- Email: [email protected]
Cognitive dysfunction in individuals with myasthenia gravis
Journal of the Neurological Sciences 179 (2000) 59–64 www.elsevier.com / locate / jns
Cognitive dysfunction in individuals with myasthenia gravis a, a a a b Robert H. Paul *, Ronald A. Cohen , James M. Gilchrist , Mark S. Aloia , Jonathan M. Goldstein a
Brown University School of Medicine, Miriam Hospital, 164 Summit Ave., Providence, RI 02906, USA b Yale University School of Medicine, Yale, NY, USA Received 23 February 2000; received in revised form 8 May 2000; accepted 6 July 2000
Abstract In the present study we administered a battery of cognitive measures that examined attention, response fluency, information processing, and verbal and visual learning and retention to 28 individuals with generalized myasthenia gravis (MG) and 18 demographically similar control subjects. Results revealed that MG patients performed significantly more poorly than control subjects on the measures of response fluency, information processing and most measures of verbal and visual learning. Significant group differences were not evident on the measure of attention span or on the indices of retention of information. Cognitive performances of the MG group were not related to mood disturbance, disease duration, or daily dose of prednisone. While these results suggest central involvement in MG, previous studies have not provided evidence that MG antibodies bind to central nicotinic receptors. Possible alternative mechanisms underlying cognitive dysfunction in MG are discussed. 2000 Elsevier Science B.V. All rights reserved. Keywords: Myasthenia gravis; Neuropsychology; Memory; Acetycholine
1. Introduction Myasthenia gravis (MG) is an autoimmune disease that involves antibody-mediated destruction of nicotinic cholinergic receptors. The typical clinical presentation of MG is associated with the prominent role of nicotinic receptors at the motor endplate of striated muscles. Specifically, MG patients report increased fatigability of voluntary muscles that worsens with exercise and attenuates with rest [1]. Most commonly, muscles of the eyes, face, neck, arms, and trunk are affected, although some patients also experience weakness of the legs and the diaphragm [1,2]. In addition to the muscular symptoms of the disease, several studies have identified central abnormalities in MG. Specifically, alterations in EEG patterns in both MG patients and animals with experimental autoimmune myasthenia gravis (EAMG) [3,4] have been reported, as well as *Corresponding author. Tel.: 11-401-793-4383; fax: 11-401-7519873. E-mail address: robert [email protected] (R.H. Paul). ]
alterations in brain stem auditory evoked potentials in individuals with MG [5]. Several studies [6–8] have reported an increased prevalence of seizures in MG compared to the general population, however, this finding has not been consistently replicated and the vast majority of MG patients do not present with abnormal ictal activity [9,10]. The central abnormalities described above have been attributed to dysfunction of central nicotinic receptor systems, however, there is limited evidence that central cholinergic receptors are affected in MG [9]. Nearly 60% of MG patients complain of cognitive difficulties [10], but few studies have empirically examined cognition in MG. Of the controlled studies that have been reported, results have not been consistent. The first controlled study of cognition in MG [11] revealed significant deficits by MG patients compared to healthy control subjects and a medical control group (non-neurological chronic diseases) on immediate and delayed recall of verbal information, but there was no group difference in the ability to retain information over time. Similar findings have been reported by Iwasaki et al. [12,13] using immediate and delayed trials of memory for prose passages.
0022-510X / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 00 )00367-1
60
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
Results from these studies revealed significant impairments among MG patients on both trials of the memory test. Impairments of visual learning and memory have also been reported in MG using measures of visual working memory, nonvisual spatial exploration, and performance in a human radial-arm maze [14]. In several studies of cognition in MG, performance has been shown to improve following plasmapheresis, suggesting that circulating antibodies to the nicotinic receptor might mediate cognitive dysfunction in MG [12,15]. In contrast to the investigations noted above, other studies have found only subtle, if any, deficits among MG patients on measures of cognitive function compared to healthy control samples. Specifically, Bartel and Lotz [16] examined cognition in MG subjects and healthy control subjects using the Randt Memory test and measures of simple and sustained attention, information processing speed, motor speed, and arithmetic. Results of the study revealed a significant group difference only on information processing speed. Other groups also have found no cognitive impairment among MG patients on measures of verbal learning and memory [17]. The conflicting results regarding cognition in MG may be due to methodological differences of previous studies. In many investigations [11–15], the effects of possible visual limitations (e.g. diplopia, opthalmoplegia), use of prednisone, or significant mood disturbance were not controlled, each of which may have resulted in significantly poorer performance on measures of cognition among MG patients than healthy control subjects. In other studies [16–18], inclusion of few subjects and / or patients with different subtypes of MG may have minimized group differences on cognitive tasks. The purpose of the present study was to examine cognitive performance of patients with MG after controlling for the methodological limitations of the previous studies described above. A total of 28 patients with generalized MG and 18 healthy control subjects were recruited for the study. Subjects were administered measures of attention, response fluency, information processing, and verbal and visual learning and retention.
2. Methods
2.1. Subjects A sample of 28 individuals with generalized MG were recruited for participation in the study. Patients were selfselected volunteers from regional chapters of the National Myasthenia Gravis Foundation and local neurologists. Diagnoses were confirmed by a positive Tensilon Test, electrodiagnositc recordings consistent with MG, or evidence of antinicotinic antibodies in serum. Patients were excluded if they reported a history of alcohol or drug abuse, loss of consciousness greater than 5 min, significant
cardiovascular disease, major psychiatric illness (e.g. schizophrenia) or additional neurological illness. Patients were also excluded if their corrected vision was worse than 20 / 70 or if they evidenced opthalmoplegia. Disease severity was determined using a Likert scale ranging from 1 to 3, where 15minimal interference of daily activities, 25moderate interference with daily activities despite medications, and 35significant interference of daily activities despite use of medications. The scale was treated as a continuous variable, thus permitting scores of 1.5, 2.5, etc. Patients in the present sample averaged 54. 71613.35 years of age (range529–77), 15.3962.73 years of education (range512–20), 7.7466.71 years since diagnosis, and 1.6460.67 on the disease severity scale (mild to moderate). A total of 16 (57%) patients reported experiencing cognitive difficulties as a regular symptom of the disease. This percentage is nearly identical to previously reported rates of self-described cognitive dysfunction in MG [10]. All patients were regularly taking medications to treat MG including Mestinon (n521), Imuran (n511) and prednisone (n513; median daily dose516.0 mg, range 2.5–60 mg). All patients were classified as Group II (i.e. generalized MG) according to modified Osserman criteria. A sample of 18 healthy control subjects were recruited from the community. Control subjects were similar in age and education and were required to meet the same exclusion criteria as the patient group. Control subjects in the present study averaged 51.16615.43 years of age (range5 29–78) and 16.462.57 years of education (range512–20). All subjects provided written informed consent after receiving a comprehensive explanation of the study. No subjects were paid for participation.
2.2. Procedure Study participants were provided the opportunity to be tested in our laboratory, in their own home, or at a public library. Nearly all subjects in both groups elected for home testing. Distractions during home testing were minimized by testing in a quiet location. All subjects were initially administered clinical and demographic questionnaires, followed by a visual acuity test to rule out severe visual disturbances, and a measure of mood. Subsequently, subjects were administered the cognitive battery, which required approximately 1.5 h to complete. To minimize the effects of fatigue developed over the course of testing, subjects were provided breaks from testing at 10-min intervals throughout the evaluation. In addition, the cognitive battery was administered in a counterbalanced order.
2.3. Measures Chicago Multiscale Depression Inventory (CMDI) [19]. This is a 42-item self-report scale of depression. The inventory assesses different aspects of depression including mood, self-evaluative, and vegetative symptoms. The
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
inventory was developed to assess mood symptoms of depression independent of vegetative symptoms, as the latter may artificially inflate the prevalence of depression in MG in neuroimmune disorders [19]. The mood subscale of the CMDI has good reliability and validity and correlates highly with standard clinical measures of depression [19]. The mood subscale includes 14 items and requires subjects to rate their degree of agreement using a Likert scale where 15no problem and 45extreme problem.
2.4. Cognitive battery 2.4.1. Attention span Digit Span Test [20]. Digits Forward and Backward were administered in the standard manner. The total number of correctly repeated trials for each version was the dependent measure. 2.4.2. Response fluency Letter Fluency: Subjects were required to generate words that began with the letters F, A, and S. Sixty seconds were allowed for each letter and proper nouns were not allowed. The total number of correct words generated across the three trials was the dependent measure. Animal Fluency: Subjects were required to name as many animals as quickly as possible for 60 s. The total number of animals generated was the dependent variable. 2.4.3. Information processing Oral version of the Symbol Digit Modalities Test (SDMT) [21]. Subjects were required to orally substitute numbers for symbols using a key held in constant view. The total number of correct substitutions made in 90 s was recorded as the dependent measure. 2.4.4. Verbal learning and retention The California Verbal Learning Test (CVLT) [22] was administered in standard fashion. Briefly, the test involves oral presentation of a 16-word list. The list items represent exemplars of one of four categories (fruit, clothing, tools, spices / herbs). Subjects were read the list five times and asked to recall as many words as possible from the whole list on each of the trials. Subjects were then read a distracter list, followed by free and cued recalls of the original list. After a 20-min delay, subjects were asked to free recall the original list, followed by a cued recall trial, and then a recognition trial. The total number of words recalled across the five learning trials was the dependent measure for verbal learning. Percent retained from the fifth learning trial to the long delay free recall trial and number correct on the recognition trial served as the dependent measures for verbal retention. 2.4.5. Visual learning and retention The Visual Reproductions I and II (VRI and VRII)
61
subtests from the Wechsler Memory Scale-Revised [23] were administered in standard fashion. Briefly, subjects were presented four stimulus cards, one at a time, each card depicting a geometric design. Subjects were asked to look at each design for 10 s, after which the card was removed from sight and subjects were required to draw the design from memory. Following a 20-min delay period, subjects were again asked to draw the designs from memory. Scoring was performed according to standard protocol. Total correct for the first trial served as the dependent measure of visual learning. Percent retained from the learning trial to the delayed trial served as the dependent measure of visual retention. The Rey Complex Figure Test [24] was also administered as a measure of visual learning and retention. Subjects were instructed to copy the geometric figure as accurately as possible. Immediately after completing the copy trial, subjects were asked to draw the figure from memory. Following a 20-min time delay, subjects were again instructed to draw the figure from memory. No time limit was given and the drawings were scored according to standardized criteria [25]. The dependent measure for visual learning was the total score on the immediate recall trial and the dependent measure for visual retention was the total score on the delayed trial.
3. Results The two groups did not differ significantly by age or education (Fs ,2). Similarly, on the mood subscale of the CMDI, the patient group reported a higher average score, but the difference between groups was not statistically significant (F52.58, P.0.05).
3.1. Attention On the Digit Span test, the two groups performed equally well on Digits Forward (F(1,44)50.86, P.0.05) and Digits Backward (1,44)50.26, P.0.05) (see Table 1).
3.2. Response fluency Performances on the measures of response fluency were significantly different by group. Specifically, the patient group generated fewer words than the control subjects on both letter fluency (F(1,44)519.08, P,0.05) and category fluency (F(1,44)58.84, P,0.05).
3.3. Information processing On the SDMT, the patient group performed significantly more poorly than the control group (F(1,44)511.37, P, 0.05).
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
62
Table 1 Performances on cognitive measures for patients and controls: Mean (SD) Measure
MG patients
Control subjects
Attention span Digits forward Digits backward
9.62 (1.88) 8.22 (2.06)
9.11 (1.74) 7.88 (2.21)
Response fluency Letter fluency Animal fluency
37.66 (12.31) 18.25 (4.61)
52.38 (8.85)** 22.33 (4.32)**
Information Processing Symbol digit modalities test
49.85 (12.46)
61.33 (8.89)**
Verbal learning and retention CVLT (5 trials; max580) Retention (long delay / trial 5) CVLT Recognition (max516)
48.40 (9.76) 0.75 (0.92) 14.83 (1.54)
57.72 (8.30)** 0.86 (0.98) 14.31 (1.28)
Visual learning and retention Visual reproductions I (max542) Retention (VRII / VRI) Rey immediate (max536) Rey delayed (max536)
33.96 0.81 17.00 15.94
37.50 0.87 19.22 18.30
(6.07) (0.20) (8.64) (7.48)
(2.74)* (0.16) (6.63) (6.68)
*P,0.05. **P,0.01.
3.4. Verbal learning and retention The two groups also differed significantly on the measure of verbal learning, as the patient group recalled fewer words across the five learning trials of the CVLT compared to the control subjects (F(1,44)511.03, P, 0.05). By contrast, the percent retained from the fifth learning trial of the CVLT to the long-delay free recall trial did not differ significantly by group (F,3). Similarly, the two groups did not differ significantly on the recognition trial of the CVLT (F,2).
3.5. Visual learning and retention On the learning trial of the Visual Reproductions test, the patient group performed significantly more poorly than the control group (F(1,44)55.34, P,0.05). By contrast, the percent retained from the learning trial to the delayed trial did not differ by group (F51). On the Rey Complex Figure, the two groups did not differ on the copy, immediate, or delayed trials (Fs ,3). Multiple regression analyses were conducted to examine the amount of variance in cognitive function accounted for by clinical variables. Time since diagnosis, disease severity, daily dose of prednisone, and the mood subscale of the CMDI was entered as a clinical variables for these analyses. The dependent variables were performances on each of the measures of cognition. Results of these analyses revealed a significant relationship between disease seventy and performance on the SDMT (R50.385, F(1,25)54.35, P,0.05). Cognitive performance was not associated with any of the other variables. For prednisone,
the greatest R value was 0.253, P.0.05 (letter fluency), while for mood the greatest R value was 0.233, P.0.05 (animal fluency).
4. Discussion Cognitive dysfunction in MG has been reported in previous investigations examining the neurobehavioral aspects of the illness, however methodological issues produced considerable limitations to these studies. After controlling for many of these concerns in the present study, MG patients were found to exhibit mild, but significant difficulty compared to control subjects on several measures of cognitive function. Specifically, while MG patients performed as well as control subjects on measures of attention span and retention of verbal and visual information, the patient group exhibited significantly poorer performance on measures of response fluency, information processing and verbal and visual learning. Importantly, these difficulties were not associated with mood disturbance or with daily dose of prednisone. It should noted, however, that the severity of these cognitive difficulties is mild and not likely to significantly affect daily function. The observation that MG patients performed as well as control subjects on Digit Span but not on the SDMT test indicates that attention span is intact in MG, while rapid processing of information is impaired. The poorer performance by the MG group on the fluency measures also likely reflects impaired information processing given the limited time constraints allowed for completion of these tasks. On the learning and memory tasks, the MG group performed as well as controls on measures that allowed self-determined exposure to the stimuli (i.e. Rey Complex Figure), but not on measures that involved limited exposure of stimuli. These findings raise the possibility that MG patients were less capable of efficiently processing information within the time constraints imposed by the CVLT and Visual Reproductions test. In addition, while MG patients typically recalled significantly less information on both the learning and delayed trials of memory tasks, they did not lose more information over time than control subjects, a pattern that is inconsistent with the pattern typical of neurodegenerative diseases involving central cholinergic systems (e.g. Dementia of the Alzheimer’s type) [26]. As this study represents the first examination of cognition in MG to simultaneously address issues such as medications, gross visual impairments and mood, replication of the current findings using an independent sample represents the most valuable follow-up to this study. Until these results are confirmed, speculation regarding the etiology of cognitive dysfunction in patients with the disease seems premature. Nevertheless, consideration of potential etiological mechanisms will help direct future
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
studies, and therefore it is appropriate to briefly mention these possibilities at the present time. In most previous studies, cognitive impairments in MG have been attributed to the effects of antibody-mediated destruction of central nicotinic receptors. However, Keesey [9] recently reviewed the literature on this topic, and based largely on the different stoichiometry of the four nicotinic receptor subtypes, provided a convincing argument against the possibility of central nicotinic dysfunction in MG. We agree with this position, and add that the lack of involvement of autonomic nicotinic receptors in the disease provides further evidence that these receptors are not uniformly affected. Another possible etiological mechanism that has been suggested by previous research [9,27] concerns the diaphragm, which like many striated muscles, is affected by the binding of MG antibodies to nicotinic receptors. In MG, many patients experience breathing abnormalities during sleep, when the diaphragm is directly involved in the regulation of airflow [28,29]. Although cognitive difficulties are common consequences of reduced oxygen saturation during sleep, only one study has examined sleep and cognition in MG [27]. Results from this study revealed no significant relationship between cognition and percent of sleep with oxygen saturation below 93% using regression analyses. However, when the patient group was subdivided into those with and without oxygen desaturation during sleep, comparisons revealed a significant difference on one of two memory tests. Unfortunately, additional analyses addressing how these two groups may have differed by age, education, mood status, or use of medications were not provided. Clearly, additional studies are needed to examine the relationship between breathing abnormalities and cognitive dysfunction in MG. A third possible explanation for cognitive dysfunction in MG relates to the effects of nonspecific immunological processes (e.g. lymphokines) triggered by the influx of antibodies to the central nervous system. This hypothesis has been raised as a possible etiology for the electrical abnormalities that have been consistently identified in the brains of MG patients and animals with EAMG [3,4]. Whether or not such immunological responses also impair cognition in MG has not been considered, but future investigations should examine this possibility. Several limitations of the present study warrant discussion. First, the measures of response fluency administered in the present study required rapid oral responses and the measure of information processing required oral responses and visual tracking. Given the high motor demands of these tasks, MG patients may have performed more poorly on these measures due to mild dysarthria or opthalmoparesis, rather than impaired cognitive abilities. However, motor limitations would not have affected performance on the learning trials of the CVLT or the Visual Reproductions task, because responses to these measures were not timed. As such, we do not believe that patients’
63
poorer cognitive performance can be attributed entirely to motor difficulties. Second, although rest breaks were provided and the order of tests was counterbalanced, we cannot completely rule out the possibility that performances by the patient group were affected by the development of fatigue over the course of testing. Considering the prominent experience of fatigue reported by most MG patients, future studies should administer measures to monitor fatigue during testing to determine if perceptions of fatigue relate to cognitive performance. Finally, our assessment of disease severity was based on a simple self-report scale rather than the more traditional method of single fiber electromyography (SFEMG). In the present study, the mean disease severity rating was mild-moderate, which is consistent with the average disease activity of most patients in the population [1]. As such, it is unlikely that the disease severity scale administered in the present study was significantly inaccurate. In summary, results from the current investigation revealed that MG patients exhibit mild, but significant deficits on measures of response fluency, information processing and verbal and visual learning. No significant deficits were evident on measures of attention span or on indices of verbal and visual retention of information. The etiology of these cognitive difficulties is unresolved, but fatigue, apnea, and indirect immune processes represent important areas of future research.
References [1] Grob D. Clinical manifestations of myasthenia gravis. In: Albuquerque E, Eldefrawi A, editors, Myasthenia Gravis, New York: Chapman Hall, 1985, pp. 319–45. [2] Hopkins L. Clinical features of myasthenia gravis. Neurol Clin N America 1994;12(2):243–61. [3] Saphier D, Birmanns B, Brenner T. Electroencephalographic changes in experimental autoimmune myasthenia gravis 1993;114(2):200–4. [4] Hokkanen B, Toivakka B. Electroencephalographic findings in myasthenia gravis. Acta Neurol Scandinav 1969;45:556–67. [5] Bergonzi P, Mazza S, Mennuni G, Morantte M, Sollazzo D, Scoppetta C. Central nervous system involvement in myasthenia gravis. Ann NY Acad Sci 1981;377:810–1. [6] Hoeffer P, Aranow H, Rowland L. Myasthenia gravis and epilepsy. Arch Neurol Psychiatry 1958;80:10. [7] Simpson J, Westerberg M, Magee K. Myasthenia gravis Acta Neurol Scand 1966;23(suppl.):1–27. [8] Hokkanen E. Myasthenia gravis, a clinical analysis of the total material from Finland with special reference to endocrinological and neurological disorders. Ann Clin Res 1969;1:94–108. [9] Keesey J. Does myasthenia gravis affect the brain? J Neurol Sci 1999;170(2):77–89. [10] Ochs C, Bradley J, Katholi C et al. Symptoms of patients with myasthenia gravis receiving treatment. J Medicine 1988;29:1–12. [11] Tucker D, Roeltgen D, Wann P, Wertheimer R. Memory dysfunction in myasthenia gravis. Evidence for central cholinergic effects. Neurology 1988;38:1173–7. [12] Iwasaki Y, Kinoshita M, Ikeda K, Takamiya K, Shiojima T. Cognitive dysfunction in myasthenia gravis. Int J Neurosci 1990;54:29–33.
64
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
[13] Iwasaki Y, Kinoshita M, Ikeda K, Shiojima T, Kurihara T. Neuropsychological function before and after plasma exchange in myasthenia gravis. J Neurol Sci 1993;114:223–6. [14] Bohbot V, Jech R, Bures J, Nadel L, Ruzicka E. Spatial and nonspatial memory involvement in myasthenia gravis. J Neurol 1997;244:529–32. [15] Aarli J, Gilhus N, Thorlacius S, Johnsen H. Recovery from global amnesia during plasma exchange in myasthenia gravis: report of a case. Acta Neurol Scand 1989;80:351–3. [16] Bartel P, Lotz B. Neuropsychological test performance and affect in myasthenia gravis. Acta Neurol Scand 1995;91(4):266–70. [17] Paradis C, Lazar R, Kula R. Cognitive function in myasthenia gravis. Neuropsych, Neuropsychol, Behav Neurol 1994;7(3):211–4. [18] Glennerster A, DPhil M, Palace B, Warburton D, Oxbury S, Newsom-Davis J. Memory in myasthenia gravis: Neuropsychological tests of central cholinergic function before and after immunologic treatment. Neurology 1996;46:1138–42. [19] Nyehnuis DL, Luchetta T, Yamamoto C, Terrien A, Bernardin L, Rao S, Garron D. The development, standardization, and initial validation of the Chicago Multiscale Depression Inventory. J Personality Assessment 1998;70(2):386–401. [20] Wechsler D. Adult intelligence scale-revised, New York: Psychological Corporation, 1981. [21] Smith A. Symbol digit modalities test manual, Los Angeles: Western Psychological Services, 1982.
[22] Delis D, Kramer J, Kaplan B, Ober B. Manual: California verbal learning test, adult version, San Antonio: The Psychological Corporation, 1987. [23] Wechsler D. Wechsler memory scale-revised, San Antonio: Psychological Corporation, 1987. [24] Corwin J, Bylsma F. Translations of excerpts from Andre Reys’ psychological examination of traumatic encephalopathy and P.A. Osterrieth’s The Complex Figure Copy Test. The Clin Neuropsychol 1993;7:3–15. [25] Lezak M. Neuropsychological assessment, 2nd ed., Oxford University Press, 1983. [26] Welsh-Bohmer K, Ogrocki P. Clinical differentiation of memory disorders in neurodegenerative disease. In: Troster A, editor, Memory in neurodegenerative disease: biological, cognitive and clinical perspectives, Cambridge University Press, 1998, pp. 290–313. [27] Stepansky R, Weber G, Zeitlhofer J. Sleep apnea and cognitive dysfunction in myasthenia gravis. Acta Medica Austriaca 1997;24(3):128–31. [28] Manni R, Piccolo G, Sartori I, Castelnovo G, Raiola E, Lombardi M, Cerveri I, Fanfulla F, Tartara A. Breathing during sleep in myasthenia gravis. Italian J Neurol Sci 1995;16(9):589–94. [29] Amino A, Shiozawa Z, Nagasaka T, Shindo K, Ohashi K, Tsunoda S, Shintanin S. Sleep apnoea in well-controlled myasthenia gravis and the effect of thymectomy. J Neurol 1998;245(2):77–80.
Cognitive dysfunction in individuals with myasthenia gravis a, a a a b Robert H. Paul *, Ronald A. Cohen , James M. Gilchrist , Mark S. Aloia , Jonathan M. Goldstein a
Brown University School of Medicine, Miriam Hospital, 164 Summit Ave., Providence, RI 02906, USA b Yale University School of Medicine, Yale, NY, USA Received 23 February 2000; received in revised form 8 May 2000; accepted 6 July 2000
Abstract In the present study we administered a battery of cognitive measures that examined attention, response fluency, information processing, and verbal and visual learning and retention to 28 individuals with generalized myasthenia gravis (MG) and 18 demographically similar control subjects. Results revealed that MG patients performed significantly more poorly than control subjects on the measures of response fluency, information processing and most measures of verbal and visual learning. Significant group differences were not evident on the measure of attention span or on the indices of retention of information. Cognitive performances of the MG group were not related to mood disturbance, disease duration, or daily dose of prednisone. While these results suggest central involvement in MG, previous studies have not provided evidence that MG antibodies bind to central nicotinic receptors. Possible alternative mechanisms underlying cognitive dysfunction in MG are discussed. 2000 Elsevier Science B.V. All rights reserved. Keywords: Myasthenia gravis; Neuropsychology; Memory; Acetycholine
1. Introduction Myasthenia gravis (MG) is an autoimmune disease that involves antibody-mediated destruction of nicotinic cholinergic receptors. The typical clinical presentation of MG is associated with the prominent role of nicotinic receptors at the motor endplate of striated muscles. Specifically, MG patients report increased fatigability of voluntary muscles that worsens with exercise and attenuates with rest [1]. Most commonly, muscles of the eyes, face, neck, arms, and trunk are affected, although some patients also experience weakness of the legs and the diaphragm [1,2]. In addition to the muscular symptoms of the disease, several studies have identified central abnormalities in MG. Specifically, alterations in EEG patterns in both MG patients and animals with experimental autoimmune myasthenia gravis (EAMG) [3,4] have been reported, as well as *Corresponding author. Tel.: 11-401-793-4383; fax: 11-401-7519873. E-mail address: robert [email protected] (R.H. Paul). ]
alterations in brain stem auditory evoked potentials in individuals with MG [5]. Several studies [6–8] have reported an increased prevalence of seizures in MG compared to the general population, however, this finding has not been consistently replicated and the vast majority of MG patients do not present with abnormal ictal activity [9,10]. The central abnormalities described above have been attributed to dysfunction of central nicotinic receptor systems, however, there is limited evidence that central cholinergic receptors are affected in MG [9]. Nearly 60% of MG patients complain of cognitive difficulties [10], but few studies have empirically examined cognition in MG. Of the controlled studies that have been reported, results have not been consistent. The first controlled study of cognition in MG [11] revealed significant deficits by MG patients compared to healthy control subjects and a medical control group (non-neurological chronic diseases) on immediate and delayed recall of verbal information, but there was no group difference in the ability to retain information over time. Similar findings have been reported by Iwasaki et al. [12,13] using immediate and delayed trials of memory for prose passages.
0022-510X / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 00 )00367-1
60
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
Results from these studies revealed significant impairments among MG patients on both trials of the memory test. Impairments of visual learning and memory have also been reported in MG using measures of visual working memory, nonvisual spatial exploration, and performance in a human radial-arm maze [14]. In several studies of cognition in MG, performance has been shown to improve following plasmapheresis, suggesting that circulating antibodies to the nicotinic receptor might mediate cognitive dysfunction in MG [12,15]. In contrast to the investigations noted above, other studies have found only subtle, if any, deficits among MG patients on measures of cognitive function compared to healthy control samples. Specifically, Bartel and Lotz [16] examined cognition in MG subjects and healthy control subjects using the Randt Memory test and measures of simple and sustained attention, information processing speed, motor speed, and arithmetic. Results of the study revealed a significant group difference only on information processing speed. Other groups also have found no cognitive impairment among MG patients on measures of verbal learning and memory [17]. The conflicting results regarding cognition in MG may be due to methodological differences of previous studies. In many investigations [11–15], the effects of possible visual limitations (e.g. diplopia, opthalmoplegia), use of prednisone, or significant mood disturbance were not controlled, each of which may have resulted in significantly poorer performance on measures of cognition among MG patients than healthy control subjects. In other studies [16–18], inclusion of few subjects and / or patients with different subtypes of MG may have minimized group differences on cognitive tasks. The purpose of the present study was to examine cognitive performance of patients with MG after controlling for the methodological limitations of the previous studies described above. A total of 28 patients with generalized MG and 18 healthy control subjects were recruited for the study. Subjects were administered measures of attention, response fluency, information processing, and verbal and visual learning and retention.
2. Methods
2.1. Subjects A sample of 28 individuals with generalized MG were recruited for participation in the study. Patients were selfselected volunteers from regional chapters of the National Myasthenia Gravis Foundation and local neurologists. Diagnoses were confirmed by a positive Tensilon Test, electrodiagnositc recordings consistent with MG, or evidence of antinicotinic antibodies in serum. Patients were excluded if they reported a history of alcohol or drug abuse, loss of consciousness greater than 5 min, significant
cardiovascular disease, major psychiatric illness (e.g. schizophrenia) or additional neurological illness. Patients were also excluded if their corrected vision was worse than 20 / 70 or if they evidenced opthalmoplegia. Disease severity was determined using a Likert scale ranging from 1 to 3, where 15minimal interference of daily activities, 25moderate interference with daily activities despite medications, and 35significant interference of daily activities despite use of medications. The scale was treated as a continuous variable, thus permitting scores of 1.5, 2.5, etc. Patients in the present sample averaged 54. 71613.35 years of age (range529–77), 15.3962.73 years of education (range512–20), 7.7466.71 years since diagnosis, and 1.6460.67 on the disease severity scale (mild to moderate). A total of 16 (57%) patients reported experiencing cognitive difficulties as a regular symptom of the disease. This percentage is nearly identical to previously reported rates of self-described cognitive dysfunction in MG [10]. All patients were regularly taking medications to treat MG including Mestinon (n521), Imuran (n511) and prednisone (n513; median daily dose516.0 mg, range 2.5–60 mg). All patients were classified as Group II (i.e. generalized MG) according to modified Osserman criteria. A sample of 18 healthy control subjects were recruited from the community. Control subjects were similar in age and education and were required to meet the same exclusion criteria as the patient group. Control subjects in the present study averaged 51.16615.43 years of age (range5 29–78) and 16.462.57 years of education (range512–20). All subjects provided written informed consent after receiving a comprehensive explanation of the study. No subjects were paid for participation.
2.2. Procedure Study participants were provided the opportunity to be tested in our laboratory, in their own home, or at a public library. Nearly all subjects in both groups elected for home testing. Distractions during home testing were minimized by testing in a quiet location. All subjects were initially administered clinical and demographic questionnaires, followed by a visual acuity test to rule out severe visual disturbances, and a measure of mood. Subsequently, subjects were administered the cognitive battery, which required approximately 1.5 h to complete. To minimize the effects of fatigue developed over the course of testing, subjects were provided breaks from testing at 10-min intervals throughout the evaluation. In addition, the cognitive battery was administered in a counterbalanced order.
2.3. Measures Chicago Multiscale Depression Inventory (CMDI) [19]. This is a 42-item self-report scale of depression. The inventory assesses different aspects of depression including mood, self-evaluative, and vegetative symptoms. The
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
inventory was developed to assess mood symptoms of depression independent of vegetative symptoms, as the latter may artificially inflate the prevalence of depression in MG in neuroimmune disorders [19]. The mood subscale of the CMDI has good reliability and validity and correlates highly with standard clinical measures of depression [19]. The mood subscale includes 14 items and requires subjects to rate their degree of agreement using a Likert scale where 15no problem and 45extreme problem.
2.4. Cognitive battery 2.4.1. Attention span Digit Span Test [20]. Digits Forward and Backward were administered in the standard manner. The total number of correctly repeated trials for each version was the dependent measure. 2.4.2. Response fluency Letter Fluency: Subjects were required to generate words that began with the letters F, A, and S. Sixty seconds were allowed for each letter and proper nouns were not allowed. The total number of correct words generated across the three trials was the dependent measure. Animal Fluency: Subjects were required to name as many animals as quickly as possible for 60 s. The total number of animals generated was the dependent variable. 2.4.3. Information processing Oral version of the Symbol Digit Modalities Test (SDMT) [21]. Subjects were required to orally substitute numbers for symbols using a key held in constant view. The total number of correct substitutions made in 90 s was recorded as the dependent measure. 2.4.4. Verbal learning and retention The California Verbal Learning Test (CVLT) [22] was administered in standard fashion. Briefly, the test involves oral presentation of a 16-word list. The list items represent exemplars of one of four categories (fruit, clothing, tools, spices / herbs). Subjects were read the list five times and asked to recall as many words as possible from the whole list on each of the trials. Subjects were then read a distracter list, followed by free and cued recalls of the original list. After a 20-min delay, subjects were asked to free recall the original list, followed by a cued recall trial, and then a recognition trial. The total number of words recalled across the five learning trials was the dependent measure for verbal learning. Percent retained from the fifth learning trial to the long delay free recall trial and number correct on the recognition trial served as the dependent measures for verbal retention. 2.4.5. Visual learning and retention The Visual Reproductions I and II (VRI and VRII)
61
subtests from the Wechsler Memory Scale-Revised [23] were administered in standard fashion. Briefly, subjects were presented four stimulus cards, one at a time, each card depicting a geometric design. Subjects were asked to look at each design for 10 s, after which the card was removed from sight and subjects were required to draw the design from memory. Following a 20-min delay period, subjects were again asked to draw the designs from memory. Scoring was performed according to standard protocol. Total correct for the first trial served as the dependent measure of visual learning. Percent retained from the learning trial to the delayed trial served as the dependent measure of visual retention. The Rey Complex Figure Test [24] was also administered as a measure of visual learning and retention. Subjects were instructed to copy the geometric figure as accurately as possible. Immediately after completing the copy trial, subjects were asked to draw the figure from memory. Following a 20-min time delay, subjects were again instructed to draw the figure from memory. No time limit was given and the drawings were scored according to standardized criteria [25]. The dependent measure for visual learning was the total score on the immediate recall trial and the dependent measure for visual retention was the total score on the delayed trial.
3. Results The two groups did not differ significantly by age or education (Fs ,2). Similarly, on the mood subscale of the CMDI, the patient group reported a higher average score, but the difference between groups was not statistically significant (F52.58, P.0.05).
3.1. Attention On the Digit Span test, the two groups performed equally well on Digits Forward (F(1,44)50.86, P.0.05) and Digits Backward (1,44)50.26, P.0.05) (see Table 1).
3.2. Response fluency Performances on the measures of response fluency were significantly different by group. Specifically, the patient group generated fewer words than the control subjects on both letter fluency (F(1,44)519.08, P,0.05) and category fluency (F(1,44)58.84, P,0.05).
3.3. Information processing On the SDMT, the patient group performed significantly more poorly than the control group (F(1,44)511.37, P, 0.05).
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
62
Table 1 Performances on cognitive measures for patients and controls: Mean (SD) Measure
MG patients
Control subjects
Attention span Digits forward Digits backward
9.62 (1.88) 8.22 (2.06)
9.11 (1.74) 7.88 (2.21)
Response fluency Letter fluency Animal fluency
37.66 (12.31) 18.25 (4.61)
52.38 (8.85)** 22.33 (4.32)**
Information Processing Symbol digit modalities test
49.85 (12.46)
61.33 (8.89)**
Verbal learning and retention CVLT (5 trials; max580) Retention (long delay / trial 5) CVLT Recognition (max516)
48.40 (9.76) 0.75 (0.92) 14.83 (1.54)
57.72 (8.30)** 0.86 (0.98) 14.31 (1.28)
Visual learning and retention Visual reproductions I (max542) Retention (VRII / VRI) Rey immediate (max536) Rey delayed (max536)
33.96 0.81 17.00 15.94
37.50 0.87 19.22 18.30
(6.07) (0.20) (8.64) (7.48)
(2.74)* (0.16) (6.63) (6.68)
*P,0.05. **P,0.01.
3.4. Verbal learning and retention The two groups also differed significantly on the measure of verbal learning, as the patient group recalled fewer words across the five learning trials of the CVLT compared to the control subjects (F(1,44)511.03, P, 0.05). By contrast, the percent retained from the fifth learning trial of the CVLT to the long-delay free recall trial did not differ significantly by group (F,3). Similarly, the two groups did not differ significantly on the recognition trial of the CVLT (F,2).
3.5. Visual learning and retention On the learning trial of the Visual Reproductions test, the patient group performed significantly more poorly than the control group (F(1,44)55.34, P,0.05). By contrast, the percent retained from the learning trial to the delayed trial did not differ by group (F51). On the Rey Complex Figure, the two groups did not differ on the copy, immediate, or delayed trials (Fs ,3). Multiple regression analyses were conducted to examine the amount of variance in cognitive function accounted for by clinical variables. Time since diagnosis, disease severity, daily dose of prednisone, and the mood subscale of the CMDI was entered as a clinical variables for these analyses. The dependent variables were performances on each of the measures of cognition. Results of these analyses revealed a significant relationship between disease seventy and performance on the SDMT (R50.385, F(1,25)54.35, P,0.05). Cognitive performance was not associated with any of the other variables. For prednisone,
the greatest R value was 0.253, P.0.05 (letter fluency), while for mood the greatest R value was 0.233, P.0.05 (animal fluency).
4. Discussion Cognitive dysfunction in MG has been reported in previous investigations examining the neurobehavioral aspects of the illness, however methodological issues produced considerable limitations to these studies. After controlling for many of these concerns in the present study, MG patients were found to exhibit mild, but significant difficulty compared to control subjects on several measures of cognitive function. Specifically, while MG patients performed as well as control subjects on measures of attention span and retention of verbal and visual information, the patient group exhibited significantly poorer performance on measures of response fluency, information processing and verbal and visual learning. Importantly, these difficulties were not associated with mood disturbance or with daily dose of prednisone. It should noted, however, that the severity of these cognitive difficulties is mild and not likely to significantly affect daily function. The observation that MG patients performed as well as control subjects on Digit Span but not on the SDMT test indicates that attention span is intact in MG, while rapid processing of information is impaired. The poorer performance by the MG group on the fluency measures also likely reflects impaired information processing given the limited time constraints allowed for completion of these tasks. On the learning and memory tasks, the MG group performed as well as controls on measures that allowed self-determined exposure to the stimuli (i.e. Rey Complex Figure), but not on measures that involved limited exposure of stimuli. These findings raise the possibility that MG patients were less capable of efficiently processing information within the time constraints imposed by the CVLT and Visual Reproductions test. In addition, while MG patients typically recalled significantly less information on both the learning and delayed trials of memory tasks, they did not lose more information over time than control subjects, a pattern that is inconsistent with the pattern typical of neurodegenerative diseases involving central cholinergic systems (e.g. Dementia of the Alzheimer’s type) [26]. As this study represents the first examination of cognition in MG to simultaneously address issues such as medications, gross visual impairments and mood, replication of the current findings using an independent sample represents the most valuable follow-up to this study. Until these results are confirmed, speculation regarding the etiology of cognitive dysfunction in patients with the disease seems premature. Nevertheless, consideration of potential etiological mechanisms will help direct future
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
studies, and therefore it is appropriate to briefly mention these possibilities at the present time. In most previous studies, cognitive impairments in MG have been attributed to the effects of antibody-mediated destruction of central nicotinic receptors. However, Keesey [9] recently reviewed the literature on this topic, and based largely on the different stoichiometry of the four nicotinic receptor subtypes, provided a convincing argument against the possibility of central nicotinic dysfunction in MG. We agree with this position, and add that the lack of involvement of autonomic nicotinic receptors in the disease provides further evidence that these receptors are not uniformly affected. Another possible etiological mechanism that has been suggested by previous research [9,27] concerns the diaphragm, which like many striated muscles, is affected by the binding of MG antibodies to nicotinic receptors. In MG, many patients experience breathing abnormalities during sleep, when the diaphragm is directly involved in the regulation of airflow [28,29]. Although cognitive difficulties are common consequences of reduced oxygen saturation during sleep, only one study has examined sleep and cognition in MG [27]. Results from this study revealed no significant relationship between cognition and percent of sleep with oxygen saturation below 93% using regression analyses. However, when the patient group was subdivided into those with and without oxygen desaturation during sleep, comparisons revealed a significant difference on one of two memory tests. Unfortunately, additional analyses addressing how these two groups may have differed by age, education, mood status, or use of medications were not provided. Clearly, additional studies are needed to examine the relationship between breathing abnormalities and cognitive dysfunction in MG. A third possible explanation for cognitive dysfunction in MG relates to the effects of nonspecific immunological processes (e.g. lymphokines) triggered by the influx of antibodies to the central nervous system. This hypothesis has been raised as a possible etiology for the electrical abnormalities that have been consistently identified in the brains of MG patients and animals with EAMG [3,4]. Whether or not such immunological responses also impair cognition in MG has not been considered, but future investigations should examine this possibility. Several limitations of the present study warrant discussion. First, the measures of response fluency administered in the present study required rapid oral responses and the measure of information processing required oral responses and visual tracking. Given the high motor demands of these tasks, MG patients may have performed more poorly on these measures due to mild dysarthria or opthalmoparesis, rather than impaired cognitive abilities. However, motor limitations would not have affected performance on the learning trials of the CVLT or the Visual Reproductions task, because responses to these measures were not timed. As such, we do not believe that patients’
63
poorer cognitive performance can be attributed entirely to motor difficulties. Second, although rest breaks were provided and the order of tests was counterbalanced, we cannot completely rule out the possibility that performances by the patient group were affected by the development of fatigue over the course of testing. Considering the prominent experience of fatigue reported by most MG patients, future studies should administer measures to monitor fatigue during testing to determine if perceptions of fatigue relate to cognitive performance. Finally, our assessment of disease severity was based on a simple self-report scale rather than the more traditional method of single fiber electromyography (SFEMG). In the present study, the mean disease severity rating was mild-moderate, which is consistent with the average disease activity of most patients in the population [1]. As such, it is unlikely that the disease severity scale administered in the present study was significantly inaccurate. In summary, results from the current investigation revealed that MG patients exhibit mild, but significant deficits on measures of response fluency, information processing and verbal and visual learning. No significant deficits were evident on measures of attention span or on indices of verbal and visual retention of information. The etiology of these cognitive difficulties is unresolved, but fatigue, apnea, and indirect immune processes represent important areas of future research.
References [1] Grob D. Clinical manifestations of myasthenia gravis. In: Albuquerque E, Eldefrawi A, editors, Myasthenia Gravis, New York: Chapman Hall, 1985, pp. 319–45. [2] Hopkins L. Clinical features of myasthenia gravis. Neurol Clin N America 1994;12(2):243–61. [3] Saphier D, Birmanns B, Brenner T. Electroencephalographic changes in experimental autoimmune myasthenia gravis 1993;114(2):200–4. [4] Hokkanen B, Toivakka B. Electroencephalographic findings in myasthenia gravis. Acta Neurol Scandinav 1969;45:556–67. [5] Bergonzi P, Mazza S, Mennuni G, Morantte M, Sollazzo D, Scoppetta C. Central nervous system involvement in myasthenia gravis. Ann NY Acad Sci 1981;377:810–1. [6] Hoeffer P, Aranow H, Rowland L. Myasthenia gravis and epilepsy. Arch Neurol Psychiatry 1958;80:10. [7] Simpson J, Westerberg M, Magee K. Myasthenia gravis Acta Neurol Scand 1966;23(suppl.):1–27. [8] Hokkanen E. Myasthenia gravis, a clinical analysis of the total material from Finland with special reference to endocrinological and neurological disorders. Ann Clin Res 1969;1:94–108. [9] Keesey J. Does myasthenia gravis affect the brain? J Neurol Sci 1999;170(2):77–89. [10] Ochs C, Bradley J, Katholi C et al. Symptoms of patients with myasthenia gravis receiving treatment. J Medicine 1988;29:1–12. [11] Tucker D, Roeltgen D, Wann P, Wertheimer R. Memory dysfunction in myasthenia gravis. Evidence for central cholinergic effects. Neurology 1988;38:1173–7. [12] Iwasaki Y, Kinoshita M, Ikeda K, Takamiya K, Shiojima T. Cognitive dysfunction in myasthenia gravis. Int J Neurosci 1990;54:29–33.
64
R.H. Paul et al. / Journal of the Neurological Sciences 179 (2000) 59 – 64
[13] Iwasaki Y, Kinoshita M, Ikeda K, Shiojima T, Kurihara T. Neuropsychological function before and after plasma exchange in myasthenia gravis. J Neurol Sci 1993;114:223–6. [14] Bohbot V, Jech R, Bures J, Nadel L, Ruzicka E. Spatial and nonspatial memory involvement in myasthenia gravis. J Neurol 1997;244:529–32. [15] Aarli J, Gilhus N, Thorlacius S, Johnsen H. Recovery from global amnesia during plasma exchange in myasthenia gravis: report of a case. Acta Neurol Scand 1989;80:351–3. [16] Bartel P, Lotz B. Neuropsychological test performance and affect in myasthenia gravis. Acta Neurol Scand 1995;91(4):266–70. [17] Paradis C, Lazar R, Kula R. Cognitive function in myasthenia gravis. Neuropsych, Neuropsychol, Behav Neurol 1994;7(3):211–4. [18] Glennerster A, DPhil M, Palace B, Warburton D, Oxbury S, Newsom-Davis J. Memory in myasthenia gravis: Neuropsychological tests of central cholinergic function before and after immunologic treatment. Neurology 1996;46:1138–42. [19] Nyehnuis DL, Luchetta T, Yamamoto C, Terrien A, Bernardin L, Rao S, Garron D. The development, standardization, and initial validation of the Chicago Multiscale Depression Inventory. J Personality Assessment 1998;70(2):386–401. [20] Wechsler D. Adult intelligence scale-revised, New York: Psychological Corporation, 1981. [21] Smith A. Symbol digit modalities test manual, Los Angeles: Western Psychological Services, 1982.
[22] Delis D, Kramer J, Kaplan B, Ober B. Manual: California verbal learning test, adult version, San Antonio: The Psychological Corporation, 1987. [23] Wechsler D. Wechsler memory scale-revised, San Antonio: Psychological Corporation, 1987. [24] Corwin J, Bylsma F. Translations of excerpts from Andre Reys’ psychological examination of traumatic encephalopathy and P.A. Osterrieth’s The Complex Figure Copy Test. The Clin Neuropsychol 1993;7:3–15. [25] Lezak M. Neuropsychological assessment, 2nd ed., Oxford University Press, 1983. [26] Welsh-Bohmer K, Ogrocki P. Clinical differentiation of memory disorders in neurodegenerative disease. In: Troster A, editor, Memory in neurodegenerative disease: biological, cognitive and clinical perspectives, Cambridge University Press, 1998, pp. 290–313. [27] Stepansky R, Weber G, Zeitlhofer J. Sleep apnea and cognitive dysfunction in myasthenia gravis. Acta Medica Austriaca 1997;24(3):128–31. [28] Manni R, Piccolo G, Sartori I, Castelnovo G, Raiola E, Lombardi M, Cerveri I, Fanfulla F, Tartara A. Breathing during sleep in myasthenia gravis. Italian J Neurol Sci 1995;16(9):589–94. [29] Amino A, Shiozawa Z, Nagasaka T, Shindo K, Ohashi K, Tsunoda S, Shintanin S. Sleep apnoea in well-controlled myasthenia gravis and the effect of thymectomy. J Neurol 1998;245(2):77–80.