Placing a checkmark on the to-do list is an extremely liberating feeling for many eager list lovers, especially when the task has been postponed for a long time.
But what happens in our brain when we have completed a postponed task? Will it be deactivated?
If so, how?
A team of scientists from the Collaborative Research Centre 940 “Volition and Cognitive Control” at TU Dresden, together with two leading international experts, Julie Bugg and Michael Scullin, investigated these questions in a systematic review article.
Headed by Dr. Marcus Möschl from the Chair of General Psychology at TU Dresden, the team analysed 20 years of research on intention deactivation and so-called aftereffects of completed intentions across different research fields.
There are many everyday examples of postponed intentions: children who put off tidying up their room until later, people planning to call their best friend tomorrow rather than now, and so on.
Intentions have been proven to influence our thoughts and actions until they are completed. Afterwards, they could be deactivated and removed from our mental to-do list.
“To our surprise, however, the studies we reviewed have repeatedly shown that completed intentions are sometimes not deactivated immediately, but continue to affect people.
For instance, when implementing new intentions,” explains Marcus Möschl.
In particular, this happens, when an action has been postponed until a certain salient event or stimulus function as a reminder.
Elderly or sick people, for instance, may postpone taking medication until they receive a calendar notification on their mobile.
If this alarm accidently rings again after having taken the medication, they may not only recall their intention, in drastic cases, they even might take their medication again.
There are many everyday examples of postponed intentions: children who put off tidying up their room until later, people planning to call their best friend tomorrow rather than now, and so on.
However, as the studies suggest, such drastic cases of aftereffects are rather rare. “Often, intentions seem to be deactivated as soon as they are completed,” continues Dr. Möschl. “However, this deactivation does not always work perfectly like switching the light on and off.
In some cases, connections between a stimulus and a completed action have to be dissolved step by step until the event or stimulus no longer trigger the retrieval of the completed intention.”
Now the scientists want to investigate in which situations it could be advantageous if completed intentions remain activated and whether these aftereffects play a role in learning new routines
One core component of executive control is the ability to schedule and sustain multiple goals at the same time or in rapid alternation.
It is now well established that this ability is trainable by specific interventions, such as task switching or multi-tasking training, which may be particularly relevant for older adults that show genuine impairments in executive behavior (Karbach and Kray, 2009; Anguera et al., 2013; Karbach and Verhaeghen, 2014).
Recent neuroimaging studies suggest fundamental changes at the neural level underlying this age-related decline in executive control. Specifically, research revealed age-associated brain activation changes on transient (i.e., brief, trial-related) and sustained (i.e., enduring, block-related) timescales in distributed cortical and sub-cortical networks (Dennis et al., 2007; Jimura and Braver, 2009), such as in the prefrontal cortex (e.g., frontal pole: Brodmann area (BA) 10; dorsolateral prefrontal cortex (PFC): BA 9, 46; ventrolateral PFC: BA 44, 45; inferior frontal junction: intersection of BA 6,8,9, and 44; anterior cingulate: BA 6, 8, and 32) as well as in parietal circuits (inferior parietal lobule: BA 39,40; superior parietal lobule: BA 5, 7) and in subcortical areas in the basal ganglia (BG), including the putamen and the caudate head (see Dosenbach et al., 2006; Dosenbach et al., 2007, 2008; Gold et al., 2010; Gazes et al., 2012; Kim et al., 2012; Nee et al., 2013). Hence, interventions aiming at improving task switching behavior in older adults, such as the present one, should be evaluated by their impact on the neural dynamics underlying performance changes.
In our recent study (Dörrenbächer et al., 2017a,b), we already examined such neural plasticity in brain activity dynamics after task-switching training in older adults during the task-switching task.
Specifically, we applied a hybrid event-related-/block fMRI design (Visscher et al., 2003) to investigate not only spatial but also temporal dynamics of neural mechanisms underlying changes after training in task switching compared to a single-task active control regime in older adults. After task-switching training, we found selective changes in the task-switching training task
(a) for trial-related brain activation in fronto-lateral (i.e., bilateral ventrolateral PFC and inferior frontal junction) and parietal regions (i.e., left superior parietal lobule), but
(b) for block-related brain activation in the basal ganglia bilaterally (see section “Materials and Methods” for details).
Hence, our results revealed spatio-temporal interactions in training-induced neural changes.
These spatially dissociable changes of trial- versus block-related brain activation were both related to improvement of task switching behavior. However, considering the behavioral level alone, training group differences were altogether small.
Hence, a second important insight of that study was that neural mechanisms might reveal more subtle training-induced effects that we may not become aware of from only examining the behavioral data.
Based on these findings, the present study aimed to explore further the nature and scope of such neural plasticity beyond behavioral changes in untrained task paradigms tapping into the updating of task sets and the inhibition of irrelevant task contents.
Empirical evidence for transfer of functional brain changes after executive-control training in older adults is scarce so far. Heinzel et al. (2016) found a change in the spatial distribution of brain activation across canonical control networks in older adults.
In that study, older adults practiced working-memory updating on an adaptive n-back training task. The authors revealed a selective decrease of neural activation in the right caudal superior frontal sulcus for both the trained n-back task as well as for an untrained Sternberg task for the older adult training group as compared to a no-contact control group.
However, one caveat of that study is that the researchers only included passive controls, so that unspecific effects associated with the training regime might have influenced changes in the treatment group. Therefore, in the present study, we included a control group that performed the same task as the treatment group but practiced the task in single-task conditions putting lower demands on executive control than mixed-task conditions (see also Kray and Dörrenbächer, 2019).
Moreover, Heinzel et al. (2016) investigated neural transfer in older adults only as a function of spatial overlap across cognitive-control networks. Indeed, neural changes in untrained tasks are usually assumed to occur if activation changes associated with the trained and untrained tasks rely on the same cognitive processes and on spatially overlapping brain regions (Dahlin et al., 2008). However, it has not yet been discussed whether the amount of neural changes in untrained tasks may also depend on overlapping transient or sustained timescales of the involved neural processes, hence on temporal overlap.
To summarize, the aims of the present study were twofold: first, to investigate whether task-switching training as compared to active-control single-task training may induce neural changes above and beyond behavioral changes in untrained cognitive tasks by considering both the spatial location and the temporal dynamics of brain-activation changes; and second, to investigate the spatial and temporal overlap of these changes with changes in the training task (Dörrenbächer et al., 2017a,b).
We administered two task paradigms different from the training task: one was sensitive to transient dynamics (i.e., a context-updating task, adapted from Schmitt et al., 2014) and the other was sensitive to sustained dynamics of task switching behavior (a delayed-recognition working-memory task, adapted from Clapp et al., 2010).
The context-updating task requires rapidly switching and updating the current task from trial to trial (see also Lenartowicz et al., 2010). Hence, this task taps into local, transient processing requirements on executive control (cf. Marí-Beffa and Kirkham, 2014) and, therefore, was modeled with an event-related fMRI design, capturing transient brain activation dynamics.
The latter delayed-recognition working-memory task requires the ability to sustain multiple task-set representations throughout a certain delay period (i.e., sustained task-set maintenance) and to select between them (i.e., sustained scheduling at the task-set level; see section “Materials and Methods” for details). Hence, this task may tap into global, sustained processing requirements on executive control (cf. Marí-Beffa and Kirkham, 2014), thus was implemented into a block-related fMRI design capturing sustained brain activation dynamics.
Regarding our first study goal, we were specifically interested in selective neural changes in the magnitude of transient activation during the context-updating task or of sustained activation during the delayed-recognition task after task-switching training compared to single-task control training.
The neural signature of optimal task processing is still a matter of debate. One candidate approach is that of neural efficiency (e.g., Neubauer and Fink, 2009), proposing reductions in the magnitude of brain activation (i.e., task energy) consumed in performing a given task associated with performance improvement.
In contrast, the cortical-effort approach suggests increased brain activation as a signature of an increased capability to recruit task-relevant neural resources (for a review, see Barulli and Stern, 2013). Given the dissent in predictions of directional changes of brain activation associated with behavioral improvement, the present study focused on task-beneficial changes in the magnitude of brain activation after training, while the direction of change remained an open question.
Regarding our second study goal, to systematically investigate spatial and temporal overlap of changes in activation magnitude during the trained and untrained tasks (cf. Dörrenbächer et al., 2017a,b), we tested one of three likely outcomes based on a region-of-interest (ROI) analysis approach (these potential outcome patterns should be considered as exploratory).
One likely outcome would suggest only spatial overlap (i.e., same ROIs but different timescale), hence an overlap in ROIs that had proven sensitive to the other timescale in the task-switching training task.
Given this overlap criterion, the transient-sensitive context-updating task would show changes in those ROIs that had been sensitive to changes in sustained activation in the task-switching training task (i.e., in basal ganglia; and vice versa for the sustained-sensitive delayed-recognition task: in ventrolateral PFC, inferior frontal junction, superior parietal lobule, cf. Dörrenbächer et al., 2017a,b).
Another likely outcome would suggest only temporal overlap (i.e., same timescale but different ROIs), hence an overlap in changes on sustained or transient activation in canonical control regions outside those obtained for the task-switching training task (e.g., frontal pole, dorsolateral PFC, anterior cingulate, inferior parietal lobule, cf. Dörrenbächer et al., 2017a,b).
A third likely outcome would suggest spatio-temporal overlap (i.e., same ROI and same timescale), hence an exact reproduction of patterns found for the task-switching training task.
Given this overlap criterion, the transient-sensitive context-updating task would show changes in exactly those regions that had also been sensitive to changes in transient activation in the task-switching training task (i.e., in ventrolateral PFC, inferior frontal junction, superior parietal lobule; and vice versa for the sustained-sensitive delayed-recognition task: in basal ganglia, cf. Dörrenbächer et al., 2017a,b).
Source:
TU Dresden
Media Contacts:
Dr. Marcus Möschl – TU Dresden
Original Research: Open access
“Aftereffects and deactivation of completed prospective memory intentions: A systematic review”. Möschl, Marcus, Fischer, Rico, Bugg, Julie M., Scullin, Michael K., Goschke, Thomas,Walser, Moritz.
Psychological Bulletin doi:10.1037/bul0000221.
