Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Oct 28.
Published before final editing as: Depress Anxiety. 2021 Apr 28:10.1002/da.23160. doi: 10.1002/da.23160

Comorbid Anxiety and Depression: Opposing Effects on the Electrocortical Processing of Negative Imagery in a Focal Fear Sample

Elizabeth A Bauer 1, Annmarie MacNamara 1
PMCID: PMC8640943  NIHMSID: NIHMS1760401  PMID: 33909324

Abstract

Background:

Anxiety and depression are highly comorbid and share clinical characteristics, such as high levels of negative emotion. Attention toward negative stimuli in anxiety and depression has been studied primarily using negative pictures. Yet, negative mental imagery – i.e., mental representations of imagined negative events or stimuli – might more closely mirror patient experience.

Methods:

The current study presents the first examination of neural response to negative imagery in fifty-seven adults (39 female) who all shared a common “focal fear” diagnosis (i.e., specific phobia or performance-only social anxiety disorder), but varied in levels of comorbid anxiety and depression. After listening to standardized descriptions of negative and neutral scenes, participants imagined these scenes as vividly as possible. Associations between categorical and continuous measures of depression, generalized anxiety disorder (GAD), social anxiety disorder with electrocortical and subjective responses to negative imagery were assessed.

Results:

Individuals who were more depressed showed reduced electrocortical processing of negative imagery, whereas those with GAD showed increased electrocortical processing of negative imagery – but only when controlling for depression. Furthermore, participants with higher levels of depression rated negative imagery as less negative and those with greater social anxiety symptoms rated negative imagery more negatively.

Conclusions:

Depression and GAD are characterized by opposing electrocortical response to negative imagery; moreover, depression may suppress GAD-related increases in the electrocortical processing of negative imagery. Results highlight distinctions between different dimensions of distress-based psychopathology, and reveal the unique and complex contribution of comorbid depression to affective response in anxiety.

Keywords: Late positive potential (LPP), negative imagery, generalized anxiety disorder (GAD), depression (MDD), social anxiety disorder (SAD), transdiagnostic, event-related potential (ERP)

Introduction

Individuals with anxiety and depression experience intrusive negative imagery (e.g., Patel et al., 2007; Wells & Hackmann, 1993). These mental representations of negative events may consume substantial processing resources and could play a causal role in these disorders (Hirsch et al., 2006). Understanding the neural correlates of negative imagery in anxiety and depression could increase insight into disorder mechanisms, yet there have been few such investigations. Moreover, despite high rates of comorbidity (Brown et al., 2001; Fava et al., 2000), no prior work has examined how anxiety and depression might simultaneously influence the processing of negative imagery.

The late positive potential (LPP), an electroencephalographic (EEG) event-related potential beginning approximately 400 ms following stimulus onset, is a reliable and cost-effective method for measuring neural response to emotional stimuli (Moran et al., 2013). The LPP is maximal at centroparietal sites and is larger for emotional versus neutral stimuli (Cuthbert et al., 2000; Hajcak et al., 2010). The LPP is also sensitive to more fine-grained distinctions in stimulus salience. For example, the LPP is larger for pictures of one’s own relatives and name (Grasso & Simons, 2011; Tacikowski & Nowicka, 2010). Therefore, the LPP can measure individual differences in the motivational salience of stimuli (Lang & Bradley, 2010). Along these lines, the LPP has been shown to reflect contributions from a number of brain regions involved in the processing of salient stimuli, including the amygdala and insula, as well as visual, temporal, and frontal cortices (Liu et al., 2012; MacNamara et al., 2018).

Larger LPPs to negative pictures have been found for individuals high in trait anxiety (Mocaiber et al., 2009) and worry (Burkhouse et al., 2015), and for individuals diagnosed with generalized anxiety disorder (GAD; MacNamara et al., 2016; MacNamara & Hajcak, 2010). Social anxiety has also been associated with larger LPPs to both generic and disorder-specific negative stimuli (Kinney et al., 2019; Moser et al., 2008). Similarly, depression might also be associated with increased LPPs to negative stimuli, particularly when stimuli are self-referential (Benau et al., 2019). However, a growing body of work has also found evidence of reduced processing of generic emotional stimuli in depression (for reviews, see Bylsma et al., 2008; Proudfit et al., 2015). For example, individuals diagnosed with major depressive disorder (MDD; Kayser et al., 2000) and individuals with greater self-reported depression (Hill et al., 2019) have been found to exhibit smaller LPPs to negative pictures compared to non-depressed individuals. Therefore, while anxiety might be characterized by increased attention towards negative pictures, depression might be associated with reduced engagement with non-idiographic negative pictorial stimuli as assessed using the LPP.

Despite high rates of comorbidity between anxiety and depression (Brown et al., 2001; Fava et al., 2000), only a small body of work has examined how these disorders simultaneously influence the processing of negative stimuli. In one such study, depression and anxiety not only showed opposing relationships with the LPP, but depression suppressed anxiety-related increases in the LPP to negative pictures: MDD was associated with smaller LPPs to negative pictures, but GAD was only associated with larger LPPs to negative pictures when controlling for MDD (MacNamara et al., 2016). Suppressor effects like this occur when the predictive power of a variable is only revealed when controlling for another predictor. Suppressor effects play an important role in unpacking the construct validity of symptom measures, which can contain opposing dimensions that are mostly hidden in the overall measures (Watson et al., 2013). For instance, anxiety and depression might share a general distress component, but might also be characterized by distinct, opposing dimensions that are only observable when using suppressor analyses (Watson et al., 2013). Importantly, suppressor effects are not to be equated with interactions, where the nature of the association between one predictor and the dependent variable varies across levels of the other predictor.

Though most work has used negative pictures to elicit negative emotion in the lab, imagined negative scenarios might more closely resemble patient experience. In the only study to-date to use standardized negative imagery to elicit the LPP (MacNamara, 2018), undergraduate participants listened to audio descriptions of negative or neutral scenarios, before imagining themselves in the scene as vividly and realistically as possible (while a fixation cross was displayed on the screen). Although no emotional stimulus was displayed, the LPP was larger for negative compared to neutral scenarios during the imagine period. Moreover, individuals higher in intolerance of uncertainty – a transdiagnostic risk factor for internalizing psychopathology, including anxiety and depression (Gentes & Ruscio, 2011) – showed reduced LPPs to negative imagery (MacNamara, 2018).

Although the processing of negative imagery has yet to be examined in clinical samples using event-related potentials, it has been examined in anxiety and depression using fear-potentiated startle (McTeague et al., 2009; McTeague & Lang, 2012). According to Lang and colleagues (2016), internalizing disorders lie along a spectrum characterized on one end by more focal/fear-based disorders (e.g., specific phobia, performance-only social anxiety disorder [SAD]), and on the other end by more diffuse/distress-based psychopathology (e.g., generalized SAD, depression). Blunted startle response to negative imagery has been observed among individuals at the far end of this spectrum, and has been attributed to distress-based psychopathology (McTeague & Lang, 2012). However, distress-based psychopathology, including high levels of diffuse anxiety, is strongly associated with depression (Lang et al., 2014). As such, depression (rather than distress, more broadly) might explain reduced attention to negative imagery across the internalizing spectrum, which would fit with findings observed using the LPP and negative pictures. Moreover, if individuals at the distress end of this spectrum lack a focal fear diagnosis, the absence of focal fear might partially explain reduced responding to negative stimuli.

Here, we sought to tease apart these issues by performing the first assessment of the neural correlates of negative imagery processing in a transdiagnostic, internalizing sample that controlled for focal fear diagnosis. To ensure similar levels of focal fear across the sample while permitting variation in levels of comorbid distress-based psychopathology (i.e., anxiety and depression), all participants were required to meet criteria for a focal fear disorder, but could in addition meet criteria for any number of additional comorbid anxiety/depressive/trauma-related diagnoses. To examine overlapping versus distinct associations between anxiety and depression in the processing of negative imagery, we examined separate and simultaneous associations between categorical and continuous GAD/diffuse anxiety, depression, and generalized SAD. Based on prior work using pictures, we hypothesized that depression would be associated with smaller LPPs to negative imagery (Proudfit et al., 2015), and that generalized SAD (Kinney et al., 2019) and GAD/diffuse anxiety would only be associated with increased LPPs to negative imagery when controlling for comorbid depression (MacNamara et al., 2016). Based on prior work, we expected that GAD/diffuse anxiety, generalized SAD, and depression would be associated with more negative ratings for negative imagery (Lang et al., 2014; McTeague et al., 2011).

Materials and Methods

Participants

Participant clinical characteristics and diagnoses are presented in Table 1. Sixty individuals recruited from the community participated in the study; three participants were excluded for having poor quality EEG data (>50% trials rejected; diagnoses for excluded participants were: specific phobia only [n=2] and fear of public speaking plus SAD [n=1]), leaving 57 individuals (39 female; age M=22.96 years, SD=8.41; 62.9% White, 17.1% Asian, 14.3% Black or African American, 2.9% American Indian or Alaska Native, 2.9% other; 25.7% Hispanic/Latino). Of these, valence ratings were missing for seven participants, leaving a sample of 50 participants available for ratings analyses. All participants were required to meet criteria for specific phobia or performance-only SAD, but comorbidities were permitted (i.e., additional anxiety, depressive or trauma-related diagnoses). Therefore, all participants shared a focal fear diagnosis, but varied in levels of comorbid distress-based psychopathology, predominantly MDD, GAD and/or generalized SAD. Analyses focused on categorical and continuous measures of depression, GAD and SAD. We did not analyze associations with other diagnoses/symptoms (e.g., post-traumatic stress disorder, PTSD), as the prevalence of other diagnoses in our sample was low (i.e., ns<8).

Table 1.

Clinical characteristics and diagnoses

M (SD)
DASS-21 Depression 4.68 (4.48)
DASS-21 Anxiety 2.84 (2.70)
SPIN 24.25 (14.83)
# of current diagnoses 2.56 (1.45)
n (%)
Current Diagnosis
 Focal fear 57 (100)
 SAD 35 (61)
 GAD 19 (33)
 MDD/PDD 13 (23)
 PTSD 7 (12)
 PMDD 5 (9)
 Substance use disorder 2 (4)
 Agoraphobia 2 (4)
 Anorexia nervosa 2 (4)
Open in a new tab

Note. DASS-21, Depression Anxiety and Stress Scales, 21-item version; SPIN, Social Phobia Inventory; SAD, social anxiety disorder; GAD, generalized anxiety disorder; MDD, major depressive disorder; PDD, persistent depressive disorder; PTSD, post-traumatic stress disorder; PMDD, premenstrual dysphoric disorder. Focal fear included specific phobia (n=20), performance-only social anxiety (n=23), or both (n=14).

Exclusionary criteria included a history of major medical/neurological illness, current/historical traumatic brain injury, bipolar disorder, psychotic disorder, mental retardation, or developmental disorders. Diagnoses were made according to the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, DSM-5 (SCID-5; First et al., 2015). The SCID-5 was performed by a licensed clinical psychologist (the senior author) and clinical psychology graduate students trained in its administration, who were supervised by the senior author. Participants were not engaged in any psychiatric treatment (medication/psychotherapy). Study procedures complied with the Helsinki Declaration of 1975 (as revised in 1983) and were approved by the Texas A&M University institutional review board.

Self-Report Measures

Self-reported depression and diffuse anxiety were assessed using the depression and anxiety subscales of the Depression Anxiety Stress Scales, 21-item version (DASS-21; Lovibond & Lovibond, 1995). Participants indicated how much each statement applied to them over the past week on a scale from 0 (did not apply to me) to 3 (applied to me very much, or most of the time). Internal consistency of the depression subscale was excellent (α=.90) and internal consistency of the anxiety subscale was good (α=.73). Scores on the depression subscale ranged from normal to extremely severe (0–18), and scores on the anxiety subscale ranged from normal to severe (0–9).

Social anxiety symptoms over the past week were assessed using the 17-item Social Phobia Inventory (SPIN; Connor et al., 2000), which uses a scale of 0 (not at all) to 4 (extremely). Internal consistency of the SPIN was excellent (α=.93), and scores ranged from normal to very severe (1–61).1

To assess associations between diagnoses and symptoms, we conducted regression analyses with diagnoses (GAD, SAD, depression) as predictors of each self-report measure. There were three multiple regressions: one with DASS-21 depression as the outcome variable, one with DASS-21 anxiety as the outcome variable, and one with SPIN as the outcome variable. All three diagnoses (i.e., depression, GAD, SAD) were entered as predictors in each model. Results showed that each diagnosis was only significantly associated with its respective self-report measure, indicating that the diagnoses were characterized by different types of symptoms. Specifically, when all three diagnoses were entered as predictors of self-reported depression, F(3,53)=14.93, p<.001, R2=.46, only depression diagnosis was associated with greater self-reported depression, β=.67, p<.001 (GAD, β=−.09, p=.422, and SAD, β=.08, p=.460). When all three diagnoses were entered as predictors of self-reported diffuse anxiety, F(3,53)=2.99, p=.039, R2=.15, only GAD was associated with greater self-reported generalized/diffuse anxiety, β=.31, p=.021 (depression, β=.12, p=.340, and SAD, β=.10, p=.442). Finally, when all three diagnoses were entered as predictors of self-reported social anxiety, F(3,53)=6.21, p=.001, R2=.26, only SAD was associated with greater self-reported social anxiety, β=.44, p=.001 (depression, β=.04, p=.739, and GAD, β=.17, p=.161).

Stimuli

Audio descriptions of scenes were drawn from the Affective Norms for English Text (Bradley & Lang, 2007) and stimuli created in-house (MacNamara, 2018). Stimuli are available upon request from the senior author. All scenes were worded to facilitate the listener imagining themselves in the scene. There were 20 negative and 20 neutral scenes; descriptions were read by a female voice.

Participants used the Self-Assessment Manikin valence scale (SAM; Bradley & Lang, 1994) to rate how negatively they felt about each scene. Response options ranged from 1 to 5, with lower numbers indicating more negative valence and higher numbers indicating more neutral valence.

Procedure

Participants were instructed to imagine negative and neutral scenes (MacNamara, 2018). There were 4 blocks of 10 trials (2 negative, 2 neutral blocks). Each trial began with a white fixation cross presented in the center of a black background for 1,000 ms; the white fixation cross was then replaced by a yellow fixation cross that lasted 20,000 ms, during which time participants heard a 10,000 ms audio description of a negative/neutral scene. Participants were instructed that they should begin imagining themselves in the scene described “as vividly and realistically as possible” as soon as possible, and that they should continue to imagine the scene until the yellow fixation cross went off screen. Immediately after fixation offset, participants rated the valence of the scene they had just imagined. Before starting the task, participants were trained based on recommendations for imagery response induction (Lang et al., 1980), in which they were verbally reinforced for reporting physiological responses and for responding to the scene as if it were a real perceptual experience. The task was presented using Presentation software (Neurobehavioral Systems Inc., Albany, CA).

Valence Ratings

Valence ratings were averaged for each stimulus type, yielding one negative and one neutral average per participant.

EEG Recording and Data Reduction

Continuous EEG recordings were collected using an ActiCap and the ActiCHamp amplifier system (Brain Products GmbH, Gilching Germany). Thirty-two electrodes were used based on the 10/20 system. Electrooculogram was recorded from four facial electrodes: two placed approximately 1 cm above and below the right eye, forming a bipolar channel to measure vertical eye movement and blinks, and two placed approximately 1 cm beyond the outer edges of each eye, forming a bipolar channel to measure horizontal eye movements. EEG data were digitized at 24-bit resolution at a sampling rate of 1000 Hz.

EEG data were processed offline using BrainVision Analyzer 2 software (Brain Products GmbH). Data were segmented for each trial beginning 200 ms before audio description onset and continuing for 20,200 ms (10,000 ms beyond audio offset, during which time participants imagined the scene described in the audio clip as vividly as possible). Baseline correction for each trial was performed using the 200 ms before audio onset. The signal from each electrode was re-referenced to the average of the left and right mastoids (TP9/10) and band-pass filtered with high-pass and low-pass filters of 0.01 and 30 Hz, respectively. Eyeblink and ocular corrections used the method developed by Miller and colleagues (1988). Artifact analysis was used to identify a voltage step of more than 50.0 μV between sample points, a voltage difference of 300.0 μV within a trial, and a maximum voltage difference of less than 0.50 μV within 100-ms intervals. Trials were also inspected visually for any remaining artifacts; data from individual channels containing artifacts were rejected on a trial-to-trial basis.

The LPP was scored by averaging amplitudes at CP1, CP2, Cz, and Pz (MacNamara et al., 2013) during the time window after audio offset (10,500–20,000 ms; MacNamara, 2018).

Data Analysis

The LPP and ratings were analyzed using multiple regression to assess suppressor effects involving a) presence/absence of categorical diagnosis of GAD, SAD and depression (MDD and persistent depressive disorder were collapsed to form a general current depression diagnosis)2 and b) dimensional self-reported generalized/diffuse anxiety, social anxiety and depression. Because we wanted to examine variance specific to negative imagery, responses to neutral imagery were controlled for in each regression. First, we performed a series of regressions with each diagnosis/symptom measure and the LPP/ratings to neutral scenes as the only predictors of response to negative imagery; next, we entered all three diagnoses/symptom measures and the LPP/ratings to neutral scenes into regression analyses simultaneously to test for suppressor effects (Watson et al., 2013). Diagnoses were dummy coded for presence/absence of each disorder. Parallel analyses for the 1,000–10,500 ms LPP are reported in the Supporting Information.

Recent work has emphasized the importance of reporting on the psychometric properties of neurobiological measures (Hajcak et al., 2017). Therefore, we report the internal consistency of the LPP, which was assessed using split-half reliability, in which correlations were performed between averages created separately for the first and second block of negative and neutral trials. Internal consistency was calculated for the individual conditions as well as the difference score (i.e., negative minus neutral imagery) and unstandardized residuals (i.e., negative imagery controlling for neutral imagery). The Spearman-Brown formula was used to correct these correlations (Nunnally, 1978). Analyses were performed using SPSS statistical software version 26.0 (IBM, Armonk, NY).

Results

LPP

Across participants, the LPP did not vary by condition, t(56)=0.37, p=.716 (negative imagery M=1.52 μV, SD=6.93; neutral imagery M=1.08 μV, SD=8.33).

Categorical Analyses

Grand-averaged waveforms and headmaps depicting the voltage distribution for neutral and negative imagery 10,500–20,000 ms following audio onset are shown separately for participants with and without depression (Figure 1A) and for participants with and without GAD (Figure 2). When only depression was entered as a predictor of the LPP, F(2,54)=5.28, p=.008, R2=.16, it was associated with smaller LPPs to negative imagery, β=−0.28, p=.027. No significant associations were observed for GAD, β=0.24, p=.068 or SAD, β=0.05, p=.711 when they were entered as individual predictors of the LPP to negative imagery.

Figure 1.

Figure 1.

Open in a new tab

A) Grand-averaged waveforms at the centroparietal pooling where the LPP was scored, shown separately for neutral and negative imagery, for participants without a diagnosis of depression (n=44) and participants diagnosed with depression (n=13); positive is plotted downwards. Headmaps depict the voltage distributions for neutral and negative imagery 10,500 to 20,000 ms after audio clip onset. B) Partial regression plot showing the association between depression symptoms and the LPP to negative imagery, controlling for the LPP to neutral imagery, diffuse anxiety, and social anxiety symptoms.

Figure 2.

Figure 2.

Open in a new tab

Grand-averaged waveforms at the centroparietal pooling where the LPP was scored, shown separately for neutral and negative imagery, for participants without a diagnosis of GAD (n=38) and participants diagnosed with GAD (n=19); positive is plotted downwards. Headmaps depict the voltage distributions for neutral and negative imagery 10,500 to 20,000 ms after audio clip onset. GAD = generalized anxiety disorder.

When depression, GAD and SAD were entered into the model simultaneously, F(4,52)=3.83, p=.008, R2=.23, depression continued to be associated with smaller LPPs to negative imagery, β=−0.30, p=.018, whereas GAD was now associated with larger LPPs to negative imagery, β=0.26, p=.047; SAD was not associated with the LPP to negative imagery, β=−0.02, p=.895. Multicollinearity was not a concern (all Tolerance>.93; all VIF<1.01).

Continuous Analyses

When only self-reported depression was entered as a predictor of the LPP, F(2,54)=5.99, p=.004, R2=.18, it was associated with smaller LPPs to negative imagery, β=−0.32, p=.014. No significant associations were observed for diffuse anxiety, β=−0.10, p=.469 or social anxiety, β=0.04, p=.791 when entered as individual predictors of the LPP to negative imagery.

Figure 1B presents a partial regression plot depicting the association between self-reported depression and the LPP to negative imagery. When self-reported depression, diffuse anxiety, and social anxiety were entered into the model simultaneously, F(4,52)=3.53, p=.013, R2=.21, depression was associated with smaller LPPs to negative imagery, β=−0.35, p=.008; diffuse anxiety, β=−0.15, p=.299 and social anxiety, β=0.19, p=.180 were not associated with the LPP to negative imagery.3 Multicollinearity was not a concern (all Tolerance>.76; all VIF<1.10).

Split-half reliability

Split-half reliability for the LPP was as follows: negative r=.38, neutral r=.56, difference score r=.48, unstandardized residual r=.40. Because the trial count per condition was rather low (i.e., 20 trials per condition), we also calculated what the split-half reliability would have been if we had included 40 instead of 20 trials per condition (using the Spearman-Brown Prophecy formula). Results showed that if we had included 40 trials per condition, reliability would have been as follows: negative r=.55, neutral r=.72, difference score r=.65, unstandardized residual r=.57.

Valence Ratings

Across participants, negative imagined scenarios were rated more negatively than neutral imagined scenarios, t(49)=33.97, p<.001, d=5.96 (negative imagery M=1.88, SD=0.60; neutral imagery M=4.76, SD=0.33).

Categorical Analyses

No significant associations were observed for depression, β=0.26, p=.059; GAD, β=−0.02, p=.905; or SAD, β=0.13, p=.358 when entered as individual predictors of valence ratings of negative imagery.

When depression, GAD, and SAD were entered into the model simultaneously, no significant associations were observed (depression, β=0.28, p=.058; GAD, β=−0.08, p=.558; SAD, β=0.14, p=.322). Multicollinearity was not a concern (all Tolerance>.92; all VIF<1.10).

Continuous Analyses

No significant associations were observed for self-reported depression, β=0.18, p=.197, diffuse anxiety, β=−0.16, p=.251, or social anxiety, β=−0.26, p=.065 when entered as individual predictors of valence ratings for negative imagery.

Figure 3 presents partial regression plots depicting associations between valence ratings for negative imagery and each of self-reported depression (3A) and social anxiety (3B). When self-reported depression, diffuse anxiety, and social anxiety were entered into the model simultaneously, F(4,45)=3.06, p=.026, R2=.21, depression was associated with less negative ratings of negative imagery, β=0.28, p=.050, while greater social anxiety was associated with more negative ratings, β=−0.33, p=.045; diffuse anxiety was not associated with ratings, β=−0.03, p=.826.4 Multicollinearity was not a concern (all Tolerance>.67; all VIF<1.49).

Figure 3.

Figure 3.

Open in a new tab

A) Partial regression plot showing the association between depression symptoms and valence ratings of negative imagery, controlling for valence ratings of neutral imagery, diffuse anxiety, and social anxiety symptoms. B) Partial regression plot showing the association between social anxiety symptoms and valence ratings of negative imagery, controlling for valence ratings of neutral imagery, depression symptoms, and diffuse anxiety. Higher valence ratings indicate more neutral valence.

Discussion

The current study presents the first examination of neural response to negative imagery in a mixed internalizing sample. Though prior work had found larger LPPs for negative versus neutral imagery in an unselected sample (MacNamara, 2018), here, we failed to observe an overall effect of condition, most likely due to the opposing influences of depression and GAD on the LPP. More specifically, depression was associated with smaller LPPs to negative imagery irrespective of GAD; however, a diagnosis of GAD was only associated with larger LPPs to negative imagery when controlling for depression (MacNamara et al., 2016). Therefore, depression suppressed the association between GAD and the LPP. In other words, by controlling for variance shared between GAD and depression (e.g., high negative affectivity), it was possible to observe an association between GAD and the LPP to negative imagery that was occluded at the bivariate level (Watson et al., 2013). This suggests that the diagnosis of GAD might contain multiple dimensions, some of which show opposing associations with neurobiological response. A more refined definition of GAD and/or its delineation into underlying dimensions (e.g., Kotov et al., 2017) might be more faithful to underlying neurobiology and might facilitate a more pathophysiologically-informed approach to clinical care.

Our observation that depression was associated with reduced response to negative imagery aligns with emotional context insensitivity theory (Rottenberg & Hindash, 2015), which posits that the evolved function of depression is to reduce action in the face of chronic adversity or repeated failure. Conversely, GAD is thought to be associated with hypervigilance for threat and worry as a means of “preparing” for the worst possible outcome (Newman & Llera, 2011). Therefore, both depression and GAD might involve cognitive processes aimed at minimizing perturbations in mood, but might employ different (even opposing) mechanisms to achieve this aim.

The current results appear to be in partly contrast with the startle literature, which found that distress-based disorders, broadly defined (i.e., to include GAD), were associated with blunted startle response to negative imagery (McTeague & Lang, 2012). While differences in the functional significance of startle and the LPP (i.e., defensive responding versus motivated attention) might in part explain these results, it is also possible that distress-based psychopathology like GAD may have been confounded with comorbid depression in prior work (e.g., Lang et al., 2014; McTeague & Lang, 2012), which could explain why both GAD and other distress-based psychopathologies were associated with blunted startle to negative imagery.

When assessing continuous symptoms measures rather than diagnosis, we found that depression was also associated with reduced LPPs to negative (versus neutral) imagery, but we failed to observe an association between generalized/diffuse anxiety and the LPP, even though we had observed an association with a diagnosis of GAD. Therefore, only relatively severe anxiety (i.e., warranting a diagnosis of GAD) might be associated with enhanced processing of negative imagery, whereas individuals with greater self-reported depression might be characterized by emotional blunting even at sub-threshold (diagnostic) levels (Rottenberg & Hindash, 2015). Alternatively, a diagnosis of GAD may capture constructs that were not represented in our measure of diffuse anxiety. Taken together, this suggests that more work is needed to parse the aspects of GAD that are (and are not) associated with enhanced processing of negative imagery.

Contrary to our hypotheses, we found no association between diagnosis or symptoms of social anxiety and the LPP to negative imagery. Notably, prior work that had found that social anxiety was associated with larger LPPs to generic negative pictures (Kinney et al., 2019; MacNamara et al., 2019) did not control for focal fear, which was controlled for in the current study (i.e., all participants – both with and without social anxiety – had a focal fear diagnosis). Therefore, our results suggest that generalized SAD may not be associated with increased attention toward standardized negative stimuli when parsing out focal fear.

In line with our results for the LPP, self-reported depression was associated with less negative ratings of negative imagery, but only when controlling for self-reported diffuse anxiety and social anxiety. In addition, social anxiety was associated with more negative ratings of negative imagery, but only when controlling for self-reported diffuse anxiety and depression. These findings also reflect suppressor effects, in which unique dimensions of depression and social anxiety were only evident when controlling for variance that was shared between these symptom scales (Watson et al., 2013). Therefore, current definitions of social anxiety and depression may contain shared variance that occludes opposing associations between underlying dimensions and subjective response. Future work aimed at parsing these dimensions might facilitate a more refined and empirically informed conceptualization of these disorders.

Limitations

The current study had several limitations. First, we did not analyze the effects of additional types of distress-based psychopathology (e.g., PTSD) on the LPP and valence ratings, due to low prevalence rates in our sample. Second, the current results can only be assumed to hold for generic/standardized negative imagery and not idiographic imagery, for which different results might be observed (Benau et al., 2019). Third, the internal consistency of the LPP was rather low, suggesting that more trials may be needed in future work. Additionally, all symptoms were self-reported (versus obtained via interview) and worry (a prominent feature of GAD), was not assessed. Finally, arousal ratings were not collected.

Conclusion

We found evidence of suppressor effects at both the neural and subjective level. Specifically, depression was associated with reduced electrocortical processing of negative mental imagery, whereas GAD was associated with increased electrocortical processing of negative imagery only when controlling for depression. Depression was also associated with reduced subjective response to negative imagery and social anxiety with increased subjective response to negative imagery, but only when controlling for the simultaneous influence of these predictors. Together, the results highlight distinctions between different dimensions of distress-based psychopathology, reveal the unique and complicating contribution of comorbid depression to affective response, and suggest continued work towards refinement of dimensions underlying affective psychopathology is warranted.

Supplementary Material

Supporting Information

Acknowledgements

This work was supported by National Institute of Mental Health grant, K23MH105553 (to AM). Thanks to Blake Barley and Jared R. Ruchensky.

Footnotes

1.

The SPIN and DASS-21 anxiety subscale correlated moderately, r=.43, p=.001, and the SPIN and DASS-21 depression subscale correlated weakly, r=.27, p=.044. The association between the DASS-21 depression and anxiety subscales did not reach significance, r=.12, p=.394.

2.

We did not consider premenstrual dysphoric disorder as part of a general current depression diagnosis, as premenstrual dysphoric disorder has a much shorter duration and may involve different mechanisms than other forms of depression – e.g., gonadal hormones (Cunningham et al., 2009).

3.

To confirm that effects were specific to the LPP elicited by negative (not neutral) imagery, we performed multiple regressions separately for the LPP to negative and neutral imagery. Categorical analyses including all three diagnoses showed that for the LPP to negative imagery, F(3,53)=3.22, p=.030, R2=.15, depression was associated with smaller LPPs, β=−0.31, p=.019; GAD was associated with larger LPPs, β=0.27, p=.044, and SAD was not associated with the LPP, β=−0.02, p=.878. No significant associations were observed for the LPP to neutral imagery, β=−0.02 to 0.05, all ps>.743. Continuous analyses including all three self-report measures showed that for the LPP to negative imagery, F(3,53)=1.76, p=.166, R2=.09, only self-reported depression was associated with smaller LPPs to negative imagery, β=−0.30, p=.032; diffuse anxiety (β=−0.09, p=.562) and social anxiety (β=0.15, p=.321) were not associated with the LPP. No significant associations were observed for the LPP to neutral imagery, β=−0.12 to 0.17, all ps>.266.

4.

Results were identical when continuous predictors were centered (as expected, given the absence of interaction terms).

Disclosures

Dr. MacNamara is a consultant for Aptinyx Inc. Ms. Bauer reported no biomedical financial interests or potential conflicts of interest.

Data Availability Statement

The data that support the findings of this study are openly available at https://osf.io/ua68j/.

References

  1. Benau EM, Hill KE, Atchley RA, O’Hare AJ, Gibson LJ, Hajcak G, Ilardi SS, & Foti D (2019). Increased neural sensitivity to self-relevant stimuli in major depressive disorder. Psychophysiology, 56(7), e13345. 10.1111/psyp.13345 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bradley MM, & Lang PJ (1994). Measuring emotion: The self-assessment manikin and the semantic differential. Journal of Behavior Therapy and Experimental Psychiatry, 25(1), 49–59. 10.1016/0005-7916(94)90063-9 [DOI] [PubMed] [Google Scholar]
  3. Bradley MM, & Lang PJ (2007). Affective norms for English text (ANET): Affective ratings of text and instruction manual (Technical report No. D-1). University of Florida. [Google Scholar]
  4. Brown TA, Campbell LA, Lehman CL, Grisham JR, & Mancill RB (2001). Current and lifetime comorbidity of the DSM-IV anxiety and mood disorders in a large clinical sample. Journal of Abnormal Psychology, 110(4), 585–599. 10.1037/0021-843X.110.4.585 [DOI] [PubMed] [Google Scholar]
  5. Burkhouse KL, Woody ML, Owens M, & Gibb BE (2015). Influence of worry on sustained attention to emotional stimuli: Evidence from the late positive potential. Neuroscience Letters, 588, 57–61. 10.1016/j.neulet.2014.11.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bylsma LM, Morris BH, & Rottenberg J (2008). A meta-analysis of emotional reactivity in major depressive disorder. Clinical Psychology Review, 28(4), 676–691. 10.1016/j.cpr.2007.10.001 [DOI] [PubMed] [Google Scholar]
  7. Connor KM, Davidson JRT, Churchill LE, Sherwood A, Weisler RH, & Foa E (2000). Psychometric properties of the Social Phobia Inventory (SPIN): New self-rating scale. British Journal of Psychiatry, 176(4), 379–386. 10.1192/bjp.176.4.379 [DOI] [PubMed] [Google Scholar]
  8. Cunningham J, Yonkers KA, O’Brien S, & Eriksson E (2009). Update on Research and Treatment of Premenstrual Dysphoric Disorder. Harvard Review of Psychiatry, 17(2), 120–137. 10.1080/10673220902891836 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cuthbert BN, Schupp HT, Bradley MM, Birbaumer N, & Lang PJ (2000). Brain potentials in affective picture processing: Covariation with autonomic arousal and affective report. Biological Psychology, 52(2), 95–111. 10.1016/S0301-0511(99)00044-7 [DOI] [PubMed] [Google Scholar]
  10. Fava M, Rankin MA, Wright EC, Alpert JE, Nierenberg AA, Pava J, & Rosenbaum JF (2000). Anxiety disorders in major depression. Comprehensive Psychiatry, 41(2), 97–102. 10.1016/s0010-440x(00)90140-8 [DOI] [PubMed] [Google Scholar]
  11. First MB, Williams JBW, Karg RS, & Spitzer RL (2015). Structured Clinical Interview for DSM 5-Research Version (SCID-5 for DSM-5, Research Version; SCID-5-RV). American Psychiatric Association. [Google Scholar]
  12. Gentes EL, & Ruscio AM (2011). A meta-analysis of the relation of intolerance of uncertainty to symptoms of generalized anxiety disorder, major depressive disorder, and obsessive–compulsive disorder. Clinical Psychology Review, 31(6), 923–933. 10.1016/j.cpr.2011.05.001 [DOI] [PubMed] [Google Scholar]
  13. Grasso DJ, & Simons RF (2011). Perceived parental support predicts enhanced late positive event-related brain potentials to parent faces. Biological Psychology, 86(1), 26–30. 10.1016/j.biopsycho.2010.10.002 [DOI] [PubMed] [Google Scholar]
  14. Hajcak G, MacNamara A, & Olvet DM (2010). Event-Related Potentials, Emotion, and Emotion Regulation: An Integrative Review. Developmental Neuropsychology, 35(2), 129–155. 10.1080/87565640903526504 [DOI] [PubMed] [Google Scholar]
  15. Hajcak G, Meyer A, & Kotov R (2017). Psychometrics and the neuroscience of individual differences: Internal consistency limits between-subjects effects. Journal of Abnormal Psychology, 126(6), 823–834. 10.1037/abn0000274 [DOI] [PubMed] [Google Scholar]
  16. Hill KE, South SC, Egan RP, & Foti D (2019). Abnormal emotional reactivity in depression: Contrasting theoretical models using neurophysiological data. Biological Psychology, 141, 35–43. 10.1016/j.biopsycho.2018.12.011 [DOI] [PubMed] [Google Scholar]
  17. Hirsch CR, Mathews A, Clark DM, Williams R, & Morrison JA (2006). The causal role of negative imagery in social anxiety: A test in confident public speakers. Journal of Behavior Therapy and Experimental Psychiatry, 37(2), 159–170. 10.1016/j.jbtep.2005.03.003 [DOI] [PubMed] [Google Scholar]
  18. Kayser J, Bruder GE, Tenke CE, Stewart JE, & Quitkin FM (2000). Event-related potentials (ERPs) to hemifield presentations of emotional stimuli: Differences between depressed patients and healthy adults in P3 amplitude and asymmetry. International Journal of Psychophysiology, 36(3), 211–236. 10.1016/S0167-8760(00)00078-7 [DOI] [PubMed] [Google Scholar]
  19. Kinney KL, Burkhouse KL, & Klumpp H (2019). Self-report and neurophysiological indicators of emotion processing and regulation in social anxiety disorder. Biological Psychology, 142, 126–131. 10.1016/j.biopsycho.2019.01.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kotov R, Krueger RF, Watson D, Achenbach TM, Althoff RR, Bagby RM, Brown TA, Carpenter WT, Caspi A, Clark LA, Eaton NR, Forbes MK, Forbush KT, Goldberg D, Hasin D, Hyman SE, Ivanova MY, Lynam DR, Markon K, … Zimmerman M (2017). The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. Journal of Abnormal Psychology, 126(4), 454–477. 10.1037/abn0000258 [DOI] [PubMed] [Google Scholar]
  21. Lang PJ, & Bradley MM (2010). Emotion and the motivational brain. Biological Psychology, 84(3), 437–450. 10.1016/j.biopsycho.2009.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lang PJ, Kozak MJ, Miller GA, Levin DN, & McLean A (1980). Emotional Imagery: Conceptual Structure and Pattern of Somato-Visceral Response. Psychophysiology, 17(2), 179–192. 10.1111/j.1469-8986.1980.tb00133.x [DOI] [PubMed] [Google Scholar]
  23. Lang PJ, McTeague LM, & Bradley MM (2014). Pathological anxiety and function/dysfunction in the brain’s fear/defense circuitry. Restorative Neurology and Neuroscience, 32(1), 63–77. 10.3233/RNN-139012 [DOI] [PubMed] [Google Scholar]
  24. Lang PJ, McTeague LM, & Bradley MM (2016). RDoC, DSM, and the reflex physiology of fear: A biodimensional analysis of the anxiety disorders spectrum: RDoC and fear reactivity. Psychophysiology, 53(3), 336–347. 10.1111/psyp.12462 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu Y, Huang H, McGinnis-Deweese M, Keil A, & Ding M (2012). Neural Substrate of the Late Positive Potential in Emotional Processing. The Journal of Neuroscience, 32(42), 14563–14572. 10.1523/JNEUROSCI.3109-12.2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lovibond PF, & Lovibond SH (1995). The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behaviour Research and Therapy, 33(3), 335–343. 10.1016/0005-7967(94)00075-U [DOI] [PubMed] [Google Scholar]
  27. MacNamara A (2018). In the mind’s eye: The late positive potential to negative and neutral mental imagery and intolerance of uncertainty. Psychophysiology, 55(5), e13024. 10.1111/psyp.13024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. MacNamara A, & Hajcak G (2010). Distinct electrocortical and behavioral evidence for increased attention to threat in generalized anxiety disorder. Depression and Anxiety, 27(3), 234–243. 10.1002/da.20679 [DOI] [PubMed] [Google Scholar]
  29. MacNamara A, Jackson TB, Fitzgerald JM, Hajcak G, & Phan KL (2019). Working Memory Load and Negative Picture Processing: Neural and Behavioral Associations With Panic, Social Anxiety, and Positive Affect. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 4(2), 151–159. 10.1016/j.bpsc.2018.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. MacNamara A, Kotov R, & Hajcak G (2016). Diagnostic and Symptom-Based Predictors of Emotional Processing in Generalized Anxiety Disorder and Major Depressive Disorder: An Event-Related Potential Study. Cognitive Therapy and Research, 40(3), 275–289. 10.1007/s10608-015-9717-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. MacNamara A, Post D, Kennedy AE, Rabinak CA, & Phan KL (2013). Electrocortical processing of social signals of threat in combat-related post-traumatic stress disorder. Biological Psychology, 94(2), 441–449. 10.1016/j.biopsycho.2013.08.009 [DOI] [PubMed] [Google Scholar]
  32. MacNamara A, Rabinak CA, Kennedy AE, & Phan KL (2018). Convergence of fMRI and ERP measures of emotional face processing in combat-exposed U.S. military veterans. Psychophysiology, 55(2). 10.1111/psyp.12988 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McTeague LM, & Lang PJ (2012). The Anxiety Spectrum and the Reflex Physiology of Defense: From Circumscribed Fear to Broad Distress. Depression and Anxiety, 29(4), 264–281. 10.1002/da.21891 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McTeague LM, Lang PJ, Laplante M-C, & Bradley MM (2011). Aversive imagery in panic disorder: Agoraphobia severity, comorbidity, and defensive physiology. Biological Psychiatry, 70(5), 415–424. 10.1016/j.biopsych.2011.03.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. McTeague LM, Lang PJ, Laplante M-C, Cuthbert BN, Strauss CC, & Bradley MM (2009). Fearful imagery in social phobia: Generalization, comorbidity, and physiological reactivity. Biological Psychiatry, 65(5), 374–382. 10.1016/j.biopsych.2008.09.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Miller GA, Gration G, & Yee CM (1988). Generalized Implementation of an Eye Movement Correction Procedure. Psychophysiology, 25(2), 241–243. 10.1111/j.1469-8986.1988.tb00999.x [DOI] [Google Scholar]
  37. Mocaiber I, Pereira MG, Erthal FS, Figueira I, Machado-Pinheiro W, Cagy M, Volchan E, & de Oliveira L. (2009). Regulation of negative emotions in high trait anxious individuals: An ERP study. Psychology & Neuroscience, 2(2), 211–217. 10.3922/j.psns.2009.2.014 [DOI] [Google Scholar]
  38. Moran TP, Jendrusina AA, & Moser JS (2013). The psychometric properties of the late positive potential during emotion processing and regulation. Brain Research, 1516, 66–75. 10.1016/j.brainres.2013.04.018 [DOI] [PubMed] [Google Scholar]
  39. Moser JS, Huppert JD, Duval E, & Simons RF (2008). Face processing biases in social anxiety: An electrophysiological study. Biological Psychology, 78(1), 93–103. 10.1016/j.biopsycho.2008.01.005 [DOI] [PubMed] [Google Scholar]
  40. Newman MG, & Llera SJ (2011). A novel theory of experiential avoidance in generalized anxiety disorder: A review and synthesis of research supporting a contrast avoidance model of worry. Clinical Psychology Review, 31(3), 371–382. 10.1016/j.cpr.2011.01.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Nunnally JC (1978). Psychometric theory. McGraw-Hill. [Google Scholar]
  42. Patel T, Brewin CR, Wheatley J, Wells A, Fisher P, & Myers S (2007). Intrusive images and memories in major depression. Behaviour Research and Therapy, 45(11), 2573–2580. 10.1016/j.brat.2007.06.004 [DOI] [PubMed] [Google Scholar]
  43. Proudfit GH, Bress JN, Foti D, Kujawa A, & Klein DN (2015). Depression and Event-related Potentials: Emotional disengagement and reward insensitivity. Current Opinion in Psychology, 4, 110–113. 10.1016/j.copsyc.2014.12.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rottenberg J, & Hindash AC (2015). Emerging evidence for emotion context insensitivity in depression. Current Opinion in Psychology, 4, 1–5. 10.1016/j.copsyc.2014.12.025 [DOI] [Google Scholar]
  45. Tacikowski P, & Nowicka A (2010). Allocation of attention to self-name and self-face: An ERP study. Biological Psychology, 84(2), 318–324. 10.1016/j.biopsycho.2010.03.009 [DOI] [PubMed] [Google Scholar]
  46. Watson D, Clark LA, Chmielewski M, & Kotov R (2013). The Value of Suppressor Effects in Explicating the Construct Validity of Symptom Measures. Psychological Assessment, 25(3), 929–941. 10.1037/a0032781 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wells A, & Hackmann A (1993). Imagery and Core Beliefs in Health Anxiety: Content and Origins. Behavioural and Cognitive Psychotherapy, 21(3), 265–273. 10.1017/S1352465800010511 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
NIHMS1760401-supplement-Supporting_Information.pdf (84.9KB, pdf)

Data Availability Statement

The data that support the findings of this study are openly available at https://osf.io/ua68j/.

RESOURCES