Impact of Upper-Body Motor Function on Verbal Free Recall

People are generally poor multitaskers, yet do so regularly. Both health and productivity interventions in the workplace (e.g., treadmill desks) have potential for unintended performance consequences, as executing multiple tasks simultaneously, even if simple individually, can be detrimental to task outcomes (Multiple Resource Theory; Wickens, 2008). Even the use of “smart” devices intending to augment performance can cause cognitive processing interference and overload, as the information presented on these devices may divert limited attentional and processing resources away from the primary objective. Understanding how different specific tasks interfere with each other, and ways to diminish this interference, is critical. The purpose of this research is to contribute to the body of knowledge that may improve safety, efficiency, performance, and wellbeing, in real-world multitasking scenarios.

Previous research has explored dual-task memory performance while outdoor running (e.g., Epling et al., 2016, 2018), rock climbing (e.g., Blakely et al., 2021; Epling et al., 2018), swimming (Stets et al., 2020), and even kayaking (Blakely et al., 2022) - the importance being these physically-demanding dual-tasks’ applicability to high-risk job settings (e.g., military, police, firefighting, search, and rescue). Much of this research is based upon the premise that such activities are more cognitively demanding than the indoor, lab-based cyclical exercises (e.g., ergometer cycling, treadmill running) pervading the exercise science literature (Lambourne & Tomporowski, 2010). In other words, navigating real terrain with complex movement likely requires more cognitive resources and thus leads to more interference with secondary cognitive tasks than much of the exercise science literature might lead one to predict (e.g., Tomporowski & Ellis, 1986; Whelan, 1996). Recently, we replicated the dual-task memory-plus-movement research paradigm with a simple indoor cycling task replacing the more complex movement tasks listed above. As expected, cycling dual-task interference was less than that of more operationally-relevant physical tasks. However, despite its simplicity, indoor cycling did cause some interference.

A remaining question has to do with the mechanism(s) of interference, as past results cannot be explained by resource-specific interference alone (e.g., verbal vs. spatial, audio vs. visual; Wickens, 2008). For example, what causes climbing and swimming to be more detrimental to verbal memory tasks than running? A neuro-anatomical link exists between language and gesture (Frick-Horbury & Guttentag, 1998; Wagner et al., 2004; Xu et al., 2009), so it is plausible that climbers and swimmers hand and arm use may impair their ability to fully process a verbal narrative. However, the posture-first hypothesis (Shumway-Cook & Woollacott, 2000) suggests that climbing and running should actually be the more interfering activities, due to evolutionary predisposition to prioritize physical safety (and the risk of falling inherent to these endeavors). The present study will replicate our recent cycling study with a key difference: participants will now “cycle” with their arms using an upper body desk ergometer, rather than their legs, to test the importance of the occupation of a one’s hands when trying to memorize a list of words, while holding general posture, fall risk, stationarity, and movement type constant. Additionally, a high versus low cognitive load manipulation will be used to explore potential interactions between dual-tasking effects and workload.

Aerobically fit and healthy volunteers (N = 26) participated in a within-subjects dual-task paradigm developed around two primary tasks: spinning on an upper-body ergometer, and listening to and freely recalling a list of words. In total, there were five, 5-min conditions that each participant completed in a counterbalanced order with a short break between each: single-task ergometer, single-task low-load word memory, single-task high-load word memory, dual-task ergometer with low-load memory, and dual-task ergometer with high-load memory. To minimize physical fatigue effects, the memory single-tasks always occurred second and fourth in the condition order.

With all ergometer conditions, participants were asked to cover as much (virtual) distance as possible by spinning at a moderate to somewhat hard pace. With all word memory conditions, participants were asked to listen carefully to the narrated words, because they would be asked to freely recall as many as possible at the end of the recording. With the dual-task conditions participants were told to give equal performance priority to both tasks.

The low-load word list recordings had 10 evenly spaced words and high-load had 20 words. There were two of each, balanced for ratings of concreteness, imagery, length, syllables, etc., and they were counterbalanced among the single- and dual-task conditions. The single-task ergometer condition had participants listen to pink noise instead of words.

After each condition, participants typed all words recalled (if applicable) and completed the Subjective Stress State Questionnaire (SSSQ; Blakely et al., 2016). Under dual-task conditions they were also asked to report their task prioritization.

An analysis of variance yielded a significant main effect of single- versus dual-task condition whereby the number of words recalled in the single-tasks was greater than the dual-tasks. A significant interaction revealed that dual-tasking only impaired memory in high-load conditions; the percent of information lost ([1 − #words dual/# words single] × 100) when under low-load conditions (M = 5.21%) was significantly less than high-load (M = 11.85%). Despite a lack of ceiling effect under low-load, participants also remembered more words when under high-load. The high-load conditions were compared to the prior studies by Helton and colleagues (cited above) using the same dual-task paradigm and comparable recall lists. Dual-tasking with the stationary cycling task led to information loss of 7.6%, and over 16%, 25%, and 40% for running, swimming, and climbing tasks respectively.

The difference in distance traveled between the ergometer single-task (M = 1.78 km) and the average distance in the dual-tasks (M = 1.73 km) was marginally significant. Task prioritization ratings revealed that in both the low- and high-word load dual-task conditions, participants prioritized word recall more than the ergometer, counter to instructions for equal prioritization.

The self-report factors of workload, spent (how burnt out or exhausted participants felt), and task focus were calculated from participants’ SSSQ responses. For both the spent and workload factors, the dual-task high-load condition yielded the highest values, the single-task high-load the next highest, and the single-ergometer the lowest. For task focus, all conditions involving word memory were significantly greater than the single-task ergometer.

Our results were consistent with hypotheses that dual-tasking with a simple physical task should contribute to comparatively low interference with a verbal memory task. Interestingly, dual-tasking only impacted word recall under high-load; when workload is low enough, people have the capacity to take on additional work without sacrificing performance. This is consistent with resource theories as well as evidence that reducing workload is not always a good thing. The fact that participants remembered more words when given more words to remember suggests finding an optimal level of challenge is important for fostering maximum performance potential.

Compared to prior studies, word memory loss was less than that caused by more operationally-relevant physical tasks, but more than the stationary cycling task. This suggests that more complex movement exerts greater cognitive demand, and also implicates the use of one’s hands as a contributor to the high interference found in past climbing, swimming, etc. dual-tasks.

Finally, the lack of equal task priority reported in this study implies that past dual-tasking results should be interpreted cautiously, as we cannot trust participants followed prioritization instructions. However, habituation to the cyclic task may have reduced perception of prioritization rather than actual division of effort. Implications of prioritization require further research.

This research will be explored in conjunction with the comparable stationary cycling study in a forthcoming publication. Together, they build upon the growing systematic investigation of key physical task parameters for making cognitive interference predictions, and contribute to improving our theoretical understanding of the cognitive processing requirements of physical movement. This in turn can contribute to the better design of operational task demands, task environments, and technological aid.