Fronto-temporal alterations within the first 200 ms during an attentional task distinguish major depression, non-clinical participants with depressed mood and healthy controls: a potential biomarker?

Attentional impairment in depression is a cardinal feature of depression and has been proposed as a candidate endophenotype for major depressive disorder. Event-related potentials (ERPs) elicited by oddball signal detection tasks provide objective markers of selective stimulus processing, and are pertinent endophenotypic markers for depression. While previous studies have sought to determine objective markers for attentional impairment in depression, evidence is inconsistent and may involve heterogeneity in relatively small samples. Here, we brought together oddball ERP recording with source localization of neural correlates of selective attention in outpatients with major depressive disorder (MDD; n = 78) and participants with depressed mood (PDM; n = 127) relative to healthy controls (CTL; n = 116). The key finding was a dimensional exaggeration of the P200 (140-270 ms) to both target (signal) and non-target (noise) stimuli, most pronounced in MDD, followed by PDM, relative to CTL. This exaggeration was coupled with slower and more variable response times, suggesting that neural systems are attempting to compensate for a difficulty in discriminating signal from noise. P200 alterations were localised to limbic (hippocampal), temporal and ventral prefrontal regions, key components of the signal detection network. A subsequent reduction and delay in the P300 was also revealed for MDD indicating that the pronounced lack of discrimination in clinical depression may also lead to impaired stimulus evaluation. This P200 increase in depression could provide a potential mechanism for the attentional impairment frequently observed in depression and consequent alterations in the P300 may differentiate clinically significant depression.

Investigating models of affect: relationships among EEG alpha asymmetry, depression, and anxiety.

The approach-withdrawal and valence-arousal models both predict that depressive and anxious profiles will be associated with relatively reduced left frontal and increased right frontal activity respectively, while the valence-arousal model also proposes a dissociation by lower and higher right parietotemporal activity, respectively. Recent work further suggests that subtypes of anxiety disorders may be characterized by distinctive patterns of activity depending on their type of arousal (anxious arousal/apprehension). The aim of this study was to investigate the relationships among nonclinical depression/anxiety and lateralized frontal/parietotemporal activity by categorizing participants (N=428) on the basis of both negative mood and alpha EEG. Key findings include: (i) greater right frontal lateralization in anxious participants, symmetrical frontal activity in depressed/comorbid, and left frontal lateralization in healthy controls; (ii) right frontal lateralization in anxious arousal participants, left frontal and right parietotemporal lateralization in anxious apprehension; (iii) bilateral increase in frontal and increased right parietotemporal activity in depressed/comorbid participants. Findings support predictions for frontal but not posterior regions. Grouping on the basis of EEG may not be reciprocally predictive of negative mood groupings, suggesting involvement of additional factors.

Resting electroencephalogram asymmetry and posttraumatic stress disorder.

The valence-arousal (W. Heller, 1993) and approach-withdrawal (R. J. Davidson, 1998a) models hypothesize that particular patterns of hemispheric brain activity are associated with specific motivational tendencies and psychopathologies. We tested several of these predictions in two groups-a posttraumatic stress disorder (PTSD) and a "supercontrol" group, selected to be maximally different from those with PTSD. Contrary to almost all hypotheses, individuals with PTSD did not differ from controls on resting electroencephalogram (EEG) asymmetry. Particular aspects of PTSD were also not related to EEG hemisphere differences. Our null findings are consistent with the few studies that have examined resting EEG asymmetries in PTSD and suggest that PTSD may be associated with different processes than psychopathologies previously examined in studies of hemispheric brain activity (e.g., major depressive disorder, panic disorder).

Integrating objective gene-brain-behavior markers of psychiatric disorders.

There is little consensus about which objective markers should be used to assess major psychiatric disorders, and predict/evaluate treatment response for these disorders. Clinical practice relies instead on subjective signs and symptoms, such that there is a "translational gap" between research findings and clinical practice. This gap arises from: a) a lack of integrative theoretical models which provide a basis for understanding links between gene-brain-behavior mechanisms and clinical entities; b) the reliance on studying one measure at a time so that linkages between markers are their specificity are not established; and c) the lack of a definitive understanding of what constitutes normative function. Here, we draw on a standardized methodology for acquiring multiple sources of genomic, brain and behavioral data in the same subjects, to propose candidate markers of selected psychiatric disorders: depression, post-traumatic stress disorder, schizophrenia, attention-deficit/hyperactivity disorder and dementia disorders. This methodology has been used to establish a standardized international database which provides a comprehensive framework and the basis for testing hypotheses derived from an integrative theoretical model of the brain. Using this normative base, we present preliminary findings for a number of disorders in relation to the proposed markers. Establishing these objective markers will be the first step towards determining their sensitivity, specificity and treatment prediction in individual patients.

Predicting severity of non-clinical depression: preliminary findings using an integrated approach.

INTRODUCTION: Depression is characterized by disturbances in affect, cognition, brain and body function, yet studies have tended to focus on single domains of dysfunction. An integrated approach may provide a more complete profile of the range of deficits characterized by depressed individuals, but it is unclear whether this approach is able to predict depression severity over and above that predicted by single tasks or domains of function. In this study, we examined the value of combining multiple domains of function in predicting depression severity. METHODS: Participants contained in the International Brain Database, (http://www.brainresource.com) had completed three testing components including a web-based questionnaire of Personal History, the Brain Resource Cognition battery of Neuropsychological tests, Personality assessment and Psychophysiological testing. Two hundred and sixty six of these participants were able to be classified as either non-depressed, mild-moderately or severely (non-clinically) depressed, based on a depression screening questionnaire. Analysis of variance identified variables on which the categorized participants differed. Significant variables were then entered into a series of stepwise regressions to examine their ability to predict depression scores. RESULTS: An integrated model including measures of affect (increased Neuroticism; decreased Emotional Intelligence), cognition (increased variability of reaction time during a working memory task; decreased "name the word component score" in the verbal interference task), brain (decreased left-lateralized P150 ERP component during a working memory task) and body function (increased negative skin conductance level gradient) were found to predict more of the variation in depression severity than any single domain of function. DISCUSSION: On the basis of behavioral as well as Psychophysiological findings reported in this study, it was suggested that deficits in subclinically depressed individuals are more pronounced during automatic stages of stimulus processing, and that performance in these individuals may improve (to the level displayed by controls) when task demands are increased. Findings also suggest that it is important to consider disturbances across different domains of function in order to elucidate depression severity. Each domain may contribute unique explanatory information consistent with an integrative model of depression, taking into account the role of both behavior and underlying neural changes.