- Email: [email protected]
Neuropsychological effects of seizures
Epilepsy & Behavior Epilepsy & Behavior 5 (2004) S21–S24 www.elsevier.com/locate/yebeh
Neuropsychological effects of seizures Carl B. Dodrill* Regional Epilepsy Center, Departments of Neurology and Neurological Surgery, University of Washington School of Medicine (Box 359745), Harborview Medical Center, 325 Ninth Avenue, Seattle, WA 98104-2499, USA Received 4 November 2003; accepted 4 November 2003
Abstract The neuropsychological effects of seizures, including an accumulation of single attacks, on mental abilities are explored through a selective review of the worldÕs literature. Each paper included in this review was longitudinal in nature, and in each case, formal psychological testing was accomplished both at the beginning and at the end of the study. Of the 22 investigations meeting all requirements for inclusion, 9 focused on children and 13 on adolescents and adults. Some difficulties were encountered in that the types and numbers of seizures occurring during the studies were rarely included. Except in three studies where there were very few seizures of any kind, it appears justifiable to assume that the patients in all other studies had active seizure disorders. Assuming this to be true, in 12 of the remaining 20 papers, a relationship was found between seizures and changes in mental abilities, in 5 cases the results were mixed or uncertain, and in 3 cases the available evidence was against such a relationship. Mild but definite relationships between seizures and mental decline are supported by this literature review, but these relationships require much more study, particularly with a clearer understanding of the types and frequencies of seizures experienced. Suggestions for future research are offered. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Epileptic seizures and cognition; Seizure-induced cognitive deterioration
1. Introduction While the possibility that seizures may damage the brain has been entertained for decades, the question is extraordinarily complex, and as of yet there is no clear and simple answer. Indeed, only recently a week-long symposium of the worldÕs experts was organized to address this question, and the resulting monograph [1] shows how complicated the question actually is. This review is selective in nature, and it includes only studies that had the following features: (1) formal tests of mental abilities were used rather than clinical judgments of mental status; and (2) patients with epilepsy were tested on at least two occasions without an intervening treatment such as surgery or systematic drug changes that may have impacted the test results. Thus, all studies reviewed in this article are longitudinal investigations where formal testing was accomplished at both the beginning and the end of the studies. It is * Fax: +206-731-4409. E-mail address: [email protected].
1525-5050/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2003.11.004
possible to insist on these criteria since there are now more than 20 published investigations that meet these requirements. Longitudinal studies are likely to present the most accurate information as crosssectional studies attribute more adverse cognitive changes to seizures than can be documented by longitudinal investigations [2]. Studies are reviewed first for children and then for adults.
2. Studies involving children Table 1 summarizes the results of the longitudinal studies done with children. A review of the table shows no fewer than five longitudinal studies with test–retest cognitive data published prior to 1940. Thus, this has been a topic of interest for decades, and as will become evident, longitudinal studies with children were done actually ahead of those with adults. A second conclusion that is obvious from Table 1 is that the studies done almost exclusively used measures of intelligence. This is definitely a limitation of this group of investigations as
S22
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
Table 1 Longitudinal cognitive studies of children with epilepsy Investigator/year
Subject group
Fox/1924 [4]
130 institutionalized Binet children with epilepsy
Patterson and Fonner/1928 [5]
98 institutionalized children and adolescents Dawson and 21 children with Conn/1929 [6] epilepsy Fetterman and 46 hospital Barnes/1934 [7] dispensary cases Sullivan and 44 children Gahagan/1935 [8] 7.5 years of age; outpatients Bourgeois et al./ 72 children 1983 [9] 7.5 years of age; outpatients Ellenberg et al./ 1986 [10]
Aldenkamp et al./ 1990 [11] Bjornaes et al./ 2001 [12]
83 children from a large developmental study 45 children 9.3 years of age; outpatients 17 children 10.2 years of age; later had surgery
Test(s) used
Test–retest interval
Seizure frequency
1 year
Unclear; probably active seizures
Dutch form of WISC-R
4.2 years on average
Norwegian form of WISC-R
3.5 years on average
All but 10 children had active seizures through follow-up All had active seizures of moderate severity
Stanford–Binet
Binet Stanford–Binet Stanford–Binet
Age-appropriate IQ and other tests
Stanford–Binet at age 4; WISC at age 7
Cognitive outcome
37% decrease>2 IQ points 41% same IQ (2 points) 22% increase>2 IQ points 8% decrease>10 IQ points 12–46 months Unclear; probably 34% decrease>2 IQ points active seizures 31% same-IQ 2 points 34% increase>2 IQ points 8–56 months Not stated Mean IQ drop from 82 to 67; 11 showed losses, 10 no change 21 months on Not stated 50% ‘‘moderate loss’’; 9% average no change; 41% ‘‘slight increase’’ 1–59 months; 14 Not well described 50% ‘‘negative change’’; 5% ‘‘no change’’; months average 45% ‘‘positive change’’; 9% clear deterioration 4 years on average Details provided No change in IQ for the group at length; active overall, but IQ>10 point loss in seizure disorders 11% who had frequent seizures, toxic drug levels, early epilepsy 3 years Patients had very Intelligence showed no losses from few seizures ages 3 to 7
measures of intelligence are not as sensitive to difficulties in brain functions as are validated neuropsychological tests [3]. Measures of intelligence were, after all, devised to predict how well children would do in school, and not to identify brain-related problems [3]. Also, neuropsychological tests clearly evaluate a broader range of functions than measures of intelligence. An additional limitation is obvious from a perusal of Table 1: very few data are presented on the type and frequency of seizures between the testings. Presumably, the children in all studies except one [10] had active seizures between the cognitive testings. Of the eight remaining studies, there was evidence for some cognitive loss in each study which exceeded any gains except for one investigation where no evidence for such a loss was found [5]. Furthermore, in about half the papers, significant IQ losses of 10 points or more were reported in about 10–25% of the children studied. While the data on seizure type and seizure frequency are so sparse that no connections between these variables and cognitive losses can be made, the weight of the evidence is that measurable losses in intelligence were experienced over time in at least a portion of the children with active seizures. The importance of this overall conclusion is underscored by the fact that all these studies had test–retest intervals
IQ scores slightly but not significantly down for total group; 24% had FSIQ losses>9 points Adverse IQ score changes found: PIQ (P < 0:05), FSIQ (P < 0:05)
of only 1–4 years. It appears that longer intervals would likely have resulted in more findings.
3. Studies involving adolescents and adults Results on the 13 longitudinal investigations with adolescents and adults are presented in Table 2. A more diverse group of tests were used in these studies, and only 5 of the 13 investigations rested on measures of intelligence alone. Also, the test–retest intervals tended to be longer, with three of the studies evaluating cognitive changes over 10-year periods. However, the seizure types and frequencies experienced during the test–retest intervals were no more precisely given than with children, with only a few exceptions [16,19,20,23]. In two studies with adolescents and adults [18,21], patients had few or no seizures during the course of the studies, and in both of these cases test performances either remained the same or improved. In the remaining 11 investigations summarized in Table 2, one can assume that most patients had at least a few seizures during the study periods even though in some studies a small number of people had complete seizure control. Of these 11 studies, evidence was found for a relationship
Table 2 Longitudinal cognitive studies of adults with epilepsy Investigator/year
Subject group
Test used
Test–retest interval
Seizure frequency
Cognitive outcome
Arieff and Yacorzynski/1942 [13]
90 adolescents and adults; 27 known etiology, 63 unknown etiology 56 adolescents and adults
Binet, 1916 and1937 versions
1–10 years
Wechsler–Bellevue, Form I (?)
At least 5 years, 7 years on average
Wide range of seizure control from complete remission to very active Wide range of seizure control
WAIS
1–5 years, 18 months
Varied by subgroup
Dodrill and Wilensky/1990 [16]
9 adults with status epilepticus during study period, and 9 closely matched controls with no history of status 69 adults who had participated in a vocational rehabilitation program 36 adults with no medication changes over a 5-year period 28 adults with TLE, typically with complaints of memory problems 35 adults with intractable partial seizures (60% also had secondary GTC seizures) 58 new cases with partial seizures 47 TLE cases
WAIS
5 years (6 months)
One group with status (4 GTC, 5 partial), one group without status
Finnish form of WAIS; tests of memory and some other functions Expanded Halstead–Reitan battery, WAIS WAIS-R, Wechsler Memory Scale (Form I)
Approximately 10 years
Improved in control over study, but only 10% controlled at end
Memory losses in 15-20%; intelligence unchanged overall; seizure frequency and losses in abilities difficult to connect
5 years (6 months)
Very few or no seizures
No changes on the tests beyond those expected by chance
2.3 years on average
8 partial seizures per month
Expanded Halstead–Reitan battery, WAIS
10 years (6 months)
9 partial and generalized seizures per month
Finnish form of the WAIS and other tests Measures of memory and other functions WAIS
5 years
Very few seizures of any kind
2–10 years, 4.7 years on average 6 years
Active partial seizures
Slight improvement in PIQ and FSIQ as would be expected by chance in a normal adult group WAIS little change; 6 of 17 other test variables showed losses (visual memory, attention, problem solving, perception) 12 of 24 test variables showed slight gains; none showed losses Verbal memory unchanged; especially figural memory showed loss (P < 0:05) No losses found; better PIQ (P < 0:01) and FSIQ (P < 0:05) may have been retest effects Normal group improved more than epilepsy group on 3 of 20 test variables; visual memory lost in both groups; partial seizures not correlated with ability changes, but 2 of 20 significantly correlated with number of GTC seizures; GTC status patients lost verbal and visual memory 50% showed significant decline in memory (verbal plus figural bined); few changes in nonmemory functions
Kalska/1991 [17]
Dodrill and Wilensky/1992 [18] Selwa et al./1994 [19]
Holmes et al./1998 [20]
Aikia and Kalviainen/1999 [21] Helmstaedter et al./2000 [22] Bjornaes et al./2001 [12]
17 adults who became surgical candidates
Dodrill/2002 [23]
35 adults with partial seizures with or without secondary generalization; 35 normal controls
Expanded Halstead–Reitan battery, WAIS
10 years (6 months)
Helmstaedter et al./2003 [24]
102 adults with TLE
Measures of memory and of some other cognitive functions
2–10 years, average of 57 months
Mild to moderate seizure severity (mean of 2.3 on 0–4 scale) Average of 1109 partial seizures and 61 GTC seizures over 10 years for epilepsy group; four patients experienced GTC status epilepticus during 10-year period
89% had more than 1/month at start, 72% had more than 1/ month at end of study
S23
47 adults; 22 improved in seizure frequency; 25 had no change or got worse
Rodin/1968 [14]
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
Seidenberg et al./1981 [15].
Significant loss of 6 IQ points if etiology was known; no loss when etiology was unknown Favorable changes in seizure frequency correlated. 33 (P < 0:02) with favorable changes in FSIQ Seizure improved group had on average 3–10 IQ point improvements; seizure unimproved group had )1 to +4 point changes The no status group improved more on FSIQ than did the status group (P < 0:05); same trend shown on VIQ
S24
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
between seizures and adverse cognitive changes in five [14–16,20,24], mixed or uncertain findings were noted in four [13,17,22,24], and no connection was found between seizures and changes in mental abilities in two [12,19]. While taken as a whole, these findings must be described as mixed; an examination of the last column in Table 2 reveals a series of findings, all of which would support a mild (but no more than a mild) connection between seizures and adverse cognitive change. Among these findings are the following: (1) statistically significant relationships between numbers of seizures (or presence of status epilepticus) and decreased scores on tests of abilities [14,16,23]; (2) changes in intelligence corresponding to changes in seizure frequency [15]; (3) better performances over time with normals than with people with epilepsy [23]; and (4) losses in mental abilities with patients having uncontrolled seizures which touch on a number of areas of cognitive functioning [20] despite the fact that a consensus exists that losses in memory are most common [17,22–24].
4. Discussion Based on the information at hand, it is concluded that losses in mental abilities do occur over time with uncontrolled seizures in patients with epilepsy. There is some evidence that adverse changes are more easily found in children than in adults, and this was shown in the review as a whole as well as in a particular study that permitted such a comparison [12]. The impression gained by this author is that losses in mental abilities are most easily connected with generalized tonic–clonic seizures, especially when experienced in a serial fashion (status epilepticus). This is difficult to demonstrate from a review of Tables 1 and 2, however. To more effectively answer the obvious unanswered questions pertaining to the cognitive effects of seizures, the major limitations in existing studies need to be addressed directly and deliberately. In particular, longitudinal prospective investigations are needed that include recording the numbers of the various types of seizures experienced throughout the study. This is the most glaring deficiency in the current investigations. It is also likely that long study periods (at least 25 years) are required before the true cognitive effects of single seizures in the lifetime of our patients can be adequately assessed. It is of importance that such studies include normal groups closely matched with the epilepsy samples for age and education at the beginning of the study as developmental and aging changes must be considered in such lengthy investigations. While these investigations will certainly prove to be laborious, their value to our patients will be great, and it is essential that they be undertaken.
References [1] Sutula T, Pitkanen A. Do seizures damage the brain? Prog Brain Res 2002;135:1–520. [2] Dodrill CB. Progressive decline in adolescents and adults with epilepsy. In: Sutula T, Pitkanen A, editors. Do seizures damage the brain? Prog Brain Res 2002;135:399–407. [3] Reitan RM, Wolfson D. The Halstead–Reitan Neuropsychological Test Battery. Tucson: Neuropsychology Press; 1993. [4] Fox JT. The response of epileptic children to mental and educational tests. Br J Med Psychol 1924;4:235–48. [5] Patterson HA, Fonner D. Some observations on the intelligence quotient in epileptics. Psychiatr Q 1928;2:542–8. [6] Dawson S, Conn JCM. The intelligence of epileptic children. Arch Dis Child 1929;4:142–51. [7] Fetterman J, Barnes MR. Serial studies of the intelligence of patients with epilepsy. Arch Neurol Psychiatry 1934;32:797–801. [8] Sullivan EB, Gahagan L. On intelligence of epileptic children. Genet Psychol Monog 1935;17:309–75. [9] Bourgeois BFD, Prensky AL, Palkes HS, Talent BK, Buxch SG. Intelligence in epilepsy: a prospective study in children. Ann Neurol 1983;14:438–44. [10] Ellenberg JH, Hirtz DG, Nelson KB. Do seizures in children cause intellectual deterioration? N Engl J Med 1986;314:1085–8. [11] Aldenkamp AP, Alpherts WCJ, Bruine-Seeder DD, Dekker MJA. Test–retest variability in children with epilepsy: a comparison of WISC-R profiles. Epilepsy Res 1990;7:165–72. [12] Bjornaes H, Stabell K, Henriksen O, Loyning Y. The effects of refractory epilepsy on intellectual functioning in children and adults: a longitudinal study. Seizure 2001;10:250–9. [13] Arieff AJ, Yacorzynski GK. Deterioration of patients with organic epilepsy. J Nerv Ment Dis 1942;96:49–55. [14] Rodin EA. The prognosis of patients with epilepsy. Springfield, IL: Charles C Thomas; 1968. [15] Seidenberg M, OÕLeary DS, Berent S, Boll T. Changes in seizure frequency and test–retest scores on the Wechsler Adult Intelligence Scale. Epilepsia 1981;22:75–83. [16] Dodrill CB, Wilensky AJ. Intellectual impairment as an outcome of status epilepticus. Neurology 1990;40(Suppl. 2):23–7. [17] Kalska H. Cognitive changes in epilepsy: a ten-year follow-up. In: Commentationes Scientiarum Socialium, vol. 44. Helsinki: Societas Scientiarum Fennica; 1991. p. 1–85. [18] Dodrill CB, Wilensky AJ. Neuropsychological abilities before and after 5 years of stable antiepileptic drug therapy. Epilepsia 1992;33:327–34. [19] Selwa LM, Berent S, Giordani B, Henry TR, Buchtel HA, Ross DA. Serial cognitive testing in temporal lobe epilepsy: longitudinal changes with medical and surgical therapies. Epilepsia 1994;35:743–9. [20] Holmes MD, Dodrill CB, Wilkus RJ, Ojemann LM, Ojemann GA. Is partial epilepsy progressive? Ten-year follow-up of EEG and neuropsychological changes in adults with partial seizures. Epilepsia 1998;39:1189–93. [21] Aikia M, Kalviainen R. Five-year follow-up of cognitive performance of adult patients with well-controlled partial epilepsy. Epilepsia 1999;40(Suppl. 2):100–1. [22] Helmstaedter C, Kurthen M, Lux S, et al. Temporal lobe epilepsy: longitudinal clinical, neuropsychological and psychosocial followup of surgically and conservatively managed patients. Nervenarzt 2000;71:639–42 [in German]. [23] Dodrill CB. Progressive cognitive decline in adolescents and adults with epilepsy. In: Sutula T, Pitkanen A, editors. Do seizures damage the brain? Prog Brain Res 2002;135:399–407. [24] Helmstaedter C, Kurthen M, Lux S, Reuber M, Elger CE. Chronic epilepsy and cognition: a longitudinal study in temporal lobe epilepsy. Ann Neurol 2003;54:425–32.
Neuropsychological effects of seizures Carl B. Dodrill* Regional Epilepsy Center, Departments of Neurology and Neurological Surgery, University of Washington School of Medicine (Box 359745), Harborview Medical Center, 325 Ninth Avenue, Seattle, WA 98104-2499, USA Received 4 November 2003; accepted 4 November 2003
Abstract The neuropsychological effects of seizures, including an accumulation of single attacks, on mental abilities are explored through a selective review of the worldÕs literature. Each paper included in this review was longitudinal in nature, and in each case, formal psychological testing was accomplished both at the beginning and at the end of the study. Of the 22 investigations meeting all requirements for inclusion, 9 focused on children and 13 on adolescents and adults. Some difficulties were encountered in that the types and numbers of seizures occurring during the studies were rarely included. Except in three studies where there were very few seizures of any kind, it appears justifiable to assume that the patients in all other studies had active seizure disorders. Assuming this to be true, in 12 of the remaining 20 papers, a relationship was found between seizures and changes in mental abilities, in 5 cases the results were mixed or uncertain, and in 3 cases the available evidence was against such a relationship. Mild but definite relationships between seizures and mental decline are supported by this literature review, but these relationships require much more study, particularly with a clearer understanding of the types and frequencies of seizures experienced. Suggestions for future research are offered. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Epileptic seizures and cognition; Seizure-induced cognitive deterioration
1. Introduction While the possibility that seizures may damage the brain has been entertained for decades, the question is extraordinarily complex, and as of yet there is no clear and simple answer. Indeed, only recently a week-long symposium of the worldÕs experts was organized to address this question, and the resulting monograph [1] shows how complicated the question actually is. This review is selective in nature, and it includes only studies that had the following features: (1) formal tests of mental abilities were used rather than clinical judgments of mental status; and (2) patients with epilepsy were tested on at least two occasions without an intervening treatment such as surgery or systematic drug changes that may have impacted the test results. Thus, all studies reviewed in this article are longitudinal investigations where formal testing was accomplished at both the beginning and the end of the studies. It is * Fax: +206-731-4409. E-mail address: [email protected].
1525-5050/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2003.11.004
possible to insist on these criteria since there are now more than 20 published investigations that meet these requirements. Longitudinal studies are likely to present the most accurate information as crosssectional studies attribute more adverse cognitive changes to seizures than can be documented by longitudinal investigations [2]. Studies are reviewed first for children and then for adults.
2. Studies involving children Table 1 summarizes the results of the longitudinal studies done with children. A review of the table shows no fewer than five longitudinal studies with test–retest cognitive data published prior to 1940. Thus, this has been a topic of interest for decades, and as will become evident, longitudinal studies with children were done actually ahead of those with adults. A second conclusion that is obvious from Table 1 is that the studies done almost exclusively used measures of intelligence. This is definitely a limitation of this group of investigations as
S22
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
Table 1 Longitudinal cognitive studies of children with epilepsy Investigator/year
Subject group
Fox/1924 [4]
130 institutionalized Binet children with epilepsy
Patterson and Fonner/1928 [5]
98 institutionalized children and adolescents Dawson and 21 children with Conn/1929 [6] epilepsy Fetterman and 46 hospital Barnes/1934 [7] dispensary cases Sullivan and 44 children Gahagan/1935 [8] 7.5 years of age; outpatients Bourgeois et al./ 72 children 1983 [9] 7.5 years of age; outpatients Ellenberg et al./ 1986 [10]
Aldenkamp et al./ 1990 [11] Bjornaes et al./ 2001 [12]
83 children from a large developmental study 45 children 9.3 years of age; outpatients 17 children 10.2 years of age; later had surgery
Test(s) used
Test–retest interval
Seizure frequency
1 year
Unclear; probably active seizures
Dutch form of WISC-R
4.2 years on average
Norwegian form of WISC-R
3.5 years on average
All but 10 children had active seizures through follow-up All had active seizures of moderate severity
Stanford–Binet
Binet Stanford–Binet Stanford–Binet
Age-appropriate IQ and other tests
Stanford–Binet at age 4; WISC at age 7
Cognitive outcome
37% decrease>2 IQ points 41% same IQ (2 points) 22% increase>2 IQ points 8% decrease>10 IQ points 12–46 months Unclear; probably 34% decrease>2 IQ points active seizures 31% same-IQ 2 points 34% increase>2 IQ points 8–56 months Not stated Mean IQ drop from 82 to 67; 11 showed losses, 10 no change 21 months on Not stated 50% ‘‘moderate loss’’; 9% average no change; 41% ‘‘slight increase’’ 1–59 months; 14 Not well described 50% ‘‘negative change’’; 5% ‘‘no change’’; months average 45% ‘‘positive change’’; 9% clear deterioration 4 years on average Details provided No change in IQ for the group at length; active overall, but IQ>10 point loss in seizure disorders 11% who had frequent seizures, toxic drug levels, early epilepsy 3 years Patients had very Intelligence showed no losses from few seizures ages 3 to 7
measures of intelligence are not as sensitive to difficulties in brain functions as are validated neuropsychological tests [3]. Measures of intelligence were, after all, devised to predict how well children would do in school, and not to identify brain-related problems [3]. Also, neuropsychological tests clearly evaluate a broader range of functions than measures of intelligence. An additional limitation is obvious from a perusal of Table 1: very few data are presented on the type and frequency of seizures between the testings. Presumably, the children in all studies except one [10] had active seizures between the cognitive testings. Of the eight remaining studies, there was evidence for some cognitive loss in each study which exceeded any gains except for one investigation where no evidence for such a loss was found [5]. Furthermore, in about half the papers, significant IQ losses of 10 points or more were reported in about 10–25% of the children studied. While the data on seizure type and seizure frequency are so sparse that no connections between these variables and cognitive losses can be made, the weight of the evidence is that measurable losses in intelligence were experienced over time in at least a portion of the children with active seizures. The importance of this overall conclusion is underscored by the fact that all these studies had test–retest intervals
IQ scores slightly but not significantly down for total group; 24% had FSIQ losses>9 points Adverse IQ score changes found: PIQ (P < 0:05), FSIQ (P < 0:05)
of only 1–4 years. It appears that longer intervals would likely have resulted in more findings.
3. Studies involving adolescents and adults Results on the 13 longitudinal investigations with adolescents and adults are presented in Table 2. A more diverse group of tests were used in these studies, and only 5 of the 13 investigations rested on measures of intelligence alone. Also, the test–retest intervals tended to be longer, with three of the studies evaluating cognitive changes over 10-year periods. However, the seizure types and frequencies experienced during the test–retest intervals were no more precisely given than with children, with only a few exceptions [16,19,20,23]. In two studies with adolescents and adults [18,21], patients had few or no seizures during the course of the studies, and in both of these cases test performances either remained the same or improved. In the remaining 11 investigations summarized in Table 2, one can assume that most patients had at least a few seizures during the study periods even though in some studies a small number of people had complete seizure control. Of these 11 studies, evidence was found for a relationship
Table 2 Longitudinal cognitive studies of adults with epilepsy Investigator/year
Subject group
Test used
Test–retest interval
Seizure frequency
Cognitive outcome
Arieff and Yacorzynski/1942 [13]
90 adolescents and adults; 27 known etiology, 63 unknown etiology 56 adolescents and adults
Binet, 1916 and1937 versions
1–10 years
Wechsler–Bellevue, Form I (?)
At least 5 years, 7 years on average
Wide range of seizure control from complete remission to very active Wide range of seizure control
WAIS
1–5 years, 18 months
Varied by subgroup
Dodrill and Wilensky/1990 [16]
9 adults with status epilepticus during study period, and 9 closely matched controls with no history of status 69 adults who had participated in a vocational rehabilitation program 36 adults with no medication changes over a 5-year period 28 adults with TLE, typically with complaints of memory problems 35 adults with intractable partial seizures (60% also had secondary GTC seizures) 58 new cases with partial seizures 47 TLE cases
WAIS
5 years (6 months)
One group with status (4 GTC, 5 partial), one group without status
Finnish form of WAIS; tests of memory and some other functions Expanded Halstead–Reitan battery, WAIS WAIS-R, Wechsler Memory Scale (Form I)
Approximately 10 years
Improved in control over study, but only 10% controlled at end
Memory losses in 15-20%; intelligence unchanged overall; seizure frequency and losses in abilities difficult to connect
5 years (6 months)
Very few or no seizures
No changes on the tests beyond those expected by chance
2.3 years on average
8 partial seizures per month
Expanded Halstead–Reitan battery, WAIS
10 years (6 months)
9 partial and generalized seizures per month
Finnish form of the WAIS and other tests Measures of memory and other functions WAIS
5 years
Very few seizures of any kind
2–10 years, 4.7 years on average 6 years
Active partial seizures
Slight improvement in PIQ and FSIQ as would be expected by chance in a normal adult group WAIS little change; 6 of 17 other test variables showed losses (visual memory, attention, problem solving, perception) 12 of 24 test variables showed slight gains; none showed losses Verbal memory unchanged; especially figural memory showed loss (P < 0:05) No losses found; better PIQ (P < 0:01) and FSIQ (P < 0:05) may have been retest effects Normal group improved more than epilepsy group on 3 of 20 test variables; visual memory lost in both groups; partial seizures not correlated with ability changes, but 2 of 20 significantly correlated with number of GTC seizures; GTC status patients lost verbal and visual memory 50% showed significant decline in memory (verbal plus figural bined); few changes in nonmemory functions
Kalska/1991 [17]
Dodrill and Wilensky/1992 [18] Selwa et al./1994 [19]
Holmes et al./1998 [20]
Aikia and Kalviainen/1999 [21] Helmstaedter et al./2000 [22] Bjornaes et al./2001 [12]
17 adults who became surgical candidates
Dodrill/2002 [23]
35 adults with partial seizures with or without secondary generalization; 35 normal controls
Expanded Halstead–Reitan battery, WAIS
10 years (6 months)
Helmstaedter et al./2003 [24]
102 adults with TLE
Measures of memory and of some other cognitive functions
2–10 years, average of 57 months
Mild to moderate seizure severity (mean of 2.3 on 0–4 scale) Average of 1109 partial seizures and 61 GTC seizures over 10 years for epilepsy group; four patients experienced GTC status epilepticus during 10-year period
89% had more than 1/month at start, 72% had more than 1/ month at end of study
S23
47 adults; 22 improved in seizure frequency; 25 had no change or got worse
Rodin/1968 [14]
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
Seidenberg et al./1981 [15].
Significant loss of 6 IQ points if etiology was known; no loss when etiology was unknown Favorable changes in seizure frequency correlated. 33 (P < 0:02) with favorable changes in FSIQ Seizure improved group had on average 3–10 IQ point improvements; seizure unimproved group had )1 to +4 point changes The no status group improved more on FSIQ than did the status group (P < 0:05); same trend shown on VIQ
S24
C.B. Dodrill / Epilepsy & Behavior 5 (2004) S21–S24
between seizures and adverse cognitive changes in five [14–16,20,24], mixed or uncertain findings were noted in four [13,17,22,24], and no connection was found between seizures and changes in mental abilities in two [12,19]. While taken as a whole, these findings must be described as mixed; an examination of the last column in Table 2 reveals a series of findings, all of which would support a mild (but no more than a mild) connection between seizures and adverse cognitive change. Among these findings are the following: (1) statistically significant relationships between numbers of seizures (or presence of status epilepticus) and decreased scores on tests of abilities [14,16,23]; (2) changes in intelligence corresponding to changes in seizure frequency [15]; (3) better performances over time with normals than with people with epilepsy [23]; and (4) losses in mental abilities with patients having uncontrolled seizures which touch on a number of areas of cognitive functioning [20] despite the fact that a consensus exists that losses in memory are most common [17,22–24].
4. Discussion Based on the information at hand, it is concluded that losses in mental abilities do occur over time with uncontrolled seizures in patients with epilepsy. There is some evidence that adverse changes are more easily found in children than in adults, and this was shown in the review as a whole as well as in a particular study that permitted such a comparison [12]. The impression gained by this author is that losses in mental abilities are most easily connected with generalized tonic–clonic seizures, especially when experienced in a serial fashion (status epilepticus). This is difficult to demonstrate from a review of Tables 1 and 2, however. To more effectively answer the obvious unanswered questions pertaining to the cognitive effects of seizures, the major limitations in existing studies need to be addressed directly and deliberately. In particular, longitudinal prospective investigations are needed that include recording the numbers of the various types of seizures experienced throughout the study. This is the most glaring deficiency in the current investigations. It is also likely that long study periods (at least 25 years) are required before the true cognitive effects of single seizures in the lifetime of our patients can be adequately assessed. It is of importance that such studies include normal groups closely matched with the epilepsy samples for age and education at the beginning of the study as developmental and aging changes must be considered in such lengthy investigations. While these investigations will certainly prove to be laborious, their value to our patients will be great, and it is essential that they be undertaken.
References [1] Sutula T, Pitkanen A. Do seizures damage the brain? Prog Brain Res 2002;135:1–520. [2] Dodrill CB. Progressive decline in adolescents and adults with epilepsy. In: Sutula T, Pitkanen A, editors. Do seizures damage the brain? Prog Brain Res 2002;135:399–407. [3] Reitan RM, Wolfson D. The Halstead–Reitan Neuropsychological Test Battery. Tucson: Neuropsychology Press; 1993. [4] Fox JT. The response of epileptic children to mental and educational tests. Br J Med Psychol 1924;4:235–48. [5] Patterson HA, Fonner D. Some observations on the intelligence quotient in epileptics. Psychiatr Q 1928;2:542–8. [6] Dawson S, Conn JCM. The intelligence of epileptic children. Arch Dis Child 1929;4:142–51. [7] Fetterman J, Barnes MR. Serial studies of the intelligence of patients with epilepsy. Arch Neurol Psychiatry 1934;32:797–801. [8] Sullivan EB, Gahagan L. On intelligence of epileptic children. Genet Psychol Monog 1935;17:309–75. [9] Bourgeois BFD, Prensky AL, Palkes HS, Talent BK, Buxch SG. Intelligence in epilepsy: a prospective study in children. Ann Neurol 1983;14:438–44. [10] Ellenberg JH, Hirtz DG, Nelson KB. Do seizures in children cause intellectual deterioration? N Engl J Med 1986;314:1085–8. [11] Aldenkamp AP, Alpherts WCJ, Bruine-Seeder DD, Dekker MJA. Test–retest variability in children with epilepsy: a comparison of WISC-R profiles. Epilepsy Res 1990;7:165–72. [12] Bjornaes H, Stabell K, Henriksen O, Loyning Y. The effects of refractory epilepsy on intellectual functioning in children and adults: a longitudinal study. Seizure 2001;10:250–9. [13] Arieff AJ, Yacorzynski GK. Deterioration of patients with organic epilepsy. J Nerv Ment Dis 1942;96:49–55. [14] Rodin EA. The prognosis of patients with epilepsy. Springfield, IL: Charles C Thomas; 1968. [15] Seidenberg M, OÕLeary DS, Berent S, Boll T. Changes in seizure frequency and test–retest scores on the Wechsler Adult Intelligence Scale. Epilepsia 1981;22:75–83. [16] Dodrill CB, Wilensky AJ. Intellectual impairment as an outcome of status epilepticus. Neurology 1990;40(Suppl. 2):23–7. [17] Kalska H. Cognitive changes in epilepsy: a ten-year follow-up. In: Commentationes Scientiarum Socialium, vol. 44. Helsinki: Societas Scientiarum Fennica; 1991. p. 1–85. [18] Dodrill CB, Wilensky AJ. Neuropsychological abilities before and after 5 years of stable antiepileptic drug therapy. Epilepsia 1992;33:327–34. [19] Selwa LM, Berent S, Giordani B, Henry TR, Buchtel HA, Ross DA. Serial cognitive testing in temporal lobe epilepsy: longitudinal changes with medical and surgical therapies. Epilepsia 1994;35:743–9. [20] Holmes MD, Dodrill CB, Wilkus RJ, Ojemann LM, Ojemann GA. Is partial epilepsy progressive? Ten-year follow-up of EEG and neuropsychological changes in adults with partial seizures. Epilepsia 1998;39:1189–93. [21] Aikia M, Kalviainen R. Five-year follow-up of cognitive performance of adult patients with well-controlled partial epilepsy. Epilepsia 1999;40(Suppl. 2):100–1. [22] Helmstaedter C, Kurthen M, Lux S, et al. Temporal lobe epilepsy: longitudinal clinical, neuropsychological and psychosocial followup of surgically and conservatively managed patients. Nervenarzt 2000;71:639–42 [in German]. [23] Dodrill CB. Progressive cognitive decline in adolescents and adults with epilepsy. In: Sutula T, Pitkanen A, editors. Do seizures damage the brain? Prog Brain Res 2002;135:399–407. [24] Helmstaedter C, Kurthen M, Lux S, Reuber M, Elger CE. Chronic epilepsy and cognition: a longitudinal study in temporal lobe epilepsy. Ann Neurol 2003;54:425–32.