This article was downloaded by: [New York University] On: 04 October 2014, At: 01:27 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Clinical and Experimental Neuropsychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ncen19
Persistent memory impairment following transient global amnesia a
John R. Hodges & Susan M. Oxbury
b
a
University Department of Clinical Neurology , The Radcliffe Infirmary, Oxford b
Department of Clinical Neuropsychology , The Radcliffe Infirmary, Oxford Published online: 04 Jan 2008.
To cite this article: John R. Hodges & Susan M. Oxbury (1990) Persistent memory impairment following transient global amnesia, Journal of Clinical and Experimental Neuropsychology, 12:6, 904-920, DOI: 10.1080/01688639008401030 To link to this article: http://dx.doi.org/10.1080/01688639008401030
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly
Downloaded by [New York University] at 01:27 04 October 2014
forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Clinical and Experimental Neumpsychology 1990, VOl. 12, NO.6,pp. 904-920
0168-8634/90/1206-O3.00 Q Swets 8 Zeitlinger
Persistent Memory Impairment Following Transient Global Amnesia" John R. Hodges University Department of Clinical Neurology
Downloaded by [New York University] at 01:27 04 October 2014
Susan M. Oxbury Department of Clinical Neuropsychology
The Radcliffe Infirmary Oxford
ABSTRACT A controlled neuropsychological study of 41 patients tested at 6 months after attacks of transient global amnesia (TGA) revealed no evidence of general intellectual, immediate (short-term) memory or nonverbal memory impairment. The patient group's performance was, however, significantly worse than that of the control's on measures of verbal memory notably immediate, 30-minute and %hour delayed paragraph recall. In addition, tests of public and personal remote memory revealed significant impairment of naming and recognition of famous faces, and of dating famous events without evidence of a temporal gradient, and impairment of cued recall of autobiographical memories on the Modified Crovitz Test. These fiidings suggest that following TGA there is persistent, albeit mild, hippocampal-diencephalic dysfunction which appears to involve left-sided structures preferentially. This impairment probably results from the attack, although a pre-existent deficit cannot be excluded.
The term transient global amnesia (TGA) was first coined by Fisher and Adams (1964) to describe a clinical syndrome, characterized by the abrupt onset of severe anterograde amnesia, usually accompanied by repetitive questioning, which occurs in the middle-aged o r elderly. During the attack patients remain alert and
* This research was supported by a grant from the Medical Research Council. We are extremely grateful to Drs Francis Marriott and Ziyah Mehta of the Oxford Department of Biomathematics and Statistics for their expert statistical advice, to Jessica Matheson and Efi Hobbs for their assistance with test administration, and David Salmon and Bill Heindel who made valuable comments on earlier versions of the paper. Address all correspondence to: John R. Hodges, M.D.. University of Cambridge Clinical School, Neurology Department, Addenbrooke's Hospital, Cambridge CB2 2QQ. United Kingdom Accepted for publication: March 12. 1990.
Downloaded by [New York University] at 01:27 04 October 2014
MEMORY AFTER TGA
905
communicative, and personal identity is preserved. Focal neurological and epileptic features are absent, and general cognition appears intact. After a few hours the ability to lay down new memories gradually recovers and subjects return to normal except for a dense amnesic gap for the duration of the attack (Caplan, 1985; Fisher 1982; Fisher & Adams, 1964; Hodges & Ward, 1989; Wilson, Koller, & Kelly, 1980). In contrast to the clinical impression that TGA is a brief self-limiting condition with rapid recovery of memory function, formal neuropsychological studies suggest that subclinical memory deficits may persist for much longer. Detailed studies of individual patients tested serially after attacks have demonstrated memory impairment for up to a month (Caffarra, Moretti, Mazzucchi & Parma, 1981; Gallassi, Lorusso, & Stracciari, 1986; Hodges & Ward, 1989; Regard & Landis, 1984). In addition, a number of authors have reported permanent memory impairment in patients studied at an interval ranging from weeks to many months after TGA (Cattaino, Querin, Pomes, & Piazza 1984; Jensen & Olivarius, 1981; Mathew & Meyer, 1974; Mazzucchi, Moretti, Caffarra, & Parma, 1980; Steinmetz & Vroom, 1972). The latter reports can, however, be criticized on a number of grounds. Firstly, in the earlier literature there is confusion between TGA and permanent amnesia of sudden onset - so-called amnestic stroke. CT scanning has led to an increased recognition of amnestic stroke syndromes, which result from bilateral thalamic infarction (Graff-Radford, Damasio, Yamada, Eslinger, & Damasio, 1985; Guberman & Stuss, 1983; Winocur, Oxbury, Roberts, Agnetti, & Davis, 1984) or occlusion of the posterior cerebral artery with resultant hippocampal damage (De Renzi, Zambolin, & Crisi, 1987). Secondly, there has been considerable variation in the use of the term TGA. Some series have included patients with a variety of focal neurological symptoms during the attack such as hemiparesis, ataxia, or impaired consciousness. Such cases would not qualify for the diagnosis as now applied (Caplan, 1985; Hinge, Jensen, Kjaer, Marquardsen, & Olivarius, 1986; Hodges & Warlow, 1990). Thirdly, the time of testing in relationship to the attack has been highly variable. Patients tested within days or weeks may still be in the recovery phase. At the other extreme, when patients were evaluated several years post-attack it is difficult to draw firm conclusions about the relationship between memory impairment and TGA. Finally, most studies reporting permanent memory impairment post-TGA have not used normal controls and as patients with TGA are elderly, a proportion may be cognitively impaired due to unrelated causes. In the only controlled study reported to date, Mazzucchi et al. (1980) compared cognitive and memory function in 16 patients with strictly defined TGA and normal controls. The patients had a significant Verbal-Performance IQ discrepancy in favor of the latter, and were impaired on all tests of verbal memory. Nonverbal memory with the exception of delayed reproduction of the Rey Complex Figure appeared intact. The authors postulated that the patients had mild left hippocampal deficit and speculated as to whether the deficit preceded and thereby
Downloaded by [New York University] at 01:27 04 October 2014
906
JOHN R. HODGES AND SUSAN M.OXBURY
predisposed to TGA, or resulted from the attack. They emphasized that further controlled studies would be required before definite conclusionscould be reached. The only criticism of this study is that the timing of the tests post-attack was highly variable, ranging from days to almost four years, so that those tested very soon post-attack may still have been in the recovery phase. Thus, there is some evidence that anterograde memory, and possibly VIQ, may be persistently impaired after TGA. However, the reported deficits may have been exaggerated for the methodological reasons discussed above. Furthermore, although anterograde memory function has been evaluated after TGA there have been no studies looking at remote memory. During TGA the anterograde amnesia is invariably associated with a retrograde deficit (Hodges CQ Ward, 1989; Kritchevsky & Squire, 1989); it is therefore quite conceivable that patients might also suffer from a permanent remote memory deficit. It is clearly important from a practical standpoint to establish whether TGA is truly transient. In addition, clarification of the neuropsychological status postattack may throw light on the site of pathology in TGA which remains unknown. The aims of the present study were to investigate anterograde and remote memory after TGA in a large group of patients at a standard time post-attack and to compare them with carefully matched normal controls. The neuropsychological battery employed was designed to include tests of verbal and spatial memory, which would be sensitiveto left- and right-sided dysfunction respectively, together with a range of tests of public and personal remote memory.
METHODS Subjects Two groups consisting of 82 subjects participated in the study: There were 41 TGA patients (28 men and 13 women, mean age 64.3 SD 8.4 years) and 41 normal control subjects (28 men and 13 women, mean age 65.5 SD 8.6 years). The patients, who were recruited into a prospective study of TGA conducred by one of the authors (JH) between 1984 and 1987, underwent neuropsychological evaluation at 6 months post-attack. This interval was chosen to allow any reversible impairment to recover under the assumption that any deficit still present at 6 months was l i e l y to be permanent and stable. All patients fulfilled strict criteria for the diagnosis of TGA. These are described in detail elsewhere (Hodges & Ward, 1989; Hodges & Warlow. 1990). Briefly, all amnesic episodes had to be witnessed and information available from a reliable observer; patients with disturbed consciousness, focal neurological features. recent head injury, or known epilepsy were excluded. All patients were tested following a single episode of TGA. The controls were selected from a larger panel of 109 community-based normal controls, used in a parallel epidemiological study (Hodges & Warlow. 1990), obtained from two local general practitioners' age/sex registers. All 109 had been given the National Adult Reading Test (NART) (Nelson & O'Connell. 1978) as an estimate of premorbid IQ. The controls and patients were matched on an individual matched-pair basis using the following criteria: sex. age (within 5 years) and FSIQ (estimated from the NART to within 5 IQ points). Subjects with known cerebral pathology or other conditions which might disqualify them as "normal" controls or interfere with testing were excluded. In all,
MEMORY AFTER TGA
907
10 potential controls from the panel of 109 were excluded; 2 were illiterate, 2 had chronic schizophrenia, 1 was a known alcoholic, 1 had severe aphasia, 1 suffered from advanced Parkinson’s Disease, 2 were crippled by rheumatological diseases and 1 was blind.
Procedure Subjects were administered the following battery of neuropsychological tests which were given in 2 sessions on consecutive days. Each session lasted approximately 2 hours.
Downloaded by [New York University] at 01:27 04 October 2014
General Intelligence Tests 1. Wechsler Adult Intelligence Test (WAIS) (Wechsler, 1955). A shortened version of the WAIS was used consisting of three verbal subtests (Similarities, Arithmetic, Digit Span) and three performance subtests (Block Design, Object Assembly, Digit Symbol). 2. National Adult Reading Test (NART) (Nelson & O’Connell, 1978). Immediate (short-term) Memory Tests 1 . Digits Forward and Backward (Wechsler, 1955). 2. Corsi Block-tapping Span (Milner, 1971). Verbal Memory Tests 1. Logical Memory (paragraph recall). Subjects’ immediate, 30-min and 24-hr delayed recall of the two stories from the Wechsler Memory Scale (Wechsler, 1945) was tested. 2. Paired Associate Learning Test (PALT). In this test, also taken from the Wechsler Memory Scale (Wechsler, 1945), immediate, 30-min and 24-hr delayed recall were tested. 3. Drilled Word List Learning Test. In this version of the test, adapted from Weintraub and Mesulam (1985), a standard 8-word list was used. The word list was read to the subjects at one word per 2 s. Immediately after the first trial subjects were asked to recall the list. If failures occurred the trials were repeated until a criterion of three consecutive correct trials was reached, or until eight trials had elapsed. Subjects were then asked to silently rehearse the words, and recall was tested after 60 s without distraction. Thereafter, subjects were distracted and recall tested, without warning, after a further 60 s, 3 min and 10 min. If failures occurred after 60 s or thereafter, subjects were again drilled until criterion was reached. Scores were thereby obtained for the number of trials to criterion, and the number of words correctly recalled at the various time intervals. Nonverbal Memory Tests The Rey-Osterrieth Complex Figure (Ostemeth, 1944). Subjects f i s t copied the figure, then 40 min later without warning, they were a k e d to reproduce the figure. The subjects’ drawings were scored for accuracy according to the criteria described by Osterrieth (1944). Corsi Block Supraspan Learning Test (De Renzi, Faglioni, & Previdi, 1977). After determination of the block-tapping span subjects were required to learned a sequence which exceeded their span by two blocks. The examiner first tapped the sequence to be learnt at a rate of one item per second. Subjects were then asked to reproduce the sequence. Trials were repeated until a criterion of two correct consecutive trials was reached or 25 trials had elapsed. The score was taken as the number of trials to the first of the two correct consecutive trials. Incidental Spatial Memory Test (Smith & Milner, 1981). Sixteen small toys, each representing a common object (e.g., washing machine, hair brush, watering can, etc.), were placed in a fixed random array on a 60 cm square board. Subjects were asked to point to each item, to name the item and then after 10 s had elapsed to estimate the
908
JOHN R. HODGES AND SUSAN M.OXBURY
Downloaded by [New York University] at 01:27 04 October 2014
price of the real objects represented by the toy. The time restraint was used to ensure equal attention to each item. Two practice items were given before the test proper. The following measures were obtained: (i) recall of object identity - upon finishing the price estimation, the subject was seated with his back to the board and was distracted by a counting task. Meanwhile the experimenter removed the toys from the board. Four minutes after the end of the price estimation subjects were asked to name as many of the objects as possible, within 1 minute. (ii) Recall of object location immediately upon finishing the object recall task subjects were given all 16 toys and asked to replace them in their original positions. Two minutes were allowed. The mean object displacement was calculated as the distance between the position assigned to an object during recall and its original position. Twenty-four hours later (Day 2). without warning the tests of object identity and location tests were repeated. Remote Memory Tests 1. Famous Faces Test. This test was based on the famous faces test of Albert, Butters, and Levin (1979) and is described elsewhere (Hodges & Ward,1989). One hundred photographs of public figures who had come to prominence between 1920 and 1980 were used. The photographs were presented so that each block of ten contained photographs from all decades. Naming, identification and forced-choice recognition were tested. Subjects were first asked to name the famous person represented. If the correct name was produced they moved on to the next item. If incorrect or unnamed, the subject proceeded to the second and third parts. In the second part (identification) they were asked to provide a detailed description of the famous person represented. No clues were given but the examiner probed for further details. Responses were scored as either nonidentifying broad category responses (e.g., “an actor’’ or “a politician”) or identifying responses with specific details (e.g.. “the Tory politician who resigned during the Suez Crisis’’ etc.). In the third part (forcedchoice recognition) subjects chose from three aurally presented alternative names from the same category and era as the test item. Results were expressed as the percentage correct per decade for naming. identification (naming or accurate identification) and forced-choice recognition. 2. Famous Events Test (Hodges & Ward, 1989). The test stimuli consisted of 50 clearly datable real events (10 per decade from the 1930s-1970s) and 50 fictitious events made-up to sound like real news events (e.g., “The Jin-Jin”, “The Vienna Airlift” etc.). The real and fictitious items were mixed so that each block of 10 items contained news items from each decade with a varying number of intervening fictitious items. Subjects were read each item in turn and then asked to signify those items remembered as real events. They were told beforehand that the list contained a mixture of genuine news items from the past 50 years and fictitious items. Whenever subjects identified an event as real they were asked to date it by assigning it to a single decade. Guessing was encouraged. The following measures were obtained: (i) recognition of events the proportion of correctly identified true items (true positives) per decade, (ii) to adjust for guessing a d‘ score was calculated by subtracting the total false positive responses (i.e., fictitious items identified as real) from the total number of true positive responses, (iii) dating - the proportion of correctly dated items from each decade, (iv) errors in dating were designated negative if an underestimate was given (e.g., the Suez Crisis 1950s dated as 1940s etc.) or positive if an overestimate was given (e.g.. Suez crisis 1950s dated as 1960s etc.). 3 . Modified Crovitz Test of remote autobiographical memory. This test was based on the method of Sagar. Cohen, Sullivan, Corkin, and Growden (1988). Subjects were asked to relate personally experienced events from any remote time period involving each of 10 high-frequency noun cues (bird, flag, car, ship etc.) and then to estimate when that event occurred. Subjects were given up to 4 min to respond including if necessary 2 min with specific cues. In the frst 2 min the examiner prompted with nonspecific
Downloaded by [New York University] at 01:27 04 October 2014
MEMORY m
R TGA
909
prompts (e.g., “Tell me more”, “Can you think of an instance?”). Responses were scored on a 0 to 3-point scale depending upon the degree of specificity to time and place of the memory recalled in accordance with the criteria previously described (Hodges & Ward, 1989). Subjects who achieved a score of 0 or 1 in the first 2 min were given a further 2 min with cues, consisting of examples of specific memories (e.g., “Have you ever been on a boar?”). Subjects were then asked to date any memories that earned a score of 3. Twenty-four hours later (Day 2). without warning, each cue word was re-presented and subjects were required to reproduce the previous day’s memory. Responses were scored on the same 0- to 3-point scale for memories that were wholly or in part reproduced on Day 2. If after 2 min they failed to regain their Day 1 score they were cued with key words from their Day 1 memories, and were given a further 2 min to respond. Subjects were again asked to date memories which earned a score of 3. Total scores for memory specificity cued and uncued were thereby obtained for Day 1 and Day 2. Statistical methods For the standard Intelligence, anterograde memory tests and the Famous Faces and Famous Events Tests measurement followed an interval scale and the data were normally distributed. Comparisons were therefore made using repeated measures design analysis of variance (ANOVAs) for related groups in which the condition was always considered as the repeated measure. Since measurement of event specificity on the Crovitz Test followed an ordinal scale and the distribution of data points did not conform to the normal distribution, nonparametric statistics (Wilcoxon’s matched-pairs signed-ranktest) were used to compare the groups’ performance (Armitage, 1971). Because not all subjects completed all tests, in order to maintain matching both the patient and their matched control were dropped from the appropriate analysis whenever data on either was unavailable.
RESULTS Complete data for NART, for Verbal, Performance and Full Scale IQs, and for the WAIS subtests are shown in Table 1. The control group had slightly higher mean IQ scores than the TGA patients but the differences were not significant (ps >.05).For both groups the WAIS IQ scores exceeded the NART scores by very similar amounts, showing that the NART had underestimated IQ. A 2 (groups) X 6 (conditions) ANOVA of the WAIS subtest data revealed a significant main effect for conditions [F (5,40) = 14.18, p
Journal of Clinical and Experimental Neuropsychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ncen19
Persistent memory impairment following transient global amnesia a
John R. Hodges & Susan M. Oxbury
b
a
University Department of Clinical Neurology , The Radcliffe Infirmary, Oxford b
Department of Clinical Neuropsychology , The Radcliffe Infirmary, Oxford Published online: 04 Jan 2008.
To cite this article: John R. Hodges & Susan M. Oxbury (1990) Persistent memory impairment following transient global amnesia, Journal of Clinical and Experimental Neuropsychology, 12:6, 904-920, DOI: 10.1080/01688639008401030 To link to this article: http://dx.doi.org/10.1080/01688639008401030
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly
Downloaded by [New York University] at 01:27 04 October 2014
forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Clinical and Experimental Neumpsychology 1990, VOl. 12, NO.6,pp. 904-920
0168-8634/90/1206-O3.00 Q Swets 8 Zeitlinger
Persistent Memory Impairment Following Transient Global Amnesia" John R. Hodges University Department of Clinical Neurology
Downloaded by [New York University] at 01:27 04 October 2014
Susan M. Oxbury Department of Clinical Neuropsychology
The Radcliffe Infirmary Oxford
ABSTRACT A controlled neuropsychological study of 41 patients tested at 6 months after attacks of transient global amnesia (TGA) revealed no evidence of general intellectual, immediate (short-term) memory or nonverbal memory impairment. The patient group's performance was, however, significantly worse than that of the control's on measures of verbal memory notably immediate, 30-minute and %hour delayed paragraph recall. In addition, tests of public and personal remote memory revealed significant impairment of naming and recognition of famous faces, and of dating famous events without evidence of a temporal gradient, and impairment of cued recall of autobiographical memories on the Modified Crovitz Test. These fiidings suggest that following TGA there is persistent, albeit mild, hippocampal-diencephalic dysfunction which appears to involve left-sided structures preferentially. This impairment probably results from the attack, although a pre-existent deficit cannot be excluded.
The term transient global amnesia (TGA) was first coined by Fisher and Adams (1964) to describe a clinical syndrome, characterized by the abrupt onset of severe anterograde amnesia, usually accompanied by repetitive questioning, which occurs in the middle-aged o r elderly. During the attack patients remain alert and
* This research was supported by a grant from the Medical Research Council. We are extremely grateful to Drs Francis Marriott and Ziyah Mehta of the Oxford Department of Biomathematics and Statistics for their expert statistical advice, to Jessica Matheson and Efi Hobbs for their assistance with test administration, and David Salmon and Bill Heindel who made valuable comments on earlier versions of the paper. Address all correspondence to: John R. Hodges, M.D.. University of Cambridge Clinical School, Neurology Department, Addenbrooke's Hospital, Cambridge CB2 2QQ. United Kingdom Accepted for publication: March 12. 1990.
Downloaded by [New York University] at 01:27 04 October 2014
MEMORY AFTER TGA
905
communicative, and personal identity is preserved. Focal neurological and epileptic features are absent, and general cognition appears intact. After a few hours the ability to lay down new memories gradually recovers and subjects return to normal except for a dense amnesic gap for the duration of the attack (Caplan, 1985; Fisher 1982; Fisher & Adams, 1964; Hodges & Ward, 1989; Wilson, Koller, & Kelly, 1980). In contrast to the clinical impression that TGA is a brief self-limiting condition with rapid recovery of memory function, formal neuropsychological studies suggest that subclinical memory deficits may persist for much longer. Detailed studies of individual patients tested serially after attacks have demonstrated memory impairment for up to a month (Caffarra, Moretti, Mazzucchi & Parma, 1981; Gallassi, Lorusso, & Stracciari, 1986; Hodges & Ward, 1989; Regard & Landis, 1984). In addition, a number of authors have reported permanent memory impairment in patients studied at an interval ranging from weeks to many months after TGA (Cattaino, Querin, Pomes, & Piazza 1984; Jensen & Olivarius, 1981; Mathew & Meyer, 1974; Mazzucchi, Moretti, Caffarra, & Parma, 1980; Steinmetz & Vroom, 1972). The latter reports can, however, be criticized on a number of grounds. Firstly, in the earlier literature there is confusion between TGA and permanent amnesia of sudden onset - so-called amnestic stroke. CT scanning has led to an increased recognition of amnestic stroke syndromes, which result from bilateral thalamic infarction (Graff-Radford, Damasio, Yamada, Eslinger, & Damasio, 1985; Guberman & Stuss, 1983; Winocur, Oxbury, Roberts, Agnetti, & Davis, 1984) or occlusion of the posterior cerebral artery with resultant hippocampal damage (De Renzi, Zambolin, & Crisi, 1987). Secondly, there has been considerable variation in the use of the term TGA. Some series have included patients with a variety of focal neurological symptoms during the attack such as hemiparesis, ataxia, or impaired consciousness. Such cases would not qualify for the diagnosis as now applied (Caplan, 1985; Hinge, Jensen, Kjaer, Marquardsen, & Olivarius, 1986; Hodges & Warlow, 1990). Thirdly, the time of testing in relationship to the attack has been highly variable. Patients tested within days or weeks may still be in the recovery phase. At the other extreme, when patients were evaluated several years post-attack it is difficult to draw firm conclusions about the relationship between memory impairment and TGA. Finally, most studies reporting permanent memory impairment post-TGA have not used normal controls and as patients with TGA are elderly, a proportion may be cognitively impaired due to unrelated causes. In the only controlled study reported to date, Mazzucchi et al. (1980) compared cognitive and memory function in 16 patients with strictly defined TGA and normal controls. The patients had a significant Verbal-Performance IQ discrepancy in favor of the latter, and were impaired on all tests of verbal memory. Nonverbal memory with the exception of delayed reproduction of the Rey Complex Figure appeared intact. The authors postulated that the patients had mild left hippocampal deficit and speculated as to whether the deficit preceded and thereby
Downloaded by [New York University] at 01:27 04 October 2014
906
JOHN R. HODGES AND SUSAN M.OXBURY
predisposed to TGA, or resulted from the attack. They emphasized that further controlled studies would be required before definite conclusionscould be reached. The only criticism of this study is that the timing of the tests post-attack was highly variable, ranging from days to almost four years, so that those tested very soon post-attack may still have been in the recovery phase. Thus, there is some evidence that anterograde memory, and possibly VIQ, may be persistently impaired after TGA. However, the reported deficits may have been exaggerated for the methodological reasons discussed above. Furthermore, although anterograde memory function has been evaluated after TGA there have been no studies looking at remote memory. During TGA the anterograde amnesia is invariably associated with a retrograde deficit (Hodges CQ Ward, 1989; Kritchevsky & Squire, 1989); it is therefore quite conceivable that patients might also suffer from a permanent remote memory deficit. It is clearly important from a practical standpoint to establish whether TGA is truly transient. In addition, clarification of the neuropsychological status postattack may throw light on the site of pathology in TGA which remains unknown. The aims of the present study were to investigate anterograde and remote memory after TGA in a large group of patients at a standard time post-attack and to compare them with carefully matched normal controls. The neuropsychological battery employed was designed to include tests of verbal and spatial memory, which would be sensitiveto left- and right-sided dysfunction respectively, together with a range of tests of public and personal remote memory.
METHODS Subjects Two groups consisting of 82 subjects participated in the study: There were 41 TGA patients (28 men and 13 women, mean age 64.3 SD 8.4 years) and 41 normal control subjects (28 men and 13 women, mean age 65.5 SD 8.6 years). The patients, who were recruited into a prospective study of TGA conducred by one of the authors (JH) between 1984 and 1987, underwent neuropsychological evaluation at 6 months post-attack. This interval was chosen to allow any reversible impairment to recover under the assumption that any deficit still present at 6 months was l i e l y to be permanent and stable. All patients fulfilled strict criteria for the diagnosis of TGA. These are described in detail elsewhere (Hodges & Ward, 1989; Hodges & Warlow. 1990). Briefly, all amnesic episodes had to be witnessed and information available from a reliable observer; patients with disturbed consciousness, focal neurological features. recent head injury, or known epilepsy were excluded. All patients were tested following a single episode of TGA. The controls were selected from a larger panel of 109 community-based normal controls, used in a parallel epidemiological study (Hodges & Warlow. 1990), obtained from two local general practitioners' age/sex registers. All 109 had been given the National Adult Reading Test (NART) (Nelson & O'Connell. 1978) as an estimate of premorbid IQ. The controls and patients were matched on an individual matched-pair basis using the following criteria: sex. age (within 5 years) and FSIQ (estimated from the NART to within 5 IQ points). Subjects with known cerebral pathology or other conditions which might disqualify them as "normal" controls or interfere with testing were excluded. In all,
MEMORY AFTER TGA
907
10 potential controls from the panel of 109 were excluded; 2 were illiterate, 2 had chronic schizophrenia, 1 was a known alcoholic, 1 had severe aphasia, 1 suffered from advanced Parkinson’s Disease, 2 were crippled by rheumatological diseases and 1 was blind.
Procedure Subjects were administered the following battery of neuropsychological tests which were given in 2 sessions on consecutive days. Each session lasted approximately 2 hours.
Downloaded by [New York University] at 01:27 04 October 2014
General Intelligence Tests 1. Wechsler Adult Intelligence Test (WAIS) (Wechsler, 1955). A shortened version of the WAIS was used consisting of three verbal subtests (Similarities, Arithmetic, Digit Span) and three performance subtests (Block Design, Object Assembly, Digit Symbol). 2. National Adult Reading Test (NART) (Nelson & O’Connell, 1978). Immediate (short-term) Memory Tests 1 . Digits Forward and Backward (Wechsler, 1955). 2. Corsi Block-tapping Span (Milner, 1971). Verbal Memory Tests 1. Logical Memory (paragraph recall). Subjects’ immediate, 30-min and 24-hr delayed recall of the two stories from the Wechsler Memory Scale (Wechsler, 1945) was tested. 2. Paired Associate Learning Test (PALT). In this test, also taken from the Wechsler Memory Scale (Wechsler, 1945), immediate, 30-min and 24-hr delayed recall were tested. 3. Drilled Word List Learning Test. In this version of the test, adapted from Weintraub and Mesulam (1985), a standard 8-word list was used. The word list was read to the subjects at one word per 2 s. Immediately after the first trial subjects were asked to recall the list. If failures occurred the trials were repeated until a criterion of three consecutive correct trials was reached, or until eight trials had elapsed. Subjects were then asked to silently rehearse the words, and recall was tested after 60 s without distraction. Thereafter, subjects were distracted and recall tested, without warning, after a further 60 s, 3 min and 10 min. If failures occurred after 60 s or thereafter, subjects were again drilled until criterion was reached. Scores were thereby obtained for the number of trials to criterion, and the number of words correctly recalled at the various time intervals. Nonverbal Memory Tests The Rey-Osterrieth Complex Figure (Ostemeth, 1944). Subjects f i s t copied the figure, then 40 min later without warning, they were a k e d to reproduce the figure. The subjects’ drawings were scored for accuracy according to the criteria described by Osterrieth (1944). Corsi Block Supraspan Learning Test (De Renzi, Faglioni, & Previdi, 1977). After determination of the block-tapping span subjects were required to learned a sequence which exceeded their span by two blocks. The examiner first tapped the sequence to be learnt at a rate of one item per second. Subjects were then asked to reproduce the sequence. Trials were repeated until a criterion of two correct consecutive trials was reached or 25 trials had elapsed. The score was taken as the number of trials to the first of the two correct consecutive trials. Incidental Spatial Memory Test (Smith & Milner, 1981). Sixteen small toys, each representing a common object (e.g., washing machine, hair brush, watering can, etc.), were placed in a fixed random array on a 60 cm square board. Subjects were asked to point to each item, to name the item and then after 10 s had elapsed to estimate the
908
JOHN R. HODGES AND SUSAN M.OXBURY
Downloaded by [New York University] at 01:27 04 October 2014
price of the real objects represented by the toy. The time restraint was used to ensure equal attention to each item. Two practice items were given before the test proper. The following measures were obtained: (i) recall of object identity - upon finishing the price estimation, the subject was seated with his back to the board and was distracted by a counting task. Meanwhile the experimenter removed the toys from the board. Four minutes after the end of the price estimation subjects were asked to name as many of the objects as possible, within 1 minute. (ii) Recall of object location immediately upon finishing the object recall task subjects were given all 16 toys and asked to replace them in their original positions. Two minutes were allowed. The mean object displacement was calculated as the distance between the position assigned to an object during recall and its original position. Twenty-four hours later (Day 2). without warning the tests of object identity and location tests were repeated. Remote Memory Tests 1. Famous Faces Test. This test was based on the famous faces test of Albert, Butters, and Levin (1979) and is described elsewhere (Hodges & Ward,1989). One hundred photographs of public figures who had come to prominence between 1920 and 1980 were used. The photographs were presented so that each block of ten contained photographs from all decades. Naming, identification and forced-choice recognition were tested. Subjects were first asked to name the famous person represented. If the correct name was produced they moved on to the next item. If incorrect or unnamed, the subject proceeded to the second and third parts. In the second part (identification) they were asked to provide a detailed description of the famous person represented. No clues were given but the examiner probed for further details. Responses were scored as either nonidentifying broad category responses (e.g., “an actor’’ or “a politician”) or identifying responses with specific details (e.g.. “the Tory politician who resigned during the Suez Crisis’’ etc.). In the third part (forcedchoice recognition) subjects chose from three aurally presented alternative names from the same category and era as the test item. Results were expressed as the percentage correct per decade for naming. identification (naming or accurate identification) and forced-choice recognition. 2. Famous Events Test (Hodges & Ward, 1989). The test stimuli consisted of 50 clearly datable real events (10 per decade from the 1930s-1970s) and 50 fictitious events made-up to sound like real news events (e.g., “The Jin-Jin”, “The Vienna Airlift” etc.). The real and fictitious items were mixed so that each block of 10 items contained news items from each decade with a varying number of intervening fictitious items. Subjects were read each item in turn and then asked to signify those items remembered as real events. They were told beforehand that the list contained a mixture of genuine news items from the past 50 years and fictitious items. Whenever subjects identified an event as real they were asked to date it by assigning it to a single decade. Guessing was encouraged. The following measures were obtained: (i) recognition of events the proportion of correctly identified true items (true positives) per decade, (ii) to adjust for guessing a d‘ score was calculated by subtracting the total false positive responses (i.e., fictitious items identified as real) from the total number of true positive responses, (iii) dating - the proportion of correctly dated items from each decade, (iv) errors in dating were designated negative if an underestimate was given (e.g., the Suez Crisis 1950s dated as 1940s etc.) or positive if an overestimate was given (e.g.. Suez crisis 1950s dated as 1960s etc.). 3 . Modified Crovitz Test of remote autobiographical memory. This test was based on the method of Sagar. Cohen, Sullivan, Corkin, and Growden (1988). Subjects were asked to relate personally experienced events from any remote time period involving each of 10 high-frequency noun cues (bird, flag, car, ship etc.) and then to estimate when that event occurred. Subjects were given up to 4 min to respond including if necessary 2 min with specific cues. In the frst 2 min the examiner prompted with nonspecific
Downloaded by [New York University] at 01:27 04 October 2014
MEMORY m
R TGA
909
prompts (e.g., “Tell me more”, “Can you think of an instance?”). Responses were scored on a 0 to 3-point scale depending upon the degree of specificity to time and place of the memory recalled in accordance with the criteria previously described (Hodges & Ward, 1989). Subjects who achieved a score of 0 or 1 in the first 2 min were given a further 2 min with cues, consisting of examples of specific memories (e.g., “Have you ever been on a boar?”). Subjects were then asked to date any memories that earned a score of 3. Twenty-four hours later (Day 2). without warning, each cue word was re-presented and subjects were required to reproduce the previous day’s memory. Responses were scored on the same 0- to 3-point scale for memories that were wholly or in part reproduced on Day 2. If after 2 min they failed to regain their Day 1 score they were cued with key words from their Day 1 memories, and were given a further 2 min to respond. Subjects were again asked to date memories which earned a score of 3. Total scores for memory specificity cued and uncued were thereby obtained for Day 1 and Day 2. Statistical methods For the standard Intelligence, anterograde memory tests and the Famous Faces and Famous Events Tests measurement followed an interval scale and the data were normally distributed. Comparisons were therefore made using repeated measures design analysis of variance (ANOVAs) for related groups in which the condition was always considered as the repeated measure. Since measurement of event specificity on the Crovitz Test followed an ordinal scale and the distribution of data points did not conform to the normal distribution, nonparametric statistics (Wilcoxon’s matched-pairs signed-ranktest) were used to compare the groups’ performance (Armitage, 1971). Because not all subjects completed all tests, in order to maintain matching both the patient and their matched control were dropped from the appropriate analysis whenever data on either was unavailable.
RESULTS Complete data for NART, for Verbal, Performance and Full Scale IQs, and for the WAIS subtests are shown in Table 1. The control group had slightly higher mean IQ scores than the TGA patients but the differences were not significant (ps >.05).For both groups the WAIS IQ scores exceeded the NART scores by very similar amounts, showing that the NART had underestimated IQ. A 2 (groups) X 6 (conditions) ANOVA of the WAIS subtest data revealed a significant main effect for conditions [F (5,40) = 14.18, p
