Posted comment on ´Impaired oxygenation of the prefrontal cortex during verbal fluency task in young adults with major depressive disorder and suicidality: a functional near-infrared spectroscopy study` written by H. Kim, JK. Choi, B.Jeong, M.Fava, D.Mischoulon, M.J.Park, H.S.Kim and H.J. Jeon and published in Frontiers in Psychiatry 2022 vol 13 doi 10.3389/fpsyt.2022.915425
SUMMARY
Kim and colleagues describe in their article their findings relating to reduced levels of oxygenated haemoglobin (oxy-Hb) in the prefrontal cortex associated with major depressive disorder (MDD) diagnosed in young adults.
In their study, Korean individuals with an average age between 24 and 27 were put into groups according to MDD status. The final results related to 55 controls (HC), 31 patients with MDD but without suicidality and 23 patients with MDD and suicidality. Controls were assessed with having no history of MDD and a score of less than 7 on the Korean version of the Hamilton Depression Rating Scale (HAM-D) whereas MDD was based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) with certain additional exclusions (eg. comorbidity with major psychiatric illness such as bipolar disorder, schizophrenia, delirium, a history of substance-abuse and brain damage) and suicidality was evaluated by the Mini International Neuropsychiatric Interview (MINI – assessment of risk scored from responses to questions, risk of self-harm, suicidal ideation, suicidal plan, suicide attempt – score 1-5 low risk, 6-9 moderate risk, greater than 10 as high risk). The study participants were evaluated to establish the baseline using psychological scales including the MINI, HAM-D, HAM-A (Hamilton anxiety rating scale – Korean version), Barratt Impulsiveness scale-11 (BIS-11 – Korean version) and Clinical Global Impression scale – severity (CGI-S).
Levels of oxy-Hb in various parts of prefrontal cortex were measured (48 sensing areas) including the left prefrontal cortex (PFC), dorsolateral prefrontal cortex (DLPFC), right ventromedial prefrontal cortex (VMPFC), right ventrolateral prefrontal cortex (VLPFC) and left frontopolar cortex (FPC). The levels of oxy-HB were measured using Functional Near-Infrared Spectroscopy (fNIRS) which has the advantages of being a method for non-invasive monitoring of oxygenated and deoxygenated haemoglobin with good levels of spatial and temporal resolution. The changes in oxy-Hb (the haemodynamic changes in each channel were extracted using Modified Beer-Lambert Law – MBLL) were measured at baseline, 4 weeks and 8 weeks when subjects participated in behavioural tests of 3 consecutively repeated 30 second trials of rest, speech (recited Korean alphabet)and verbal fluency tests (VFT – generation of words beginning with specific consonant – a marker for depression since level is known to be sensitive to depression because of known cognitive deficits such as initiation, attention, retrieval and persistence). Kim and team measured the prefrontal activities to give values for baseline, speaking and cognitive executive status during each trial and the values were averaged after appropriate permutation tests. Various statistical analyses were carried out, eg. repeated-measures ANOVA (RM-ANOVA)
The first set of results given showed that the although the PFC activities increased with the cognitive task, the level of the increases were smaller in the young adults diagnosed with MDD than the controls. According to Kim, changes in oxy-Hb were found during the VFT between the groups. A significant difference was found between the groups in the left FPC and this was maintained when the permutation test was performed. When the groups were adjusted for age, sex, systolic blood pressure, heart rate, alcohol intake, smoking status and years in education, there was also a significant difference in the left FPC. Including a further mathematical analysis of VFT values comparing only MDD without suicidality to MDD with suicidality showed that there were significant differences in the values relating to the left PFC, the right VMPFC, right VLPFC and the left FPC when adjustments for the covariates was made.
Following a description of their results, Kim and colleagues began their discussion with first a general summary of the main points of their findings, ie. MDD causes a generally lower prefrontal activation, the impairments were more prominent with MDD with suicidality and that adjusting the values for the covariates, significance was found in the left frontopolar cortex. This led to the conclusion that fNIRS could be used as a diagnostic tool for patients with MDD. Kim and colleagues stated that the left frontopolar cortex findings were supported by other researchers using PET or electroencephalography methods and that the involvement of this area would be consistent with the role of FPC in higher cognitive functions such as planning, problem-solving, reasoning and episodic memory retrieval. However, Kim pointed out that even though the changes in oxy-Hb were smaller in all prefrontal regions with MDD with suicidality only the right VLPFC gave a statistically significant difference.
The discussion concluded with a description of the study`s limitations (eg. use of VFT only, limited temporal measurement, disregard of handedness, and lack of matching for age or sex in the subject`s enrolment process) but stated that this study was the first to identify the value of FNIRS for diagnosis and characterisation of MDD with or without suicidality in young adults.
COMMENT
What makes this topic interesting is that it shows that changes in prefrontal cortex (PFC) firing and connectivity can occur in individuals as more negative emotional behaviours develop. The article written by Kim and colleagues on which this comment is based describes the extent of firing of various PFC regions in the condition of major depressive disorder (MDD) with or without attempts of suicide. Kim and colleagues, limiting their studies to participants between 24 and 27 years of age, showed that activation of the various PFC regions could be divided essentially into 5 groups and these were: a decline in activation of the PFC area which was decreased further in those participants who suffer from diagnosed MDD and had attempted suicide with this group including the regions of left PFC, right ventromedial PFC (vmPFC) and left frontopolar cortex (FPC); no change in PFC activation for MDD sufferers but suicide attempts caused a decrease in activity and this group included the regions of right PFC, right dorsolateral PFC (dlPFC), right and left ventrolateral PFC (vlPFC) and right FPC; a decrease in PFC regional activity with MDD but no change for the suicide attempters and this only related to the region of the left PFC; a decline in activation for the MDD sufferer and an increase in activation for those who had diagnosed MDD and who had attempted suicide and this applied to the left vmPFC region; and finally, the fifth group where MDD with or without suicide attempts led to no change being seen and this was the left dlPFC. The differences in area activity between the MDD with or without suicide attempts groups indicated to the Kim and colleagues that this could be a possible clinical diagnostic tool to validate behavioural assessments and possibly identify further emotional behavioural decline in patients.
Before a discussion on why various PFC regions would be affected by this type of mental illness, it should be noted that Kim`s experiments measured the level of brain activation by near-infrared spectroscopy (NIRS) of blood oxygenated haemoglobin. This was based on the premise that neuronal cell firing would require oxygen for ´household` energy production and hence, activation would need oxygenated haemoglobin (oxy-Hb) and decreased activation would show less oxy-Hb was required. The cognitive test undertaken by the participants which would stimulate the PFC activation and hence, be indicative of any changes in energy needs, was the Verbal Fluency Test. This tests the individual`s language capability plus memory recall as it requires the participant to generate and speak words according to the task given.
The physiological requirements for this type of task requires right brain hemispheric function for general aspects of speech, intonation and emphasis and left brain hemispheric function for grammar and word production. Specific areas affected include the right hemispheric, Wernicke area, and the left Broca region. These are in addition to hemispheric differences where the left (right side of body, right field of vision) focuses on details, sequences and logical tasks, logic numbers, words and core aspects of speech (grammar and words) and the right side (left side of body, left field of vision) focuses on colour, intonation, emphasis, shape, movement and rhythm. The expectation by Kim that MDD would have an negative effect on performance was validated in that this is already known in the cases of stroke and Alzheimer`s disease for example. Further analysis of the words formulated by the participants could aid determination of exact areas of dysfunction for example since it is known that Alzheimer`s disease sufferers have difficulty with irregular verbs due to effects on areas of the brain responsible for explicit memory and Parkinson`s sufferers have difficulties with regular past tenses, – ed endings and others that reflect damage of the brain areas responsible for procedural memory. Damage to the left hemisphere temporal lobe decreases verbal performance whereas damage to the right has an effect on words but not on grammar. However, from a cognitive perspective, there are limitations to the Verbal Fluency Test in that creativity is not being tested and neither is the decision-making capability relating to assessment of risk/reward as well as forward planning.
This comment concentrates on why activations of two PFC areas, the vmPFC and the vlPFC change in the directions reported by Kim and colleagues in their study of participants with MDD with or without suicide attempts.In general, the areas of the PFC are locationally and functionally dependent and this comes from different neurons, activation levels, connectivity and responses (Bear, Wang, Onn) and these aid its wide-ranging functionality such as cognition, working memory, central executive, attention and emotions. Therefore, it is likely that depression which is defined as a serious disorder of mood and has symptoms from an ´emotional side` such as feelings of sadness and rejection, loss of interest in pleasurable activities and a prolonged approach to negative experiences (Robson) and cognitively as an inability to concentrate, difficulty in learning and even ´failure to spot the difference` (Hamzelou) can lead to effects on PFC activation. Behavioural traits of depression come from physiological, neurochemical and firing connectivity changes. In general, depression is associated with: an overactive PFC (not actually supported by Kim`s study which shows for MDD decreased overall functionality for the left PFC for MDD and further decline for suicide attempters, and the right PFC only with suicide attempters); and decreased serotonergic and noradrenergic systems. Associated systems and brain areas also playing a role in depression are: the hippocampus – smaller in depression whereby Prozac can stimulate neurogenesis; occipital functioning which appears to be decreased (Bruder) and likely contributes to the ´failure to spot the difference` (Hamzelou) described above; brain area shrinkage in general as seen by overproduction of the protein KEDD1 which decreases myelin (Duman) and shrivelled dendrites; and an overactive immune system with higher levels of cytokines and where a drug trial with infliximab, a powerful anti-inflammatory, improves symptoms of low mood, fatigue, depression and suicide.
Theories of depression relate to the altered HPA axis as well as serotonergic and noradrenergic systems. In the case of the HPA axis, the theory of depression termed the diathesis-stress hypothesis of mood disorders (Nemeroff), is linked to an overactive hypothalamic-pituitary-adrenal gland system controlling the release of cortisol. This increased HPA activity is linked to stress, anxiety and severe depression. Activation of hippocampal glucocorticoid receptors by cortisol normally leads to feedback inhibition of the HPA axis. Therefore, changes in the hippocampus (smaller and decreased levels of neurogenesis) as seen in depressed patients leads to changes in HPA axis activity. In the case of depression and the neurotransmitter serotonergic and noradrenergic systems, low levels of noradrenaline and 5HT are thought to play a major role in depressive illness. Kety (1975) put forward his permissive amine theory where the level of noradrenaline is greatly controlled by the level of 5HT. Low 5HT means that the noradrenaline levels are less controlled and so can become higher or lower than normal. Drugs that increase the level of 5HT in depressed patients, such as the administration of antidepressant monamine oxidase inhibitors, or 5HT uptake inhibitors would lead to alleviation of the symptoms, including brain memory improvement. However, the time delay between administration and observed effects suggests that the antidepressant mechanism does not rely just on ´real-time` neurotransmitter neuronal levels and that long-term changes in neuronal sensitivity due to receptor number for example may apply. Both 5HT depletion and specific 5HT agonists show the reverse and reduce brain memory performance. In this case, these effects are said to be largely dependent on transmission via the 5-HT1A and 5-HT3 receptors, which regulate the selectivity of retrieval. Therefore, brain memory impairment in depressed subjects may be related to 5HT under-transmission and support the notion that 5HT1A agonists ameliorate memory deficits in depression (Meeter). From the noradrenergic perspective, depression appears to be the result of decreased noradrenaline (Bears) and antidepressants lead to an increase in this particular neurotransmitter probably via the 5HT control described above as well as reducing noradrenaline degradation as in iproniazid (a MAO inhibitor) administration or blocking reuptake of released noradrenaline as in the administration of a tricyclic antidepressant. A receptor effect is also observed with a reduction in the response to alpha2 auto-adrenergic receptors following clonidine which results in a gradual increase in firing and increasing noradrenaline synthesis. Therefore, it is assumed that these described neurochemical effects linked to depression occur in Kim`s study subjects suffering from MDD.
Major Depressive Disorder (MDD) itself is a more severe form of persistent depression according to clinical diagnosis and symptoms of MDD are often grouped according to type, eg. cognitive/affective subtype (eg. lack of interest, guilt, suicidality and suicidal ideation) and somatic/affective subtype (eg. fatigue, psychomotor agitation/retardation, sleep and appetite problems). The brain areas involved in MDD include PFC (Kim`s study), ACC, medial temporal areas including the amygdala and the hippocampus. What Kim and colleagues study found was, with reference to our first area of investigation, the ventromedial PFC (vmPFC) that the right hemispheric vmPFC belonged to the group of PFC brain areas demonstrating a decline in activation which was decreased further in those participants who suffer from diagnosed MDD and had attempted suicide. This group of brain areas included others such as the left PFC, left frontopolar cortex (FPC) whereas the left vmPFC was the only area to demonstrate a decline in activation for the MDD sufferer and an increase in activation for those of MDD who had attempted suicide. Kim`s findings show clear differences between the vmPFC hemispheric areas when suicide was involved. The fact that there are differences in blood flow with MDD is not novel since reports already show decreased blood flow in MDD in other brain areas such as the amygdala, dorsal ACC, insula, superior and middle temporal gyrus, precentral gyrus and caudate and increased blood flow in the pulvinar nucleus of the thalamus.
The question to be answered is whether the changes in vmPFC activation observed in MDD and demonstrated by the NIRS oxy-Hb measurements are caused by the changes in functional activities and manifested in the behaviour are linked to this particular area. The answer is affirmative because the prime function of the vmPFC is value-guided decision-making which is found to be negatively changed in MDD. The vmPFC (anatomically synonymous with the orbitofrontal cortex OFC, Brodmann areas 10, 11, 47 in primates) plays a significant role in emotions and values (courtesy of structural ventral versus dorsal organization – McNamee; or antero-posterior gradient – Li) with activation changing accordingly. For example: increased activity is observed during encoding goal values (Hare) and hence, likely to be decreased in MDD; recording of subjective values of outcomes of reward (Wallis, Harrick) as well as risk and reward levels (Lak) also likely to be absent or changed in MDD; and linking of events to reward values (Favonik) and representing contexts that guide memory retrieval with updating of values (Rudebeck) which are also negatively impacted in MDD. Therefore, changes to vmPFC activation would lead to alterations in values and also in its associated function that of value-guided decision-making. In this role, OFC is known to be involved in comparing options of various values in decision-making (Wunderlich, Hornick). For a subjective choice, values of each option are compared in order to make a decision and the OFC alternates between states associated with the options (ensembles of value-encoding neurons) (Rich). It is likely the strongest firing assembly is chosen and it has been found that there is problem with decision-making when two options are in the presence of very good third alternative (Chau). This type of value-guided decision-making is linked to decreased vmPFC activation (Chau) and likely to occur in MDD.
Courtesy of this link to the emotions/value-guided decision-making function, the vmPFC is known to be associated with several social behaviour roles which could/are affected in MDD and therefore, would be, according to Kim`s, negatively affected in MDD due to the decreased activation observed in the vmPFC. For example: social decision-making role (Croft) eg. acquisition and updating of complex social information in order to form normal moral and social judgements about people; hostile attribution role (Lyu) – where individuals assume others and the actions of others are hostile and this appears to be localised in the left vmPFC; threat and safe outcome discrimination role (Wiemer) – vmPFC activity required for memory of safe associations whereas the left dlPFC is required for general outcomes; and disgust assessment role (Ciaramelli) – avoidance of behaviours that normally elicit interpersonal disgust but lesions less frequent avoidance of core and moral disgust elicitors.
The changes to the vmPFC activity observed in MDD are likely to be linked neurochemically to changes to the neurotransmitters of the area or connectivity to other areas. It is suggested that the dopamine-based pleasure system in depressed individuals is ´skewed` and a different value/worth system is exhibited. One of the key components in the physiological systems resulting in emotions is the prefrontal cortex and 5HT is found in this area to promote the release of dopamine. Therefore, lowered levels of 5HT in the prefrontal cortex would mean that less dopamine is released and the dopamine system would under-perform. Long-term underperformance would lead to a different value system and hence, it is unlikely that any normal activity would result in a pleasure value for that individual. Also, GABA and glutamate might be involved because it has been found that levels of both predict the value comparison signal (Jocham) which could be affected in MDD. The other major contributor to vmPFC functionality performance is its connectivity to other brain areas. Two particular connectivities are relevant to the discussion here, that of the dlPFC which shows reciprocal connections to the vmPFC and is discussed later and the connectivity between the vmPFC and the amygdala (BLA). This particular connectivity (either direct or indirect via the magnocellular mediodorsal thalamic nucleus – Barbas) is involved in the value-guided decision-making process and is important for appropriate assessment of reward value during the choice stage (Zeeb, Rhodes). Whereas BLA lesions increase the choice of large risky rewards, lesions of the OFC significantly decrease risk-taking (Orsini). Other important OFC pathways linked to emotional value and which could influence value-guided decision-making in ways relevant to MDD behavioural traits are: interconnectivity of the OFC with the dorsal raphe nucleus (mainly serotoninergic) where the DRN can modulate and organize the anticipatory responses of the OFC and hence, the reward values allocated (Zhou); the OFC and rhinal cortices where interaction influences judgement of reward size and retrieval (Elridge); and the OFC to the striatum where white matter structural integrity of the connectivity determines the degree of model-based control in decision-making (goal directed, habitual – Piray).
It should also be noted that if vmPFC activation is decreased with MDD then other functionalities known to be associated with this particular area will probably also be negatively impacted. These could include selection and action since lateral OFC lesions produce deficits in response inhibition (Bryden). This is said to be an executive role of the OFC to distinguish between two perceptually similar options during a response conflict and this may be the cause of lack of decision-making in MDD patients. Another area affected may be memory since the OFC exhibits experience-based plasticity (Moghaddan) and therefore, has a role in integrating previous experience into ongoing memory processes (Spalding, Rudebeck). Attention is also linked to OFC functioning and may be affected by this area`s changed levels of activation. It is known that the OFC is involved in mechanisms for visual attention allocated to stimulus features that have been associated with reward on previous trials (Arvaidya). In this case, OFC activation guides attention to reward-predictive stimulus features and therefore, decreases in activity as observed with MDD could cause the sufferers to have deficits in allocating attention to relevant or desired features (Arvaidya).
Therefore, the reduced levels of activation of the vmPFC (left or right) observed with MDD in Kim`s experiments were, according to the known functionalities of the area, responsible for the behavioural traits associated with MDD such as feelings of sadness and rejection, loss of interest in pleasurable activities, value-guided decision-making problems, inability to concentrate and difficulty in learning. However, Kim`s study expanded the knowledge of this association by looking at what happens to vmPFC activation when their MDD study subjects included those that had already attempted suicide or exhibited suicide ideation. In this case, Kim found there were hemispheric differences in vmPFC activations in that the right vmPFC showed decline in oxy-Hb levels between healthy controls to MDD sufferers and a further decline to MDD sufferers who had attempted suicide whereas the left vmPFC showed decline in oxy-Hb levels between healthy controls to MDD sufferers but then MDD sufferers who had attempted suicide demonstrated increased levels of oxy-Hb to those who had not (non-attempters). This shows clear differences between the areas when suicide was involved.
In order to explain this difference, we must first consider what behaviour and cognitive characteristics are linked to suicide attempts. There are multiple theories about the neurochemistry that distinguish suicidal individuals to others, eg. Integrated Motivational Model and Jollant`s Neurocognitive model (Dobbertin). In Van Heeringen`s Neurobiological Model which is based on predictive encoding, it is proposed that suicidal individuals compared to non-suicidal individuals, have impaired systems of evaluating old beliefs and the importance and certainty (precision) of new information (ie. emotional values) as well as an impairment of appropriately changing current beliefs and strategies accordingly (updating). Learning is directly affected with it being biased towards more negatively valued behaviours. Therefore, compared to MDD without suicidality it can be assumed that the physical mechanisms and cognitive capabilities associated with vmPFC functionality are the same even if not under the same demands. Hence, the decreased activation of the right vmPFC observed by Kim is explainable. However, an increase in activation was observed in the left vmPFC in the MDD with suicide attempt rather than the increased decrease observed with the right. A search for the reason for this increase has given no definitive conclusions, but three suggestions are made which might be possible explanations.
The first of these suggestions relates to updating of information. Suicide attempters are said to display difficulty in updating beliefs and strategies and since belief updating is dependent on the right inferior frontal gyrus (IFG) and bilateral superior frontal gyrus (SFG) for positive valued information and for negative information the left IFG and right inferior parietal lobule (IPL), then it could be that increasing activation of the left vmPFC leads to inhibition of these areas. This appears to be supported by the observation that general firing patterns/connectivity with suicide attempters shows that increased activation in the right lateral OFC is connected to decreased activation in the right superior frontal gyrus during decision-making with angry/neutral faces (Brown). Therefore, although this study is not related directly to updating it does show that OFC-frontal gyrus connectivity does exist and could provide an explanation for the increase in activation seen in MDD with suicide attempts.
The second possible explanation relates to the PFC hostile attribution role which is shown to be localised to the left vmPFC (Lyu). Hostile attribution, where individuals show a cognitive bias by interpreting the behaviour of others in various situations as threatening, aggressive or both, may be increased in MDD with suicide attempts whereas ambivalence or lack of interest in self or others may be the behavioural trait associated with MDD without suicide attempt. Further experimentation would have to be carried out to verify this.
And the third possible explanation may be that MDD with suicide attempts is associated with increased levels of anger compared to MDD without where just like, hostile attribution, MDD individuals exhibit ambivalence or lack of interest in Self or others. A positron emission tomography study looking at regional cerebral blood flow and autobiographical narrative scripts and anger induction found that controls during anger induction had significantly greater regional cerebral blood flow increases in the left vmPFC than MDD with anger induction (Dougherty). Therefore, in the control individuals, anger leads to increased activation in the left vmPFC and an inverse relationship to the left amygdala. However, in patients with MDD there is no significant relationship between left vmPFC and amygdala activation but in MDD individuals with anger induction then there is a positive correlation between left vmPFC and amygdala (Dougherty). Just like Kim`s study this difference was suggested as distinct enough pathophysiologically to be of clinical use. Therefore, in MDD with suicide attempts then there would be no vmPFC activation, but MDD with suicide attempts would be linked to anger inductions. The increased amygdala activation associated with this as suggested by Dougherty`s study is supported when connectivity/firing patterns in suicide attempters is studied. For example, there is increased activation in the ACC, dlPFC, sensory cortex, temporal cortex during angry trials with adolescents but decreased activation in all areas with neutral/happy face trials (Pan) but with MDD with suicide attempt, individuals show increased functional connectivity between the amygdala, dlPFC, dmPFC and precuneus during emotional face recognition tasks when faces are considered to look like them independent of emotional expression (Alarcon). However, against this hypothesis is the view that depression results from anger turned inward (Freud) and that depressed individuals have higher rates of anger and aggression than controls, eg. ´anger attacks` where the anger is inappropriate to the situation are observed in patients with MDD at a 30-40% frequency rate. Also, it is difficult to ascertain if anger is a valid reason for the increase in vmPFC activation since the areas believed to be involved in MDD (eg. PFC, medial temporal areas such as amygdala and hippocampus) are also believed to be activated in anger and therefore, it is difficult to assign specific functionality to the activation observed.
Therefore, to summarise Kim`s results relating to vmPFC activation in MDD with or without suicide attempt, the decrease in vmPFC activation observed with the verbal fluency task correlates to the behavioural traits attributed to the conditions. This is because lack of interest, dysfunctional decision-making, attentional problems, memory difficulties for example are functions allocated to the vmPFC and its connectivity to other brain areas. The increase in activation in the vmPFC with MDD with suicide attempts is suggested as possibly due to updating problems and connectivity to the frontal gyrus, and an increase in anger and hostility towards others and linked to the action of the amygdala.
The second aspect of PFC activation relating to MDD with or without suicide attempt highlighted by Kim`s study relate to the lateral PDC area. The researchers found that the right dlPFC exhibited no change in MDD, but MDD with suicide attempts showed decreased activation as observed by decreased oxy-Hb levels. There were no changes reported with the left dlPFC with MDD with or without suicide attempts. In the case of the right and left vlPFC, both hemispheres exhibited no change in PFC activation in MDD, but MDD with suicide attempt showed decreased activation as observed by decreased oxy-Hb levels. Therefore, the disruption in activity is associated with MDD with suicide and not particularly with MDD alone. In general, the lateral PFC is said to be functionally involved in strategic control, ie. thinking and planning (Macdonald) and receives motivational inputs from the ACC as well as in the representation of cognitive information from memory (Ding). It is also said to be involved in threat and safe outcome discrimination (Wiemer) with particularly the left dlPFC being required for general outcomes. Therefore, together the lateral PFC area can relate various information to outcomes and exert cognitive control. This probably involves the third functional role of the lateral PFC, that of working memory. The lateral PFC plays a role in the mediation of complex tasks (Medalla) with the strength and integrity of connectivity between the lateral PFC and posterior parietal cortex predicting the ability to maintain and update working memory items (Ekman). Therefore, the lack of activation of this area observed by Kim in the MDD condition can be correlated to the behavioural traits of general lack of interest and a reluctance to participate in decision-making and strategic control. However, with MDD with suicide attempts the decrease in activation observed relative to controls could be attributed to the subject`s further lack of engagement of strategic control and outcome monitoring when suicide has been attempted.
A look at the functionality of the vlPFC areas (right and left) and right dlPFC area specifically questions Kim`s findings. This is because vlPFC is known to be involved in emotion, reward, motivation, threat detection and fear (Koenigs) and supports decision-making (right middle vlPFC responds to decision uncertainty). The vlPFC is also known to be involved in memory such as that with behavioural reversal activity (Evers) and prospective memory (event greater than time based – Cheng) and the updating of action plans (right posterior vlPFC). It is also involved in visual and perceptual processing since the vlPFC is the end point of the ventral pathway of inputted visual information processing and is responsible for transmitting stimulus characteristics. For example, the vlPFC shows multisensory responses to simultaneously presented faces and vocalisations of non-human primates (Huang) and whereas some neurons are selective for a particular face or vocalisation, others exhibit activity patterns relating to working memory combining information from both auditory and visual sources. The area, also like vmPFC, is involved in selection and action where the vlPFC plays a role in processing and integrity of contextual information and directs manipulative action, eg. eat and grasp (Brians). Related to this, the right vlPFC is said to be linked to control of motor inhibition where control is engaged to stop or to override motor responses. It also has a role in right lateralised ventral attention network governing reflexive reorienting (Corbetta) – a function requiring connectivity to the temporoparietal junction and inferior parietal lobe (Corbetta).
Therefore, with regards to this long list of functions that match behaviour known to be affected in MDD, it would be expected that MDD subjects would demonstrate decreased activation of the vlPFC just like with the vmPFC. This, however, was not observed by Kim. One possible explanation is that the natural decrease in vlPFC was masked by increased excitation from one of its connected excitatory partners, the amygdala. Studies show that in some conditions the amygdala-PFC pathway is increased. For example, it was found that reminders of death often lead to defensive responses in individuals especially in those with low self-esteem. Individuals with high-esteem faced with mortality threat had increased amygdala-vmPFC connectivity during the processing of the threat stimuli compared to the individuals with low self-esteem. Also, rumination in response to stress in adolescent girls is found to be linked to high levels of depressive symptoms linked to functional connectivity between the amygdala and vlPFC (Fowler). This suggests that stress reactive rumination interferes with effective emotion regulation which possibly heightens the depressive symptoms (Fowler). And also, abnormal functional connectivity between the amygdala (more significant connections with right amygdala) and vlPFC is observed in patients with higher anxiety of PTSD both in resting state and threatening conditions (Tetereva). In contrast, however, antisocial personality disorder (APD) individuals show a reduction in overall PFC gray matter of 11 -17% depending on degree of violent/aggressive/impulsive aggression tendencies and antisocial behaviour is linked to decreased amygdala functioning. This shows that the connectivity and functionality is not clear cut for all behavioural conditions. Therefore, in Kim`s study under MDD the decreased activation of the vlPFC according to functional demands is possibly masked by increased activity induced by amygdala pathway activation as shown by most behavioural traits associated with the condition.
Another ´masking` hypothesis relates to the use of Verbal Fluency Test. For example, it has been shown that the vlPFC is important for cognitive reappraisal (Cheng) and therefore, in an experiment where TMS is applied to left or right vlPFC during a reappraisal task there are changes in activation when participants are asked to down-regulate or attend to pictures showing social exclusion scenarios (Cheng). The results showed that TMS applied to either hemisphere increases reappraisal difficulty but the outcome of the reappraisal (measured by negative feelings) is significantly worse when TMS is applied to the right vlPFC whereas verbal fluency during oral reporting of the reappraisal is worse when the TMS is applied to the left. Cheng and team suggested that distinctive roles occur with the left vlPFC responsible for the linguistic task demand and the right vlPFC becomes responsible for inhibiting inappropriate negative emotions and thoughts generated by the effective scenarios (Cheng). Therefore, in Kim`s experiments fulfilment of the Verbal Fluency Task demand would stimulate activation of the left vlPFC and hence, mask the natural reduction in activation seen in MDD.
Therefore, this leads on to the question of what happens in MDD with suicide attempts. In this case, the vlPFC activity is decreased as expected. This negates the hypothesis about the VFT masking the natural reduction by causing vlPFC activation, because the VFT demand is still there. However, with regards to the amygdala hypothesis suggested above, the decreased activation observation implies that the amygdala ´excitatory` influence is removed. (Dobbertin). This is possible as shown by the observations of antisocial personality disorder (APD) individuals where antisocial behaviour is linked to decreased amygdala functioning.
Another possible explanation for a change in amygdala influence or even in suicide attempters alone is physiological. It is possible that the vlPFC and dlPFC areas may undergo negative physiological changes in connection to suicide attempts. For example: cortical thinning is reported in the left dlPFC and left vlPFC areas (Wagner); there is a more generalised reduced cortical volume in left vlPFC, dorsal PFC (Ding); and further gray matter volume reductions in the left and right dlPFC and right vlPFC in MDD with suicide attempts than MDD without (Zhang). This morphological decline may cause the decreased activation observed in these regions by Kim and colleagues in their studies. This view is supported by the decrease in gray matter volume in the vlPFC observed with ageing and correlating to irrational behaviour (Chung) and a reduction of gray matter of 11-17% in antisocial personality disorder (APD) individuals.
The left dlPFC region produced in Kim`s study unlike the vlPFC region surprisingly showed no change in activation levels in MDD with or without suicide attempts. Just like with the ventrolateral areas, the dlPFC is functionally important in various cognitive capabilities and therefore, decreases in activation would be expected in these MDD conditions. For example, the dlPFC is known to be involved in the encoding of sensory information with feature characteristics and magnitude of reward (Donahue). The dlPFC encodes the signals relating to both task-relevant and task irrelevant features, but only mnemonic signals encode congruently with the choice signals. Task relevant signals relating to previous events are more robustly encoded following rewarded outcomes with distractor responses more strongly suppressed in the dlPFC than the LIP (Suzuki). The dlPFC is also known to have an executive or control role (D´Espisito) with more activity shown in reasoning tasks with right side activation in plan generation and left side for plan execution. Also, with reference to MDD, the dlPFC is known to be involved in value assignment (Asaad) and value comparison (Wan) as well as memory retrieval (eg. disruption of dlPFC leads to improvement of recognition accuracy only in responses where the subject`s awareness of memory retrieval is absent – Glee) and prospective memory (activity in right dlPFC, vlPFC areas with event based memories causing greater activation than time based – Cheng). Therefore, just like for the vlPFC areas it would be expected that both dorsolateral areas would also demonstrate reduced activation in MDD with or without suicide attempts due to the behaviours exhibited. Kim`s study however showed that in the case of dlPFC, the right side demonstrated decreased activation only in MDD with suicide and in the case of the left dlPFC, there was no change in either MDD with or without suicide attempts. This is in contrast to other reports which show that there is in fact increased dlPFC activation in individuals with depression and suicide thoughts and behaviour (STB) (Miller).
The possible explanations for the lack of effect in the right dlPFC in MDD without suicide attempt and decreased activation in MDD with suicide attempt could be, like the vlPFC that there is ´masking` of the natural decrease in vlPFC by increased excitation from the amygdala which is connected to it. Studies show that in some conditions the amygdala-PFC pathway is increased. For example, using NIRS like Kim and team, Matsuoka found that patients with schizophrenia who had a history of suicide attempts showed decreased activation in the right dlPFC due to cortical thinning and this was accompanied by increased right amygdala volume. But the decreased activation observation observed by Kim implies that the amygdala influence is removed (Dobbertin). This is possible because of MDD with suicide attempt shows increased functional activity between the amygdala, dlPFC during some forms of emotional task (Alarcon). However, again the observation with antisocial personality disorder (APD) individuals showing a reduction in PFC gray matter and decreased amygdala function contradicts this. However, in the dlPFC case, there may be another limbic area involved since it is known that there is competition between the dlPFC and striatal systems (Daw) and structural and functional connectivity exists between the OFC, dlPFC and posterior parietal cortex in the striatum responsible for information integration (Jarbo).
Another possible explanation for a change in amygdala influence on the dlPFC, like for the vlPFC, for the suicide attempters may be physiological with the dlPFC areas undergoing negative morphological changes in connection to suicide attempts. For example: cortical thinning is reported in the left dlPFC (Wagner); more generalised reduced cortical volume in the dorsal PFC (Ding); and further gray matter volume reductions in the left and right dlPFC in MDD with suicide attempts than MDD without (Zhang). This physiological reduction in functioning brain cells may cause the decreased activation observed in these regions by Kim and colleagues in their studies simply by the cells not being there. In addition and in support of the striatum link, the reduced volume in the right dlPFC is reported to be associated with an increased volume in the caudate part of the striatum. It has also been found that there are further GMV reductions in the right dlPFC in suicide attempters to those with just MDD and controls (Zhang).
However, the left dlPFC exhibits different results to the right dlPFC, ie. there is no difference in activation linked to MDD with or without to the controls.This could mean that if there is increased dlPFC activation in individuals with depression and suicide thoughts and behaviour (STB) (Miller) then this is masked by decreased morphology as suggested by cortical thinning in the left dlPFC (Wagner) and further GMV reductions in the left dlPFC in suicide attempters to those with just MDD and controls (Zhang). However, it is possible that the left dlPFC is less sensitive to change than other PFC areas and can either through the fact it is neurochemically basically inhibitory (GABA firing where higher concentrations improve working memory load processing capability – Yoon) or its connectivity be adaptive enough to provide the brain with cognitive protection. This is suggested since for example: stress decreases working memory performance due to a decrease in activity of the dlPFC which can be counteracted by tDCS (Bogdanov); Cognitive Training can improve multitasking working memory performance and the volume of the rostral part of the left dlPFC predicts response to training (Verghese); and ageing leads to a slowing of visual processing speed but enhanced recruitment of the dlPFC in older adults is an adaptive response (Voytek). However, further research of this area would have to be carried out to elucidate the exact reason for its lack of change in MDD with or without suicide attempt.
Therefore, in summary, the various PFC regions show different effects with regards to MDD with or without suicide attempts with these effects not really always correlating between functionality of the area in question and the behavioural traits associated with the condition. Kim and colleagues proposed that their NIRS experimental set-up measuring oxy-HB changes in response to demand from the Verbal Fluency Test could be used as a clinical diagnostic test for MDD with or without suicide attempt. However, it can be seen from the discussion above that individual PFC areas play multiple cognitive roles and therefore, performance of any task can cause activations in many areas that may confuse what is actually going on and which changes are being specifically induced. Although the NIRS experimental set-up with the Verbal Fluency Task may show changes in some PFC areas in MDD with and without suicide attempt this particular task places limited demands on cognitive capabilities ie. visual processing, working memory, memory retrieval. The one cognitive capability probably more severely affected by MDD, that of value-guided decision-making is not really required for its successful completion. Therefore, probably a more effective cognitive test would be the Iowa Gambling Task although this may unfortunately then have the disadvantage of being too complicated when studying activation levels of multiple PFC areas. This topic is certainly an area where further research is required with the prize of accurate non-invasive clinical diagnostic testing making it worthwhile.
Since we`re talking about the topic………………..
………there is reported hyperactivity of the HPA axis in anxiety and severe depression with increased blood cortisol levels and increased CHR in the CSF being measured. Injected CHR into the brain is said to cause insomnia, anxiety and decreased appetite in individuals. If CHR is administered and the individual required to perform reward-based decision-making task such as the Iowa Gambling Task whilst NIRS is carried out according to Kim`s study, can it be presumed that vmPFC activation would be more clearly defined under these conditions than with the original Kim experiments?
……… if Kim`s experiments were repeated, but instead of the Verbal Fluency Task, a text was given which required reading out aloud, would measurements of brain waves show that MDD causes a change from the two distinct frequency bands normally observed (ie. theta for word locked saccades and linked to visuospatial attention and fixed durations at approx. 1HZ for multi-word linked formations)?
………if Kim`s experiments were repeated but instead of the VFT using a reward-allocation task that involved encoding the outcomes of rewards to the individuals themselves against rewards to other individuals, would differences to vmPFC activation be observed since it is reported that encoding of self-reward is is carried out in the vmPFC whereas rewards to others are encoded in the ACC (Chang)?
…….if Kim`s experiments were repeated with the time delay between testing extended and whole-brain voxel based morphology analyses of gray matter and white matter integrity (via edge-weight measurement for example) carried out as well, would the morphology testing support the ideas that decreases in PFC area activation seen with MDD are due to reduced brain volume particularly with the vlPFC (Chung) and vmPFC (Jia)?
neurochitchat 