Magnocellular and parvocellular contributions to backward masking dysfunction in schizophrenia

Schizophrenia Research 64 (2003) 91 – 101 www.elsevier.com/locate/schres

Magnocellular and parvocellular contributions to backward masking dysfunction in schizophrenia Isaac Schechter *, Pamela D. Butler, Gail Silipo, Vance Zemon, Daniel C. Javitt Program in Cognitive Neuroscience and Schizophrenia, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Road, Orangeburg, NY 10962, USA Received 12 July 2002; received in revised form 6 December 2002; accepted 23 December 2002

Abstract Patients with schizophrenia have repeatedly shown deficits in visual processing. These deficits have been well documented using visual backward masking (VBM). The VBM deficit in schizophrenia is thought to be due to aberrant interactions between magnocellular (M) and parvocellular (P) visual pathways. To date, no study has studied these claims with rigorous stimuli isolating M and P pathway responses. This study examined the function of each pathway and their interactions by creating Mand P-biased targets based on their known physiological properties. The M system responds to very low luminance contrast whereas the P system does not, and the P system responds to color contrast whereas the M system generally does not. Thus, to activate the P system, target letters and masks utilized color contrast, and to activate the M system, target letters and masks utilized very low luminance contrast. Four conditions were presented such that M- and P-biased targets were paired with both M- and P-biased masks. A significant Group  Mask Condition interaction was found when a P target was used in combination with an M or P mask, but not when an M target was used. In particular, schizophrenia patients needed significantly longer interstimulus intervals (ISIs) than controls to escape from masking in the P target/M mask condition, but not in any of the other three conditions. In addition, the critical stimulus durations (CSDs) for unmasked stimuli were significantly increased for both M and P targets in patients relative to controls. These findings demonstrate a significant impairment in M, but not P pathway, function in patients with schizophrenia. Furthermore, deficits of letter identification, including those of P targets, may also reflect impairment of the M pathway given the priming function of the dorsal stream. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Schizophrenia; Backward masking; Visual pathways; Information processing

1. Introduction Deficits in early visual processing have been demonstrated repeatedly using the visual backward * Corresponding author. Tel.: +1-845-398-6617; fax: +1-845398-6545. E-mail address: [email protected] (I. Schechter).

masking (VBM) paradigm in schizophrenia (Braff et al., 1991; Butler et al., 1996; Green and Nuechterlein, 1999; McClure, 2001; Rund 1993; Saccuzo and Braff, 1986; Slaghuis and Bakker, 1995; Weiner et al., 1990). VBM occurs when two stimuli are presented in rapid succession. Under such circumstances, recognition of the first (target) stimulus is hampered by the presentation of the second (masking) stimulus. In

0920-9964/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0920-9964(03)00008-2

92

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

schizophrenia, visual targets require a longer processing time to ‘‘escape’’ the effects of a subsequent mask (Braff and Saccuzzo, 1982; Butler et al., 1996; Green and Nuechterlein, 1999; Rund, 1993; Saccuzzo and Braff, 1986; Slaghuis and Bakker, 1995; Weiner et al., 1990). VBM occurs because the second stimulus interferes with the trace (icon) produced by the first stimulus. In the traditional backward masking task, the targets are letters and the mask is a cross of Xs which spatially overlap the target. In such a paradigm, patients with schizophrenia require approximately 300 ms between target and mask to successfully identify the target, whereas comparison volunteers need only about 120 ms (Braff and Saccuzzo, 1982; Butler et al., 1996; Saccuzzo and Braff, 1986; Weiner et al., 1990). In the classic literature, the phenomenon of backward masking has been attributed to interactions between two visual channels termed ‘‘transient’’ and ‘‘sustained’’ (Breitmeyer and Ganz, 1976; Green, 1984; Keesey, 1972; Kulikowski and Tolhurst, 1973; Tolhurst, 1973). The transient channel responds rapidly and briefly to stimulation, whereas the sustained channel shows a longer duration response related to stimulus identification. Masking by interruption occurs in this model because the more rapid transient response ‘‘catches up’’ to the continued processing within the sustained channel. In light of this model, VBM deficits in schizophrenia have been attributed to hyperactivity of the transient channel, leading to greater transient channel interference with ongoing sustained channel processing (Green et al., 1994b; Merritt and Balogh, 1989; Schuck and Lee, 1989; Slaghuis and Curran, 1999). The transient channel is known to be activated preferentially by low spatial frequency stimuli (i.e., those with large internal elements), whereas the sustained channel is activated preferentially by high spatial frequency stimuli (i.e., those with small internal elements). The hypothesis of transient channel hyperactivity in schizophrenia has been supported over recent years by studies using low and high spatial frequency masks (Butler et al., 2002; Slaghuis and Curran, 1999). Similarly, deficits have been found on tasks that increased reliance on transient channel activity by using a target location task or a blurred target, but not on tasks that increased reliance on sustained channel activity (Cadenhead et al., 1998;

Green et al., 1994b; Keri et al., 2001). However, conflicting results have also been reported (Chen et al., 1999c; Keri et al., 2002; Schwartz et al., 1987; Schwartz and Winstead, 1985; Slaghuis, 1998), and mechanisms underlying transient channel dysfunction remain to be determined. Over the past decade, a more complex picture of visual processing has emerged. The psychophysically defined transient and sustained channels have been shown to correspond to the neuronal properties of distinct visual pathways, termed magnocellular (M) and parvocellular (P). The M pathway is composed of large, rapidly conducting neurons that project primarily to the dorsal visual stream (the ‘‘where’’ system), which is primarily involved in processing of motion and location of salient stimuli. The P pathway is composed of smaller, more slowly conducting neurons that project primarily to the ventral visual stream (the ‘‘what’’ system), which is responsible primarily for object recognition. Despite the preferential targeting of dorsal and ventral streams by M and P neurons, however, information cross-over does occur (Felleman and Van Essen, 1991; Merigan and Maunsell, 1993). Properties of the M and P neurons correspond roughly to properties of the transient and sustained channels, respectively (Cadenhead et al., 1998). Properties of M and P neurons have been extensively characterized over recent decades, permitting development of stimuli that differentially activate the two systems. Conscious object recognition occurs primarily within the ventral visual stream. Decoding of simple stimulus properties (e.g., spatial frequency) occurs when information first enters the ventral stream following stimulation (feedforward sweep). However, decoding of complex information, such as letter identity, requires several ‘‘sweeps’’ of information through the visual system (Lamme and Roelfsema, 2000). This recurrent activity, which involves both dorsal and ventral stream structures, may correspond to the classically defined sustained response. Because transmission is more rapid through the dorsal, than ventral, visual stream, high intensity M-biased stimuli are particularly effective as masks. However, masking may occur both within and across visual pathways. Differential properties of M and P cells make it possible to evaluate the relative contributions of M

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

and P pathway dysfunction to VBM deficits in schizophrenia. M cells are more sensitive than P cells to stimuli with low luminance contrast, but saturate more quickly as a function of increasing luminance contrast (Kaplan, 1991). In fact, the cortical neurons that receive P input do not respond below 8% contrast (Tootell et al., 1988). Thus, low luminance contrast stimuli bias processing towards the M pathway, whereas stimuli that are modulated around a high luminance contrast ‘‘pedestal’’ bias processing towards the P pathway. M and P cells also show differential sensitivity to size. M cells are activated vigorously by relatively large (i.e., low spatial frequency) stimulus elements, whereas P cells are activated more strongly by relatively small (i.e., high spatial frequency) stimulus elements (Kaplan, 1991; Merigan and Maunsell, 1993). In addition, M cells respond to low contrast objects and movement of these objects in the visual field, whereas P cells are relatively insensitive to movement (Merigan and Maunsell, 1993). Finally, M cells are not highly responsive to chromatic (color) contrast whereas P cells are (Kaplan, 1991; Merigan and Maunsell, 1993). We have recently observed that patients with schizophrenia show reduced electrophysiological activation to M-biased stimuli (Butler et al., 2001), consistent with prior literature showing contrast sensitivity deficits to M-biased stimuli (Schwartz et al., 1987; Schwartz and Winstead, 1985; Slaghuis, 1998). Based upon transient/sustained channel interaction models of VBM dysfunction in schizophrenia, it was predicted that patients would show deficits under conditions where an M-biased stimulus is used as a mask, but not under conditions in which a P-biased stimulus is used. This study utilizes the physiological properties of M and P pathways to examine the role of parallel pathways in VBM dysfunction. Luminance contrast was used to emphasize M pathway activity and color contrast was utilized to emphasize P pathway activity. The relationship of backward masking to clinical symptoms was also examined. Based on previous studies, it was hypothesized that a negative correlation would be found such that patients with the highest level of negative symptoms would show the greatest deficit in masking (Braff, 1989; Butler et al., 2002; Cadenhead et al., 1997; Green and Walker, 1984; Slaghuis and Bakker, 1995; Weiner et al., 1990).

93

2. Methods 2.1. Participants Twenty patients meeting the criteria for DSM-IV diagnosis of schizophrenia (n = 19) or schizoaffective disorder (n = 1) provided written informed consent and participated in the study. Patients were recruited from the Clinical Research and Evaluation Facility (CREF) at the Nathan Kline Institute (NKI), a state psychiatric facility for assisted living and an outpatient medication clinic. Diagnosis was obtained by means of chart review and the Structured Clinical Interview for DSM-IV (SCID; First et al., 1997). All patients were rated using the Positive and Negative Syndrome Scale (PANSS; Kay et al., 1992). Patients were excluded if they met criteria for alcohol or substance dependence, current alcohol or substance abuse or any major neurological disorder. All patients were taking atypical antipsychotic medication at the time of testing. Two were also taking typical antipsychotics. Mean F S.E.M. chlorpromazine equivalents were 1262.7 F 70.2. Eighteen control volunteers also participated in the study. Controls with a history of major psychiatric or neurological disorder, or a history of alcohol or substance dependence or current alcohol or substance abuse were excluded. Patients and controls did not differ significantly in age (patients: 41.6 F 1.5; controls: 38.3 F 2.2 years), though they differed in gender composition (patients: 3 females/17 males; controls: 11 females/7 males). PANSS, negative, positive, cognitive, excitement and depression/anxiety factor scores (Lindenmayer et al., 1994) were 18.8 F 1.3, 12.8 F 1.0, 14.7 F 0.9, 7.6 F 0.5 and 12.4 F 0.6, respectively. 2.2. Apparatus Stimuli were presented using a VENUS system (Neuroscientific, Farmingdale, NY) and RGB monitor with a frame rate of f 119 Hz (noninterlaced). Use of the VENUS system allows precise control over luminance and chromatic contrast. Viewing distance was 114 cm and the stimulus field subtended 8j  8j of visual angle. Targets and masks appeared on a yellow background. Luminance of the yellow background was 60 cd/m2.

94

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

2.3. Stimuli There were 12 target letters (E, H, I, K, L, M, T, V, W, X, Y, Z). The pattern mask consisted of an X that fell on the target letter. Targets and masks subtended 5j of visual angle. Targets and masks were designed to bias processing towards M or P pathways (Table 1). M-biased targets and masks were low luminance contrast dark yellow letters on a yellow background (M target = 3% luminance contrast; M mask = 8% luminance contrast using the formula for Michaelson contrast). P-biased targets and masks were isoluminant chromatic contrast red letters on a yellow background (P target = 8% chromatic contrast; P mask = 20% chromatic contrast). Isoluminance is the point at which there is no difference in luminance between two stimuli that differ in chromatic contrast. To determine the isoluminant point for each participant, a psychophysical adaptation of the electrophysiological technique of Zemon et al. (1991) was used, and isoluminance was estimated through manipulation of the ratio of red and green guns of the RGB display monitor. Signals from the two guns were modulated in counterphase (i.e., 180j out of phase with respect to one another). The red/green ratio at which no apparent flicker was reported was used as the isoluminant point (for detailed methodology, see Greenstein et al., 1998). Ten repetitions of stimuli were used to confirm the isoluminant point. Each individual’s isoluminant point was used in presenting chromatic contrast stimuli. Presentation of chromatic contrast stimuli involved counterphase modulation of the red and green guns with the blue gun set to zero. Chromatic contrast was defined by the depth of modulation of the red gun. The green gun was always modulated in counterphase with the red

Table 1 M and P target and mask combinations Target

Mask

M-biased—3% luminance contrast M-biased—3% luminance contrast P-biased—8% chromatic contrast P-biased—8% chromatic contrast

M-biased—8% luminance contrast P-biased—20% chromatic contrast M-biased—8% luminance contrast P-biased—20% chromatic contrast

gun at a depth of modulation used to yield isoluminance for each participant. The depth of modulation of the red gun for the P target was 8% and for the P mask was 20%. 2.4. Procedure 2.4.1. Critical stimulus duration Critical stimulus durations (CSDs) were determined for each subject prior to VBM testing. An up-down transformed response method was used to obtain a 79% level of accuracy for M and P target stimuli (Wetherill and Levitt, 1965). Stimulus duration was increased for every wrong response and decreased for every three consecutive correct responses. Duration was increased in 8 ms steps because smaller increments were not possible given the frame rate limitation of the VENUS system. A reversal occurs every time the direction is changed. The mean of eight reversals was used to obtain threshold. 2.4.2. General procedure Testing was performed in either one or two sessions. Participants were tested first in the P target conditions and then in the M target conditions. Target letters were presented in a predetermined random order. For all four masking conditions, the blank yellow background screen was presented for 952 ms. Target letters were presented at each participant’s CSD followed by the mask at one of six interstimulus intervals (ISIs: 8, 32, 64, 96, 128 and 256 ms). The blank yellow screen was presented during the ISI. In order to ensure salience of the mask, duration of Pbiased masks was presented at each individual’s Pbiased target CSD, and duration of the M-biased masks for the M target condition was presented at each individual’s M-biased target CSD. However, for the P target/M mask condition, M masks were presented at the P target CSD because participants had not yet received M CSD testing. The 128 ms ISI was presented first, followed by the 256, 96, 64, 32 and 8 ms ISIs. Data are presented as percent correct. There were a maximum of 15 trials per ISI block. However, to maximize testing efficiency and prevent participant frustration, an algorithm was employed. If a participant had consistent performance on the first five presentations (all correct

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

or all wrong) that ISI block was ended and scored as either 0% or 100% correct. If either eight or nine responses of the first 10 presentations were consistent (8/9 right or 8/9 wrong), presentation at that ISI was halted and scored as the corresponding percentage correct (10%, 20% or 80%, 90% correct). All participants were prompted to respond even when unsure and were given a sheet with all 12 possible letters to choose from. In addition to percent correct, emergence from masking was also determined and refers to the shortest ISI at which subjects achieved at least 20% correct performance.

95

tion coefficients were used to examine relationships between CSDs and between percent correct on backward masking or emergence from total masking and clinical variables. Data from points showing greatest between-group differences were used for correlational analyses. In the P target conditions, data was not collected from one patient and one control. In the M target conditions, data were not collected from four patients and two controls. For the 96 ms ISI, data were interpolated for two patients and two controls using surrounding data points.

2.5. Statistical analysis 3. Results The primary dependent variable consisted of percent correct performance at each ISI in each target/ mask condition. Masking effects were determined by separate repeated measures MANOVA (rmMANOVA) for each of the two target conditions. Thus, each rmMANOVA included one between-group factor (patient vs. control) and two within-group factors (P vs. M mask; ISI). Follow-up rmMANOVAs with one between-group factor (patient/control) and one withingroup factor (ISI) were performed for each of the four target-mask combinations. The range of active masking for each condition determined the ISI ranges used in each ANOVA. Post hoc t-tests were performed when ANOVA revealed significant effects or interactions. In order to facilitate comparison across conditions, a secondary outcome variable, emergence from masking, was also calculated for each subject. Given the number of trials per sequence (15) and chance performance level on each trial (1/12), chance performance rate per ISI was 1 –2 correct (8– 16%). Emergence from masking was defined as the briefest ISI at which subjects achieved 3/15 (20%) correct responses. Between-group differences in emergence from masking in each of the two target conditions were determined by rmMANOVA with a between-group factor of patient vs. control and a within-group factor of mask type (P vs. M). Post hoc t-tests were performed when ANOVA revealed significant main or interaction effects. Demographic comparisons were analyzed for between-group differences using two-tailed t-tests. CSD data were compared between groups using two-tailed t-tests. Pearson’s product moment correla-

3.1. Demographics Patients did not differ in age from the control group (t = 1.32, df = 35, p = 0.20). Although more females were in the control group, dependent measures did not differ between males and females in the control group. 3.2. CSDs CSDs for P- and M-biased targets were increased by 20% and 70%, respectively, for patients as compared with controls. The main effect of group was highly significant ( F = 17.29, df = 1/32, p < 0.0001), with no significant group X target interaction ( F = 1.29, df = 1/32, p = 0.27). CSDs were significantly decreased for patients versus controls in both P and M target conditions (t = 2.89, df = 36, p = 0.006; t = 3.41, df = 32, p = 0.002, respectively) (Fig. 1). A significant positive correlation was found between M- and P-biased CSDs (r = 0.43, df = 32, p = 0.01). 3.3. P target conditions In the P target conditions (Fig. 2), patients showed significantly greater susceptibility to M than P masks, as reflected in a significant group X mask interaction ( F = 13.27, df = 1/34, p = 0.001). In contrast, the main effect of group was not significant ( F = 0.92, df = 1/ 34, p = 0.34). A follow-up rmMANOVA focusing on the P target/M mask condition showed a significant main effect of group over the range of ISIs from 8 to 64 ms ( F = 5.57, df = 1/34, P = 0.02), indicating

96

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

of group ( F = 0.28, df = 1/34, p = 0.6) or group X ISI interaction ( F = 0.5, df = 2/33, p = 0.61) was observed in the P target/P mask condition. 3.4. M target conditions

Fig. 1. Bar graph of critical stimulus duration (CSD) to P and M targets for patients and controls. P CSD: n = 20 schizophrenia patients; n = 18 controls; M CSD: n = 18 schizophrenia patients; n = 16 controls. **p < 0.01.

greater susceptibility to masking in patients vs. controls. Follow-up t-tests showed that at ISIs of 8 and 32 ms patients had significantly lower percent correct detection of the target than controls in the P target/M mask condition. In contrast, no significant main effect

In the M-target condition (Fig. 3), there was no significant group ( F = 1.47, df = 1/28, p = 0.24) or group X mask ( F = 1.13, df = 1/28, p = 0.3) interaction. ANOVA showed a trend for a significant group effect in performance over the range of active masking (8– 128 ms) in the M target/M mask condition ( F = 3.94, df = 1/30, p = 0.06). No significant group X ISI interaction ( F = 0.65, df = 4/27, p = 0.63) was found. No significant group ( F = 0.21, df = 1/28, p = 0.65) or group X ISI interaction ( F = 0.26, df = 4/25, p = 0.9) was observed in the M target/P mask condition. 3.5. Emergence from masking The overall rmANOVA revealed a significant group X mask interaction ( F = 4.13, df = 1/27, p = 0.05), but no significant group X target interaction ( F = 0.03, df = 1/27, p = 0.87). In the P target conditions, ANOVA demonstrated a significant group X

Fig. 2. Graph shows backward masking of P targets with M and P masks for patients and controls. P target conditions: n = 19 schizophrenia patients; n = 17 controls. *p < 0.05.

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

97

Fig. 3. Graph shows backward masking of M targets with M and P masks for patients and controls. M target/M mask condition: n = 16 schizophrenia patients; n = 16 controls; M target/P mask condition: n = 16 schizophrenia patients; n = 14 controls.

Fig. 4. Bar graph showing emergence from masking for M and P target conditions for patients and controls. P target conditions: n = 19 schizophrenia patients; n = 17 controls; M target conditions: n = 16 schizophrenia patients; n = 15 controls. *p < 0.05.

98

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

mask effect ( F = 5.57, df = 1/34, p = 0.02), indicating greater susceptibility of patients to M vs. P masks, similar to results of the primary analysis. In addition, follow-up t-tests showed that patients emerged from total masking significantly later than controls in the P target/M mask condition (t = 2.31, df = 34, p = 0.03). In the M target conditions, group ( F = 0.48, df = 1/29, p = 0.49) and group X mask ( F = 0.48, df = 1/29, p = 0.49) effects were not significant (Fig. 4). 3.6. Correlational analysis A significant positive correlation was found between the M CSD and severity of PANSS negative (r = 0.59, df = 16, p = 0.01) and cognitive (r = 0.46, df = 16, p = 0.05) factors. No significant correlations were found between PANSS factors and the P CSD. Patient performance on the P target/M mask condition at an ISI of 8 and 32 ms was not significantly correlated with any of the five factors emerging from the PANSS. Similarly, no correlation was found between the emergence from masking in the P target/M mask condition and any of the five PANSS factors. No correlation was found between either CSD, percent correct or emergence from masking and medication levels.

4. Discussion Deficits in VBM in schizophrenia are well documented (Braff et al., 1991; Butler et al., 1996; Green and Nuechterlein, 1999; McClure, 2001; Rund, 1993; Saccuzzo and Braff, 1986; Slaghuis and Bakker, 1995; Weiner et al., 1990). It has been hypothesized that the backward masking deficit in schizophrenia is due to an aberrant transient channel response to the mask disrupting the sustained channel response necessary for target identification (Green et al., 1994b; Merritt and Balogh, 1989; Schuck and Lee, 1989; Slaghuis and Curran, 1999). Previous masking studies have utilized masks that are a cross of Xs (Cadenhead et al., 1998; Green and Nuechterlein, 1999; Green et al., 1994b) or are low and high spatial frequency gratings (Butler et al., 2002; Slaghuis and Curran, 1999), which can activate both sustained and transient channels (Kaplan, 1991), leaving unresolved the issue of whether backward masking deficits in schizophrenia reflect an aberrant transient or sustained response

to the mask. This is the first study we are aware of to evaluate visual backward masking using the physiological properties of luminance and chromatic contrast to bias processing towards M or P pathways. Transient and sustained channels are defined psychophysically and are somewhat similar to the M and P visual pathways which have been anatomically and physiologically defined (Kaplan, 1991; Merigan and Maunsell, 1993). Consistent with a model of M or ‘‘transient channel’’ dysfunction in this disorder, patients showed a deficit relative to controls only when an M-biased mask was used. These results suggest significant impairment in M, but not P, function in patients with schizophrenia. These results also support earlier backward masking studies based on the transient/sustained distinction indicating that there is a transient channel deficit (Butler et al., 2002; Cadenhead et al., 1998; Green et al., 1999, 1994a,b; Keri et al., 2001; Slaghuis and Curran, 1999). A growing literature supports M pathway/dorsal stream dysfunction in schizophrenia. Patients with schizophrenia consistently show deficits on psychophysical tasks that tap dorsal stream function including velocity discrimination, coherent motion, and trajectory tasks (Chen et al., 1999a,b,c; O’Donnell et al., 1996; Stuve et al., 1997). In addition, several recent electrophysiological studies from our laboratory also support M and/or dorsal stream dysfunction. Using the differential properties of magnocellular and parvocellular cells (i.e., luminance and chromatic contrast), we have recently reported (Butler et al., 2001) a preferential M versus P deficit in patients with schizophrenia using steady-state VEPs. In a perceptual closure task, P1 and N1 amplitude were examined using transient VEPs (Foxe et al., 2001; Doniger et al., 2002). P1 arises from multiple generators over dorsal and ventral streams, whereas N1 is localized more specifically over ventral visual stream areas (Doniger et al., 2001; Murrary et al., 2001; Simpson et al., 1995; Woldorff et al., 1997). There was greater P1 amplitude reduction over dorsal, rather than ventral, scalp whereas N1 amplitude did not differ between groups (Doniger et al., 2002; Foxe et al., 2001) supporting the view that schizophrenia is associated with impairment of M pathway/dorsal stream processing. A second finding of the present study is that patients showed longer CSDs even to unmasked targets than did controls. This finding is in agreement

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

with much prior literature (Butler et al., 1996, 2002; Saccuzzo and Braff, 1986; Slaghuis and Bakker, 1995; Weiner et al., 1990). However, in previous studies (Butler et al., 1996, 2002; Saccuzzo and Braff, 1986; Slaghuis and Bakker, 1995; Weiner et al., 1990), targets were almost always high contrast black and white stimuli, which would activate both M and P pathways. This is the first study to our knowledge in which CSDs were obtained to M- or P-biased letters using the physiological properties of luminance and chromatic contrast. Patients showed deficits in both the M and P CSD conditions, but no differential deficit was found. It should be emphasized that deficits in letter recognition, as reflected in elevated CSDs, were statistically as robust as deficits in VBM and showed a greater degree of correlation with symptoms. The elevated CSD shows primary visual system pathology irrespective of increased VBM susceptibility. VBM deficits, however, occurred even following adjustment of CSD in the two groups, and so reflect an additional level of dysfunction. Although the M pathway projects predominately to the dorsal stream, there are numerous points at which M-inputs have access to the ventral stream (Merigan and Maunsell, 1993). Thus, ultimately the ventral stream is necessary for object identification in both M and P CSD conditions. In the case of the M CSD, the P pathway/ventral stream may be intact, and the M CSD longer due to M pathway dysfunction. With regard to the P CSD, since information transfer through the M pathway is generally faster, a crucial role of the M pathway may be to modulate and/or prime the ventral pathway or frame the visual scene (Doniger et al., 2002; Schroeder et al., 1998). Thus, the P CSD deficit may be due to lack of M priming or framing of the ventral stream to attend to the incoming stimulus. Both M and P CSD deficits being secondary to M pathway dysfunction may also explain the significant correlation between performance on the M- and P-biased CSDs. Schwartz et al. (1999) have recently shown that when letters were made up of moving dots, patients with schizophrenia showed letter identification deficits only at high, not low, dot velocity. Since the M pathway mediates processing of rapidly moving stimuli, this finding also suggests that letter recognition deficits in schizophrenia may be secondary to M pathway dysfunction.

99

In the present study, deficits on the M CSD, but not P CSD, were correlated with negative and cognitive components of the PANSS. This indicates that M dysfunction in schizophrenia is related to negative symptoms and cognitive dysfunction. These results are in agreement with previous studies showing that VBM deficits, which are thought to rely more on transient channel function, are more pronounced in patients with negative symptoms or poor prognosis than in patients with positive symptoms or good prognosis (Braff, 1989; Cadenhead et al., 1997; Green and Walker, 1984; Saccuzzo and Braff, 1981; Slaghuis and Bakker, 1995; Weiner et al., 1990). However, this study is the first to use physiological definitions of M and P pathway function (i.e., luminance and chromatic contrast) to examine the relationship with symptomatology. Somewhat surprising, given the previous data showing relationships between backward masking and negative symptoms, there were no significant correlations between backward masking performance and symptomatology. A limitation of this study is that all patients were receiving antipsychotic medication at the time of testing. Thus, a medication effect cannot be excluded since dopamine is found in the visual system and plays a role in visual processing including spatial frequency tuning (Bodis-Wollner and Tzelepi, 1998; Masson et al., 1993). However, several backward masking studies suggest that antipsychotic medications are unlikely to account for the current findings. For instance, in previous studies, we did not find a difference in backward masking performance in the same patients tested on and off medication (Butler et al., 1996, 2002). Furthermore, studies utilizing different groups of patients tested on and off antipsychotic medications showed that antipsychotic medications either reduced the deficit (Braff and Saccuzzo, 1982; Brody et al., 1980) or produced no difference between groups (Cadenhead et al., 1997; Harvey et al., 1990). In the current study, no significant correlations were found between CPZ equivalents and any measures of backward masking. Thus, while the issue of antipsychotic medication is important, it is unlikely that the magnocellular dysfunction in backward masking performance reported here is a medication effect. In conclusion, consistent with a model of magnocellular or ‘‘transient channel’’ dysfunction in this disorder, patients showed a deficit relative to controls

100

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101

only when an M-biased mask was used. These results suggest significant impairment in M pathway, but not P pathway, function in patients with schizophrenia. Correlations between the M CSD and PANSS negative symptom and cognitive impairment factors suggest that magnocellular dysfunction in schizophrenia may be related to negative symptoms and cognitive dysfunction.

Acknowledgements Supported by grants R01 MH49334 and the Burroughs Wellcome Fund to DCJ.

References Bodis-Wollner, I., Tzelepi, A., 1998. The push – pull action of dopamine on spatial tuning of the monkey retina: the effects of dopaminergic deficiency and selective D1 and D2 receptor ligands on the pattern electroretinogram. Vis. Res. 38, 1479 – 1487. Braff, D.L., 1989. Sensory input deficits and negative symptoms in schizophrenic patients. Am. J. Psychiatry 146, 1006 – 1011. Braff, D.L., Saccuzzo, D.P., 1982. Effect of antipsychotic medication on speed of information processing in schizophrenic patients. Am. J. Psychiatry 139, 1127 – 1130. Braff, D.L., Saccuzzo, D.P., Geyer, M.A., 1991. Information processing dysfunctions in schizophrenia: studies of visual backward masking, sensorimotor gating, and habituation. In: Steinhauer, S.R., Gruzelier, J.H., Zubin, J. (Eds.), Handbook of Schizophrenia, vol. 5. Elsevier, New York, pp. 303 – 334. Breitmeyer, B., Ganz, L., 1976. Implications of sustained and transient channels for theories of visual pattern masking, saccadic suppression and information processing. Psychol. Rev. 83, 1 – 36. Brody, D., Saccuzzo, D.P., Braff, D.L., 1980. Information processing for masked and unmasked stimuli in schizophrenia and old age. J. Abnorm. Psychology 89, 617 – 622. Butler, P.D., Harkavy-Friedman, J.M., Amador, X.F., Gorman, J.M., 1996. Backward masking in schizophrenia: relationship to medication status, neuropsychological functioning, and dopamine metabolism. Biol. Psychiatry 40, 295 – 298. Butler, P.D., Schechter, I., Zemon, V., et al., 2001. Dysfunction of early stage visual processing in schizophrenia. Am. J. Psychiatry 158, 1126 – 1133. Butler, P.D., DeSanti, L.A., Maddox, J., et al., 2002. Visual backward masking deficits in schizophrenia: relationship to visual pathway dysfunction and symptomatology. Schizophr. Res. 59, 199 – 209. Cadenhead, K.S., Geyer, M.A., Butler, R.W., Perry, W., Sprock, J., Braff, D.L., 1997. Information processing deficits of schizophrenia patients: relationship to clinical ratings, gender and medication status. Schizophr. Res. 28, 51 – 62.

Cadenhead, K.S., Serper, Y., Braff, D.L., 1998. Transient versus sustained visual channels in the visual backward masking deficits of schizophrenia patients. Biol. Psychiatry 43, 132 – 138. Chen, Y., Levy, D.L., Nakayama, K., Matthysse, S., Palafox, G., Holzman, P.S., 1999a. Dependence of impaired eye tracking on deficient velocity discrimination in schizophrenia. Arch. Gen. Psychiatry 56, 155 – 161. Chen, Y., Nakayama, K., Levy, D.L., Matthysse, S., Holzman, P.S., 1999b. Psychophysical isolation of a motion-processing deficit in schizophrenics and their relatives and its association with impaired smooth pursuit. Proc. Natl. Acad. Sci. U. S. A. 96, 4724 – 4729. Chen, Y., Palafox, G.P., Nakayama, K., Levy, D.L., Matthysse, S., Holzman, P.S., 1999c. Motion perception in schizophrenia. Arch. Gen. Psychiatry 56, 149 – 154. Doniger, G.M., Foxe, J.J., Schroeder, C.E., Murray, M.M., Higgins, B.A., Javitt, D.C., 2001. Visual perceptual learning in human object recognition areas: a repetition priming study using highdensity electrical mapping. NeuroImage 13, 305 – 313. Doniger, G.M., Foxe, J.J., Murray, M.M., Higgins, B.A., Javitt, D.C., 2002. Impaired visual object recognition and dorsal/ventral stream interaction in schizophrenia. Arch. Gen. Psychiatry 59, 1011 – 1020. Felleman, D., Van Essen, D., 1991. Distributed hierarchical processing in the primate cerebral cortex. Cereb. Cortex 1, 1 – 47. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1997. Structured Clinical Interview for DSM-IV Axis I Disorders— Patient Edition. New York State Psychiatric Institute, New York. Foxe, J.J., Doniger, G.M., Javitt, D.C., 2001. Early visual processing deficits in schizophrenia: impaired PI generation revealed by high-density electrical mapping. NeuroReport 12. Green, M., 1984. Masking by light and the sustained-transient dichotomy. Percept. Psychophys. 35, 519 – 535. Green, M.F., Nuechterlein, K.H., 1999. Backward masking performance as an indicator of vulnerability to schizophrenia. Acta Psychiatr. Scand., Suppl. 395, 34 – 40. Green, M., Walker, E., 1984. Susceptibility to backward masking in schizophrenic patients with positive or negative symptoms. Am. J. Psychiatry 141, 1273 – 1275. Green, M.F., Nuechterlein, K.H., Mintz, J., 1994a. Backward masking in schizophrenia and mania: I. Specifying a mechanism. Arch. Gen. Psychiatry 51, 939 – 944. Green, M.F., Nuechterlein, K.H., Mintz, J., 1994b. Backward masking in schizophrenia and mania: II. Specifying the visual channels. Arch. Gen. Psychiatry 51, 945 – 951. Green, M.F., Nuechterlein, K.H., Breitmeyer, B., Mintz, J., 1999. Backward masking in unmedicated schizophrenic patients in psychotic remission: possible reflection of aberrant cortical oscillation. Am. J. Psychiatry 156, 1367 – 1373. Greenstein, V.C., Seliger, S., Zemon, V., Ritch, R., 1998. Visual evoked potential assessment of the effects of glaucoma on visual subsystems. Vis. Res. 38, 1901 – 1911. Harvey, P.D., Keefe, R.S., Moskowitz, J., Putnam, K.M., Mohs, R.C., Davis, K.L., 1990. Attentional markers of vulnerability to schizophrenia: performance of medicated and unmedicated patients and normals. Psychiatry Res. 33, 179 – 188.

I. Schechter et al. / Schizophrenia Research 64 (2003) 91–101 Kaplan, E., 1991. The Receptive Field Structure of Retinal Ganglion Cells in Cat and Monkey. CRC Press, Boston. Kay, S., Opler, L., Fiszbein, A., 1992. The Positive and Negative Syndrome Scale (PANSS) Manual. Multi-Health Systems, Toronto. Keesey, U.T., 1972. Flicker and pattern detection: a comparison of thresholds. J. Opt. Soc. Am. 62, 446 – 448. Keri, S., Szendi, I., Kelemen, O., Benedek, G., Janka, Z., 2001. Remitted schizophrenia-spectrum patients with spared working memory show information processing abnormalities. Eur. Arch. Psychiatry Clin. Neurosci. 251, 60 – 65. Keri, S., Antal, A., Szekeres, G., Benedek, G., Janka, Z., 2002. Spatiotemporal visual processing in schizophrenia. J. Neuropsychiatry Clin. Neurosci. 14, 190 – 196. Kulikowski, J.J., Tolhurst, D.J., 1973. Psychophysical evidence for sustained and transient detectors in human vision. J. Physiol. 232, 149 – 162. Lamme, V.A.F., Roelfsema, P.R., 2000. The distinct modes of vision offered by feedforward and recurrent processing. Trends Neurosci. 23, 571 – 579. Lindenmayer, J.P., Bernstein-Hyman, R., Grochowski, S., 1994. A new five factor model of schizophrenia. Psychiatr. Q. 65, 299 – 322. Masson, G., Mestre, D., Blin, O., 1993. Dopaminergic modulation of visual sensitivity in man. Fundam. Clin. Pharmacol. 7, 449 – 463. McClure, R.K., 2001. The visual backward masking deficit in schizophrenia. Prog. Neuro-psychopharmacol. Biol. Psychiatry 25, 301 – 311. Merigan, W.H., Maunsell, J.H.R., 1993. How parallel are the primate visual pathways? In: Cowan, W.M., Shooter, E.M., Stevens, C.F., Thompson, R.F. (Eds.), Ann. Rev. Neuroscience. Annual Reviews, Palo Alto, CA, pp. 369 – 402. Merritt, R.D., Balogh, D.W., 1989. Backward masking spatial frequency effects among hypothetically schizotypal individuals. Schizophr. Bull. 15, 573 – 583. Murrary, M.M., Foxe, J.J., Higgins, B.A., Javitt, D.C., Schroeder, C.E., 2001. Visuo-spatial neural response interactions in early cortical processing during a simple reaction time task: a high-density electrical mapping study. Neuropsychologia 39, 828 – 844. O’Donnell, B.F., Swearer, J.M., Smith, L.T., Nestor, P.G., Shenton, M.E., McCarley, R.W., 1996. Selective deficits in visual perception and recognition in schizophrenia. Am. J. Psychiatry 153, 687 – 692. Rund, B.R., 1993. Backward-masking performance in chronic and nonchronic schizophrenics, affectively disturbed patients, and normal control subjects. J. Abnorm. Psychology 102, 74 – 81. Saccuzzo, D.P., Braff, D.L., 1981. Early information processing deficit in schizophrenia. New findings using schizophrenic subgroups and manic control subjects. Arch. Gen. Psychiatry 38, 175 – 179. Saccuzzo, D.P., Braff, D.L., 1986. Information-processing abnor-

101

malities: trait- and state-dependent components. Schizophr. Bull. 12, 447 – 459. Schroeder, C.E., Mehta, A.D., Givre, S.J., 1998. A spatiotemporal profile of visual system activation revealed by current source density analysis in the awake macaque. Cereb. Cortex 8, 575 – 592. Schuck, J.R., Lee, R.G., 1989. Backward masking, information processing, and schizophrenia. Schizophr. Bull. 15, 491 – 500. Schwartz, B.D., Winstead, D.K., 1985. Icon formation in chronic schizophrenics. Biol. Psychiatry 20, 1015 – 1018. Schwartz, B.D., McGinn, T., Winstead, D.K., 1987. Disordered spatiotemporal processing in schizophrenics. Biol. Psychiatry 22, 688 – 698. Schwartz, B.D., Maron, B.A., Evans, W.J., Winstead, D.K., 1999. High velocity transient visual processing deficits diminish ability of patients with schizophrenia to recognize objects. Neuropsychiatry Neuropsychol. Behav. Neurol. 12, 170 – 177. Simpson, G.V., Foxe, J.J., Vaughan Jr., H.G., Mehta, A.D., Schroeder, C.E., 1995. Integration of electrophysiological source analyses, MRI and animal models in the study of visual processing and attention. EEG Clin. Neurophysiol. 44 (Suppl.), 76 – 92. Slaghuis, W.L., 1998. Contrast sensitivity for stationary and drifting spatial frequency gratings in positive- and negative-symptom schizophrenia. J. Abnorm. Psychology 107, 49 – 62. Slaghuis, W.L., Bakker, V.J., 1995. Forward and backward visual masking of contour by light in positive- and negative-symptom schizophrenia. J. Abnorm. Psychology 104, 41 – 54. Slaghuis, W.L., Curran, C.E., 1999. Spatial frequency masking in positive- and negative-symptom schizophrenia. J. Abnorm. Psychology 108, 42 – 50. Stuve, T.A., Friedman, L., Jesberger, J.A., Gilmore, G.C., Strauss, M.E., Meltzer, H.Y., 1997. The relationship between smooth pursuit performance, motion perception and sustained visual attention in patients with schizophrenia and normal controls. Psychol. Med. 27, 143 – 152. Tolhurst, D.J., 1973. Separate channels for the analysis of shape and movement of a moving stimulus. J. Physiol. 231, 385 – 402. Tootell, R.B., Hamilton, S.L., Switkes, E., 1988. Functional anatomy of macaque striate cortex: IV. Contrast and magno-parvo streams. J. Neurosci. 8, 1594 – 1609. Weiner, R.U., Opler, L.A., Kay, S.R., Merriam, A.E., Papouchis, N., 1990. Visual information processing in positive, mixed, and negative schizophrenic syndromes. J. of Nerv. Ment. Dis. 178, 616 – 626. Wetherill, G.B., Levitt, H., 1965. Sequential estimation of points on a psychometric function. Br. J. Math. Stat. Psychol. 18, 1 – 10. Woldorff, M.G., Fox, P.T., Matzke, M., et al., 1997. Retinotopic organization of early visual spatial attention effects as revealed by PET and ERPs. Hum. Brain Mapp. 5, 280 – 286. Zemon, V., Siegfried, J., Gordon, J., 1991. Magno and parvo pathways in humans studied using VEPs to luminance and chromatic contrast (abstract). Invest. Opthalmol. Vis. Sci. Suppl. 32, 1033.