- Email: [email protected]
Span of apprehension deficits in older outpatients with schizophrenia
SCHIZOPHRENIA RESEARCH ELSEVIER
Schizophrenia Research 20 (1996) 51 56
Span of apprehension deficits in older outpatients with schizophrenia Eric Granholm a,,, Robert F. Asarnow b, Steven P. Verney ¢, Peter Nelson a, Dilip V. Jested a Department of Psychiatry, University of California, San Diego Psychology Service, San Diego VA Medical Center ( l16B), 3350 La Jolla Village Drive, San Diego, CA 92161, USA b Department of Psychiatry andBiobehavioral Sciences, University of California, Los Angeles, CA, USA c San Diego State University~University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA, USA d Departments of Psychiatry and Neurosciences, University of California, San Diego Psychiatry Service, San Diego VA Medical Center ( l16B), 3350 La Jolla Village Drive, San Diego, CA 92161, USA Received 24 July 1995; revision 14 November 1995; accepted 22 November 1995
Abstract
Performance on the span of apprehension task, a well-studied information processing task in schizophrenia research, was examined in 11 schizophrenia patients and 11 normal comparison participants, all over the age of 45 years. Subjects detected 'T' and 'F' targets in briefly-flashed arrays of 1, 6, and 12 letters on the span task. Consistent with previously reported findings in younger schizophrenia patients, the older patients detected significantly fewer targets in the larger (12-letter), but not smaller (1-, or 6-letter), arrays. The older schizophrenia patients also showed significantly slower reaction times in all array-size conditions. Neither age of onset nor duration of illness was significantly correlated with span task performance. The characteristic span of apprehension task deficit found in the older schizophrenia patients suggests that late-life schizophrenia shares a common cognitive impairment with childhood and young adulthood schizophrenia, and provides supportive evidence for a possible stable vulnerability trait deficit in schizophrenia that is independent of age of onset and duration of illness. Keywords: Psychosis; Information-processing; Aging; Cognition; Attention
1. Introduction
Impaired performance on the partial-report span of apprehension (SOA) task has been frequently reported in schizophrenia patients and provides evidence of information-processing deficits in schizophrenia (for a review, see Asarnow et al., * Corresponding author. E-mail: [email protected]. 0920-9964/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved SSDI 0920-9964(95)00106-9
1991). On the SOA task, participants must report whether specific target letters are among a group of distractor letters flashed briefly in a visual display. SOA task studies have shown that younger adults (age 20-40) and children (age 6-15, with onset of illness before age 12) with schizophrenia detect fewer target stimuli than nonpsychiatric controls as well as other psychiatric patients with 8-, 10- or 12-letter arrays, but not with smaller
52
Eric Granholm et al./Schizophrenia Research 20 (1996) 51 56
array sizes (Neale et al., 1969; Asarnow et al., 1991). Impairments on the SOA task have also been observed in schizophrenia patients in a remitted state, nonpsychotic biological mothers of schizophrenia patients, foster children whose biological mothers have schizophrenia, and individuals in the general population with schizotypal personality characteristics (see Asarnow et al., 1991). Although some conflicting results have been reported (Harvey et al., 1985; Strauss et al., 1987), the majority of the studies suggest that the impaired performance on the SOA task may reflect subtle underlying stable trait and/or genetic vulnerability marker of schizophrenia (Nuechterlein and Dawson, 1984; Asarnow et al., 1991). Information processing in older patients with schizophrenia has been examined in only a few studies (Saccuzzo, 1977; McDowd et al., 1993; O'Donnell et al., 1995). Late-life schizophrenia has been a neglected but important condition for study (Jeste, 1993). Many of the informationprocessing impairments found in schizophrenia are also commonly found in normal aging (Saccuzzo, 1977). Information-processing abilities in older patients with schizophrenia may differ from those of younger patients not only because of aging but also due to chronicity of illness and treatment. If impairment on the SOA task reflects a stable vulnerability trait in schizophrenia, then older schizophrenia patients should show the characteristic SOA task impairment in larger array-size conditions, when compared to similarly aged and educated normal participants, and this SOA impairment should not correlate with age of onset or duration of illness. The present study examined this hypothesis by comparing the SOA task performance of middle-aged and elderly schizophrenia patients and normal comparison participants.
2. Methods
2.1. Participants Twenty-two ambulatory, community-dwelling, older (>45 years) participants signed a written informed consent to participate in the study. These included 11 patients with schizophrenia and 11
normal comparison subjects. All the participants were recruited from the Clinical Research Center on Late-Life Psychosis and Antipsychotics at the University of California, San Diego. The Structured Clinical Interview for DSM-III-R or SCID (Spitzer et al., 1990) was used to make DSM-III-R diagnoses (American Psychiatric Association, 1987). Subjects from both groups were excluded for past or present major head injuries (i.e., with loss of consciousness > 30 min), serious medical problems, neurological disorders (e.g., stroke; dementia; seizures), alcohol or substance abuse or dependence (SCID/DSM-III-R criteria) during the past 6 months, or other Axis I or II disorders, and were also screened for at least 20/50 vision using a Snellen wall chart. Nine schizophrenia patients were receiving antipsychotic medications (mean chlorpromazine equivalent 483.3mg daily; SD=990.9) (Jeste and Wyatt, 1982) and six patients were also receiving anticholinergic medications in low dosages (typically 2 mg/day of benztropine mesylate). Participants were evaluated by 'blind' raters using an 18-item version of the Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham, 1962), with items rated from 1 (absent) to 7 (severe), the Scale for the Assessment of Positive Symptoms (SAPS) (Andreasen and Olsen, 1982), with global scores summed (range=0-20) for hallucinations, delusions, bizarre behavior and thought disorder, and the Scale for the Assessment of Negative Symptoms (SANS) (Andreasen and Olsen, 1982), with global scores summed (range=0-25) for affective flattening, alogia, avolition/apathy, anhedonia/asociality and attention.
2.2. Partial-report span of apprehension task Subjects sat with their head stabilized in a chinrest facing a Dell 333 computer screen at a distance of 77 cm. Subjects were told that either a 'T' or an 'F' would be flashed briefly in a group of other letters on a computer screen and that, using the index finger of their dominant hand, they should press one of the two buttons on a joystick box corresponding to the correct target letter as quickly as possible. A central fixation point remained on the screen during a five-second inter-
53
Eric Granholm et al./Schizophrenia Research 20 (1996) 51-56
trial interval. The target stimulus was embedded in arrays containing either 0 (l-letter condition), 5 (6-letter condition), or I 1 (12-letter condition) distractor letters, which were randomly selected from all the other letters of the English alphabet. Fifty trials of each array-size condition were presented in counterbalanced blocks of 10 trials per block in the following sequence: 1, 6, 12, 1, 12, 6, 6, 12, 1, 6, 1, 12, 12, 1, 6. Distractor and target letters were randomly assigned to locations in a 5 × 5 matrix, with two factors constraining array construction: an equal number of T's and F's were presented within each array-size condition; and targets appeared an equal number of times in each matrix location (one F and one T per location) within each condition. The visual angle subtended by each element in the array was 5.2°× 5.4 ° and by the entire 5 × 5 matrix was 33.4°× 28.3 °. The display duration was 70 ms. Percentage of targets correctly detected and median reaction time were calculated for each subject within each array-size condition.
2.3. Statistical analyses A separate split-plot (group × array size) analysis of variance (ANOVA) was carried out on the detection accuracy and reaction time scores on the SOA task. Bonferroni-corrected, two-tailed t-tests were used for all post-hoc comparisons. Spearman rank-order correlations were performed between detection accuracy on the SOA task and certain clinical characteristics only within the schizophrenia group.
3. Results
As shown in Table 1, the two groups did not differ significantly with regard to age, education, or gender. Table 1 also shows age of onset of illness (defined according to DSM-III-R as the age of onset of prodromal symptoms or functional decline), and symptom ratings on the BPRS, SAPS and SANS. For the accuracy data, significant effects were found for array size (F(2,40)=95.94, p<0.001), group (F(1,20) = 7.64, p < 0.05), and for the group-
Table 1 Demographic and clinical characteristics Variable
Normal comparison participants (n=ll)
Schizophrenia patients (n=ll)
Age (years) Education (years) Gender (% male) Age of onset of schizophrenia (years) BPRS totaP SAPS totaP SANS totaP
65.3 (5.1) 13.3 (2.0) 64
59.0 (9.0) 12.6 (2.4) 73 28.4 (11.2)
20.5 (2.6) 0.82 (0.8) 1.4 (2.2)
30.7 (10.2) b 4.6 (4.4) b 8.2 (3.3) b
The values (except for gender) represent means (with standard deviations). ap<0.05 (two-tailed t-test). bn = 10.
by-array size interaction (F(2,40)= 3.70, p<0.05) (Fig. 1). Between-group comparisons showed that the schizophrenia patients detected significantly fewer targets than the normal comparison participants only in the 12-letter condition (t(20)= 3.19, p<0.05). For the reaction time data, significant effects were found for array size (F(2,40)=75.18, p < 0.001 ), and for group (F(1,20) = 6.54, p < 0.05), but not for the interaction between group and array size (Fig. 2). Reaction time scores for all 22 participants (i.e., collapsing across groups) showed faster scores for l-letter arrays than for 6-letter (t(21)= - 11.65, p<0.01), and 12-letter (t(21)= 10.18, p < 0.01) arrays, but the difference between 6 - and 12-letter arrays was not significant. Schizophrenia patients showed significantly longer reaction times than the normal participants for all array sizes. Correlations were performed for only the schizophrenia patients to examine the relationships between certain patient characteristics and detection accuracy on the SOA task. Due to the small sample size and significant skew in the clinical ratings on several psychopathology scale scores, Spearman nonparametric rank-order correlation coefficient (rs) was used. To reduce the number of correlations carried out, only SOA task accuracy in the 12-letter array-size condition was used in all
54
Eric Granholm et al./Schizophrenia Research 20 (1996) 51-56 100
-
95
~, - Late-Life Schizophrenia Patients
]
• _ pNO~tin~ia~aCtsmP aris°n
90
85
80 0 75 o~
u
B_
70
65
60
55
50 I
6
12
Array Size Fig. I. Percentage of correctly detected targets for l-, 6-, and 12-letter arrays on the span of apprehension task for schizophrenia patients and normal comparison subjects. See the text for statistical analyses.
1300
~- ~ - "
1200
1100
/
//
/
/
,/ 1000
1/
E
i.T.
900
//
/
/
800
70O -- -e -- Late-Life Schizophrenia Patients • Normal Comparison Participants
600
500
I
1
ArraSyize 6
: 12
Fig. 2. Average median reaction time for l-, 6-, and 12-letter arrays on the span of apprehension task for schizophrenia patients and normal comparison subjects. See the text for statistical analyses.
Eric Granholm et at~Schizophrenia Research 20 (1996) 51 56
correlations, because this was the only condition that significantly discriminated between the two groups. The SOA scores did not correlate significantly with age of onset ( r s = - 0 . 2 9 , p=0.412), duration of illness (r s = -0.02, p = 0.947), psychopathology ratings on the BPRS ( r s = - 0 . 1 9 , p = 0.604), SAPS (rs = -0.11, p=0.762), and SANS ( r s = - 0 . 2 3 , p=0.518) or daily neuroleptic dose (mg chlorpromazine equivalent) (rs=0.36, p = 0.310).
4. Discussion
The partial-report SOA task was used to investigate information-processing impairment in middleaged and elderly patients with schizophrenia. Consistent with numerous previously reported findings in younger schizophrenia patients (for a review, see Asarnow et al., 1991), the older schizophrenia patients in this study demonstrated impaired detection accuracy in larger, but not smaller, array-size conditions when compared to the normal participants with similar age and education level. In addition, the correlations between SOA task performance and age of onset and duration of illness were not significant; indeed the correlation between the SOA task detection accuracy and age of onset of illness was negative. These findings are consistent with neuropsychological studies showing that the basic cognitive impairment in schizophrenia is largely independent of age and chronicity of illness (Heaton et al., 1994). Taken together, these findings are more consistent with the hypothesis that basic information-processing impairment in schizophrenia is a stable trait of the disease, rather than a degenerative (or synergistic) impairment that interacts with aging and duration of illness. The reaction time data may provide some insight into the nature of information-processing deficits in schizophrenia. In previous research using the SOA task (Asarnow et al., 1991; Granholm et al., 1996), we have hypothesized that schizophrenia patients may be slow to initiate their covert scanning of the array elements or carry out cognitive operations more slowly within each scan move (e.g., slowed rate of comparison of items in
55
memory with display-set items), and, hence, do not complete their search for targets within the time limits of the iconic store. If the cognitive operations involved in covert search were carried out slower within each scan move, reaction times would have increased disproportionately more for schizophrenia patients in the larger array conditions, which require more scan moves (Granholm et al., 1996). Yet, the schizophrenia patients showed slower reaction times overall, and a significant group by array size interaction was not found for reaction time. This uniform slowing of reaction times across array-size conditions is more consistent with the hypothesis that the schizophrenia patients are slower to initiate search or are slower to terminate search effectively (e.g., may be slow to organize and initiate their responsesl. The correlations revealed that the SOA task impairment could not be attributed to neuroleptic medications. On the contrary, a higher daily dosage of neuroleptic was moderately (but nonsignificantly) associated with better SOA task performance. This finding is similar to several other reports that neuroleptic medications may enhance the performance of schizophrenia patients on certain attention and information-processing tasks (Spohn and Strauss, 1989). The severity SOA task impairment did not correlate significantly with the severity or type (positive or negative) of psychiatric symptoms. This study has several limitations. The schizophrenia sample was skewed toward higher functioning outpatients and may not generalize to more severely ill inpatients. We did not use a longitudinal design or younger comparison participants, which would have strengthened interpretations regarding whether information-processing impairment in schizophrenia is best characterized as a stable trait or a degenerative process associated with aging and chronicity. In addition, different versions of the SOA task are not equally sensitive to impairments in schizophrenia. For example, full-report versions (Cash et al., 1972) and wider-visual-angle versions (Granholm et al., 1996) may be less sensitive to impairment in schizophrenia. The sample size was small. Interpretations of negative findings (e.g., nonsignificant correlations between symptom ratings or age of onset, and SOA task
56
Eric Granholm et al./Schizophrenia Research 20 (1996) 51 56
performance) should, therefore, be viewed cautiously. On the other hand, this was a difficult sample to obtain, because the participants were well-characterized older patients and normal comparison participants free of any major medical or neurologic illness and current substance abuse. We believe that the research strategy of examining schizophrenia in relationship to aging used in this report and other studies (Saccuzzo, 1977; McDowd et al., 1993; O'Donnell et al., 1995) may prove to be a productive approach to examining developmental, trait/state, and disease course issues (e.g., whether vulnerability markers change with age: progressive cognitive deterioration versus stable impairment hypothesis in schizophrenia). Cross-sectional studies comparing younger versus older patients and normal comparison subjects, as well as longitudinal investigations are needed to examine the interactions, if any, between aging and schizophrenia more directly.
Acknowledgment Portions of this research were presented at the 148th Annual Meeting of the American Psychiatric Association in Miami, Florida, May, 1995. This research was supported, in part, by NIMH Grants MH49671, MH43693 and MH45131, the Scottish Rite Benevolent Foundation's Schizophrenia Research Program, NMJ, USA, the National Alliance for Research on Schizophrenia and Depression (NARSAD), and by the Department of Veterans Affairs.
References American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. American Psychiatric Press, Washington, DC. Andreasen NC, Olsen S (1982) Negative vs positive schizophre-
nia: definition and validation. Arch Gen Psychiatry 39, 789-794. Asarnow RF, Granholm E, Sherman T (1991) Span of apprehension in schizophrenia. In: Steinhauer S, Gruzelier JH, Zubin J (eds) Handbook of Schizophrenia, Vol. 4, Neuropsychology, Psychophysiology, and Information Processing. Elsevier, Amsterdam, pp 335-370. Cash TF, Neale JM, Cromwell RL (1972) Span of apprehension in acute schizophrenics: full-report technique. J Abnorm Psychol 79, 322-326. Granholm EL, Asarnow RF, Marder SR (1995) Display visual angle and attentional scanpaths on the span of apprehension task in schizophrenia. J Abnorm Psychol 105, 17-24. Harvey PD, Weintraub S, Neale JM (1985) Span of apprehension deficits in children vulnerable to psychopathology: a failure to replicate. J Abnorm Psychol 94, 410 413, Heaton R, Paulsen J, McAdams LA, Kuck J, Zisook S, Braff D, Harris M J, Jeste DV (1994) Neuropsychological deficits in schizophrenia: relationship to age, chronicity and dementia. Arch Gen Psychiatry 51,469 476. Jeste DV (1993) Late-life schizophrenia: Editor's introduction. Schizophr Bull 19, 687-689. Jeste DV, Wyatt RJ (1982) Understanding and Treating Tardive Dyskinesia. Guilford Press, New York. McDowd JM, Filion DL, Harris J, Braff DL (1993) Sensory gating and inhibitory function in late-life schizophrenia. Schizophr Bull 19(4), 733 746. Neale JM, Mclntyre CW, Fox R, Cromwell RL (1969) Span of apprehension in acute schizophrenia. J Abnorm Psychol 74, 593-599. Nuechterlein KH, Dawson ME (1984) Information processing and attentional functioning in the course of schizophrenic disorder. Schizophr Bull 10, 160-203. O'Donnell BF, Faux SF, McCarley RW, Kimble MO, Salisbury DF, Nestor PG, Kikinis R, Jolesz FA, Shenton ME (1995) Increased rate of P300 latency prolongation with age in schizophrenia. Electrophysiological evidence for a neurodegenerative process. Arch Gen Psychiatry 52, 544-549. Overall JE, Gorharn DR (1962) The Brief Psychiatric Rating Scale. Psychol Rep 10, 799 812. Saccuzzo DP (1977) Bridges between schizophrenia and gerontology: generalized or specific deficits? Schizophr Bull 84, 595--600. Spitzer RL, Williams JBW, Gibbon M, First MB (1990) User's Guide for the Structured Clinical Interview for DSM-III-R. American Psychiatric Press, Washington, DC. Spohn HE, Strauss ME (1989) Relation of neuroleptic and anticholinergic medication to cognitive functions in schizophrenia. J Abnorm Psychol 98, 367 380. Strauss ME, Prescott CA, Gutterman DF, Tune LE (1987) Span of apprehension deficits in schizophrenia and mania. Schizophr Bull 13, 699-704.
Schizophrenia Research 20 (1996) 51 56
Span of apprehension deficits in older outpatients with schizophrenia Eric Granholm a,,, Robert F. Asarnow b, Steven P. Verney ¢, Peter Nelson a, Dilip V. Jested a Department of Psychiatry, University of California, San Diego Psychology Service, San Diego VA Medical Center ( l16B), 3350 La Jolla Village Drive, San Diego, CA 92161, USA b Department of Psychiatry andBiobehavioral Sciences, University of California, Los Angeles, CA, USA c San Diego State University~University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA, USA d Departments of Psychiatry and Neurosciences, University of California, San Diego Psychiatry Service, San Diego VA Medical Center ( l16B), 3350 La Jolla Village Drive, San Diego, CA 92161, USA Received 24 July 1995; revision 14 November 1995; accepted 22 November 1995
Abstract
Performance on the span of apprehension task, a well-studied information processing task in schizophrenia research, was examined in 11 schizophrenia patients and 11 normal comparison participants, all over the age of 45 years. Subjects detected 'T' and 'F' targets in briefly-flashed arrays of 1, 6, and 12 letters on the span task. Consistent with previously reported findings in younger schizophrenia patients, the older patients detected significantly fewer targets in the larger (12-letter), but not smaller (1-, or 6-letter), arrays. The older schizophrenia patients also showed significantly slower reaction times in all array-size conditions. Neither age of onset nor duration of illness was significantly correlated with span task performance. The characteristic span of apprehension task deficit found in the older schizophrenia patients suggests that late-life schizophrenia shares a common cognitive impairment with childhood and young adulthood schizophrenia, and provides supportive evidence for a possible stable vulnerability trait deficit in schizophrenia that is independent of age of onset and duration of illness. Keywords: Psychosis; Information-processing; Aging; Cognition; Attention
1. Introduction
Impaired performance on the partial-report span of apprehension (SOA) task has been frequently reported in schizophrenia patients and provides evidence of information-processing deficits in schizophrenia (for a review, see Asarnow et al., * Corresponding author. E-mail: [email protected]. 0920-9964/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved SSDI 0920-9964(95)00106-9
1991). On the SOA task, participants must report whether specific target letters are among a group of distractor letters flashed briefly in a visual display. SOA task studies have shown that younger adults (age 20-40) and children (age 6-15, with onset of illness before age 12) with schizophrenia detect fewer target stimuli than nonpsychiatric controls as well as other psychiatric patients with 8-, 10- or 12-letter arrays, but not with smaller
52
Eric Granholm et al./Schizophrenia Research 20 (1996) 51 56
array sizes (Neale et al., 1969; Asarnow et al., 1991). Impairments on the SOA task have also been observed in schizophrenia patients in a remitted state, nonpsychotic biological mothers of schizophrenia patients, foster children whose biological mothers have schizophrenia, and individuals in the general population with schizotypal personality characteristics (see Asarnow et al., 1991). Although some conflicting results have been reported (Harvey et al., 1985; Strauss et al., 1987), the majority of the studies suggest that the impaired performance on the SOA task may reflect subtle underlying stable trait and/or genetic vulnerability marker of schizophrenia (Nuechterlein and Dawson, 1984; Asarnow et al., 1991). Information processing in older patients with schizophrenia has been examined in only a few studies (Saccuzzo, 1977; McDowd et al., 1993; O'Donnell et al., 1995). Late-life schizophrenia has been a neglected but important condition for study (Jeste, 1993). Many of the informationprocessing impairments found in schizophrenia are also commonly found in normal aging (Saccuzzo, 1977). Information-processing abilities in older patients with schizophrenia may differ from those of younger patients not only because of aging but also due to chronicity of illness and treatment. If impairment on the SOA task reflects a stable vulnerability trait in schizophrenia, then older schizophrenia patients should show the characteristic SOA task impairment in larger array-size conditions, when compared to similarly aged and educated normal participants, and this SOA impairment should not correlate with age of onset or duration of illness. The present study examined this hypothesis by comparing the SOA task performance of middle-aged and elderly schizophrenia patients and normal comparison participants.
2. Methods
2.1. Participants Twenty-two ambulatory, community-dwelling, older (>45 years) participants signed a written informed consent to participate in the study. These included 11 patients with schizophrenia and 11
normal comparison subjects. All the participants were recruited from the Clinical Research Center on Late-Life Psychosis and Antipsychotics at the University of California, San Diego. The Structured Clinical Interview for DSM-III-R or SCID (Spitzer et al., 1990) was used to make DSM-III-R diagnoses (American Psychiatric Association, 1987). Subjects from both groups were excluded for past or present major head injuries (i.e., with loss of consciousness > 30 min), serious medical problems, neurological disorders (e.g., stroke; dementia; seizures), alcohol or substance abuse or dependence (SCID/DSM-III-R criteria) during the past 6 months, or other Axis I or II disorders, and were also screened for at least 20/50 vision using a Snellen wall chart. Nine schizophrenia patients were receiving antipsychotic medications (mean chlorpromazine equivalent 483.3mg daily; SD=990.9) (Jeste and Wyatt, 1982) and six patients were also receiving anticholinergic medications in low dosages (typically 2 mg/day of benztropine mesylate). Participants were evaluated by 'blind' raters using an 18-item version of the Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham, 1962), with items rated from 1 (absent) to 7 (severe), the Scale for the Assessment of Positive Symptoms (SAPS) (Andreasen and Olsen, 1982), with global scores summed (range=0-20) for hallucinations, delusions, bizarre behavior and thought disorder, and the Scale for the Assessment of Negative Symptoms (SANS) (Andreasen and Olsen, 1982), with global scores summed (range=0-25) for affective flattening, alogia, avolition/apathy, anhedonia/asociality and attention.
2.2. Partial-report span of apprehension task Subjects sat with their head stabilized in a chinrest facing a Dell 333 computer screen at a distance of 77 cm. Subjects were told that either a 'T' or an 'F' would be flashed briefly in a group of other letters on a computer screen and that, using the index finger of their dominant hand, they should press one of the two buttons on a joystick box corresponding to the correct target letter as quickly as possible. A central fixation point remained on the screen during a five-second inter-
53
Eric Granholm et al./Schizophrenia Research 20 (1996) 51-56
trial interval. The target stimulus was embedded in arrays containing either 0 (l-letter condition), 5 (6-letter condition), or I 1 (12-letter condition) distractor letters, which were randomly selected from all the other letters of the English alphabet. Fifty trials of each array-size condition were presented in counterbalanced blocks of 10 trials per block in the following sequence: 1, 6, 12, 1, 12, 6, 6, 12, 1, 6, 1, 12, 12, 1, 6. Distractor and target letters were randomly assigned to locations in a 5 × 5 matrix, with two factors constraining array construction: an equal number of T's and F's were presented within each array-size condition; and targets appeared an equal number of times in each matrix location (one F and one T per location) within each condition. The visual angle subtended by each element in the array was 5.2°× 5.4 ° and by the entire 5 × 5 matrix was 33.4°× 28.3 °. The display duration was 70 ms. Percentage of targets correctly detected and median reaction time were calculated for each subject within each array-size condition.
2.3. Statistical analyses A separate split-plot (group × array size) analysis of variance (ANOVA) was carried out on the detection accuracy and reaction time scores on the SOA task. Bonferroni-corrected, two-tailed t-tests were used for all post-hoc comparisons. Spearman rank-order correlations were performed between detection accuracy on the SOA task and certain clinical characteristics only within the schizophrenia group.
3. Results
As shown in Table 1, the two groups did not differ significantly with regard to age, education, or gender. Table 1 also shows age of onset of illness (defined according to DSM-III-R as the age of onset of prodromal symptoms or functional decline), and symptom ratings on the BPRS, SAPS and SANS. For the accuracy data, significant effects were found for array size (F(2,40)=95.94, p<0.001), group (F(1,20) = 7.64, p < 0.05), and for the group-
Table 1 Demographic and clinical characteristics Variable
Normal comparison participants (n=ll)
Schizophrenia patients (n=ll)
Age (years) Education (years) Gender (% male) Age of onset of schizophrenia (years) BPRS totaP SAPS totaP SANS totaP
65.3 (5.1) 13.3 (2.0) 64
59.0 (9.0) 12.6 (2.4) 73 28.4 (11.2)
20.5 (2.6) 0.82 (0.8) 1.4 (2.2)
30.7 (10.2) b 4.6 (4.4) b 8.2 (3.3) b
The values (except for gender) represent means (with standard deviations). ap<0.05 (two-tailed t-test). bn = 10.
by-array size interaction (F(2,40)= 3.70, p<0.05) (Fig. 1). Between-group comparisons showed that the schizophrenia patients detected significantly fewer targets than the normal comparison participants only in the 12-letter condition (t(20)= 3.19, p<0.05). For the reaction time data, significant effects were found for array size (F(2,40)=75.18, p < 0.001 ), and for group (F(1,20) = 6.54, p < 0.05), but not for the interaction between group and array size (Fig. 2). Reaction time scores for all 22 participants (i.e., collapsing across groups) showed faster scores for l-letter arrays than for 6-letter (t(21)= - 11.65, p<0.01), and 12-letter (t(21)= 10.18, p < 0.01) arrays, but the difference between 6 - and 12-letter arrays was not significant. Schizophrenia patients showed significantly longer reaction times than the normal participants for all array sizes. Correlations were performed for only the schizophrenia patients to examine the relationships between certain patient characteristics and detection accuracy on the SOA task. Due to the small sample size and significant skew in the clinical ratings on several psychopathology scale scores, Spearman nonparametric rank-order correlation coefficient (rs) was used. To reduce the number of correlations carried out, only SOA task accuracy in the 12-letter array-size condition was used in all
54
Eric Granholm et al./Schizophrenia Research 20 (1996) 51-56 100
-
95
~, - Late-Life Schizophrenia Patients
]
• _ pNO~tin~ia~aCtsmP aris°n
90
85
80 0 75 o~
u
B_
70
65
60
55
50 I
6
12
Array Size Fig. I. Percentage of correctly detected targets for l-, 6-, and 12-letter arrays on the span of apprehension task for schizophrenia patients and normal comparison subjects. See the text for statistical analyses.
1300
~- ~ - "
1200
1100
/
//
/
/
,/ 1000
1/
E
i.T.
900
//
/
/
800
70O -- -e -- Late-Life Schizophrenia Patients • Normal Comparison Participants
600
500
I
1
ArraSyize 6
: 12
Fig. 2. Average median reaction time for l-, 6-, and 12-letter arrays on the span of apprehension task for schizophrenia patients and normal comparison subjects. See the text for statistical analyses.
Eric Granholm et at~Schizophrenia Research 20 (1996) 51 56
correlations, because this was the only condition that significantly discriminated between the two groups. The SOA scores did not correlate significantly with age of onset ( r s = - 0 . 2 9 , p=0.412), duration of illness (r s = -0.02, p = 0.947), psychopathology ratings on the BPRS ( r s = - 0 . 1 9 , p = 0.604), SAPS (rs = -0.11, p=0.762), and SANS ( r s = - 0 . 2 3 , p=0.518) or daily neuroleptic dose (mg chlorpromazine equivalent) (rs=0.36, p = 0.310).
4. Discussion
The partial-report SOA task was used to investigate information-processing impairment in middleaged and elderly patients with schizophrenia. Consistent with numerous previously reported findings in younger schizophrenia patients (for a review, see Asarnow et al., 1991), the older schizophrenia patients in this study demonstrated impaired detection accuracy in larger, but not smaller, array-size conditions when compared to the normal participants with similar age and education level. In addition, the correlations between SOA task performance and age of onset and duration of illness were not significant; indeed the correlation between the SOA task detection accuracy and age of onset of illness was negative. These findings are consistent with neuropsychological studies showing that the basic cognitive impairment in schizophrenia is largely independent of age and chronicity of illness (Heaton et al., 1994). Taken together, these findings are more consistent with the hypothesis that basic information-processing impairment in schizophrenia is a stable trait of the disease, rather than a degenerative (or synergistic) impairment that interacts with aging and duration of illness. The reaction time data may provide some insight into the nature of information-processing deficits in schizophrenia. In previous research using the SOA task (Asarnow et al., 1991; Granholm et al., 1996), we have hypothesized that schizophrenia patients may be slow to initiate their covert scanning of the array elements or carry out cognitive operations more slowly within each scan move (e.g., slowed rate of comparison of items in
55
memory with display-set items), and, hence, do not complete their search for targets within the time limits of the iconic store. If the cognitive operations involved in covert search were carried out slower within each scan move, reaction times would have increased disproportionately more for schizophrenia patients in the larger array conditions, which require more scan moves (Granholm et al., 1996). Yet, the schizophrenia patients showed slower reaction times overall, and a significant group by array size interaction was not found for reaction time. This uniform slowing of reaction times across array-size conditions is more consistent with the hypothesis that the schizophrenia patients are slower to initiate search or are slower to terminate search effectively (e.g., may be slow to organize and initiate their responsesl. The correlations revealed that the SOA task impairment could not be attributed to neuroleptic medications. On the contrary, a higher daily dosage of neuroleptic was moderately (but nonsignificantly) associated with better SOA task performance. This finding is similar to several other reports that neuroleptic medications may enhance the performance of schizophrenia patients on certain attention and information-processing tasks (Spohn and Strauss, 1989). The severity SOA task impairment did not correlate significantly with the severity or type (positive or negative) of psychiatric symptoms. This study has several limitations. The schizophrenia sample was skewed toward higher functioning outpatients and may not generalize to more severely ill inpatients. We did not use a longitudinal design or younger comparison participants, which would have strengthened interpretations regarding whether information-processing impairment in schizophrenia is best characterized as a stable trait or a degenerative process associated with aging and chronicity. In addition, different versions of the SOA task are not equally sensitive to impairments in schizophrenia. For example, full-report versions (Cash et al., 1972) and wider-visual-angle versions (Granholm et al., 1996) may be less sensitive to impairment in schizophrenia. The sample size was small. Interpretations of negative findings (e.g., nonsignificant correlations between symptom ratings or age of onset, and SOA task
56
Eric Granholm et al./Schizophrenia Research 20 (1996) 51 56
performance) should, therefore, be viewed cautiously. On the other hand, this was a difficult sample to obtain, because the participants were well-characterized older patients and normal comparison participants free of any major medical or neurologic illness and current substance abuse. We believe that the research strategy of examining schizophrenia in relationship to aging used in this report and other studies (Saccuzzo, 1977; McDowd et al., 1993; O'Donnell et al., 1995) may prove to be a productive approach to examining developmental, trait/state, and disease course issues (e.g., whether vulnerability markers change with age: progressive cognitive deterioration versus stable impairment hypothesis in schizophrenia). Cross-sectional studies comparing younger versus older patients and normal comparison subjects, as well as longitudinal investigations are needed to examine the interactions, if any, between aging and schizophrenia more directly.
Acknowledgment Portions of this research were presented at the 148th Annual Meeting of the American Psychiatric Association in Miami, Florida, May, 1995. This research was supported, in part, by NIMH Grants MH49671, MH43693 and MH45131, the Scottish Rite Benevolent Foundation's Schizophrenia Research Program, NMJ, USA, the National Alliance for Research on Schizophrenia and Depression (NARSAD), and by the Department of Veterans Affairs.
References American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. American Psychiatric Press, Washington, DC. Andreasen NC, Olsen S (1982) Negative vs positive schizophre-
nia: definition and validation. Arch Gen Psychiatry 39, 789-794. Asarnow RF, Granholm E, Sherman T (1991) Span of apprehension in schizophrenia. In: Steinhauer S, Gruzelier JH, Zubin J (eds) Handbook of Schizophrenia, Vol. 4, Neuropsychology, Psychophysiology, and Information Processing. Elsevier, Amsterdam, pp 335-370. Cash TF, Neale JM, Cromwell RL (1972) Span of apprehension in acute schizophrenics: full-report technique. J Abnorm Psychol 79, 322-326. Granholm EL, Asarnow RF, Marder SR (1995) Display visual angle and attentional scanpaths on the span of apprehension task in schizophrenia. J Abnorm Psychol 105, 17-24. Harvey PD, Weintraub S, Neale JM (1985) Span of apprehension deficits in children vulnerable to psychopathology: a failure to replicate. J Abnorm Psychol 94, 410 413, Heaton R, Paulsen J, McAdams LA, Kuck J, Zisook S, Braff D, Harris M J, Jeste DV (1994) Neuropsychological deficits in schizophrenia: relationship to age, chronicity and dementia. Arch Gen Psychiatry 51,469 476. Jeste DV (1993) Late-life schizophrenia: Editor's introduction. Schizophr Bull 19, 687-689. Jeste DV, Wyatt RJ (1982) Understanding and Treating Tardive Dyskinesia. Guilford Press, New York. McDowd JM, Filion DL, Harris J, Braff DL (1993) Sensory gating and inhibitory function in late-life schizophrenia. Schizophr Bull 19(4), 733 746. Neale JM, Mclntyre CW, Fox R, Cromwell RL (1969) Span of apprehension in acute schizophrenia. J Abnorm Psychol 74, 593-599. Nuechterlein KH, Dawson ME (1984) Information processing and attentional functioning in the course of schizophrenic disorder. Schizophr Bull 10, 160-203. O'Donnell BF, Faux SF, McCarley RW, Kimble MO, Salisbury DF, Nestor PG, Kikinis R, Jolesz FA, Shenton ME (1995) Increased rate of P300 latency prolongation with age in schizophrenia. Electrophysiological evidence for a neurodegenerative process. Arch Gen Psychiatry 52, 544-549. Overall JE, Gorharn DR (1962) The Brief Psychiatric Rating Scale. Psychol Rep 10, 799 812. Saccuzzo DP (1977) Bridges between schizophrenia and gerontology: generalized or specific deficits? Schizophr Bull 84, 595--600. Spitzer RL, Williams JBW, Gibbon M, First MB (1990) User's Guide for the Structured Clinical Interview for DSM-III-R. American Psychiatric Press, Washington, DC. Spohn HE, Strauss ME (1989) Relation of neuroleptic and anticholinergic medication to cognitive functions in schizophrenia. J Abnorm Psychol 98, 367 380. Strauss ME, Prescott CA, Gutterman DF, Tune LE (1987) Span of apprehension deficits in schizophrenia and mania. Schizophr Bull 13, 699-704.