Hostname: page-component-6898cfcb55-zkb6p Total loading time: 0 Render date: 2026-11-11T17:38:01.183Z Has data issue: false hasContentIssue false
  • English
  • Français

Associations between aberrant working memory-related neural activity and cognitive impairments in somatically healthy, remitted patients with mood disorders

Published online by Cambridge University Press:  13 April 2023

Julian MacoveanuOpen the ORCID record for Julian Macoveanu[Opens in a new window],
Jeff Zarp PetersenOpen the ORCID record for Jeff Zarp Petersen[Opens in a new window],
Patrick M. FisherOpen the ORCID record for Patrick M. Fisher[Opens in a new window],
Lars Vedel KessingOpen the ORCID record for Lars Vedel Kessing[Opens in a new window],
Gitte Moos KnudsenOpen the ORCID record for Gitte Moos Knudsen[Opens in a new window] and
Kamilla Woznica MiskowiakOpen the ORCID record for Kamilla Woznica Miskowiak[Opens in a new window]
Julian Macoveanu
Affiliation:
Neurocognition and Emotion in Affective Disorders (NEAD) Centre, Psychiatric Centre Copenhagen, Mental Health Services, Capital Region of Denmark, and Department of Psychology, University of Copenhagen, Copenhagen, Denmark
Jeff Zarp Petersen
Affiliation:
Neurocognition and Emotion in Affective Disorders (NEAD) Centre, Psychiatric Centre Copenhagen, Mental Health Services, Capital Region of Denmark, and Department of Psychology, University of Copenhagen, Copenhagen, Denmark
Patrick M. Fisher
Affiliation:
Neurobiology Research Unit and Center for Integrated Molecular Imaging, Rigshospitalet, Copenhagen, Denmark
Lars Vedel Kessing
Affiliation:
Department of Clinical Medicine, University of Copenhagen, Copenhagen, DenmarkCopenhagen Affective Disorder Research Centre (CADIC), Psychiatric Centre Copenhagen, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark
Gitte Moos Knudsen
Affiliation:
Neurobiology Research Unit and Center for Integrated Molecular Imaging, Rigshospitalet, Copenhagen, DenmarkDepartment of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark
Kamilla Woznica Miskowiak*
Affiliation:
Neurocognition and Emotion in Affective Disorders (NEAD) Centre, Psychiatric Centre Copenhagen, Mental Health Services, Capital Region of Denmark, and Department of Psychology, University of Copenhagen, Copenhagen, DenmarkCopenhagen Affective Disorder Research Centre (CADIC), Psychiatric Centre Copenhagen, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark
*
Corresponding author: Kamilla Woznica Miskowiak; Email: kamilla.miskowiak@regionh.dk
Rights & Permissions [Opens in a new window]

Abstract

Background

Persistent cognitive deficits are prevalent in patients with bipolar disorder (BD) and unipolar disorder (UD), but treatments effectively targeting cognition in these mood disorders are lacking. This is partly due to poor insight into the neuronal underpinnings of cognitive deficits.

Methods

The aim of this functional magnetic resonance imaging (fMRI) study was to investigate the neuronal underpinnings of working memory (WM)-related deficits in somatically healthy, remitted patients with BD or UD (n = 66) with cognitive and functional impairments compared to 38 healthy controls (HC). The participants underwent neuropsychological testing and fMRI, while performing a visuospatial and a verbal N-back WM paradigm.

Results

Relative to HC, patients exhibited hypo-activity across dorsolateral prefrontal cortex as well as frontal and parietal nodes of the cognitive control network (CCN) and hyper-activity in left orbitofrontal cortex within the default mode network (DMN) during both visuospatial and verbal WM performance. Verbal WM-related response in the left posterior superior frontal gyrus (SFG) within CCN was lower in patients and correlated positively with out-of-scanner executive function performance across all participants.

Conclusions

Our findings suggest that cognitive impairments across BD and UD are associated with insufficient recruitment of task-relevant regions in the CCN and down-regulation of task-irrelevant orbitofrontal activity within the DMN during task performance. Specifically, a lower recruitment of the left posterior SFG within CCN during verbal WM was associated with lower cognitive performance.

Keywords

Bipolar disordercognitive impairmentdepressionfMRI

Information

Type
Original Article
Information
Psychological Medicine , Volume 53 , Issue 15 , November 2023, pp. 7203 - 7213
Check for updates
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press

Background

Enduring cognitive impairment even after remission is prevalent across bipolar disorder (BD) and unipolar disorder (UD) and results in functional disability (Depp et al., Reference Depp, Mausbach, Harmell, Savla, Bowie, Harvey and Patterson2012; Jaeger, Berns, Loftus, Gonzalez, & Czobor, Reference Jaeger, Berns, Loftus, Gonzalez and Czobor2007; McIntyre et al., Reference McIntyre, Cha, Soczynska, Woldeyohannes, Gallaugher, Kudlow and Baskaran2013; Torrent et al., Reference Torrent, Martinez-Arán, del Mar Bonnin, Reinares, Daban, Solé and Vieta2012; Tse, Chan, Ng, & Yatham, Reference Tse, Chan, Ng and Yatham2014). Cognition thus constitutes a central treatment target to advance functional recovery. However, the neural correlates of these deficits remain unclear, which impedes discovery of novel pro-cognitive interventions (Miskowiak et al., Reference Miskowiak, Burdick, Martinez-Aran, Bonnin, Bowie, Carvalho and Vieta2017). A recent systematic review of functional magnetic resonance imaging (fMRI) studies emphasized aberrant (predominantly hypo-) activity in a medial and dorsal cognitive control network (CCN) and failure to deactivate the default mode network (DMN) as the most consistent neural correlates of cognitive impairment in mood disorders (Miskowiak & Petersen, Reference Miskowiak and Petersen2019). Nevertheless, findings were inconsistent with respect to the regions affected and direction of aberrant activity in patients v. healthy controls (HC). In particular, there seems to be more evidence for hypo-activation in CCN regions like the dorsal prefrontal cortex (dPFC) in patients with BD, while patients with UD (with no objective cognitive performance decline) show dPFC hyper-activity (Miskowiak & Petersen, Reference Miskowiak and Petersen2019). This pattern of findings highlights task-relevant CCN hypo-activity to underlie reduced cognitive performance in patients with detectable deficits (Zarp Petersen et al., Reference Zarp Petersen, Varo, Skovsen, Ott, Kjærstad, Vieta and Miskowiak2022) as more often the case in patients with BD (Gualtieri & Morgan, Reference Gualtieri and Morgan2008). In this way, a putative reason for previous inter-diagnostic neural activity inconsistency could be that patients with UD do not present with the same magnitude of cognitive impairment as patients with BD at a group level. At the same time, another reason could be that fMRI studies generally not account for the cognitive or functional status of the investigated patients. Nevertheless, a growing body of literature documents substantial cognitive heterogeneity across BD and UD (Bora et al., Reference Bora, Hıdıroğlu, Özerdem, Kaçar, Sarısoy, Arslan and Atalay2016; Burdick et al., Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014; Gualtieri & Morgan, Reference Gualtieri and Morgan2008; Jensen, Knorr, Vinberg, Kessing, & Miskowiak, Reference Jensen, Knorr, Vinberg, Kessing and Miskowiak2016; Kjaerstad, Eikeseth, Vinberg, Kessing, & Miskowiak, Reference Kjaerstad, Eikeseth, Vinberg, Kessing and Miskowiak2019; Pu, Noda, Setoyama, & Nakagome, Reference Pu, Noda, Setoyama and Nakagome2018). Hence, comparison of neural activity in (cognitively impaired and cognitively normal) patients and HC hamper detection of brain changes that give rise to cognitive deficits. To address this challenge, we compared neural activity patterns between subgroups of remitted BD patients neuropsychologically defined as ‘cognitively impaired’ or ‘cognitively normal’ based on hierarchical cluster analysis. Our findings revealed CCN hypo-activity, including dorsolateral prefrontal cortex (dlPFC), and DMN hyper-activity during N-back working memory (WM) performance in the cognitively impaired v. cognitively normal subgroup (Zarp Petersen et al., Reference Zarp Petersen, Varo, Skovsen, Ott, Kjærstad, Vieta and Miskowiak2022). However, it remains to be explored whether these abnormalities can be replicated across BD and UD and whether they scale with cognitive and functional difficulties. Additionally, investigation of the precise illness-related neurobiological basis of cognitive impairment in mood disorders has been impeded by confounding effects of mood symptoms and comorbidity with somatic disorders as well as alcohol- and substance use disorders (Cerullo & Strakowski, Reference Cerullo and Strakowski2007; Davis, Uezato, Newell, & Frazier, Reference Davis, Uezato, Newell and Frazier2008).

In this fMRI study, we therefore investigated neuronal underpinnings of cognitive and functional impairments in remitted somatically healthy patients with BD or UD, who presented with clinically relevant cognitive impairments and reduced functioning.

Aims of the study

The aims were to: (i) map WM-related functional abnormalities in our patient sample compared to matched HC, and (ii) investigate whether WM-related activity in regions showing significant group differences were associated with the severity of cognitive impairment and functional disability. Based on emerging evidence, we hypothesized that our sample of cognitively impaired patients would display reduced WM-related activity within the CCN and increased activity within the DMN during visuospatial and verbal WM relative to HC, and that such aberrant neural activity would correlate with cognitive performance deficits and functional disability as measured outside the scanner.

Methods

Participants

The patient sample was recruited as part of our randomized, controlled intervention trial with erythropoietin (EPO) to target cognitive impairment in patients with mood disorders, Prefrontal Target Engagement as a biomarker model for Cognitive improvement – Erythropoietin (PRETEC-EPO) trial (Petersen et al., Reference Petersen, Schmidt, Vinberg, Jørgensen, Hageman, Ehrenreich and Miskowiak2018). Baseline fMRI and neuropsychological test data from 49% of participants in this report (patients: 21/66; HC: 30/38) were also included in our recent study investigating neural underpinnings of cognitive impairment in BD (Zarp Petersen et al., Reference Zarp Petersen, Varo, Skovsen, Ott, Kjærstad, Vieta and Miskowiak2022).

Inclusion criteria comprised: 18–65 years of age, ICD-10 diagnosis of BD or recurrent UD confirmed with Schedules for Clinical Assessment in Neuropsychiatry (SCAN version 2.1) (Wing et al., Reference Wing, Babor, Brugha, Burke, Cooper, Giel and Sartorius1990), partial or full remission [⩽14 or ⩽7 on Hamilton Depression Rating Scale 17-items (HDRS-17; Hamilton, Reference Hamilton1960) and Young Mania Rating Scale (YMRS; Young, Biggs, Ziegler, & Meyer, Reference Young, Biggs, Ziegler and Meyer1978)], fluent in Danish language, and objective cognitive impairment [total score ⩽77 or ⩾1 standard deviations (s.d.) below the norm on ⩾2/5 sub-tests on Screen for Cognitive Impairment in Psychiatry (SCIP) (Jensen et al., Reference Jensen, Støttrup, Nayberg, Knorr, Ullum, Purdon and Miskowiak2015; Purdon, Reference Purdon2005)]. Main exclusion criteria included significant medical conditions (including diabetes and cardiovascular disease), comorbid schizophrenia/schizoaffective disorder, current alcohol/substance abuse disorder, neurological disorder (including dementia), and history of severe head trauma. Maximum daily use of benzodiazepines allowed was 22.5 mg oxazepam (access ClinicalTrials.gov: NCT03315897 for all exclusion criteria related to ensuring safety throughout the EPO treatment course).

Neuroimaging and neuropsychological test data were included from age-, gender-, and intelligence-matched HC from the Bipolar Illness Onset (BIO) study (Kessing et al., Reference Kessing, Munkholm, Faurholt-Jepsen, Miskowiak, Nielsen, Frikke-Schmidt and Vinberg2017). The HC had no personal or first-degree family history of psychiatric disorder, neurological illness (including dementia), personal history of relevant medical illness, nor current alcohol or substance abuse disorder as verified through SCAN (Wing et al., Reference Wing, Babor, Brugha, Burke, Cooper, Giel and Sartorius1990) and a thorough clinical interview.

The study was approved by the Danish Research Ethics Committee for the Capital Region of Denmark (PreTEC-EPO protocol no.: H-16043370; BIO protocol no.: H-7-2014-007) and The Danish Data Protection Agency Capital Region of Denmark (PreTEC-EPO protocol no.: RHP-2017-020; BIO protocol no.: RHP-2015-023). Written consent was obtained from all participants.

Procedures

The fMRI scan and neuropsychological assessment took place at Copenhagen Affective Disorder Research Centre, Psychiatric Centre Copenhagen, and Copenhagen University Hospital, Rigshospitalet, respectively (Petersen et al., Reference Petersen, Schmidt, Vinberg, Jørgensen, Hageman, Ehrenreich and Miskowiak2018). The scanning acquisition protocol can be found in the online Supplementary methods section. Most participants (98%) were neuropsychologically assessed and fMRI scanned 0–3 days apart (days, mean ± s.d.: 1.5 ± 1.2).

Measures

Objective cognitive functioning

For this study, we assessed cognition with a comprehensive neuropsychological test battery including Rey Auditory Verbal Learning Test (RAVLT) (Schmidt, Reference Schmidt1996), Trail Making Test Part A & B (TMT-A, TMT-B) (Battery, Reference Battery1944), The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) Digit Span & Coding (Randolph, Tierney, Mohr, & Chase, Reference Randolph, Tierney, Mohr and Chase1998), Wechsler Adult Intelligence Scale Third edition (WAIS-III) Letter-Number Sequencing (Wechsler, Reference Wechsler1997), verbal fluency (letters ‘S’, ‘D’) (Borkowski, Benton, & Spreen, Reference Borkowski, Benton and Spreen1967), and the Rapid Visual (Information) Processing (RVP) and Spatial Working Memory (SWM) tests, respectively, from the Cambridge Neuropsychological Test Automated Battery (CANTAB®, Cambridge Cognition Ltd.). We estimated verbal intelligence using the Danish Adult Reading Task (DART) (Nelson & Willison, Reference Nelson and Willison1991).

Level of functioning

Interview-based level of functioning was assessed with the Functioning Assessment Short Test (FAST) (Rosa et al., Reference Rosa, Sánchez-Moreno, Martínez-Aran, Salamero, Torrent, Reinares and Vieta2007), which contains 24 items that cover the following six functional domains: autonomy, occupational functioning, cognitive functioning, financial issues, interpersonal relationships, and leisure time.

Visuospatial N-back fMRI task

During fMRI, participants first completed a visuospatial-variant of the N-back WM task with two levels of WM load (1-back, and 2-back) and a control condition (0-back). Each of the three conditions was performed in blocks where a yellow dot appeared in a sequence of 14 random locations in a 5 × 5 grid for 300 ms followed by an empty grid displayed for 1200 ms. In 1- and 2-back conditions, participants had to indicate with a right index finger button press on an MRI-compatible press-pad whenever the current dot appeared in the same field as 1 or 2 steps earlier, respectively (target stimulus). In the 0-back condition, patients were instructed to indicate each time the dot appeared in one of the four grid corners. The condition blocks included an average of three target trials and were shown five times each (15 blocks in total) interleaved by an 8 s fixation cross. Total task length was 7 min and 35 s.

Verbal N-back fMRI task

The verbal letter variant of the N-back task included three levels of WM load (1-back, 2-back, and 3-back) and a control condition (0-back). In this task, participants viewed a series of consonant letters (b, B, d, D, p, P, t, T, v, V) and had to indicate with a right-hand index finger button press whenever a letter stimulus matched the target stimulus from N (i.e. 1, 2, or 3) steps back in the sequence. The 0-back condition was a sensorimotor control task requiring participants to respond when they saw the letter x/X. Each condition was presented in four blocks in a fixed pseudo-random order preceded by instruction screens presented for 12 s. Each block contained 10 stimuli (3 target stimuli) each displayed for 0.5 s with a fixed interstimulus interval of 1.5 s. Blocks were separated by a 5000 ms fixation cross. Task length was 9 min and 52 s.

Across both WM paradigms, response time latency and accuracy were recorded with ePrime 2.0 (Psychological Software Tools, Pittsburgh, PA, USA).

Statistical procedures

Behavioral data were analyzed with IBM Statistical Package for the Social Sciences (SPSS), version 28 (IBM Corporation, Armonk, NY, USA) with α-level = 0.05. Visual inspection of Q-Q plots and Shapiro–Wilk test (Shapiro & Wilk, Reference Shapiro and Wilk1965) was carried out to assess data normality distribution for each variable included in SPSS analyses (normally distributed data: p ⩾ 0.05; non-normally distributed: p < 0.05). Independent samples t test, Pearson's χ2, and Mann–Whitney U test were conducted to assess group differences in demographic and clinical characteristics as well as cognitive performance and functioning.

Neuropsychological test data

We z-transformed patients' neuropsychological test raw scores using means and s.d. from HC. Standardized scores were averaged and classified within three domain-based composite scores: attention and psychomotor speed, verbal learning and memory, and WM and executive functioning (online Supplementary Table S1). Additionally, these domain composite scores were averaged to calculate an overall global cognitive composite score. The z scores for TMT-A, TMT-B, RVP A’, RVP Mean Latency, SWM Strategy, and SWM Between Errors were inversed (*−1) to ensure uni-directionality across all of the standardized cognitive test variables.

fMRI data analysis

Definition of the region of interest anatomical masks: Based on our hypotheses, we constructed two volumes of interest (VOI) masks for (i) the DMN, and (ii) the dorsal CCN, respectively, in accordance with prior neuroimaging findings (Emch, Von Bastian, & Koch, Reference Emch, Von Bastian and Koch2019; Owen, McMillan, Laird, & Bullmore, Reference Owen, McMillan, Laird and Bullmore2005) and consistent with our previous methodological approach (Petersen et al., Reference Petersen, Macoveanu, Kjærstad, Knudsen, Kessing and Miskowiak2021; Zarp Petersen et al., Reference Zarp Petersen, Varo, Skovsen, Ott, Kjærstad, Vieta and Miskowiak2022). The following regions were included in the DMN: hippocampal complex, cingulate gyrus (posterior division), frontal medial cortex. The CCN mask included postcentral gyrus, precuneus, angular gyrus, supramarginal gyrus (anterior and posterior divisions), parietal operculum cortex, superior parietal lobule, superior frontal gyrus (SFG), and middle frontal gyrus. Regions included in both VOI masks were based on the Harvard-Oxford Cortical Structural Atlas (Desikan et al., Reference Desikan, Segonne, Fischl, Quinn, Dickerson, Blacker and Killiany2006), thresholded at 25%, and defined on the standard MNI template available in the FSL Package and constructed using FSLEyes Tool in FSL Version 6.01. In addition, based on previous reports, we hypothesized that dlPFC activity during N-back performance would be sensitive to changes in related cognitive processes (Emch et al., Reference Emch, Von Bastian and Koch2019; Owen et al., Reference Owen, McMillan, Laird and Bullmore2005). We therefore performed region-of-interest (ROI) analyses using the extracted BOLD signal change from (i) a 10 mm radius sphere centered on MNI coordinates: x = −44, y = 18, z = 22 for the verbal N-back task and (ii) 8 mm radius sphere centered on MNI coordinates: x = 40, y = 34, z = 29 for the visuospatial N-back task (Emch et al., Reference Emch, Von Bastian and Koch2019; Owen et al., Reference Owen, McMillan, Laird and Bullmore2005; Smith et al., Reference Smith, Browning, Conen, Smallman, Buchbjerg, Larsen and Deakin2018).

fMRI pre-processing and first-level analyses: Functional MRI data were analyzed using FMRI Expert Analysis Tool (FEAT) (Woolrich, Ripley, Brady, & Smith, Reference Woolrich, Ripley, Brady and Smith2001), part of FSL version 6.0.1 (Jenkinson, Beckmann, Behrens, Woolrich, & Smith, Reference Jenkinson, Beckmann, Behrens, Woolrich and Smith2012). We carried out standard pre-processing steps which comprised non-brain tissue removal, linear and non-linear registration to structural space, normalization to the Montreal Neurological Institute (MNI) standard space, motion correction, and spatial smoothing using a Gaussian Kernel of 5 mm FWHM. Geometric distortions related to the B0 field were corrected based on the obtained B0 field map. All subjects' registration to the MNI template were visually inspected for data quality assessment.

In the first-level/subject-based analysis, we modelled (i) the verbal N-back task using a general linear model (GLM) with four conditions: 0-back, 1-back, 2-back, and 3-back, respectively, which were convolved with double-gamma hemodynamic response function. These conditions were used to compute a WM-load contrast that reflected a linear increase in activation from 0- to 3-back, and (ii) the visuospatial N-back task using a GLM with three conditions: 0-back, 1-back, and 2-back that were convolved with double-gamma hemodynamic response function and used to compute two contrasts of interest: 2 > 0-back (general WM) and 2 > 1-back (high-load specific WM). Both tasks further included the six basic motion regressors in the GLM models.

ROI analysis: The predefined left and right ROIs were used to extract the BOLD signal change to WM-load and general WM for the verbal and visuospatial N-back tasks, respectively. Group differences in BOLD response were assessed using independent samples t tests implemented in SPSS.

fMRI second-level analyses: For the group-level analysis, we set up GLM models in FEAT and applied the FMRIB's Local Analysis of Mixed Effects (FLAME) estimation method (Woolrich, Behrens, Beckmann, Jenkinson, & Smith, Reference Woolrich, Behrens, Beckmann, Jenkinson and Smith2004). The GLM models included the contrasts of interest calculated for the first-level analysis within each N-back task. The search volume was limited to either the CCN or the DMN masks. The significance at cluster levels was set at p < 0.05, corrected for multiple comparisons with a cluster-forming threshold of Z = 2.57 (p < 0.005). Mean percentage BOLD signal change was extracted from clusters showing significant group differences in WM-related activity for illustrative purposes and post-hoc sensitivity analyses.

Post-hoc exploratory correlation analyses

To aid interpretation of the WM-related abnormalities identified in patients v. HC, we explored whether any of the regions showing group differences for the two n-back tasks (Tables 2 and 3) correlated with either the (i) WM and executive functions composite score, or (ii) level of functioning (FAST total score). For this, we set up a second-level analysis in FEAT as above which added these two scores as covariates in addition to the group factors. Upon the identification of regions correlating with a cognitive score, we performed a conjunction analysis with cluster-level inference (Nichols, Brett, Andersson, Wager, & Poline, Reference Nichols, Brett, Andersson, Wager and Poline2005) assessing regions showing both group differences and correlation with cognitive or functioning scores.

Post-hoc sensitivity analyses

We investigated whether significant effects of group on WM-related BOLD response were independent of mood symptoms by entering extracted BOLD responses in univariate GLM models adjusted for total HDRS and YMRS scores. Within the patient sample, we further explored using independent samples t tests possible differences in BOLD responses extracted from the above-mentioned clusters between patients with BD v. UD, medicated v. unmedicated, and treated v. not treated with either antipsychotic, antidepressants, anticonvulsants, or lithium, respectively.

Results

Demographic and clinical characteristics and group comparisons

Demographic, clinical, and cognitive data across groups are presented in Table 1. Behavioral and neuroimaging data from a total of N = 104 participants [n = 66 patients; BD, n = 47 (71%), UD, n = 19 (29%); n = 38 HC] were included in this study. Patients and HC were matched for age, sex distribution, and verbal IQ (ps ⩾ 0.1), but patients had fewer years of education and greater subsyndromal mood symptoms than HC (ps ⩽ 0.05). As expected, patients also performed worse than HC across all domain-based and global cognitive composite scores (all p < 0.001) and presented with greater functional disability [mean FAST total score above defined impairment-threshold total score >11 (Rosa et al., Reference Rosa, Sánchez-Moreno, Martínez-Aran, Salamero, Torrent, Reinares and Vieta2007) (p < 0.001)]. At the time of inclusion, ~75% of patients were actively employed or studying.

Table 1.

Demographic and clinical characteristics at baseline in patient v. HC group

BD, bipolar disorder; UD, unipolar disorder; M, mean; s.d., standard deviation; IQ, intelligence quotient; HDRS-17: Hamilton Depression Rating Scale 17-items version; YMRS, Young Mania Rating Scale; FAST, Functioning Assessment Short Test.

Notes: χ2 for categorical variables, Mann–Whitney for non-parametric data [median (IQR)], independent t tests for normally distributed data [M (s.d.)].

a Data missing for n = 2 patients.

b Employment status ‘Other’ included current sick leave, vocational rehabilitation program participation, receiving early retirement pension, or working in flexible job. *p < 0.05, ***p < 0.001 (two-tailed).

Seven patients had comorbid psychiatric disorder(s): n = 3 ADHD, n = 1 social anxiety/OCD, n = 2 anorexia nervosa, and n = 1 borderline personality disorder. Finally, most patients received psychotropic medication (80%) at the time of assessment (Table 1; online Supplementary Table S3).

Due to a technical issue during fMRI sessions, behavioral responses [discriminability index (d′) data] (Grier, Reference Grier1971) were only partially recorded by the ePrime software for 15 (of 66) patients across both tasks and 27 (of 31) HC for the visuospatial paradigm and n = 35 (of 38) HC for the verbal paradigm. In-scanner behavioral data are therefore not presented.

Working memory-related activations

Dorsolateral prefrontal cortices region-of-interest analyses. Patients displayed reduced WM-load response in left dlPFC ROI (BOLDM = 0.33, BOLDs.d. = 0.29) v. HC (BOLDM = 0.43, BOLDs.d. = 0.19) (t (100.37) = −2.12, p = 0.037). There were no statistically significant differences between patients and HC in right dlPFC ROI BOLD response to general or high-load-specific WM in the visuospatial paradigm (Fig. 1).

Figure 1.

Findings for the visuospatial N-back task. Activations to general WM (2-back > 0-back) in the predefined right dorsolateral prefrontal cortex (dlPFC) region and clusters showing significant differences (marked with *) between cognitively impaired patients relative to healthy controls. ROI: right dlPFC; cognitive control network clusters: left and right superior frontal gyri (SFG). Default mode network clusters: left orbitofrontal cortex (OFC). Error bars represent standard error of the mean (s.e.m.). Statistics, cluster size, peak Z values, and peak MNI coordinates are shown in Table 2.

CCN and DMN volume-of-interest analyses

Visuospatial N-back paradigm. The HC engaged an expected widespread fronto-parietal network during general N-back WM (2-back > 0-back) (Table 2). Within the CCN VOI, the patient group displayed reduced activation across WM task-positive frontal regions, including bilateral SFG compared with HC (Table 2, Fig. 1). Within the DMN VOI, patients showed higher activity in the left OFC (decreased deactivation) compared to HC during general N-back WM (Table 2, Fig. 1). No statistically significant group differences were observed for the high-load-specific WM condition (2-back > 1-back) in the VOI analyses.

Table 2.

Visuospatial n-back fMRI task

VOI, volume of interest; BA, Brodmann area; cluster size, number of voxels in the significant cluster; p, corrected p value of the cluster; x, y, z, MNI coordinates of local maxima; Z stat, max statistical Z values for voxel.

Peak cluster activation (corrected p < 0.005) to the ‘2-back > 0-back’ contrast and group differences within the predefined regions of interest (VOIs): the default mode network and cognitive control network.

Verbal N-back paradigm. The HC showed significant fronto-parietal activations to the WM-load contrast (Table 3). Within the CCN, patients displayed reduced activation compared to HC in the following fronto-parietal clusters: left posterior SFG (dorsomedial prefrontal cortex; dmPFC), left medial SFG, bilateral middle frontal gyrus (MFG) (dlPFC), right supramarginal gyrus (SMG) (posterior), and right precuneus compared to HC (Table 3, Fig. 2). Patients also showed higher WM-load activity compared to HC within the CCN VOI in left SMG (anterior) and left postcentral gyrus and higher activity in the left OFC (decreased deactivation) within the DMN VOI (Table 3, Fig. 2).

Figure 2.

Findings for the verbal N-back task. Activations to the WM-load contrast (linear increase in response from 0-back to 3-back) in the predefined left dorsolateral prefrontal cortex (dlPFC) region and clusters showing significant differences (marked with *) between cognitively impaired patients relative to healthy controls. ROI: left dlPFC; cognitive control network clusters: left medial and posterior superior frontal gyri (SFG), left and right middle frontal gyrus (MFG), supramarginal gyrus (SMG) (posterior), and precuneus; default mode network clusters: left supramarginal gyrus (anterior), left postcentral gyrus, and left orbitofrontal cortex (OFC). The encircled regions represent the left posterior SFG region identified by the conjunction analysis assessing regions showing both hypo-activity in patients and correlation with cognitive scores. Error bars represent standard error of the mean (s.e.m.). Statistics, cluster size, peak Z values, and peak MNI coordinates are shown in Table 3.

Table 3.

Verbal n-back fMRI task

VOI, volume of interest; BA, Broadman area; cluster size, number of voxels in the significant cluster; p, corrected p value of the cluster; x, y, z, MNI coordinates of local maxima; Z-stat, max statistical Z values for voxel.

Peak cluster activation (corrected p < 0.005) to increased working memory (WM) load and group differences within the predefined regions of interest (VOIs): the default mode network and cognitive control network.

a Region identified by the conjunction analysis assessing regions showing both hypoactivity in patients and correlation with cognitive scores.

Post-hoc exploratory correlation analyses between working memory-related activations and cognition and level of functioning

For the verbal n-back task, we identified a network consisting of six dorsal fronto-parietal clusters (Table 3) showing a significant positive correlation with WM-load and executive function domain composite performance scores. The conjunction analysis assessing regions showing both task-related hypo-activity in patients and a correlation between task-activity and cognitive scores identified one common cluster in the left posterior SFG region (Table 3, Fig. 2).

We found no significant correlations between WM-related activations in the visuospatial n-back task and either cognition or level of functioning scores.

Post-hoc sensitivity analysis of the significant effects of group on working memory-related activations

Post-hoc analysis of the extracted BOLD responses from the 12 clusters showing an effect of group on WM activations across both paradigms (Tables 2 and 3) adjusting for mood symptoms (HDRS and YMRS total scores) showed that the effect of group remained significant for all clusters (all p ⩽ 0.007), suggesting that the aberrant WM activations in patients were not dependent on subsyndromal symptoms.

There were no significant differences in WM-related activations across the 12 clusters when comparing patients with BD v. UD (all p ⩾ 0.2). Detailed statistical data on the effect of diagnosis and on the effect of medication on WM-related activations is presented in the online Supplementary material.

Discussion

This is the first fMRI study to investigate functional neural correlates of objectively verified cognitive and functional impairments across remitted patients with BD or UD with no somatic comorbidity. The study revealed consistent WM-related hypo-activity within left dlPFC and frontal and parietal nodes of the CCN, including dlPFC, dmPFC, SFG, MFG, SMG, and precuneus, and hyper-activity in OFC within the DMN across visuospatial and verbal N-back WM paradigms. Importantly, the observed lower left posterior SFG activity within the CCN correlated with lower executive function performance. None of the identified regions correlated with functional disability.

The finding of reduced CCN WM-related activity adds to existing literature demonstrating hypo-activity within similar neurocircuitries during cognitive performance in mood disorders (Frangou, Kington, Raymont, & Shergill, Reference Frangou, Kington, Raymont and Shergill2008; Joshi et al., Reference Joshi, Vizueta, Foland-Ross, Townsend, Bookheimer, Thompson and Altshuler2016; McKenna, Sutherland, Legenkaya, & Eyler, Reference McKenna, Sutherland, Legenkaya and Eyler2014; Penfold, Vizueta, Townsend, Bookheimer, & Altshuler, Reference Penfold, Vizueta, Townsend, Bookheimer and Altshuler2015; Townsend, Bookheimer, Foland-Ross, Sugar, & Altshuler, Reference Townsend, Bookheimer, Foland-Ross, Sugar and Altshuler2010; Weathers et al., Reference Weathers, Brotman, Deveney, Kim, Zarate, Fromm and Leibenluft2013). Notably, Alonso-Lana et al. (Reference Alonso-Lana, Goikolea, Bonnin, Sarró, Segura, Amann and McKenna2016) revealed dlPFC hypo-activity during N-back WM performance in euthymic BD patients classified as cognitively ‘impaired’ v. ‘preserved’ (based on pre-defined percentile cut-offs). These associations were verified in our hierarchical cluster analysis-based fMRI study demonstrating strong support for WM-related dorsal PFC hypo-activity and DMN hyper-activity, which correlated with greater executive deficits, in cognitively ‘impaired’ v. ‘normal’ patients with remitted BD (Zarp Petersen et al., Reference Zarp Petersen, Varo, Skovsen, Ott, Kjærstad, Vieta and Miskowiak2022). However, studies have also revealed hyper-activity or lack of activity differences between patients and HC (Fitzgerald et al., Reference Fitzgerald, Srithiran, Benitez, Daskalakis, Oxley, Kulkarni and Egan2008; Harvey et al., Reference Harvey, Fossati, Pochon, Levy, LeBastard, Lehéricy and Dubois2005; Norbury, Godlewska, & Cowen, Reference Norbury, Godlewska and Cowen2014; Walter, Wolf, Spitzer, & Vasic, Reference Walter, Wolf, Spitzer and Vasic2007). According to the putative model by Miskowiak and Petersen (Reference Miskowiak and Petersen2019) both hypo- and hyper-activity represent aberrant neural functioning with (i) hypo-activity reflecting impaired performance (i.e. reduced cognitive capacity) and (ii) hyper-activity reflecting preserved cognitive performance (i.e. reduced cortical efficiency). Accordingly, prior conflicting findings could be partially explained by the cognitive heterogeneity in these patient groups (Bora et al., Reference Bora, Hıdıroğlu, Özerdem, Kaçar, Sarısoy, Arslan and Atalay2016; Burdick et al., Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014; Gualtieri & Morgan, Reference Gualtieri and Morgan2008; Jensen et al., Reference Jensen, Knorr, Vinberg, Kessing and Miskowiak2016; Kjaerstad et al., Reference Kjaerstad, Eikeseth, Vinberg, Kessing and Miskowiak2019; Pu et al., Reference Pu, Noda, Setoyama and Nakagome2018). Indeed, since we controlled for cognitive functioning prior to enrolment, the current findings add further evidence to support the hypothesis that WM-related hypo-activity specifically underlies reduced cognitive performance in patients with detectable deficits. In particular, the identified associations between reduced left SFG activity and poorer cognitive performance are interesting because of a central role of this subregion in WM performance (including monitoring and manipulation of task target stimulus) (du Boisgueheneuc et al., Reference du Boisgueheneuc, Levy, Volle, Seassau, Duffau, Kinkingnehun and Dubois2006). Our finding of DMN hyper-activity (OFC) corroborates previous findings of DMN hyper-activity across BD and UD (Miskowiak & Petersen, Reference Miskowiak and Petersen2019) and that less DMN suppression is related to less successful goal-directed cognitive performance (Fernández-Corcuera et al., Reference Fernández-Corcuera, Salvador, Monté, Sarró, Goikolea, Amann and Vieta2013; Sheline et al., Reference Sheline, Barch, Price, Rundle, Vaishnavi, Snyder and Raichle2009; Weissman, Roberts, Visscher, & Woldorff, Reference Weissman, Roberts, Visscher and Woldorff2006). Indeed, the DMN is usually activated during internal self-referential activities, subserving depressive rumination, and deactivated during active engagement in externally oriented cognitive processes (Anticevic et al., Reference Anticevic, Cole, Murray, Corlett, Wang and Krystal2012; Sheline et al., Reference Sheline, Barch, Price, Rundle, Vaishnavi, Snyder and Raichle2009; Zhou et al., Reference Zhou, Chen, Shen, Li, Chen, Zhu and Yan2020). Interestingly, recent neuroimaging studies show effects of psychedelic substances (e.g. psilocybin or LSD) on reducing DMN connectivity (Carhart-Harris et al., Reference Carhart-Harris, Bolstridge, Day, Rucker, Watts, Erritzoe and Pilling2018; Madsen et al., Reference Madsen, Stenbæk, Arvidsson, Armand, Marstrand-Joergensen, Johansen and Fisher2021; Speth et al., Reference Speth, Speth, Kaelen, Schloerscheidt, Feilding, Nutt and Carhart-Harris2016), which deserves further investigation in future neuroimaging studies.

The exact pathophysiological mechanisms involved in the observed neural engagement deficits associated with cognitive impairment remain elusive. Nonetheless, this study provides novel insight suggesting that patients' deficits may be speculated to arise from a combination of (i) inefficient attentional capacity (manifest as task-relevant hypo-activity in CCN areas) and (ii) failure to disengage from irrelevant self-referential thoughts (manifest as hyper-activity in DMN areas) while engaging in external, cognitively demanding activities. Importantly, the detected correlation between verbal WM-related left posterior SFG activity and out-of-scanner WM and executive functions performance suggests that patients' cognitive impairments may originate from failures to activate relevant neural networks during active cognitive performance. It is therefore possible that targeting cognitive impairment through combining pharmacological with non-pharmacological interventions (e.g. cognitive remediation) would facilitate effects on brain function that would generate complementary and more robust effects than unimodal interventions (Miskowiak et al., Reference Miskowiak, Burdick, Martinez-Aran, Bonnin, Bowie, Carvalho and Vieta2017; Vieta & Torrent, Reference Vieta and Torrent2016). In fact, a study revealed that employed patients with UD facing continuous cognitive challenges at work exhibited superior cognitive benefits of vortioxetine treatment than unemployed patients (McIntyre et al., Reference McIntyre, Florea, Tonnoir, Loft, Lam and Christensen2017).

From a treatment implications perspective, our findings thus highlight the relevance of implementing intervention designs in future cognition trials that aid direct cognitive improvement through training of cognitive functions (and thereby facilitate enhancement of task-relevant CCN hypo-activity) in this subgroup of patients.

A strength of this study was the inclusion of patients that were cognitively and functionally impaired but without comorbid somatic (e.g. diabetes, cardiovascular disorder) or alcohol/substance abuse disorder; conditions known per se to have independent negative effects on cognition (Hadley et al., Reference Hadley, Zhang, Harris-Skillman, Alexopoulou, DeLuca and Pendlebury2022; Harrison et al., Reference Harrison, Ding, Tang, Siervo, Robinson, Jagger and Stephan2014; Ko et al., Reference Ko, Ridley, Bryce, Allott, Smith and Kamminga2021). Additionally, all patients were in full or partial remission, minimizing symptomatic mood-dependent confounding effects on the results. As a limitation, the cross-sectional study design hampered causal inferences between aberrant CCN and DMN activity and cognitive impairment. In addition, most patients received psychotropic medication at the time of assessment, which may have had non-specific effects on the identified neural activity patterns and cognitive functioning (Campbell et al., Reference Campbell, Boustani, Limbil, Ott, Fox, Maidment and Gulati2009; Dias et al., Reference Dias, Balanzá-Martinez, Soeiro-de-Souza, Moreno, Figueira, Machado-Vieira and Vieta2012), albeit this subject is controversial (Sanches, Bauer, Galvez, Zunta-Soares, & Soares, Reference Sanches, Bauer, Galvez, Zunta-Soares and Soares2015). Nevertheless, it is noteworthy that only 30% of patients received lithium, based on studies suggesting putative neuroprotective effects of lithium treatment (Kessing, Forman, & Andersen, Reference Kessing, Forman and Andersen2010; Kessing, Søndergård, Forman, & Andersen, Reference Kessing, Søndergård, Forman and Andersen2008). Another limitation was the incomplete d′ data, which hindered validation of in-scanner performance differences between groups. However, during fMRI scans, we monitored that participants continuously pressed the pad button in response to targets, thereby ensuring performance attainment throughout each sequence. A further limitation was that we did not collect data on psychotic symptom history that could have influenced the findings (Brandt et al., Reference Brandt, Eichele, Melle, Sundet, Server, Agartz and Andreassen2014). Finally, grouping BD and UD into one sample may be considered a limitation, particularly since most patients were diagnosed with BD (71%), which limits result generalization to UD. Nonetheless, post-hoc analyses revealed no significant differences in WM-related response between patients with BD or UD.

In conclusion, this fMRI study indicates that task-related CCN hypo-activity and DMN hyper-activity may represent key neural correlates of cognitive impairments in mood disorders. Multimodal treatment approaches targeting these abnormalities may produce synergistic effects and possibly constitute an important next step toward more effective treatments for cognitive impairment in this group of patients for whom available treatments are insufficient.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S0033291723000715

Acknowledgements

We wish to thank research assistants in the NEAD Group, CADIC, for helping with data collection for the current report.

Financial support

The study was supported by a 5-year Fellowship (grant no. R215-2015-4121) awarded to K. W. M. by the Lundbeck Foundation.

Conflict of interest

K. W. M. has received consultancy fees from Lundbeck and Janssen in the past 3 years. J. Z. P., J. M., and P. M. F. report no conflicts of interest. L. V. K. has within recent 3 years been a consultant for Lundbeck and Teva. G. M. K. has received honoraria as a speaker for Sage Biogen and as a consultant for Sanos.

Footnotes

*

Authors contributed equally as shared first authors.

References

Alonso-Lana, S., Goikolea, J. M., Bonnin, C. M., Sarró, S., Segura, B., Amann, B. L., … McKenna, P. J. (2016). Structural and functional brain correlates of cognitive impairment in euthymic patients with bipolar disorder. PLoS ONE, 11(7), e0158867.CrossRefGoogle ScholarPubMed
Anticevic, A., Cole, M. W., Murray, J. D., Corlett, P. R., Wang, X.-J., & Krystal, J. H. (2012). The role of default network deactivation in cognition and disease. Trends in Cognitive Sciences, 16(12), 584–592.CrossRefGoogle ScholarPubMed
Battery, A. I. T. (1944). Manual of directions and scoring. Washington, DC: War Department, Adjutant General's Office.Google Scholar
Bora, E., Hıdıroğlu, C., Özerdem, A., Kaçar, ÖF, Sarısoy, G., Arslan, F. C., … Atalay, A. (2016). Executive dysfunction and cognitive subgroups in a large sample of euthymic patients with bipolar disorder. European Neuropsychopharmacology, 26(8), 1338–1347.CrossRefGoogle Scholar
Borkowski, J. G., Benton, A. L., & Spreen, O. (1967). Word fluency and brain damage. Neuropsychologia, 5(2), 135–140.CrossRefGoogle Scholar
Brandt, C. L., Eichele, T., Melle, I., Sundet, K., Server, A., Agartz, I., … Andreassen, O. A. (2014). Working memory networks and activation patterns in schizophrenia and bipolar disorder: Comparison with healthy controls. The British Journal of Psychiatry, 204(4), 290–298.CrossRefGoogle ScholarPubMed
Burdick, K. E., Russo, M., Frangou, S., Mahon, K., Braga, R. J., Shanahan, M., & Malhotra, A. K. (2014). Empirical evidence for discrete neurocognitive subgroups in bipolar disorder: Clinical implications. Psychological Medicine, 44(14), 3083–3096.CrossRefGoogle ScholarPubMed
Campbell, N., Boustani, M., Limbil, T., Ott, C., Fox, C., Maidment, I., … Gulati, R. (2009). The cognitive impact of anticholinergics: A clinical review. Clinical Interventions in Aging, 4, 225–233. doi: 10.2147/cia.s5358.Google ScholarPubMed
Carhart-Harris, R. L., Bolstridge, M., Day, C., Rucker, J., Watts, R., Erritzoe, D., … Pilling, S. (2018). Psilocybin with psychological support for treatment-resistant depression: Six-month follow-up. Psychopharmacology, 235(2), 399–408.CrossRefGoogle ScholarPubMed
Cerullo, M. A., & Strakowski, S. M. (2007). The prevalence and significance of substance use disorders in bipolar type I and II disorder. Substance Abuse Treatment, Prevention, and Policy, 2(1), 1–9.CrossRefGoogle ScholarPubMed
Davis, L., Uezato, A., Newell, J. M., & Frazier, E. (2008). Major depression and comorbid substance use disorders. Current Opinion in Psychiatry, 21(1), 14–18.CrossRefGoogle ScholarPubMed
Depp, C. A., Mausbach, B. T., Harmell, A. L., Savla, G. N., Bowie, C. R., Harvey, P. D., & Patterson, T. L. (2012). Meta-analysis of the association between cognitive abilities and everyday functioning in bipolar disorder. Bipolar Disorders, 14(3), 217–226. doi: 10.1111/j.1399-5618.2012.01011.xCrossRefGoogle ScholarPubMed
Desikan, R. S., Segonne, F., Fischl, B., Quinn, B. T., Dickerson, B. C., Blacker, D., … Killiany, R. J. (2006). An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage, 31(3), 968–980. doi: 10.1016/j.neuroimage.2006.01.021CrossRefGoogle ScholarPubMed
Dias, V. V., Balanzá-Martinez, V., Soeiro-de-Souza, M., Moreno, R., Figueira, M., Machado-Vieira, R., & Vieta, E. (2012). Pharmacological approaches in bipolar disorders and the impact on cognition: A critical overview. Acta Psychiatrica Scandinavica, 126(5), 315–331.CrossRefGoogle ScholarPubMed
du Boisgueheneuc, F., Levy, R., Volle, E., Seassau, M., Duffau, H., Kinkingnehun, S., … Dubois, B. (2006). Functions of the left superior frontal gyrus in humans: A lesion study. Brain, 129(Pt 12), 3315–3328. doi: 10.1093/brain/awl244CrossRefGoogle ScholarPubMed
Emch, M., Von Bastian, C. C., & Koch, K. (2019). Neural correlates of verbal working memory: An fMRI meta-analysis. Frontiers in Human Neuroscience, 13, 180. doi: 10.3389/fnhum.2019.00180CrossRefGoogle ScholarPubMed
Fernández-Corcuera, P., Salvador, R., Monté, G. C., Sarró, S. S., Goikolea, J. M., Amann, B., … Vieta, E. (2013). Bipolar depressed patients show both failure to activate and failure to de-activate during performance of a working memory task. Journal of Affective Disorders, 148(2–3), 170–178.CrossRefGoogle ScholarPubMed
Fitzgerald, P. B., Srithiran, A., Benitez, J., Daskalakis, Z. Z., Oxley, T. J., Kulkarni, J., & Egan, G. F. (2008). An fMRI study of prefrontal brain activation during multiple tasks in patients with major depressive disorder. Human Brain Mapping, 29(4), 490–501.CrossRefGoogle ScholarPubMed
Frangou, S., Kington, J., Raymont, V., & Shergill, S. S. (2008). Examining ventral and dorsal prefrontal function in bipolar disorder: A functional magnetic resonance imaging study. European Psychiatry, 23(4), 300–308.CrossRefGoogle ScholarPubMed
Grier, J. B. (1971). Nonparametric indexes for sensitivity and bias: Computing formulas. Psychological Bulletin, 75, 424–429. doi: 10.1037/h0031246CrossRefGoogle ScholarPubMed
Gualtieri, C. T., & Morgan, D. W. (2008). The frequency of cognitive impairment in patients with anxiety, depression, and bipolar disorder: An unaccounted source of variance in clinical trials. Journal of Clinical Psychiatry, 69(7), 1122–1130.CrossRefGoogle ScholarPubMed
Hadley, G., Zhang, J., Harris-Skillman, E., Alexopoulou, Z., DeLuca, G. C., & Pendlebury, S. T. (2022). Cognitive decline and diabetes: A systematic review of the neuropathological correlates accounting for cognition at death. Journal of Neurology, Neurosurgery & Psychiatry, 93(3), 246–253.CrossRefGoogle ScholarPubMed
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, and Psychiatry, 23(1), 56. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/14399272.CrossRefGoogle ScholarPubMed
Harrison, S. L., Ding, J., Tang, E. Y., Siervo, M., Robinson, L., Jagger, C., & Stephan, B. C. (2014). Cardiovascular disease risk models and longitudinal changes in cognition: A systematic review. PLoS ONE, 9(12), e114431.CrossRefGoogle ScholarPubMed
Harvey, P.-O., Fossati, P., Pochon, J.-B., Levy, R., LeBastard, G., Lehéricy, S., … Dubois, B. (2005). Cognitive control and brain resources in major depression: An fMRI study using the n-back task. Neuroimage, 26(3), 860–869.CrossRefGoogle ScholarPubMed
Jaeger, J., Berns, S., Loftus, S., Gonzalez, C., & Czobor, P. (2007). Neurocognitive test performance predicts functional recovery from acute exacerbation leading to hospitalization in bipolar disorder. Bipolar Disorders, 9(1–2), 93–102.CrossRefGoogle ScholarPubMed
Jenkinson, M., Beckmann, C. F., Behrens, T. E., Woolrich, M. W., & Smith, S. M. (2012). Fsl. Neuroimage, 62(2), 782–790.CrossRefGoogle ScholarPubMed
Jensen, J. H., Knorr, U., Vinberg, M., Kessing, L. V., & Miskowiak, K. W. (2016). Discrete neurocognitive subgroups in fully or partially remitted bipolar disorder: Associations with functional abilities. Journal of Affective Disorders, 205, 378–386.CrossRefGoogle ScholarPubMed
Jensen, J. H., Støttrup, M. M., Nayberg, E., Knorr, U., Ullum, H., Purdon, S. E., … Miskowiak, K. W. (2015). Optimising screening for cognitive dysfunction in bipolar disorder: Validation and evaluation of objective and subjective tools. Journal of Affective Disorders, 187, 10–19.CrossRefGoogle ScholarPubMed
Joshi, S. H., Vizueta, N., Foland-Ross, L., Townsend, J. D., Bookheimer, S. Y., Thompson, P. M., … Altshuler, L. L. (2016). Relationships between altered functional magnetic resonance imaging activation and cortical thickness in patients with euthymic bipolar I disorder. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 1(6), 507–517.Google ScholarPubMed
Kessing, L. V., Forman, J. L., & Andersen, P. K. (2010). Does lithium protect against dementia? Bipolar Disorders, 12(1), 87–94.CrossRefGoogle ScholarPubMed
Kessing, L. V., Munkholm, K., Faurholt-Jepsen, M., Miskowiak, K. W., Nielsen, L. B., Frikke-Schmidt, R., … Vinberg, M. (2017). The bipolar illness onset study: Research protocol for the BIO cohort study. BMJ Open, 7(6), e015462. doi: 10.1136/bmjopen-2016-015462CrossRefGoogle ScholarPubMed
Kessing, L. V., Søndergård, L., Forman, J. L., & Andersen, P. K. (2008). Lithium treatment and risk of dementia. Archives of General Psychiatry, 65(11), 1331–1335.CrossRefGoogle ScholarPubMed
Kjaerstad, H. L., Eikeseth, F. F., Vinberg, M., Kessing, L. V., & Miskowiak, K. W. (2019). Neurocognitive heterogeneity in patients newly diagnosed with bipolar disorder and their unaffected relatives: Associations with emotional cognition. Psychological Medicine, 2021 Mar, 51(4):668–679. doi: 10.1017/S0033291719003738. Epub 2019 Dec 26. PMID: 31875793.CrossRefGoogle ScholarPubMed
Ko, K. Y., Ridley, N., Bryce, S. D., Allott, K., Smith, A., & Kamminga, J. (2021). Screening tools for cognitive impairment in adults with substance use disorders: A systematic review. Journal of the International Neuropsychological Society, 28(7), 1–24.Google ScholarPubMed
Madsen, M. K., Stenbæk, D. S., Arvidsson, A., Armand, S., Marstrand-Joergensen, M. R., Johansen, S. S., … Fisher, P. M. (2021). Psilocybin-induced changes in brain network integrity and segregation correlate with plasma psilocin level and psychedelic experience. European Neuropsychopharmacology, 50, 121–132.CrossRefGoogle ScholarPubMed
McIntyre, R. S., Cha, D. S., Soczynska, J. K., Woldeyohannes, H. O., Gallaugher, L. A., Kudlow, P., … Baskaran, A. (2013). Cognitive deficits and functional outcomes in major depressive disorder: Determinants, substrates, and treatment interventions. Depression and Anxiety, 30(6), 515–527.CrossRefGoogle ScholarPubMed
McIntyre, R. S., Florea, I., Tonnoir, B., Loft, H., Lam, R. W., & Christensen, M. C. (2017). Efficacy of vortioxetine on cognitive functioning in working patients with major depressive disorder. The Journal of Clinical Psychiatry, 78(1), 22276.CrossRefGoogle ScholarPubMed
McKenna, B. S., Sutherland, A. N., Legenkaya, A. P., & Eyler, L. T. (2014). Abnormalities of brain response during encoding into verbal working memory among euthymic patients with bipolar disorder. Bipolar Disorders, 16(3), 289–299.CrossRefGoogle ScholarPubMed
Miskowiak, K. W., Burdick, K. E., Martinez-Aran, A., Bonnin, C. M., Bowie, C. R., Carvalho, A. F., … Vieta, E. (2017). Methodological recommendations for cognition trials in bipolar disorder by the international society for bipolar disorders targeting cognition task force. Bipolar Disorders, 19(8), 614–626.CrossRefGoogle ScholarPubMed
Miskowiak, K. W., & Petersen, C. S. (2019). Neuronal underpinnings of cognitive impairment and-improvement in mood disorders. CNS Spectrums, 24(1), 30–53.CrossRefGoogle ScholarPubMed
Nelson, H. E., & Willison, J. (1991). National adult reading test (NART) (pp. 1–26). Windsor: Nfer-Nelson.Google Scholar
Nichols, T., Brett, M., Andersson, J., Wager, T., & Poline, J. B. (2005). Valid conjunction inference with the minimum statistic. Neuroimage, 25(3), 653–660.CrossRefGoogle ScholarPubMed
Norbury, R., Godlewska, B., & Cowen, P. J. (2014). When less is more: A functional magnetic resonance imaging study of verbal working memory in remitted depressed patients. Psychological Medicine, 44(6), 1197–1203.CrossRefGoogle Scholar
Owen, A. M., McMillan, K. M., Laird, A. R., & Bullmore, E. (2005). N-back working memory paradigm: A meta-analysis of normative functional neuroimaging studies. Human Brain Mapping, 25(1), 46–59.CrossRefGoogle ScholarPubMed
Penfold, C., Vizueta, N., Townsend, J. D., Bookheimer, S. Y., & Altshuler, L. L. (2015). Frontal lobe hypoactivation in medication-free adults with bipolar II depression during response inhibition. Psychiatry Research: Neuroimaging, 231(3), 202–209.CrossRefGoogle ScholarPubMed
Petersen, J. Z., Macoveanu, J., Kjærstad, H. L., Knudsen, G. M., Kessing, L. V., & Miskowiak, K. W. (2021). Assessment of the neuronal underpinnings of cognitive impairment in bipolar disorder with a picture encoding paradigm and methodological lessons learnt. Journal of Psychopharmacology, 35(8), 983–991.CrossRefGoogle ScholarPubMed
Petersen, J. Z., Schmidt, L. S., Vinberg, M., Jørgensen, M. B., Hageman, I., Ehrenreich, H., … Miskowiak, K. W. (2018). Effects of recombinant human erythropoietin on cognition and neural activity in remitted patients with mood disorders and first-degree relatives of patients with psychiatric disorders: A study protocol for a randomized controlled trial. Trials, 19, 1–14.CrossRefGoogle ScholarPubMed
Pu, S., Noda, T., Setoyama, S., & Nakagome, K. (2018). Empirical evidence for discrete neurocognitive subgroups in patients with non-psychotic major depressive disorder: Clinical implications. Psychological Medicine, 48(16), 2717–2729.CrossRefGoogle ScholarPubMed
Purdon, S. E. (2005). The screen for cognitive impairment in psychiatry (SCIP): Instructions and three alternate forms. Edmonton, Alberta, Canada: PNL Inc.Google Scholar
Randolph, C., Tierney, M. C., Mohr, E., & Chase, T. N. (1998). The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): Preliminary clinical validity. Journal of Clinical and Experimental Neuropsychology, 20(3), 310–319. doi: 10.1076/jcen.20.3.310.823CrossRefGoogle ScholarPubMed
Rosa, A. R., Sánchez-Moreno, J., Martínez-Aran, A., Salamero, M., Torrent, C., Reinares, M., … Vieta, E. (2007). Validity and reliability of the Functioning Assessment Short Test (FAST) in bipolar disorder. Clinical Practice and Epidemiology in Mental Health, 3(1), 1–8.CrossRefGoogle ScholarPubMed
Sanches, M., Bauer, I. E., Galvez, J. F., Zunta-Soares, G. B., & Soares, J. C. (2015). The management of cognitive impairment in bipolar disorder: Current status and perspectives. American Journal of Therapeutics, 22(6), 477.CrossRefGoogle Scholar
Schmidt, M. (1996). Rey auditory verbal learning test: A handbook (pp. 1996). Los Angeles, CA: Western Psychological Services.Google Scholar
Shapiro, S. S., & Wilk, M. B. (1965). An analysis of variance test for normality (complete samples). Biometrika, 52(3/4), 591–611.CrossRefGoogle Scholar
Sheline, Y. I., Barch, D. M., Price, J. L., Rundle, M. M., Vaishnavi, S. N., Snyder, A. Z., … Raichle, M. E. (2009). The default mode network and self-referential processes in depression. Proceedings of the National Academy of Sciences, 106(6), 1942–1947.CrossRefGoogle ScholarPubMed
Smith, J., Browning, M., Conen, S., Smallman, R., Buchbjerg, J., Larsen, K., … Deakin, J. (2018). Vortioxetine reduces BOLD signal during performance of the N-back working memory task: A randomised neuroimaging trial in remitted depressed patients and healthy controls. Molecular Psychiatry, 23(5), 1127–1133.CrossRefGoogle ScholarPubMed
Speth, J., Speth, C., Kaelen, M., Schloerscheidt, A. M., Feilding, A., Nutt, D. J., & Carhart-Harris, R. L. (2016). Decreased mental time travel to the past correlates with default-mode network disintegration under lysergic acid diethylamide. Journal of Psychopharmacology, 30(4), 344–353.CrossRefGoogle Scholar
Torrent, C., Martinez-Arán, A., del Mar Bonnin, C., Reinares, M., Daban, C., Solé, B., … Vieta, E. (2012). Long-term outcome of cognitive impairment in bipolar disorder. The Journal of Clinical Psychiatry, 73(7), 10736.CrossRefGoogle ScholarPubMed
Townsend, J., Bookheimer, S. Y., Foland-Ross, L. C., Sugar, C. A., & Altshuler, L. L. (2010). fMRI abnormalities in dorsolateral prefrontal cortex during a working memory task in manic, euthymic and depressed bipolar subjects. Psychiatry Research: Neuroimaging, 182(1), 22–29.CrossRefGoogle ScholarPubMed
Tse, S., Chan, S., Ng, K. L., & Yatham, L. N. (2014). Meta-analysis of predictors of favorable employment outcomes among individuals with bipolar disorder. Bipolar Disorders, 16(3), 217–229.CrossRefGoogle ScholarPubMed
Vieta, E., & Torrent, C. (2016). Functional remediation: The pathway from remission to recovery in bipolar disorder. World Psychiatry, 15(3), 288–289.CrossRefGoogle ScholarPubMed
Walter, H., Wolf, R. C., Spitzer, M., & Vasic, N. (2007). Increased left prefrontal activation in patients with unipolar depression: An event-related, parametric, performance-controlled fMRI study. Journal of Affective Disorders, 101(1–3), 175–185.CrossRefGoogle ScholarPubMed
Weathers, J., Brotman, M. A., Deveney, C. M., Kim, P., Zarate, C. Jr., Fromm, S., … Leibenluft, E. (2013). A developmental study on the neural circuitry mediating response flexibility in bipolar disorder. Psychiatry Research: Neuroimaging, 214(1), 56–65.CrossRefGoogle Scholar
Wechsler, D. (1997). WAIS-III: Wechsler adult intelligence scale. Psychological Corporation.Google Scholar
Weissman, D. H., Roberts, K., Visscher, K., & Woldorff, M. (2006). The neural bases of momentary lapses in attention. Nature Neuroscience, 9(7), 971–978.CrossRefGoogle ScholarPubMed
Wing, J. K., Babor, T., Brugha, T. S., Burke, J., Cooper, J. E., Giel, R., … Sartorius, N. (1990). SCAN: Schedules for clinical assessment in neuropsychiatry. Archives of General Psychiatry, 47(6), 589–593.CrossRefGoogle ScholarPubMed
Woolrich, M. W., Behrens, T. E., Beckmann, C. F., Jenkinson, M., & Smith, S. M. (2004). Multilevel linear modelling for FMRI group analysis using Bayesian inference. Neuroimage, 21(4), 1732–1747.CrossRefGoogle ScholarPubMed
Woolrich, M. W., Ripley, B. D., Brady, M., & Smith, S. M. (2001). Temporal autocorrelation in univariate linear modeling of FMRI data. Neuroimage, 14(6), 1370–1386.CrossRefGoogle ScholarPubMed
Young, R. C., Biggs, J. T., Ziegler, V. E., & Meyer, D. A. (1978). A rating scale for mania: Reliability, validity and sensitivity. The British Journal of Psychiatry, 133(5), 429–435.CrossRefGoogle ScholarPubMed
Zarp Petersen, J., Varo, C., Skovsen, C. F., Ott, C. V., Kjærstad, H. L., Vieta, E., … Miskowiak, K. W. (2022). Neuronal underpinnings of cognitive impairment in bipolar disorder: A large data-driven functional magnetic resonance imaging study. Bipolar Disorders, 24(1), 69–81.CrossRefGoogle ScholarPubMed
Zhou, H. X., Chen, X., Shen, Y. Q., Li, L., Chen, N. X., Zhu, Z. C., … Yan, C. G. (2020). Rumination and the default mode network: Meta-analysis of brain imaging studies and implications for depression. Neuroimage, 206, 116287.CrossRefGoogle ScholarPubMed
Figure 0

Table 1.Demographic and clinical characteristics at baseline in patient v. HC group

Figure 1

Figure 1.Findings for the visuospatial N-back task. Activations to general WM (2-back > 0-back) in the predefined right dorsolateral prefrontal cortex (dlPFC) region and clusters showing significant differences (marked with *) between cognitively impaired patients relative to healthy controls. ROI: right dlPFC; cognitive control network clusters: left and right superior frontal gyri (SFG). Default mode network clusters: left orbitofrontal cortex (OFC). Error bars represent standard error of the mean (s.e.m.). Statistics, cluster size, peak Z values, and peak MNI coordinates are shown in Table 2.

Figure 2

Table 2.Visuospatial n-back fMRI task

Figure 3

Figure 2.Findings for the verbal N-back task. Activations to the WM-load contrast (linear increase in response from 0-back to 3-back) in the predefined left dorsolateral prefrontal cortex (dlPFC) region and clusters showing significant differences (marked with *) between cognitively impaired patients relative to healthy controls. ROI: left dlPFC; cognitive control network clusters: left medial and posterior superior frontal gyri (SFG), left and right middle frontal gyrus (MFG), supramarginal gyrus (SMG) (posterior), and precuneus; default mode network clusters: left supramarginal gyrus (anterior), left postcentral gyrus, and left orbitofrontal cortex (OFC). The encircled regions represent the left posterior SFG region identified by the conjunction analysis assessing regions showing both hypo-activity in patients and correlation with cognitive scores. Error bars represent standard error of the mean (s.e.m.). Statistics, cluster size, peak Z values, and peak MNI coordinates are shown in Table 3.

Figure 4

Table 3.Verbal n-back fMRI task

Supplementary material: File

Macoveanu et al. supplementary material
Download undefined(File)
File 17.7 KB