dorsolateral prefrontal cortex role in decision-making

Posted comment on ´Stimulation of dorsolateral prefrontal cortex enhances adaptive cognitive control: a high-definition transcranial direct current stimulation study` by O. Gbadeyan, K. McMahon, M. Steinhauser and M. Meinzer and published in Journal of Neuroscience 2016 36 (50) p.12530; doi.org/10.1523/JNEUROSCI.2450-16.2016

SUMMARY

Gbadeyan and colleagues investigated in their experiments the role of the dorsolateral prefrontal cortex (DLPFC) in the adjustment of strategic control to real-time conflict caused by previously experiences of conflict. The authors support conflict monitoring theory where the anterior cingulate cortex (ACC) adjusts attentional resources to goal-directed information and ignores irrelevant information – a process believed to involve the lateral prefrontal cortex (LPFC). Their research centred on the role of the dorsolateral prefrontal cortex in strategic control in particular since most research using human subjects was said to concentrate on the activity of the ACC and any findings on the role of the DLPFC appears to be contradictory, eg. different views on the contributions made by the two hemispheres.

In their experiments, Gbadeyan and colleagues used 120 healthy participants and performed high-definition transcranial direct current stimulation (HD-tDCS) under 4 conditions – left or right DLPFC, or left or right primary motor cortex (M1). The M1 area was used to assess the specificity of the HD-tDCS performed since the area is thought to play no part in the response to conflict. Active and sham HD-tDCS were performed on each group in crossover and double blind tests. The adaptation to conflict was assessed during a visual flanker test ie. assessment of modulation of the flanker effect as a function of previous response to conflict. The participants were required to respond to a centrally presented target (a centre row of arrows) for 100ms after being presented with first a fixation cross (300ms) and then the distracting material. This was 4 flanker arrows which were presented (100ms) as either pointing in the same direction as the target stimulus (congruent), or in the opposite direction (incongruent). The participants were required to ignore these flanker arrows and to respond by indicating the direction of the target row arrows by pressing either the left or right response key. Each subject performed 5 blocks of trials.  The HD-tDCS used was a concentric set up with a 20 min long administration before ramping down. Nine potential side effects were recorded by all participants. Data analysis involved the mean response times (RTs) and varying ANOVA. Adverse effects between the active and sham trials were compared using paired t tests.

The experiments were performed to measure the conflict adaptation effect. The authors expected that there would be a flanker effect because responses are slower for incongruent trials than congruent. The responses were also expected to be slower due to the modulation of the response from the directly preceding trials and this is the so-called conflict adaptation effect to anticipated levels of conflict. On analysis of the mean RTs results of their experiments, Gbadeyan and colleagues found that there was a sizeable conflict adaptation effect in all groups and in all conditions (greater than 30%). ANOVA showed a significant 2 way interaction between current congruency and previous congruency indicating a strong conflict adaptation effect. There were also significant results for an ANOVA 4 way interaction between region, stimulation, current congruency and previous congruency. Separate ANOVAs between stimulation, current congruency and previous congruency found the largest conflict adaptation effect in the active condition than in the sham group, but the effect was not further modulated by laterality ie. both DLPFC groups had the same influence. With the M1 directed HD-tDCS then the conflict adaptation effect was reflected by significant interaction between current congruency and previous congruency, but was not further influenced by stimulation.

Gbadeyan and colleagues reanalysed the DLPFC with additional variable response repetition (repetition, switch) to see if the effects of active stimulation on conflict adaptation were larger for, or were restricted to response repetitions. This would indicate a priming effect. The authors found that the conflict adaptation effects were larger on response repetition trials than on response switches, but did not further interact with stimulation. This indicated that the effects did not rely on priming. To see if the effect of stimulation significantly changed the course of the experiment the researchers used an additional variable Block. No significant results were obtained under these conditions.

The authors also investigated whether the conflict adaptation effect demonstrated regional specificity (left vs right hemisphere) and motor cortex involvement. A 5 way ANOVA showed that there was significant interaction between region and laterality for DLPFC. It was found that the overall mean RTs were all higher when stimulation was applied to the right hemisphere as compared to left, whereas the opposite was observed for the M1 brain area.

Gbadeyan and authors concluded from their investigations that the DLPFC has a causal role in adaptive cognitive control. Their observations support previous evidence obtained from BOLD responses and also reports about DLPFC activity of current trials matching the activity of ACC in previous trials.  The activity appeared not to be restricted to either the left or right hemisphere which was contrary to reports from other researchers who had linked adaptive control to activity in one hemisphere or another. The authors concluded their article by continuing to describe the beneficial effects of tDCS stimulation on cognition generally and how HD-tDCS in particular could modulate cognition in a regionally specific manner.

COMMENT

What makes this article interesting is it further describes the subtleties of the decision-making process and the various roles that the brain area, the prefrontal cortex (PFC), plays in it. In the last Neurochitchat Blog post, we described the ventral medial prefrontal cortex (VMPFC – or orbitofrontal area, OFC) area and its multiple roles in various cognitive processes such as memory, conditioning, attention, emotions and consciousness. That blog post also described the PFC`s roles in general and the OFC roles in particular in the decision- making System 1 mechanism (the rapid decision-making mechanism) with a link to the striatum and also in System 2 (the slow mechanism) relating to values, conscious choice and the effect that values play on that choice. Research demonstrating these various roles include for example: the orbitofrontal area-basal amygdala pathway and the personal assessment of values given to events and the later recording of updates to that value; comparison of values as in choosing which action to be followed (Rich); the alternating process between two options, also responsibility of the frontal cortex-basal amygdala pathway (Rich); the switching of attention;  the link to reward with encoding predictions of reward values (Zhou); the allocation of rewards, the responsibility of the orbitofrontal cortex-anterior cingulate cortex pathway (Chang); the awareness of choice, involving the VMPFC and DLPFC (De Martino); and the calculation of delay in respect to expected reward/risk (Rudebeck) with feedback of the event responsibility of the VMPFC and rostrolateral prefrontal cortex connectivity.

In this article, Gbadeyan and colleagues discuss the role of a further part of the PFC brain area, that of the dorsolateral prefrontal cortex (DLPFC) and its role in particular in decision-making. This area (designated Brodmann areas 9 and 46) is connected to the post-parietal cortex and is known to be involved in updating information and maintaining the working memory state. Research has shown its activity in working memory is dependent on connectivity with the frontoparietal area (Ekman) and that activity is NMDA dependent. However, the DLPFC area is also known for its effects on decision-making – a function predominantly known to be linked to the VMPFC – with activity not vital to decision-making since it can be carried out without it, but influential.

Therefore, the question is ´what roles does the DLPFC play then in decision-making?` We suggest here that in this context DLPFC acts as a neuronal ´signal amplifier` that allows differences in event values to be more easily distinguished by the sender region, the VMPFC. What do we mean by that? Imagine hitting a ball against a wall where the wall doubles the force of that ball so that when it rebounds back it comes back at double the speed. If we hit several balls with different forces at the wall then the returning balls would be more easily distinguished if the wall can alter the return speed. Doubling numbers increases the difference between consecutive numbers more than between consecutive ones eg. 2 squared is 4, 3 squared is 9 so the difference between the two is 5 compared to just 1 if consecutive numbers are only considered. If a brain area receives a neuronal firing signal and amplifies it to double or triple its original firing strength and then returns it back to its sender region then distinguishing between competing signals by that sender is improved. An alternative option is that the ´amplifying cell` could dismiss the weakest signals and only return the stronger ones. If the signals sent are the values of the various options in the decision-making task and the sender is the ventromedial prefrontal cortex then by amplifying the received signals the dorsolateral prefrontal cortex would have aided the ventromedial prefrontal cortex in its choice of the optimal course of action. The DLPFC area is itself not responsible for the formation of event values or their assessment, both of which are jobs of the VMPFC and other brain areas, it just makes the task of the next stage in the decision-making physiological mechanism, that of choice, by these other decision-making areas easier.

There are several areas of support for such an action of the DLPFC and these are:

  • It is known that there is reciprocal connectivity between the VMPFC and DLPFC areas. Therefore, the VMPFC could send its signal to the DLPFC and the DLPFC could return it – the strength of the signal is not measurable so it does not directly prove that the DLPFC acts as an amplifier. This could explain why low motivational decisions lead to earlier activity in the DLPFC – the DLPFC strengthens the signal even though it is not warranted by the normal valuation process since it is of low motivation. It could also explain why high rewards are favoured since the signal is strong when it is sent to the DLPFC so amplifying would mean further strengthening which is then received and re-registered by the VMPFC.
  • It is also known that there is increased activity in this DLPFC area when actions are selected and initiated (Spence). Therefore, the DLPFC is involved in decision-making, but the exact stage is not determined. It is also known that the DLPFC is involved in reasoning tasks (D`Espito) with the right hemisphere thought to be involved in plan generation and the left in plan execution. A role in strategic control is also reported with the VMPFC known for comparing values as in choosing which action to be followed (Rich), but the awareness of choice the responsibility of the VMPFC and linked DLPFC (De Martino) and feedback of the event the responsibility of the VMPFC and rostrolateral PFC connectivity. Therefore, it would seem that DLPFC plays a role in strategic control.
  • Support also comes from the observations that the activity of the DLPFC leads to irrelevant information being ignored. The lateral intraparietal area (LIP) is also involved in ignoring distracting information, but this could relate to the functioning of the attentional system only. The role of the DLPFC in this function could relate to the other informational situations, eg. sensory input and memory formation. Repetition leads to a shift away from task relevance. In the case of DLPFC, then distracting information is part of the decision-making process since it can be considered as what forms the other options before the choice is made. Since we know that DLPFC activity suppresses distracting information (Gbadeyan) then this implies that the suggestion that the DLPFC only sends back the strongest option is likely.
  • It is also known that activity of this area is linked to conscious awareness. If this suggestion is valid then the incoming signal is amplified and strengthened by the DLPFC so that the information reaches conscious awareness. This information takes priority over the distracting, irrelevant information which according to point (3) above could be suppressed by the DLPFC.

Therefore, the suggested role of signal amplification by the DLPFC implies that this area is part of the decision-making process and does not act just as assurance or verification that the process carried out by other brain areas particularly the VMPFC is correct. It implies that the DLPFC can play a role in both the slow decision-making System 2 (here, strengthens the best value option) and the fast System 1 ´magic answer` decision-making process (here, strengthens the strongest firing option). It can also explain the disputed role of DLPFC in spatial working memory (eg. Platke – yes, Mackay – no with roles in accuracy only and a requirement for precentral sulcus activity). If we look at spatial working memory as being an example of part of a decision-making process undertaken when the route is followed and a decision is required as to the next step, if the above hypothesis is true then the options (ie. turn left or turn right) require the retrieved memory of the correct option to be a stronger firing assembly compared to the unreal option. Amplification of the firing of the possible options by the DLPFC would mean that the correct route learnt from previous experiences would be chosen. This could explain why Mackay said the DLPFC is involved in accuracy only. If the DLPFC is not involved then previous experience would still present the option with the highest value, but it would be more difficult to recognise it from other options presented that represent previous unsuccessful and successful attempts carried out during the learning process. This supports the idea that DLPFC is part of the strategic control mechanism.

This proposed signal amplification role of the DLPFC means that the area is fully functioning for all aspects of decision-making, but at different levels for different tasks. It is therefore, necessary to investigate the relationship between the two for optimal efficiency in all circumstances. The rebound nature of the area may mean that its efficiency can be affected by factors such as tiredness or stress which may imply that only the simple ´magic answer` decision-making mechanism can be used and not the more complicated option assessment and choice mechanism. Therefore, factors influencing the activity of the area may be important and certainly it highlights the need for care in interpreting brain area connectivity and cognitive functioning.

Since we`re talking about the topic………………

……it is said that mindful meditation increases awareness of unexpected distractors. If the role of the DLPFC in amplification of neuronal signals is valid, would an effect on activity of this area in decision-making be observed if mindful meditation is carried out during or before problem solving?

….tDCS is known to decrease stress, but has a negative effect on working memory. What happens to DLPFC activity in an anxiety situation? Would more errors be observed as the ability to suppress distractors cannot be overcome?

…..spatial memory performance is said to be impaired in a rat model of neuropathic pain and this is associated with a reduced hippocampus-prefrontal cortex connectivity? What would happen to DLPFC activity in this situation? Does the pain cause the DLPFC to preferentially strengthen and send back signals relevant to the removal of the pain stimulus whether they are relevant to the decision-making task at hand or not?

…….the administration of a drug that temporarily prevents the opening of K+ channels in the PFC leads to a restoration of working memory activity in ageing (Laubach). Also, administration of estrogen is known to restore multi-synaptic boutons in the DLPFC area in ageing and hence, an increase in working memory is observed. Do these factors have an effect on DLPFC functioning in decision-making?

Advertisements
This entry was posted in decision-making, emotions, prefrontal cortex, Uncategorized and tagged , , . Bookmark the permalink.