Advertisement

Medial Prefrontal Cortex Dysfunction Mediates Working Memory Deficits in Patients with Schizophrenia

  • John C. Williams
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, 11794
    Search for articles by this author
  • Zu Jie Zheng
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794
    Search for articles by this author
  • Philip N. Tubiolo
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, 11794
    Search for articles by this author
  • Jacob R. Luceno
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794
    Search for articles by this author
  • Roberto B. Gil
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Psychiatry, Columbia University Vagelos College of Physicians and Surgeons, New York-Presbyterian / Columbia University Irving Medical Center, New York, NY, 10032

    New York State Psychiatric Institute, New York, NY, 10032
    Search for articles by this author
  • Ragy R. Girgis
    Affiliations
    Department of Psychiatry, Columbia University Vagelos College of Physicians and Surgeons, New York-Presbyterian / Columbia University Irving Medical Center, New York, NY, 10032

    New York State Psychiatric Institute, New York, NY, 10032
    Search for articles by this author
  • Mark Slifstein
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Psychiatry, Columbia University Vagelos College of Physicians and Surgeons, New York-Presbyterian / Columbia University Irving Medical Center, New York, NY, 10032

    New York State Psychiatric Institute, New York, NY, 10032
    Search for articles by this author
  • Anissa Abi-Dargham
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, 11794

    Department of Psychiatry, Columbia University Vagelos College of Physicians and Surgeons, New York-Presbyterian / Columbia University Irving Medical Center, New York, NY, 10032

    New York State Psychiatric Institute, New York, NY, 10032
    Search for articles by this author
  • Jared X. Van Snellenberg
    Correspondence
    To whom correspondence may be addressed. , 101 Nicolls Rd., Health Sciences Center T10-087J, Stony Brook, NY 11794
    Affiliations
    Department of Psychiatry and Behavioral Health, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, 11794

    Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, 11794

    Department of Psychiatry, Columbia University Vagelos College of Physicians and Surgeons, New York-Presbyterian / Columbia University Irving Medical Center, New York, NY, 10032

    New York State Psychiatric Institute, New York, NY, 10032

    Department of Psychology, Stony Brook University, Stony Brook, NY, 11794
    Search for articles by this author
Open AccessPublished:October 25, 2022DOI:https://doi.org/10.1016/j.bpsgos.2022.10.003

      Abstract

      Background

      Schizophrenia (SCZ) is marked by working memory (WM) deficits, which predict poor functional outcome. While most functional magnetic resonance imaging (fMRI) studies of WM in SCZ focus on dorsolateral prefrontal cortex (dlPFC), recent work suggests medial PFC (mPFC) may play a role. We investigated whether task-evoked mPFC deactivation is associated with WM performance, and whether it mediates deficits in SCZ. Additionally, we investigated associations between mPFC deactivation and cortical dopamine release.

      Methods

      Patients with SCZ (N=41) and HC (N=40) performed a visual object n-back task during fMRI. Dopamine release capacity in mPFC was quantified with [11C]FLB457 in a subset of participants (9 SCZ, 14 HC) using an amphetamine challenge. Correlations between task-evoked deactivation and performance were assessed in mPFC and dlPFC masks, and was further examined for relationships with diagnosis and dopamine release.

      Results

      MPFC deactivation was associated with WM task performance, but dlPFC activation was not. Deactivation in mPFC was reduced in SCZ relative to HC and mediated the relationship between diagnosis and WM performance. Additionally, mPFC deactivation was significantly and inversely associated with dopamine release capacity across groups and in HCs alone, but not in patients.

      Conclusions

      Reduced WM task-evoked mPFC deactivation is a mediator of, and potential substrate for, WM impairment in SCZ, although our study design does not rule out that these findings could relate to cognition in general rather than WM specifically. We further present preliminary evidence of an inverse association between deactivation during WM tasks and dopamine release capacity in mPFC.

      Keywords

      Abbreviations:

      BOLD (blood-oxygen-level-dependent), BPND (binding potential relative to the non-displaceable compartment), dlPFC (dorsolateral prefrontal cortex), D2/3R (dopamine receptor D2/3), DF (drug free), DMN (default mode network), DN (drug naïve), FD (framewise displacement), fMRI (functional magnetic resonance imaging), FWER (family-wise error rate), HC (healthy control), HRF (hemodynamic response function), lPFC (lateral prefrontal cortex), mPFC (medial prefrontal cortex), PANSS (Positive and Negative Syndrome Scale), PET (positron emission tomography), RMS (root-mean-square), ROI (region of interest), SCZ (schizophrenia), SANS (Scale for the Assessment of Negative Symptoms), SAPS (Scale for the Assessment of Positive Symptoms), SD (standard deviation), SES (socioeconomic status), PALM (Permutation Analysis of Linear Models), TFCE (threshold-free cluster enhancement), TMS (transcranial magnetic stimulation), WM (working memory)

      Introduction

      Cognitive deficit has long been recognized as a hallmark of the schizophrenia syndrome (

      Kraepelin E, Barclay RM, Robertson GM (1919): Dementia præcox and paraphrenia. Edinburgh,: E. & S. Livingstone.

      ,

      Bleuler E (1950): Dementia praecox; or, The group of schizophrenias. New York: International Universities Press.

      ,
      • Palmer B.W.
      • Dawes S.E.
      • Heaton R.K.
      What do we know about neuropsychological aspects of schizophrenia?.
      ,
      • Green M.F.
      • Harvey P.D.
      Cognition in schizophrenia: Past, present, and future.
      ,
      • Heinrichs R.W.
      • Zakzanis K.K.
      Neurocognitive deficit in schizophrenia: a quantitative review of the evidence.
      ), but is only modestly improved with antipsychotic treatment (
      • Keefe R.S.
      • Bilder R.M.
      • Davis S.M.
      • Harvey P.D.
      • Palmer B.W.
      • Gold J.M.
      • et al.
      Neurocognitive effects of antipsychotic medications in patients with chronic schizophrenia in the CATIE Trial.
      ,
      • Heinrichs R.W.
      Cognitive improvement in response to antipsychotic drugs: neurocognitive effects of antipsychotic medications in patients with chronic schizophrenia in the CATIE Trial.
      ,
      • Keefe R.S.
      • Sweeney J.A.
      • Gu H.
      • Hamer R.M.
      • Perkins D.O.
      • McEvoy J.P.
      • et al.
      Effects of olanzapine, quetiapine, and risperidone on neurocognitive function in early psychosis: a randomized, double-blind 52-week comparison.
      ,
      • Green M.F.
      Cognition, drug treatment, and functional outcome in schizophrenia: a tale of two transitions.
      ,
      • Mishara A.L.
      • Goldberg T.E.
      A meta-analysis and critical review of the effects of conventional neuroleptic treatment on cognition in schizophrenia: opening a closed book.
      ,
      • Bowie C.R.
      • Harvey P.D.
      Cognitive deficits and functional outcome in schizophrenia.
      ). Patients with schizophrenia (SCZ) exhibit particularly severe impairments in working memory (WM) (
      • Lee J.
      • Park S.
      Working memory impairments in schizophrenia: a meta-analysis.
      ,
      • Forbes N.F.
      • Carrick L.A.
      • McIntosh A.M.
      • Lawrie S.M.
      Working memory in schizophrenia: a meta-analysis.
      ), which are highly predictive of patients’ functional outcomes (
      • Green M.F.
      • Harvey P.D.
      Cognition in schizophrenia: Past, present, and future.
      ,
      • Bowie C.R.
      • Harvey P.D.
      Cognitive deficits and functional outcome in schizophrenia.
      ,

      Green MF (1998): Schizophrenia from a neurocognitive perspective : probing the impenetrable darkness. Boston: Allyn and Bacon.

      ,
      • Kern R.S.
      • Gold J.M.
      • Dickinson D.
      • Green M.F.
      • Nuechterlein K.H.
      • Baade L.E.
      • et al.
      The MCCB impairment profile for schizophrenia outpatients: results from the MATRICS psychometric and standardization study.
      ). WM performance in humans and nonhuman primates depends on prefrontal cortical circuits and their dopaminergic modulation (
      • Mehta M.A.
      • Sahakian B.J.
      • McKenna P.J.
      • Robbins T.W.
      Systemic sulpiride in young adult volunteers simulates the profile of cognitive deficits in Parkinson's disease.
      ,
      • Cools R.
      • D'Esposito M.
      Inverted-U-shaped dopamine actions on human working memory and cognitive control.
      ,
      • Egan M.F.
      • Goldberg T.E.
      • Kolachana B.S.
      • Callicott J.H.
      • Mazzanti C.M.
      • Straub R.E.
      • et al.
      Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia.
      ,
      • Brozoski T.J.
      • Brown R.M.
      • Rosvold H.E.
      • Goldman P.S.
      Cognitive deficit caused by regional depletion of dopamine in prefrontal cortex of rhesus monkey.
      ,
      • Arnsten A.F.
      • Cai J.X.
      • Murphy B.L.
      • Goldman-Rakic P.S.
      Dopamine D1 receptor mechanisms in the cognitive performance of young adult and aged monkeys.
      ) and pathology of the prefrontal cortex is implicated in the etiology of SCZ (
      • Dempster K.
      • Norman R.
      • Theberge J.
      • Densmore M.
      • Schaefer B.
      • Williamson P.
      Cognitive performance is associated with gray matter decline in first-episode psychosis.
      ,
      • Haijma S.V.
      • Van Haren N.
      • Cahn W.
      • Koolschijn P.C.
      • Hulshoff Pol H.E.
      • Kahn R.S.
      Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects.
      ,
      • Weinberger D.R.
      • Egan M.F.
      • Bertolino A.
      • Callicott J.H.
      • Mattay V.S.
      • Lipska B.K.
      • et al.
      Prefrontal neurons and the genetics of schizophrenia.
      ,
      • Wible C.G.
      • Anderson J.
      • Shenton M.E.
      • Kricun A.
      • Hirayasu Y.
      • Tanaka S.
      • et al.
      Prefrontal cortex, negative symptoms, and schizophrenia: an MRI study.
      ,
      • Glantz L.A.
      • Lewis D.A.
      Decreased dendritic spine density on prefrontal cortical pyramidal neurons in schizophrenia.
      ,
      • Smucny J.
      • Dienel S.J.
      • Lewis D.A.
      • Carter C.S.
      Mechanisms underlying dorsolateral prefrontal cortex contributions to cognitive dysfunction in schizophrenia.
      ,
      • Lewis D.A.
      • Curley A.A.
      • Glausier J.R.
      • Volk D.W.
      Cortical parvalbumin interneurons and cognitive dysfunction in schizophrenia.
      ,
      • Abi-Dargham A.
      • Mawlawi O.
      • Lombardo I.
      • Gil R.
      • Martinez D.
      • Huang Y.
      • et al.
      Prefrontal dopamine D1 receptors and working memory in schizophrenia.
      ,
      • Abi-Dargham A.
      • Xu X.
      • Thompson J.L.
      • Gil R.
      • Kegeles L.S.
      • Urban N.
      • et al.
      Increased prefrontal cortical D(1) receptors in drug naive patients with schizophrenia: a PET study with [(1)(1)C]NNC112.
      ,
      • Abi-Dargham A.
      • Javitch J.A.
      • Slifstein M.
      • Anticevic A.
      • Calkins M.E.
      • Cho Y.T.
      • et al.
      Dopamine D1R Receptor Stimulation as a Mechanistic Pro-cognitive Target for Schizophrenia.
      ,
      • Kuperberg G.R.
      • Broome M.R.
      • McGuire P.K.
      • David A.S.
      • Eddy M.
      • Ozawa F.
      • et al.
      Regionally localized thinning of the cerebral cortex in schizophrenia.
      ), including reduced cortical dopamine release (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ). Consequently, identifying WM task-evoked functional deficits in neural activation that could explain impaired WM and provide a link to molecular abnormalities is a major goal in SCZ research.
      The medial prefrontal cortex (mPFC, which includes portions of the pregenual and subgenual cingulate) is a major component of the default mode network (DMN), active in the resting state and during autobiographical, self-referential, or otherwise internally-directed cognition (
      • Smallwood J.
      • Bernhardt B.C.
      • Leech R.
      • Bzdok D.
      • Jefferies E.
      • Margulies D.S.
      The default mode network in cognition: a topographical perspective.
      ,
      • Mittner M.
      • Boekel W.
      • Tucker A.M.
      • Turner B.M.
      • Heathcote A.
      • Forstmann B.U.
      When the brain takes a break: a model-based analysis of mind wandering.
      ,
      • Buckner R.L.
      • DiNicola L.M.
      The brain's default network: updated anatomy, physiology and evolving insights.
      ,
      • Yeshurun Y.
      • Nguyen M.
      • Hasson U.
      The default mode network: where the idiosyncratic self meets the shared social world.
      ,
      • Vatansever D.
      • Menon D.K.
      • Stamatakis E.A.
      Default mode contributions to automated information processing.
      ). Conversely, mPFC is deactivated during externally-oriented, goal-directed, tasks such as WM, and mPFC deactivation magnitude is correlated with WM performance (
      • Anticevic A.
      • Cole M.W.
      • Murray J.D.
      • Corlett P.R.
      • Wang X.J.
      • Krystal J.H.
      The role of default network deactivation in cognition and disease.
      ,
      • Anticevic A.
      • Repovs G.
      • Shulman G.L.
      • Barch D.M.
      When less is more: TPJ and default network deactivation during encoding predicts working memory performance.
      ,
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ). Patients with SCZ consistently show reduced DMN deactivation, including mPFC, during WM relative to healthy controls (HC), which may reflect a reduced ability to suppress internally-focused cognition or a general inability to shift from task-negative to optimal task-positive brain states in response to task demands (
      • Anticevic A.
      • Cole M.W.
      • Murray J.D.
      • Corlett P.R.
      • Wang X.J.
      • Krystal J.H.
      The role of default network deactivation in cognition and disease.
      ,
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ,
      • Braun U.
      • Harneit A.
      • Pergola G.
      • Menara T.
      • Schafer A.
      • Betzel R.F.
      • et al.
      Brain network dynamics during working memory are modulated by dopamine and diminished in schizophrenia.
      ,
      • Pomarol-Clotet E.
      • Salvador R.
      • Sarro S.
      • Gomar J.
      • Vila F.
      • Martinez A.
      • et al.
      Failure to deactivate in the prefrontal cortex in schizophrenia: dysfunction of the default mode network?.
      ,
      • Whitfield-Gabrieli S.
      • Thermenos H.W.
      • Milanovic S.
      • Tsuang M.T.
      • Faraone S.V.
      • McCarley R.W.
      • et al.
      Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
      ,
      • Hahn B.
      • Harvey A.N.
      • Gold J.M.
      • Ross T.J.
      • Stein E.A.
      Load-dependent hyperdeactivation of the default mode network in people with schizophrenia.
      ,
      • Kim D.I.
      • Manoach D.S.
      • Mathalon D.H.
      • Turner J.A.
      • Mannell M.
      • Brown G.G.
      • et al.
      Dysregulation of working memory and default-mode networks in schizophrenia using independent component analysis, an fBIRN and MCIC study.
      ,
      • Metzak P.D.
      • Riley J.D.
      • Wang L.
      • Whitman J.C.
      • Ngan E.T.
      • Woodward T.S.
      Decreased efficiency of task-positive and task-negative networks during working memory in schizophrenia.
      ,
      • Wu D.
      • Jiang T.
      Schizophrenia-related abnormalities in the triple network: a meta-analysis of working memory studies.
      ,
      • Zhou L.
      • Pu W.
      • Wang J.
      • Liu H.
      • Wu G.
      • Liu C.
      • et al.
      Inefficient DMN Suppression in Schizophrenia Patients with Impaired Cognitive Function but not Patients with Preserved Cognitive Function.
      ). However, few studies (e.g., 40, 43) have examined associations between mPFC hypodeactivation and WM performance in SCZ, which is necessary for group differences in activation to have explanatory power over WM impairment in SCZ. Critically, one of these studies (
      • Whitfield-Gabrieli S.
      • Thermenos H.W.
      • Milanovic S.
      • Tsuang M.T.
      • Faraone S.V.
      • McCarley R.W.
      • et al.
      Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
      ) was an investigation of only 13 patients and employed now-outdated statistical methodologies, while the other (
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ) used a relatively under-studied WM task, the Self-Ordered WM Task. In addition, the mechanisms by which cortical hypodopaminergia (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ,

      Weinberger DR, Berman KF, Daniel DG (1992): Mesoprefrontal cortical dopaminergic activity and prefrontal hypofunction in schizophrenia. Clin Neuropharmacol. 15 Suppl 1 Pt A:568A-569A.

      ) may produce WM deficits in SCZ, including potential relationships with mPFC hypodeactivaiton, are unclear.
      Furthermore, the role of dopamine in DMN function and deactivation during WM tasks is poorly understood. In dorsolateral prefrontal cortex (dlPFC), dopamine is thought to mediate the balance of excitation and inhibition via D1 and D2 receptors by activating GABAergic and pyramidal cells, thus increasing signal to noise and neuronal activity needed for WM function (
      • Seamans J.K.
      • Gorelova N.
      • Durstewitz D.
      • Yang C.R.
      Bidirectional dopamine modulation of GABAergic inhibition in prefrontal cortical pyramidal neurons.
      ,
      • Seamans J.K.
      • Yang C.R.
      The principal features and mechanisms of dopamine modulation in the prefrontal cortex.
      ,
      • O'Donnell P.
      Cortical disinhibition in the neonatal ventral hippocampal lesion model of schizophrenia: new vistas on possible therapeutic approaches.
      ,
      • O'Donnell P.
      Adolescent onset of cortical disinhibition in schizophrenia: insights from animal models.
      ,
      • Arnsten A.F.
      • Wang M.J.
      • Paspalas C.D.
      Neuromodulation of thought: flexibilities and vulnerabilities in prefrontal cortical network synapses.
      ). However, its effects in mPFC are less clear, and to our knowledge it remains unknown whether the circuit dynamics in mPFC are similar to dlPFC or if other mechanisms are at play. For example, it is conceivable that mPFC activation at rest could occur with lower dopamine levels than those needed in dlPFC, allowing broader representations and less network sculpting (i.e., a noisier network). Regardless, if mPFC deactivation impacts cognitive deficits in SCZ, understanding dopamine’s effect on mPFC microcircuitry may be critical in understanding WM impairment.
      Here, we performed a secondary analysis of data from medicated and unmedicated patients with SCZ and matched HC (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ,
      • Cassidy C.M.
      • Van Snellenberg J.X.
      • Benavides C.
      • Slifstein M.
      • Wang Z.
      • Moore H.
      • et al.
      Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
      ) who performed a visual n-back task (
      • Cassidy C.M.
      • Van Snellenberg J.X.
      • Benavides C.
      • Slifstein M.
      • Wang Z.
      • Moore H.
      • et al.
      Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
      ) during fMRI. First, we sought to identify mPFC voxels that both deactivate during task and show a relationship between deactivation and task performance across participants. We then assessed whether hypodeactivation in these voxels is a statistical mediator of WM impairment in SCZ. Second, we conducted the same analyses to investigate lateral PFC (lPFC; which includes dlPFC and ventrolateral PFC, which is also commonly activated by WM tasks) activation, a major target of investigations into WM impairments for several decades (
      • Smucny J.
      • Dienel S.J.
      • Lewis D.A.
      • Carter C.S.
      Mechanisms underlying dorsolateral prefrontal cortex contributions to cognitive dysfunction in schizophrenia.
      ,
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ,
      • Goldman-Rakic P.S.
      Working memory dysfunction in schizophrenia.
      ,
      • Weinberger D.R.
      • Berman K.F.
      Prefrontal function in schizophrenia: confounds and controversies.
      ,
      • Carter C.S.
      • Perlstein W.
      • Ganguli R.
      • Brar J.
      • Mintun M.
      • Cohen J.D.
      Functional hypofrontality and working memory dysfunction in schizophrenia.
      ,
      • Weinberger D.R.
      • Berman K.F.
      • Zec R.F.
      Physiologic dysfunction of dorsolateral prefrontal cortex in schizophrenia. I. Regional cerebral blood flow evidence.
      ,
      • Manoach D.S.
      Prefrontal cortex dysfunction during working memory performance in schizophrenia: reconciling discrepant findings.
      ,
      • Van Snellenberg J.X.
      • Torres I.J.
      • Thornton A.E.
      Functional neuroimaging of working memory in schizophrenia: task performance as a moderating variable.
      ,
      • Van Snellenberg J.X.
      • Slifstein M.
      • Read C.
      • Weber J.
      • Thompson J.L.
      • Wager T.D.
      • et al.
      Dynamic shifts in brain network activation during supracapacity working memory task performance.
      ,
      • Eryilmaz H.
      • Tanner A.S.
      • Ho N.F.
      • Nitenson A.Z.
      • Silverstein N.J.
      • Petruzzi L.J.
      • et al.
      Disrupted Working Memory Circuitry in Schizophrenia: Disentangling fMRI Markers of Core Pathology vs Other Aspects of Impaired Performance.
      ,
      • Lewis D.A.
      • Sweet R.A.
      Schizophrenia from a neural circuitry perspective: advancing toward rational pharmacological therapies.
      ,
      • Barch D.M.
      • Carter C.S.
      • Braver T.S.
      • Sabb F.W.
      • MacDonald 3rd, A.
      • Noll D.C.
      • et al.
      Selective deficits in prefrontal cortex function in medication-naive patients with schizophrenia.
      ,
      • Williams G.V.
      • Goldman-Rakic P.S.
      Modulation of memory fields by dopamine D1 receptors in prefrontal cortex.
      ,
      • Sawaguchi T.
      • Goldman-Rakic P.S.
      D1 dopamine receptors in prefrontal cortex: involvement in working memory.
      ). Our hypotheses were that: 1) activation in lPFC, and deactivation in mPFC, would be associated with n-back performance; and 2) activation in lPFC, and deactivation in mPFC, would be impaired in patients and that these impairments would mediate group differences in performance.
      Finally, we explored the relationship between task-induced changes in activation and dopamine release in a subset of participants who underwent positron emission tomography (PET) imaging with the dopamine receptor D2/3 (D2/3R) antagonist FLB457 ([11C]FLB457) before and after an amphetamine challenge, enabling quantification of prefrontal synaptic dopamine release (
      • Narendran R.
      • Frankle W.G.
      • Mason N.S.
      • Rabiner E.A.
      • Gunn R.N.
      • Searle G.E.
      • et al.
      Positron emission tomography imaging of amphetamine-induced dopamine release in the human cortex: a comparative evaluation of the high affinity dopamine D2/3 radiotracers [11C]FLB 457 and [11C]fallypride.
      ,
      • Narendran R.
      • Mason N.S.
      • May M.A.
      • Chen C.M.
      • Kendro S.
      • Ridler K.
      • et al.
      Positron emission tomography imaging of dopamine D(2)/(3) receptors in the human cortex with [(1)(1)C]FLB 457: reproducibility studies.
      ,
      • Narendran R.
      • Jedema H.P.
      • Lopresti B.J.
      • Mason N.S.
      • Gurnsey K.
      • Ruszkiewicz J.
      • et al.
      Imaging dopamine transmission in the frontal cortex: a simultaneous microdialysis and [11C]FLB 457 PET study.
      ). We evaluated the relationship between task-induced deactivation and dopamine release in mPFC and lPFC. We predicted that prefrontal dopamine release would be associated with greater activation and deactivation within performance-associated subregions of lPFC and mPFC, respectively.

      Materials and Methods

      Participants

      All procedures described in this study were approved by the New York State Psychiatric Institute Institutional Review Board, and by the Yale University Human Investigation committee and Yale Radiation Safety Committee. This study is a secondary analysis of fMRI, PET, and clinical data collected from
      patients with SCZ and matched HC that were published previously (see 32, 55, 71).
      From an initial sample of 18 unmedicated patients, 39 medicated patients, and 43 matched HC, quality control procedures resulted in the exclusion of 4 unmedicated patients, 12 medicated patients, and 3 HC for failing to meet performance and/or fMRI data quality requirements; the final sample included 14 unmedicated patients, 27 medicated patients, and 40 HC. Table 1 contains a summary of the clinical measures and demographic data for these participants, as well as a subgroup of individuals for whom [11C]FLB457 PET data were available. Details regarding recruitment, informed consent, inclusion and exclusion criteria, clinical assessments, and requirements for task performance and fMRI data quality are detailed in Supplementary Materials and Methods.

      Visual n-back Working Memory fMRI Task Procedures

      All participants completed a visual n-back working memory task (
      • Cassidy C.M.
      • Van Snellenberg J.X.
      • Benavides C.
      • Slifstein M.
      • Wang Z.
      • Moore H.
      • et al.
      Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
      ) that uses simple isometric line drawings of 3D objects (
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ,
      • Van Snellenberg J.X.
      • Slifstein M.
      • Read C.
      • Weber J.
      • Thompson J.L.
      • Wager T.D.
      • et al.
      Dynamic shifts in brain network activation during supracapacity working memory task performance.
      ,
      • Curtis C.E.
      • Zald D.H.
      • Pardo J.V.
      Organization of working memory within the human prefrontal cortex: a PET study of self-ordered object working memory.
      ); see Figure 1A. During the n-back task (
      • Moore M.E.
      • Ross B.M.
      Context effects in running memory.
      ,
      • Ross B.M.
      Serial order as a unique source of error in running memory.
      ,
      • Ross B.M.
      Serial-order effects in two-channel running memory.
      ,
      • Gevins A.S.
      • Bressler S.L.
      • Cutillo B.A.
      • Illes J.
      • Miller J.C.
      • Stern J.
      • et al.
      Effects of prolonged mental work on functional brain topography.
      ), participants are presented stimuli sequentially and must respond with a button press to indicate whether the current stimulus (the probe) matches the one presented n trials previously (the target). WM load was modulated by changing the number of trials between probe and target (n), resulting in a 1-back condition (Figure 1A, top) and 2-back condition (Figure 1A, bottom).
      Figure thumbnail gr1
      Figure 1The visual n-back task. A) Sample sequence of stimuli presented during visual n-back task blocks. Top: 1-back condition, where participants are instructed to determine whether the currently presented stimulus matches the stimulus that was presented one trial previously. Bottom: 2-back condition, where participants are instructed to determine whether the currently presented stimulus matches the stimulus that was presented two trials previously. Each stimulus was presented for 2 seconds and separated temporally from others by a 2-second fixation cross display. B) Schematic diagram of a run of the visual n-back task. Participants were asked to complete 4 task runs, each composed of 2 blocks of the 1-back condition and 2 blocks of the 2-back condition, in a repeating AABB BBAA design. Blue bars represent condition A, and red bars represent condition B; assignment of 1-back and 2-back to A and B blocks was performed pseudorandomly across subjects. Each block comprised 10 stimuli, totaling 40 seconds per block. Before the start of each block, participants were provided a rest period of 9 seconds, followed by 8 seconds during which task instructions were displayed (gold bars), and another rest period of 8 seconds before the first stimulus.
      Participants completed four 290-second runs of the n-back task during fMRI (Figure 1B) with a 2 second repetition time and 3 mm isotropic voxels. Two HC and 4 SCZ completed 3 runs (average 3.95 runs HC, 3.90 runs SCZ), and all other participants completed 4 full runs. Within each run, each participant completed 2 full blocks of both 1-back and 2-back task conditions (i.e., no partial completion of task blocks or runs). Data were pre-processed using fMRIPrep v20.0.7 (
      • Esteban O.
      • Markiewicz C.J.
      • Blair R.W.
      • Moodie C.A.
      • Isik A.I.
      • Erramuzpe A.
      • et al.
      fMRIPrep: a robust preprocessing pipeline for functional MRI.
      ,
      • Esteban O.
      • Ciric R.
      • Finc K.
      • Blair R.W.
      • Markiewicz C.J.
      • Moodie C.A.
      • et al.
      Analysis of task-based functional MRI data preprocessed with fMRIPrep.
      ,

      Esteban O, Markiewicz CJ, Goncalves M, DuPre E, Kent JD, Salo T, et al. (2021): fMRIPrep: a robust preprocessing pipeline for functional MRI. 20.0.7 ed: Zenodo.

      ). Details of the task design, fMRI acquisition, pre-processing, and post-processing are further described in Supplementary Materials and Methods.

      Positron Emission Tomography (PET) Acquisition and Analysis

      The PET data used in this study are a subset of a previously published investigation with detailed methods described in that report (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ). PET imaging data were acquired using [11C]FLB457 on a Siemens HR+ PET scanner at the Yale University PET Center. Participants completed two scan sessions in a single day. After an initial 90-minute baseline scan, a second 90-minute scan was performed 3 hours following oral administration of 0.5 mg/kg amphetamine. Binding potential relative to the non-displaceable compartment (BPND) was quantified within an a priori mPFC region of interest (ROI) drawn on each participant’s high-resolution T1-weighted MR image using MEDx (Medical Numerics, Inc., Germantown, MD), as has been described in detail previously (
      • Abi-Dargham A.
      • Xu X.
      • Thompson J.L.
      • Gil R.
      • Kegeles L.S.
      • Urban N.
      • et al.
      Increased prefrontal cortical D(1) receptors in drug naive patients with schizophrenia: a PET study with [(1)(1)C]NNC112.
      ,
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ,
      • Abi-Dargham A.
      • Martinez D.
      • Mawlawi O.
      • Simpson N.
      • Hwang D.R.
      • Slifstein M.
      • et al.
      Measurement of striatal and extrastriatal dopamine D1 receptor binding potential with [11C]NNC 112 in humans: validation and reproducibility.
      ,
      • Kegeles L.S.
      • Slifstein M.
      • Xu X.
      • Urban N.
      • Thompson J.L.
      • Moadel T.
      • et al.
      Striatal and extrastriatal dopamine D2/D3 receptors in schizophrenia evaluated with [18F]fallypride positron emission tomography.
      ). It should be noted that these individualized, hand-drawn mPFC ROIs did not completely overlap the fMRI ROI, as they were defined using operational criteria rather than being taken from a parcellation in MNI space. However, a direct comparison indicated that 80% of voxels in the PET ROI did fall within the fMRI ROI (on average across participants), so PET imaging results in this region should predominantly reflect findings within the same mPFC region. Dopamine release capacity within the mPFC ROI was quantified as the percentage change in BPND following amphetamine (ΔBPND), with more negative ΔBPND values indicating greater dopamine release. From the 81 total participants whose fMRI data were analyzed in the current study, PET data were available for 23 participants (14 HC and 9 unmedicated patients; see Table 1).

      N-back fMRI Task Performance

      T-tests for the difference in task performance between groups were employed after testing for normality using Welch’s unequal variances t-test or Mann–Whitney–Wilcoxon rank-sum test (

      Gibbons JD, Chakraborti S (2021): Nonparametric statistical inference. 6th edition. ed. Boca Raton: CRC Press.

      ,

      Hollander M, Wolfe DA, Chicken E (2014): Nonparametric statistical methods. Third edition / ed. Hoboken, New Jersey: John Wiley & Sons, Inc.

      ), as appropriate. Normality was assessed using a Lilliefors test (
      • Lilliefors H.W.
      On the Kolmogorov-Smirnov Test for Normality with Mean and Variance Unknown.
      ,

      Conover WJ (1999): Practical nonparametric statistics. 3rd ed. New York: Wiley.

      ). Associations between task performance and both gender and symptom severity (PANSS subscores) were assessed separately as described in Supplementary Materials and Methods.

      Within-Participants (First-Level) Modeling

      Within-participants modeling of the n-back task was performed using a block design generalized linear model (GLM) in SPM12 (

      Friston KJ, Ashburner J, Kiebel S, Nichols T, Penny WD (2007): Statistical parametric mapping : the analysis of funtional brain images. First edition. ed. Amsterdam ; Boston: Elsevier/Academic Press, pp 1 online resource (vii, 647 pages).

      ,
      • Friston K.J.
      • Jezzard P.
      • Turner R.
      Analysis of functional MRI time-series.
      ,
      • Friston K.J.
      • Holmes A.P.
      • Worsley K.J.
      • Poline J.P.
      • Frith C.D.
      • Frackowiak R.S.J.
      Statistical parametric maps in functional imaging: A general linear approach.
      ), version 7771, in MATLAB R2018a (The MathWorks, Inc., Natick, MA), with a three-parameter hemodynamic response function (HRF), which included canonical HRF (height), time-to-peak (temporal derivative), and full-width-half-maximum (dispersion derivative) terms (
      • Friston K.J.
      • Fletcher P.
      • Josephs O.
      • Holmes A.
      • Rugg M.D.
      • Turner R.
      Event-related fMRI: characterizing differential responses.
      ,
      • Friston K.J.
      • Penny W.
      • David O.
      Modeling brain responses.
      ,
      • Henson R.N.
      • Price C.J.
      • Rugg M.D.
      • Turner R.
      • Friston K.J.
      Detecting latency differences in event-related BOLD responses: application to words versus nonwords and initial versus repeated face presentations.
      ). A regressor for each of the six motion parameters and their six derivatives was included, as were spike regressors to remove volumes during which excessive motion occurred (using run-adaptive generalized extreme value-DVARS thresholds; see 98), without being convolved with the HRF. A block design GLM was selected due to the relatively rapid trials (stimulus duration and inter-stimulus interval of 2 seconds each, with no jitter) that would have rendered it difficult to model hemodynamic responses associated with individual events. Complete first-level modeling design details are provided in Supplementary Materials and Methods.

      Across-Participants (Second-Level) Modeling

      Across-participants modeling was performed using Permutation Analysis of Linear Models (PALM; 99, 100-103) alpha119. All analyses in PALM were carried out with 20,000 permutations or sign-flips, using built-in family-wise error rate correction (FWER; 104), threshold-free cluster enhancement (TFCE; 105) and multiple testing correction over contrasts (
      • Alberton B.A.V.
      • Nichols T.E.
      • Gamba H.R.
      • Winkler A.M.
      Multiple testing correction over contrasts for brain imaging.
      ). Regressors were mean-centered before calculation of interaction terms.

      ROI Analyses of Activation and Deactivation

      We produced volumetric ROIs of mPFC and lPFC, shown in Figure 2, using the cortical area parcellation developed by Gordon and associates (
      • Gordon E.M.
      • Laumann T.O.
      • Adeyemo B.
      • Huckins J.F.
      • Kelley W.M.
      • Petersen S.E.
      Generation and Evaluation of a Cortical Area Parcellation from Resting-State Correlations.
      ), as described in Supplementary Materials and Methods. Within these ROIs, we identified voxels that demonstrated significant activation in lPFC, or deactivation in mPFC, during the 2-back task condition, either across all participants, or within HC or SCZ groups separately, using PALM. T-scores were thresholded (FWER corrected p < 0.05, over 3 contrasts) to generate masks of voxels with significant activation during the 2-back task condition (relative to implicit baseline) within mPFC and lPFC. The 2-back contrast was examined relative to baseline, rather than to the 1-back task condition, because during prior work with this dataset (
      • Cassidy C.M.
      • Van Snellenberg J.X.
      • Benavides C.
      • Slifstein M.
      • Wang Z.
      • Moore H.
      • et al.
      Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
      ) we had observed relatively minimal activation (and deactivation) differences between 1-back and 2-back conditions, especially in the DMN. Thus, because DMN deactivation was of specific interest for our analyses here, we decided to focus on 2-back activation relative to an implicit baseline due to the lack of a 0-back condition in the original experimental design.
      Figure thumbnail gr2
      Figure 2Medial prefrontal cortex (mPFC; top) and lateral prefrontal cortex (lPFC; bottom) masks, generated from the Gordon (2016) parcellation. mPFC masks were generated by including all parcels in prefrontal cortex (PFC) parcels that are members of the community, “Default Mode Network.” lPFC masks were generated by including all PFC parcels that are members of either the “Dorsal Attention Network” community or the “Frontoparietal” community.

      Relationships Between Task Performance and Activation or Deactivation

      Next, voxels showing significant activation or deactivation during 2-back were examined to identify relationships between activation magnitude and performance (percent correct), while controlling for group membership and group-performance interaction effects, using PALM, as detailed in Supplementary Materials and Methods.
      Results of the above analysis were then thresholded (FWER corrected p < 0.05, tested separately in the positive and negative direction and corrected over 2 contrasts in order to obtain valid TFCE statistics) and a scalar value representing mPFC deactivation for each participant was quantified by taking the spatial average of deactivation during the 2-back task condition in the performance-associated mPFC subregion for each participant. Note that greater mPFC deactivation magnitude is indicated by a more negative activation in this region. To optimally describe the data extracted after hypothesis testing in PALM, performance was then modeled as the dependent variable in an initial full GLM that included regressors for intercept, SCZ diagnosis, mean mPFC deactivation, and covariates of age, handedness, and biological sex, as well as all possible interactions (through fifth order). All regressors except for performance were mean-centered before calculating interaction terms. An optimal reduced model was then produced by selecting terms through backward elimination (see Supplementary Materials and Methods).
      Group differences (HC vs. SCZ) in mPFC subregion deactivation were assessed for significance using Welch’s t-test after testing for normality. Deactivation in the mPFC subregion was evaluated as a mediator of the impact of diagnosis on task performance using the CANlab M3 Mediation Toolbox (
      • Wager T.D.
      • Davidson M.L.
      • Hughes B.L.
      • Lindquist M.A.
      • Ochsner K.N.
      Prefrontal-subcortical pathways mediating successful emotion regulation.
      ,
      • Geuter S.
      • Reynolds Losin E.A.
      • Roy M.
      • Atlas L.Y.
      • Schmidt L.
      • Krishnan A.
      • et al.
      Multiple Brain Networks Mediating Stimulus-Pain Relationships in Humans.
      ,
      • Shrout P.E.
      • Bolger N.
      Mediation in experimental and nonexperimental studies: new procedures and recommendations.
      ,
      • Kenny D.A.
      • Korchmaros J.D.
      • Bolger N.
      Lower level mediation in multilevel models.
      ). and the significance of each path was assessed using a bias-corrected and accelerated bootstrap procedure (

      Efron B, Tibshirani R (1993): An introduction to the bootstrap. New York: Chapman & Hall.

      ,
      • DiCiccio T.J.
      • Efron B.
      Bootstrap confidence intervals.
      ) with 107 resamples. To show the effect of controlling for group and group-performance interaction effects, these analyses were repeated while excluding those two regressors when identifying voxels with significant activation-performance associations.
      Several supplementary analyses were performed, as detailed in Supplementary Materials and Methods. Associations between mPFC subregion deactivation and gender, as well as PANSS positive and negative subscores, were assessed using robust regression. To assess for the possibility that observed differences between groups are not the result of differences in participant motion or partial volume related signal loss, we additionally performed group comparisons in median framewise displacement (
      • Power J.D.
      • Barnes K.A.
      • Snyder A.Z.
      • Schlaggar B.L.
      • Petersen S.E.
      Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion.
      ) and root-mean-square (RMS) signal intensity within all ROIs used in main analyses. Finally, to confirm that observed group differences in mPFC deactivation were robust to the cluster selection procedure employed and did not originate from a “double-dipping” confound (see 114), we additionally isolated voxels associated with task performance as described in this section, but directly using the mPFC mask created from the Gordon parcellation (
      • Gordon E.M.
      • Laumann T.O.
      • Adeyemo B.
      • Huckins J.F.
      • Kelley W.M.
      • Petersen S.E.
      Generation and Evaluation of a Cortical Area Parcellation from Resting-State Correlations.
      ).

      PET-fMRI Analysis of [11C]FLB457 ΔBPND

      Associations between mPFC [11C]FLB457 ΔBPND and average activation in the performance-associated mPFC subregion were assessed using robust regression with mPFC deactivation as the dependent variable with regressors for intercept, SCZ diagnosis, mPFC [11C]FLB457 ΔBPND, diagnosis*ΔBPND interactions, and age as a covariate, in order to control for the potential effect of age on cortical D2R parameters (
      • Inoue M.
      • Suhara T.
      • Sudo Y.
      • Okubo Y.
      • Yasuno F.
      • Kishimoto T.
      • et al.
      Age-related reduction of extrastriatal dopamine D2 receptor measured by PET.
      ,
      • Kaasinen V.
      • Vilkman H.
      • Hietala J.
      • Nagren K.
      • Helenius H.
      • Olsson H.
      • et al.
      Age-related dopamine D2/D3 receptor loss in extrastriatal regions of the human brain.
      ). Age and diagnosis were mean-centered.

      Results

      Task Performance

      Accuracy (percent of correct trials) and reaction time on the n-back task is shown for each group in Figure 3 and Table S1. HC (N = 40) exhibited significantly greater accuracy than patients with SCZ (N = 41) on both the 1-back (p = 0.0131) and 2-back task condition (p = 0.0015); HC reaction time was significantly faster than patients during 1-back trials (p = 0.0395) and were faster during 2-back trials with trend-level significance (p = 0.0938). Unmedicated patients (N = 17) and medicated patients (N = 24) did not differ significantly on either accuracy or reaction time. No significant associations were found between task performance and gender (p = 0.6230; see Table S2) or PANSS subscores (postive subscore, p = 0.8381; negative subscore, p = 0.7061; see Table S3).
      Figure thumbnail gr3
      Figure 3Participant performance on the n-back task. Solid fills show kernel density estimates within the range of observed data. Solid lines denote group median and dashed lines denote upper and lower quartiles. Top, percent of responses correct during the 1-back condition (A) and 2-back condition (B). Bottom, reaction time during the 1-back condition (C) and 2-back condition (D). Asterisks (*) denote a significant difference between groups as assessed by either Welch’s unequal variances t-test or Mann-Whitney-Wilcoxon rank-sum test as indicated after testing for normality by Lilliefors test (α = 0.05, uncorrected). P-values generated using non-parametric tests are indicated by dagger (†) superscripts.

      Working Memory Task Functional Magnetic Resonance Imaging

      Activation to Task
      Within mPFC, a total of 5,727 voxels (154.63 cm3) showed significant deactivation during 2-back task blocks (Figure 4A; cluster coordinates, extent, and greatest magnitude t-scores shown in Table S4). In lPFC, 4,738 voxels (127.93 cm3) were found to show significant activation during 2-back task blocks (Figure 4B and Table S5).
      Figure thumbnail gr4
      Figure 4T-scores within voxels showing significant deactivation during 2-back task blocks in medial prefrontal cortex (mPFC; 5,727 voxels; top) and activation in lateral prefrontal cortex (lPFC; 4,738 voxels; bottom). T-scores were determined by testing across all participants, within healthy controls, and within patients with schizophrenia, and combining contrasts using non-parametric combination (NPC). Significance was assessed after threshold-free cluster enhancement (TFCE), and family-wise error rate (FWER) correction over voxels and contrasts (FWER corrected p < 0.05), using Permutation Analysis of Linear Models (PALM) with null distributions generated from 20,000 sign-flips.

      Activation-Performance Associations

      Medial prefrontal cortex (mPFC)
      From among the 5,727 mPFC voxels significantly deactivating during 2-back blocks, 558 voxels (15.07 cm3) were found to additionally display associations between deactivation magnitude and task performance, while controlling for main effects of diagnosis and for diagnosis-performance interaction effects. Figure 5A and Table S6 show the subregion of mPFC in which significant activation-performance associations were observed. Associations between WM load and deactivation in this region were assessed in Supplementary Materials and Methods are discussed in Supplementary Results and shown in Figure S1. Median and interquartile range of average mPFC deactivation for each task condition, and the change across task conditions, are additionally shown in Table S7.
      Figure thumbnail gr5
      Figure 5A) T-scores in the 558 voxels (15.07 cm3) of medial prefrontal cortex (mPFC) within which magnitude of deactivation during 2-back task blocks was significantly associated with task performance (percentage of trials correct) across participants, controlling for diagnosis (healthy controls vs. patients with schizophrenia) and diagnosis-performance interaction effects. B) Relationship between the spatial average of mPFC subregion activation magnitude and task performance (percent correct trials) during 2-back blocks, after hypothesis testing in PALM. Performance was modeled as the dependent variable as selected by backward elimination from an initial model including an intercept, diagnosis, mPFC deactivation, age, gender, and all possible interaction terms. The optimal model included terms for intercept, diagnosis, mPFC activation magnitude, and age. 77 degrees of freedom; R2 = 0.364; adjusted R2 = 0.339. Model parameters shown in Table S8. C) Violin plots of mPFC subregion activation in each group. Significance of group differences determined by Welch’s unequal variances t-test. Solid lines indicate group means. D) Assessment of mPFC subregion deactivation as a mediator of working memory performance impairment in schizophrenia. Mean and standard error of path coefficients are shown. Note that hypothesis testing for the effect of mPFC activation on 2-back performance was performed in PALM prior to this analysis. Asterisks denote significance (α = 0.05) as assessed by a bias-corrected and accelerated bootstrap procedure (
      • Petrides M.
      • Milner B.
      Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man.
      ,
      • Gold J.M.
      • Hahn B.
      • Zhang W.W.
      • Robinson B.M.
      • Kappenman E.S.
      • Beck V.M.
      • et al.
      Reduced capacity but spared precision and maintenance of working memory representations in schizophrenia.
      ) with 107 resamples.
      Activation was averaged across these voxels within each participant and used in a regression model to predict task performance, with terms included for intercept (p = 6.686 × 10-58), diagnosis (p = 0.0152), mPFC subregion activation (p = 2.975 × 10-5), and age (p = 0.0510). Note that these p-values are not representative of the model’s level of significance and are included only for completeness; second-level analyses that were performed in PALM are the appropriate unbiased null-hypothesis tests for this result. Model coefficients and statistics are shown in Table S8, and model fit in each group is shown in Figure 5B. Patients with SCZ displayed significantly reduced task-evoked mPFC subregion deactivation relative to HC (p = 0.0351), as shown in Figure 5C. No significant associations were found between activation and gender (p = 0.8152; see Table S9) or PANSS subscores (postive subscore, p = 0.2073; negative subscore, p = 0.3408; see Table S10). Follow-up analysis aiming to detect clusters of deactivation-performance association within in the Gordon parcellation-derived mPFC mask directly revealed 506 such voxels (13.66 cm3), shown in Figure S2 and Table S11; patients with SCZ likewise showed reduced deactivation in this region relative to HC (p = 0.0438). Participant motion (median FD) and RMS signal intensity did not differ significantly between groups, as shown in Figure S3 and Figure S4, respectively.
      Finally, we evaluated the effect of mPFC subregion deactivation as a mediator of the WM impairment observed in patients with SCZ, shown in Figure 5D. Mediation analysis revealed a significant mediated (indirect) effect (p = 0.0431) of diagnosis on WM task performance, and a direct effect with only trend level significance (p = 0.0545). This suggests that failure to sufficiently deactivate mPFC may partially explain the observed impairment in task performance in patients with SCZ.
      The analyses shown in Figure 5 were additionally performed without controlling for diagnosis and diagnosis-performance interactions when assessing for significant activation-performance associations to show the effect of controlling for these variables. As described in Supplementary Results and shown in Figure S5, Table S12, and Table S13, these results largely similar, in that 1,125 voxels of showing significance were found in mPFC and zero were found in lPFC (minimum p-value of 0.371).

      Lateral prefrontal cortex (lPFC)

      From among the 4,738 voxels showing significant activation during 2-back task blocks, none showed significant associations between activation magnitude and task performance (2-back percent correct); the minimum TFCE-enhanced, FWER-corrected p-value was 0.278. To ensure this was not due to a potential confounding effect of differences in hemispheric dominance across participants, we repeated all second-level modeling in lPFC (i.e., localizing regions with significant activation during 2-back task blocks, and attempting to further detect voxels with significant activation-performance associations), but after removing all 8 left-handed participants (7 SCZ, 1 HC; remaining total N = 73). This analysis similarly yielded no voxels showing significant associations between activation magnitude and performance.

      Cross-Modality [11C]FLB457 PET and Working Memory Task fMRI Results

      Next, we sought to examine the relationship between task-induced mPFC deactivation and dopamine release in mPFC in a subset of HC and unmedicated patients with SCZ who completed both fMRI during the n-back task and [11C]FLB457 PET with an amphetamine challenge, controlling for age and diagnosis as shown in in Figure 6. Coefficients and statistics for the model are shown in Table S14, which included terms for intercept (p = 0.0269), diagnosis (p = 0.504), [11C]FLB457 ΔBPND (p = 0.0.0257), and age (p = 0.569), as well as a ΔBPND*Diagnosis (p = 0.311) interaction term. Across all participants, greater mPFC deactivation during task blocks was significantly associated with reduced dopamine release (ΔBPND = (post-amphetamine BPND/baseline BPND) -1; more negative values correspond to greater dopamine release). When evaluated in each group separately (retaining and age as a covariate), ΔBPND was significantly associated with mPFC deactivation in HC (p = 0.0152), but not in SCZ (p = 0.617).
      Figure thumbnail gr6
      Figure 6Activation in medial prefrontal cortex (mPFC) during 2-back working memory (WM) task blocks as a function of mPFC dopamine release capacity quantified as the fractional change in binding potential, (ΔBPND) using radiotracer [11C]FLB457 and an amphetamine challenge paradigm. Model coefficients and statistics are shown in Table S12. Note that greater dopamine release capacity is indicated by a more negative ΔBPND and greater mPFC deactivation magnitude is indicated by more negative activation. R2 = 0.341; adjusted R2 = 0.195.

      Supplementary Whole-Brain fMRI Analyses

      In order to comprehensively characterize our dataset for future use (e.g., to support meta-analyses), we also performed several separate follow-up analyses within all cortical grey matter rather than a priori ROIs. These analyses are described in Supplementary Materials and Methods and Supplementary Results, and shown in Tables S15-S35 and Figures S6-S24.

      Discussion

      The results presented here suggest that mPFC deactivation during n-back task performance plays a meaningful role in accurate responding by participants and may, at least in part, explain WM deficits in SCZ. We demonstrated that while dlPFC, ventrolateral PFC, and mPFC all responded to task blocks (Figure 4), a relationship between activation and task performance was only observed in mPFC (Figure 5A-B). Within the performance-associated subregion of mPFC, deactivation magnitude during the n-back task was significantly reduced in SCZ (Figure 5C) and mediates group differences in task performance (Figure 5D). This mediation effect indicates that a portion of the deficit in n-back performance observed in SCZ is associated with a failure to deactivate mPFC to the same extent as HCs. Although this does not necessarily indicate a causal relationship between mPFC deactivation and WM deficit in SCZ, and may reflect elements of cognition that are not necessarily specific to WM task engagement (see Limitations), it is consistent with this possibility and warrants further study.
      In addition, in a subset of participants we demonstrated that dopamine release capacity in mPFC, as quantified by the amphetamine-induced ΔBPND of the PET radiotracer [11C]FLB457, is negatively associated with the magnitude of mPFC deactivation (Figure 5), which did not support our original hypothesis. This dopamine release-deactivation relationship differed between patients and controls at a trend level. In HC, mPFC deactivation was associated with reduced release capacity, while in SCZ this relationship was not observed; potentially due to blunted mPFC deactivation, and thus reduced range in mPFC deactivation magnitudes, observed in SCZ. However, this finding should be considered preliminary due to the small sample size employed here.
      Our null result in dlPFC, taken in the context of a general lack of studies demonstrating a statistical association between dlPFC activation and WM task performance in the literature, challenges the notion that deficient task-induced dlPFC activation (or hypofrontality, broadly) is a robust biomarker of WM impairment in SCZ that is generalizable across tasks and load conditions (a concern that has been raised previously; see, e.g., 40, 117), but is consistent with mixed observations throughout the literature of both hypo- and hyper-activation of dlPFC during WM by patients that have been noted for decades (
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ,
      • Manoach D.S.
      Prefrontal cortex dysfunction during working memory performance in schizophrenia: reconciling discrepant findings.
      ,
      • Van Snellenberg J.X.
      • Torres I.J.
      • Thornton A.E.
      Functional neuroimaging of working memory in schizophrenia: task performance as a moderating variable.
      ,
      • Van Snellenberg J.X.
      • Slifstein M.
      • Read C.
      • Weber J.
      • Thompson J.L.
      • Wager T.D.
      • et al.
      Dynamic shifts in brain network activation during supracapacity working memory task performance.
      ,
      • Hahn B.
      • Bae G.Y.
      • Robinson B.M.
      • Leonard C.J.
      • Luck S.J.
      • Gold J.M.
      Cortical hyperactivation at low working memory load: A primary processing abnormality in people with schizophrenia?.
      ,
      • Callicott J.H.
      • Mattay V.S.
      • Verchinski B.A.
      • Marenco S.
      • Egan M.F.
      • Weinberger D.R.
      Complexity of prefrontal cortical dysfunction in schizophrenia: more than up or down.
      ). As has been noted previously, care must be taken not to apply “hypofrontality” as a paradigm for explaining WM deficits observed in SCZ across tasks with varying psychometric properties (
      • Van Snellenberg J.X.
      • Torres I.J.
      • Thornton A.E.
      Functional neuroimaging of working memory in schizophrenia: task performance as a moderating variable.
      ,
      • Gur R.C.
      • Gur R.E.
      Hypofrontality in schizophrenia: RIP.
      ). One view is that varying findings could be due to an “inverted-U” relationship between dlPFC activation magnitude and WM load, which is left-shifted, attenuated, or blunted in SCZ (
      • Manoach D.S.
      Prefrontal cortex dysfunction during working memory performance in schizophrenia: reconciling discrepant findings.
      ,
      • Hahn B.
      • Bae G.Y.
      • Robinson B.M.
      • Leonard C.J.
      • Luck S.J.
      • Gold J.M.
      Cortical hyperactivation at low working memory load: A primary processing abnormality in people with schizophrenia?.
      ,
      • Callicott J.H.
      • Mattay V.S.
      • Verchinski B.A.
      • Marenco S.
      • Egan M.F.
      • Weinberger D.R.
      Complexity of prefrontal cortical dysfunction in schizophrenia: more than up or down.
      ). Using the self-ordered WM task (
      • Curtis C.E.
      • Zald D.H.
      • Pardo J.V.
      Organization of working memory within the human prefrontal cortex: a PET study of self-ordered object working memory.
      ,
      • Van Snellenberg J.X.
      • Conway A.R.
      • Spicer J.
      • Read C.
      • Smith E.E.
      Capacity estimates in working memory: Reliability and interrelationships among tasks.
      ,
      • Petrides M.
      • Milner B.
      Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man.
      ), which allows for the modulation of WM load from 0 through 7 items, this “inverted-U” has been observed and replicated in HCs (
      • Van Snellenberg J.X.
      • Slifstein M.
      • Read C.
      • Weber J.
      • Thompson J.L.
      • Wager T.D.
      • et al.
      Dynamic shifts in brain network activation during supracapacity working memory task performance.
      ), and blunting of the “inverted-U” has been observed in SCZ (
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ). The etiology of this blunting remains to be determined, but could reflect reduced WM capacity (
      • Gold J.M.
      • Hahn B.
      • Zhang W.W.
      • Robinson B.M.
      • Kappenman E.S.
      • Beck V.M.
      • et al.
      Reduced capacity but spared precision and maintenance of working memory representations in schizophrenia.
      ), inefficient allocation of cognitive resources (
      • Callicott J.H.
      • Mattay V.S.
      • Verchinski B.A.
      • Marenco S.
      • Egan M.F.
      • Weinberger D.R.
      Complexity of prefrontal cortical dysfunction in schizophrenia: more than up or down.
      ), or hyperfocusing of attention onto an aberrantly narrow range of task stimuli (
      • Hahn B.
      • Bae G.Y.
      • Robinson B.M.
      • Leonard C.J.
      • Luck S.J.
      • Gold J.M.
      Cortical hyperactivation at low working memory load: A primary processing abnormality in people with schizophrenia?.
      ,
      • Luck S.J.
      • Hahn B.
      • Leonard C.J.
      • Gold J.M.
      The Hyperfocusing Hypothesis: A New Account of Cognitive Dysfunction in Schizophrenia.
      ,
      • Gray B.E.
      • Hahn B.
      • Robinson B.
      • Harvey A.
      • Leonard C.J.
      • Luck S.J.
      • et al.
      Relationships between divided attention and working memory impairment in people with schizophrenia.
      ); however, this highlights that the expected direction, magnitude, and relationship between dlPFC activation and WM impairment in SCZ is likely difficult to predict a priori and may change substantively as a function of task properties.
      Conversely, our findings of an association between mPFC deactivation and WM performance, mPFC hypodeactivation in SCZ, and mediation of WM impairment in SCZ by mPFC deactivation, suggest that failure to deactivate mPFC during WM could be a robust marker of WM impairment. Similar associations between mPFC deactivation, WM performance, and SCZ diagnosis were observed in a small study (N = 12 in each group) of patients with SCZ, first-degree relatives of patients with SCZ, and HC (
      • Whitfield-Gabrieli S.
      • Thermenos H.W.
      • Milanovic S.
      • Tsuang M.T.
      • Faraone S.V.
      • McCarley R.W.
      • et al.
      Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
      ). Although published over a decade ago, mPFC has been the subject of considerably less investigation than dlPFC; nevertheless, a handful of studies have revealed similar findings (
      • Anticevic A.
      • Repovs G.
      • Shulman G.L.
      • Barch D.M.
      When less is more: TPJ and default network deactivation during encoding predicts working memory performance.
      ,
      • Van Snellenberg J.X.
      • Girgis R.R.
      • Horga G.
      • van de Giessen E.
      • Slifstein M.
      • Ojeil N.
      • et al.
      Mechanisms of Working Memory Impairment in Schizophrenia.
      ,
      • Eryilmaz H.
      • Tanner A.S.
      • Ho N.F.
      • Nitenson A.Z.
      • Silverstein N.J.
      • Petruzzi L.J.
      • et al.
      Disrupted Working Memory Circuitry in Schizophrenia: Disentangling fMRI Markers of Core Pathology vs Other Aspects of Impaired Performance.
      ). Our work bolsters this small literature with a moderately large sample of both unmedicated and medicated patients with SCZ who performed a well-studied WM task, and further establishes mPFC as a mediator of patient deficits in WM for the first time. Activation of the human mPFC, as a component of the DMN, is associated with task-negative states as well as engagement of internally-focused cognition such as autobiographical memory, social or self-referential inference, perspective-taking, future thinking, moral judgment, and mind wandering (
      • Smallwood J.
      • Bernhardt B.C.
      • Leech R.
      • Bzdok D.
      • Jefferies E.
      • Margulies D.S.
      The default mode network in cognition: a topographical perspective.
      ,
      • Mittner M.
      • Boekel W.
      • Tucker A.M.
      • Turner B.M.
      • Heathcote A.
      • Forstmann B.U.
      When the brain takes a break: a model-based analysis of mind wandering.
      ,
      • Buckner R.L.
      • DiNicola L.M.
      The brain's default network: updated anatomy, physiology and evolving insights.
      ,
      • Yeshurun Y.
      • Nguyen M.
      • Hasson U.
      The default mode network: where the idiosyncratic self meets the shared social world.
      ). WM impairments associated with reduced mPFC deactivation could thus result, at least in part, from an inability to suppress “task-irrelevant” cognitive processes or, relatedly, to insulate WM processes from internally-generated “distraction”.
      Finally, in a small sample of HCs (N = 14) and unmedicated SCZ (N = 9), we observed a relationship between greater amphetamine-induced dopamine release and reduced WM task-evoked deactivation in mPFC. This result is in direct contrast with our hypothesis, wherein we expected greater dopamine release capacity to be directly, rather than inversely, associated with greater deactivation. We also observed this relationship in HC alone, but not in unmedicated SCZ (N = 9), possibly due to either the small sample size or to the limited range of dopamine release in SCZ, as observed here and previously reported in the full sample (N = 20) (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ).
      Although speculative, our finding in HCs could reflect a role for prefrontal dopamine release in which, beyond a baseline level of dopaminergic stimulation, certain features of WM ability are enhanced, but at the expense of others (

      Cools R (2016): The costs and benefits of brain dopamine for cognitive control. Wiley Interdiscip Rev Cogn Sci. 7:317-329.

      ). This is consistent with an extension of the dual-state theory of mutually antagonistic D1 and D2 stimulation (
      • Durstewitz D.
      • Seamans J.K.
      The dual-state theory of prefrontal cortex dopamine function with relevance to catechol-o-methyltransferase genotypes and schizophrenia.
      ), the dynamics of which produce a balance between cognitive stability and flexibility in healthy individuals. In this view, “D1-dominant” states allow for superior stabilization of task stimuli, resistance to distractors, and set maintenance, but at the expense of conferring reduced efficiency in updating and manipulating items in WM, as well as set switching, which are conversely more optimal in “D2-dominant” states (

      Cools R (2016): The costs and benefits of brain dopamine for cognitive control. Wiley Interdiscip Rev Cogn Sci. 7:317-329.

      ,
      • Fallon S.J.
      • Zokaei N.
      • Norbury A.
      • Manohar S.G.
      • Husain M.
      Dopamine Alters the Fidelity of Working Memory Representations according to Attentional Demands.
      ,
      • Roberts A.C.
      • De Salvia M.A.
      • Wilkinson L.S.
      • Collins P.
      • Muir J.L.
      • Everitt B.J.
      • et al.
      6-Hydroxydopamine lesions of the prefrontal cortex in monkeys enhance performance on an analog of the Wisconsin Card Sort Test: possible interactions with subcortical dopamine.
      ,
      • Broadway J.M.
      • Frank M.J.
      • Cavanagh J.F.
      Dopamine D2 agonist affects visuospatial working memory distractor interference depending on individual differences in baseline working memory span.
      ). It is possible that updating efficiency in the HC sample may be inversely associated with mPFC dopamine release due to excess stability of internal representations of task stimuli; this would be consistent with a previous finding where administration of methylphenidate to HC participants prior to a visual WM task resulted in improved performance on trials involving distractor stimuli, but reduced performance on trials requiring unexpected updates to items stored in WM (
      • Fallon S.J.
      • Zokaei N.
      • Norbury A.
      • Manohar S.G.
      • Husain M.
      Dopamine Alters the Fidelity of Working Memory Representations according to Attentional Demands.
      ).
      In SCZ, patients exhibit an upregulation in cortical D1R (28, though a follow-up study found upregulation only in antipsychotic-naïve patients; see 29), and blunted cortical dopamine release capacity (
      • Slifstein M.
      • van de Giessen E.
      • Van Snellenberg J.
      • Thompson J.L.
      • Narendran R.
      • Gil R.
      • et al.
      Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
      ), and insufficient prefrontal D1R stimulation may produce in SCZ states that are too D2-dominant for optimal task performance. In this case, the result would be poor maintenance of items sored in WM, in addition to their loss and replacement with irrelevant stimuli, and mPFC hypodeactivation could reflect a general failure to suppress internally-generated distractors. In addition, a paucity of D1R stimulation could preclude the stabilization of cognitive attractor states that are more metabolically demanding, but optimal for WM task performance (
      • Braun U.
      • Harneit A.
      • Pergola G.
      • Menara T.
      • Schafer A.
      • Betzel R.F.
      • et al.
      Brain network dynamics during working memory are modulated by dopamine and diminished in schizophrenia.
      ). In mPFC, hypodeactivation could thus represent more generalized brain states that are less tailored to specific task demands, which could produce task-positive neural states that are less differentiable from the resting state in patients as compared to HCs. This reduced ability to enter optimized task-positive states could also explain the previous observation of reduced associations between WM task load and changes in interhemispheric frontoparietal network (FPN) connectivity in SCZ relative to HCs (
      • Cassidy C.M.
      • Van Snellenberg J.X.
      • Benavides C.
      • Slifstein M.
      • Wang Z.
      • Moore H.
      • et al.
      Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
      ), as load-dependent connectivity changes may result from refinement of a particular task-positive state that is generally less accessible to patients.

      Limitations

      Several limitations of this study warrant consideration. For example, inclusion of medicated patients presents the possibility of a medication confound, although medicated and unmedicated patients showed no difference in either task performance (Figure 3) or mPFC deactivation (Figure S1), making this unlikely. Further, the cross-modality PET+fMRI analysis was carried out using unmedicated patients and HC only. Sample size for PET/fMRI analysis was small, so these results should be interpreted cautiously and confirmation in an independent cohort should be sought. Additionally, individually hand- drawn mPFC ROIs for PET data analysis were not fully overlap with the fMRI mPFC ROI, averaging approximately 80% of PET ROI voxels encompassed by the fMRI ROI.
      Additionally, two features of the task design present limitations. First, the block design used here optimizes power but does not allow for more selective analysis of neural activity, such as isolation of correctly performed trials or detection of effects related to WM processes like encoding, maintenance, and retrieval. Second, due to the absence of a 0-back (no WM load) control task condition, neural activity during WM task blocks were compared relative to implicit baseline. This precludes differentiating effects specific to WM task engagement from those related to other, more general components of cognition, including sustained attention or visual processing effects, via cognitive subtraction. As a result, the impairment in mPFC deactivation detected in patients with SCZ relative to HC could reflect dysfunction either in processes specific to WM task engagement, in those generally associated with engagement in cognitively demanding (WM and non-WM) tasks, or both in combination. It is of note, however, that comparing task engagement to an implicit baseline may provide a clearer picture of general task-related impairments in SCZ as opposed to only those impairments associated with effects of WM load.

      Conclusion

      These findings suggest that failure to deactivate the mPFC during performance of a WM task is an important correlate, or even mechanism, of cognitive impairment in patients with SCZ. Furthermore, these results suggest an important role for mPFC deactivation in healthy individuals, with greater deactivation associated with better task performance regardless of diagnosis. Finally, PET-fMRI analysis revealed an inverse relationship between dopamine release capacity and WM task-induced mPFC deactivation in HC, but not in SCZ (possibly due to small sample size and range restriction); a larger study is needed to confirm these findings. Together, our results suggest that a deficiency in the ability of patients to suppress mPFC activity while engaged in WM tasks is a potential mechanism for WM impairment in SCZ, and thus a potential target for novel therapeutics, including pharmacotherapies and transcranial magnetic stimulation.

      Uncited reference

      • Girgis R.R.
      • Van Snellenberg J.X.
      • Glass A.
      • Kegeles L.S.
      • Thompson J.L.
      • Wall M.
      • et al.
      A proof-of-concept, randomized controlled trial of DAR-0100A, a dopamine-1 receptor agonist, for cognitive enhancement in schizophrenia.
      ,
      • Oldfield R.C.
      The assessment and analysis of handedness: the Edinburgh inventory.
      ,

      AB H (1975): Four factor index of social status. Working paper published by the author.

      ,
      • Kay S.R.
      • Fiszbein A.
      • Opler L.A.
      The positive and negative syndrome scale (PANSS) for schizophrenia.
      ,
      • Andreasen N.C.
      The Scale for the Assessment of Negative Symptoms (SANS): conceptual and theoretical foundations.
      ,

      Andreasen NC (1983): The Scale for the Assessment of Negative Symptoms (SANS). Department the Psychiatry: Iowa City.

      ,

      Andreasen NC (1984): The Scale for the Assessment of Positive Symptoms (SAPS). Department the Psychiatry: Iowa City.

      ,

      Williams JC, Tubiolo PN, Luceno JR, Van Snellenberg JX (2022): Advancing motion denoising of multiband resting-state functional connectivity fMRI data. Neuroimage.118907.

      ,
      • Winkler A.M.
      • Ridgway G.R.
      • Webster M.A.
      • Smith S.M.
      • Nichols T.E.
      Permutation inference for the general linear model.
      ,
      • Winkler A.M.
      • Webster M.A.
      • Brooks J.C.
      • Tracey I.
      • Smith S.M.
      • Nichols T.E.
      Non-parametric combination and related permutation tests for neuroimaging.
      ,
      • Winkler A.M.
      • Webster M.A.
      • Vidaurre D.
      • Nichols T.E.
      • Smith S.M.
      Multi-level block permutation.
      ,
      • Winkler A.M.
      • Ridgway G.R.
      • Douaud G.
      • Nichols T.E.
      • Smith S.M.
      Faster permutation inference in brain imaging.
      ,
      • Holmes A.P.
      • Blair R.C.
      • Watson J.D.
      • Ford I.
      Nonparametric analysis of statistic images from functional mapping experiments.
      ,
      • Smith S.M.
      • Nichols T.E.
      Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference.
      ,
      • Kriegeskorte N.
      • Simmons W.K.
      • Bellgowan P.S.
      • Baker C.I.
      Circular analysis in systems neuroscience: the dangers of double dipping.
      .

      Acknowledgements

      Assistance
      The authors would like to thank Dr. Anderson Winkler for his generous and comprehensive support throughout our use of the Permutation Analysis of Linear Models (PALM) software (99-103). We additionally acknowledge the substantial computing resources and technical assistance provided by Stony Brook Research Computing and Cyberinfrastructure, and the Institute for Advanced Computational Science (IACS) at Stony Brook University for access to the high-performance SeaWulf computing system, which was made possible by a $1.4M National Science Foundation grant (NSF Award #1531492; PI: Dr. Robert Harrison, Co-PI: Dr. Yuefan Deng).

      References

      1. Kraepelin E, Barclay RM, Robertson GM (1919): Dementia præcox and paraphrenia. Edinburgh,: E. & S. Livingstone.

      2. Bleuler E (1950): Dementia praecox; or, The group of schizophrenias. New York: International Universities Press.

        • Palmer B.W.
        • Dawes S.E.
        • Heaton R.K.
        What do we know about neuropsychological aspects of schizophrenia?.
        Neuropsychol Rev. 2009; 19: 365-384
        • Green M.F.
        • Harvey P.D.
        Cognition in schizophrenia: Past, present, and future.
        Schizophr Res Cogn. 2014; 1: e1-e9
        • Heinrichs R.W.
        • Zakzanis K.K.
        Neurocognitive deficit in schizophrenia: a quantitative review of the evidence.
        Neuropsychology. 1998; 12: 426-445
        • Keefe R.S.
        • Bilder R.M.
        • Davis S.M.
        • Harvey P.D.
        • Palmer B.W.
        • Gold J.M.
        • et al.
        Neurocognitive effects of antipsychotic medications in patients with chronic schizophrenia in the CATIE Trial.
        Arch Gen Psychiatry. 2007; 64: 633-647
        • Heinrichs R.W.
        Cognitive improvement in response to antipsychotic drugs: neurocognitive effects of antipsychotic medications in patients with chronic schizophrenia in the CATIE Trial.
        Arch Gen Psychiatry. 2007; 64: 631-632
        • Keefe R.S.
        • Sweeney J.A.
        • Gu H.
        • Hamer R.M.
        • Perkins D.O.
        • McEvoy J.P.
        • et al.
        Effects of olanzapine, quetiapine, and risperidone on neurocognitive function in early psychosis: a randomized, double-blind 52-week comparison.
        Am J Psychiatry. 2007; 164: 1061-1071
        • Green M.F.
        Cognition, drug treatment, and functional outcome in schizophrenia: a tale of two transitions.
        Am J Psychiatry. 2007; 164: 992-994
        • Mishara A.L.
        • Goldberg T.E.
        A meta-analysis and critical review of the effects of conventional neuroleptic treatment on cognition in schizophrenia: opening a closed book.
        Biol Psychiatry. 2004; 55: 1013-1022
        • Bowie C.R.
        • Harvey P.D.
        Cognitive deficits and functional outcome in schizophrenia.
        Neuropsychiatr Dis Treat. 2006; 2: 531-536
        • Lee J.
        • Park S.
        Working memory impairments in schizophrenia: a meta-analysis.
        Journal of abnormal psychology. 2005; 114: 599-611
        • Forbes N.F.
        • Carrick L.A.
        • McIntosh A.M.
        • Lawrie S.M.
        Working memory in schizophrenia: a meta-analysis.
        Psychological medicine. 2009; 39: 889-905
      3. Green MF (1998): Schizophrenia from a neurocognitive perspective : probing the impenetrable darkness. Boston: Allyn and Bacon.

        • Kern R.S.
        • Gold J.M.
        • Dickinson D.
        • Green M.F.
        • Nuechterlein K.H.
        • Baade L.E.
        • et al.
        The MCCB impairment profile for schizophrenia outpatients: results from the MATRICS psychometric and standardization study.
        Schizophrenia research. 2011; 126: 124-131
        • Mehta M.A.
        • Sahakian B.J.
        • McKenna P.J.
        • Robbins T.W.
        Systemic sulpiride in young adult volunteers simulates the profile of cognitive deficits in Parkinson's disease.
        Psychopharmacology (Berl). 1999; 146: 162-174
        • Cools R.
        • D'Esposito M.
        Inverted-U-shaped dopamine actions on human working memory and cognitive control.
        Biological psychiatry. 2011; 69: e113-125
        • Egan M.F.
        • Goldberg T.E.
        • Kolachana B.S.
        • Callicott J.H.
        • Mazzanti C.M.
        • Straub R.E.
        • et al.
        Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia.
        Proceedings of the National Academy of Sciences of the United States of America. 2001; 98: 6917-6922
        • Brozoski T.J.
        • Brown R.M.
        • Rosvold H.E.
        • Goldman P.S.
        Cognitive deficit caused by regional depletion of dopamine in prefrontal cortex of rhesus monkey.
        Science. 1979; 205: 929-932
        • Arnsten A.F.
        • Cai J.X.
        • Murphy B.L.
        • Goldman-Rakic P.S.
        Dopamine D1 receptor mechanisms in the cognitive performance of young adult and aged monkeys.
        Psychopharmacology (Berl). 1994; 116: 143-151
        • Dempster K.
        • Norman R.
        • Theberge J.
        • Densmore M.
        • Schaefer B.
        • Williamson P.
        Cognitive performance is associated with gray matter decline in first-episode psychosis.
        Psychiatry Res Neuroimaging. 2017; 264: 46-51
        • Haijma S.V.
        • Van Haren N.
        • Cahn W.
        • Koolschijn P.C.
        • Hulshoff Pol H.E.
        • Kahn R.S.
        Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects.
        Schizophr Bull. 2013; 39: 1129-1138
        • Weinberger D.R.
        • Egan M.F.
        • Bertolino A.
        • Callicott J.H.
        • Mattay V.S.
        • Lipska B.K.
        • et al.
        Prefrontal neurons and the genetics of schizophrenia.
        Biological psychiatry. 2001; 50: 825-844
        • Wible C.G.
        • Anderson J.
        • Shenton M.E.
        • Kricun A.
        • Hirayasu Y.
        • Tanaka S.
        • et al.
        Prefrontal cortex, negative symptoms, and schizophrenia: an MRI study.
        Psychiatry Res. 2001; 108: 65-78
        • Glantz L.A.
        • Lewis D.A.
        Decreased dendritic spine density on prefrontal cortical pyramidal neurons in schizophrenia.
        Arch Gen Psychiatry. 2000; 57: 65-73
        • Smucny J.
        • Dienel S.J.
        • Lewis D.A.
        • Carter C.S.
        Mechanisms underlying dorsolateral prefrontal cortex contributions to cognitive dysfunction in schizophrenia.
        Neuropsychopharmacology. 2021;
        • Lewis D.A.
        • Curley A.A.
        • Glausier J.R.
        • Volk D.W.
        Cortical parvalbumin interneurons and cognitive dysfunction in schizophrenia.
        Trends Neurosci. 2012; 35: 57-67
        • Abi-Dargham A.
        • Mawlawi O.
        • Lombardo I.
        • Gil R.
        • Martinez D.
        • Huang Y.
        • et al.
        Prefrontal dopamine D1 receptors and working memory in schizophrenia.
        The Journal of neuroscience : the official journal of the Society for Neuroscience. 2002; 22: 3708-3719
        • Abi-Dargham A.
        • Xu X.
        • Thompson J.L.
        • Gil R.
        • Kegeles L.S.
        • Urban N.
        • et al.
        Increased prefrontal cortical D(1) receptors in drug naive patients with schizophrenia: a PET study with [(1)(1)C]NNC112.
        Journal of psychopharmacology. 2012; 26: 794-805
        • Abi-Dargham A.
        • Javitch J.A.
        • Slifstein M.
        • Anticevic A.
        • Calkins M.E.
        • Cho Y.T.
        • et al.
        Dopamine D1R Receptor Stimulation as a Mechanistic Pro-cognitive Target for Schizophrenia.
        Schizophrenia bulletin. 2021;
        • Kuperberg G.R.
        • Broome M.R.
        • McGuire P.K.
        • David A.S.
        • Eddy M.
        • Ozawa F.
        • et al.
        Regionally localized thinning of the cerebral cortex in schizophrenia.
        Arch Gen Psychiatry. 2003; 60: 878-888
        • Slifstein M.
        • van de Giessen E.
        • Van Snellenberg J.
        • Thompson J.L.
        • Narendran R.
        • Gil R.
        • et al.
        Deficits in prefrontal cortical and extrastriatal dopamine release in schizophrenia: a positron emission tomographic functional magnetic resonance imaging study.
        JAMA psychiatry. 2015; 72: 316-324
        • Smallwood J.
        • Bernhardt B.C.
        • Leech R.
        • Bzdok D.
        • Jefferies E.
        • Margulies D.S.
        The default mode network in cognition: a topographical perspective.
        Nat Rev Neurosci. 2021; 22: 503-513
        • Mittner M.
        • Boekel W.
        • Tucker A.M.
        • Turner B.M.
        • Heathcote A.
        • Forstmann B.U.
        When the brain takes a break: a model-based analysis of mind wandering.
        J Neurosci. 2014; 34: 16286-16295
        • Buckner R.L.
        • DiNicola L.M.
        The brain's default network: updated anatomy, physiology and evolving insights.
        Nat Rev Neurosci. 2019; 20: 593-608
        • Yeshurun Y.
        • Nguyen M.
        • Hasson U.
        The default mode network: where the idiosyncratic self meets the shared social world.
        Nat Rev Neurosci. 2021; 22: 181-192
        • Vatansever D.
        • Menon D.K.
        • Stamatakis E.A.
        Default mode contributions to automated information processing.
        Proc Natl Acad Sci U S A. 2017; 114: 12821-12826
        • Anticevic A.
        • Cole M.W.
        • Murray J.D.
        • Corlett P.R.
        • Wang X.J.
        • Krystal J.H.
        The role of default network deactivation in cognition and disease.
        Trends in cognitive sciences. 2012; 16: 584-592
        • Anticevic A.
        • Repovs G.
        • Shulman G.L.
        • Barch D.M.
        When less is more: TPJ and default network deactivation during encoding predicts working memory performance.
        NeuroImage. 2010; 49: 2638-2648
        • Van Snellenberg J.X.
        • Girgis R.R.
        • Horga G.
        • van de Giessen E.
        • Slifstein M.
        • Ojeil N.
        • et al.
        Mechanisms of Working Memory Impairment in Schizophrenia.
        Biological psychiatry. 2016; 80: 617-626
        • Braun U.
        • Harneit A.
        • Pergola G.
        • Menara T.
        • Schafer A.
        • Betzel R.F.
        • et al.
        Brain network dynamics during working memory are modulated by dopamine and diminished in schizophrenia.
        Nat Commun. 2021; 12: 3478
        • Pomarol-Clotet E.
        • Salvador R.
        • Sarro S.
        • Gomar J.
        • Vila F.
        • Martinez A.
        • et al.
        Failure to deactivate in the prefrontal cortex in schizophrenia: dysfunction of the default mode network?.
        Psychol Med. 2008; 38: 1185-1193
        • Whitfield-Gabrieli S.
        • Thermenos H.W.
        • Milanovic S.
        • Tsuang M.T.
        • Faraone S.V.
        • McCarley R.W.
        • et al.
        Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
        Proc Natl Acad Sci U S A. 2009; 106: 1279-1284
        • Hahn B.
        • Harvey A.N.
        • Gold J.M.
        • Ross T.J.
        • Stein E.A.
        Load-dependent hyperdeactivation of the default mode network in people with schizophrenia.
        Schizophr Res. 2017; 185: 190-196
        • Kim D.I.
        • Manoach D.S.
        • Mathalon D.H.
        • Turner J.A.
        • Mannell M.
        • Brown G.G.
        • et al.
        Dysregulation of working memory and default-mode networks in schizophrenia using independent component analysis, an fBIRN and MCIC study.
        Hum Brain Mapp. 2009; 30: 3795-3811
        • Metzak P.D.
        • Riley J.D.
        • Wang L.
        • Whitman J.C.
        • Ngan E.T.
        • Woodward T.S.
        Decreased efficiency of task-positive and task-negative networks during working memory in schizophrenia.
        Schizophr Bull. 2012; 38: 803-813
        • Wu D.
        • Jiang T.
        Schizophrenia-related abnormalities in the triple network: a meta-analysis of working memory studies.
        Brain Imaging Behav. 2020; 14: 971-980
        • Zhou L.
        • Pu W.
        • Wang J.
        • Liu H.
        • Wu G.
        • Liu C.
        • et al.
        Inefficient DMN Suppression in Schizophrenia Patients with Impaired Cognitive Function but not Patients with Preserved Cognitive Function.
        Sci Rep. 2016; 621657
      4. Weinberger DR, Berman KF, Daniel DG (1992): Mesoprefrontal cortical dopaminergic activity and prefrontal hypofunction in schizophrenia. Clin Neuropharmacol. 15 Suppl 1 Pt A:568A-569A.

        • Seamans J.K.
        • Gorelova N.
        • Durstewitz D.
        • Yang C.R.
        Bidirectional dopamine modulation of GABAergic inhibition in prefrontal cortical pyramidal neurons.
        J Neurosci. 2001; 21: 3628-3638
        • Seamans J.K.
        • Yang C.R.
        The principal features and mechanisms of dopamine modulation in the prefrontal cortex.
        Prog Neurobiol. 2004; 74: 1-58
        • O'Donnell P.
        Cortical disinhibition in the neonatal ventral hippocampal lesion model of schizophrenia: new vistas on possible therapeutic approaches.
        Pharmacol Ther. 2012; 133: 19-25
        • O'Donnell P.
        Adolescent onset of cortical disinhibition in schizophrenia: insights from animal models.
        Schizophr Bull. 2011; 37: 484-492
        • Arnsten A.F.
        • Wang M.J.
        • Paspalas C.D.
        Neuromodulation of thought: flexibilities and vulnerabilities in prefrontal cortical network synapses.
        Neuron. 2012; 76: 223-239
        • Cassidy C.M.
        • Van Snellenberg J.X.
        • Benavides C.
        • Slifstein M.
        • Wang Z.
        • Moore H.
        • et al.
        Dynamic Connectivity between Brain Networks Supports Working Memory: Relationships to Dopamine Release and Schizophrenia.
        The Journal of neuroscience : the official journal of the Society for Neuroscience. 2016; 36: 4377-4388
        • Goldman-Rakic P.S.
        Working memory dysfunction in schizophrenia.
        The Journal of neuropsychiatry and clinical neurosciences. 1994; 6: 348-357
        • Weinberger D.R.
        • Berman K.F.
        Prefrontal function in schizophrenia: confounds and controversies.
        Philos Trans R Soc Lond B Biol Sci. 1996; 351: 1495-1503
        • Carter C.S.
        • Perlstein W.
        • Ganguli R.
        • Brar J.
        • Mintun M.
        • Cohen J.D.
        Functional hypofrontality and working memory dysfunction in schizophrenia.
        The American journal of psychiatry. 1998; 155: 1285-1287
        • Weinberger D.R.
        • Berman K.F.
        • Zec R.F.
        Physiologic dysfunction of dorsolateral prefrontal cortex in schizophrenia. I. Regional cerebral blood flow evidence.
        Archives of general psychiatry. 1986; 43: 114-124
        • Manoach D.S.
        Prefrontal cortex dysfunction during working memory performance in schizophrenia: reconciling discrepant findings.
        Schizophrenia research. 2003; 60: 285-298
        • Van Snellenberg J.X.
        • Torres I.J.
        • Thornton A.E.
        Functional neuroimaging of working memory in schizophrenia: task performance as a moderating variable.
        Neuropsychology. 2006; 20: 497-510
        • Van Snellenberg J.X.
        • Slifstein M.
        • Read C.
        • Weber J.
        • Thompson J.L.
        • Wager T.D.
        • et al.
        Dynamic shifts in brain network activation during supracapacity working memory task performance.
        Human brain mapping. 2015; 36: 1245-1264
        • Eryilmaz H.
        • Tanner A.S.
        • Ho N.F.
        • Nitenson A.Z.
        • Silverstein N.J.
        • Petruzzi L.J.
        • et al.
        Disrupted Working Memory Circuitry in Schizophrenia: Disentangling fMRI Markers of Core Pathology vs Other Aspects of Impaired Performance.
        Neuropsychopharmacology. 2016; 41: 2411-2420
        • Lewis D.A.
        • Sweet R.A.
        Schizophrenia from a neural circuitry perspective: advancing toward rational pharmacological therapies.
        J Clin Invest. 2009; 119: 706-716
        • Barch D.M.
        • Carter C.S.
        • Braver T.S.
        • Sabb F.W.
        • MacDonald 3rd, A.
        • Noll D.C.
        • et al.
        Selective deficits in prefrontal cortex function in medication-naive patients with schizophrenia.
        Archives of general psychiatry. 2001; 58: 280-288
        • Williams G.V.
        • Goldman-Rakic P.S.
        Modulation of memory fields by dopamine D1 receptors in prefrontal cortex.
        Nature. 1995; 376: 572-575
        • Sawaguchi T.
        • Goldman-Rakic P.S.
        D1 dopamine receptors in prefrontal cortex: involvement in working memory.
        Science. 1991; 251: 947-950
        • Narendran R.
        • Frankle W.G.
        • Mason N.S.
        • Rabiner E.A.
        • Gunn R.N.
        • Searle G.E.
        • et al.
        Positron emission tomography imaging of amphetamine-induced dopamine release in the human cortex: a comparative evaluation of the high affinity dopamine D2/3 radiotracers [11C]FLB 457 and [11C]fallypride.
        Synapse. 2009; 63: 447-461
        • Narendran R.
        • Mason N.S.
        • May M.A.
        • Chen C.M.
        • Kendro S.
        • Ridler K.
        • et al.
        Positron emission tomography imaging of dopamine D(2)/(3) receptors in the human cortex with [(1)(1)C]FLB 457: reproducibility studies.
        Synapse. 2011; 65: 35-40
        • Narendran R.
        • Jedema H.P.
        • Lopresti B.J.
        • Mason N.S.
        • Gurnsey K.
        • Ruszkiewicz J.
        • et al.
        Imaging dopamine transmission in the frontal cortex: a simultaneous microdialysis and [11C]FLB 457 PET study.
        Mol Psychiatry. 2014; 19: 302-310
        • Girgis R.R.
        • Van Snellenberg J.X.
        • Glass A.
        • Kegeles L.S.
        • Thompson J.L.
        • Wall M.
        • et al.
        A proof-of-concept, randomized controlled trial of DAR-0100A, a dopamine-1 receptor agonist, for cognitive enhancement in schizophrenia.
        Journal of psychopharmacology. 2016; 30: 428-435
        • Oldfield R.C.
        The assessment and analysis of handedness: the Edinburgh inventory.
        Neuropsychologia. 1971; 9: 97-113
      5. AB H (1975): Four factor index of social status. Working paper published by the author.

        • Kay S.R.
        • Fiszbein A.
        • Opler L.A.
        The positive and negative syndrome scale (PANSS) for schizophrenia.
        Schizophrenia bulletin. 1987; 13: 261-276
        • Andreasen N.C.
        The Scale for the Assessment of Negative Symptoms (SANS): conceptual and theoretical foundations.
        Br J Psychiatry. 1989; Suppl: 49-58
      6. Andreasen NC (1983): The Scale for the Assessment of Negative Symptoms (SANS). Department the Psychiatry: Iowa City.

      7. Andreasen NC (1984): The Scale for the Assessment of Positive Symptoms (SAPS). Department the Psychiatry: Iowa City.

        • Curtis C.E.
        • Zald D.H.
        • Pardo J.V.
        Organization of working memory within the human prefrontal cortex: a PET study of self-ordered object working memory.
        Neuropsychologia. 2000; 38: 1503-1510
        • Moore M.E.
        • Ross B.M.
        Context effects in running memory.
        Psychological Reports. 1963; 12: 451-465
        • Ross B.M.
        Serial order as a unique source of error in running memory.
        Percept Mot Skills. 1966; 23: 195-209
        • Ross B.M.
        Serial-order effects in two-channel running memory.
        Percept Mot Skills. 1966; 23: 1099-1107
        • Gevins A.S.
        • Bressler S.L.
        • Cutillo B.A.
        • Illes J.
        • Miller J.C.
        • Stern J.
        • et al.
        Effects of prolonged mental work on functional brain topography.
        Electroencephalogr Clin Neurophysiol. 1990; 76: 339-350
        • Esteban O.
        • Markiewicz C.J.
        • Blair R.W.
        • Moodie C.A.
        • Isik A.I.
        • Erramuzpe A.
        • et al.
        fMRIPrep: a robust preprocessing pipeline for functional MRI.
        Nature methods. 2019; 16: 111-116
        • Esteban O.
        • Ciric R.
        • Finc K.
        • Blair R.W.
        • Markiewicz C.J.
        • Moodie C.A.
        • et al.
        Analysis of task-based functional MRI data preprocessed with fMRIPrep.
        Nat Protoc. 2020; 15: 2186-2202
      8. Esteban O, Markiewicz CJ, Goncalves M, DuPre E, Kent JD, Salo T, et al. (2021): fMRIPrep: a robust preprocessing pipeline for functional MRI. 20.0.7 ed: Zenodo.

        • Abi-Dargham A.
        • Martinez D.
        • Mawlawi O.
        • Simpson N.
        • Hwang D.R.
        • Slifstein M.
        • et al.
        Measurement of striatal and extrastriatal dopamine D1 receptor binding potential with [11C]NNC 112 in humans: validation and reproducibility.
        J Cereb Blood Flow Metab. 2000; 20: 225-243
        • Kegeles L.S.
        • Slifstein M.
        • Xu X.
        • Urban N.
        • Thompson J.L.
        • Moadel T.
        • et al.
        Striatal and extrastriatal dopamine D2/D3 receptors in schizophrenia evaluated with [18F]fallypride positron emission tomography.
        Biological psychiatry. 2010; 68: 634-641
      9. Gibbons JD, Chakraborti S (2021): Nonparametric statistical inference. 6th edition. ed. Boca Raton: CRC Press.

      10. Hollander M, Wolfe DA, Chicken E (2014): Nonparametric statistical methods. Third edition / ed. Hoboken, New Jersey: John Wiley & Sons, Inc.

        • Lilliefors H.W.
        On the Kolmogorov-Smirnov Test for Normality with Mean and Variance Unknown.
        Journal of the American Statistical Association. 1967; 62: 399-402
      11. Conover WJ (1999): Practical nonparametric statistics. 3rd ed. New York: Wiley.

      12. Friston KJ, Ashburner J, Kiebel S, Nichols T, Penny WD (2007): Statistical parametric mapping : the analysis of funtional brain images. First edition. ed. Amsterdam ; Boston: Elsevier/Academic Press, pp 1 online resource (vii, 647 pages).

        • Friston K.J.
        • Jezzard P.
        • Turner R.
        Analysis of functional MRI time-series.
        Human Brain Mapping. 1994; 1: 153-171
        • Friston K.J.
        • Holmes A.P.
        • Worsley K.J.
        • Poline J.P.
        • Frith C.D.
        • Frackowiak R.S.J.
        Statistical parametric maps in functional imaging: A general linear approach.
        Human Brain Mapping. 1994; 2: 189-210
        • Friston K.J.
        • Fletcher P.
        • Josephs O.
        • Holmes A.
        • Rugg M.D.
        • Turner R.
        Event-related fMRI: characterizing differential responses.
        Neuroimage. 1998; 7: 30-40
        • Friston K.J.
        • Penny W.
        • David O.
        Modeling brain responses.
        Int Rev Neurobiol. 2005; 66: 89-124
        • Henson R.N.
        • Price C.J.
        • Rugg M.D.
        • Turner R.
        • Friston K.J.
        Detecting latency differences in event-related BOLD responses: application to words versus nonwords and initial versus repeated face presentations.
        Neuroimage. 2002; 15: 83-97
      13. Williams JC, Tubiolo PN, Luceno JR, Van Snellenberg JX (2022): Advancing motion denoising of multiband resting-state functional connectivity fMRI data. Neuroimage.118907.

        • Winkler A.M.
        • Ridgway G.R.
        • Webster M.A.
        • Smith S.M.
        • Nichols T.E.
        Permutation inference for the general linear model.
        NeuroImage. 2014; 92: 381-397
        • Winkler A.M.
        • Webster M.A.
        • Brooks J.C.
        • Tracey I.
        • Smith S.M.
        • Nichols T.E.
        Non-parametric combination and related permutation tests for neuroimaging.
        Hum Brain Mapp. 2016; 37: 1486-1511
        • Alberton B.A.V.
        • Nichols T.E.
        • Gamba H.R.
        • Winkler A.M.
        Multiple testing correction over contrasts for brain imaging.
        Neuroimage. 2020; 216116760
        • Winkler A.M.
        • Webster M.A.
        • Vidaurre D.
        • Nichols T.E.
        • Smith S.M.
        Multi-level block permutation.
        Neuroimage. 2015; 123: 253-268
        • Winkler A.M.
        • Ridgway G.R.
        • Douaud G.
        • Nichols T.E.
        • Smith S.M.
        Faster permutation inference in brain imaging.
        Neuroimage. 2016; 141: 502-516
        • Holmes A.P.
        • Blair R.C.
        • Watson J.D.
        • Ford I.
        Nonparametric analysis of statistic images from functional mapping experiments.
        J Cereb Blood Flow Metab. 1996; 16: 7-22
        • Smith S.M.
        • Nichols T.E.
        Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference.
        Neuroimage. 2009; 44: 83-98
        • Gordon E.M.
        • Laumann T.O.
        • Adeyemo B.
        • Huckins J.F.
        • Kelley W.M.
        • Petersen S.E.
        Generation and Evaluation of a Cortical Area Parcellation from Resting-State Correlations.
        Cereb Cortex. 2016; 26: 288-303
        • Wager T.D.
        • Davidson M.L.
        • Hughes B.L.
        • Lindquist M.A.
        • Ochsner K.N.
        Prefrontal-subcortical pathways mediating successful emotion regulation.
        Neuron. 2008; 59: 1037-1050
        • Geuter S.
        • Reynolds Losin E.A.
        • Roy M.
        • Atlas L.Y.
        • Schmidt L.
        • Krishnan A.
        • et al.
        Multiple Brain Networks Mediating Stimulus-Pain Relationships in Humans.
        Cereb Cortex. 2020; 30: 4204-4219
        • Shrout P.E.
        • Bolger N.
        Mediation in experimental and nonexperimental studies: new procedures and recommendations.
        Psychol Methods. 2002; 7: 422-445
        • Kenny D.A.
        • Korchmaros J.D.
        • Bolger N.
        Lower level mediation in multilevel models.
        Psychol Methods. 2003; 8: 115-128
      14. Efron B, Tibshirani R (1993): An introduction to the bootstrap. New York: Chapman & Hall.

        • DiCiccio T.J.
        • Efron B.
        Bootstrap confidence intervals.
        Statistical Science. 1996; 11: 189-228
        • Power J.D.
        • Barnes K.A.
        • Snyder A.Z.
        • Schlaggar B.L.
        • Petersen S.E.
        Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion.
        NeuroImage. 2012; 59: 2142-2154
        • Kriegeskorte N.
        • Simmons W.K.
        • Bellgowan P.S.
        • Baker C.I.
        Circular analysis in systems neuroscience: the dangers of double dipping.
        Nat Neurosci. 2009; 12: 535-540
        • Inoue M.
        • Suhara T.
        • Sudo Y.
        • Okubo Y.
        • Yasuno F.
        • Kishimoto T.
        • et al.
        Age-related reduction of extrastriatal dopamine D2 receptor measured by PET.
        Life Sci. 2001; 69: 1079-1084
        • Kaasinen V.
        • Vilkman H.
        • Hietala J.
        • Nagren K.
        • Helenius H.
        • Olsson H.
        • et al.
        Age-related dopamine D2/D3 receptor loss in extrastriatal regions of the human brain.
        Neurobiol Aging. 2000; 21: 683-688
        • Gur R.C.
        • Gur R.E.
        Hypofrontality in schizophrenia: RIP.
        Lancet. 1995; 345: 1383-1384
        • Hahn B.
        • Bae G.Y.
        • Robinson B.M.
        • Leonard C.J.
        • Luck S.J.
        • Gold J.M.
        Cortical hyperactivation at low working memory load: A primary processing abnormality in people with schizophrenia?.
        Neuroimage Clin. 2020; 26102270
        • Callicott J.H.
        • Mattay V.S.
        • Verchinski B.A.
        • Marenco S.
        • Egan M.F.
        • Weinberger D.R.
        Complexity of prefrontal cortical dysfunction in schizophrenia: more than up or down.
        The American journal of psychiatry. 2003; 160: 2209-2215
        • Van Snellenberg J.X.
        • Conway A.R.
        • Spicer J.
        • Read C.
        • Smith E.E.
        Capacity estimates in working memory: Reliability and interrelationships among tasks.
        Cognitive, affective & behavioral neuroscience. 2014; 14: 106-116
        • Petrides M.
        • Milner B.
        Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man.
        Neuropsychologia. 1982; 20: 249-262
        • Gold J.M.
        • Hahn B.
        • Zhang W.W.
        • Robinson B.M.
        • Kappenman E.S.
        • Beck V.M.
        • et al.
        Reduced capacity but spared precision and maintenance of working memory representations in schizophrenia.
        Arch Gen Psychiatry. 2010; 67: 570-577
        • Luck S.J.
        • Hahn B.
        • Leonard C.J.
        • Gold J.M.
        The Hyperfocusing Hypothesis: A New Account of Cognitive Dysfunction in Schizophrenia.
        Schizophr Bull. 2019; 45: 991-1000
        • Gray B.E.
        • Hahn B.
        • Robinson B.
        • Harvey A.
        • Leonard C.J.
        • Luck S.J.
        • et al.
        Relationships between divided attention and working memory impairment in people with schizophrenia.
        Schizophr Bull. 2014; 40: 1462-1471
      15. Cools R (2016): The costs and benefits of brain dopamine for cognitive control. Wiley Interdiscip Rev Cogn Sci. 7:317-329.

        • Durstewitz D.
        • Seamans J.K.
        The dual-state theory of prefrontal cortex dopamine function with relevance to catechol-o-methyltransferase genotypes and schizophrenia.
        Biol Psychiatry. 2008; 64: 739-749
        • Fallon S.J.
        • Zokaei N.
        • Norbury A.
        • Manohar S.G.
        • Husain M.
        Dopamine Alters the Fidelity of Working Memory Representations according to Attentional Demands.
        J Cogn Neurosci. 2017; 29: 728-738
        • Roberts A.C.
        • De Salvia M.A.
        • Wilkinson L.S.
        • Collins P.
        • Muir J.L.
        • Everitt B.J.
        • et al.
        6-Hydroxydopamine lesions of the prefrontal cortex in monkeys enhance performance on an analog of the Wisconsin Card Sort Test: possible interactions with subcortical dopamine.
        J Neurosci. 1994; 14: 2531-2544
        • Broadway J.M.
        • Frank M.J.
        • Cavanagh J.F.
        Dopamine D2 agonist affects visuospatial working memory distractor interference depending on individual differences in baseline working memory span.
        Cogn Affect Behav Neurosci. 2018; 18: 509-520