Error-related psychophysiology and negative affect☆
Introduction
The error-related negativity (ERN or Ne) is a response-locked event-related brain potential (ERP) observed at fronto-central (Fz, FCz, Cz) recording sites that begins around the time of an erroneous response, and peaks 50–100 ms later (Falkenstein, Hohnsbein, Hoormann, & Blanke, 1991; Falkenstein, Hoormann, Christ, & Hohnsbein, 2000; Gehring, Coles, Meyer, & Donchin, 1990; Gehring, Goss, Coles, Meyer, & Donchin, 1993). The ERN has been observed across a variety of stimulus and response modalities, and appears to reflect the activity of a generic response-monitoring system (Bernstein, Scheffers, & Coles, 1995; Falkenstein et al., 1991; Holroyd, Dien, & Coles, 1998; Van ‘t Ent & Apkarian, 1999). In terms of the source of the ERN, studies utilizing source localization have consistently found that error-related brain activity can be best described by a neural generator in the anterior cingulate cortex (Dehaene, Posner, & Tucker, 1994; Holroyd et al., 1998).
Numerous studies have found that the ERN is also sensitive to motivational and contextual factors. For instance, increased focus on accuracy over speed has been found to increase the magnitude of the ERN; similarly, the ERN appears larger when subjects are more certain that they have made a mistake—suggesting that the response-monitoring system is sensitive to motivational factors during performance (Falkenstein et al., 2000; Gehring et al., 1993).
Individual differences in anxiety have also been found to influence the ERN. Specifically, Gehring, Himle, and Nisenson (2000) found that patients with obsessive-compulsive disorder (OCD) have significantly larger ERNs than age-matched control subjects (for similar results in patients with OCD, see Johannes et al., 2001). In studies from our own laboratory, we have likewise reported enhanced error-related brain activity in obsessive-compulsive and worried undergraduates when compared to both control subjects and subjects with specific phobias (Hajcak & Simons, 2002; Hajcak, McDonald, & Simons, 2003b).
These results, linking anxiety to enhanced response monitoring, are consistent with both anxious behavior and the pathophysiology of anxiety. Specifically, OCD and pathological worry have been shown to relate to increased doubts about actions and concerns over mistakes (Frost & Steketee, 1997; Hajcak & Huppert, 2004; Stober & Joormann, 2001). Furthermore, the anterior cingulate cortex has been shown to be hyperactive in many anxiety disorders and this has led some researchers to postulate that anterior cingulate cortex dysfunction may be related to the experience of symptoms common to all anxiety disorders (Kimbrell et al., 1999; Malizia, 1999).
In addition to individual differences in anxiety, an increased ERN-like component to negative feedback has also been found in subjects diagnosed with clinical depression (Tucker, Luu, Frishkoff, Quiring, & Poulsen, 2003). In considering the substantial symptom overlap and diagnostic comorbidity between anxiety and depression, Clark and Watson (1991) proposed that both disorders are characterized by high levels of affective distress, or negative affect (NA). In terms of the differentiation between anxiety and depression, the tripartite model suggests that only depression is characterized by low levels of positive affect (PA), whereas anxiety is uniquely characterized by physiological hyperarousal (Clark & Watson, 1991; for empirical support, see Brown, Chorpita, & Barlow, 1998). Within this framework, it is possible that an enhanced ERN is not a function of either anxiety or depression specifically, but relate to the underlying high NA characteristic of both syndromes.
Some support for this possibility was reported by Luu, Collins, and Tucker (2000), who found significant correlations between NA and ERN amplitude in college students. That is, ERNs were enhanced in college students who were high on self-reported NA. High-NA students in the Luu et al. study, however, had larger ERNs only in the first testing quartile; in fact, the relationship between NA and ERN was in the opposite direction for the remaining three quartiles. Luu et al. interpret this result in terms of task disengagement over time in the high-NA students despite their initial higher than normal concerns with performance. Because the task-disengagement explanation was based solely on an increased reduction in post-error slowing across time in the high-NA group and was not accompanied by any change in performance (e.g., errors or RT), the relationship between ERN and negative affect must be regarded as theoretically plausible, but not well substantiated at this point.
Far less studied than the ERN, the error positivity (Pe) is an ERP component that also appears to be related to response monitoring processes. The Pe has a slightly more posterior scalp distribution, and follows the ERN—peaking approximately 200–400 ms after subjects make a mistake (Falkenstein et al., 2000; Hohnsbein, Falkenstein, & Hoormann, 1989; Nieuwenhuis, Ridderinkhof, Blom, Band, & Kok, 2001). Falkenstein et al. (2000) suggested that the Pe may index processing that occurs after error detection, such as error recognition or error salience. This possibility was supported by a Nieuwenhuis et al. (2001) study, which reported a relationship between Pe and the subjective awareness of making a mistake. Nieuwenhuis et al. found that when subjects were unaware of their mistakes the ERN was unaffected, but the Pe was substantially reduced. In the several studies that have related ERN to differences in affective variables such as anxiety, there have been no formal evaluations of the anxiety variable and its impact on the Pe. Visual inspection of the relevant data suggests either no differences in Pe (Hajcak et al., 2003b; Luu et al., 2000) or smaller Pe in high-anxious subjects (Gehring et al., 2000; Hajcak & Simons, 2002).
In addition to the ERN-Pe ERP complex, a number of recent studies have reported that autonomic nervous system (ANS) activity is also sensitive to response monitoring. For instance, Somsen, Van der Molen, Jennings, and van Beek (2000) found that cardiac deceleration was related to negative feedback in a Wisconsin card sorting task; similar data relating HR deceleration to negative feedback during response monitoring has also been reported by Crone et al. (2003). Finally, data from our own lab indicates that both HR and SCR are also sensitive to endogenous error detection (Hajcak, McDonald, & Simons, 2003a). Like the Pe, these ANS responses to errors have not been evaluated with respect to individual differences in anxiety or NA.
The present study was conducted to more comprehensively evaluate the relationship between self-reported affective experiences and error-related psychophysiology in a between-groups design. The primary aim was to compare the ERN, Pe and ANS responses from subjects that were either high or low in self-reported NA. A secondary aim was to compare the same variables in subgroups of high-NA subjects with different levels of positive affect (PA). All subjects were selected on the basis of the NA and PA subscales of the Positive and Negative Affect Scale (PANAS; Watson, Clark, & Tellegen, 1988). To this end, we measured ERPs and ANS activity while subjects performed a speeded reaction time task in which they had to respond to Stroop stimuli. In particular, subjects saw color words (e.g., “red”) presented in either a congruent (red) or incongruent (blue) color; subjects were instructed to respond based on whether or not the color and name of the stimuli matched as quickly and accurately as possible.
Section snippets
Subjects
Undergraduate students in an introductory psychology class completed the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988). The PANAS is a 20-item self-report measure that measures two mood dimensions: positive affect (PA; 10 items) and negative affect (NA; 10 items). All items are rated on a Likert-type scale ranging from 1 (very slightly or not at all) to 5 (extremely). Although the PANAS can be administered in terms of various time frames, subjects responded to PANAS items
Personality measures
A number of analyses were conducted to confirm that subject assignment resulted in distinct groups and subgroups. In terms of the PANAS scores, there was a highly significant difference in NA between the forty high-NA and 20 low-NA subjects (t(58) = 17.87, p < .001). The two groups also differed on PA, with the high-NA group scoring lower on PA than the low-NA group (t(58) = 7.29, p < .001). These two groups also differed from each other on both the depression (t(58) = 4.71, p < .001) and anxiety (t(58) =
Discussion
We found that college students scoring high on self-reported negative affect were characterized by enhanced ERN and CRN, larger error-related SCR, and reduced post-ERN/CRN positive activity. These results are consistent with Luu et al. (2000), who report evidence for a relationship between ERN and NA in college students. Also consistent with Luu et al., we found that within the context of high NA, PA did not have a moderating role on response monitoring psychophysiology. Because only NA is
References (44)
The central autonomic network: Functional organization, dysfunction, and perspective
Mayo Clinic Proceedings
(1993)- et al.
Cognitive and emotional influences in anterior cingulate cortex
Trends in Cognitive Sciences
(2000) - et al.
Effects of cross-modal divided attention on late ERP components: II. Error processing in choice reaction tasks
Electroencephalography and Clinical Neurophysiology
(1991) - et al.
ERP components on reaction errors and their functional significance: A tutorial
Biological Psychology
(2000) - et al.
A comparison of two measures of perfectionism
Personality and Individual Differences
(1993) - et al.
Perfectionism in obsessive-compulsive disorder patients
Behavior Research & Therapy
(1997) - et al.
A new method for off-line removal of ocular artifact
Electroencephalography and Clinical Neurophysiology
(1983) - et al.
Error-related brain activity in obsessive-compulsive undergraduates
Psychiatry Research
(2002) - et al.
Anxiety and error-related brain activity
Biological Psychology
(2003) - et al.
Error-related scalp potentials elicited by hand and foot movements: Evidence for an output-independent error-processing system in humans
Neuroscience Letters
(1998)
Discrepant target detection and action monitoring in obsessive-compulsive disorder
Psychiatry Research: Neuroimaging
Regional brain activity during transient self-induced anxiety and anger in healthy adults
Biological Psychiatry
The structure of negative emotional states: Comparison of the depression anxiety stress scales (DASS) with the beck depression and anxiety inventories
Behavior Research and Therapy
Errors are foreshadowed in brain potentials associated with action monitoring in cingulate cortex in humans
Neuroscience Letters
Wisconsin card sorting in adolescents: Analysis of performance, response times and heart rate
Acta Psychologica
A model of neurovisceral integration in emotion regulation and dysregulation
Journal of Affective Disorders
Motoric response inhibition in finger movement and saccadic eye movement: A comparative study
Clinical Neurophysiology
Is the ‘error negativity’ specific to errors?
Biological Psychology
Error negativity on correct trials: A reexamination of available data
Biological Psychology
The central autonomic network
“Where did I go wrong?” A psychophysiological analysis of error detection
Journal of Experimental Psychology: Human Perception & Performance
Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal
Journal of Abnormal Psychology
Cited by (0)
- ☆
Portions of this paper were presented both at the 43rd annual meeting of the Society for Psychophysiological Research, Chicago, IL, October, 2003 and at Errors, Conflict, and the Brain: Current Opinions on Performance Monitoring, Dortmund, July, 2003.