Epidemiology of epilepsy in developing countries.

Reviews/Analyses Epidemiology of epilepsy in developing countries N. Senanayake1 & G.C. Roman2 Epilepsy is an important health problem in developing countries, where its prevalence can be up to 57 per 1000 population. This article reviews the epidemiology of epilepsy in developing countries in terms of its incidence, prevalence, seizure type, mortality data, and etiological factors. The prevalence of epilepsy is particularly high in Latin America and in several African countries, notably Liberia, Nigeria, and the United Republic of Tanzania. Parasitic infections, particularly neurocysticercosis, are important etiological factors for epilepsy in many of these countries. Other reasons for the high prevalence include intracranial infections of bacterial or viral origin, perinatal brain damage, head injuries, toxic agents, and hereditary factors. Many of these factors are, however, preventable or modifiable, and the introduction of appropriate measures to achieve this could lead to a substantial decrease in the incidence of epilepsy in developing countries. Introduction In 1978 a WHO Study Group identified epilepsy to be a disorder whose control should receive top pri- ority, in view of its high prevalence in developing countries and potentially severe consequences (1). At that time, most of the information available on epi- lepsy in developing countries had been derived from hospital-based studies, which made extrapolation to the general population highly conjectural. Since then, new data accumulated from different parts of the developing world give a better understanding of the magnitude of the problem, based on community stud- ies of the prevalence and the possible etiological fac- tors. There has also been a shift of emphasis from hospital-based epilepsy management to community- based epilepsy control (2). This article reviews the epidemiology of epilepsy in developing countries in the light of these new findings and changing attitudes. I Visiting Scientist, Neuroepidemiology Branch, National Institute of Neurological Disorders and Stroke (NINDS), National Insti- tutes of Health, Bethesda, MD, USA, and Professor of Medicine, University of Peradeniya, Peradeniya, Sri Lanka. Requests for reprints should be sent to Professor Senanayake at the latter address. 2 Chief, Neuroepidemiology Branch, NINDS, National Institutes of Health, Bethesda, Maryland, USA. Reprint No. 5380 Definition of epilepsy The WHO Neurosciences Research Protocol for studying the prevalence of neurological disorders in developing countries, which was developed in col- laboration with the Neuroepidemiology Branch of the U.S. National Institute for Neurological Dis- orders and Stroke (NINDS) (3), defines epilepsy as two or more afebrile seizures unrelated to acute metabolic disorders or to withdrawal of drugs or alcohol. Patients who have had a seizure within the last 2-5 years and those on anticonvulsant medica- tion are considered to have active epilepsy. Incidence and prevalence of epilepsy Incidence In industrialized countries, population studies defining epilepsy as recurrent unprovoked seizures have yielded age-adjusted incidences that vary from 28.9 to 53.1 per 100000 person-years (4). Higher incidences occur in children, age being an important factor in determining the incidence. Of the few studies carried out in developing countries, two from the Mariana Islands have reported annual incidences of 30-47.3 per 100 000 population (5, 6). Bulletin of the World Health Organization, 71 (2): 247-258 (1993) 247 N. Senanayake & G.C. Roman Prevalence The prevalence of epilepsy in industrialized coun- tries is about 3-9 per 1000 population. Most of the prevalence data available for developing countries (Fig. 1) are based on community surveys of rural populations. Such surveys, in general, have employed a two-phase design, the first phase consist- ing of screening interviews by field workers and the second phase comprising medical evaluation by neu- rologists. Common methodological problems of epi- demiological studies of epilepsy include deficiencies in the following: differential diagnosis, case ascer- tainment methods, definition and classification of epilepsy, and selection bias. The possible exclusion of minor seizures, such as absences and partial (tem- poral lobe) seizures, and underreporting, because of the stigma attached to the disease, were some of the difficulties encountered in one community survey conducted in Sri Lanka (7). In some communities, epileptic patients may be expelled from their homes as outcasts, and hence not be available for case ascertainment (8). Notwithstanding these short- comings, this form of survey is the optimum meth- od for detecting all active cases of epilepsy for prevalence studies. Fig. 1. Map showing the prevalences of epilepsy (per U = prevalence in urban areas. Several studies have reported low prevalences. For example, a house-to-house survey in rural Kash- mir, India, found a prevalence of 2.47 per 1000 for active epilepsy (9). This arose because mild cases were not identified; also the infant mortality rate was high (104.8 per 1000 live births per year), suggesting that many patients with epilepsy died early in child- hood. However, in a community with high infant mortality and morbidity, the prevalence of epilepsy would be expected also to be high. The prevalences for some other regions of India (Baroda, Calcutta, Patiala) are also quite low, and this merits further evaluation. Prevalences greater than 3-9 per 1000 have been found in several African (10-13) and Latin Ameri- can countries (14-18). The rates in some of these studies are not very different from those in industrial- ized countries (14). The prevalence of epilepsy in the Wroughbarh clan of the Grand Bassa County, Liberia, is among the highest in the world (49 per 1000 in one study (10) and 28 per 1000 in another (11)). Seizure disorders appear to have been intro- duced to this area approximately 30 years ago, as indicated by the period of onset of epilepsy and by retrospective information gathered from community 1000 population) in tropical and subtropical countries. WHO Bulletin OMS. Vol 71 1993248 Epidemiology of epilepsy In developing countries elders. An environmental factor, possibly interacting with genes that determine the susceptibility of the host, was postulated to be involved in the etiology (11). Aiyete-a rural community of about 2000 inhabitants, about 100-km south-west of Ibadan, Nigeria-has a high prevalence of active epilepsy (37 per 1000) (12). In contrast, in the Nigerian town of Igbo-Ora, which was surveyed using the same WHO protocol that was employed in Aiyete, the prevalence was only 5.3 per 1000 (19). The prin- cipal reasons for the lower prevalence in Igbo-Ora were held to be the availability of effective primary health care with emphasis on prevention of childhood infectious diseases, improved antenatal care, and a system of health education (19). In the United Republic of Tanzania, a Bantu population of about 10 000 Wapogoro tribesmen in the Mahenge region had 201 diagnosed cases of epi- lepsy (prevalence, 20 per 1000) (13, 20). The causes suggested included asphyxia among newborns caused by poor obstetrics management, syphilis, parasitic infections, meningoencephalitis, and chron- ic malnutrition. Owing to its marked geographical, cultural and social isolation, the Wapogoro society is traditionally endogamous, and marital union even between first cousins is encouraged. This trend is more pronounced among families with epileptic members, whose low social prestige and poor finan- cial status prevent them from acquiring brides from healthy families. Of the 201 patients, a family history of epilepsy was identified for 154 (76.6%). These findings strongly support a genetic basis for the high prevalence of epilepsy in this tribe. Studies from Latin American countries, in general, have reported high prevalences, e.g., 22 per 1000 in Panama City (17) and 57 per 1000 among the Guaymi Indians in Changuinola, Panama (18). The two main risk factors identified in the latter study were a family history of epilepsy and a history of febrile convulsions. Some of these high prevalences of epilepsy in developing countries probably reflect local or regional characteristics that do not apply to the countries as a whole. It should also be noted that several studies reporting high prevalences used small population samples, some less than 1000 individuals (12, 17). Sex-specific prevalence. Most studies of epilepsy in industrialized countries report that males are more frequently affected than females, although the difference is seldom statistically significant. Results from developing countries are similar, although, some studies in Nigeria (12, 19) and Latin America (15, 21) have found higher prevalences for females. Age-specific prevalence. The prevalence of epilepsy increases with age, reaching a peak in the third and fourth decades of life. In Ethiopia (22), Nigeria (12, 19), and Sri Lanka (7) the highest rates occur in the second decade of life; in Guam (23) in the third decade; and in Ecuador (16) in the fifth decade. Seizure type The International League against Epilepsy (ILAE) classifies seizures into two major groups: partial or generalized, according to the presence or absence of localized or focal onset (24). Partial seizures originate in a circumscribed area of the cerebral cortex and tend to be due to local causes such as trauma, embol- ic stroke, or tumour. Complex partial seizures, usually originating in the temporal lobe, may present with a complex variety of symptoms, including prominent psychic sensations such as fear, panic, and dream-like states. Rapid spreading of seizures with focal onset may result in generalized motor seizures. These are classified as partial seizures with secondary gener- alization. The term primary generalized seizures is reserved for convulsive or nonconvulsive seizures that are bilateral, symmetrical, and without focal onset, presumably arising from deep brain struc- tures. The commonest forms are tonic-clonic (grand mal) seizures, absence (petit mal) seizures, myo- clonic seizures, and atonic seizures. Most studies in industrialized countries have reported a higher proportion of primary generalized seizures. In contrast, in developing countries the trend seems to be a higher frequency of partial seizures, usually secondarily generalized (Table 1). It should be borne in mind, however, that the investi- gations listed in Table 1 are a mixture of community- based (9, 19, 25, 30) and institutional-based studies (26, 27, 32).a The selection bias in the institutional studies may have influenced the findings; also, some of the studies were based on data from as few as 19 patients. Furthermore, there is a lack of uniformity in the classification of seizure types, particularly in the earlier studies. These methodological differences may explain some of the variations in the frequen- cies of seizure types between the studies. The general tendency for a higher frequency of partial seizures in developing countries could be an indication of the high incidence of symptomatic epilepsy in the tropics, caused by cortical damage resulting from secondary factors. Of the generalized seizures, tonic-clonic are by far the commonest. a Sakamoto, A.C. [Clinical study of and prognosis for epileptic fits that began in infancy in the Brazilian population]. Doctoral thesis, University of Sao Paulo, 1985 (in Portuguese). WHO Bulletin OMS. Vol 71 1993 249 N. Senanayake & G.C. Roman CD. U)U)N.N.~ Io. . O N oC00) 'CON N NsIN.0 N. 0 N- urI O a) c o wcm_ . .C.O.C. _ 't uN cv c .. V . 0) C"Dco,o0 U 0 tcoN o cxP, oIn It Vco__ . LO o It cv If- o- cro CD aw 00LO LO CD CD_ 0 N 0 0 0cmN - .0) 0) CDo U) 0 o-t N. U) CD N' C - C) CD CDM 0 'C, -0N o 0 C C)0 i iiIC\o ai C) C0 _ 0)-CD N O D )Ut CD N.CO _ CD N CD N LO ND U w D CD N. CD 0 N. CV) 0 CD U) 0) 0r CD CCD0 -'- ONC o ao r, 0 IU) CQoqN U U) U) 0 N. _-0CO CM CO CDco t U) N NCm) Io'-° I II0 0 co cm o 1I0cM I Ic6CM _ cm o 0co CD -CO N.CD C -mI IO U) CD U) 0) U) N. CDa''-NM CV)'- C CV) c co .i0EE uco cloCOco D0 co co co0 co o co o2 -6 0N its 0a) C 0 U)Cu C COD0o0EX UIT, . O E U)ao0, 0)1L1D _ <(5 Q 1a WHO Bulletin OMS. Vol 71 1993 -o CO) CO M a cb - O CO .a) ._N t-a. -a 0 F- CC (L x 'C E 0 0 0) E cn -a 0 U,0 N.0 c a) a1) 0 -c 0 C)0 U,00 cn 0 U)Q .0 c C .c 0I- 0 .2 0 0 .2 00 6 0. .06 0 N 0 o0 2 6 C 0 .> 0 0) 3 cu 0 U,0) U, C) 0 z 250 Epidemiology of epilepsy in developing countries Absences are underrepresented in developing coun- tries, the frequency of these seizures in indus- trialized countries being approximately 6% (33); such seizures, because of the mild nature of their symptomatology, are either not recognized as a form of epilepsy, or there is a genuine, low incidence of this disorder in the tropics. An electroencephalo- graphic study in Nigeria has corroborated the latter explanation (34). Mortality data Mortality data, based on a single underlying cause of death listed on the death certificate, considerably underestimate the number of deaths among people with epilepsy. Furthermore, the case-fatality ratio for epilepsy is low, so that mortality statistics are not a good indicator of the disease frequency. Also, the accuracy of death certificate diagnosis varies, parti- cularly from one country to another and over long periods of time. Males have a higher age-adjusted mortality rate than females, and developing countries have higher mortality rates than industrialized coun- tries (35). In a recent community-based study in rural Ethiopia, 20 of the 316 persons with epilepsy (6.3%) died over a 2-year period (22); death was related to epilepsy in nine instances, eight individuals died in status epilepticus, and one person died from severe bums sustained by falling into a domestic fire during a seizure. Etiological factors Studies in industrialized countries have recorded a specific etiology for epilepsy in about 60-70% of cases. In developing countries, despite the abundance of potential causes of epilepsy, a known etiology has been reported for less than 40% of cases (Table 2). Clinicians in most developing countries may have ready access to only the most basic investigations, hence the etiological diagnosis is often made on clinical grounds. If such diagnoses were supplemen- ted with those made using the technology available in industrialized countries, the percentage of epilep- sy cases with an identifiable etiology would in- crease considerably. Parasitic infections Cysticercosis. Cysticercosis is caused by the larva of the pork tapeworm, Taenia solium, and is the most frequent parasitosis of the central nervous system in several countries in Africa, Latin America, and Asia. Seizures are the commonest manifestation of neuro- cysticercosis (38, 39). In Mexico it was the main Table 2: Distribution of etiological factors for epilepsy in developing countries Etiological factor (%):a No. of cases PBD HI ICI CC ICT CVD TM MIS FS FH NK Africa Ghana (36) Nigeria (19) Nigeria (27) United Republic of Tanzania (28) Asia India (29) India (29) India (31) Sri Lankab Latin America Brazil c Brazil (32) Uruguay (37) 204 101 945 428 3378 1804 1000 1287 5.4 2.0 3.8 4.4 3.1 1.9 2.0 1.2 455 d 14.0 147 e 11.0 500 d 8.5 5.4 5.9 8.4 5.1 4.8 1.7 1.4 1.7 2.5 1.6 6.3 0.9 0.9 0.5 - 0.5 - 1.0 1.4 - 2.8 [3.5]' 0.4 3.0 5.0 13.0 13.0 - 4.0 8.0 4.0 - a PBD = perinatal brain damage; Hi = head injuries; ICI = intracranial infections; CC = cerebral cysticercosis; ICT = intracranial tumour; CVD = cerebrovascular disease; TM = toxic/metabolic including alcohol; MIS = miscellaneous; FS = febrile seizures; FH = family history of epilepsy; and NK = not known. b Unpublished data. c See footnote a, p. 249. d Age: 1 month to 15 years. e Age >13 years. f Figures in square brackets denote the sum of the two adjacent etiological factors. WHO Bulletin OMS. Vol 71 1993 5.4 82.4 60.4 66.9 65.9 0.5 0.8 0.1 0.2 0.3 1.6 2.9 1.0 0.3 3.4 2.0 1.1 2.6 0.5 1.1 1.3 0.2 2.5 5.0 0.1 1.9 1.2 0.2 23.8 16.3 11.2 1.2 25.0 - 85.1 - 91.6 78.2 10.1 - 4.0 1.5 6.5 - 63.0 [68.0]' 9.0 59 251 N. Senanayake & G.C. RomAn cause of late-onset epilepsy, accounting for half the cases (40, 41); in Ecuador, nearly a quarter of all new cases of epilepsy were due to this cause (42); and in Brazil, it was the single most frequently identifiable cause of epilepsy.b In Africa, neurocysticercosis caused epilepsy among Bantus and Southern Rhodesians (43, 44); in northern Togo, the cause of epilepsy in a third of patients aged over 15 years and of two-thirds of those whose epilepsy began after the age of 50 years was attributed to cysticercosis (45). In India, 2% of cases of focal epilepsy (46) reported in one hospital in New Delhi were due to cysticerc- osis; it also accounted for 5.1% of 253 patients aged .25 years who were evaluated for recent-onset epilepsy (47). Cerebral cysticercosis has also been suggested to be the cause of many cases of epilepsy in South India (48). In south-east Asia, cysticercosis is the commonest cause of late-onset epilepsy (49). In Bali, for example, 8% of the population in one village had epilepsy possibly related to neurocysticercosis (50); and this has been identified as a frequent cause of epileptic seizures among the Kapadoku people of Indonesia (51) and in New Guinea (52); recently, among Nepalese (Gurkhas) in Hong Kong who presented with adult-onset epilepsy, seven out of eight cases had cerebral cysticercosis (53). Since no control studies were carried out, some of these findings may reflect local or regional character- istics within countries. However, with the increased availability of computed tomograph (CT) scans and magnetic resonance imaging (MRI) facilities, neurocysticercosis is being recognized more and more frequently in the tropics. Schistosomiasis. Schistosomiasis, caused by Schistosoma japonicum, which is endemic in south- east Asia, causes the majority of cases of cerebral schistosomiasis. Acute schistosomiasis (Katayama fever) may occasionally produce an encephalopathy that manifests as impairment of consciousness, papilloedema, partial and generalized seizures, and focal neurological signs. Chronic forms of cerebral schistosomiasis caused by embolized schistosoma eggs are commonly manifested by epileptic attacks, usually in the form of partial seizures (54). Paragonimiasis. This condition, which is caused by Paragonimus westermani, is prevalent in the Far East, south-east Asia, some parts of Africa, and South America. The lung is the primary site of infec- tion but of the other sites, the brain is involved with the highest frequency. Cerebral paragonimiasis is one of the major neurological conditions in endemic b See footnote a, p. 249. areas (55). For example, in Korea it accounts for nearly a quarter of "brain tumours" (56). Seizures, usually focal motor, are the commonest manifesta- tion (57). Toxoplasmosis. Toxoplasmosis has recently assum- ed special status because it is a frequent opportun- istic infection in patients with acquired immuno- deficiency syndrome (AIDS) (58). Epilepsy is a well-recognized consequence of toxoplasmosis and occurs in about 25% of affected individuals (59). With the increasing incidence of AIDS in the tropics, toxoplasmosis may become a more important cause of epilepsy in developing countries. Approximately 40-60% of patients with congenital toxoplasmosis also have seizures. African trypanosomiasis. This condition (sleeping sickness), caused by Trypanosoma brucei, is widely distributed throughout sub-Saharan Africa. The chronic stage of the disease is characterized by pro- gressive neurological involvement, including gener- alized and partial seizures. Congenital cases may also occur and manifest as mental retardation and epilepsy. American trypanosomiasis. This form of trypanoso- miasis (Chagas disease) is caused by T. cruzi and is a public health problem in rural areas of Central and South America. The heart, oesophagus, and colon are the sites primarily affected. Cerebral involvement secondary to embolization of cardiac blood clots is a cause of late-onset epilepsy with a high frequency of partial seizures having elementary symptomatology (60). Malaria. Malaria is endemic in tropical Africa, America, and Asia, as well as subtropical areas of the Eastern Mediterranean. Epilepsy has long been recognized as one of its late sequelae. Malaria is also a common cause of febrile seizures in the tropics. For example, in Congo, 9.6% of all children admitted to Brazzaville General Hospital in the period 1981-83 presented with seizures (61). Status epilepticus occurred in 13.6% of the cases and 67% of these were related to benign malaria; febrile seizures occurred in 73.5% of all cases, and 81% of them were related to malaria. Approximately 60% of all seizure disorders in children aged 1 month to 6 years in one large general hospital in Congo were related to benign or malignant forms of malaria, and seizures were the reason for admission of 10% of all children in this age range (59). Other parasitic diseases. Diseases such as spargano- sis, hydatid disease, toxocariasis, and ascariasis may also contribute to the etiology of epilepsy in the tropics. 252 WHO Bulletin OMS. Vol 71 1993 Epidemiology of epilepsy in developing countries Bacterial infections Tuberculous meningitis. This form of meningitis can cause epilepsy as a late sequela in 8-14% of patients and is frequently associated with hemiplegia (62, 63). Epilepsy is also a common manifestation of intracranial tuberculomas that present as slow- growing, space-occupying lesions (64). CT scans often identify small multiple tuberculomas in patients with tuberculous meningitis; these can develop and enlarge after tuberculous meningitis has been successfully treated (65). Tuberculomas can also be discovered accidentally in asymptomatic individuals and in children with focal or generalized seizures (66, 67). In India, ring or disc-like enhancing lesions, which are thought to be of tuberculous origin, have been identified in CT scans of patients who presented with simple partial epilepsy (68). Pyogenic meningitis. This is common in the tropics, and epidemics of meningococcal meningitis occur from time to time in sub-Saharan Africa and Brazil. In the USA, among 734 survivors of intracranial infections the overall risk for epilepsy was increased sevenfold; for individuals with a diagnosis of brain abscess the risk was more than 40 times greater (69). In a study of 185 children who were followed up after bacterial meningitis, 13 developed one or more afebrile seizures after the initial hospitalization (70). These findings confirm that intracranial infections are important risk factors for epilepsy. This risk remains elevated for at least 20 years after the infec- tion, although in most instances the onset is within 5 years of the onset of the acute illness (69). Viral infections Japanese encephalitis. Japanese encephalitis is perhaps the most important and best documented form of epidemic viral encephalitis in developing countries. At present, its incidence appears to be increasing and it is spreading across parts of Bangladesh, India, Myanmar, Nepal, Thailand, and Vietnam (71). Neurological sequelae, including intellectual and emotional changes and motor impair- ment, occur in up to 80% of survivors, especially children (72). In an outbreak that affected 128 patients in Bangalore, 1.3% of the 78 survivors developed epilepsy (73). In Chiang Mai, Thailand, of 59 children affected by Japanese encephalitis, 57% of the survivors had neurological sequelae-including 20% with epilepsy (74). Perinatal factors Data from case-control studies prove convincingly that perinatal pathology is a frequent cause of epilep- sy. In developed countries, bearing in mind that acquired epilepsy constitutes about 40% of all cases, perinatal pathology should be the cause of about 13-14% of all epilepsies (75). The proportions from developing countries (Table 2), which vary from 1% to 14% seem to be an underestimate. Hypoxaemic-ischaemic encephalopathy is the most frequent type of perinatal pathology that predis- poses to epilepsy. The causes of perinatal hypoxia include severe maternal cardiovascular diseases, placental and umbilical cord diseases, prolonged labour, and airway obstruction at birth (75). In dystocic deliveries, the excessive reduction of fetal cranial diameters may lead to hemiation of hippo- campal regions through the incisura tentorium, causing ischaemia, atrophy, and subsequent gliosis of these structures ("mesial sclerosis"), which predis- poses to epilepsy (75). Prematurity is also an etiologi- cal factor for epilepsy because of its strong associa- tion with neonatal cerebral lesions. In many developing countries, most deliveries in rural areas are conducted by traditional birth atten- dants (76). Families are large, in certain areas the rate of twinning is high (77), and the incidence of pre-term deliveries is at least twice that in developed countries (78). Also, many of the mothers are multi- parous, and it is not unusual to find women having babies well into their late forties (76). The incidence of brain damage in babies is five times greater if the mother is above 30 years of age (78). The mothers may also be malnourished, anaemic, and exposed to a variety of infections that could affect the baby in utero or at delivery. All these factors can contribute to perinatal brain damage. Neonatal bilirubinaemic encephalopathy is a further disorder that can lead to epilepsy (75). Although relatively uncommon in developed coun- tries, it is a particular problem in many developing countries; for example, in Nigeria it is the com- monest cause of cerebral palsy, accounting for 50% of the cases (77); and in Singapore, it is the com- monest cause of death in babies under the age of 1 week (79). Causative diseases such as glucose-6- phosphate dehydrogenase deficiency (43% of cases in the Singapore study) (79) and umbilical cord sepsis probably explain its high incidence in the tropics. Febrile seizures Febrile seizures are a common acute neurological disturbance of childhood, affecting about 2-5% of children (4). In view of the high frequency of febrile illnesses caused by diseases such as malaria, a higher prevalence of febrile convulsions would be expected in the tropics. The reported frequency of such convulsions varies from 64 per 100 000 in a semi- WHO Bulletin OMS. Vol 71 1993 253 N. Senanayake & G.C. RomAn urban and rural area of South India (80), 2.4 per 1000 in the Parsi community of Bombay (81), and 5.4 per 1000 in Ecuador (16), to 14% in the Mariana Islands (5). Febrile seizures are associated with a six- fold increased risk of subsequent epilepsy. The two conditions are probably not causally related, but rather are an independent outcome of a common underlying factor-either genetically determined or antecedent-localized abnormalities (82). It is worth- while to investigate whether the seizures that accom- pany fevers such as malaria in the tropics are "febrile seizures" in the sense that the term is used in developed countries or whether they are the result of direct cerebral insult from infection. Also, the toxici- ty of some of the home remedies, such as cow's urine and herbal medicine used to treat convulsions in the tropics, may contribute to brain damage (76, 77). Head injuries Head injuries have been reported to be an important cause of epilepsy in adults in certain communities, particularly in Africa (83, 84). Such injuries are attributed to fighting and jousting with heavy sticks and knobkerries, a pastime that is both a national sport and a means of settling grievances (85). Cranial trauma related to personal attacks or to road traffic accidents is a very common cause of epilepsy in Brazil and Uruguay (37, 86). Also, in some tropical countries people climb trees to gather coconuts or tap toddy, and accidental falls may result in head injuries. Traffic accidents associated with poor regu- lation of motor vehicle transit, lack of seat belts, and lack of helmets for motor-cycle riders are a growing problem in developing countries. Toxic agents The neurotoxic potential of some home remedies used in the treatment of convulsions has been men- tioned above. Toxicity, mainly from lead, often causes seizures in children. Refining of jeweller's wastes to recover gold and the reconditioning and smelting of lead-containing batteries to obtain scrap lead are the commonest sources of childhood plumbism in Sri Lanka (87). Burning lead-containing batteries for cooking purposes or to provide heat have also caused lead poisoning in children in poor communities (88). A condition referred to as "epidemic epilepsy" in the Lakhimpur Kheri district of India has been attributed to the toxicity of the pesticide benzene hexachloride, which is used as a food grain preserva- tive (89, 90). Chloroquine, which is widely used in the tropics to treat malaria and hepatic amoebiasis, has been reported to cause seizures. This could be an idiosyn- cratic reaction; however, a causal relationship has not been clearly established (91). Alcoholism is another important cause of seizures and also exacerbates primary epilepsy and triggers seizures in epileptic patients (92). Consump- tion of alcohol, illicit liquor in particular, is a grow- ing problem in many developing countries, and its contribution as an etiological agent of epilepsy in the tropics merits further evaluation. Genetic factors The risk of epilepsy is increased threefold for indivi- duals with a first-degree relative who has the condi- tion (93). Also, a number of diseases that follow Mendelian patterns of inheritance may have seizures as one of their manifestations. In many developing countries, in-breeding is relatively common, and cus- tomary; and in some communities, the stigma of epi- lepsy prevents those affected from finding marriage partners from healthy families (13). Such practices are likely to increase the risk of seizure disorders in any offspring. In isolated communities, specific inherited diseases may contribute to the etiology of epilepsy, as in the Grand Bassa County of Liberia (11) and among the Wapogoro tribe in the United Republic of Tanzania (20). Conclusions Although over the past decade there has been signifi- cant progress in defining the magnitude of the prob- lem, and identifying the likely etiological factors, our understanding of the epidemiology of epilepsy in developing countries is far from complete. The prob- lem is compounded, particularly in areas where physicians and neurologists do not have ready access to large segments of the population, and where diagnostic methods such as CT scans and electro- encephalography (EEG) may not be available. Other difficulties include the following: the diagnosis of epilepsy is not always straightforward and may require EEG confirmation; the classification of epilepsy and epileptic syndromes has become com- plicated and poorly suited for field studies; the definitions used vary in different studies; and the inclusion criteria are not always clear. The potential number of etiological factors for epilepsy in developing countries is considerable, and many are preventable. Intracranial infections are of particular importance in this respect. Where simple preventive measures fail, the solution may lie in vac- cination, e.g., the control of epidemic meningococcal meningitis. Epilepsy and other long-term sequelae of intracranial infections can be minimized by early WHO Bulletin OMS. Vol 71 1993254 Epidemiology of epilepsy in developing countries detection and prompt and adequate treatment. Enforcement of strict traffic regulations, such as speed limits and the wearing of helmets by motor- cycle riders, and also increasing public awareness, are urgent needs in many developing countries that would slow the rapid increase in the number of head injuries caused by road accidents and the resultant post-traumatic epilepsy. Also, some form of genetic counselling should be made available at least in areas where a specific hereditary predisposition to epilepsy has been established. Acknowledgement N.S. received a WHO/Fogarty International Neurosciences Fellowship. Resume Epidemiologie de l'6pilepsie dans les pays en d6veloppement L'6pilepsie constitue un probleme de sant6 impor- tant dans les pays en d6veloppement, ou elle est a l'origine d'une morbidite 6lev6e et de cons6- quences potentiellement graves. Sa pr6valence est extremement variable, allant de 1,3 a 57 pour 1000 habitants. Cette prevalence est g6n6rale- ment faible dans les r6gions du nord de l'Inde et 6Iev6e dans plusieurs pays africains, notamment au Lib6ria (28/1000, 49/1000), au Nig6ria (37/1000) et en R6publique-Unie de Tanzanie (20/1000), ainsi qu'en Am6rique latine - Colombie (17/1000 et 19,5/1000), Equateur (17/1000), Panama (22/1000 et 57/1000) et Venezuela (4,1 a 22,6/1000). La plupart des 6tudes rapportent un pourcentage plus 6leve de crises d6pilepsie partielle, probable- ment en raison de la forte incidence de l'6pilepsie symptomatique provoqu6e par des facteurs secon- daires. A cet 6gard, les infestations parasitaires et en particulier la cysticercose, en sont l'une des principales causes. Au Mexique, la cysticercose est a l'origjne de la moiti6 des cas d'epilepsie d'installation tardive et en Equateur, d'un quart de tous les nouveaux cas. Au Br6sil, c'est la cause d'6pilepsie la plus facile a identifier. D'autres infec- tions, telles que la m6ningite tuberculeuse et les abces cerebraux sont 6galement des facteurs d6clenchants importants. Dans de nombreux pays en d6veloppement, une forte proportion des accouchements a lieu a domicile, sans I'aide d'une sage-femme qualifi6e. Dans certaines r6gions, les proportions de gros- sesses multiples et d'accouchements avant terme sont 6lev6es. De nombreuses meres sont 6gale- ment multipares, malnutries, an6mi6es et expo- sees a toutes sortes d'infections pouvant toucher leurs enfants in utero. L'ictere nucl6aire provoqu6 par un d6ficit en glucose-6-phosphate-d6shydro- g6nase pose des problemes particuliers dans cer- taines r6gions. Tous ces facteurs peuvent entral- ner des I6sions c6r6brales dans la p6riode p6rinatale et donc une 6pilepsie. En raison de la fr6quence 6lev6e des maladies f6briles, l'inciden- ce des convulsions febriles est elev6e sous les tropiques; ces dernieres peuvent etre associ6es par la suite a un risque six fois plus 6lev6 d'6pi- lepsie. Les traumatismes craniens dus a des agres- sions, a des accidents de la circulation, ou a des chutes sont une autre cause importante d'6pilep- sie dans les pays en d6veloppement. Le potentiel neurotoxique de certains remedes prepares a domicile et les poisons pr6sents dans 1'environ- nement tels que le plomb et les insecticides doivent 6galement etre pris en compte. Dans de nombreux pays, la consanguinit6 relativement fr6quente et la stigmatisation associ6e a l'6pilep- sie peuvent empecher ceux qui en souffrent de trouver un conjoint dans des familles en bonne sante; il peut s'ensuivre un risque augment6 de troubles comitiaux dans leur descendance. Les facteurs 6tiologiques potentiels de l'6pi- lepsie dans les pays en d6veloppement sont nom- breux, meme si beaucoup d'entre eux sont 6vi- tables. Une d6tection pr6coce et un traitement ad6quat peuvent r6duire le risque d'6pilepsie et les autres s6quelles a long terme des infections intracraniennes. L'application d'une r6glementa- tion routiere stricte, par exemple, limitation de vitesse, port du casque obligatoire pour les moto- cyclistes, et le fait de faire prendre conscience au public de ces problemes, ralentirait I'augmentation rapide du nombre de l6sions c6r6brales dues a des accidents de la route et de l'6pilepsie post- traumatique. Certaines formes de conseil g6n6- tique devraient etre mises a la disposition du public, du moins dans les r6gions ou il existe une pr6disposition h6r6ditaire particuliere a l'6pilepsie. References 1. WHO Technical Report Series No. 629, 1978 (Appli- cation of advances in neurosciences for the control of neurological disorders), pp. 9-20. 2. Senanayake, N. & Meinardi, H. Improvement of the care of people with epilepsy in rural areas of a developing country. In: Manelis, J. et al., ed. Advances in epileptology, vol. 17. New York, Raven Press, 1989, pp. 441-444. 3. Li, S. et al. Epidemiology of epilepsy in urban areas of the People's Republic of China. Epilepsia, 26: WHO Bulletin OMS. Vol 71 1993 255 N. Senanayake & G.C. Roman 391-394 (1985). 4. Hauser, W.A. & Hesdorffer, D.C. Epidemiology of epilepsy. In: Anderson, D.W., ed. Neuroepidemi- ology: a tribute to Bruce Schoenberg. Boca Raton, FL, CRC Press, 1991, pp. 97-119. 5. Mathai, K.V. et al. Convulsive disorders in the Mariana islands. Epilepsia, 9: 77-85 (1968). 6. Chen, K. et al. Patterns of neurologic diseases in Guam, I. Epidemiologic aspects. Archives of neu- rology, 19: 573-578 (1968). 7. Senanayake, N. Epilepsy control in a developing country-the challenge of tomorrow. Ceylon medical journal, 32: 181-199 (1987). 8. Giel, R. The problem of epilepsy in Ethiopia. Tropi- cal and geographical medicine, 22: 439-442 (1970). 9. Koul, R. et al. Prevalence and pattern of epilepsy (Lath/Mirgi/Laran) in rural Kashmir, India. Epilepsia, 29: 116-122 (1988). 10. Gerrits, C.A. West African epilepsy focus. Lancet, 1: 358 (1983). 11. Goudsmit, J. et al. Epilepsy in the Gbawein and Wroughbarh clan of Grand Bassa County, Liberia: the endemic occurrence of "See-ee" in the native population. Neuroepidemiology, 2: 24-34 (1983). 12. Osuntokun, B.O. et al. Research protocol for measuring the prevalence of neurologic disorders in developing countries-results of a pilot study in Nigeria. Neuroepidemiology, 1: 143-153 (1982). 13. Aall-Jilek, L.M. Epilepsy in the Wapogoro tribe in Tanganyika. Acta psychiatrica et neurologica Scan- dinavica, 41: 57-86 (1965). 14. Marino, Jr. R. et al. Epidemiological aspects of epi- lepsy in Sao Paulo, Brazil: a prevalence rate study. In: Wolf, P. et al., ed. Advances in epileptology, vol. 16. New York, Raven Press, 1987, pp. 759-764. 15. Gomez, J.G. et al. Prevalence of epilepsy in Bogota, Colombia. Neurology (Minneap.), 28: 90-94 (1978). 16. Cruz, M.E. et al. Pilot study to detect neurologic disease in Ecuador among a population with a high prevalence of endemic goiter. Neuroepidemiology, 4: 108-116 (1985). 17. Gracia, F.J. et al. [Prevalence of neurological ill- nesses in the Corregimiento Belisario Porras district of San Miguelito, Panama, 1986]. Revista Medicina de Panama, 13: 40-45 (1988) (in Spanish). 18. Gracia, F. et al. Epidemiology of epilepsy in Guaymi Indians from Bocas del Toro Province, Republic of Panama. Epilepsia, 31: 718-723 (1990). 19. Osuntokun, B.O. et al. Prevalence of the epilep- sies in Nigerian Africans: a community-based study. Epilepsia, 28: 272-279 (1987). 20. Jilek-Aall, L. et al. Clinical and genetic aspects of seizure disorders prevalent in an isolated African population. Epilepsia, 20: 613-622 (1979). 21. Cruz, M.E. et al. Epidemiology of epilepsy. In: Poeck, K. et al., ed. Neurology. Berlin, Springer Verlag, 1986, pp. 229-239. 22. Tekle-Haimanot, R. et al. Clinical and electroen- cephalographic characteristics of epilepsy in rural Ethiopia: a community-based study. Epilepsy research, 7: 230-239 (1990). 23. Stanhope, J.M. et al. Convulsions among the Cha- morro people of Guam, Mariana islands, I. Seizure disorders. American journal of epidemiology, 95: 292-298 (1972). 24. Dreifuss, F.E. The epilepsies: clinical implications of the international classification. Epilepsia, 31(suppl.): S3-S10 (1990). 25. van der Waals, F.W. et al. See-ee: clinical charac- teristics of highly prevalent seizure disorders in the Gbawein and Wroughbarh Clan region of Grand Bassa County, Liberia. Neuroepidemiology, 2: 35-44 (1983). 26. Ogunniyi, A. et al. Risk factors for epilepsy: case-control study in Nigerians. Epilepsia, 28: 280-285 (1987). 27. Danesi, M.A. Classification of the epilepsies: an investigation of 945 patients in a developing coun- try. Epilepsia, 26: 131-136 (1985). 28. Matuja, W.B.P. Aetiological factors in Tanzanian epileptics. East African medical journal, 66: 343-348 (1989). 29. Mani, K.S. & Rangan, G. Epilepsy in the third world-Asian aspects. In: Dam, M. & Gram, L., ed. Comprehensive epileptology. New York, Raven Press, 1991, pp. 781-793. 30. Bharucha, N.E. et al. Prevalence of epilepsy in the Parsi community of Bombay. Epilepsia, 29: 111-115 (1988). 31. Joshi, V. et al. Profile of epilepsy in a developing country: a study of 1000 patients based on the inter- national classification. Epilepsia, 18: 549-554 (1977). 32. Bittencourt, P.R.M. Epilepsy in Latin America. In: Laidlaw, J. et al., ed. A textbook of epilepsy, 3rd ed. Edinburgh, Churchill Livingstone, 1988, pp. 518-528. 33. Hauser, W.A. & Kurland, L.T. The epidemiology of epilepsy in Rochester, Minnesota, 1935 through 1967. Epilepsia, 16: 1-66 (1975). 34. Osuntokun, B.O. et al. Electroencephalographic correlates of epilepsy in Nigerian children. Develop- mental medicine and child neurology, 16: 659-663 (1974). 35. Massey, E.W. & Schoenberg, B.S. Mortality from epilepsy: international patterns and changes over time. Neuroepidemiology, 4: 65-70 (1985). 36. Edoo, B.B. & Haddock, D.R.W. Epilepsy in Accra, Ghana: a report on classification and etiology. Afri- can journal of medical sciences, 1: 207-212 (1970). 37. de Pasquet, E.G. et al. [Epidemiological study of 500 epileptic adults drawn from a hospital popula- tion]. Acta neurologica Latinoamericana, 22: 50-65 (1976) (in Spanish). 38. Bruck, I. et al. [Neurocysticercosis in infancy]. Arquivos de neuropsiquiatria, 49: 43-46 (1991) (in Portuguese). 39. Wei, G.Z. et al. Cysticercosis of the central nervous system: a clinical study of 1400 cases. Chinese medical journal, 101: 493-500 (1988). 40. Medina, M.T. et al. Neurocysticercosis as the main cause of late-onset epilepsy in Mexico. Archives of internal medicine, 150: 325-327 (1990). 41. Sotelo, J. et al. Neurocysticercosis: a new classifi- cation based on active and inactive forms. Archives of internal medicine, 145: 442-445 (1985). 42. Shorvon, S.D. & Farmer, P.J. Epilepsy in develop- 256 WHO Bulletin OMS. Vol 71 1993 Epidemiology of epilepsy in developing countries ing countries: a review of epidemiological, sociocul- tural, and treatment aspects. Epilepsia, 29(suppl.): S36-S54 (1988). 43. Levy, L.F. et al. Epilepsy in Africans. Central Afri- can journal of medicine, 10: 241 (1964). 44. Powell, S.J. et al. Cysticercosis and epilepsy in Africans. Annals of tropical medicine and parasitol- ogy, 60: 152-615 (1966). 45. Dumas, M. et al. Epidemiological study of neuro- cysticercosis in northern Togo (West Africa). Acta Leidensia, 57: 191-196 (1989). 46. Wani, M.A. et al. Neurocysticercosis-some uncom- mon presentations. Neurology (India), 29: 58-63 (1981). 47. Ahuja, G.K. & Mohanta, A. Late-onset epilepsy: a prospective study. Acta neurologica Scandinavica, 66: 216-226 (1982). 48. Veliath, A.J. et al. Cysticercosis in South India. Journal of tropical medicine and hygiene, 88: 25-29 (1985). 49. Vejajiva, A. Neurological disorders in South-east Asia. In: Weatherall, D.J. et al., ed. Oxford textbook of medicine. Oxford, Oxford University Press, 1987, pp. 21.269-21.272. 50. Simanjuntak, G.M. et al. An investigation of taenia- sis and cysticercosis in Bali. South-east Asian jour- nal of tropical medicine and public health, 8: 494-497 (1977). 51. Subianto, D.B. et al. Burns and epileptic fits asso- ciated with cysticercosis in mountain people of Irian Jaya. Tropical and geographical medicine, 30: 275-278 (1978). 52. Gadjusek, D.C. Introduction of Taenia solium into West New Guinea with a note on an epidemic of burns from cysticercus epilepsy in the Ekari people of the Wissel Lake area. Papua New Guinea medi- cal journal, 21: 329-342 (1978). 53. Heap, B.J. Cerebral cysticercosis as a common cause of epilepsy in Gurkhas in Hong Kong. Journal of the Royal Army Medical Corps, 136: 146-149 (1 990). 54. Chang, Y.C. et al. Cerebral schistosomiasis. Chinese medical journal, 75: 892-907 (1957). 55. Oh, S.J. The rate of cerebral involvement in cere- bral paragonimiasis-an epidemiological study. Japanese journal of parasitology, 18: 211-215 (1969). 56. Cho, W.J. & Kim, H.J. Ocular symptoms of intra- cranial tumour (clinical survey of 93 Korean cases). Journal of the Korean Ophthalmological Society, 5: 35-41 (1964). 57. Oh, S.J. Cerebral paragonimiasis. Journal of the neurological sciences, 8: 27-48 (1968). 58. Navia, B.A. et al. Cerebral toxoplasmosis complicat- ing the acquired immune deficiency syndrome: clini- cal and neurological findings in 27 patients. Annals of neurology, 19: 224-238 (1986). 59. Bittencourt, P.R.M. et al. Epilepsy and parasitosis of the central nervous system. In: Pedley, T.A. & Meldrum, B.S., ed. Recent advances in epilepsy, vol. 4. Edinburgh, Churchill Livingstone, 1988, pp. 123-159. 60. Jardim, E. & Takayanagui, O.M. [Epilepsy and chronic Chagas disease]. Arquivos de neuropsiquia- tria, 39: 32-41 (1981) (in Portuguese). 61. Senga, P. et al. Profile of seizures in infants and young children in Brazzaville (Congo). Annales de p6diatrie, 32: 477-480 (1985). 62. Lorber, J. Long-term follow-up of 100 children who recovered from tuberculous meningitis. Pediatrics, 28: 778-782 (1961). 63. Donner, M. & Wasz-Hockert, 0. Late neurological sequelae of tuberculous meningitis. Acta paediatrica, 51(suppl.): 34-38 (1962). 64. Bahemuka, M. & Murungi, H. Tuberculosis of the nervous system: a clinical, radiological and patho- logical study of 39 consecutive cases in Riyadh, Saudi Arabia. Journal of the neurological sciences, 90: 67-76 (1989). 65. Teoh, R. et al. Symptomatic intracranial tuberculo- ma developing during treatment of tuberculosis: a report of 10 patients and review of the literature. Quarterly journal of medicine, 63(241): 449-460 (1987). 66. Bharucha, N.E. & Bharucha, E.P. Neurology in India. In: Bradley, W.G. et al., ed. Neurology in clini- cal practice. Boston, Butterworth-Heinemann, 1991, pp. 1925-1941. 67. Vyravanathan, S. & Senanayake, N. Tuberculosis presenting with hemiplegia. Journal of tropical medi- cine and hygiene, 82: 38-40 (1979). 68. Wadia, R.S. et al. Focal epilepsy in India with special reference to lesions showing ring or disc- like enhancement on contrast computed tomogra- phy. Journal of neurology, neurosurgery and psychia- try, 50: 1298-1301 (1987). 69. Annegers, J.F. et al. The risk of unprovoked seizures after encephalitis and meningitis. Neurol- ogy, 38: 1407-1410 (1988). 70. Pomeroy, S.L. et al. Seizures and other neurologic sequelae of bacterial meningitis in children. New England journal of medicine, 323: 1651-1657 (1 990). 71. Umenal, T. et al. Japanese encephalitis: current worldwide status. Bulletin of the World Health Organization, 63: 625-631 (1985). 72. Monath, T.P. Japanese encephalitis-a plague of the Orient. New England journal of medicine, 319: 641-643 (1988). 73. Gourie-Devi, M. & Deshpande, D.H. Japanese encephaltis. In: Prasad, L.S. & Kulczycki, L.L., ed. Paediatric problems. New Delhi, S. Chand, 1982, pp. 340-356. 74. Poneprasert, B. Japanese encephalitis in children in northern Thailand. South-east Asian journal of tropical medicine and public health, 20: 599-603 (1989). 75. Bergamasco, B. et al. Perinatal pathology and epi- lepsy. Progress in clinical and biological research, 124: 185-198 (1983). 76. Asirifi, Y. Aetiology of cerebral palsy in developing countries. Developmental medicine and child neu- rology, 14: 230-232 (1972). 77. Animashaun, A. Aetiology of cerebral palsy in Afri- can children. African journal of medical sciences, 2: 165-171 (1971). 78. Malamud, N.H.H. et al. An etiologic and diagnostic study of cerebral palsy. Journal of pediatrics, 65: WHO Bulletin OMS. Vol 71 1993 257 N. Senanayake & G.C. Roman 270-293 (1964). 79. Boon, W.H. A surveillance system to prevent kernicterus in Singapore infants. Journal of the Singapore Paediatric Society, 17: 1-9 (1975). 80. Gourie-Devi, M. et al. Neuroepidemiological study in semiurban and rural areas in south India: pattern of neurological disorders including motor neurone disease. In: Gourie-Devi, M., ed. Motor neurone disease. New Delhi, Oxford & IBH Publishing Co., 1987, pp. 11-21. 81. Bharucha, N.E. et al. Pilot survey of the prevalence of neurologic disorders in the Parsi community of Bombay. American joumal of preventive medicine, 3: 293-299 (1987). 82. Hauser W.A. & Nelson K.B. Epidemiology of epi- lepsy in children. Cleveland Clinic journal of medi- cine, 56: S185-S194 (1989). 83. Haddock, D.R.W. An attempt to assess the preva- lence of epilepsy in Accra. Ghana medical journal, 6: 140 (1967). 84. Tellang, B.V. & Hettiaratchi, E.S.G. Patterns of epilepsy in Kenya-a clinical analysis of 115 cases. East African medical journal, 58: 437-444 (1981). 85. Cosnett, J.E. Neurological disorders in the Zulu. Neurology (Minneap.), 14: 443-454 (1964). 86. Bittencourt, P.R.M. & Turner, M. Epilepsy in the third world: Latin American aspects. In: Dam, M. & Gram, L., ed. Comprehensive epileptology. New York, Raven Press, 1991, pp. 807-820. 87. Mirando, E.H. & Ranasinghe, L. Lead encephalo- pathy in children: uncommon clinical aspects. Medi- cal journal of Australia, 2: 966-968 (1970). 88. Greengard, J. et al. Lead encephalopathy in child- ren: intravenous use of urea and its management. New England journal of medicine, 264: 1027-1030 (1961). 89. Khare, S.B. et al. Epidemic outbreak of neuromus- cular manifestations due to chronic BHC poisoning. Joumal of the Association of Physicians of India, 25: 216-222 (1977). 90. Nag, D. et al. Epilepsy epidemic due to benzahexa- chlorine. Tropical and geographical medicine, 29: 229-232 (1977). 91. Heligren, U. & Rombo, L. Malaria prophylaxis and epilepsy. British medical journal, 297: 1267 (1988). 92. Heckmatt, J. et al. Seizure induction by alcohol in patients with epilepsy: experience in two hospital clinics. Journal of the Royal Society of Medicine, 83: 6-9 (1990). 93. Annegers, J.F. et al. The risks of seizure disorders among relatives of patients with childhood-onset epilepsy. Neurology, 32: 174-179 (1982). 258 WHO Bulletin OMS. Vol 71 1993