1. Department of Medicine, University Teaching Hospital, Benin City, Nigeria
  2. Division of Neurology, Southern Illinois University Chicago Illinois
  3. Neurology Unit, Dept. of Medicine, UCH, PMB 5116, Ibadan, Oyo State, Nigeria

E-Mail Contact - OGUNRIN Olubunmi : bfunmi@uniben.edu



Epilepsy is particularly highly prevalent in developing African countries and has been associated with cognitive disturbances, but more importantly is the contribution of the anti-epileptic drugs (AEDs).


This study aimed at comparing the effects of AEDs on the cognitive functions of Nigerian epileptic patients.


This is a prospective study of 55 consecutive patients with epilepsy, aged 14 years and above, over a two year period (October 2000 to October 2002), recruited from the Neurology Clinic of the University Teaching Hospital, Benin City, Nigeria. Anti-epileptic treatment with either carbamazepine (19 patients), phenytoin (18 patients), or phenobarbitone (18 patients) which was randomly assigned constituted the interventional measure.
Cognitive testing, using the Iron Psychology (FePsy) a computerized neuro-psychological test battery, measured the visual and auditory reaction times, the continuous performance test and the recognition memory test to assess the mental speed, attention and memory respectively.


The effect of the individual drug on cognitive performance revealed significant impairment of mental speed (p<0.001) with the exemption of improved performance with phenytoin on auditory reaction time (p>0.05). Carbamazepine did not significantly affect the verbal (Words section) memory scores (p>0.05) implying better performance in tasks of verbal memory (p<0.05). All the three anti-epileptic drugs strongly reduced the attention abilities of the patients (p<0.001). Patients on phenobarbitone had the worst scores in both the verbal and non-verbal memory tasks. Conclusion
The results of this study will be useful in the rationale selection of anti-epileptic drugs with the objective of minimizing, as much as possible, their cognitive side effects.



La prévalence de l’épilepsie est élevée en Afrique de même que les troubles cognitifs associés à l’utilisation des antiépileptiques (AEDs).


Le but de l’étude est de comparer les effets des AEDs sur les fonctions cognitives d’une population épileptique nigériane.


L’étude prospective, randomizée, concerne 55 patients épileptiques âgés de 14 ans et plus, étudiés sur une période de 2 ans (octobre 2000 à octobre 2002) et recrutés dans le service de neurologie du CHU de Benin City. Le traitement était soit la carbamazépine, (19 patients) soit la phénitoïne (18 patients), soit du phénobarbital (18 patients)
Les tests cognitifs utilisant le Iron Psychology (FePsy), tests psychologiques, ont mesuré les temps de réaction auditif et visuels, les tests de performance de la mémoire mentale, de l’attention et de la mémoire.


Chacun des médicaments a entrainé une perturbation de la vivacité mentale (p < 0.001) à l'exception d'une amélioration du temps de réaction auditive avec la phénitoïne (p > 0.05). La carbamazepine n’a pas affecté de manière significative la mémoire verbale (mots) (p > 0.05). Tous les autres antiépileptiques ont réduit les capacités d’attention des patients ( p < 0.001) Les patients sous phénobabital ont eu les plus mauvais scores. Conclusion
Les résultats de cette étude aideront à une utilistion rationnelle des anti-épileptiques afin d’éviter les effets cognitifs.

Keywords : Africa, Anti-epileptic drugs, Epilepsy, Memory, Mental speed, Nigeria, Afrique, Antiépileptiques, Epilepsie, Mémoire. Nigéria, Tests psychologiques


Epilepsy is the second most common disorder of the central nervous system, affecting 1% of the human population (5). It is particularly highly prevalent in developing African countries (10,12). Epilepsy has been associated with cognitive disturbances, but more importantly is the contribution of the anti-epileptic drugs (AEDs).
Anti-epileptic drugs (AEDs) have adverse effects on cognitive functions such as attention, memory and psychomotor speed. There is evidence that drug-induced cognitive impairment has great impact on critical daily life function of patients with epilepsy (3,7). The cognitive effects of AEDs are of special concern because they are iatrogenically induced (3).
Comparative studies of AEDs in developed countries have shown that while carbamazepine had minimal side effects, phenytoin and phenobarbitone have less favorable cognitive side effects with impairments of mental speed, memory and attention (4,13,14). These latter two drugs are however the most widely used in developing countries, because of good efficacy, broad spectrum of activity and low cost (11).
There is little information available in the literature on the effect of AEDs on the cognitive functions of African epileptic patients. This study was aimed at comparing the effects of the commonly used AEDs on the cognitive performances of Nigerian patients with epilepsy. The results of such a study will be useful in the rational selection of anti-epileptic drugs and in the educational and vocational counseling of patients. Moreover, this study provides another opportunity to gain more information about the older AEDs, i.e phenobarbitone and phenytoin.


Patient’s Selection

All the patients that presented with recurrent afebrile seizures in the outpatient neurology clinic of the University Teaching Hospital Benin City, Nigeria between October 2000 and October 2002 were recruited i.e. a total of 147 patients.

A total of 55 consecutive patients aged 14 years and above, satisfied the inclusion criteria. Patients with psychiatric illness, mental sub-normality and those with progressive neurological disorders were excluded from the study. An EEG was not required for diagnosis, as epilepsy is a clinical diagnosis. These patients have not been on AED prior to presentation.

Patient’s Demographic Data

All patients completed questionnaires designed to obtain demographic information on the age, sex, level of education, age at onset of seizures, frequency of seizures and type of seizures.

The seizure types were classified clinically, based on the International League against Epilepsy (ILAE) classification of 1981 (6). The patients comprised 35 men and 20 women, with a mean age of 29.8 (+/- 12.47) years. Seventeen patients had primary education, 26 had secondary education and 12 had post- secondary education. Forty-five patients had primary or secondary generalized seizures while 10 had partial seizures. Laboratory investigations done on the patients included:

(a)Liver function tests (b) Urea and electrolytes (c) Blood sugar estimations and (d) if indicated, lumbar puncture for cerebrospinal analysis.

Patients and Drug Therapy

Patients and Drug Therapy

After initial diagnoses, the patients were randomly assigned to anti-epileptic medications, with nineteen (19) patients receiving carbamazepine (CBZ), eighteen (18) receiving phenobarbitone (PB) and eighteen (18) phenytoin (PHT).

The treatment was chosen at random.

(A) Carbamazepine group

The median dosage for these patients was 600 mg per day (range 400-1200 mg) with a modal seizure frequency of 3-4 attacks per month, and mean duration of seizures of 6.5 years.

(B) Phenytoin group

The median dosage was 300 mg per day (range 200-300 mg) with a modal seizure frequency of 2 attacks per month, and mean duration of seizures of 5.5 years.

(C) Phenobarbitone group

The median dosage was 120 mg per day (range 60-180mg) with a modal seizure frequency of 3-4 attacks per month and a mean duration of 7.5 years.

There was no significant difference in the three groups of patients at recruitment in terms of seizure variables or demographic data. (Refer Table 1)
The patients’ cognitive performances were assessed after three months of AED therapy, following the initial pre-medication assessment at recruitment. AEDs were available freely to all patients in the study. At the initial visit each patient was given a health talk on AED compliance and was well motivated to take AEDs because of the stigma locally attached to epilepsy.
There were no facilities for AED serum level estimations in our hospital where the study was conducted.

Cognitive Testing

The administration of tests was done with the FePsy computerized neuropsychological test battery (1, 3, 8, 9). Test presentation and response registration were controlled by a microcomputer, but one of the authors (O.O) was always present to adjust instructions to the individual performance level of the patients. The cognitive assessment was not blind as the author (O.O) was aware of the patients’ drug status. However, this did not affect result of neuropsychological testing as it is computer-controlled, and the evaluation of performance is not dependent on the author, but on the testees’ cognitive abilities.
(A)Short-term Memory was assessed using the Recognition Memory test (RMT). The test involves the use of study items which consist of 3 or 4 figures (for the visual, non-verbal memory test) and 4 or 6 words (for the verbal memory test) which are presented simultaneously.

The task is divided into a study phase in which the material to be remembered is presented and a test phase in which recognition is tested. A delay of two seconds is allowed between study phase and the test phase. In the test phase, the figures or words are presented again and the target item has to be recognized. Patients with primary school education were tested using 3 figures and 4 words, while those with secondary and post-secondary education were tested using 4 figures and 6 words. The results were calculated as percentage of correct responses. The evaluation of the recognition task was performed in the context of the short-term memory function (8).
(B)Psychomotor Speed was assessed by the Simple Reaction Time. In the auditory version the testee was asked to react as quickly as possible to sound stimuli of 800 hertz generated by the computer. For the visual version, the testee reacts as quickly as possible to a white square in the middle of the computer screen. In both versions the inter-stimulus interval is randomly varied form 2.5 to 4 seconds. The evaluation of the results was done within the context of speed of information processing and alertness functions (8).

(C)Attention was assessed using the Continous Performance Test, which involves the display of a string of eight characters, either ‘XXXXXXXX’ or ‘XAXXXXXX’. The stimuli are presented during a short (200 msec) period. The testee has to respond (by pressing a key on the keyboard) to the appearance of a character ‘A’ at a random position in the stimulus string. The results are computed by the FePsy software, according to a signal detection model (8).

Statistical Analysis

The patients were divided into three categories based on the type of AEDs received. The data from each cognitive task was separately submitted to a one-way analysis of co-variance with cognitive performance as the dependent variable and AED medication as the independent variable. Significance level was set at 5%.
Data comparing performances pre- and post- medication were tested using the Student’s ‘t’ test with level of significance set at p< 0.05. RESULTS

Reaction Times (Mental speed)

The comparison of the data on the cognitive performances of the patients pre- and post medication revealed statistically significant adverse effect on the mental speed, with phenobarbitone group having the worst scores ( p< 0.001 ). On the other hand, phenytoin had a favorable effect on the auditory reaction time (p> 0.05). Refer Tables 11b and 11c. But when the effects of all the three AEDs were compared, there was no statistically significant difference between the three drug groups in both auditory and visual reaction times. Refer Table 111.


The carbamazepine group performed better than those on phenytoin and phenobarbitone in tasks of both verbal Words and visual, non-verbal memory Figures but only the difference in verbal memory performance reached statistical significance (p < 0.05). Refer Table 111. The effect of carbamazepine on verbal and visual memory was also favorable when the three AEDs were compared post medication (p>0.05) refer table 11a. Patients on phenobarbitone had the worst scores in both verbal and non-verbal memory tasks. Refer Table 111.


All the three drug groups showed significant worse performances in attention when the pre- and post- medication Continuous Performance test scores d’ were compared (p< 0.05) without significant difference in their response bias(p>0.05). But the scores reflected better perceptual sensitivity and sustained attention in patients on phenytoin and carbamazepine compared to those on phenobarbitone; though these differences did not reach statistical significance. Refer Table 111.


This study was aimed at assessing the effects of phenobarbitone, phenytoin and carbamazepine on cognitive function in newly diagnosed Nigerians with epilepsy. The patients were randomly assigned to one of the three AEDs and their cognitive function assessed by computer at recruitment and three months after commencement of medications. The computer-based test battery has been shown to be highly sensitive in detecting cognitive dysfunction in patients with epilepsy (8, 9). This setting is unique as similar studies have not been carried out in Africans with epilepsy.

The study showed significant deterioration in cognitive performances with all the three drugs when their scores before and after medication were compared, with the only exemption being the improved performance in the auditory reaction time in patients on phenytoin and significant improvement in verbal memory scores in the carbamazepine group.

However, there were no significant differences in auditory and visual reaction times between the three AEDs. Although on the whole, reaction times were longer for the patients on phenobarbitone. Patients on carbamazepine performed significantly better in verbal memory tasks, and also had a trend towards better performances in non-verbal memory. Patients on phenobarbitone had the worst performance in the tests of sustained attention, though there was no difference in the response bias of the three groups of patients. This implies that this observation is not due to chance.

Our study reconfirms in Africans the observations in the literature that carbamazepine has a relatively favorable cognitive profile compared to phenytoin and phenobarbitone (1, 2, 4, 13, 14). Though it would have been desirable to correlate the cognitive performances with serum levels of these drugs, this facility was not available in our center.


Our study has confirmed the presence of cognitive impairments in epilepsy prior to administration of, and the worsening of these by, antiepileptic medications commonly used in Nigeria and other African countries. There is significant deterioration in attention abilities, mental speed and memory performance. However, carbamazepine appears to have a more favorable profile than phenytoin and phenobarbitone.

Coping with day-to-day activities demands unimpaired cognitive abilities including memory, attention and psychomotor speed. For instance, the learning process requires sustained attention and intact memory. Although costs and the relative unavailability of health care systems will not allow the ideal selection of AEDs in all circumstances, the results of this study suggests the need for careful follow up of African patients on AEDs, especially phenobarbitone, in other to minimize cognitive side effects and to maximize quality of life.
There is need for simple tests of cognitive function that can screen patients with epilepsy and determine the need for more sophisticated tests of cognitive function, such as computerized test used in this study.

TABLE 1 – Descriptive details of the patients

Variables Carbamazepine (Group N= 19) Phenobarbitone (Group N= 18) PHENYTOIN (Group N=18)
Sex M/F 12 / 7 12 / 6 11 / 7
Mean age [years] 27.8 +/- 8.9 18.7 +/- 5.6 35 +/- 14.7
Range [years] 14 – 38 15 – 26 29 – 55
Level of Education
Nil 0 0 0
Primary 5 6 6
Secondary 10 8 8
Post-Secondary 4 4 4
Mean seizure frequency 3-4 attacks/month 2 attacks/month 3 attacks/month
Mean duration of epilepsy 6.5 years 5.5 years 7.5 years
Seizure Type
Generalized 16 15 14
Partial 3 3 4

TABLE 2 – The Effect of AEDs on Mean Cognitive Performances – [pre- and post-medication]

TABLE 2a – Carbamazepine Group ( n = 19)

Cognitive Tests Pre Drug Post Drug Level of significance
Auditory Reaction Time (milliseconds) 473.13 500.27 P < 0.001
Visual Reaction time (milliseconds) 423.05 461.60 P < 0.001
Recognition Memory Test (% correct)
Words (verbal) 41.97 44.0 P > 0.05 (better response)
Figures (non-verbal) 34.17 40.31 P > 0.05 (better response)
Continuous Performance Test
Perceptual sensitivity 0.63 0.69 P > 0.05
Response Bias 0.79 0.75 P > 0.05

TABLE 2b – Phenytoin Group ( n = 18)

Cognitive Tests Pre Drug Post Drug Level of significance
Auditory Reaction Time (milliseconds) 443.28 442.50 P > 0.05
Visual Reaction time (milliseconds) 307.5 370.29 P < 0.001
Recognition Memory Test (% correct)
Words (verbal) 44.86 34.10 P < 0.05
Figures (non-verbal) 38.05 32.01 P < 0.05
Continuous Performance Test
Perceptual sensitivity 0.94 0.87 P < 0.05
Response Bias 0.81 0.79 P > 0.05

TABLE 2c – Phenobarbitone Group ( n = 18)

Cognitive Tests Pre Drug Post Drug Level of significance
Auditory Reaction Time (milliseconds) 400.20 503.00 P < 0.001
Visual Reaction time (milliseconds) 435.90 544.20 P < 0.001
Recognition Memory Test (% correct)
Words (verbal) 43.80 18.75 P < 0.001
Figures (non-verbal) 26.19 22.38 P < 0.05
Continuous Performance Test
Perceptual sensitivity 0.25 0.15 P < 0.01

TABLE 3 – The Effect of Anti-Epileptic Drugs on Mean Cognitive Performances

Cognitive Tests Carbamazepine (N=19) Phenobarbitone (N=18) Phenytoin (N=18)
Auditory Reaction Time (Milliseconds)
Dominant Hand 500.27 503.0 442.50 NS
Non-Dominant Hand 434.00 532.00 420.20 NS
Visual Reaction Time (Milliseconds)
Dominant Hand 461.60 544.20 370.29 NS
Non-Dominant Hand 439.3 453.20 357.20 NS
Recognition Memory Test (% Correct)
Words 44.00 18.75 34.10 P<0.05 S
Figures 40.31 26.19 # 32.01 # NS #
Continuous Performance Test – Vigilance
Perceptual Sensitivity 0.69 0.15 0.87 NS
Response Bias 0.75 0.52 0.79 NS
Response Bias 0.55 0.52 P > 0.05


  1. ALDENKAMP AP, ALPHERTS WCJ, BLENNOW G, et al: Withdrawal of antiepileptic medication-effects on cognitive function in children- the results of the multicenter ‘Holmfrid’ study. Neurology 1993;43(1):41-51.
  2. ALDENKAMP AP, VERMEILEN J. Phenytoin and Carbamazepine: Differential effects on cognitive function. Seizures 1995; 4: 95-104.
  3. ALPHERT WCJ, ALDENKAMP AP. Neuropsychological assessment of cognitive functioning in children with epilepsy. Epilepsia 1990; 31(suppl. 4) S35-S40.
  4. ANDREWS DG, BULLEN JG, TOMLINSON L, et al: A comparative study of the cognitive effects of phenytoin and carbamazepine in new referrals with epilepsy. Epilepsia 1986; 27: 128-34.
  5. CHANG B, LOWENSTEIN D. Mechanisms of disease: epilepsy. N Engl J Med. 2003; 349: 1257-1266.
  6. Commission on Classification and Terminology, International League against Epilepsy. Proposed revisions of clinical and electro-encephalographic classification of epileptic seizures. Epilepsia 1981; 22: 480-501.
  7. DRANE DL, MEADOR KJ. Epilepsy, anticonvulsant drugs and cognition. Clin Neurol. 1996; 5(4): 877-85.
  8. MOERLAND MC, ALDENKAMP AP, ALPHERTS WCJ. Computerized psychological testing in epilepsy. In: Moarse F.J Mulder L.J.M. Syouw W.P.B. and Aldenkamp A.P. Computers in Psychology, Methods, Instrumentation and Psychodiagnostics. Swets and Zertlinger, 1988. Lisse, USA. 157-164.
  9. OGUNRIN O, ADAMOLEKUN B, OGUNNNIYI A, ALDENKAMP AP. Cognitive functions in Newly Diagnosed Nigerians with Epilepsy. Can J Neurol Sci. 2000; Vol. 27(2): 148-151.
  10. OSUNTOKUN BO. Epilepsy in Africa. Trop Geogr Med. 1978; 31: 24-31.
  11. OSUNTOKUN BO, ADEUJA AO, NOTTIDGE VA, et al: Prevalence of the epilepsy in Nigerian Africans: a community-based study. Epilepsia 1987; 28(3), 272-79.
  12. SENANAYAKE N, ROMAN GC. Epidemiology of epilepsy in developing countries. Bull World Health Organization. 1993; 71(2):247-58.
  13. SMITH DB, MATTSON RH, CRAMER TA, et al: Results of a Nationwide Veterans Administration Cooperative Study comparing the efficacy and toxicity of carbamazepine, phenobarbital, phenytoin and primidone. Epilepsia 1987; 28(suppl. 3) S50- S58.
  14. THOMPSON PJ, TRIMBLE MR. Anticonvulsant drugs and cognitive functions. Epilepsia 1982; 23: 531-44.

© 2002-2018 African Journal of Neurological Sciences.
All rights reserved. Terms of use.
Tous droits réservés. Termes d'Utilisation.
ISSN: 1992-2647