Psychiatry Research 230 (2015) 262–270

Contents lists available at ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Transdiagnostic psychiatric symptoms related to visual evoked potential abnormalities Jeffrey S. Bedwell a,n, Pamela D. Butler b, Chi C. Chan a, Benjamin J. Trachik a a b

Department of Psychology, University of Central Florida, Orlando, FL, USA Nathan S. Kline Institute for Psychiatric Research, Orangeburg, NY, USA

art ic l e i nf o

a b s t r a c t

Article history: Received 30 January 2015 Received in revised form 28 July 2015 Accepted 2 September 2015 Available online 8 September 2015

Visual processing abnormalities have been reported across a range of psychotic and mood disorders, but are typically examined within a particular disorder. The current study used a novel transdiagnostic approach to examine diagnostic classes, clinician-rated current symptoms, and self-reported personality traits in relation to visual processing abnormalities. We examined transient visual-evoked potentials (VEPs) from 48 adults (56% female), representing a wide range of psychotic and mood disorders, as well as individuals with no history of psychiatric disorder. Stimuli were low contrast check arrays presented on green and red backgrounds. Pairwise comparisons between individuals with schizophrenia-spectrum disorders (SSD), chronic mood disorders (CMD), and nonpsychiatric controls (NC) revealed no overall differences for either P1 or N1 amplitude. However, there was a significant interaction with the color background in which the NC group showed a significant increase in P1 amplitude to the red, vs. green, background, while the SSD group showed no change. This was related to an increase in social anhedonia and general negative symptoms. Stepwise regressions across the entire sample revealed that individuals with greater apathy and/or eccentric behavior had a reduced P1 amplitude. These relationships provide clues for uncovering the underlying causal pathology for these transdiagnostic symptoms. & 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: Schizophrenia Mood disorders Red light Disorganized Apathy ERP Social anhedonia

1. Introduction There is considerable evidence that at least a subset of individuals with schizophrenia have abnormalities in visual processing (Butler et al., 2008; Yoon et al., 2013). As many of these abnormalities have been replicated in samples of first-degree relatives of individuals with schizophrenia (Bedwell et al., 2003; Green et al., 2006; Bakanidze et al., 2013; Sponheim et al., 2013) and nonpsychiatric schizotypy (Aichert et al., 2012; Bedwell et al., 2013), they may reflect underlying vulnerability for schizophrenia and serve as useful biomarkers. However, visual processing deficits have also been reported in other psychiatric disorders, including autism (Marco et al., 2011), bipolar disorder (Yeap et al., 2009), and unipolar depression (Normann et al., 2007), and therefore may not be specific to the diagnostic category of schizophrenia. Existing studies on these relationships have primarily examined visual processing deficits within a single diagnostic category or construct, as compared to nonpsychiatric controls. This strategy limits our understanding of how such endophenotypes may reflect vulnerability factors that do not respect the divisions in current n

Corresponding author. E-mail address: [email protected] (J.S. Bedwell).

http://dx.doi.org/10.1016/j.psychres.2015.09.004 0165-1781/& 2015 Elsevier Ireland Ltd. All rights reserved.

psychiatric classification system. To address these limitations, the current study uses a broad transdiagnostic approach, consistent with the Research Domain Criteria (RDoC) initiative from the National Institute of Mental Health (Cuthbert and Insel, 2013), to examine how past and present psychiatric symptoms relate to visual processing. Visual processing is a subconstruct in the RDoC matrix (under Perception: Cognitive Systems). Over the past decade, much of the research in this area has assessed visual-evoked potentials (VEPs), obtained from electroencephalography. The current study will focus on the P1 and N1 VEP components. P1 represents a positive voltage peak approximately 90–160 ms following stimulus onset, which is thought to originate from the dorsal and ventral extrastriate cortex (Di Russo et al., 2002). N1 is a negative voltage deflection that often shows multiple peaks in the range of 140–220 ms following stimulus onset and is thought to reflect primarily a ventral stream generator (Di Russo et al., 2002). Within psychiatric research, the majority of transient VEP studies have assessed schizophrenia-spectrum conditions. The most consistent and replicated finding is that the P1 component shows a reduced amplitude in individuals with schizophrenia (Schechter et al., 2005; Yeap et al., 2008b; Butler et al., 2013; Verleger et al., 2013) or psychometrically-defined schizotypy (Koychev et al., 2010; Bedwell et al., 2013) as compared to nonpsychiatric controls. While the majority of studies

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

that have examined N1 in schizophrenia/schizotypy have found no significant difference from nonpsychiatric controls (Foxe et al., 2005; Butler et al., 2013; Verleger et al., 2013), some have reported that the N1 amplitude was also reduced in schizophrenia (Butler and Javitt, 2005; Martinez et al., 2012). Transient VEPs to basic visual stimuli have also been examined in other diagnostic groups. Two studies reported a reduced P1 amplitude in bipolar I disorder (Yeap et al., 2009; Verleger et al., 2013), and that bipolar I disorder did not differ from schizophrenia on P1 amplitude (Verleger et al., 2013). However, a study on bipolar II disorder found no difference from controls on P1 amplitude (Elvsashagen et al., 2012). Two studies on unipolar depression found a reduction in both P1 and N1 amplitude (Normann et al., 2007), which may be particularly true for the subtype of “anxiousagitated” depression (Pierson et al., 1996). Diffuse red light is known to suppress the magnocellular/dorsal pathway (de Monasterio, 1978; Livingstone and Hubel, 1984). A recent study found that while control participants showed the expected reduction in P1 amplitude in response to a high-contrast stimulus on a red background, individuals with high-schizotypy showed no change (Bedwell et al., 2013). Consistent with this finding, a different pattern of behavioral accuracy change to a red background on visual backward masking tasks has been previously reported in individuals with schizophrenia (Bedwell et al., 2011b) and their first degree relatives (Bedwell et al., 2003). Given the findings of VEP abnormalities across multiple psychiatric disorders, it is possible that a shared phenomenological characteristic and/or underlying vulnerability factor accounts for particular VEP abnormalities. Therefore, the current study uses a novel approach of recruiting a wide range of psychiatric disorders and focusing on particular symptoms that relate to the VEP abnormalities. Most existing studies have either not reported symptom relationships, or only reported VEP relationships with broad categories of current symptoms such as current total positive or negative symptom ratings in schizophrenia, which have generally shown no relationship to P1 amplitude (Schechter et al., 2005; Obayashi et al., 2009). Based on previous findings within samples of a single diagnosis, the current study examined the diagnostic classes of schizophrenia-spectrum disorders and chronic mood disorders in order to assess the relationship of these broad diagnostic relationships with VEP amplitudes. We complemented this methodology with clinician ratings of particular current symptoms, and self-reported psychometric trait measures of specific schizotypy features and subtypes of anhedonia. We expected that this novel combination of lifetime vs. current symptoms and self-reported vs. clinician-rated symptoms would optimize our ability to examine symptom correlates of VEP abnormalities across diagnoses. We also included red and comparison (green) backgrounds to examine the previously reported red light effect for the first time in a transdiagnostic sample. As this approach has not been previously published in the VEP literature to our knowledge, analyses of transdiagnostic symptom relationships with VEP amplitudes was exploratory. In terms of diagnostic class differences, we hypothesized that, across stimuli conditions, the schizophrenia-spectrum group and chronic mood disorder groups would have a smaller P1, but not N1, mean amplitude when compared to nonpsychiatric controls, but would not differ from each other on either P1 or N1 amplitude (e.g., Verleger et al., 2013). We additionally hypothesized that the schizophreniaspectrum group in particular would show a group by color interaction on the P1 amplitude when compared to the nonpsychiatric controls, such that controls would show a reduction in P1 amplitude to the red background, while the schizophrenia-spectrum group would show no change.

263

2. Methods 2.1. Participants Participants were recruited with attempts to include a wide range of psychiatric disorders, ranging from psychotic disorders, nonpsychotic mood and anxiety disorders, to no diagnosis. Attempts were made to recruit a sufficient number of individuals with schizophrenia-spectrum disorders and chronic mood disorders (e.g., bipolar disorder), given the existing findings of P1 amplitude abnormalities. This broad diagnostic inclusion was chosen to be consistent with the NIMH RDoC aim of examining dimensional symptom relationships that cut across multiple disorders, and to help ensure a wide range of severity across the symptoms measured. Participants were recruited from the local community using advertisements such as flyers in psychiatric clinics, newspaper advertisements, and online advertisements (e.g., Craigslist). Some of these advertisements mentioned that we are looking for individuals with “schizophrenia, schizoaffective disorder, or bipolar disorder” in order to recruit a sufficient number of these individuals. All participants completed informed consent prior to data collection and were paid a cash stipend for time spent participating in this study. Exclusionary criteria included self-reported history of significant neurological injury or symptoms (e.g., seizures), substance abuse in the past month or dependence in the past 6 months, medical conditions that may affect brain functioning, or English not being the native language. In addition, we administered the Reading subtest from the Wide Range Achievement Test 3rd edition (WRAT-3; Wilkinson, 1993) and excluded those with estimated IQ o70, as well as those who showed indication of color blindness using the Ischihara Color Blindness Test, or an estimated corrected visual acuity worse than 20/40 using a Snellen wall chart. Following exclusions, including exclusions for poor accuracy on the task of interest (described below), the final sample used in the data analyses consisted of 48 individuals (see Table 1 for demographics and descriptive statistics). Based on the Structured Clinical Interview for DSM-IV Axis-I Disorders (SCID-I) and the Avoidant, Paranoid, and Schizotypal sections of the Structured Clinical Interview for DSM-IV Axis-II Disorders (SCID-II), the sample consisted of the following primary diagnoses: 14 (29.2%) no diagnosis, 10 (20.8%) bipolar I (7 of which had history of psychosis), 6 (12.5%) schizoaffective disorder, 4 (8.3%) major depressive disorder (MDD), 3 (6.3%) schizophrenia, 2 (4.2%) delusional disorder, 2 (4.2%) social phobia, 1 (2%) posttraumatic stress disorder, 1 (2%) generalized anxiety disorder, 1 (2%) dysthymic disorder, 1 (2%) bipolar II (nonpsychotic), 1 (2%) schizotypal and avoidant personality disorders and MDD, 1 (2%) paranoid personality disorder and MDD, 1 (2%) avoidant personality disorder and MDD. 2.2. Procedures Following informed consent and administration of the structured interviews and visual/cognitive screens, participants then completed the Structured Clinical Interview for Positive and Negative Syndrome Scale (SCI-PANSS; Kay et al., 1987), the Schizotypal Personality Questionnaire-Brief Revised (SPQ-BR; Cohen et al., 2010; Callaway et al., 2014), the Temporal Experience of Pleasure Scale (TEPS; Gard et al., 2007), and the Anticipatory and Consummatory Interpersonal Pleasure Scale (ACIPS; Gooding et al., 2014; Gooding and Pflum, 2014). Participants then completed the VEP task. 2.2.1. VEP “Check” task Stimuli were 8
8 arrays of isolated checks, similar to those used in transient VEP studies by others (Butler et al., 2007; Yeap

264

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

Table 1 Descriptive statistics of demographic, clinical, and cognitive variables. Sample size Age Sex (% female) Race (% non-Hispanic white) Estimated IQn Years of education PANSS positive symptoms PANSS negative symptoms PANSS general symptoms SPQ-BR social anxiety SPQ-BR no close friends/constricted affect SPQ-BR magical thinking SPQ-BR unusual perceptual experiences SPQ-BR ideas of reference/suspiciousness SPQ-BR eccentric behavior SPQ- BR odd speech ACIPS total score TEPS total score

48 35.88 (9.11); 19–55 56% 62.5% 101.27 (10.76); 72–117 14.35 (2.56); 6–20 11.52 (4.86); 7–24 12.25 (5.11); 7–28 25.27 (7.71); 16–52 7.09 (4.74); 0–16 4.13 (2.77); 0–10 5.60 (3.92); 0–15 4.06 (3.82); 0–15 4.98 (3.44); 0–12 7.26 (4.44); 0–14 9.49 (4.46); 0–16 79.19 (14.92); 40–102 78.48 (14.03); 39–100

Unless otherwise noted, descriptive data are in the format of [mean (standard deviation); range]. PANSS ¼ Positive and Negative Symptom Scale; SPQ-BR ¼Schizotypal Personality Questionnaire-Brief Revised; ACIPS¼ Anticipatory and Consummatory Interpersonal Pleasure Scale total score; TEPS ¼ Temporal Experience of Pleasure Scale total score. n IQ was estimated using the standard sore from the Reading subtest of the Wide Range Achievement Test – 3rd edition.

et al., 2008a, 2009). We included two background colors (green and red) that were appoximately matched on physical luminance (15.5 cd/m2) using a spot photometer placed near center of monitor, and two contrast conditions for the checks (isoluminant and 12%; based on Michelson's contrast for the physical luminance). Green was chosen as the baseline color condition as it is the opponent color of red, and therefore represents an optimal contrast. The isoluminant condition was chosen as it minimizes the contribution from the dorsal pathway and results in predominant ventral pathway signaling (Schechter et al., 2005), while the 12% contrast includes relatively more contribution from the dorsal pathway along with the ventral pathway (Butler et al., 2007). The background color surrounded and appeared between the inner 8
8 check stimulus, which subtended 8.41
8.41 degrees of visual angle. The outer edges of the color on the widescreen monitor subtended 18.64
32.78 degrees of visual angle. E-Prime 2.0 Professional was used for stimulus presentation. For each block, 120 check stimuli were presented for 60 ms each, with a random variable interstimulus interval that ranged between 740 and 1540 ms. The 120 check stimuli in each block contained a random presentation of the two contrast conditions (isoluminant and 12%), with 60 check stimuli from each of the two contrast levels. Also consistent with several earlier studies, we included 30 pictures of four types of animals in each block, which were randomly interleaved with the check images. Participants were asked to press a button each time they saw a particular target animal during the task. The accuracy data from the animal targets was examined to help ensure adequate attention. The two color backgrounds alternated in consecutive blocks, with a 30 s rest break between blocks. Each color block was repeated seven times in order to collect 420 VEP responses to each of the two contrasts within each of the two color conditions. The task lasted approximately 48 minutes. EEG was acquired using the Neuroscan Synamps2 32 channel system, using DC recording digitized at 1000 Hz, with a 100 Hz low pass filter, no high-pass filter, and impedance of less than 10 kOhms at each electrode. DC recording was chosen as it allows for later highpass filtering at any cutoff using better quality software filters (Luck, 2014). The averaged bilateral mastoid electrodes were used as the

active reference, and vertical and horizontal ocular electrodes were used to estimate blinks and large eye movements. 2.2.2. VEP analysis VEP data were analyzed using Brain Vision Analyzer 2.0 software, and were filtered offline using a high-pass filter of 0.10 Hz and a low pass filter of 45 Hz, both with a roll off of 48 db/oct. Data were segmented using a 100 ms baseline and a 500 ms post-stimulus interval. A semi-automated procedure was used to identify blinks and large eye movements using the ocular channels, and these segments were removed from further analysis. Artifact rejection was then conducted on individual scalp electrodes for any segment that contained more than a 120 mV max/min difference within 200 ms or less than 0.5 mV within 100 ms. Baseline correction was conducted using the interval of  80 to þ20 ms around stimulus onset. Methods for baseline correction (e.g., Yeap et al., 2006, 2009), artifact rejection (e.g., Yeap et al., 2006, 2009; Butler et al., 2013), and active reference location (e.g., Martinez et al., 2012) were chosen to be consistent with previous studies examining VEP in schizophrenia and bipolar disorder to allow more direct comparisons of findings to those studies. Data were then further segmented and averaged separately for each color and contrast condition. Examination of the voltage topography map from all 48 participants across all conditions for all parietal and occipital electrodes during the VEP time window for P1 (100–150 ms) indicated that electrodes P4 and P8 had the strongest positive deflection. These locations are relatively consistent with previous literature on the P1 component, including right hemisphere dominance for extrastriate visual processing. Examination of the grand average waveforms revealed that the N1 peak latencies from the parietal electrodes were closest to the time frame of N1 reported by others (170–190 ms). As the N1 window voltage topography over the parietal electrodes was roughly equivalent between the hemispheres, we chose P4 and P8 to be consistent with the P1 analysis. Therefore we chose to use a linear derivation of electrodes P4 and P8 to measure both the P1 and N1 components. All participants had at least 250 segments that survived exclusion criteria at these selected electrodes (P4 and P8) within each contrast by color condition. P1 was defined as the largest positive voltage peak within the latency range of 75–175 ms, while N1 was defined as the largest negative voltage peak within the range of 100 to 240 ms. All participants had a visible P1 and N1 in each condition using the P4/P8 electrode average. All peak amplitudes were defined by the mean amplitude from a 10 ms window centered over the local maxima. 2.2.3. Statistical analysis approach For each VEP (P1 and N1), we first used 2 (color) by 2 (contrast) repeated-measure ANOVAs to verify expected effects of our manipulations of color and contrast on each VEP mean amplitude across all participants. We also repeated these analyses in just the subset with no psychiatric disorder (N ¼14) in order to clarify any “healthy” effects that may have been masked by the inclusion of those with psychiatric disorders. As we had a priori hypotheses about VEP differences between the three groups, we examined these hypotheses with pairwise ANCOVAs. These diagnostic classes included: (1) schizophreniaspectrum disorders (N ¼15; mean age¼36.93, SD ¼6.86; 47% male; median task accuracy¼94.00%): 6 with schizoaffective disorder, 3 schizophrenia, 2 delusional disorder, 1 schizotypal and avoidant personality disorders, 2 paranoid personality disorder, and 1 avoidant personality disorder; and (2) chronic mood disorders (N ¼13; mean age¼36.31, SD ¼10.77; 31% male; median task accuracy ¼99.00%): 9 with bipolar I disorder, 2 recurrent major depressive disorder (MDD), 1 dysthymic disorder and

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

recurrent MDD, 1 bipolar II disorder. We did not include individuals with a single past major depressive episode in this “chronic” group. These groups were compared to participants with no history of a psychiatric disorder (N ¼ 14; mean age¼ 33.07, SD ¼9.60; 50% male; median task accuracy ¼95.50%). While there were no statistically significant differences in age, sex, or behavioral accuracy on the task across the three groups, the groups were not carefully matched on these factors during recruitment. Therefore, we included these three variables as covariates in ANCOVAs for all group comparisons. When examining symptom correlates for the VEPs, we chose an exploratory statistical approach to map onto our objective of examining potential particular symptom correlates of the VEPs across a wide range of disorders. These analyses were conducted across the entire sample of 48 participants, which included six participants with disorders that did not fall into the categories of interest used in the group comparisons (e.g., anxiety disorders). We chose stepwise regressions (entry alpha ¼0.05, exit alpha ¼0.10), as this allowed us to enter several potential symptoms to examine which symptom or subset accounted for the

265

largest unique variance in the VEPs, thereby partially controlling for the large number of predictors. A systematic approach was used within each of the two VEP components in which we first examined clinican-rated current symptoms using all of the individual items from the PANSS. A second stepwise regression examined the seven subscales from the SPQ-BR (see Table 1), along with the total TEPS and ACIPS scores, as these are all self-reported psychometric trait measures of interest. There did not appear to be any notable problems with collinearity among the predictors in these regressions (e.g., Condition Index never exceeded 3.58; see Table 2 for intercorrelation matrix of all symptom scales).

3. Results The behavioral accuracy to the attention check was negatively skewed. We excluded five participants with o75% accuracy, due to concern about inconsistent attention to the stimuli of interest. We chose 75% based on inspection of the distribution of accuracy across all participants, as it appeared to be the last value before

Table 2 Zero-order correlations among symptom scales across entire sample (N ¼ 48). Self-report scales 1. SPQ-BR social anxiety 2. SPQ-BR no close friends/constricted affect 3. SPQ-BR magical thinking 4. SPQ-BR unusual perceptual experiences 5. SPQ-BR ideas of reference/suspiciousness 6. SPQ-BR eccentric behavior 7. SPQ-BR odd speech 8. ACIPS total score (interpersonal pleasure) 9. TEPS total score (general Pleasure) Individual PANSS (clinician-rated) items Delusions (P1) Conceptual disorganization (P2) Hallucinatory behavior (P3) Excitement (P4) Grandiosity (P5) Suspiciousness/persecution (P6) Hostility (P7) Blunted affect (N1) Emotional withdrawal (N2) Poor rapport (N3) Passive/apathetic social withdrawal (N4) Difficulty in abstract thinking (N5) Lack spontaneity/flow of conversation (N6) Stereotyped thinking (N7) Somatic concern (G1) Anxiety (G2) Guilt feelings (G3) Tension (G4) Mannerisms & postering (G5) Depression (G6) Motor retardation (G7) Uncooperativeness (G8) Unusual thought content (G9) Disorientation (G10) Poor attention (G11) Lack of judgement and insight (G12) Disturbance of volition (G13) Poor impulse control (G14) Preoccupation (G15) Active social avoidance (G16)

1

2 –

0.36n  0.04 0.39nn 0.26 0.24 0.59nnn 0.13 0.30n 0.13 0.07 0.38nn 0.12 0.15 0.15 0.29 0.35n 0.36n 0.24 0.07 0.25 0.23  0.07 0.17  0.14  0.04 0.07  0.02 0.03 0.10 0.46nn

3 0.63nnn –

0.50nnn 0.23 0.20 0.21 0.41nn 0.59nnn 0.22 0.30n 0.36n 0.12 0.47nnn 0.02 0.12 0.30n 0.23 0.23 0.52nnn 0.19 0.14 0.40nn 0.16 0.14 0.37nn  0.11 0.26 0.29n 0.15 0.30n 0.09 0.52nnn

4 0.38nn 0.30n –

0.3n 0.21 0.48nnn 0.33n 0.34n 0.34n 0.19 0.26 0.18 0.15 0.22 0.08 0.24  0.09 0.18 0.19  0.05 0.06 0.08 0.15 0.28  0.25 0.40nn  0.04 0.08 0.20  0.02 0.11 0.11 0.14

5

6

7

0.57nnn 0.52nnn 0.62nnn –

0.66nnn 0.56nnn 0.54nnn 0.73nnn –

0.34n 0.52nnn 0.53nnn 0.53nnn 0.45nn –

0.58nnn 0.28 0.75nnn 0.54nnn 0.41nn 0.38nn 0.28 0.41nn 0.22 0.19 0.30n 0.05 0.06 0.23  0.07 0.24 0.20 0.34n 0.23 0.01 0.26  0.14 0.46nnn  0.03 0.08 0.06  0.01 0.08 0.01 0.25

0.49nnn 0.26 0.59nnn 0.49 0.41nn 0.62nnn 0.25 0.44nn 0.17 0.24 0.35n 0.16 0.14 0.26 0.32n 0.40nn 0.23 0.26 0.27 0.23 0.27  0.18 0.36n 0.08 0.16 0.23 0.16 0.23 0.002 0.41nn

0.38nn 0.43nn 0.39nn 0.27 0.36n 0.34n 0.19 0.43nn 0.41nn 0.19 0.40nn 0.17 0.44nn 0.24 0.16 0.09 0.06 0.16 0.18 0.30n 0.33n 0.18 0.45nnn  0.02 0.33n 0.22 0.13 0.22 0.06 0.31n

8 0.49nnn 0.43nn 0.43nn 0.44nn 0.40nn 0.65nn

0.20 0.23 0.30n 0.24 0.15 0.32n 0.09 0.24 0.18 0.13 0.28 0.12 0.12 0.07 0.21 0.22 0.20 0.18 0.05 0.32n 0.20 0.17 0.18 0.03 0.13 0.17 0.08 0.13  0.06 0.31n

9

 0.57nnn  0.69nnn  0.22  0.46nn  0.52nnn  0.46nn  0.31n –

 0.31n  0.54nnn 0.02  0.26  0.25  0.37n  0.23 0.80nnn –

 0.46nnn  0.15  0.29n  0.36n  0.25  0.49nnn  0.10  0.44nn  0.45nnn  0.16  0.57nnn 0.004  0.31n  0.30n  0.07  0.18  0.41nn  0.14 o 0.001  0.28  0.21  0.17  0.36n  0.18  0.17  0.04  0.16  0.15  0.005  0.57nnn

 0.36n  0.19  0.12  0.36n  0.18  0.20  0.14  0.29n  0.38nn  0.02  0.40nn 0.02  0.30n  0.30n 0.09  0.06  0.23  0.01  0.03  0.10  0.11  0.30n  0.27 0.21  0.19 0.01  0.15  0.19  0.05  0.38nn

SPQ-BR ¼Schizotypal Personality Questionnaire-Brief Revised; ACIPS ¼Anticipatory and Consummatory Interpersonal Pleasure Scale total score; TEPS ¼Temporal Experience of Pleasure Scale total score; PANSS ¼Positive and Negative Symptom Scale (item label is followed by PANSS item # in parentheses). Note: inter-correlations of the PANSS items are omitted to reduce complexity of the table. n

= P o0.05. Po 0.01 nnn Po 0.001. nn

266

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

Table 3 Peak latencies and mean amplitudes of the visual-evoked potential components by diagnostic classes of interest. Diagnostic class of interest

P1 Lat. Green (ms)

P1 Amp. Green (uV)

P1 Lat. Red (ms)

P1 Amp. Red (uV)

N1 Lat. (ms)

N1 Amp. (uV)

Entire sample (N ¼48)a Schizophrenia spectrum disorders (N ¼ 15) Chronic mood disorders (N ¼13) Nonpsychiatric controls (N ¼ 14)

122.46 129.77 111.65 122.25

6.11 5.34 5.20 6.75

111.07 116.43 106.04 107.93

7.27 5.63 6.54 8.66

169.18 174.27 159.98 168.04

 4.59  4.42  5.03  4.37

(21.54) (24.67) (19.84) (15.70)

(5.39) (3.94) (6.29) (6.20)

(17.42) (17.54) (20.27) (13.41)

(6.03) (3.48) (7.24) (6.84)

(28.10) (30.63) (26.17) (28.04)

(4.35) (3.28) (6.01) (3.19)

Descriptive statistics in format of [mean (standard deviation)]. Both P1 and N1 values are averaged across the two contrast conditions as they did not significantly differ by contrast. N1 values are additionally averaged across the color conditions as they did not significantly differ by color. Lat. ¼Peak Latency; Amp.¼ Mean Amplitude. a

The overall sample included an additional 6 participants who had diagnoses that did not fall into one of the diagnostic categories of interest (e.g., anxiety disorder).

lower values began to decrease at a noticeably greater slope. This value remains well above chance performance (i.e., approximately 50% accuracy), suggesting that all remaining participants exhibited acceptable attention during the task. This resulted in 48 participants used for the final data analysis, who had a median accuracy of 96%; range 75–100%. Descriptive statistics for peak latencies and mean amplitudes are listed in Table 3. VEP waveforms are depicted in Fig. 1. 3.1. P1 component Across all participants (N ¼48), the mean amplitude was larger in the red, compared to green, background condition, F(1,47) ¼ 12.33, p ¼0.001, ή2 ¼0.21 (see Table 3). However, there was no main effect of contrast and no interaction of contrast and color.

When we only examined participants with no psychiatric diagnosis (N ¼ 14), we found the same pattern of results for color and contrast. Therefore, for the purposes of further analyses, the P1 mean amplitude was averaged across contrast conditions within each color condition. When comparing the schizophrenia-spectrum participants to nonpsychiatric controls using repeated-measure (i.e., color) ANCOVAs, there was no main effect of group, F(1,24) ¼0.70, p ¼0.41, η2 ¼0.03, but there was a significant interaction of group by color F (1,24)¼ 5.03, p ¼0.03, η2 ¼0.17. The simple effects of this interaction were explored with repeated-measures ANCOVAs separately within each group, examining the effect of color background with the covariate of behavioral accuracy. Within the control group, the P1 mean amplitude was significantly higher on the red, as compared to green, background, F(1,12) ¼ 4.89, p ¼0.05, η2 ¼ 0.29.

Fig. 1. Grand average event-related potential waveforms depicting P1 and N1 components from the green and red background in individuals with schizophrenia-spectrum disorders as compared to nonpsychiatric controls.

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

However, within the schizophrenia-spectrum group, there was no difference in P1 mean amplitude between the color background conditions, F(1,13) ¼0.001, p ¼0.97, η2 o.001 (see Fig. 1). This appears to be driven by more individuals in the schizophreniaspectrum group showing a decrease, rather than increase, in P1 amplitude to red. This resulted in a nonsignificant group mean decrease in P1 to red, rather than the significant increase seen in controls (see Table 3 and Fig. 1). Given this finding, we then examined the relationship between a difference score for P1 amplitude (red minus green background conditions) and the three SPQ-BR factor scores, the TEPS and ACIPS total scores, and the PANSS positive and negative symptom factors, using partial correlations that included covariates of accuracy, sex, age, and P1 amplitude from the green (baseline) background. As this was a secondary analysis based on this particular pairwise group by color interaction, only participants from the schizophrenia-spectrum and nonpsychiatric control groups were included in this analysis (N ¼ 29). We found that a smaller increase and/or decrease in P1 amplitude to the red background related to increased self-reported social anhedonia (i.e., lower ACIPS score), r (23) ¼0.44, p ¼0.03, and other general negative symptoms (i.e., higher score on SPQ-BR Interpersonal factor), r(23) ¼  0.52, p ¼0.008. When comparing the chronic mood disorders group to the nonpsychiatric controls, there was no main effect or interaction with color for P1 mean amplitude (ps 4 .26). Similarly, when the schizophrenia-spectrum group was compared to the chronic mood disorders group, there was no main effect or interaction with color for P1 mean amplitude (ps 4.41). We then examined symptom correlates of P1 mean amplitude using stepwise regressions across the entire sample (N ¼48), as these symptoms and traits can occur across traditional diagnostic class boundaries (see Table 4). Behavioral accuracy was entered in an initial block as a covariate for each regression. These series of regressions revealed a significant negative relationship between P1 mean amplitude with the green background and the PANSS (clinician-rated state measure) item of Emotional Withdrawal, r (48) ¼ 0.31, p ¼0.03 (see Fig. 2). The interaction between diagnostic class membership and Emotional Withdrawal in relation to P1 amplitude from the green background was not statistically significant, suggesting a relatively similar relationship across the two diagnostic classes and controls, F(2,35)¼ 0.18, p ¼0.84, η2 ¼ 0.01. In addition, the P1 mean amplitude from both color backgrounds showed a negative relationship with the SPQ-BR (self-report psychometric scale) subscale of Eccentric Behavior, a Table 4 Stepwise regression outcomes for relationships between visual-evoked potential mean amplitudes with psychiatric symptoms – across entire sample (N ¼ 48). VEP component

PANSS Items

Psychometric self-report

P1 from green background P1 from red background N1 from both backgrounds

Emotional withdrawal (β ¼  0.29; p ¼0.05) None entered model

SPQ eccentric behavior (β ¼  0.35; p ¼ 0.02) SPQ eccentric behavior (β¼  0.37; p ¼ 0.01) None entered model

None entered model

All stepwise regressions used entry alpha of 0.05 and exit of 0.10, and included behavioral accuracy in an initial block as a covariate. Each cell represents a separate stepwise regression. All beta values listed are standardized. VEP ¼visual-evoked potential; PANSS ¼Positive and Negative Symptom Scale – all individual symptoms were included in model; Psychometric Self-Report Scales include the 7 subscales from Schizotypal Personality Questionnaire-Brief Revised (SPQ) along with Temporal Experience of Pleasure Scale and the Anticipatory and Consummatory Interpersonal Pleasure Scale total scores. Note: None of the regression models showed problems with collinearity among the predictors (all Condition Index scoreso3.59).

267

disorganized symptom. When P1 was averaged from both backgrounds, the negative relationship with Eccentric Behavior was somewhat stronger, r(48) ¼  0.40, p¼ 0.006 (see Fig. 3). The interaction between diagnostic class membership and the Eccentric Behavior subscale in relation to P1 amplitude across both backgrounds was not statistically significant, suggesting a relatively similar relationship across the two diagnostic classes and controls, F(2,34) ¼ 1.05, p ¼0.36, η2 ¼ 0.06. The anhedonia scales from TEPS and ACIPS did not enter the model. 3.2. N1 component Across all participants (N ¼48), there were no main effects of color or contrast on N1 mean amplitude, and no interaction between color and contrast. When we only examined participants with no psychiatric diagnosis (N ¼14), we found the same pattern of results for color and contrast. Therefore, for the purposes of analyses, N1 amplitude was averaged across color and contrast conditions. Pairwise diagnostic class comparisons using univariate ANCOVAs revealed no significant group differences on N1 mean amplitude (ps 4.86). Similarly, N1 mean amplitude did not relate to any of the symptoms examined in stepwise regressions (see Table 4).

4. Discussion Pairwise comparisons between schizophrenia-spectrum, chronic mood disorder, and nonpsychiatric control groups did not find any significant group difference on P1 mean amplitude. While this is partially consistent with one study that found no difference in P1 amplitude in individuals with bipolar II disorder (Elvsashagen et al., 2012), it is inconsistent with our hypothesis, which was based on many studies finding a reduced P1 amplitude in schizophrenia, schizotypy, bipolar I disorder, and unipolar depression samples (see Introduction). One possible explanation is our use of only low contrast stimuli, as the other studies have typically used high-contrast stimuli that may be more sensitive to P1 amplitude reductions found in these disorders. In addition, this appears to be the first VEP study to examine broad diagnostic classes rather than single disorders. We partially supported our hypothesis regarding the red background, as we replicated an earlier study which reported that nonpsychiatric adults scoring higher on the SPQ total score showed no significant change in P1 amplitude to a red, vs. green, background, while participants scoring lower on the SPQ showed a significant reduction in P1 amplitude to the red background (Bedwell et al., 2013). Similar to that finding, our schizophreniaspectrum group showed no change in P1 amplitude to the red background, while our nonpsychiatric group showed a significant increase in P1 amplitude (see Fig. 1). We were surprised that the P1 amplitude was larger rather than smaller for the red background in our nonpsychiatric controls considering the findings in the low schizotypy controls in the previous study (Bedwell et al., 2013), which seems more consistent with the known property of diffuse red light suppressing the dorsal visual pathway in healthy adults. However, that study only included high-contrast check stimuli (64%), which were not included in the current study. The current contrasts were much lower, which elicits less activity in both pathways. Therefore, it is possible that this overall increase in P1 may reflect a compensatory increase in the ventral input in response to the selective suppression (by red light) of the already weak dorsal signal. However, this is a theory that needs to be directly tested in future studies which include both low and highcontrast stimuli. Despite the difference in the direction of change in the P1

268

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

Fig. 2. Scatterplot depicting the relationship of P1 mean amplitude from the green background with scores on the emotional withdrawal item from the positive and negative symptom scale across the entire sample. The shaded region represents the 95% confidence interval. The category of “Other Psychiatric Disorders” includes the six participants with disorders outside of the diagnostic classes of interest (e.g., anxiety disorders).

amplitude in controls between the studies, both found that red light modulated P1 amplitude in nonpsychiatric controls, but had no significant effect in individuals with schizophrenia-spectrum disorders or personality traits. The earlier study found that the SPQ subscale of Ideas of Reference (a positive symptom) was related to this red light modulation (Bedwell et al., 2013), while the current study found that social anhedonia and general negative symptoms were related. While these two studies were inconsistent in regard to symptom relationships, the current finding of negative symptoms showing the strongest relationship to the differential modulation by red light is consistent with two earlier studies examining behavioral backward masking accuracy change to a red background in individuals with schizophrenia (Bedwell et al., 2011b) and nonpsychiatric male participants with schizotypy (Bedwell et al., 2011a). Thus, the overall pattern suggests that a lack of modulation in the visual system by red light in individuals with schizophrenia-spectrum conditions is primarily related to negative symptoms such as social anhedonia, constricted affect, few close friends, social anxiety, and apathy. When examining psychometric self-reported schizotypy and anhedonia features across the entire sample (N ¼48), we found that individuals scoring higher on the SPQ-BR Eccentric Behavior scale had reduced P1 amplitudes across both background colors (see Fig. 2). In addition, when examining clinician-rated current symptoms from the PANSS, we found that individuals with higher scores on the Emotional Withdrawal item (i.e., lack of interest/ apathy to life's events) had reduced P1 amplitudes in the green

condition (see Fig. 3). As both of these dimensional findings were exploratory and there does not appear to be any existing transdiagnostic studies to compare these particular findings, we consider them preliminary. However, apathy and eccentric behavior may share underlying vulnerability factors with a reduced P1 amplitude, which provides a model for future studies to test. Finally, when we examined N1 mean amplitudes, there were no relationships found with our diagnostic classes or psychiatric symptoms. Some studies have reported reduced N1 amplitude in schizophrenia (Butler and Javitt, 2005; Martinez et al., 2012) and unipolar depression (Pierson et al., 1996; Normann et al., 2007). However, the relative lack of relationships found with N1 in our study is consistent with the majority of the literature reporting more specific differences in P1 in schizophrenia and bipolar I (Foxe et al., 2005; Yeap et al., 2009; Butler et al., 2013; Verleger et al., 2013). We were surprised that the contrast manipulation did not affect the P1 amplitude. This may be due to our choice of isoluminant and 12% contrasts that were based on physical (i.e., screen), not on individualized psychophysical (i.e., perceived), luminance. The “isoluminant” stimuli were likely perceived by most participants as a slight contrast difference, producing two conditions that were both perceived as “low contrast,” and resulting in the observed similarilty in P1 amplitude. To the authors’ knowledge, this is the first study to examine VEPs in a broadly-defined transdiagnostic sample with a focus on larger diagnostic classes and individual psychiatric symptoms that relate to VEP abnormalities. The current study is limited by a

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

269

Fig. 3. Scatterplot depicting the relationship of P1 mean amplitude from all conditions with scores on the eccentric behavior subscale of the schizotypal personality disorder brief revised scale across the entire sample. The shaded region represents the 95% confidence interval. The category of “Other Psychiatric Disorders” includes the six participants with disorders outside of the diagnostic classes of interest (e.g., anxiety disorders).

relatively small sample size, an exploratory approach that increases Type I error probability, and the lack of a high-contrast stimulus. However, despite these limitations, our findings provide important clues to VEP abnormalities reported in earlier studies which assessed particular psychiatric disorders. While preliminary, it appears that within schizophrenia-spectrum disorders, a lack of modulation in P1 amplitude to a red background is related to increased social, but not general, anhedonia, along with other negative symptoms. In addition, regardless of diagnostic class, a reduced P1 amplitude may be particularly prominent in individuals who have high-levels of emotional withdrawal/apathy (a particular negative symptom) and/or eccentric/ disorganized behavior. This study also provides an example of a novel methodological framework for examining a broader array of biomarkers using an RDoC approach. It is possible that the same brain network disturbance that elicits these particular symptoms also changes the visual networks in a specific manner. Thus, examining the visual system response to diffuse red light, whether through VEPs or behavioral methods such as backward masking, may be a sensitive biomarker to negative symptoms found in schizophrenia-spectrum conditions. These relationships provide clues for uncovering the underlying causal pathology for these transdiagnostic symptoms.

Acknowledgments The authors thank the following undergraduate research assistants who were instrumental to the success of this project: Lisa Kadison, Maricel Chacon, and Carlos Puentes. We also thank Christopher C. Spencer, a graduate student who created the scatterplot figures. This project was funded by an internal university “SEED” Grant to Dr. Bedwell from the University of Central Florida College of Sciences and Office of Research and Commercialization (Research ID: 1053929).

References Aichert, D.S., Williams, S.C., Moller, H.J., Kumari, V., Ettinger, U., 2012. Functional neural correlates of psychometric schizotypy: an fMRI study of antisaccades. Psychophysiology 49, 345–356. Bakanidze, G., Roinishvili, M., Chkonia, E., Kitzrow, W., Richter, S., Neumann, K., Herzog, M.H., Brand, A., Puls, I., 2013. Association of the nicotinic receptor alpha7 subunit gene (CHRNA7) with schizophrenia and visual backward masking. Front. Psychiatry 4, 133. Bedwell, J.S., Brown, J.M., Miller, L.S., 2003. The magnocellular visual system and schizophrenia: what can the color red tell us? Schizophr. Res. 63, 273–284. Bedwell, J.S., Chan, C.C., Trachik, B.J., Rassovsky, Y., 2013. Changes in the visualevoked P1 potential as a function of schizotypy and background color in healthy young adults. J. Psychiatry Res. 47, 542–547. Bedwell, J.S., Hernandez, D.C., Ranieri, A.Y., 2011a. The backward masking red light

270

J.S. Bedwell et al. / Psychiatry Research 230 (2015) 262–270

effect and schizotypy: the influence of sex. Psychiatry Res. 189, 228–232. Bedwell, J.S., Rassovsky, Y., Orem, D.M., Kamath, V., 2011b. The backward masking red light effect in schizophrenia: Relationship to clinical features and neurocognitive performance. J. Abnormal Psychol. 120, 490–496. Butler, P.D., Abeles, I.Y., Silverstein, S.M., Dias, E.C., Weiskopf, N.G., Calderone, D.J., Sehatpour, P., 2013. An event-related potential examination of contour integration deficits in schizophrenia. Front. Psychol. 4, 132. Butler, P.D., Javitt, D.C., 2005. Early-stage visual processing deficits in schizophrenia. Curr. Opinion Psychiatry 18, 151–157. Butler, P.D., Martinez, A., Foxe, J.J., Kim, D., Zemon, V., Silipo, G., Mahoney, J., Shpaner, M., Jalbrzikowski, M., Javitt, D.C., 2007. Subcortical visual dysfunction in schizophrenia drives secondary cortical impairments. Brain: J. Neurol. 130, 417–430. Butler, P.D., Silverstein, S.M., Dakin, S.C., 2008. Visual perception and its impairment in schizophrenia. Biol. Psychiatry 64, 40–47. Callaway, D.A., Cohen, A.S., Matthews, R.A., Dinzeo, T., 2014. Schizotypal Personality Questionnaire-Brief Revised: psychometric replication and extension. Personal. Disord. 5, 32–38. Cohen, A.S., Matthews, R.A., Najolia, G.M., Brown, L.A., 2010. Toward a more psychometrically sound brief measure of schizotypal traits: introducing the SPQBrief Revised. J. Personal. Disord. 24, 516–537. Cuthbert, B.N., Insel, T.R., 2013. Toward the future of psychiatric diagnosis: the seven pillars of RDoC. BMC Med. 11, 126. de Monasterio, F.M., 1978. Properties of concentrically organized X and Y ganglion cells of macaque retina. J. Neurophysiol. 41, 1394–1417. Di Russo, F., Martinez, A., Sereno, M.I., Pitzalis, S., Hillyard, S.A., 2002. Cortical sources of the early components of the visual evoked potential. Hum. Brain Mapp. 15, 95–111. Elvsashagen, T., Moberget, T., Boen, E., Boye, B., Englin, N.O., Pedersen, P.O., Andreassen, O.A., Dietrichs, E., Malt, U.F., Andersson, S., 2012. Evidence for impaired neocortical synaptic plasticity in bipolar II disorder. Biol. Psychiatry 71, 68–74. Foxe, J.J., Murray, M.M., Javitt, D.C., 2005. Filling-in in schizophrenia: a high-density electrical mapping and source-analysis investigation of illusory contour processing. Cereb. Cortex 15, 1914–1927. Gard, D.E., Kring, A.M., Gard, M.G., Horan, W.P., Green, M.F., 2007. Anhedonia in schizophrenia: distinctions between anticipatory and consummatory pleasure. Schizophr. Res. 93, 253–260. Gooding, D.C., Cohen, A.S., Pflum, M.J., 2014. Hedonic capacity and schizotypy: evidence for the criterion validity of the ACIPS. Compr. Psychiatry 55, 1455–1461. Gooding, D.C., Pflum, M.J., 2014. The assessment of interpersonal pleasure: introduction of the Anticipatory and Consummatory Interpersonal Pleasure Scale (ACIPS) and preliminary findings. Psychiatry Res. 215, 237–243. Green, M.F., Nuechterlein, K.H., Breitmeyer, B., Mintz, J., 2006. Forward and backward visual masking in unaffected siblings of schizophrenic patients. Biol. Psychiatry 59, 446–451. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261–276.

Koychev, I., El-Deredy, W., Haenschel, C., Deakin, J.F.W., 2010. Visual information processing deficits as biomarkers of vulnerability to schizophrenia: an eventrelated potential study in schizotypy. Neuropsychologia 48, 2205–2214. Livingstone, M.S., Hubel, D.H., 1984. Anatomy and physiology of a color system in the primate visual cortex. J. Neurosci. 4, 309–356. Luck, S.J., 2014. An Introduction to the Event-Related Potential Technique, 2nd edition. MIT Press, Cambridge. Marco, E.J., Hinkley, L.B., Hill, S.S., Nagarajan, S.S., 2011. Sensory processing in autism: a review of neurophysiologic findings. Pediatr. Res. 69, 48R–54R. Martinez, A., Hillyard, S.A., Bickel, S., Dias, E.C., Butler, P.D., Javitt, D.C., 2012. Consequences of magnocellular dysfunction on processing attended information in schizophrenia. Cereb. Cortex 22, 1282–1293. Normann, C., Schmitz, D., Furmaier, A., Doing, C., Bach, M., 2007. Long-term plasticity of visually evoked potentials in humans is altered in major depression. Biol. Psychiatry 62, 373–380. Obayashi, C., Nakashima, T., Onitsuka, T., Maekawa, T., Hirano, Y., Hirano, S., Oribe, N., Kaneko, K., Kanba, S., Tobimatsu, S., 2009. Decreased spatial frequency sensitivities for processing faces in male patients with chronic schizophrenia. Clin. Neurophysiol. 120, 1525–1533. Pierson, A., Ragot, R., Van Hooff, J., Partiot, A., Renault, B., Jouvent, R., 1996. Heterogeneity of information-processing alterations according to dimensions of depression: an event-related potentials study. Biol. Psychiatry 40, 98–115. Schechter, I., Butler, P.D., Zemon, V.M., Revheim, N., Saperstein, A.M., Jalbrzikowski, M., Pasternak, R., Silipo, G., Javitt, D.C., 2005. Impairments in generation of early-stage transient visual evoked potentials to magno- and parvocellularselective stimuli in schizophrenia. Clin. Neurophysiol. 116, 2204–2215. Sponheim, S.R., Sass, S.M., Noukki, A.L., Hegeman, B.M., 2013. Fragile early visual percepts mark genetic liability specific to schizophrenia. Schizophr. Bull. 39, 839–847. Verleger, R., Talamo, S., Simmer, J., Smigasiewicz, K., Lencer, R., 2013. Neurophysiological sensitivity to attentional overload in patients with psychotic disorders. Clin. Neurophysiol. 124, 881–892. Wilkinson, G.S., 1993. Wide Range Achievement Test (WRAT3) Administration Manual. Wide Range, Inc., Wilmington, DE. Yeap, S., Kelly, S.P., Reilly, R.B., Thakore, J.H., Foxe, J.J., 2009. Visual sensory processing deficits in patients with bipolar disorder revealed through high-density electrical mapping. J. Psychiatry Neurosci. 34, 459–464. Yeap, S., Kelly, S.P., Sehatpour, P., Magno, E., Garavan, H., Thakore, J.H., Foxe, J.J., 2008a. Visual sensory processing deficits in Schizophrenia and their relationship to disease state. Eur. Arch. Psychiatry Clin. Neurosci. 258, 305–316. Yeap, S., Kelly, S.P., Sehatpour, P., Magno, E., Javitt, D.C., Garavan, H., Thakore, J.H., Foxe, J.J., 2006. Early visual sensory deficits as endophenotypes for schizophrenia: high-density electrical mapping in clinically unaffected first-degree relatives. Arch. Gen. Psychiatry 63, 1180–1188. Yeap, S., Kelly, S.P., Thakore, J.H., Foxe, J.J., 2008b. Visual sensory processing deficits in first-episode patients with Schizophrenia. Schizophr. Res. 102, 340–343. Yoon, J.H., Sheremata, S.L., Rokem, A., Silver, M.A., 2013. Windows to the soul: vision science as a tool for studying biological mechanisms of information processing deficits in schizophrenia. Front. Psychol. 4, 681.