4 Discussion
The present study focused on the effects of acute stress on pilots’ WM. The results showed that the stress group scored significantly higher on the S-AI after the TSST than the control group; compared to baseline, the stress group had elevated cortisol concentrations after the TSST, with no significant differences observed for the control group. This suggests that the modified version of the TSST paradigm used in this study was effective in inducing stress in pilots. The results of the WM test showed that as the WM load increased, the participants’ reaction time increased, and accuracy showed a gradual decrease. This is consistent with the results of most previous studies, where the strain on WM was elevated when the memory load in the N-back task was increased(Itthipuripat et al., 2012). Studies that have used EEG techniques found that the EEG components associated with WM updating were impaired at high-load levels in participants under acute stress(Gärtner et al., 2014), which may help explain the decreased performance on the N-back task.
It was also found that there were no significant differences in reaction times between the stress and control groups for the three memory load conditions. In terms of accuracy results, there were no significant differences between the two groups of pilots with the low/high memory load, but there was a significant difference with the medium memory load, with the accuracy being higher in the stress group than in the control group. In other words, when the pilots were under acute stress, their accuracy was significantly enhanced when they completed the medium memory load task. The effect of acute stress on WM is moderated by WM load. This is partially consistent with the results of previous related studies. For example, Cornelisse (2011) et al. found that acute stress could increase WM levels in a group of male participants under moderate memory load conditions. Specifically, male participants had faster reaction times in the moderate memory load condition after completing the TSST. Although there was no significant increase in accuracy, the performance of the stress group was better than that of the control group. Duncko et al. (2009) similarly found that the stress group had significantly faster reaction times in the high cognitive load condition than the control group. When reviewing the controversial results regarding the effects of acute stress on WM under different memory loads, one researcher found that acute stress impaired WM under high memory load conditions(Schoofs et al., 2008). Some researchers have also found that stress impaired cognitive tasks only when memory load was low(Yu et al., 2015). We speculate that three main causes are responsible for this.
First, acute stress is induced in different ways. Previous studies have found that different methods (psychological, physiological, and pharmacological) can lead to different intensities and durations of acute stress(Li and Ku, 2020), which may affect the response time and accuracy of WM. In this study, we used the modified version of the TSST, which is psychological. The induced acute stress levels were considered moderate compared to the potential levels that could be induced by physical and pharmacological interventions. Pessoa et al. (2009) found that a moderate intensity of emotion enhanced cognitive task performance, while a high intensity of emotion impaired task performance(Pessoa, 2009). Therefore, we hypothesized that moderate acute stress might produce similar results. It is worth noting that the relationship between individual stress levels and job performance follows an inverted U-shaped curve and is also influenced by task difficulty(Broadhurst, 1957). In this study, moderate stress levels led to better job performance when performing moderately difficult tasks. Thus, the stress flight group had higher accuracy WM when completing the moderately difficult task (1-back) than the control group.
Second, the participants were different. Some researchers have found that sex, occupation, and mental state also influence the effects of acute stress on WM(Cornelisse et al., 2011; Gärtner et al., 2014). The participants of this study are pilots whose long-term training helped them to have better WM. In addition, the study used a digital N-back task, where pilots are more sensitive to digital information such as radio frequency, altitude, and speed. Therefore, we assume that the pilots in this study had a different difficulty level for the memory load task than other subjects. This may explain the inconsistency with previous studies.
Finally, the present study explored the behavioral results of WM. However, some studies have found that there were no significant differences in behavioral results, but there were significant differences found in neurological results(Han et al., 2013; Yu et al., 2015). For example, in a study by Zhang et al. (2015) examining whether the effects of acute stress on WM were moderated by WM load, no interaction was found between the group and WM load on reaction time. However, EEG results revealed that the P3 component induced in the acute stress group was found to be significantly greater than that of the control group in the low WM load condition. We speculate that the behavioral results may not yet reflect differences in WM because brain neurological techniques such as EEG and MRI were not used in this study.
We acknowledge that our sample selection has some limitations, as it does not represent all pilots or other occupational groups, and lacks diversity in terms of gender, age, education, etc. This may affect the generalizability and universality of our research findings. Future studies can consider increasing the number of female or other background participants to improve the quality and validity of the research.
In summary, we found that acute stress enhances the accuracy of WM for pilots under moderate memory load. According to the Yerkes-Dodson law, the effect of stress on hippocampal and prefrontal cortex-dependent memory follows an inverted U-shaped curve(Broadhurst, 1957). According to the Yerkes-Dodson law, moderate levels of stress promote memory function, while high levels of stress impair memory function(Luethi and Mathias, 2008). After the flight stress group completed the TSST task, moderate stress levels instead promoted them to perform better on moderately difficult WM tasks. The ability of pilots to correctly recall SOPs and operate them accurately after an unexpected event is particularly important for flight safety.