Asian Journal of Psychiatry 7 (2014) 22–27
Contents lists available at ScienceDirect
Asian Journal of Psychiatry journal homepage: www.elsevier.com/locate/ajp
Utility of a computerized, paced semantic verbal fluency paradigm in differentiating schizophrenia and healthy subjects Mohammed K. Shakeel a, Harsha N. Halahalli b,1, Kiran Kumar b, Sanjeev Jain c, John P. John b,c,d,* a
Schizophrenia Research Lab, Department of Psychology, Kent State University, P.O. Box 5190, Kent, OH 44242-0001, USA Multimodal Brain Image Analysis Laboratory (MBIAL), NIMHANS, Bangalore 560029, India Department of Psychiatry, NIMHANS, Bangalore 560029, India d Department of Clinical Neuroscience, NIMHANS, Bangalore 560029, India b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 July 2013 Received in revised form 8 September 2013 Accepted 22 September 2013
Functional magnetic resonance imaging (fMRI) paradigms designed to study word generation traditionally utilize a computerized paced version of the verbal fluency task (VFT) comprising ‘blocks’ of word generation and a baseline word repetition task. The utility of the above paced VFT paradigm in differentiating neuropsychiatric patients from healthy subjects has not been systematically examined. We administered a computerized, paced version of the semantic VFT comprising word generation and word repetition blocks to 24 schizophrenia and 24 matched healthy subjects, both before and during fMRI acquisition. The performance of patients with schizophrenia was significantly inferior to that of healthy control subjects in both the ‘pre-scan’ and ‘intra-scan’ sessions of the computerized paced semantic VFT. Specifically, schizophrenia patients generated significantly fewer total responses (VFTR) as well as correct responses (VFCR), but a larger number of ‘no response’ trials. However, there were no significant group differences in perseverative responses in the pre-scan session or ‘intra-scan’ sessions. The above computerized task has been reported by us previously to generate a behavioral performance index with hemodynamic correlates (John et al., 2011). Thus, our findings support the use of computerized paced VFT comprising word generation and word repetition blocks in both clinical and research settings. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Word generation Word repetition VFT Hemodynamic fMRI
1. Introduction Verbal fluency, or the capacity for word generation, has been used as a test of frontal and temporal lobe functioning in various neuropsychiatric conditions like fronto-temporal dementia (Rascovsky et al., 2007), Alzheimer’s disease (Henry and Crawford, 2004; Rascovsky et al., 2007), Parkinson’s disease (Henry and Crawford, 2004), Huntington’s disease (Ho et al., 2002) and schizophrenia (Ojeda et al., 2008; Curtis et al., 1998). Frontal lobes are involved in initiation and retrieval of information, while the role of temporal lobes is more specific to semantic memory in verbal fluency (Rascovsky et al., 2007). Patients with frontal lesions, therefore, show deficits in both semantic and letter verbal fluency
* Corresponding author at: Multimodal Brain Image Analysis Laboratory (MBIAL), Neurobiology Research Center (NRC), PB No. 2900, Dharmaram P.O., Hosur Road, Bangalore, Karnataka 560029, India. Tel.: +91 80 26995329; fax: +91 80 26564822. E-mail addresses: [email protected], [email protected] (J.P. John). 1 Current affiliation: Department of Physiology, K.S. Hegde Medical Academy, Nitte University, Mangalore, Karnataka 575018, India. 1876-2018/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ajp.2013.09.010
tasks (VFT) (Baldo et al., 2001). Verbal fluency paradigms require subjects to generate words in response to phonological (words beginning with a particular letter, e.g., F, A, S) or semantic (words belonging to a particular category, e.g., animals, birds, flowers) cues (Lezak, 1995). In conventional semantic VFT paradigms, the subject is asked to generate as many examples that belong to a specific category like animals, fruits, or objects, in a given period of time (one or two minutes). With the advent of functional magnetic resonance imaging (fMRI) in the 1990s, verbal fluency paradigms were adapted to suit the image acquisition requirements during cognitive activation experiments. Since the un-paced responses in conventional paradigms posed problems for the detection of task-related blood oxygenation level dependent (BOLD) signals (Basho et al., 2007), fMRI paradigms used paced production of words for obtaining more robust activations. Paced production of words is characterized by superior error detection and greater inhibitory control associated with greater anterior cingulate activations, when compared to un-paced word generation (Frith et al., 1995; Fu et al., 2002; Basho et al., 2007). This could result in improved
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
accuracy of responses as well as reduced number of perseverative responses with paced verbal fluency paradigms in comparison to un-paced paradigms. A study of the effects of generation mode in semantic fluency by Basho et al. (2007) concluded that the paced, overt adaptation of the conventional VFT is best suited for use in fMRI paradigms, owing to their superiority with regard to control over and monitoring of behavioral responses. Thus, there could be some merit in using the above computerized paced, overt adaptation of the conventional VFT even in clinical situations. Another advantage of the computerized version of the VFT is that the verbal responses of subjects can be recorded along with generating behavioral response parameters such as response latency, while this is not possible with the conventional VFT used in clinical situations. Moreover, audio-taping the responses reduces the dependence on the examiner’s memory and writing speed for accurate scoring of correct, incorrect and perseverative responses (e.g., Chan et al., 2003). Among psychiatric disorders, verbal fluency deficits have been most consistently demonstrated in patients with schizophrenia in behavioral (Ojeda et al., 2008) as well as fMRI (Curtis et al., 1998) experiments. Semantic verbal fluency has been shown to have higher discriminating power to elicit significantly larger group differences between individuals with schizophrenia and healthy controls than phonological fluency (Melinder et al., 2005). Semantic verbal fluency deficits, unlike phonemic and design fluency deficits, were found even in early onset schizophrenia patients (Phillips et al., 2004). Moreover, relatives of schizophrenia patients were also found to perform less well than controls on semantic and (to a lesser extent) phonemic verbal fluency (Szoke et al., 2005). Thus semantic verbal fluency has been suggested as the best candidate for cognitive endophenotype of schizophrenia (Phillips et al., 2004; Szoke et al., 2005). Using a semantic verbal fluency paradigm, we have recently reported that schizophrenia subjects showed aberrant fMRI BOLD activations and deficient deactivations during semantic word generation in comparison to matched healthy subjects (John et al., 2011). Word generation paradigms optimized for fMRI experiments typically comprise at least two tasks – a word generation condition as well as a baseline condition (e.g., word repetition task) administered alternately in ‘‘blocks’’ using some stimulus presentation software. As mentioned above, the paced version of the VFT used in such experiments differs substantially from the conventional un-paced versions with respect to its underlying physiology and the cognitive processing. Despite VFT studies using such paced fMRI paradigms being in vogue since the early 90s (e.g., Hugdahl et al., 1999), no study has so far reported whether this computerized paced-overt version of the VFT comprising both word generation and baseline tasks could reliably differentiate schizophrenia from healthy subjects. In the present paper, we report the utility of the computerized paced semantic verbal fluency paradigm that was used in the above-cited fMRI study (John et al., 2011) in differentiating schizophrenia and healthy subjects, both in the fMRI-experimental setting as well as outside the scanner, and highlight its potential advantages over conventional VFT paradigms. We compared the behavioral responses during the ‘‘intra-scan’’ session, of the same set of subjects who participated in the above fMRI study with that during the ‘‘pre-scan’’ session (i.e., following the pre-scan training) to examine the utility of the computerized paced, overt version of the VFT task in differentiating schizophrenia subjects from matched healthy comparison subjects. We hypothesized that the computerized paced, overt semantic VFT would discriminate schizophrenia subjects from matched healthy comparison subjects, both during the ‘pre-scan’ blocks as well as the ‘intra-scan’ blocks. We further hypothesized, based on a large number of previous studies (Curtis et al., 1998; Melinder et al.,
23
2005; Szoke et al., 2005; Ojeda et al., 2008), that schizophrenia subjects would perform worse than healthy control subjects in both conditions. 2. Methods The study was carried out at the National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore, India, with due approval from the NIMHANS ethics committee, thus conforming to the ethical standards laid down in the 1964 Declaration of Helsinki. Written informed consent was obtained from all subjects (and their legally qualified representatives in case of patients) prior to enrollment into the study. 2.1. Subjects The study samples comprised of 24 healthy subjects (HS) (mean age = 27.42 years (S.D. = 6.27; 19 males) who were recruited by word of mouth and constituted predominantly by hospital staff and their contacts, as well as 24 schizophrenia patients (SZ) (mean age = 30.13 years (S.D. = 9.48; 16 males), recruited from the outpatient department of NIMHANS by purposive sampling. The socio-demographic and clinical characteristics of the study samples are detailed in John et al. (2011). Only right handed [as determined by modified Annett’s inventory (Annett, 1976)] subjects in the age group of 17–50, with a Mini Mental Status Examination (MMSE) score of 23 or more were included in the study. All participants were native speakers of various South Indian Dravidian languages (Table 1) and could read their preferred language with ease. None of these subjects were formally exposed to a second language other than their mother tongue before age six. A study-specific questionnaire to assess the speaking, reading and writing skills in different languages as well as to ascertain the primary language was administered prior to recruiting the subjects for the fMRI experiment. The matrix reasoning module of Wechsler’s Adult Intelligence Scale-III (WAIS-III) (Wechsler, 1997) was administered on all subjects to determine their perceptual organization ability. This provided an index of their intellectual ability, independent of verbal abilities. The samples were matched for age, education status, handedness, and WAIS-III scores. Group comparisons of age, WAIS-III scores, Hindi Mental State Exam (HMSE) scores, years of education and gender distribution are given in Table 1. There were no significant group differences on these variables. The Diagnostic and Statistical Manual for Mental DisordersFourth Edition (DSM-IV) (American Psychiatric Association, 2000) was used to arrive at a diagnosis of schizophrenia or schizophreniform disorder based on the consensus of the research psychiatrist who conducted a semi-structured interview and a trained research assistant who administered the Mini International Neuropsychiatric Interview (MINI) Plus (Sheehan et al., 1998). Two trained raters with good inter-rater reliability evaluated the baseline severity of schizophrenia psychopathology using the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987). The healthy comparison subjects were ascertained to be free from Axis I or II psychiatric disorders using the MINI-Plus (Sheehan et al., 1998). The presence of other medical/neurological conditions requiring continuous medications, current use of psychotropic drugs and history of psychiatric illness in first-degree relatives were ruled out by an unstructured clinical interview. 2.2. Computerized verbal fluency paradigm The computerized paced, overt verbal fluency protocol comprised of two tasks, viz., the ‘‘word repetition’’ condition and the ‘‘semantic category word generation’’ condition (Supplementary
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
24
Table 1 Socio-demographic and clinical characteristics of the study samples. Socio-demographic and illness characteristics
Subjects (N = 48) Healthy subjects (n = 24)
Schizophrenia (n = 24)
Group comparisons
Gender Male
19 (79.20%)
16 (66.70%)
Pearson Chi-square = 0.949; p < 0.0.330
Female Age (years)
5 (20.80%) 27.42 (SD = 6.27); range: 19–44
8 (33.30%) 30.13 (SD = 9.48); range: 17–47
F = 1.211 (p < 0.277)
0.026
Education No. of years of formal education HMSE WAIS-III
13.50 (SD = 2.27); range 10–17 30.67 (1.63) 12.50 (4.62)
12.21 (SD = 2.98); range 6–18 30.41 (1.10) 10.25 (3.72)
F = 2.86 (p < 0.10) F = 0.387 (p < 0.537) F = 3.46 (p < 0.07)
0.06 0.008 0.07
Diagnosis Paranoid Schizophrenia Undifferentiated schizophrenia Age of onset of illness (years) Duration of illness (months)
19 5 29.75 (SD = 9.69); range 15–46 32.12 (SD = 17.92); range 6–60
Medication status Neuroleptic naı¨ve/free On medication Mean risperidone dose (mg/day)
6/3 15 2.46 (SD = 1.30)
PANSS Positive score Negative score Global psychopathology score PANSS total score
13.96 15.29 25.21 54.50
Effect size (partial eta squared)
(SD = 6.03) (SD = 7.36) (SD = 7.62) (SD = 16.45)
HMSE, Hindi Mental State Exam; WAIS-III, Wechsler’s Adult Intelligence Scale-III; PANSS, positive and negative syndrome scale
Fig. 1). These were administered alternately in blocks as a paced, overt and visually-cued paradigm, using the E-Prime stimulus presentation software, version 2.0. (PST Inc., PA, USA). During fMRI acquisition, this paradigm was administered within the Integrated Functional Imaging System (IFIS-SA) (Invivo, Orlando, FL, USA). Each block consisted of seven trials of either word repetition or word generation. During the word repetition blocks, subjects were instructed to repeat the word ‘‘pass’’ every time it appeared on the screen. During the word generation blocks, the subjects were shown a category word-cue on the screen and were instructed to overtly produce an exemplar of the category. The ‘pre-scan’ session of the test comprised a short four-block version of the task (two blocks each for word repetition and word generation, administered alternately). The two word generation blocks during the ‘pre-scan’ session comprised of semantic categories other than those included in the intra-scan session, viz., ‘‘wooden articles’’ and ‘‘kitchen articles’’. The ‘pre-scan’ session was administered by H.N.H. for all subjects. In the intrascan session, six semantic categories were used, viz., ‘‘animals’’, ‘‘vegetables’’, ‘‘birds’’, ‘‘fruits’’, ‘‘flowers’’ and ‘‘trees’’. The above six semantic category word generation blocks were administered alternately with six word repetition blocks. The order of the word generation and repetition blocks were counterbalanced. Subjects were instructed to keep their gaze fixed on the center of the screen and, as soon as the category-word cue was presented, respond with unique examples from the category. They were also instructed to say ‘‘pass’’ if they could not think of an appropriate example, in order to minimize across-block differences in vocalization responses according to the number of exemplars produced. Each trial began with the display of a central fixation cross lasting for 2000 ms followed by the cue word for 1000 ms which was replaced by the fixation cross for a further 1000 ms. During the ‘intra-scan’ session, the overt responses were timed to occur during the ‘silent’ periods in between image acquisition, thereby making the ‘on-line’
measurement of behavioral performance possible (Fu et al., 2002; Abrahams et al., 2003). While the cue word for the word generation blocks was the name of the category (e.g., ‘‘animal’’, ‘‘vegetable’’, etc.), the word ‘‘pass’’ was used for the baseline word repetition blocks. The stimuli were presented in white font over a black background. During the fMRI experiment, all presented stimuli measured 2.5 cm vertically on the screen subtending a visual angle of about 2.48, and were displayed using a fiber optic LCD-based video system controlled by the IFIS-SA system. The responses were picked up by a magnetic resonance (MR) compatible microphone integrated within the projection hood, recorded on to a computer hard disk using Microsoft Windows Media Encoder, and were played back later for scoring as ‘‘correct’’, ‘‘incorrect’’, ‘‘repetitive’’, ‘‘pass’’ or ‘‘no responses’’. Similarly, the responses during the ‘pre-scan’ session were also recorded. The verbal fluency correct responses (VFCR) and the verbal fluency total responses (VFTR: correct + incorrect + repetitions + ‘pass’ responses) across the six word generation blocks in the intrascan session and two blocks during the pre-scan session were calculated. Response latencies were measured during the ‘intra-scan’ session for both word generation and word repetition trials. From the mean response latencies thus obtained, we have earlier reported that the difference in mean latencies between the word generation and word repetition trials (latgen latrep) showed a significant inverse relationship with task-induced fMRI BOLD deactivation in HS, while such a relationship was not observed in the poorly performing SZ (John et al., 2011). 2.3. Statistical analysis The mean VFTR, VFCR, no responses and repetitive responses for healthy and schizophrenia subjects in both the ‘pre-scan’ and ‘intrascan’ sessions were compared across the subject groups. Since the variables were non-normally distributed, as assessed using Shapiro– Wilk’s test, non-parametric statistical test (Mann–Whitney U) was
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
25
Table 2 Behavioral performance of patients with schizophrenia and healthy control subjects on verbal fluency task in the pre-scan and intra-scan sessions.a Pre-scan
Intra-scan
Behavioral measures
Healthy subjects M (SD)
Schizophrenia subjects M (SD)
MWU
P
Effect size (r)
Healthy subjects M (SD)
Schizophrenia subjects M (SD)
MWU
P
Effect size (r)
TR CR NR PR
12.92 11.29 1.00 0.13
11.50 10.38 2.29 0.17
136b 202.5 153.5b 266.00
Contents lists available at ScienceDirect
Asian Journal of Psychiatry journal homepage: www.elsevier.com/locate/ajp
Utility of a computerized, paced semantic verbal fluency paradigm in differentiating schizophrenia and healthy subjects Mohammed K. Shakeel a, Harsha N. Halahalli b,1, Kiran Kumar b, Sanjeev Jain c, John P. John b,c,d,* a
Schizophrenia Research Lab, Department of Psychology, Kent State University, P.O. Box 5190, Kent, OH 44242-0001, USA Multimodal Brain Image Analysis Laboratory (MBIAL), NIMHANS, Bangalore 560029, India Department of Psychiatry, NIMHANS, Bangalore 560029, India d Department of Clinical Neuroscience, NIMHANS, Bangalore 560029, India b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 July 2013 Received in revised form 8 September 2013 Accepted 22 September 2013
Functional magnetic resonance imaging (fMRI) paradigms designed to study word generation traditionally utilize a computerized paced version of the verbal fluency task (VFT) comprising ‘blocks’ of word generation and a baseline word repetition task. The utility of the above paced VFT paradigm in differentiating neuropsychiatric patients from healthy subjects has not been systematically examined. We administered a computerized, paced version of the semantic VFT comprising word generation and word repetition blocks to 24 schizophrenia and 24 matched healthy subjects, both before and during fMRI acquisition. The performance of patients with schizophrenia was significantly inferior to that of healthy control subjects in both the ‘pre-scan’ and ‘intra-scan’ sessions of the computerized paced semantic VFT. Specifically, schizophrenia patients generated significantly fewer total responses (VFTR) as well as correct responses (VFCR), but a larger number of ‘no response’ trials. However, there were no significant group differences in perseverative responses in the pre-scan session or ‘intra-scan’ sessions. The above computerized task has been reported by us previously to generate a behavioral performance index with hemodynamic correlates (John et al., 2011). Thus, our findings support the use of computerized paced VFT comprising word generation and word repetition blocks in both clinical and research settings. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Word generation Word repetition VFT Hemodynamic fMRI
1. Introduction Verbal fluency, or the capacity for word generation, has been used as a test of frontal and temporal lobe functioning in various neuropsychiatric conditions like fronto-temporal dementia (Rascovsky et al., 2007), Alzheimer’s disease (Henry and Crawford, 2004; Rascovsky et al., 2007), Parkinson’s disease (Henry and Crawford, 2004), Huntington’s disease (Ho et al., 2002) and schizophrenia (Ojeda et al., 2008; Curtis et al., 1998). Frontal lobes are involved in initiation and retrieval of information, while the role of temporal lobes is more specific to semantic memory in verbal fluency (Rascovsky et al., 2007). Patients with frontal lesions, therefore, show deficits in both semantic and letter verbal fluency
* Corresponding author at: Multimodal Brain Image Analysis Laboratory (MBIAL), Neurobiology Research Center (NRC), PB No. 2900, Dharmaram P.O., Hosur Road, Bangalore, Karnataka 560029, India. Tel.: +91 80 26995329; fax: +91 80 26564822. E-mail addresses: [email protected], [email protected] (J.P. John). 1 Current affiliation: Department of Physiology, K.S. Hegde Medical Academy, Nitte University, Mangalore, Karnataka 575018, India. 1876-2018/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ajp.2013.09.010
tasks (VFT) (Baldo et al., 2001). Verbal fluency paradigms require subjects to generate words in response to phonological (words beginning with a particular letter, e.g., F, A, S) or semantic (words belonging to a particular category, e.g., animals, birds, flowers) cues (Lezak, 1995). In conventional semantic VFT paradigms, the subject is asked to generate as many examples that belong to a specific category like animals, fruits, or objects, in a given period of time (one or two minutes). With the advent of functional magnetic resonance imaging (fMRI) in the 1990s, verbal fluency paradigms were adapted to suit the image acquisition requirements during cognitive activation experiments. Since the un-paced responses in conventional paradigms posed problems for the detection of task-related blood oxygenation level dependent (BOLD) signals (Basho et al., 2007), fMRI paradigms used paced production of words for obtaining more robust activations. Paced production of words is characterized by superior error detection and greater inhibitory control associated with greater anterior cingulate activations, when compared to un-paced word generation (Frith et al., 1995; Fu et al., 2002; Basho et al., 2007). This could result in improved
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
accuracy of responses as well as reduced number of perseverative responses with paced verbal fluency paradigms in comparison to un-paced paradigms. A study of the effects of generation mode in semantic fluency by Basho et al. (2007) concluded that the paced, overt adaptation of the conventional VFT is best suited for use in fMRI paradigms, owing to their superiority with regard to control over and monitoring of behavioral responses. Thus, there could be some merit in using the above computerized paced, overt adaptation of the conventional VFT even in clinical situations. Another advantage of the computerized version of the VFT is that the verbal responses of subjects can be recorded along with generating behavioral response parameters such as response latency, while this is not possible with the conventional VFT used in clinical situations. Moreover, audio-taping the responses reduces the dependence on the examiner’s memory and writing speed for accurate scoring of correct, incorrect and perseverative responses (e.g., Chan et al., 2003). Among psychiatric disorders, verbal fluency deficits have been most consistently demonstrated in patients with schizophrenia in behavioral (Ojeda et al., 2008) as well as fMRI (Curtis et al., 1998) experiments. Semantic verbal fluency has been shown to have higher discriminating power to elicit significantly larger group differences between individuals with schizophrenia and healthy controls than phonological fluency (Melinder et al., 2005). Semantic verbal fluency deficits, unlike phonemic and design fluency deficits, were found even in early onset schizophrenia patients (Phillips et al., 2004). Moreover, relatives of schizophrenia patients were also found to perform less well than controls on semantic and (to a lesser extent) phonemic verbal fluency (Szoke et al., 2005). Thus semantic verbal fluency has been suggested as the best candidate for cognitive endophenotype of schizophrenia (Phillips et al., 2004; Szoke et al., 2005). Using a semantic verbal fluency paradigm, we have recently reported that schizophrenia subjects showed aberrant fMRI BOLD activations and deficient deactivations during semantic word generation in comparison to matched healthy subjects (John et al., 2011). Word generation paradigms optimized for fMRI experiments typically comprise at least two tasks – a word generation condition as well as a baseline condition (e.g., word repetition task) administered alternately in ‘‘blocks’’ using some stimulus presentation software. As mentioned above, the paced version of the VFT used in such experiments differs substantially from the conventional un-paced versions with respect to its underlying physiology and the cognitive processing. Despite VFT studies using such paced fMRI paradigms being in vogue since the early 90s (e.g., Hugdahl et al., 1999), no study has so far reported whether this computerized paced-overt version of the VFT comprising both word generation and baseline tasks could reliably differentiate schizophrenia from healthy subjects. In the present paper, we report the utility of the computerized paced semantic verbal fluency paradigm that was used in the above-cited fMRI study (John et al., 2011) in differentiating schizophrenia and healthy subjects, both in the fMRI-experimental setting as well as outside the scanner, and highlight its potential advantages over conventional VFT paradigms. We compared the behavioral responses during the ‘‘intra-scan’’ session, of the same set of subjects who participated in the above fMRI study with that during the ‘‘pre-scan’’ session (i.e., following the pre-scan training) to examine the utility of the computerized paced, overt version of the VFT task in differentiating schizophrenia subjects from matched healthy comparison subjects. We hypothesized that the computerized paced, overt semantic VFT would discriminate schizophrenia subjects from matched healthy comparison subjects, both during the ‘pre-scan’ blocks as well as the ‘intra-scan’ blocks. We further hypothesized, based on a large number of previous studies (Curtis et al., 1998; Melinder et al.,
23
2005; Szoke et al., 2005; Ojeda et al., 2008), that schizophrenia subjects would perform worse than healthy control subjects in both conditions. 2. Methods The study was carried out at the National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore, India, with due approval from the NIMHANS ethics committee, thus conforming to the ethical standards laid down in the 1964 Declaration of Helsinki. Written informed consent was obtained from all subjects (and their legally qualified representatives in case of patients) prior to enrollment into the study. 2.1. Subjects The study samples comprised of 24 healthy subjects (HS) (mean age = 27.42 years (S.D. = 6.27; 19 males) who were recruited by word of mouth and constituted predominantly by hospital staff and their contacts, as well as 24 schizophrenia patients (SZ) (mean age = 30.13 years (S.D. = 9.48; 16 males), recruited from the outpatient department of NIMHANS by purposive sampling. The socio-demographic and clinical characteristics of the study samples are detailed in John et al. (2011). Only right handed [as determined by modified Annett’s inventory (Annett, 1976)] subjects in the age group of 17–50, with a Mini Mental Status Examination (MMSE) score of 23 or more were included in the study. All participants were native speakers of various South Indian Dravidian languages (Table 1) and could read their preferred language with ease. None of these subjects were formally exposed to a second language other than their mother tongue before age six. A study-specific questionnaire to assess the speaking, reading and writing skills in different languages as well as to ascertain the primary language was administered prior to recruiting the subjects for the fMRI experiment. The matrix reasoning module of Wechsler’s Adult Intelligence Scale-III (WAIS-III) (Wechsler, 1997) was administered on all subjects to determine their perceptual organization ability. This provided an index of their intellectual ability, independent of verbal abilities. The samples were matched for age, education status, handedness, and WAIS-III scores. Group comparisons of age, WAIS-III scores, Hindi Mental State Exam (HMSE) scores, years of education and gender distribution are given in Table 1. There were no significant group differences on these variables. The Diagnostic and Statistical Manual for Mental DisordersFourth Edition (DSM-IV) (American Psychiatric Association, 2000) was used to arrive at a diagnosis of schizophrenia or schizophreniform disorder based on the consensus of the research psychiatrist who conducted a semi-structured interview and a trained research assistant who administered the Mini International Neuropsychiatric Interview (MINI) Plus (Sheehan et al., 1998). Two trained raters with good inter-rater reliability evaluated the baseline severity of schizophrenia psychopathology using the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987). The healthy comparison subjects were ascertained to be free from Axis I or II psychiatric disorders using the MINI-Plus (Sheehan et al., 1998). The presence of other medical/neurological conditions requiring continuous medications, current use of psychotropic drugs and history of psychiatric illness in first-degree relatives were ruled out by an unstructured clinical interview. 2.2. Computerized verbal fluency paradigm The computerized paced, overt verbal fluency protocol comprised of two tasks, viz., the ‘‘word repetition’’ condition and the ‘‘semantic category word generation’’ condition (Supplementary
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
24
Table 1 Socio-demographic and clinical characteristics of the study samples. Socio-demographic and illness characteristics
Subjects (N = 48) Healthy subjects (n = 24)
Schizophrenia (n = 24)
Group comparisons
Gender Male
19 (79.20%)
16 (66.70%)
Pearson Chi-square = 0.949; p < 0.0.330
Female Age (years)
5 (20.80%) 27.42 (SD = 6.27); range: 19–44
8 (33.30%) 30.13 (SD = 9.48); range: 17–47
F = 1.211 (p < 0.277)
0.026
Education No. of years of formal education HMSE WAIS-III
13.50 (SD = 2.27); range 10–17 30.67 (1.63) 12.50 (4.62)
12.21 (SD = 2.98); range 6–18 30.41 (1.10) 10.25 (3.72)
F = 2.86 (p < 0.10) F = 0.387 (p < 0.537) F = 3.46 (p < 0.07)
0.06 0.008 0.07
Diagnosis Paranoid Schizophrenia Undifferentiated schizophrenia Age of onset of illness (years) Duration of illness (months)
19 5 29.75 (SD = 9.69); range 15–46 32.12 (SD = 17.92); range 6–60
Medication status Neuroleptic naı¨ve/free On medication Mean risperidone dose (mg/day)
6/3 15 2.46 (SD = 1.30)
PANSS Positive score Negative score Global psychopathology score PANSS total score
13.96 15.29 25.21 54.50
Effect size (partial eta squared)
(SD = 6.03) (SD = 7.36) (SD = 7.62) (SD = 16.45)
HMSE, Hindi Mental State Exam; WAIS-III, Wechsler’s Adult Intelligence Scale-III; PANSS, positive and negative syndrome scale
Fig. 1). These were administered alternately in blocks as a paced, overt and visually-cued paradigm, using the E-Prime stimulus presentation software, version 2.0. (PST Inc., PA, USA). During fMRI acquisition, this paradigm was administered within the Integrated Functional Imaging System (IFIS-SA) (Invivo, Orlando, FL, USA). Each block consisted of seven trials of either word repetition or word generation. During the word repetition blocks, subjects were instructed to repeat the word ‘‘pass’’ every time it appeared on the screen. During the word generation blocks, the subjects were shown a category word-cue on the screen and were instructed to overtly produce an exemplar of the category. The ‘pre-scan’ session of the test comprised a short four-block version of the task (two blocks each for word repetition and word generation, administered alternately). The two word generation blocks during the ‘pre-scan’ session comprised of semantic categories other than those included in the intra-scan session, viz., ‘‘wooden articles’’ and ‘‘kitchen articles’’. The ‘pre-scan’ session was administered by H.N.H. for all subjects. In the intrascan session, six semantic categories were used, viz., ‘‘animals’’, ‘‘vegetables’’, ‘‘birds’’, ‘‘fruits’’, ‘‘flowers’’ and ‘‘trees’’. The above six semantic category word generation blocks were administered alternately with six word repetition blocks. The order of the word generation and repetition blocks were counterbalanced. Subjects were instructed to keep their gaze fixed on the center of the screen and, as soon as the category-word cue was presented, respond with unique examples from the category. They were also instructed to say ‘‘pass’’ if they could not think of an appropriate example, in order to minimize across-block differences in vocalization responses according to the number of exemplars produced. Each trial began with the display of a central fixation cross lasting for 2000 ms followed by the cue word for 1000 ms which was replaced by the fixation cross for a further 1000 ms. During the ‘intra-scan’ session, the overt responses were timed to occur during the ‘silent’ periods in between image acquisition, thereby making the ‘on-line’
measurement of behavioral performance possible (Fu et al., 2002; Abrahams et al., 2003). While the cue word for the word generation blocks was the name of the category (e.g., ‘‘animal’’, ‘‘vegetable’’, etc.), the word ‘‘pass’’ was used for the baseline word repetition blocks. The stimuli were presented in white font over a black background. During the fMRI experiment, all presented stimuli measured 2.5 cm vertically on the screen subtending a visual angle of about 2.48, and were displayed using a fiber optic LCD-based video system controlled by the IFIS-SA system. The responses were picked up by a magnetic resonance (MR) compatible microphone integrated within the projection hood, recorded on to a computer hard disk using Microsoft Windows Media Encoder, and were played back later for scoring as ‘‘correct’’, ‘‘incorrect’’, ‘‘repetitive’’, ‘‘pass’’ or ‘‘no responses’’. Similarly, the responses during the ‘pre-scan’ session were also recorded. The verbal fluency correct responses (VFCR) and the verbal fluency total responses (VFTR: correct + incorrect + repetitions + ‘pass’ responses) across the six word generation blocks in the intrascan session and two blocks during the pre-scan session were calculated. Response latencies were measured during the ‘intra-scan’ session for both word generation and word repetition trials. From the mean response latencies thus obtained, we have earlier reported that the difference in mean latencies between the word generation and word repetition trials (latgen latrep) showed a significant inverse relationship with task-induced fMRI BOLD deactivation in HS, while such a relationship was not observed in the poorly performing SZ (John et al., 2011). 2.3. Statistical analysis The mean VFTR, VFCR, no responses and repetitive responses for healthy and schizophrenia subjects in both the ‘pre-scan’ and ‘intrascan’ sessions were compared across the subject groups. Since the variables were non-normally distributed, as assessed using Shapiro– Wilk’s test, non-parametric statistical test (Mann–Whitney U) was
M.K. Shakeel et al. / Asian Journal of Psychiatry 7 (2014) 22–27
25
Table 2 Behavioral performance of patients with schizophrenia and healthy control subjects on verbal fluency task in the pre-scan and intra-scan sessions.a Pre-scan
Intra-scan
Behavioral measures
Healthy subjects M (SD)
Schizophrenia subjects M (SD)
MWU
P
Effect size (r)
Healthy subjects M (SD)
Schizophrenia subjects M (SD)
MWU
P
Effect size (r)
TR CR NR PR
12.92 11.29 1.00 0.13
11.50 10.38 2.29 0.17
136b 202.5 153.5b 266.00
