1. Department of Neuropsychiatry, University Hospital, Mahajanga, Madagascar

E-Mail Contact - ANDRIANTSEHENO L. M. :


A retrospective study on 213 files of epileptic patients issued from a sample of 741 referrals to the University Hospital ofMajunga, Neuropsychiatry Unit, between June 1, 1993 December, 31, 1995 was reported. Mahajanga, the chief-town of the Province is a hot and dry region where malaria, shistosomiasis, tuberculosis and syphilis are endemic. As in the majority of tropical series, the most involved age-group was between 10 and 30 years; partial seizures had a higher rate compared to the generalized forms (52% vs 46%). The most prominent risk factors were linked to parasitosis (neurocysticercosis), to defective obstetrical cares (perinatal injuries) and infatile conditions (febrile convulsions). The lackofneurodiagnosis facilities (EEG, EMG, brain imaging) restrains notably the aetiological researches. In Madagascar, the compliance to a long lasting and expensive treatment set a problem (due to the indigence of the population, and a bad distribution of antiepileptic drugs).


Une etude retrospective des dossiers de 213 epiletiques tires d’une population de 741 malades admis au niveau d’un service de Neuropsychiatrie de Mahajanga (du 01.06.93 au 31.12.95) est rapportee. Cette ville, chef-lieu de la province du meme nom a un climat et sec ou Ie paludisme, la bilharziose, la tuberculose et la syphilis sont endemiques. Comme dans la plupart des series tropicales, la tranche d’age la plus atteint est celle comprise entre 10 et 30 ans; les crises partielles predominent sur les formes generalisees (52% vs 46%). Les facteurs de risque les plus importants sont d’origine parasitaire (neurocysticercose) ou lies aux conditions obstetricales (souffrances perinatales) et pediatriques precaires (convulsions febriles). Le manque d’equipements de neurodiagnostic (EEG, EMG, imageries cerebrales) limite considerablement les recherches etiologiques. A Madagascar se pose aussi le probleme de la compliance a un traitement prolonge et couteux (indigence de la population, distribution defectueuse des antiepileptiques).

Keywords : Epidemiology, Epilepsy, Madagascar, Africa


In a recent Press release (1), the World Health Organisation (WHO) launched, jointly with two non-governmental organisations (the International League Agaisnt Epilepsy, ILAE, and the International Bureau for Epilepsy, IBE), a worldwide campaign against epilepsy. The ILAE represents doctors and scientists, whereas the Bureau acts for patients and their families. According to this influential institution, the average prevalence of active epilepsy in the general population (i.e. continuing seizures or needing for treatment) is around 7 per 1,000 inhabitants (Fact Sheet/WHO n° 165, June 1997). The burden of epilepsy is considerable: a study from UK has estimated the cost of active epilepsy per patient at £ 4167 per annum, that of inactive epilepsy at £ 1630 per annum (Fact Sheet/WHO no166, June 1997). If indirect cost was added, the expenditure would be far larger. In developing countries, where the prevalence rate is twice-fold higher (10-15 per 1,000 inhabitants) than in developed countries [2|, epidemiological data are insufficient or of uneven interest.

Anyhow, three-quarters of the affected people (up to 30 millions, 90% of whom are in developing countries) do not receive any treatment at all, and must be brought out of the shadows (1).

In Madagascar, a tropical Island in the Indian Ocean, no epidemiological data from the general population is available. In the present hospital-based study, we report some aspects of epilepsy from the Northwestern part of the island.


The study took place in the only Department of Neurospsychiatry, expected to serve the whole province of Mahajanga (1,379,000 inhabitants). In fact, geographical isolation (seasonal roads), economic restraints, preference for traditional healers and sometimes pure fatalism, limited the access of the Department to the only inhabitants of Mahajanga City and its nearest surroundings. The province has a dry and hot climate, where malaria, schistosomiasis, tuberculosis and syphilis are endemic. The study population (3) was young (44% of the people were less than 15 years-old, only 3.20% were 65 years-old or more), and had a low instruction level (table 1).

Table 1: Comparative instruction level of patients versus that of the general population.

Instrucion level of patients Epileptic patients (Number of cases, n=213) Epileptic patients (Percentage, %) General population (n=1.379000)
Illiterate & Primary school at the best 119 56 84.9
Secondary school & High school 94 44 15.1

Ethnical groups distribution is plotted on table 2. Certain ethnics (Comorois, Sakalava...) have a strong Islamic influence.

Table 2: Comparative distribution of different ethnical groups (in %), between epileptic patients and the population of MAHAJANGA City.

Ethnical Groups Epileptic Patients (Number of cases) Epileptic Patients (Percentage) Population of Mahajanga city (percentage)
Merina 85 40 36.1
Betsileo 30 14 16.1
Antaisaka 27 12.5 12.5
Tsimihety 27 12.5 13.9
Sakalava 15 7 9.6
Other groups 29 13 11

The study itself consisted in a retrospective review of patients’ files. Neurological patients of 1 year-old or more, who have been attending the Department, between June 1, 1993 and December 31,1995, were included.

Basic neurodiagnosis facilities (EEG, EMG, brain imageries) were lacking, so case ascertainment-were set up only with clinical criteria, routine laboratory exams and therapeutic tests.

We used the International Classification of seizures, ICES, 1981 version (4) without EEG data, recognised by some authorities (5).

Term definitions were extracted from the guidelines of the ILAE (5)]: febrile seizures, single or isolate seizures were thus excluded.


The epilepsies 0^=213) represented 28.75% of all new neurological referrals (n=741) for the period (table 3).

Table 3: Aetiological repartition of neurological referrals (in percentage) during the study period.

Nature of the disease Number Rate (%)
Total n=741 100
Epilepsies 213 28.75
Chronic headaches 155 20.95
Peripheral neuropathies 102 13.75
Cerebrovascular diseases 83 11.3
Infection of the CNS 69 9.35
Cerebral palsy 42 5.6
Degenerative diseases 26 3.5
Tumoral sydromes of the CNS 24 3.2
Sleep disorders 6 0.8
Traumatic brain injuries 3 0.4
Myopathies 3 2
Unclassifiable disorders 15 2

It was possible to classify 98% (208/213) of the seizures (table 4).

Table 4: Comparative patterns of epilepsies between series from developed countries.

Seizure types Present series (Number of cases) Present series (Percentage %) Indian series (11)(%) Nigerian series (10)(%) Caucasian series (13) (%)
Total 213 100 100 100 100
Unclassifiable seizures 5 2 19 3 24
Classifiable seizures 208 98 81 97 76
- Generalised seizures 97 46 20 23.4 38
Tonic clonic (Grand Mal) 89 42 9.8 - -
Absence (Petit Mal) 2 1 3 1.2 10
Myoclonic 2 1 2 0.98 4
Others - - - - -
- Partial seizures 111 52 80 76.6 62
Simple 42 20 58 25.4 10
Complex 10 4.5 7 34 40
Secondarily generalised 59 27.5 15 17.2 12

The mean age of the patients was 23.96 years with a dispersion varying between 1 year and 73 years. More than 1 patient out of 2 were less than 20 years-old (table 5).

Table 5: Age distribution of epileptic patients (in percentage).

Age (years) Epileptic patients (Number of cases) Epileptic patients (percentage, %)
0-9 46 22
10-19 64 30
20-29 54 25
30-39 26 12
40-49 15 7
+50 8 4

Assumed secondary epilepsies supervened significally later than primary ones Mean age of occurrence respectively 30.56 years and 17.36 years).

There was male predominance (sex ratio=l 27/86=1.48).

Certain ethnical groups (Merina, Betsileo) were over represented (table 1). Forty two per cent (42%) of the patients (90/213) were students, 36% were unemployed (77/213).

Farmers represented only 6,8% of care-attenders (14/213); it confirms the preferential recruitment of urban city dwellers. Epileptic patients had a significantly higher level of instruction, compared to the general population (table 1).

Eighty per cent (80%) of cases (170/213) had at least one crisis per year. Partial seizures accounted for 52% of cases (111/213); among them, the simple type was prevalent (table 4). Forty four per cent (44%) of the epilepsies (93/213) were deemed as primary. Among the 120 remaining cases, 55 were cryptogenic, whereas 65 had an imputable cause among which neurocysticercosis was the most frequent (table 6).

Table 6: Likely aetiologies of secondary epilepsies, referred to partial and all types of seizures.

Assumed aetiologies Number of cases Percentage, referred to partial seizures (n=1110) Percentage, referred to all types of seizures (n=213)
Neurocysticercosis 27 24.32% 12.67%
Perinatal brain insults 20 18.00% 9.38%
Traumatic brain injuries 7 6.30% 3.28%
Miscellaneous encephalopathies 6 5.40% 2.82%
Alcoholism 4 3.60% 1.87%
Phakomatosis 1 0.90% 0.46%

Sixteen per cent (165%) of the patients (34/213) had a previous history of febrile convulsion, 13% of them (28/213) had a minor or specific cerebral dsyfunctions and 5.5% (12/213) had a mental retardation.

The physician rarely witnessed an attack, so the clinical disgnosis of epilepsy was essentially anamnestic.

In regard to « Grand Mal» seizure, ten criteria were used to identify it in rural area surveys in Africa, Asia, and South America (6).

The basic sign was the loss of consciousness for 1 to 30 mn:, only 15% of our crisis attenders were able to give the right duration of the seizures.

The second major criteria was the sequence of a tonic, then a clonic phase of movements: here again, only 25% of witnesses could describe it accurately. We find it more informative to make people mimick crisis.

Foaming at the mouth was reported in 80% of cases, but it was also observed in some hysterical attacks.

Lateral tongue biting (57%), sphincter disturbances (47%) and fall stigma (40%) are more specific but in our observations they were not very sensitive.

Phenobarbital (PB) was prescribed in 83.50% of cases (178 prescriptions out of 213), and was generally well tolerated (in 85% of patients).

Carbamazepine (7.50% of prescriptions) exposed to further treatment gap, because of its high cost and its bad distribution.

The effect of the treatment was good; 55% of the patients (118/213) had 50% of their seizures controlled; yet, long term follow-up was uncertain.


Our hospital-based study is not designed to assess morbidity rates (incidence, prevalence), because of selection biases. Moreover, they are not representative of all the Malagasy population, which is made of slightly different cultural and ecosystemic subunits. In spite of these drawbacks, the first rank occupied by epilepsies among our neurological referrals, indirectly emphasizes their importance for the Public Health. Like in most of the studies from developing countries (6,7,8), epilepsies involve younger peoples; in industrialized countries, they display a bimodal distribution (9). Our 0-9 years-old class is surely underrated because of parallel recruitment in Paediatry. As in the majority of series, there is male predominance. Neither sex, nor ethnic factors have any influence; the apparent overrepresentation of certain ethnies (Merina, Betsileo) results of the cumulative effect of their numerical superiority in the general population and their preference for medical resources. The distribution pattern of patients’ profession and their academic level can be explained by socioeconomical factors: recruitment of underprivileged people (students, unemployed persons) in public health informations, low hospital frequentation of country-people for different reasons afore-mentioned. The characteristics of epilepsies are similar to those described in most tropical seizures, and the underevaluation of « Petit Mal ». European data display a far higher rate for this latter (13). In the absence of sophisticated investigations, the careful research of focal neurological symptoms is helpful to diagnose focal crisis secondarily generalized. (table 3). The features of partial seizures vary according to the prevailing risk factor: Nigerian searchers report a higher rate of complex partial seizures (10), which they associate with a past history of febrile convulsions; recent findings support this assumption (14). Contrariwise, simple partial seizures are predominant in Indian series (11), where head injuries and CNS infections (including neurocysticercosis) are the main risk factors. Our pattern is closer to the Indian type, and neurocysticercosis is the first aetiology disclosed, however the Islamic influence on behaviours (prohibition of pork consumption) mitigate the importance of this parasitosis in Mahajanga City: 12.67% of cases (table 6) versus 51% from Ecuador (15), 6% of pathologic specimen vs 40% in Antananarivo (16), were pork consumption is larger. Cautious anamnesis and examination are fundamental for epilepsy diagnosis in the absence of instrumental investigations. Population knowledge, local representation, beliefs and attitudes toward the disease must be taken in account to gather relevant informations. Bad compliance remains a thorny problem for the long term treatment of epilepsy in the Third World. Drug unavailability and expensiveness, lack of inadequacy of informations, and illiteracy are the main contributors to this drawback. The worse outcome of our cases compared to that of some comprehensive studies supports these conclusions: 55.40% of good results versus 78% for Feski’s(17).


The global campaign launched by the WHO against epilepsy, expresses the importance of this disease. This stigmatizing illness constitute a real Public Health problem for the Third World. Madagascar, globally shares the same lack of facilities and the same patterns of the disease as other poor tropical countries. The high rate of partial seizures results from avoidable causes such as, febrile convulsions, perinatal and infectious brain insults. Struggle against known risk factors, education and sensitization of the both the population and the primary caredeliverers, endowment of essential drugs, are the best options recommended to face this disease, in developing countries.


  1. Bringing out of the shadows. A global campaign is launched. PRESS RELEASE WHO/48, June 1997 -.1-3.
  2. Commission on Tropical diseases of the 1LAE. Relationships between epilepsy and tropical diseases. EPILEPSIA 1994; 35(1):89-93.
  3. Commission Nationale de Recensement General la Population et de 1’habitat. Recensement general de la Population et de 1’habitat de Madagascar (1993). INSTAT, Vol 1 (4), Farinaty ede Mahajanga.Nov 1996.
  4. Commission on classification and Terminology of 1LAE. Proposal for revised clinical and EEG classification of epileptic seizures. EPILEPSIA 1981; 22:489-501.
  5. Commission on Epidemiology and prognosis of the 1LAE. Guidelines for epidemiological studies on Epilepsy. EPILEPSIA1993;34(4): 592-596.
  6. SHORVON SD, FARMER PJ. Epilepsy in developing countries: A review of epidemiological, sociocultural and treatment aspects. EPILEPSIA 1988,29 (supp I): S36-S54.
  7. OGUNNIYI A, OSUNTOKUN BO, BADEMOSI 0 et al. Risk factors for epilepsy case control study in Nigerians. EPILEPSIA 1987;28(3):280-285.
  8. FARNARIER G, MOUBEKA-MOUNGUENGUI M, KOUNA P, ASSENGONE-ZEH Y, GUEYE L. Epilepsies dans les pays tropicaux en vote de developpement: etude de quelques indicateurs de sante. EPILEPSIES 1996.8:189-213.9. Durand G, Jallon P. Epidemiologie et etiologies des epilepsies. Ed. Techniques -ENCYCL. MED CHIP. (Paris). Neurologie, 17045.A.35,1994,10 P.
  9. DANESI MA. Classification of the Epilepsy. An investigation of 945 patients in a developing country. EPILEPSIA 1985;26(2):131-136.
  10. JOSHI B, KATIYAR C, MOHAN PK et al. Profile of epilepsy in a developing country: a study of 1,000 patients based on the International Classification. EPILEPSIA 1977, 18 (4):549-554.
  11. SENANAYAKE N. Classification of epileptic seizures. EPILEPSIA 1993;34(5):812-818.
  12. GASTAUT H, GONCALVES et al. Relative frequence of different types of epilepsy: a study employing the Classification of the IEAE. EPILEPSIA 1975,16:457-467.
  13. SAGAR HJ, OXBURY JM. Hippocampal neuron loss in temporal lobe epilepsy: correlation with early childhood convulsions. ANN NEUROL 1987;22:334-340.
  14. DEL BRUTO OH, SANTIBANEZ R, NOBOA CA et al. Epilepsy due to neurocysticercosis: analysis of 203 patients. NEUROLOGY 1992;42:389-392.
  15. MICHEL P, MICHAULT A, GRUEL JC et al. La cysticercose a Madagascar. Diagnostic et epidemiologie. MED & ARMEES 1990;18(6):379-383.
  16. Commission (HI & IV) on epileptic drugs of the 1LAE. Availability and distribution of antiepileptic drugs in developing countries. EPILEPSIA 1985;26(2):117-121.
  17. FEKSI AT, KAAMUGISHA J, SANDER JWAS et al. Comprehensive primary health care on antiepileptic drug treatment programme in rural and semi-urban Kenya. THE LANCET 1991 ;vol 337(16):406-409.

© 2002-2012 African Journal of Neurological Sciences.
All rights reserved. Terms of use.
Tous droits réservés. Termes d'Utilisation.
ISSN: 1992-2647