Schizophrenia Research 56 92002) 171±185 www.elsevier.com/locate/schres

Positive and negative schizotypy in a student sample: neurocognitive and clinical correlates

Wayne M. Dinna,*, Catherine L. Harrisa, Ayse Aycicegib,1, Paul Greenea, Margaret S. Andovera

aDepartment of Psychology, Boston University, Boston, MA, USA bIstanbul University, Istanbul, Turkey Received 15 November 2000; revised 15 March 2001; accepted 26 March 2001

Abstract Positive and negative schizotypy may represent discrete factors or dimensions. To determine if distinct neurocognitive pro®les are associated with these dimensions or factors, we classi®ed university students on the basis of positive and negative schizotypal symptoms and conducted separate analyses. Following prior work in the neuropsychiatric literature, we predicted that subtle prefrontal de®cits would be selectively associated with negative schizotypal personality features in a nonclinical student sample. We also investigated the relationship between positive/negative schizotypy and associated clinical states or personality dimensions including antisocial personality disorder, obsessive±compulsive personality traits, generalized and social anxiety, empathy, and impulsivity. Classi®cation of subjects into positive and negative schizotypy groups revealed distinct neurocognitive and clinical pro®les. We observed a positive relation between measures of temporolimbic dysfunction, impulsivity, antisocial behavior, and positive schizotypal phenomena. Negative schizotypy was associated with subtle perfor- mance de®cits on measures of frontal executive function, increased social anxiety, and obsessive±compulsive phenomena. Findings are consistent with the contention that positive and negative schizotypy represent discrete factors. q 2002 Elsevier Science B.V. All rights reserved.

Keywords: Schizotypal personality; Prefrontal; Temporolimbic; Executive function; Neuropsychological; Positive and negative schizotypy

1. Introduction eccentric behavior, lack of close friends, and exces- sive social anxiety 9American Psychiatric Associa- Schizotypal personality disorder 9SPD) has been tion, 1994). conceptualized as an attenuated form or phenotypic Subtle prefrontal de®cits are associated with variant of schizophrenia. Diagnostic criteria for SPD schizotypal personality. Individuals receiving a SPD include ideas of reference, magical ideation, unusual diagnosis demonstrate a greater degree of impairment perceptual experiences, odd thinking and speech, on tests assessing frontal executive function including suspiciousness, inappropriate or constricted affect, the Wisconsin Card Sorting Test 9Diforio et al., 2000; Trestman et al., 1995; Voglmaier et al., 1997), Cali- fornia Verbal Learning Test 9Bergman et al., 1998; * Corresponding author. Address: 42 Washington Terrace, Voglmaier et al., 1997), and the Trail-Making Test Whitman, MA 02382, USA. Tel.: 11-781-447-6058. E-mail address: [email protected] 9W.M. Dinn). 9Trestman et al., 1995). Schizotypic subjects also 1 Visiting Scholar Ð Boston University. exhibit performance de®cits on visuospatial working

0920-9964/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S0920-9964901)00230-4 172 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 memory tasks 9Farmer et al., 2000; Park and McTi- symptoms including magical ideation, ideas of refer- gue, 1997; Roitman et al., 2000). Moreover, schizo- ence, and unusual perceptual experiences. Several typic subjects manifest volumetric abnormalities 9i.e. researchers suggest that positive and negative schizo- volume reduction) in temporal and prefrontal regions typy represent distinct factors or dimensions. Siever 9Dickey et al., 1999; Buchsbaum et al., 1997; Raine et 91995) noted that these dimensions or factors are inde- al., 1992a). pendently heritable and speculated that distinct patho- University students psychometrically de®ned as physiologies underlie each dimension. schizotypic or `psychosis-prone' also exhibit perfor- To determine if distinct neurocognitive pro®les are mance de®cits on tasks assessing frontal executive associated with these factors, we classi®ed subjects on function including the Wisconsin Card Sorting Test the basis of positive and negative schizotypal symp- 9Lenzenweger and Kor®ne, 1994; Lyons et al., 1991; toms and conducted separate analyses of neurocogni- Poreh et al., 1995; Suhr, 1997), Trail-Making Test tive test performance. 9Poreh et al., 1995), and the Booklet Category Test Negative schizotypy may be selectively associated 9Poreh et al., 1995). with executive dysfunction. Diforio et al. 92000) reported that adolescents with SPD demonstrated 1.1. Con¯icting ®ndings performance de®cits on the modi®ed WCST. More- over, impaired performance on the WCST was asso- Schizotypy is not invariably associated with ciated with negative, rather than positive, symptoms. impaired performance on tasks assessing frontal As noted previously, Lenzenweger and Gold 92000) executive function. Lenzenweger and Gold 92000) found that students psychometrically identi®ed as found that students psychometrically identi®ed as schizotypic did not exhibit performance de®cits on schizotypic did not demonstrate performance de®cits working memory tasks relative to controls. It is on verbal and auditory working memory tasks in important to note that Lenzenweger and Gold comparison to controls. How can we account for employed the Perceptual Aberration Scale to classify these con¯icting results? subjects. The Perceptual Aberration Scale taps posi- One possibility is that students psychometrically tive schizotypal symptoms 9e.g. perceptual distor- de®ned as schizotypic present with a less severe tions). However, Lyons et al. 91991) recruited form of schizotypal personality relative to subjects schizotypal subjects via newspaper advertisements diagnosed with SPD. However, several studies which highlighted positive symptoms 9i.e. advertise- found that subjects diagnosed with SPD did not ments seeking individuals who had experienced para- demonstrate performance de®cits on tests assessing normal phenomena such as ESP and telepathy) and frontal executive functioning 9i.e. the WCST) 9Raine found that individuals with positive symptom schizo- et al., 1992b; Condray and Steinhauer, 1992). typy demonstrated performance decrements on the A second possibility is that neurocognitive de®cits WCST relative to control subjects. are selectively associated with negative, rather than The present study uses a student sample to investi- positive, symptoms. A discussion of the two-factor gate whether subtle prefrontal de®cits are selectively model 9or two-syndrome construct) of schizotypy associated with negative schizotypal personality may clarify this issue. A two-factor model of schizo- features. Following prior work in the neuropsychiatric typy is based on Crow's two-syndrome concept of literature, we predicted that negative schizotypy schizophrenia 9i.e. Type I Ð positive symptoms and would be associated with impaired performance on Type II Ð negative symptoms) 9Crow, 1980, 1985). tests assessing frontal executive function. Variations in dopaminergic activity are associated with both positive and negative schizotypy 9Siever, 1.2. Patterns of comorbidity 1995). Hypodopaminergia in prefrontal cortex is asso- ciated with negative symptoms such as affective ¯at- We also investigated the relationship between posi- tening, avolition and apathy, asociality, and cognitive tive/negative schizotypy and associated clinical states impairment. Hyperdopaminergia in subcortical meso- or personality dimensions including antisocial limbic structures may generate positive schizotypal personality disorder 9APD), obsessive±compulsive W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 173 personality traits, generalized and social anxiety, de®cits on tests of frontal executive functioning). empathy, and impulsivity. In prior work, we observed OCPTs may represent adaptive or compensatory stra- an association between schizotypal personality tegies which develop in response to executive func- features, obsessive±compulsive personality traits, tion de®cits. That is, OC symptoms may represent and excessive social anxiety. Are these clinical states adaptations to subtle prefrontal de®cits associated selectively associated with positive or negative with negative schizotypy. In the present study, we schizotypy? predict that students obtaining high scores on a measure of negative schizotypal symptoms will also 1.2.1. Antisocial personality, impulsivity, empathy, achieve clinically signi®cant scores on a measure of and schizotypy obsessive±compulsive personality. Meehl 91989) suggested that a signi®cant subset of However, in prior work, we also observed a strong psychopathic subjects are schizotaxic. Siever et al. relation between social anxiety and obsessive± 91990) found that approximately 20% of SPD patients compulsive phenomena. Thus, OC symptoms, such also met diagnostic criteria for APD. We predicted as compulsive rituals and checking behaviors, may that individuals psychometrically de®ned as schizoty- represent compensatory mechanisms for reducing pic would achieve signi®cantly higher scores on a anxiety. Given that both OCD and SPD are associated measure of antisocial personality relative to compar- with social anxiety, it is important to understand the ison subjects. Although schizotypy may be associated relation between schizotypy and anxiety states, and with antisocial personality, it has not been determined separately examine social vs. generalized anxiety. whether antisocial behavior is selectively associated Indeed, excessive social anxiety is a characteristic with positive or negative schizotypal personality clinical feature of SPD 9American Psychiatric features and we make no a priori predictions. Association, 1994). In prior unpublished work, we observed an associa- tion between measures of impulsivity, empathy, and 1.3. Research goals schizotypal personality. We found that subjects psychometrically de®ned as schizotypic obtained Our principal research objectives were: signi®cantly higher scores on measures of impulsivity and signi®cantly lower scores on measures of empa- 1. an exploration of the relationship between positive thy relative to matched controls. In the present study, and negative schizotypy, and neuropsychological we included self-report measures of impulsivity and test performance in a student sample; empathy. 2. an investigation of the relation between schizotypy and associated clinical states or personality dimen- 1.2.2. Obsessive±compulsive personality and sions including antisocial behavior, obsessive± schizotypy compulsive phenomena, generalized and social In prior work, we observed an association between anxiety, empathy, and impulsivity. schizotypal personality features and obsessive± compulsive personality traits 9OCPTs). Moreover, both OCPTs and schizotypal personality features 2. Method were associated with performance de®cits on tests assessing frontal executive function 9Aycicegi et al., 2.1. Subjects 2001; Dinn et al., 2001a,b). We suggested that the preoccupation with rules and organization, perfection- One hundred and three undergraduate students ism, and in¯exibility displayed by subjects exhibiting served as participants for this study. Participants OCPTs may represent behavioral strategies which were drawn from introductory psychology courses evolve in response to executive function de®cits. and received course credit. Why would OCPD and schizotypal personality The sample comprised 75 female and 28 male covary? As noted previously, negative symptoms are students. Female participants had a mean age of associated with cognitive de®cits 9i.e. performance 18.6 years SD ˆ 1:0† and had completed 13.4 years 174 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185

Table 1 Positive Schizotypy: Clinical and Neurocognitive Test Findings: mean 9SD) High-, median-, and low-positive schizotypy

High Median Low F 2; 100† pfpa n 17 60 26 Age 18.5 90.9) 18.9 91.3) 19.1 91.3) 1.09 0.338 ± Education 13.4 90.9) 13.6 90.9) 13.7 91.1) 0.50 0.602 ± Gender 9Female/Male) 11/6 47/13 17/9 Handedness 9Right/Left) 17/0 55/5 21/5 Positive Symptoms 6.3 90.4) 3.9 90.8) 1.3 90.7) 231.33 0.0001 Negative Symptoms 3.7 92.0) 3.0 92.3) 2.8 92.1) 0.77 0.464 ± Limbic System Checklist-33 LSCL-Total 38.9 915.9) 25.0 911.5) 18.1 911.3) 14.77 0.0001 0.54 Somatic 10.8 95.5) 9.3 93.9) 7.8 94.8) 2.30 0.104 0.21 Hallucinatory 9.3 96.1) 5.9 94.2) 3.1 92.8) 10.35 0.0001 0.45 Automatisms 9.0 95.2) 4.7 93.5) 3.7 93.0) 11.08 0.0001 0.47 Dissociative 9.7 93.9) 5.1 93.1) 3.3 92.7) 21.24 0.0001 0.65 I 7 Questionnaire Empathy 15.9 92.2) 14.9 92.6) 13.2 93.2) 5.32 0.007 0.32 Venturesomeness 9.4 93.6) 9.1 93.1) 7.8 93.6) 1.58 0.210 0.17 Impulsivity 9.0 93.9) 7.4 93.8) 3.9 93.2) 11.61 0.0001 0.48

Fear Survey Subscales Social Anxiety 16.7 97.5) 17.6 95.7) 16.3 96.0) 0.49 0.611 ± Speci®c Phobia 23.1 98.2) 21.9 98.9) 22.6 97.9) 0.15 0.855 ± Personality Diagnostic Questionnaire $PDQ-4) OCPD 4.4 91.5) 3.8 91.3) 3.2 91.1) 4.04 0.0204 0.28 APD 2.3 90.9) 1.3 91.2) 0.8 91.0) 8.17 0.0005 0.40 FAS Test 40.1 97.6) 35.9 97.8) 37.6 98.4) 1.95 0.147 ± Divergent Thinking Task 6.0 92.4) 6.2 92.1) 6.8 94.4) 0.43 0.648 ± Porteus Maze Task 37.7s 920.8) 37.9 921.6) 39.2s 923.0) 0.03 0.963 ± Trail-Making Test Trail-Making 9A) 28.3s 97.0) 29.7s 98.7) 31.0s 911.8) 0.44 0.642 ± Trail-Making 9B) 50.2s 99.3) 50.0s 913.5) 52.4s 912.8) 0.31 0.728 ± Stroop Color±Word Test StroopWord-c 464.1ms 964) 513.7ms 9103) 532.2ms 976) 2.93 0.057 0.24 StroopWord-i 474.9ms 975) 517.0ms 9102) 543.1ms 993) 2.57 0.080 0.22 StroopColor-c 660.1ms 994) 694.4ms 998) 683.4ms 9114) 0.75 0.472 ± StroopColor-i 785.2ms 9114) 777.5ms 9112) 782.0ms 9122) 0.03 0.965 ± Rey±Osterrieth Complex Figure Test Copy Organization 4.7 91.5) 4.2 91.8) 4.6 91.6) 0.70 0.494 ± Recall Accuracy 21.7 95.7) 22.7 95.3) 24.6 94.8) 1.86 0.159 0.19

Frontal Lobe Personality Scale-PV Disinhibition 34.8 95.0) 31.5 95.8) 27.2 95.7) 9.84 0.0001 0.44 Exec. Dysfunction 40.5 97.4) 38.3 97.5) 35.0 96.8) 3.19 0.0450 0.25 Apathy 25.8 96.6) 25.6 95.5) 24.2 96.0) 0.56 0.568 ±

a fp ˆ Effect Size 9Cohen's f). Note. Personality Diagnostic Questionnaire 9PDQ-4)-OCPD ˆ Obsessive±compulsive Personality Disorder Subscale; APD ˆ Antisocial Personality Disorder Subscale; Stroop Color±Word Test 9blocks 1 and 2 ˆ word naming, blocks 3 and 4 ˆ color naming), c ˆ congruent, i ˆ incongruent; FAS Test ˆ Controlled Word Fluency Test. W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 175 of education SD ˆ 0:7†: Sixty-seven female subjects Color±Word Test. Participants were instructed to were right-handed and eight were left-handed, as deter- read words describing colors as rapidly as possible. mined by self-report. Mean age of male participants was Words were displayed one at a time on the computer 19.6 years SD ˆ 1:7† and their mean educational level monitor, with the ink color being either consistent or was 14.1 years SD ˆ 1:1†: Twenty-six male subjects inconsistent with the given word. During the ®rst were right-handed and two were left-handed, as deter- block 9non-con¯ict block) the subject was asked to mined by self-report. Written informed consent was read the word displayed on the monitor and ignore obtained from all participants. Demographic data are the ink color 940 trials). During the second block presented in Tables 1 and 2. 9con¯ict block) the participant was asked to identify the ink color 940 trials). Response latencies were 2.2. Procedure recorded using a voice-activated millisecond timer. The dependent measure was response time. We administered a neuropsychological test battery consisting of measures sensitive to prefrontal 2.3.2. Porteus maze task 2Porteus, 1955) dysfunction, and a battery of personality question- Participants were required to ®nd the exit route naires and clinical scales. The neurocognitive test from a relatively complex maze. Ef®cient perfor- battery consisted of a computer version of the Stroop mance requires planning and anticipation of blocked Color±Word Test, Controlled Word Fluency Test routes. Time to completion was the dependent 9FAS Test) 9Goodglass and Kaplan, 1972), Porteus measure. Maze Task 9Porteus, 1955), Trail-Making Test 9Parts A and B) 9Reitan and Wolfson, 1985), Diver- 2.3.3. Word Fluency Test 2FAS test) 2Goodglass and gent Thinking Task 9based on Guilford and Hoepfner, Kaplan, 1972) 1971), and the Rey Complex Figure Test 9Lezak, During the Word Fluency Test 9FAS Test) the 1995). Personality questionnaires and clinical scales subject was asked to write down as many words as included: the I 7 Questionnaire 9Eysenck et al., 1985), possible that begin with a speci®c letter 9F, A, or S) the Frontal Lobe Personality Scale 9FLPS-Patient during three one-minute trials. Total number of words Version) 9Grace and Malloy, 1992), the Schizotypal produced was the dependent measure. Personality Questionnaire-B 9SPQ-B) 9Raine and Benishay, 1995), Limbic System Checklist-33 2.3.4. Divergent Thinking Task 2based on Guilford 9LSCL-33) 9Teicher et al., 1993), the Personality and Hoepfner, 1971) Diagnostic Questionnaire 9PDQ-4) 9Hyler, 1994), During this task subjects were asked to name as and a modi®ed version of the Fear Survey Schedule many different uses of a newspaper as possible during 9Wolpe and Lang, 1964). a one-minute trial. They were provided with the All computerized tasks were administered on a following example: one use is rolling up the news- Macintosh IIci. Reaction time and voice onset laten- paper to swat a mosquito. Number of alternate uses cies were collected using a millisecond timer that was the dependent measure. interfaces with PsyScope; experimental design soft- ware developed by Cohen et al. 91993). Neurocogni- 2.3.5. Trail-Making Test 2Parts A and B) 2Reitan and tive tests were administered by trained research Wolfson, 1985) technicians who followed a standardized testing During Part A of the Trail-Making Test subjects protocol. Technicians were blind to the group status were instructed to connect 25 numbered circles 91± of participants 9i.e. SPQ-B scores). Personality ques- 25) randomly distributed over an 8 £ 11 sheet of tionnaires and clinical scales were administered and paper. Subjects were instructed to connect circles as scored by computer. rapidly as possible. During Part B subjects were 2.3. Prefrontal measures required to connect 25 circles which contain numbers 91±13) or letters 9A±L) and must sequentially alter- 2.3.1. Stroop Color±Word Test nate between numbers and letters 9that is, 1±A±2±B± We developed a computer version of the Stroop 3±C, and so forth). Subjects received feedback when 176 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185

Table 2 Negative Schizotypy: Clinical and Neurocognitive Test Findings: mean 9SD) High-, median-, and low-positive schizotypy

High Median Low F 2; 100† pfpa n 17 57 29 Age 19.1 91.5) 18.8 91.3) 18.9 91.2) 0.37 0.688 ± Education 13.7 90.9) 13.5 90.9) 13.6 91.0) 0.09 0.908 ± Gender 9Female/Male) 12/5 39/18 24/5 Handedness 9Right/Left) 16/1 50/7 27/2 Positive Symptoms 4.1 91.5) 3.7 91.9) 3.5 91.6) 0.61 0.542 ± Negative Symptoms 6.5 90.6) 3.4 91.1) 0.5 90.5) 222.62 0.0001 Limbic System Checklist-33 LSCL-Total 26.8 913.9) 26.5 915.2) 23.1 910.7) 0.66 0.516 ± Somatic 9.2 96.0) 9.4 94.3) 8.6 94.0) 0.27 0.760 ± Hallucinatory 5.4 93.5) 6.1 95.4) 5.5 93.9) 0.22 0.801 ± Automatisms 5.7 94.2) 5.4 94.4) 4.3 93.2) 0.84 0.432 ± Dissociative 6.5 93.7) 5.5 94.2) 4.6 92.6) 1.42 0.246 ± I 7 Questionnaire Empathy 14.7 92.5) 14.5 93.0) 14.9 92.7) 0.20 0.814 ± Venturesomeness 8.3 93.1) 8.9 93.6) 9.1 93.1) 0.28 0.750 ± Impulsivity 6.3 94.2) 6.8 93.9) 6.9 94.3) 0.12 0.881 ±

Fear Survey Subscales Social Anxiety 22.0 94.6) 16.5 96.2) 15.4 95.3) 7.42 0.001 0.38 Speci®c Phobia 24.3 97.5) 20.6 98.6) 24.4 98.2) 2.65 0.075 ± Personality Diagnostic Questionnaire $PDQ-4) OCPD 4.5 91.4) 3.6 91.2) 3.7 91.5) 3.01 0.05 0.24 APD 1.2 91.3) 1.4 91.2) 1.2 91.1) 0.46 0.631 ± FAS Test 34.7 99.4) 37.7 97.5) 37.1 98.0) 0.91 0.404 0.13 Divergent Thinking Task 5.5 92.8) 6.0 92.1) 7.6 93.8) 4.12 0.02 0.28 Porteus Maze Task 41.7s 919.1) 34.3s 917.4) 43.9s 928.6) 2.21 0.114 ± Trail-Making Test Trail-Making 9A) 31.4s 99.8) 31.1s 99.5) 26.2s 97.6) 3.19 0.05 0.25 Trail-Making 9B) 52.8s 99.9) 52.7s 913.6) 45.5s 911.1) 3.51 0.04 0.26 Stroop Color±Word Test StroopWord-c 534.1ms 9105) 503.9ms 989) 508.5ms 995) 0.68 0.507 ± StroopWord-i 552.4ms 9118) 506.8ms 985) 514.9ms 9105) 1.44 0.241 ± StroopColor-c 696.1ms 994) 672.4ms 9107) 706.6ms 993) 1.18 0.310 ± StroopColor-i 778.2ms 9104) 779.0ms 9117) 782.8ms 9118) 0.01 0.986 ± Rey±Osterrieth Complex Figure Test Copy Organization 3.7 91.8) 4.6 91.6) 4.4 91.7) 2.05 0.134 0.20 Recall Accuracy 22.1 96.3) 23.0 95.8) 23.4 93.3) 0.33 0.715 ±

Frontal Lobe Personality Scale-PV Disinhibition 32.0 95.8) 30.4 96.3) 31.5 95.9) 0.60 0.549 ± Exec. Dysfunction 42.7 98.7) 36.7 96.3) 37.3 98.0) 4.63 0.0110 0.30 Apathy 32.0 94.1) 24.6 95.3) 22.7 94.7) 19.59 0.0001 0.62

a fp ˆ Effect size 9Cohen's f). Note. Personality Diagnostic Questionnaire 9PDQ-4)-OCPD ˆ Obsessive±compulsive Personality Disorder Subscale; APD ˆ Antisocial Personality Disorder Subscale; Stroop Color±Word Test 9blocks 1 and 2 ˆ word naming, blocks 3 and 4 ˆ color naming), c ˆ congruent, i ˆ incongruent; FAS Test ˆ Controlled Word Fluency Test. W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 177 circles were connected out of order. Time to including: 91) behavioral disinhibition; 92) executive completion was the dependent measure. function de®cits; and 93) apathy, re¯ecting orbitofron- tal, dorsolateral±prefrontal, and mesial±prefrontal/ 2.3.6. Rey±Osterrieth Complex Figure Test 2ROCT) anterior cingulate dysfunction, respectively. 2Lezak, 1995) Participants copied a complex geometric ®gure 2.4.3. Personality Diagnostic Questionnaire 2PDQ-4) consisting of 18 separate elements. Subjects were 2Hyler, 1994) required to reproduce the ®gure from memory after The PDQ-4 is a true±false, self-report instrument. a 1-minute delay. We employed the 36-point itemized Items re¯ect DSM-IV diagnostic criteria for person- scoring system described by Lezak 91995) to evaluate ality disorders. Questions were adapted from DSM-IV constructional accuracy. We also used the test proce- diagnostic criteria. The APD and OCPD Subscales dure and scoring approach described by Savage and were administered to participants. colleagues 9Savage et al., 1999) to assess construc- tional organization during the copy condition 96-point scoring system). 2.4.4. Fear Survey Schedule-Modi®ed Version 2FSS-MV) 2based onWolpe and Lang, 1964) 2.4. Clinical scales and personality questionnaires We administered a modi®ed version of the Fear Survey Schedule. The FSS-MV is a self-report index 2.4.1. Schizotypal Personality Questionnaire-B of anxiety and avoidance behavior. Participants were 2SPQ-B) 2Raine and Benishay, 1995) asked to indicate the degree of anxiety and avoidance The SPQ-B is a 22-item, forced-choice question- behavior associated with speci®c social situations or naire. Scores range from 0 to 22. The SPQ-B is a when exposed to speci®c stimuli. The FSS-MV yields brief, self-report screening instrument used to evalu- a total score and two subscale scores. During our ate respondents for the presence of schizotypal modi®ed version of the Fear Survey participants personality features. A total score and three subscale were instructed to indicate the degree of avoidance scores were obtained. behavior associated with speci®c situations or stimuli The SPQ-B yields three subscale scores re¯ecting: presented on the computer screen because of fear or anxiety. Stimuli were classi®ed in the following 1. cognitive or perceptual distortions 9e.g. ªHave you manner: 91) social situations; and 92) speci®c phobic ever had the sense that some person or force is situations. around you, even though you cannot see anyone?º); 2. interpersonal dif®culties 9e.g. ªDo you feel that you 2.4.5. I-7 Questionnaire 2Eysenck et al., 1985) are unable to get `close' to people?º); and The I 7 Questionnaire is a forced-choice instru- 3. disorganization 9e.g. ªI sometimes use words in ment. Respondents were asked to indicate whether unusual ways.º) 9Raine and Benishay, 1995, p. they agree or disagree with a series of statements 351). related to three personality dimensions 9impulsive- ness, venturesomeness, and empathy). Items correspond to DSM-IV diagnostic criteria for SPD. The SPQ-B is a psychometrically sound instru- 2.4.6. Limbic System Checklist-33 2LSCL-33) 2Teicher ment which compares favorably to established et al., 1993) measures. The Limbic System Checklist is a 33-item symp- tom inventory. Respondents were instructed to indi- 2.4.2. Frontal Lobe Personality Scale 2FLPS-patient cate how frequently they experience symptoms version) 2Grace and Malloy, 1992) associated with temporolimbic dysfunction including Respondents were instructed to indicate how `paroxysmal somatic disturbances, brief hallucinatory frequently they experience symptoms or exhibit events, visual disturbances, automatisms, and disso- behaviors associated with frontal lobe syndromes ciative disturbances' 9Teicher et al., 1993, p. 302). 178 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 2.5. Do neuropsychological tests possess localizing pating in the Stroop Color±Word Test demonstrated value? right orbitofrontal activation as well as increased activity in bilateral parietal structures 9Bench et al., Converging lines of evidence including functional 1993). During a second experiment, Bench et al. neuroimaging research and human lesion studies 91993) documented right frontal polar and right ante- suggest that the tasks employed in our protocol are rior cingulate activation during Stroop task perfor- sensitive measures of prefrontal dysfunction. Of mance. Ef®cient performance on the Stroop task course, we must proceed cautiously when we argue requires sustained attention and impulse control. that variations in brain function correspond to patterns Therefore, the Stroop Color±Word Test should be of neuropsychological impairment and clinical considered a broadly frontal task. Nevertheless, orbi- presentation. Nevertheless, several lines of evidence tofrontal or ventral prefrontal systems may be impli- suggest that many of the experimental tasks employed cated when presenting symptoms include impaired in this study possess localizing value. Researchers performance on neurocognitive tasks which require have come to appreciate that the prefrontal region is the subject to suppress a prepotent response pattern. not a unitary structure; rather, it is fractionable into In summary, the imaging and human lesion litera- anatomically and functionally distinct subsystems. ture is broadly consistent with the notion that the Converging lines of evidence suggests that dorsolat- prefrontal region is fractionable into anatomically eral±prefrontal cortex mediates executive functions and functionally distinct subsystems and that speci®c 9Fletcher et al., 1998; Smith and Jonides, 1999), neurocognitive tasks may tap speci®c prefrontal while orbitofrontal cortex modulates sensitivity to subsystems. However, our ®ndings must be inter- reinforcement contingencies 9Rolls, 1995) and plays preted with caution. It is important to bear in mind a major role in behavioral inhibition. Performance that neurocognitive tests are only indirect measures of de®cits on tests assessing executive functions may neurophysiological function and the localizing value re¯ect dorsolateral± or mesial±prefrontal dysfunc- of such tasks is uncertain. tion. Impaired performance on verbal ¯uency and divergent thinking tasks may re¯ect dorsolateral± 2.6. Data analysis plan prefrontal dysfunction. Functional neuroimaging studies revealed signi®cant ¯ow augmentation and As noted previously, several researchers suggest increased activity in dorsolateral±prefrontal cortex that positive and negative schizotypy represent during verbal ¯uency tasks 9Cantor-Graae et al., distinct factors or dimensions. To determine if distinct 1993; Warkentin et al., 1991) and classical tests of neurocognitive pro®les are associated with these frontal executive function. An extensive body of dimensions or factors, we classi®ed subjects on the evidence suggests that DLPF cortex mediates execu- basis of positive and negative schizotypy symptoms tive functions 9e.g. planning, organization, and keep- and conducted separate analyses. Students were ing in mind diverse future consequences) 9Fletcher et assigned to low-, median-, and high-positive schizo- al., 1998; Smith and Jonides, 1999). Moreover, typy groups on the basis of scores on the Cognitive± patients with damage to the dorsolateral aspect of Perceptual Subscale 9Factor 1) from the SPQ-B. the prefrontal region display performance de®cits on Students were assigned to low-, median-, and high- the classical tests of frontal executive functioning 9e.g. negative schizotypy groups on the basis of perfor- the WCST). Ef®cient performance on the Trail- mance on the Interpersonal Subscale 9Factor 2) from Making Test 9Part B), Porteus Maze, and on the the SPQ-B, a measure of negative schizotypal symp- Rey±Osterrieth Complex Figure Test requires the toms. Separate one-way analyses of variance use of organizational strategies. For example, ef®cient 9ANOVAs) were performed with schizotypy group performance on the Porteus Maze Task requires plan- assignment serving as the independent variable. ning and anticipation of blocked routes. Following univariate analyses, we conducted post Dif®culty inhibiting a prepotent response during the hoc comparisons of means. Since the groups were Stroop task may indicate ventral/orbitofrontal not equal in size, we used the Scheffe test. dysfunction. Subjects undergoing PET while partici- Effect sizes were calculated to ascertain the W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 179 strength of group differences. Cohen 91988) suggested frequency of automatisms and dissociative distur- that an effect size of 0.10 represents a small effect, bances in comparison to the median- and low-positive while an effect size of 0.25 represents a moderate schizotypy groups. The high-positive symptom group effect. Cohen 91988) maintained that effect sizes of also obtained signi®cantly higher scores on the brief 0.40 or greater represent large effects. Effect sizes hallucinatory events subscale in comparison to the 9Cohen's f ) for signi®cant differences and for group median-positive symptom group, while the median- differences which approached, but did not achieve, schizotypy group scored signi®cantly higher on statistical signi®cance are presented in Tables 1±2. these measures relative to the low-positive schizotypy The Pearson product±moment correlation was used group. to determine the strength of association between Positive schizotypy scores 9SPQ-B Factor 1) were variables. signi®cantly associated scores on the Limbic System Checklist including total score r ˆ 0:50; p , 0:01† and LSCL subscales assessing the frequency of disso- 3. Results ciative disturbances r ˆ 0:47; p , 0:01†; automa- tisms r ˆ 0:40; p , 0:01†; and brief hallucinatory Subtyping schizotypy: positive and negative dimen- events r ˆ 0:45; p , 0:01†; and were marginally sions. Classi®cation of subjects into positive and associated with performance on the subscale assessing negative schizotypy groups revealed distinct neuro- the frequency of paroxysmal somatic disturbances cognitive and clinical pro®les. Interestingly, there r ˆ 0:23; p , 0:02†: was no relationship between the positive and negative schizotypy dimensions in the student sample r ˆ 3.1.2. Clinical/personality measures 0:11; ns†: Findings are consistent with the contention Positive schizotypy groups also differed on self- that positive and negative schizotypy represent report measures of impulsivity and empathy, with discrete factors 9Siever, 1995). F 2; 100†ˆ11:61; p , 0:0001 and F 2; 100†ˆ 3.1. Positive schizotypy 5:32; p , 0:007; respectively. As shown in Table 1, post hoc analysis revealed that high-positive schizo- We observed a positive relation between measures typy subjects obtained signi®cantly higher scores on of temporolimbic dysfunction, impulsivity, antisocial the impulsivity and empathy subscales 9I 7 Question- behavior, and positive schizotypal phenomena. naire) in comparison to the low-positive schizotypy group. Moreover, the median-positive symptom group 3.1.1. Temporolimbic dysfunction also obtained signi®cantly higher scores on the impul- Group differences on the Limbic System Checklist sivity and empathy subscales relative to low-positive 9total score) F 2; 100†ˆ14:77; p , 0:0001† and schizotypy subjects. Positive schizotypy was signi®- subscales assessing the frequency of brief hallucina- cantly associated with impulsivity r ˆ 0:42; p , tory events F 2; 100†ˆ10:35; p , 0:0001†; auto- 0:01† and empathy r ˆ 0:30; p , 0:01† subscale matisms F 2; 100†ˆ11:08; p , 0:0001†; and scores. dissociative disturbances F 2; 100†ˆ21:24; p , Positive schizotypy subjects also obtained higher 0:0001† were signi®cant. scores on a self-report measure of antisocial person- Post hoc comparisons revealed that the high-posi- ality 9PDQ-4) in comparison to median- and low-posi- tive symptom group obtained signi®cantly higher tive symptom groups, F 2; 100†ˆ8:17; p , 0:0005; scores on the Limbic System Checklist 9total) in while median- and low-positive schizotypy groups did comparison to the median- and low-positive symptom not differ on the APD Subscale. Scores on the Cogni- groups. Median- and low-positive schizotypy groups tive±Perceptual Subscale of the SPQ-B were signi®- did not differ signi®cantly on the Limbic System cantly correlated with performance on the APD Checklist 9total score); however, group differences Subscale r ˆ 0:34; p , 0:01†: Unexpectedly, group were in the expected direction. The high-positive differences on the OCPD Subscale were signi®cant. symptom group scored signi®cantly higher on the Post hoc analysis revealed that high-positive symptom Limbic System Checklist subscales assessing the subjects obtained higher scores on the OCPD 180 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 Subscale relative to low-positive schizotypy Cognitive±Perceptual Subscale of the SPQ-B and subjects F 2; 100†ˆ4:04; p , 0:03†: Low-, scales tapping a common psychological construct? median-, and high-positive symptom groups did Of course, LSCL-33 subscales assessing the not differ on measures of generalized and social frequency of dissociative disturbances and brief anxiety 9ps . 0.60). Groups did not differ on hallucinatory events, and the Cognitive±Perceptual remaining clinical/personality measures and did Subscale of the SPQ-B may tap the same under- not differ in age or educational level 9all lying construct. ps . 0.20). Low-, median-, and high-positive schi- Does the observed relation between impulsivity, zotypy groups did not differ on the Interpersonal disinhibition, antisocial personality, and positive schi- subscale of the SPQ-B 9i.e. a measure of negative zotypy also re¯ect an association between the Cogni- schizotypy) 9p . 0.40). tive±Perceptual Subscale of the SPQ-B and scales tapping a similar construct? To address this issue, 3.1.3. Prefrontal measures regression analyses were performed. Scores on self- Positive schizotypy was associated with increased report measures of impulsivity, antisocial personality, scores on the Disinhibition Subscale of the Frontal and disinhibition were used as independent predictor Lobe Personality Scale r ˆ 0:44; p , 0:01†: Group variables, and positive schizotypy score was used as differences on the Disinhibition Subscale were signif- the dependent variable. Multiple regression analysis icant, with F 2; 100†ˆ9:84; p , 0:0001: Post hoc showed that predictor variables made unique contri- analysis revealed that the high-positive symptom butions. Scores on measures of impulsivity, antisocial group obtained signi®cantly higher scores on the personality, and disinhibition signi®cantly predicted Disinhibition Subscale in comparison to the median- positive schizotypy scores r ˆ 0:51; F ˆ 11:542; positive symptom group, while the median-schizo- df ˆ 3; 99; p , 0:0001† with each independent typy group scored signi®cantly higher on this measure predictor adding signi®cantly to the prediction. We relative to the low-positive schizotypy group. Low-, regressed Disinhibition Subscale and APD Subscale median-, and high-positive symptom groups did not scores on positive schizotypy scores r ˆ 0:48; F ˆ differ on tests of frontal executive function including 15:296; df ˆ 2; 100; p , 0:0001† and found that the the Trail-Making Test Ð Parts A and B 9ps . 0.64), predictor variables made unique contributions, with Controlled Word Fluency Test 9p . 0.14), Divergent p , 0.001 and p , 0.03, respectively. Bivariate Thinking Task 9p . 0.64), Porteus Maze Task regression analysis revealed that performance on the 9p . 0.96), and the Rey Complex Figure Task Disinhibition Subscale r ˆ 0:44; df ˆ 1; 101; F ˆ 9ps . 0.15). As shown in Table 1, low-, median-, 24:718; p , 0:0001†; impulsivity scale r ˆ 0:43; and high-positive symptom groups did not differ on F ˆ 23:108; df ˆ 1; 101; p , 0:0001† and APD the Apathy Subscale of the Frontal Lobe Personality Subscale r ˆ 0:37; F ˆ 16:623; df ˆ 1; 101; p , Scale 9p . 0.56). Although group differences on the 0:0001† predicted positive schizotypy scores. We Executive Dysfunction Subscale were marginally regressed Disinhibition and APD Subscale scores, signi®cant, as determined by an omnibus F statistic and Limbic System Checklist totals on positive schi- F 2; 100†ˆ3:19; p , 0:045†; post hoc comparisons zotypy scores r ˆ 0:58; df ˆ 3; 99; F ˆ 16:397; using the more conservative Scheffe test revealed that p , 0:0001† and found that the predictor variables group differences were not signi®cant. High-positive made independent contributions to the prediction of schizotypy subjects displayed faster mean reaction positive schizotypy scores, with p , 0.05, p , 0.04, times during word-naming trials of the Stroop and p , 0.0001, respectively. The observed relation Color±Word Test. Group differences were marginally between scores on self-report measures of disinhibi- signi®cant 9ps , 0.08). tion, antisocial personality, temporolimbic dysfunc- We observed a positive relation between tion, and positive schizotypy does not appear to measures of disinhibition, impulsivity, antisocial re¯ect an association between the Cognitive± behavior, and positive schizotypal phenomena. Do Perceptual Subscale of the SPQ-B and psychometric intercorrelations between these variables and posi- measures tapping a common psychological tive schizotypy re¯ect an association between the construct. W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 181 3.2. Negative schizotypy on the Executive Dysfunction r ˆ 0:24; p , 0:02† and Apathy Subscales r ˆ 0:50; p , 0:01†: The We observed an association between subtle high-negative schizotypy group scored signi®cantly performance de®cits on measures of frontal executive higher on these subscales in comparison to median- function, increased social anxiety, and negative schizo- and low-schizotypy subjects, with F 2; 100†ˆ4:63; typal symptoms. p , 0:02 and F 2; 100†ˆ19:59; p , 0:0001; respec- tively. Group differences on the Disinhibition 3.2.1. Clinical/personality measures Subscale were not signi®cant 9p . 0.54). High-negative symptom subjects obtained clini- We included a self-report index of anxiety and cally signi®cant scores on the OCPD Subscale relative avoidance behavior 9i.e. the FSS-MV) in order to to comparison groups F 2; 100†ˆ3:01; p , 0:05†: examine the relation between social and generalized Negative schizotypy was also associated with anxiety, and schizotypy. We observed a linear corre- increased social anxiety r ˆ 0:36; p , 0:01†: Group lation between social anxiety and negative schizo- differences on the social anxiety subscale were signif- typy; therefore, it is possible that performance icant, with F 2; 100†ˆ7:420; p , 0:001: Post hoc de®cits on tests assessing frontal executive function analysis revealed that the high-negative schizotypy may be associated with performance or test anxiety group scored signi®cantly higher on the social anxiety rather than negative schizotypy per se. For this reason, subscale in comparison to median- and low-schizo- we conducted analyses of covariance 9ANCOVA), typy subjects. with total Fear Survey score used as a covariate. Low-, median-, and high-negative schizotypy The association between negative schizotypy and groups did not differ on the Cognitive±Perceptual subtle performance de®cits on the Trail-Making subscale of the SPQ-B 9p . 0.54). Low-, median-, 9Part B) and Divergent Thinking tasks remained and high-negative schizotypy groups did not differ signi®cant controlling for the effect of anxiety, with on self-report measures of impulsivity, empathy, F 2; 99†ˆ3:76; p , 0:027 and F 2; 99†ˆ4:10; venturesomeness, and antisocial behavior 9all p , 0:019; respectively. ps . 0.63). Group differences on Limbic System The high-negative symptom group obtained lower Checklist subscales were also not signi®cant 9all Rey copy organization scores relative to median- and ps . 0.24). low-negative schizotypy subjects; however, this difference did not achieve statistical signi®cance 3.2.2. Prefrontal measures 9p , 0.14). Low-, median-, and high-negative schizo- The high-negative symptom group demonstrated typy groups did not differ on the recall component of subtle performance de®cits on tests of executive func- the Rey±Osterrieth Complex Figure Test 9p . 0.70) tion including the Trail-Making Test 9Parts A and B) and on the Porteus Maze 9p , 0.12), and Stroop and Divergent Thinking Task 9see Table 2). Group Color±Word Test 9all ps . 0.24). differences on the Trail-Making Test 9Parts A and B) were signi®cant, with F 2; 100†ˆ3:19; p , 0:05 and F 2; 100†ˆ3:51; p , 0:04†; respectively. 4. Handedness and schizotypy Post hoc comparisons revealed that the high- and median-negative symptom groups demonstrated In prior work, researchers reported an association greater mean reaction times during the Trail-Making between mixed handedness and schizotypal personal- Test 9Parts A and B) relative to low-negative schizo- ity features among nonclinical subjects and students typy subjects. Group differences on the Divergent psychometrically identi®ed as schizotypic or `psycho- Thinking Task were also signi®cant, with sis-prone' 9Kim et al., 1992; Chapman and Chapman, F 2; 100†ˆ4:12; p , 0:02: Post hoc analysis 1987). We compared the neurocognitive test perfor- revealed that high-negative schizotypy subjects mance of right- and left-handed students. Ninety-three generated fewer alternate uses relative to comparison subjects were right-handed and 10 subjects were left- subjects. On the Frontal Lobe Personality Scale, nega- handed, as determined by self-report. Neurocognitive tive schizotypy was associated with increased scores and clinical pro®les were remarkably similar. Groups 182 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 did not differ on the SPQ-B 9ps . 0.26), LSCL differences. Male students obtained signi®cantly 9ps . 0.06), Fear Survey 9ps . 0.24), APD Subscale higher scores on the Executive Dysfunction 9p . 0.24), OCPD Subscale 9p . 0.65), Disinhibition 9p , 0.04) and Apathy 9p , 0.02) Subscales of the Subscale 9p . 0.07), and measures of empathy FLPS relative to female students. Male students 9p . 0.35), venturesomeness 9p . 0.54), and impul- scored signi®cantly higher on the venturesomeness sivity 9p . 0.19). Groups also did not differ on the subscale 9p , 0.008) and obtained signi®cantly Stroop task 9ps . 0.63), Trail-Making Test 9Parts A lower scores on the empathy subscale 9p , 0.0009) and B) 9ps . 0.54), Divergent Thinking Task relative to female students. Male students obtained 9p . 0.28), Porteus Maze Task 9p . 0.10), Controlled higher scores on the Interpersonal Subscale 9Factor Word Fluency Test 9p . 0.91), and the Rey±Oster- 2) from the SPQ-B, a measure of negative schizotypal rieth Complex Figure Test 9ps . 0.51). Right-handed symptoms, relative to female subjects; however, subjects obtained signi®cantly higher scores on the group differences did not attain statistical signi®- Executive Dysfunction 9p , 0.02) and Apathy cance. Group differences on the remaining clinical 9p , 0.03) subscales from the FLPS-PV. In summary, measures did not approach signi®cance. we found no evidence of an association between Male subjects generated fewer words 9FAS Test) left-handedness and schizotypy. and alternate uses 9Divergent Thinking Task) in comparison to female students; although, differences were not statistically signi®cant. Male students 5. Gender and schizotypy demonstrated signi®cant performance de®cits during the Trail-Making Test 9Part B) 9p , 0.006) relative to Research examining neurocognitive function in female subjects. Group differences on the Rey± schizophrenia reveals signi®cant gender-based differ- Osterrieth Complex Figure and Stroop Color±Word ences. Seidman et al. 91997) found that male schizo- Test did not approach signi®cance. It is important to phrenia patients demonstrated performance de®cits on emphasize that group differences which approached, measures considered sensitive to orbitofrontal 9i.e. an but did not attain, statistical signi®cance may re¯ect a odour discrimination task) and dorsolateral±prefron- lack of power. The sample comprised 75 female and tal 9i.e. the WCST) dysfunction relative to female 28 male students. This gender distribution re¯ects the schizophrenia subjects; however, both male and gender ratio in the College of Arts and Sciences. female schizophrenia patients demonstrated substan- tial neuropsychological de®cits in comparison to 6. Discussion healthy control subjects. To address the issue of gender effects on neurocognitive test performance in Classi®cation of subjects into positive and negative a nonclinical student sample, we compared the neuro- schizotypy groups revealed distinct neurocognitive cognitive and clinical pro®les of male and female and clinical pro®les. There was no relationship students. between the positive and negative schizotypy dimen- Male and female students exhibited remarkably sions. This ®nding is consistent with the proposal that similar within-group performance patterns. That is, positive and negative schizotypy represent discrete we observed a positive relation between measures of dimensions or factors. Of course, multiple compari- temporolimbic dysfunction, impulsivity, antisocial sons increase the danger of Type I error and our ®nd- behavior, and positive schizotypal phenomena within ings should be interpreted with caution. Nevertheless, the male and female samples. Similarly, we observed results form a meaningful pattern and are consistent an association between performance de®cits on with prior research which found an association measures of frontal executive function, increased between performance de®cits on tests of frontal social anxiety, and negative schizotypal symptoms executive function and schizotypal personality. within female and male groups. Strikingly similar performance patterns were observed when we exam- 6.1. Negative schizotypy ined male and female groups separately. However, there were signi®cant between-group Negative schizotypy was associated with subtle W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 183 performance de®cits on measures of frontal executive measuring the frequency of compulsive checking, function, increased social anxiety, and obsessive± doubting, hoarding, ordering, obsessive ideation, and compulsive phenomena. Negative schizotypy was washing. Interestingly, positive schizotypy groups associated with increased scores on the Executive differed signi®cantly only on the subscales measuring Dysfunction and Apathy Subscales of the Frontal the frequency of obsessive ideation and doubting. Lobe Personality Scale. However, low-, median-, and high-negative schizotypy groups did not differ on measures of temporolimbic dysfunction, impulsiv- 6.2. Positive schizotypy ity, and antisocial behavior. Group differences on the Disinhibition Subscale were not signi®cant. In the present study, positive schizotypy groups did The relation between schizotypy, obsessive- not differ on tests of frontal executive function, and on compulsive phenomena, and executive function de®- measures of generalized and social anxiety; however, cits merits further study. We observed that the positive schizotypy was associated with increased presence of schizotypal personality features was asso- scores on self-report measures of disinhibition, impul- ciated with obsessive±compulsive personality traits. sivity, antisocial personality, and temporolimbic The preoccupation with rules and organization, dysfunction. Do intercorrelations between these vari- perfectionism, and in¯exibility displayed by subjects ables and positive schizotypy re¯ect an association exhibiting obsessive±compulsive personality traits between the Cognitive±Perceptual Subscale of the and schizotypal personality features may represent SPQ-B and scales tapping a common psychological behavioral strategies which evolve in response to construct? To address this issue, regression analyses executive function de®cits. In the present study, posi- were performed. Scores on self-report measures of tive schizotypy groups also differed on the OCPD impulsivity, antisocial personality, and disinhibition Subscale. This ®nding is not consistent with the notion were used as independent predictor variables, and that OCPTs represent compensatory strategies which positive schizotypy score was used as the dependent develop in response to executive function de®cits. In a variable. Multiple regression analyses showed that second study 9unpublished ®ndings), we sought to predictor variables made unique contributions. Scores characterize, with greater precision, the relationship on measures of impulsivity, antisocial personality, between schizotypy, mood, anxiety, and obsessive± and disinhibition signi®cantly predicted positive schi- compulsive phenomena in a student sample n ˆ zotypy scores with each independent predictor adding 109†: In this study, positive schizotypy groups did signi®cantly to the prediction. It is tempting to not differ on measures of inattention, apathy, execu- conclude that the relation between these variables tive dysfunction, obsessive±compulsive personality, re¯ects a common neurophysiological process. Of and social anxiety. Positive schizotypy groups did course, we must proceed cautiously when we argue not differ signi®cantly on the OCPD Subscale that variations in brain function correspond to patterns 9p . 0.91). This ®nding suggests OCPTs may be of clinical presentation. Nevertheless, it is interesting selectively associated with negative schizotypy and to note that orbitofrontal dysfunction is associated supports the contention that negative schizotypal with syndromes of disinhibition including antisocial symptoms are associated with subtle cognitive de®- personality disorder. Recent neuroimaging studies cits, which, in turn, are associated with the develop- and neuropsychological test ®ndings support the ment of adaptive or compensatory strategies which contention that prefrontal dysfunction 9particularly are collectively labeled OCPD. orbitofrontal) is associated with psychopathic person- In the unpublished study, negative schizotypy was ality traits and antisocial behavior 9Raine et al., 1998, strongly associated with elevated scores on measures 2000; Dinn and Harris, 2000; Lapierre et al., 1995). of inattention, executive dysfunction, obsessive± Antisocial personality, disinhibition and impulsivity, compulsive phenomena, and social anxiety. Negative and positive schizotypy may re¯ect, at least in part, schizotypy groups differed signi®cantly on the PDQ-4 prefrontal 9particularly orbitofrontal) hypofunction. Subscale assessing OCPD 9p , 0.006). Negative schi- Of course, the present study does not directly test zotypy groups also differed signi®cantly on scales this hypothesis. 184 W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185

6.3. Conclusion similarities and contrasts in neurocognitive and clinical func- tioning. Schizophr. Res. 7, 33±41. Crow, T.J., 1980. Molecular pathology of schizophrenia: more than Our ®ndings lend partial support to the suggestion one disease process? Br. Med. J. 280, 66±68. that positive and negative schizotypy represent Crow, T.J., 1985. The two-syndrome concept: origins and current discrete neurobehavioral dimensions. Investigating status. Schizophr. Bull. 11, 471±486. brain±behavior relationships in a nonclinical sample Dickey, C.C., McCarley, R.W., Voglmaier, M.M., Niznikiewicz, 9such as university students) may advance our under- M.A., Seidman, L.J., Hirayasu, Y., Fischer, I., Teh, E.K., Van standing of the mechanisms that underlie schizo- Rhoads, R., Jakab, M., Kikinis, R., Jolesz, F.A., Shenton, M.E., 1999. Schizotypal personality disorder and MRI abnormalities phrenic phenomena. Establishing neurocognitive/ of temporal lobe gray matter. Biol. Psychiatry 45, 1392±1402. clinical correlations in nonclinical samples relatively Diforio, D., Walker, E.F., Kestler, L.P., 2000. Executive functions free of possible confounding variables such as lengthy in adolescents with schizotypal personality disorder. Schizophr. psychoactive medication use, generalized intellectual Res. 42, 125±134. de®cits, knowledge of diagnostic status, and the Dinn, W.M., Harris, C.L., 2000. Neurocognitive function in anti- social personality disorder. Psychiatry Res. 97, 173±190. effects of institutionalization, may shed light on the Dinn, W.M., Harris, C.L., Aycicegi, A., 2001a. Primary obsessive± origins of these complex states. compulsive disorder and obsessive±compulsive disorder with schizotypal personality features: neuropsychological test ®nd- ings, submitted for publication. Dinn, W.M., Robbins, N.C., Harris, C.L., 2001b. Adult attention- References de®cit/hyperactivity disorder: neuropsychological correlates and clinical presentation 9in press). American Psychiatric Association, 1994. Diagnostic and Statistical Eysenck, S., Pearson, P., Easting, G., Allsopp, J., 1985. Age norms Manual of Mental Disorders, 4th ed. American Psychiatric for impulsiveness, venturesomeness, and empathy in adults. Association, Washington, DC. Pers. Individ. Differ. 6, 613±619. Aycicegi, A., Dinn, W.M., Harris, C.L., 2001. Neuropsychological Farmer, C.M., O'Donnell, B.F., Niznikiewicz, M.A., Voglmaier, function in obsessive±compulsive personality disorder with M.M., McCarley, R.W., Shenton, M.E., 2000. Visual perception schizotypal personality features, submitted for publication. and working memory in schizotypal personality disorder. Am. Bench, C.J., Frith, C.D., Grasby, P.M., Friston, K.J., Paulesu, E., J. Psychiatry 157, 781±786. Frackowiak, R.S.J., Dolan, R.J., 1993. Investigations of the Fletcher, P.C., Shallice, T., Dolan, R.J., 1998. The functional roles functional anatomy of attention using the Stroop test. of prefrontal cortex in episodic memory. I. Encoding. Brain 121, Neuropsychologia 31, 907±922. 1239±1248. Bergman, A.J., Harvey, P.D., Roitman, S.L., Mohs, R.C., Marder, Goodglass, H., Kaplan, E., 1972. The Assessment of Aphasia and D., Silverman, J.M., Siever, L.J., 1998. Verbal learning and Related Disorders. Lea and Febiger, Philadelphia. memory in schizotypal personality disorder. Schizophr. Bull. Grace, J., Malloy, P., 1992. Frontal Lobe Personality Scale. Brown 24, 635±641. University, Providence, RI. Buchsbaum, M.S., Yang, S., Hazlett, E., Siegel Jr, B.V., Germans, Guilford, J.P., Hoepfner, R., 1971. The Analysis of Intelligence. M., Haznedar, M., O'Flaithbheartaigh, S., Wei, T., Silverman, McGraw-Hill, New York. J., Siever, L.J., 1997. Ventricular volume and asymmetry in Hyler, S.E., 1994. Personality Diagnostic Questionnaire 9PDQ-4), schizotypal personality disorder and schizophrenia assessed New York. with magnetic resonance imaging. Schizophr. Res. 27, 45±53. Kim, D., Raine, A., Triphon, N., Green, M.F., 1992. Mixed hand- Cantor-Graae, E., Warkentin, S., Franzen, G., Risberg, J., 1993. edness and features of schizotypal personality in a nonclinical Frontal lobe challenge: a comparison of activation procedures sample. J. Nerv. Mental Dis. 180, 133±135. during rCBF measurements in normal subjects. Neuropsychiatry Lapierre, D., Braun, C.M.J., Hodgins, S., 1995. Ventral frontal de®- Neuropsychol. Behav. Neurol. 6, 83±92. cits in psychopathy: neuropsychological test ®ndings. Chapman, J.P., Chapman, L.J., 1987. Handedness of hypothetically Neuropsychologia 33, 139±151. psychosis-prone subjects. J. Abnormal Psychol. 96, 89±93. Lenzenweger, M.F., Gold, J.M., 2000. Auditory working memory Cohen, J., 1988. Statistical Power Analysis for the Behavioral and verbal recall memory in schizotypy. Schizophr. Res. 42, Sciences, 2nd ed. Erlbaum, Hillsdale, NJ. 101±110. Cohen, J.D., MacWhinney, B., Flatt, M.R., Provost, J., 1993. Lenzenweger, M.F., Kor®ne, L., 1994. Perceptual aberrations, schi- PsyScope: an interactive graphic system for designing and zotypy, and the Wisconsin Card Sorting Test. Schizophr. Bull. controlling experiments in the psychology laboratory using 20, 345±357. Macintosh computers. Behav. Res. Methods Instrum. Comput. Lezak, M.D., 1995. Neuropsychological Assessment. 3rd ed. 25, 257±271. Oxford University Press, New York. Condray, R., Steinhauer, S.R., 1992. Schizotypal personality disor- Lyons, M.J., Merla, M.E., Young, L., Kremen, W.S., 1991. der in individuals with and without schizophrenic relatives: Impaired neuropsychological functioning in symptomatic W.M. Dinn et al. / Schizophrenia Research 56 22002) 171±185 185

volunteers with schizotypy: preliminary ®ndings. Biol. Savage, C.R., Baer, L., Keuthen, N.J., Brown, H.D., Rauch, S.L., Psychiatry 30, 424±426. Jenike, M.A., 1999. Organizational strategies mediate Meehl, P.E., 1989. Schizotaxia revisited. Arch. Gen. Psychiatry 46, nonverbal memory impairment in obsessive±compulsive 935±944. disorder. Biol. Psychiatry 45, 905±916. Park, S., McTigue, K., 1997. Working memory and the syndromes Seidman, L.J., Goldstein, J.M., Goodman, J.M., Koren, D., Turner, of schizotypal personality. Schizophr. Res. 26, 213±220. W.M., Faraone, S.V., Tsuang, M.T., 1997. Sex differences in Poreh, A.M., Ross, T.P., Whitman, R.D., 1995. Reexamination of olfactory identi®cation and Wisconsin Card Sorting executive functions in psychosis-prone college students. Pers. Performance in schizophrenia: relationship to attention and Individ. Differ. 18, 535±539. verbal ability. Biol. Psychiatry 42, 104±115. Porteus, S.D., 1955. Porteus Maze Tests: Fifty Years Application. Siever, L.J., 1995. Brain structure/function and the dopamine Paci®c Books, Palo Alto, CA. system in schizotypal personality disorder. In: Raine, A., Raine, A., Benishay, D., 1995. The SPQ-B: a brief screening instru- Lencz, T., Mednick, S.A. 9Eds.), Schizotypal Personality. ment for schizotypal personality disorder. J. Pers. Disord. 9, Cambridge University Press, New York, pp. 272±286. 346±355. Siever, L.J., Keefe, R., Bernstein, D.P., Coccaro, E.F., Klar, H.M., Raine, A., Sheard, C., Reynolds, G.P., Lencz, T., 1992a. Pre-frontal Zemishlany, Z., Peterson, A.E., Davidson, M., Mahon, T., structural and functional de®cits associated with individual differ- Horvath, T., Mohs, R., 1990. Eye tracking impairment in ences in schizotypal personality. Schizophr. Res. 7, 237±247. clinically identi®ed patients with schizotypal personality Raine, A., Triphon, N., Kim, D., Hesler, A., Bird, L., Lencz, T., disorder. Am. J. Psychiatry 147, 740±745. Redmon, M., Scerbo, A., 1992b. Schizotypal personality: factor Smith, E.E., Jonides, J., 1999. Storage and executive processes in structure, sex differences, psychiatric differences, genetics, the frontal lobes. Science 283, 1657±1661. psychophysiology, and neuropsychology. Schizotypal Personality. Suhr, J.A., 1997. Executive functioning de®cits in hypothetically Cambridge University Press, New York, Paper presented at the psychosis-prone college students. Schizophr. Res. 27, 29±35. 1992 Western Psychological Association meeting, Portland, OR. Teicher, M.H., Glod, C.A., Surrey, J., Swett, C., 1993. Early child- Cited in Raine, A., Lencz, T., Mednick, S.A., 9Eds.), 91995). hood abuse and limbic system ratings in adult psychiatric outpa- Raine, A., Stoddard, J., Bihrle, S., Buchsbaum, M., 1998. Prefrontal tients. J. Neuropsychiatry Clin. Neurosci. 5 93), 301±306. glucose de®cits in murderers lacking psychosocial deprivation. Trestman, R.L., Keefe, R.S., Mitropoulou, V., Harvey, P.D., Neuropsychiatry Neuropsychol. Behav. Neurol. 11, 1±7. deVegvar, M.L., Lees-Roitman, S., Davidson, M., Aronson, Raine, A., Lencz, T., Bihrle, S., LaCasse, L., Colletti, P., 2000. A., Silverman, J., Siever, L.J., 1995. Cognitive function and Reduced prefrontal gray matter volume and reduced autonomic biological correlates of cognitive performance in schizotypal activity in antisocial personality disorder. Arch. Gen. Psychiatry personality disorder. Psychiatry Res. 59, 127±136. 57, 119±127. Voglmaier, M.M., Seidman, L.J., Salisbury, D., McCarley, R.W., Reitan, R.M., Wolfson, D., 1985. The Halstead±Reitan Neuropsy- 1997. Neuropsychological dysfunction in schizotypal personal- chological Test Battery. Neuropsychology Press, Tucson. Roitman, S.E.L., Mitropoulou, V., Keefe, R.S.E., Silverman, J.M., ity disorder: a pro®le analysis. Biol. Psychiatry 41, 530±540. Warkentin, S., Risberg, J., Nilsson, A., Karlson, S., et al., 1991. Serby, M., Harvey, P.D., Reynolds, D.A., Mohs, R.C., Siever, L.J., 2000. Visuospatial working memory in schizotypal Cortical activity during speech production: a study of regional personality disorder patients. Schizophr. Res. 41, 447±455. cerebral blood ¯ow in normal subjects performing a word Rolls, E.T., 1995. A theory of emotion and consciousness, and its ¯uency task. Neuropsychiatry Neuropsychol Behav Neurol. 4, application to understanding the neural basis of emotion. In: 305±316. Gazzaniga, M.S. 9Ed.), The Cognitive Neurosciences. MIT Wolpe, J., Lang, P.J., 1964. A fear survey schedule for use in Press, Cambridge, MA, pp. 1091±1106. behavior therapy. Behav. Res. Ther. 2, 27±30.