
| A1- | interference is only visible in the cognitive task, |
| A2- | interference scales with the overall difficulty (e.g. speed) of the motor task, |
| A3- | the magnitude of interference might be different for different temporal locations. |
| B1- | interference may be observed in both tasks, cognitive and motor, |
| B2- | interference scales with the overall difficulty (e.g. speed) of the motor task, |
| B3- | the magnitude of interference might be different for different temporal locations. |
| C1- | interference might be reduced or even completely absent in case of an obvious temporal regularity, however… |
| C2- | becomes larger, when the regularity is removed. |
| C3- | No clear predictions can be made on the dependency on changes in difficulty of the motor task within (different moments in the cycle) and across motor tasks (e.g. different speeds). |
Table 1
Stimulus frequencies for different treadmill speeds.
| Speed [km h–1] | Range [Hz] | Mean [Hz] | Standard deviation [Hz] |
|---|---|---|---|
| 3 | 0.339–0.446 | 0.397 | 0.025 |
| 4 | 0.388–0.482 | 0.445 | 0.023 |
| 5 | 0.436–0.507 | 0.478 | 0.020 |
Figure 5
Experimental procedure. Top panel shows the sequence of blocks within a testing day. Each day started with an ST-Walking test to measure subject’s preferred cadence for the given speed. The following sequence of seven test blocks starts and ends with a Sitting block. In between subjects perform five Walking blocks for the different SORs. Sequence of SORs is counterbalanced across subjects. Each block consists of three ST and three DT trials in alternation. Boxes below depict the timeline of an ST trial (left) and a DT trial (right).
Figure 6
Exemplary data of a single subject. The upper row shows reaction times recorded over time and the calculated regression function (black dotted line). Background colors indicate the Stimulus Onset Regimes decoded in the according boxes in the lower row. Bars in the lower row represent distance of median to predicted mean performance.
Figure 7
Relative reaction times in the Sitting condition. Means and standard errors including 95% confidence interval are presented separately for the three testing days.
Figure 8
Result overview of cognitive and motor performance for different Stimulus Onset Regimes and Walking Speeds. Cognitive performance is depicted in the upper two rows. The first row represents the residual reaction times (RTresidual). The second row shows the accuracy measure for 2-back matchings (ACCmatch). The third row represents the difference between dual-task and single-task [Δ(DT-ST)] performance in step duration. Negative values indicate shorter durations under dual-task condition. Mean values of dependent variables across conditions. Error lines represent standard error including 95% confidence interval.
Figure 9
Profile of relative RTs for different SOR conditions. RTresidual values are integrated across walking speeds and z-transformed.
Table 2
Results of ANOVA-tests for walking parameters. SD = stride duration, SL = stride length, SDCV = coefficient of variation of SD, SLCV = coefficient of variation of SL, Δ = single minus dual-task performance in the dependent parameter, SOR = Stimulus Onset Regime.
| Dual-task… | Parameter | Factor | F | η2 | p |
|---|---|---|---|---|---|
| …walking pattern | SD | Walking Speed | 332.797 | 0.679 | <0.001 |
| SOR | 0.575 | <0.01 | 0.618 | ||
| Walking Speed × SOR | 1.404 | <0.01 | 0.240 | ||
| SDCV | Walking Speed | 21.766 | 0.267 | <0.001 | |
| SOR | 0.477 | <0.01 | 0.711 | ||
| Walking Speed × SOR | 0.518 | <0.01 | 0.721 | ||
| SL | Walking Speed | 325.299 | 0.537 | <0.001 | |
| SOR | 1.352 | <0.01 | 0.274 | ||
| Walking Speed × SOR | 1.877 | <0.01 | 0.122 | ||
| SLCV | Walking Speed | 38.420 | 0.404 | <0.001 | |
| SOR | 2.672 | 0.015 | 0.099 | ||
| Walking Speed × SOR | 1.683 | 0.018 | 0.207 | ||
| …costs | ΔSD | Walking Speed | 18.121 | 0.194 | <0.001 |
| SOR | 0.928 | 0.011 | 0.423 | ||
| Walking Speed × SOR | 0.831 | 0.015 | 0.458 | ||
| ΔSDCV | Walking Speed | 5.203 | 0.117 | 0.036 | |
| SOR | 0.062 | <0.01 | 0.978 | ||
| Walking Speed × SOR | 0.835 | 0.013 | 0.514 | ||
| ΔSL | Walking Speed | 8.184 | 0.105 | 0.002 | |
| SOR | 1.131 | 0.009 | 0.351 | ||
| Walking Speed × SOR | 0.340 | 0.006 | 0.864 | ||
| ΔSLCV | Walking Speed | 0.591 | 0.010 | 0.561 | |
| SOR | 0.729 | 0.014 | 0.467 | ||
| Walking Speed × SOR | 0.376 | 0.011 | 0.748 |