Dysfunction of External Switching and Backward Inhibition in Major Depressive Disorder：General or Emotional specific?

doi:10.21203/rs.3.rs-2951263/v1

Download PDF

Article

Dysfunction of External Switching and Backward Inhibition in Major Depressive Disorder：General or Emotional specific?

https://doi.org/10.21203/rs.3.rs-2951263/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

It remains controversial whether individuals with major depressive disorder (MDD) have dysfunctions in switching and/or inhibition when shifting their attention from one task set to another and whether such dysfunctions, if present, are general or specific to emotional content. In the present study, we investigated these questions with a modified task-switching task to measure the switch cost (the ability to activate a new set) and backward inhibition (the ability to inhibit a task set no longer relevant) in both clinical MDD patients and non-depressive control (NDC) individuals. The dysfunctions were identified to be in the emotional domain. Compared with NDCs, MDDs showed general and emotion-specific dysfunctions in external switching, especially difficulty switching attention from an emotional task to a non-emotional task. Further, this dysfunction was independent of rumination. In contrast, although emotional dysfunction in backward inhibition was also observed in MDD, it was closely related to elevated rumination scores. The results suggest that different symptoms of MDD were associated with different cognitive dysfunctions.

Biological sciences/Psychology/Human behaviour

Health sciences/Diseases/Psychiatric disorders/Depression

major depressive disorder

task switch

dysfunction of executive control

backward inhibition

rumination

Cognitive impairments are now widely acknowledged as an important aspect of major depressive disorder (MDD) ^1,2. Most of these impairments are considered to reflect fundamental problems in executive function (EF) ^3–5. Vital to the control and regulation of cognitive processes and response selection, EF includes three subcomponents: inhibition (e.g., inhibiting prepotent responses), updating (e.g., removing no longer relevant information from working memory), and switching (shifting between tasks or mental sets) ⁶.

Mounting empirical evidence has demonstrated depression-related executive dysfunctions ^4,7,8, for example, in tasks that require suppressing negative stimuli, memory, emotion, and thought ^9,10, updating/removing information in working memory ^11–13 and switching between memory sets or response rules ^14–16. A meta-analysis by Snyder ⁴ concluded that MDD patients are significantly impaired in all tasks requiring inhibition, updating, and switching.

Concerning the dysfunction in switching, which is tightly associated with attentional disengagement, might provide a perspective for understanding the mechanism of a hallmark feature in MDD, i.e., difficulty in disengaging from negative emotions ¹⁷. Early studies assessed dysfunctions in switching among MDD patients usually using Wisconsin Card Sorting Test ^18–21 and Making Test part B ²². But both the tasks involved various psychological components that hard to separate from switching, such as learning from feedback, categorization, and rule induction in Wisconsin Card Sorting Test ⁴ and psychomotor speed, scanning, and visuomotor tracking and divided attention in Making Test part B ²³. In recent years more rigorous controlled experimental paradigms have been widely used, such as the internal shift task and the task-switching task ^{14–16,24−30}. However, the majority of these studies investigated subclinical participants with strong ruminative tendencies. Among the few studies with clinical MDD patients, Lo and Allen ¹⁵ and De Lissnyder, et al. ²⁵ consistently found that depressive symptoms were tightly associated with dysfunctions of switching in the internal shift task. In contrast, using the task-switching task, Meiran, et al. ³¹ did not find any impairment of switching in patients with unipolar depression. Whitmer and Gotlib ¹⁴ also failed to find such impairment in clinical MDD patients unless they were induced to ruminate. Adopting the same task as Whitmer and Gotlib ¹⁴ except replacing the non-emotional stimuli with emotional faces, De Lissnyder, et al. ²⁷ found that subclinical undergraduates with moderate or severe depressive scores (score > 20 on Beck Depression Inventory-II, Netherlands version) showed few impairments in both general switching function (i.e., switching between non-emotional task) and emotional specific switching function (i.e., switching between non-emotional task and emotional task). These contradictory results indicated the issue of whether clinical MDD is related to switching dysfunction is still elusive.

One of the potential factors attributed to these discrepant results was the distinct cognitive functions underlying the two experimental paradigms. The task-switching task measures the switching between perceptual attributes of visual stimuli ³², while the internal shift task measures the switching of attentional focus among different representations in working memory ³³. With one reflecting external switching and the other internal switching ^15,28, there is much evidence for the dissociated mechanism underlying the two tasks ^34,35. The results concerning the switching dysfunction among clinical MDD so far were only observed in the internal shift task ^15,25 but not in the external switching task ^14,27,31. Moreover, many hallmark features of depression are related to operations in working memory, such as the reduced ability to regulate and control negative materials in working memory and impaired EF underlying various cognitive biases (e.g., mood-congruent interpretation biases, memory biases, and attentional biases) ^36,37. Consequently, some researchers ^24,38 considered the internal shift task more suitable than the external switching task for testing switching dysfunctions in MDD (e.g., internally switching the focus of attention away from negative information).

Still, there is a possibility of dysfunction in external switching in clinical MDD patients. Firstly, both Whitmer and Gotlib ¹⁴ and De Lissnyder, et al. ²⁷ adopted a visual search task displaying the target with 3 distracters in a 2×2 matrix, as in Wang, et al. ³⁹ and Kristjansson and Driver ⁴⁰. Given the introduction of distracters, the switch cost in such tasks can be confounded by other priming effects between successive trials, such as target repetition, distractor repetition and role reversal (i.e., target in trial N-1 become distracter in trial N) ⁴⁰. Secondly, the results in De Lissnyder, et al. ²⁷ may not be generalized to clinical MDD patients, because the high BDI scores of subclinical participants in their study had relatively good EF. Thirdly, the MDD patients in the study of Whitmer and Gotlib ¹⁴ were either induced to ruminate or distracted from rumination. With this procedure, although the MDD-rumination group exhibited significantly larger switch costs (i.e., poorer switching) than did either the MDD-distraction group or the NDC-rumination group, it remains unknown whether there would be a dysfunction in non-emotional switching for MDD patients when they are neither induced to ruminate nor distracted from rumination. Finally, the non-emotional task in De Lissnyder, et al. ²⁷ contained emotional faces which can induce automatic emotional processing in MDD patients ⁴¹ with a strong bias toward negative information ^42,43. Briefly, their design did not make a clear distinction between emotional and non-emotional task switching. Therefore, it is remaining necessary to drive us here to testify the issue of whether there is a dysfunction in external switching among MDD and if yes, to what extent such dysfunctions are emotion-specific.

In the present study, we also measured backward inhibition (BI) and rumination, since the two factors were usually involved in previous studies and proved to mediate depressive-related switching dysfunction ^14,24,25,29. BI reflects the inhibitory process of the previous tasks or mental sets to accelerate the rate of deactivation, allowing quicker switching to the new target in switching task ⁴⁴. Many studies have observed depressive-related BI dysfunction in switching tasks ^14,16. Rumination, defined as “behaviours and thoughts that focus one's attention on one's depressive symptoms and the implications of those symptoms” ⁴⁵, is a core symptom of depression. It has been shown to be related to dysfunction of switching and/or BI in both clinical and subclinical depressive individuals ^{14,24–27,29}.

In summary, the present study attempted to compare MDD patients and non-depressive control (NDC) participants on switch cost and BI effect by measuring with a classic task-switching task. Following this task, only one target with multi-valent dimensions would be presented at any given time. This was to prevent confounds from irrelevant priming effects as in Whitmer and Gotlib ¹⁴ and De Lissnyder, et al. ²⁷. Both emotional (facial emotion) and non-emotional (gender, age) tasks were used to assess whether the dysfunctions in switching and BI in MDD were emotional-specific. No emotional stimuli were used in the non-emotional tasks to avoid confounds from automatic emotional processing. Faces were used as materials as in previous studies ^24,25,27, given that they are ecologically valid interpersonal stimuli for studying depression.

Since the depression was found to be tightly related to difficulties with disengaging attention away from negative information ¹⁷ and inhibiting negative emotion ³⁶, we predicted an emotion-related dysfunction of switching and BI in MDD. However, due to the confounding factors described above in the most relevant studies of De Lissnyder, et al. ²⁷ and Whitmer and Gotlib ¹⁴, it was hard to predict whether this emotion-related dysfunction was caused by general cognitive deficits and what role rumination may play in the dysfunctions. In detail, the following questions could be tested: (1) whether MDDs have emotion-related dysfunctions in external switching and BI (by comparing the group differences of switch costs and/or BI effects related to the emotional task); (2) whether MDDs have general dysfunctions in external switching and BI (by comparing the group differences of switch costs and/or BI effects in non-emotional tasks); (3) whether emotion-related dysfunctions are (partly) caused by general cognitive dysfunction (by the group differences of difference between emotional and non-emotional switch costs and/or BI effects) and (4) whether rumination plays a role in dysfunctions of switching and BI (by estimating rumination scores in the switch cost and BI effect with analysis of regression and covariance).

2.1 Participants

The sample size of participants in the present study was referred to that of Whitmer and Gotlib ¹⁴ who observed robust switching cost and BI effect with about twenty MDD patients per group. Twenty-five MDD patients were recruited from the outpatients of the Psychiatric Clinical Department in Guangzhou Huiai Hospital, China. Data from 5 of them were discarded due to the failure to complete the whole experiment. None of the 20 valid participants were under treatment with antidepressants at the time of testing, but 12 of them had previously been under pharmacological treatment. Twenty NDC participants were recruited through advertising from the community.

All participants underwent an interview using the 21-item BDI–II ⁴⁶ to assess the severity of their depressive symptoms. The inclusion criteria for the MDD patients were scoring above 19 on BDI–II and meeting the criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), as diagnosed by two psychiatrists with Structured Clinical Interview for Axis I DSM-IV Disorders (SCID-I). The inclusion criteria for the NDC participants were scoring below 14 on BDI–II and having no history of depressive disorders or any other psychological disturbance as defined by DSM-IV.

For the 20 MDD participants (9 males), the mean age was 28.3 years (range: 18–51) and the mean years of education was 13.6 years (range: 9–19). For the 20 NDC participants (11 males), the mean age was 29.4 years (range: 19–50) and the mean years of education was 14.1 years (range: 6–19). The two groups were matched in age (t (38) = 0.39, p = 0.7] and years of education (t (38) = 0.44, p = 0.66). The details of the demographic information are listed in Table 1.

All participants were right-handed with normal or corrected-to-normal vision. None were color-blinded. Prior to participating, written informed consent was obtained from each participant, and the research protocol followed the guidelines outlined in the Declaration of Helsinki. The study was approved and conducted in accordance with the regulations set by the Ethics Committee of The Centre of Psychology and Brain Science, Guangzhou University.

Table 1

*Demographic and Clinical Characteristics of Participants*
	Group		t-test
	MDD	NDC	t	P
N	20	20
Gender(%Male)	45%	55%
Age	28.3(8.2)	29.3(9.4)	0.39	0.7
Education (Years)	13.60(3.17)	14.05(3.32)	0.44	0.66
BDI-II	31.75(12.09)	1.70(1.22)	11.06	0.00
Rumination	26.05(5.28)	17.90(3.11)	5.95	0.00
Note. Standard deviation in parentheses. BDI-II = Beck Depression Inventory-Second Edition. Rumination = scores for brooding and reflection subscales of the Ruminative Response Scale.

[insert Table 1 about here]

2.2 Materials and Procedure

Ruminative Response Scale (RRS)

The RRS was used to measure rumination ⁴⁷. The RRS consists of 26 items that describe responses focusing on the self and the causes or consequences of a depressed mood. Each item is self-reported on a four-point Likert scale ranging from 1 (almost never) to 4 (almost always). The RRS has been demonstrated to have good psychometric properties (Treynor et al., 2003).

Task-switching task

The task was programmed using the E-Prime 1.0 software package. All visual stimuli were presented against a grey background on a 19-inch colour monitor (resolution: 1024 * 768; frame rate: 75 Hz) approximately 60 cm away from the viewer.

Thirty-two angry faces and 64 non-emotional faces with more than 80% identity rate were selected from a face database ^{[insert Footnote 1 about here]}. All faces were rated on valence (range: 1 [very unpleasant] to 9 [very pleasant]) and arousal (1 [calm] to 9 [excited]). Referring to prior studies ^9,30, the inclusion criteria were as follows: for the neutral faces, valence (3 < valence < 5) and limited arousal value (arousal < 3); for the angry faces, negative valence (valence < 3) and moderate-high arousal value (arousal > 4). The two types of faces were matched in age (young or old) and gender (male or female). As in De Lissnyder, et al. ²⁵, angry faces were selected as negative stimuli because they have a direct relevance to depression and rumination which is characterized by fear of social rejection ⁴⁸.

Each participant completed 1 practice block prior to 18 test blocks, with a short break between two test blocks. As shown in Fig. 1, each block began with a cue of “start” presented at the center of the screen. Once the participant pressed the SPACE bar on a standard keyboard, 32 trials were then presented sequentially. Each trial contained a 500-ms central fixation, a target display, and finally a 250-ms feedback. The target display was a face at the screen center and a task cue (“emotion”, “age” or “gender”) below the face with a 6º center-to-center distance. Participants were to classify whether the face was angry or neutral for the “emotion” cue, young or old for the “age” cue, and male or female for the “gender” cue. They responded by pressing the “F” or “J” key as quickly and accurately as possible within a time window of 3000 ms.

[insert Fig. 1 about here]

For the emotional trials, i.e., when the task cue was “emotion”, the face could be an angry face or a neutral face with equal probability. For the non-emotional trials, i.e., when the task cue was either “age” or “gender”, all faces used were neutral faces to avoid the automatic processing of emotional information on faces. The three types of trials were pseudo-randomly mixed in each block.

For two successive trials, if the task cue was “emotion” for both the present trial and the preceding trial, the present trial would be defined as an “R-EE” trial (i.e., repeat trial, emotion cue to emotion cue). If the task cue was “age” or “gender” for both trials, the present trial would be defined as an “R-NN” trial (i.e., repeat trial, non-emotional cue to non-emotional cue). In total, there were 72 “R-EE” trials and 144 “R-NN” trials. In contrast, if the task cue changed from one non-emotional cue to another non-emotional cue, the trial would be defined as an “S-NN” trial (switch trial, non-emotional cue to non-emotional cue). Two more types of switch trials “S-EN” trials and “S-NE” trials were similarly defined. In total, there were 36 “S-NN” trials and 36 “S-EN” trials and 72 “S-NE” trials.

BI was measured following Koch, et al. ⁴⁴ by comparing reaction times (RTs) for A-B-A sequences and C-B-A control sequences. In an A-B-A sequence, it is necessary to relieve the inhibition of task A when switching from task A to B previously. As in Whitmer and Gotlib ¹⁴, two types of BI were measured: (1) BI of emotional tasks (E-ABA vs. E-CBA) where A was an emotional task and B and C were non-emotional tasks; (2) BI of non-emotional tasks (N-ABA vs. N-CBA) where A and C were non-emotional tasks and B was an emotional task. There were 32 trials for each type of BI.

3.1 Switch cost

A mixed-design ANOVA with the trial type (“R-EE”, “R-NN”, “S-EN”, “S-NE”, “S-NN”) as a within-subject factor and group (MDD and NDC) as a between-subject factor was firstly carried out on accuracy and RT. If the main effect of trial type was significant, indicating the presence of switch cost, switch costs would then be calculated by computing the difference in RT/accuracy between switch trials and repeat trials. There were three types of switch cost: “Cost-NN” (“S-NN” - “R-NN”) was the cost of switching from one non-emotional task to another; “Cost-EN” (“S-EN” - “R-NN”) was the cost of switching from an emotional task to a non-emotional task; and “Cost-NE” (“S-NE” - “R-EE”) was the cost of switching from a non-emotional task to an emotional task.

The switch costs were then entered a mixed-design ANOVA with switch type (“Cost-NN”, “Cost-EN” and “Cost-NE”) as a within-subject factor and group (MDD and NDC) as a between-subject factor to investigate whether there were switching impairments in the MDD and whether such impairments were modulated by emotional information.

3.1.1 Accuracy

The ANOVA with trial type and group factors on mean accuracy yielded a main effect of trial type (F(4, 152) = 15.23, p = 0.000, η_p² = 0.29, Greenhouse-Geisser corrected) and an interaction effect between trial type and group (F (4, 152) = 3.48, p = 0.009, η_p² = 0.08, Greenhouse-Geisser corrected), but no main effect of group (F (1, 38) = 0.08, p > 0.5). The planned pair-wise comparisons showed that the differences were significant between “S-NN” and “R-NN” (89.7% vs. 93.0%, p = 0.001) but not between “S-NE” and “R-EE” (95.0% vs. 95.4%, p = 1.00), nor between “S-EN” and “R-NN” (91.3% vs. 93.0%, p > 0.1). The results indicated the cost of switching between two different non-emotional tasks but not in switching between emotional and non-emotional tasks. No group difference was found in switch costs.

3.1.2 RTs

Median RTs from correct response trials were used to reduce the effects of outliers among the within-subject data ²⁵. The ANOVA with trial type and group factors on median RTs revealed the main effects of trial type (F (4, 152) = 52.76, p < 0.001, η_p² = 0.58, Greenhouse-Geisser corrected), group (F (1, 38) = 25.07, p < 0.001, η_p² = 0.40) and their interaction (F (4, 152) = 6.57, p < 0.005, η_p² = 0.15, Greenhouse-Geisser corrected). Response speed was significantly slower in the MDD patients than in the NDC participants (1497 ms vs. 1018 ms), and there were significant switch costs in RT data.

The three types of switch costs were calculated for MDD and NDC and analyzed with ANOVA, as described above. The results revealed the main effects of switch type (F (2, 76) = 28.48, p < 0.001, η_p² = 0.43), group (F (1, 38) = 8.90, p < 0.005, η_p² = 0.19), and their interaction effect (F (2, 76) = 9.05, p < 0.001, η_p² = 0.19). As showed in Fig. 2, the subsequent simple effects test revealed that most switch costs were significantly greater for MDD than NDC (“Cost-NN”: 327 ms. vs. 213 ms, p < 0.05; “Cost-EN”: 438 ms vs. 150 ms, p < 0.001; “Cost-NE”: 161 ms vs. 72 ms, p = 0.08), suggesting a general dysfunction in task switching among MDD patients.

[insert Fig. 2 about here]

Critically, the group difference in “Cost-EN” (288 ms) was significantly greater than that in “Cost-NN” (114 ms) and “Cost-NE” (89 ms). The group difference between “Cost-NN” and “Cost-NE” was not significant. Further comparison showed a significant difference in switch costs between “Cost-EN” and “Cost-NN” in MDD (438 ms vs. 327 ms, p < 0.05). That is, the patients had a greater switch cost when switching from an emotional task to a non-emotional task, compared with when switching from another non-emotional task. For NDC, there was no such difference (150 ms vs. 212 ms, p = 0.31).

3.2 BI effect

As in Whitmer and Gotlib ¹⁴, BI was analyzed to investigate whether in the MDD patients there was a dysfunction of inhibition in task switching. This was done by entering accuracy and RT of the “N-ABA^”, “N-CBA^”, “E-ABA” and “E-CBA^” trials into a mixed-design 2×2×2 ANOVA with emotion type (non-emotional BI vs. emotional BI) and sequence type (A-B-A sequence vs. C-B-A sequence) as within-subject factor and group (MDD vs. NDC) as between-subject factor.

3.2.1 Accuracy

ANOVA on mean accuracies revealed the main effect of emotion type (F (1, 38) = 15.03, p < 0.001, η_p² = 0.28) and interaction of emotion type and group (F (1, 38) = 3.62, p = 0.06, η_p² = 0.09) but no other effects (ps > 0.1).

3.2.2 RT

ANOVA on medium RTs revealed the main effect of emotion type (F (1, 38) = 63.99, p < 0.001, η_p² = 0.63), group (F (1, 38) = 24.27, p < 0.001, η_p² = 0.39), and their interaction (F (1, 38) = 9.33, p < 0.005, η_p² = 0.20). The two-way interaction between emotion type and sequence type (F (1, 38) = 3.68, p = 0.06, η_p² = 0.09) and between sequence type and group (F (1, 38) = 3.37, p = 0.07, η_p² = 0.08) were marginally significant. The three-way interaction was not significant (F (1, 38) = 0.01, p = 0.91).

As showed in Fig. 3, the simple effects test indicated that the RT difference between “E-ABA” and “E-CBA^” trials was significant for NDC (1021 ms vs. 941 ms, p < 0.05) but not for MDD (1416 ms vs. 1425 ms, p > 0.5), indicating a significant emotional BI effect for the NDC (79 ms) but not MDD (-9 ms). The RT differences between “N-ABA” and “N-CBA^” trials were not significant for either NDC (1158 ms vs. 1185 ms, p > 0.5) or MDD (1783 ms vs. 1911 ms, p > 0.1), indicating that non-emotional BI effect was not observed for either MDD or NDC.

[insert Fig. 3 about here]

3.3 Effect of rumination on switch cost and BI

The scores for the brooding and reflection subscales of the rumination scale were merged as an overall rumination score, as they were difficult to distinguish among currently depressed individuals ⁴⁹. Regression analyses with rumination score as an independent factor were used to estimate the effect of rumination on switch cost and BI effect (i.e., “Cost-NN”, “Cost-EN”, “Cost-NE”, non-emotional “BI” and emotional “BI” were separately used as the dependent variable) in MDD and NDC. The regression did not reveal any effects on the accuracy of data. For the RT data, only emotional “BI” was found to be negatively related to the rumination score in MDD (t (19) = -2.75, β = -17.24, p < 0.05).

To further examine the emotional “BI” effect and rumination score, an analysis of covariance (ANCOVA) was conducted on RTs with sequence type (“E-ABA” vs. “E-CBA^”) and group (MDD vs. NDC) as independent factors and rumination score as a covariate. The results showed significant main effects of sequence type (F(1, 37) = 5.92, p < 0.05, η_p² = 0.14) and group (F(1, 37) = 23.76, p < 0.001, η_p² = 0.39) and the interaction effect between sequence type and rumination score (F(1, 37) = 4.80, p < 0.05, η_p² = 0.12). Critically, the interaction between sequence type and the group was not significant (F (1, 37) = 0.04, p = 0.85). This suggested that the group difference between MDD and NDC in the emotional BI effect was eliminated after controlling for the effect of rumination.

The present study adopted the task-switching paradigm to investigate whether clinically depressed patients have dysfunctions in external switching and backward inhibition and whether these dysfunctions are specific to emotional content. In line with Whitmer and Gotlib ¹⁴, switching cost and BI were calculated based on comparisons between specific trial types.

Our Results first showed that the switch cost in non-emotional tasks was significantly greater for MDD than for NDC, suggesting a general dysfunction of external switching among MDD. Switch costs also clearly demonstrated emotion-related switching dysfunction in MDD relative to NDC participants, and critically, the switching cost was greater for the emotional-related task than the non-emotional task, indicating both general and emotional-specific external switching dysfunction was found in the current study. This result was inconsistent with prior studies of Whitmer and Gotlib ¹⁴, who failed to find any non-emotional switching dysfunction in MDD except when the subjects were induced to ruminate. This result also contradicted the results of De Lissnyder, et al. ²⁷, who failed to find any general external switching dysfunction and emotional-related dysfunction of switching in participants with moderate or severe depressive symptoms. Note that those two studies suffered from various problems that would confuse their results as analyzed in the introduction and our finding was more reliable. Further, we found that the dysfunction in MDD was mainly manifested in switching from the emotional task to the non-emotional task, rather than the reversed switching, suggesting a valence-specific dysfunction. The results that the difference between “Cost-EN” and “Cost-NN” was significant for MDD but not for NDC indicated the valence-specific dysfunction in MDD may be related to their difficulty in suppressing the emotional task when switching to a non-emotional task. The finding of valence-specific dysfunction of switching in MDD conforms theoretically to the impaired disengagement hypothesis ⁵⁰, which postulates that depression is associated with impaired attentional disengagement from negative self-referent information. The results were also confirmed by the evidence from eye-movement studies, that is, MDD showed an impairment of attentional disengagement from negative emotional information. For example, Eizenman, et al. ⁵¹ found that MDD patients spent significantly more time looking at negative emotional information than did NDCs. Recently, Sanchez, et al. ⁵² found that MDD patients were normal and similar to NDCs in their initial orientation to negative stimuli but had specific difficulties in disengaging their attention from such stimuli.

Secondly, our results showed that the BI effect in emotional tasks was robust in NDC but not in MDD, suggesting dysfunction of emotional BI in MDD. However, when the effect of rumination was controlled, the group difference was eliminated, i.e., comparable emotional BI effects were observed in MDD and NDC. These results indicated that rumination but not depression was associated with dysfunction of the emotional BI effect in external switching among MDD patients. The relationships between rumination, switching and BI in the present study was in line with those reported by Whitmer and Gotlib ¹⁴, who found that trait rumination was also related to BI dysfunction but not to increased switch costs.

Most critically, however, the BI in the study of Whitmer and Gotlib ¹⁴ was about inhibition in non-emotional tasks, while in the present study, it was about inhibition in emotional tasks. we did not find any significant BI effect except a reversed BI effect in non-emotional tasks (-128 ms and − 27 ms, respectively, for MDD and NDC). One possible explanation is that with the sequence of trials used in measuring the non-emotional BI effect, in both the A-B-A and C-B-A sequences, “B” was an emotional task in trial N-1. Since the emotional task will automatically attract attention and processing capacity ⁵³ in trial N-1, participants would not need to suppress information in the non-emotional task (“A” or “C”) performed in trial N-2. That is, the residual activation of the “A” task would facilitate the “A” task when the participant switched the task again for trial N, displaying a reversed BI effect.

From the magnitude of the reversed BI effect, one can also infer that the easier the attentional bias to the emotional face in trial N-1 (e.g., MDD, ⁵⁴), the greater the facilitation of the same task from trial N-2 to trial N. Considering that the non-emotional BI effect was influenced by the emotional task in the present study, it remained unknown whether there is a general dysfunction of BI in MDD. Future studies could test this issue with a purely non-emotional task-switching task.

Briefly, although we failed to replicate the non-emotional BI dysfunction in Whitmer and Gotlib 17, our results extended the association between trait rumination and BI dysfunction in MDD to the emotional domain, i.e., MDD patients have an emotional dysfunction of BI, negatively associated with rumination but independent of depression.

The cause of such BI dysfunction may be the failure in the suppression of negative emotional faces. Ruminative thinking can overload limited executive resources and therefore impair task-relevant processing depending on those resources ⁵⁵. Available executive resources to inhibit the previous, abandoned negative task when switching to the new task is hence reduced. In support of this possibility, multiple studies have shown that rumination is related to dysfunction in the inhibition of negative emotional information ^56,57.

Another possibility is related to the “automaticity” of emotion processing. Piguet, et al. ⁵⁸ found a reduction in BI effect in emotional trials with normal participants. Since emotion can facilitate perception, attention, and memory in various situations ⁵⁹, and since social emotional processing (such as face processing) tends to operate automatically ⁴¹, the impact of inhibition from the preceding task on the current target will be reduced when the target is an emotional face.

This explanation is supported by the neural deactivation of the anterior cingulate cortex, which is positively related to attentional control and effort to overcome the inhibition of the preceding task during the switch to a new task. Rumination in MDD was shown to correlate with an attentional bias for negative emotional faces even after controlling for BDI scores ⁶⁰. This may render the processing less susceptible to interference from inhibition of the preceding task for emotional faces than from non-emotional faces, hence eliminating the emotional BI effect for MDD.

In fact, both can be explained by the recent Attentional Scope Model of Rumination ⁵⁷. According to this model, MDD patients with stronger ruminative tendencies have a narrower attentional scope; when a negative emotional face appears in the visual field, the depressed mood creates a negative processing bias, and then the narrowed attentional scope will lead MDD patients to increase cognitive resources devoted to deep encoding and activating this negative information (i.e., the “automaticity” of emotion processing). This shift in resources, in turn, increases the difficulty of inhibiting the activation when this negative information stops being relevant (i.e., failed suppression).

Thirdly, compared with the effect of rumination on BI, the results of regression in the present study did not show any effect of rumination on switch costs. That is, rumination in MDD plays a major role in the dysfunction of emotional BI, but irrelevant to the dysfunction of emotional switching. Obviously, this conclusion is consistent with that of Whitmer and Gotlib ¹⁴. Moreover, as discussed before, our results extended their conclusion to the emotional domain.

In addition, by adopting the Internal Shift Task, prior studies by Lo and Allen ¹⁵ and De Lissnyder, et al. ²⁵ also demonstrated the dysfunction of switching in MDD. However, their results were conflicting, i.e., Lo and Allen ¹⁵ observed emotion-specific rather than general dysfunction with affective/neutral words, while De Lissnyder, et al. ²⁵ observed general but not emotion-specific dysfunction in MDD. Although there has been no research addressing this discrepancy, we speculate that it may be related to one problem in the experiment of De Lissnyder, et al. ²⁵. Specifically, although the task for the participants was to identify the gender of the face in the non-emotional task, the expression of the face may also be angry. Hence, MDD patients could automatically process the negative face, potentially blurring the difference between emotional and non-emotional (general) task switching. The same problem applies to the study of De Lissnyder, et al. ²⁷. In addition, many prior studies have provided evidence to indicate the distinct cognitive functions underlying the task-switching task and the Internal Shift Task ^34,35. Therefore, it is possible to obtain distinct patterns of results in MDD patients with the two paradigms.

The results of the present study have important clinical implications for improving the understanding of cognitive vulnerability factors for depression, such as impaired cognitive control. If cognitive processes are causally linked to the root problems, modifying these underlying cognitive processes holds promise as a way to improve depressive symptoms ⁶¹. Cognitive control training is a new intervention for depression that could facilitate rehabilitation and remission depression, helping improve patients’ quality of life and prevent relapse ³. Our results can help by providing a reference regarding the target of training.

The present study also has several limitations. First, as only angry faces were used here as emotional stimuli, we cannot conclude that the impairments can be generalized to other negative material (such as sad emotions). It is necessary to compare the effects of different negative stimuli in the future. Second, because the non-emotional BI effect was impacted by the emotional task, it remains unclear whether there is a general dysfunction of BI in MDD. Future studies can test this issue with a purely non-emotional task-switching task. Third, because of the limitations of the participant sample, we cannot generalize the conclusion to other depressed populations, such as the non-medicated depressed population. Fourth, the impact of several important potential factors (such as anxiety score, comorbidity, frequency and history of medication use, and age) on the results could not be excluded. Finally, the association between depression and dysfunction in external switching and BI in the present study does not provide information about causality. Future research should focus on the functional relationship between these constructs.

In conclusion, the present study provides more insight into the relationship between depression and dysfunctional cognitive control, especially in external switching and BI. The results suggested that MDD patients have both general dysfunction and emotion-specific in external switching and emotional dysfunction in BI. In addition, rumination plays a vital role in BI dysfunction but not in external switching dysfunction.

Conflict of Interest

The authors declare no competing interests.

Author Contributions

SH, XL, and BZ designed the experiments, carried out the analyses, wrote and reviewed the manuscript; JZ and YZ collected the data.

Funding

The work was supported by the Projects of the Social Sciences and Humanities on Young Fund of the Ministry of Education, China (19YJC880035) for SH.

Data Availability Statement

The datasets analyzed for this study can be found in the figshare. https://doi.org/10.6084/m9.figshare.21638198.v5.

Rock, P., Roiser, J., Riedel, W. & Blackwell, A. Cognitive impairment in depression: a systematic review and meta-analysis. Psychological medicine44, 2029-2040 (2014).
Vives, M., López-Navarro, E., García-Campayo, J. & Gili, M. Cognitive impairments and depression: a critical review. Actas Esp Psiquiatr43, 187-193 (2015).
Koster, E. H., Hoorelbeke, K., Onraedt, T., Owens, M. & Derakshan, N. Cognitive control interventions for depression: a systematic review of findings from training studies. Clin Psychol Rev53, 79-92 (2017).
Snyder, H. R. Major depressive disorder is associated with broad impairments on neuropsychological measures of executive function: a meta-analysis and review. Psychological bulletin139, 81 (2013).
Villalobos, D., Pacios, J. & Vazquez, C. Cognitive Control, Cognitive Biases and Emotion Regulation in Depression: A New Proposal for an Integrative Interplay Model. Front. Psychol.12, 628416, doi:10.3389/fpsyg.2021.628416 (2021).
Miyake, A. et al. The unity and diversity of executive functions and their contributions to complex "frontal lobe" tasks: a latent variable analysis. Cognitive Psychology41, 49-100 (2000).
DeBattista, C. Executive dysfunction in major depressive disorder. Expert Rev Neurother5, 79-83, doi:10.1586/14737175.5.1.79 (2005).
Hammar, A. & Ardal, G. Cognitive functioning in major depression--a summary. Front Hum Neurosci3, 26, doi:10.3389/neuro.09.026.2009 (2009).
Dai, Q. & Feng, Z. Deficient interference inhibition for negative stimuli in depression: An event-related potential study. Clinical Neurophysiology122, 52-61, doi:http://dx.doi.org/10.1016/j.clinph.2010.05.025 (2011).
Joormann, J. & Gotlib, I. H. Emotion regulation in depression: Relation to cognitive inhibition. Cognition & Emotion24, 281-298, doi:10.1080/02699930903407948 (2010).
Joormann, J. & Gotlib, I. H. Updating the contents of working memory in depression: interference from irrelevant negative material. J Abnorm Psychol117, 182-192 (2008).
Levens, S. M. & Gotlib, I. H. Updating Positive and Negative Stimuli in Working Memory in Depression. Journal of Experimental Psychology: General139, 654-664, doi:10.1037/a0020283 (2010).
Yoon, K. L., LeMoult, J. & Joormann, J. Updating emotional content in working memory: A depression-specific deficit? Journal of Behavior Therapy and Experimental Psychiatry45, 368-374, doi:http://dx.doi.org/10.1016/j.jbtep.2014.03.004 (2014).
Whitmer, A. J. & Gotlib, I. H. Switching and backward inhibition in major depressive disorder: the role of rumination. J. Abnorm. Psychol.121, 570-578, doi:10.1037/a0027474 (2012).
Lo, B. C. Y. & Allen, N. B. Affective bias in internal attention shifting among depressed youth. Psychiatry Res187, 125-129, doi:http://dx.doi.org/10.1016/j.psychres.2010.10.001 (2011).
Chen, X., Feng, Z., Wang, T., Su, H. & Zhang, L. Internal switching and backward inhibition in depression and rumination. Psychiatry Res.243, 342-348, doi:10.1016/j.psychres.2016.06.014 (2016).
Joormann, J. & Quinn, M. E. Cognitive processes and emotion regulation in depression. Depress. Anxiety31, 308-315, doi:10.1002/da.22264 (2014).
Borkowska, A., Drozdz, W., Jurkowski, P. & Rybakowski, J. K. The Wisconsin Card Sorting Test and the N-back test in mild cognitive impairment and elderly depression. World J Biol Psychiatry10, 870-876, doi:10.1080/15622970701557985 (2009).
Fossati, P., Ergis, A. M. & Allilaire, J. F. Problem-solving abilities in unipolar depressed patients: comparison of performance on the modified version of the Wisconsin and the California sorting tests. Psychiatry Res104, 145-156 (2001).
Grant, M. M., Thase, M. E. & Sweeney, J. A. Cognitive disturbance in outpatient depressed younger adults: evidence of modest impairment. Biological Psychiatry50, 35-43 (2001).
Merriam, E. P., Thase, M. E., Haas, G. L., Keshavan, M. S. & Sweeney, J. A. Prefrontal cortical dysfunction in depression determined by Wisconsin Card Sorting Test performance. The American Journal of Psychiatry156, 780-782 (1999).
Austin, M. P. et al. Cognitive function in major depression. J. Affect. Disord.25, 21-29, doi:10.1016/0165-0327(92)90089-o (1992).
Wajdik, C. et al. No change in neuropsychological functioning after receiving repetitive transcranial magnetic stimulation treatment for major depression. J. ECT30, 320-324, doi:10.1097/YCT.0000000000000096 (2014).
De Lissnyder, E., Koster, E. H. & De Raedt, R. Rumination mediates the relationship between impaired cognitive control for emotional information and depressive symptoms: A prospective study in remitted depressed adults. Behaviour Research & Therapy50, 292-297 (2012).
De Lissnyder, E. et al. Internal cognitive control in clinical depression: general but no emotion-specific impairments. Psychiatry Res199, 124-130, doi:10.1016/j.psychres.2012.04.019 (2012).
De Lissnyder, E. et al. Cognitive control moderates the association between stress and rumination. Journal of behavior therapy and experimental psychiatry43, 519-525, doi:10.1016/j.jbtep.2011.07.004 (2012).
De Lissnyder, E., Koster, E. H. W., Derakshan, N. & Raedt, R. D. The association between depressive symptoms and executive control impairments in response to emotional and non-emotional information. Cogn. Emot.24, 264-280 (2010).
Lo, B. C. Y., Lau, S., Cheung, S.-h. & Allen, N. B. The impact of rumination on internal attention switching. Cognition & Emotion26, 209-215, doi:10.1080/02699931.2011.574997 (2012).
Whitmer, A. J. & Banich, M. T. Inhibition versus switching deficits in different forms of rumination. Psychological Sciencevolume 18, 546-553 (2007).
Koster, E. H., De Lissnyder, E. & De Raedt, R. Rumination is characterized by valence-specific impairments in switching of attention. Acta Psychologica144, 563–570 (2013).
Meiran, N., Diamond, G. M., Toder, D. & Nemets, B. Cognitive rigidity in unipolar depression and obsessive compulsive disorder: Examination of task switching, Stroop, working memory updating and post-conflict adaptation. Psychiatry Res185, 149-156, doi:http://dx.doi.org/10.1016/j.psychres.2010.04.044 (2011).
Monsell, S. Task switching. Trends Cogn. Sci.7, 134-140 (2003).
Oberauer, K. Access to information in working memory: exploring the focus of attention. J Exp Psychol Learn Mem Cogn28, 411-421 (2002).
Unsworth, N. & Engle, R. W. Speed and accuracy of accessing information in working memory: an individual differences investigation of focus switching. Journal of Experimental Psychology: Learning, Memory, and Cognition34, 616-630, doi:10.1037/0278-7393.34.3.616 (2008).
Verhaeghen, P. & Hoyer, W. J. Aging, focus switching, and task switching in a continuous calculation task: evidence toward a new working memory control process. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition14, 22-39, doi:10.1080/138255890969357 (2007).
Joormann, J., Yoon, K. L. & Zetsche, U. Cognitive inhibition in depression. Applied and preventive psychology12, 128-139 (2007).
Joormann, J. in Cognitive limitations in aging and psychopathology. 275-312 (Cambridge University Press, 2005).
De Lissnyder, E., Koster, E. H. & De Raedt, R. Emotional Interference in Working Memory is Related to Rumination. Cognitive Therapy and Research36, 348-357, doi:10.1007/s10608-011-9352-4 (2012).
Wang, D., Kristjansson, A. & Nakayama, K. J. P. Efficient visual search without top-down or bottom-up guidance. J Perception Psychophysics67, 239-253, doi:10.3758/bf03206488 (2005).
Kristjansson, A. & Driver, J. Priming in visual search: separating the effects of target repetition, distractor repetition and role-reversal. Vision Res48, 1217-1232, doi:10.1016/j.visres.2008.02.007 (2008).
Lieberman, M. D. Social Cognitive Neuroscience: A Review of Core Processes. Annual Review of Psychology58, 259-289, doi:10.1146/annurev.psych.58.110405.085654 (2007).
Koster, E. H. & Hoorelbeke, K. Cognitive bias modification for depression. Current Opinion in Psychology4, 119-123, doi:http://dx.doi.org/10.1016/j.copsyc.2014.11.012 (2015).
Peckham, A. D., McHugh, R. K. & Otto, M. W. A meta-analysis of the magnitude of biased attention in depression. Depression & Anxiety27, 1135-1142 (2010).
Koch, I., Gade, M., Schuch, S. & Philipp, A. M. The role of inhibition in task switching: A review. Psychonomic Bulletin & Review17, 1-14, doi:10.3758/PBR.17.1.1 (2010).
Nolen-Hoeksema, S. Responses to depression and their effects of the duration of the episodes. Journal of Abnormal Psychology100, 569-582 (1991).
Beck, A. T., Steer, R. A. & Brown, G. K. Manual for the Beck Depression Inventory - II. (Psychological Corporation, 1996).
Nolen-Hoeksema, S. & Morrow, J. A prospective study of depression and posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology61, 115-121, doi:10.1037/0022-3514.61.1.115 (1991).
Barnett, P. A. & Gotlib, I. H. Psychosocial functioning and depression: Distinguishing among antecedents, concomitants, and consequences. Psychological bulletin104, 97 (1988).
Whitmer, A. J. & Gotlib, I. H. Brooding and Reflection Reconsidered: A Factor Analytic Examination of Rumination in Currently Depressed, Formerly Depressed, and Never Depressed Individuals. Cognit. Ther. Res.35, 99-107, doi:10.1007/s10608-011-9361-3 (2011).
Koster, E. H., De Lissnyder, E., Derakshan, N. & De Raedt, R. Understanding depressive rumination from a cognitive science perspective: the impaired disengagement hypothesis. Clin Psychol Rev31, 138-145, doi:10.1016/j.cpr.2010.08.005 (2011).
Eizenman, M. et al. A naturalistic visual scanning approach to assess selective attention in major depressive disorder. Psychiatry Res118, 117-128 (2003).
Sanchez, A., Vazquez, C., Marker, C., LeMoult, J. & Joormann, J. Attentional disengagement predicts stress recovery in depression: An eye-tracking study. J Abnorm Psychol122, 303 (2013).
Cooper, R. M. & Langton, S. R. Attentional bias to angry faces using the dot-probe task? It depends when you look for it. Behaviour Research & Therapy44, 1321-1329, doi:10.1016/j.brat.2005.10.004 (2006).
Leyman, L., De Raedt, R., Schacht, R. & Koster, E. H. Attentional biases for angry faces in unipolar depression. Psychological medicine37, 393-402, doi:10.1017/s003329170600910x (2007).
Philippot, P. & Brutoux, F. Induced rumination dampens executive processes in dysphoric young adults. Journal of behavior therapy and experimental psychiatry39, 219-227 (2008).
Joormann, J., Nee, D. E., Berman, M. G., Jonides, J. & Gotlib, I. H. Interference resolution in major depression. Cogn Affect Behav Neurosci10, 21-33, doi:10.3758/cabn.10.1.21 (2010).
Whitmer, A. J. & Gotlib, I. H. An attentional scope model of rumination. Psychological Bulletin139, 1036-1061, doi:10.1037/a0030923 (2013).
Piguet, C. et al. Neural substrates of cognitive switching and inhibition in a face processing task. NeuroImage82, 489-499, doi:https://doi.org/10.1016/j.neuroimage.2013.06.015 (2013).
Vuilleumier, P. How brains beware: neural mechanisms of emotional attention. Trends Cogn. Sci.9, 585-594 (2005).
Joormann, J., Dkane, M. & Gotlib, I. H. Adaptive and maladaptive components of rumination? Diagnostic specificity and relation to depressive biases. Behavior therapy37, 269-280 (2006).
Koster, E. H., Fox, E. & MacLeod, C. Introduction to the special section on cognitive bias modification in emotional disorders. J Abnorm Psychol118, 1-4, doi:10.1037/a0014379 (2009).
Bradley, M. M. & Lang, P. J. Measuring emotion: The self-assessment manikin and the semantic differential. J. Behav. Ther. Exp. Psychiatry25, 49-59, doi:https://doi.org/10.1016/0005-7916(94)90063-9 (1994).

Footnote 1: According to the experimental design, the face images used in the present study should be classified into three dimensions, emotion type, age and gender. The existing databases of faces in China did not satisfy this requirement. Therefore, prior to the study, we established a face database. The database consisted of 1200 faces classified into 48 groups based on four dimensions of emotion (happy, neutral, sad, angry), age (child, youth, elder), gender (male or female) and skin color (yellow or white). All of the selected faces were assessed by 60 undergraduates with the Identification task (i.e., classifying the face in each of the abovementioned dimensions) and the Self-Assessment Manikin procedure i.e., rating the face from valence, arousal and dominance respectively by self-report in a 9-point rating scale, ⁶². All faces were front-view images downloaded from the internet excluding public figures such as politicians, celebrities, and scientists. All faces were adjusted to the same size of 6.5 cm×7.5 cm with an ellipse model and cropped to remove hair, collars and other surrounding details.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Dysfunction of External Switching and Backward Inhibition in Major Depressive Disorder：General or Emotional specific?

Status:

Version 1

Abstract

Figures

1 Introduction

2 Methods

2.1 Participants

2.2 Materials and Procedure

3 Results

3.1 Switch cost

3.1.1 Accuracy

3.1.2 RTs

3.2 BI effect

3.2.1 Accuracy

3.2.2 RT

3.3 Effect of rumination on switch cost and BI

4 Discussion

Declarations

Conflict of Interest

Author Contributions

Funding

Data Availability Statement

References

Footnotes

Additional Declarations

Status:

Version 1