- Email: [email protected]
The prevalence of epilepsy amongst school children in Izmir, Turkey
Seizure 2002; 11: 392–396
doi:10.1053/seiz.2002.0684, available online at http://www.idealibrary.com on
The prevalence of epilepsy amongst school children in Izmir, Turkey ¨ ERGOR‡ & ERAY DIRIK† ADEM AYDIN† , ALP ERGOR‡ , GUL Departments of † Pediatric Neurology and ‡ Public Health, Dokuz Eylul ¨ University Faculty of Medicine, Izmir, Turkey Correspondence to: Adem Aydin MD, 1826/3 Sok. No: 2/1, 35530 Kars¸iyaka, Izmir, Turkey. E-mail: [email protected]
Objective. The aim of this study was to assess the prevalence of epilepsy among school children between the ages of 7–17 in Izmir province, in Turkey. Methods. A cross sectional study was conducted. Sample size is calculated as 4654, from the target population of 420 054 students. A stratified random sampling technique was used to select the schools located in Izmir Metropolitan area. For the 130 (2.8%) students whose families gave a positive ‘epilepsy history’ a telephone interview was conducted in order to verify the diagnosis and to evaluate a etiological factors. Results. Response rate was 90.5% (4216 students). Out of the 130 students who had a positive epilepsy history, 47 (36.2%) were accepted as epilepsy ‘cases’. The crude prevalence rates for females, males and the total study population were found respectively to the 11.3, 11.1 and 11.2 per thousand. Prevalence of active epilepsy rates for females, males and total study population were found respectively to be 4.5, 7.0 and 5.6 per thousand. Conclusion. The prevalence of epilepsy is higher in school-age children in Izmir province compared to that in developed countries. c 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.
Key words: school-children; epilepsy; prevalance.
Epilepsy is a widespread neurological disorder with a relatively high prevalence. The prevalence of epilepsy has been reported in a very wide range, between 2.3– 44 per thousand in all age groups in different parts of the world 1 . Among the paediatric age group, the prevalence of epilepsy is 18.5 (0.8–49.0) per thousand 2–20 . This wide range is probably due to the difference between study designs and population groups. A limited number of studies have been reported from Turkey, in which all age groups were covered. Therefore the data on the prevalence and demographic characteristics of people with epilepsy in Turkey is insufficient. Prevalence rates have been reported as 7.39, 8.80, 10.20 and 6.10 per thousand 21–23 . In Okan et al.’s population based study, which was designed to examine the prevalence of neurological disorders during childhood period, the prevalence of epilepsy has been reported as 9.0 per thousand 24 . However, none of these studies gave a prevalence 1059–1311/02/060392 + 05
$35.00/0
rate specific for school aged-children. In order to define the age specific epilepsy prevalence rate among children aged 7–17, a study was conducted on a sample drawn from primary and high school students of Izmir province.
METHOD Izmir is the third biggest Turkish city located on the west coast of Turkey. Socioeconomic indicators of the province are amongst the highest in the country. Schooling until the 8th grade has been obligatory since 1997. However not all the children in this age group are attending school, even in Izmir province. A cross sectional study was conducted in school children in the metropolitan area of Izmir. A stratified random sample was drawn from a sample frame of 420 054 students in all 443 public and private primary and high schools, excluding schools for those with
c 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.
Epilepsy prevalence in Izmir
393
mental retardation and developmental disorders 25 . The information on the study population was obtained from the Bureau of Statistics of the Provincial Education Administration of Izmir. The sample frame consisted 328 888 primary and 91 116 high school students. Stratification was done according to the subdivisions of the province and the sample was selected proportional to the primary and high school distribution in the subdivisions. The sample size was estimated as 4654 students with a confidence level of 95% using EPI Info 6 27 . The school were randomly chosen among 281 primary and 162 high schools, and they were represented in the sample proportional to their size. Each primary school had 8 levels with approximately 40 students in each class, while high schools had 3 levels with a similar number of students as the primary schools. Therefore, each primary school were represented by 320 students and each high school by 120 students in the sample (Fig. 1). Target Population: 420,004 Primary Schools (PS) n: 281 Number of students: 328,888
High Schools (HS) n: 162 Number of students: 91,116
Sample size n: 4.654 (α: 0.5; d: 0.4) 4900 questionairres was given out
12 PS 8 classes of each (first through eighth)
9 HS 3 classes of each (first, second, third)
PS HS 3666 (78%) Students 1034 (22%) Students
to the families by the school administration, and were collected within 3 days. The response rate was 90.5%, and 4216 questionnaires were returned out of 4654 that were sent. The dependent variable of the study was ‘positive epilepsy history’. Socio-demographic status of the families, family history of epilepsy and febrile convulsion history, age of the onset, seizure history within the last year were also independent variables. In the questionnaire epilepsy is defined as ‘fainting of the child with involuntary movements of extremities or the body, fixed eye movements, licking, with or without loss of consciousness’ 26 . Children’s families, who answered the question of epilepsy history as ‘YES’ were accepted as ‘probable cases’. Probable cases had a telephone interview by a paediatric neurologist in order to confirm the diagnosis. During the telephone interviews the criteria for diagnosis were questioned in detail for verification of epilepsy cases. If the diagnosis had been made by a neurologist and listed in the follow-up procedure for epilepsy for a certain period of time, the child was accepted as an ‘confirmed epilepsy case’. Therapeutics given by the specialist, birth history, tumours, trauma, cerebrovascular diseases, blood diseases were also recorded. Due to the prognostic differences, cases with a single seizure or a neonatal epileptic seizure history were excluded from the study. Five out of the 130 ‘probable cases’ could not be reached by telephone or refused to answer. In order to calculate the age specific prevalence, Izmir Provincial Health Administration’s 1999 data of population, and 1990 census data were used. SPSS 8.0 software package was used for statistical analysis. Differences between groups were statistically tested using chi-square and confidence intervals were calculated by using EPI Info 6 27 .
Telephone interview for VALIDATION
RESULTS 130 cases with + seizure history
83 were excluded according to the study criteria
47 epilepsy cases
22 active epilepsy
Fig. 1: Sampling method.
Written permission was obtained by the provincial education administration of Izmir. A letter was sent to seek the consent of the families of the sample population. The questionnaire, which had 22 questions, was prepared according to the prevalence study criteria of the guideline for epidemiologic studies which was proposed by The Commission on Epidemiology and Prognosis, International League Against Epilepsy (CEP/ILAE) in 1993 26 . Questionnaires were sent out
Mean age was 11.0 (SD 3.0) for 1996 male students, and 11.4 (SD 3.1) for 2220 female students, cumulative mean of age for the study sample was 11.3 (SD 3.1). Out of 4216 respondents, 38 females and 28 males were classified as ‘cases’. 14 students out of 66 cases, who had a positive seizure history, had their first seizure during the neonatal period (before the end of the first month); 5 of them described only one seizure; therefore, they were excluded from the epilepsy group. 47 responders, 25 female (53.2%) and 22 male (46.8%) were finally accepted as confirmed cases. Age and gender specific prevalence rates for males, females and total population is shown in Table 1. Prevalence rates for females, males and the total study
394
A. Aydin et al.
Table 1: Age and gender specific epilepsy prevalence (EP) and active epilepsy prevalence (AEP) of total study populationa . Male n: 1996 Age
Female n: 2220 AEP
EP
AEP
EP
7–9 (n: 1506)
0.9 CI: 0.39–1.92
0.5 CI: 0.16–1.41
0.9 CI: 0.41–2.03
10–14 (n: 1981)
0.86 CI: 0.39–1.74
0.6 CI: 0.26–1.48
15+ (n: 669)
2.5 CI: 1.12–5.43
Total (n: 4156)
1.1 CI: 0.71–1.69
Total n: 4216 EP
AEP
0.4 CI: 0.11–1.3
0.9 CI: 0.52–1.58
0.4 CI: 0.20–0.99
1.5 CI: 0.88–2.47
0.5 CI: 0.23–1.30
1.2 CI: 0.78–1.80
0.6 CI: 0.32–1.07
1.5 CI: 0.40–4.20
0.5 CI: 0.85–1.97
0.2 CI: 0.10–1.50
1.3 CI: 0.64–2.59
0.7 CI: 0.27–1.81
0.7 CI: 0.40–1.21
1.1 CI: 0.74–1.69
0.4 CI: 0.23–0.86
1.1 CI: 0.83–1.49
0.6 CI: 0.37–0.86
a Rates and confidence intervals of prevalances were shown as percent (%).
population were found respectively to be 11.3, 11.1 and 11.2 per thousand. Prevalence of active epilepsy rates for females, males and total study population were found respectively to be 4.5, 7.0 and 5.6 per thousand population (Table 1). There were no statistical difference between male and female children with epilepsy according to the age of onset of epilepsy (Table 2). Among epileptic ‘cases’ 28 (59.6%) had not had an epileptic seizure during the previous year, while 6 cases (12.8%) had had 1, 4 had had (8.5%) 2, 6 had had (12.8%) 3 and 2 had had (4.3%) more than 4 seizures during the same period of time (Table 3).
injury history (3 of them were male), 6 cases had had a central nervous system infection in the past. Thirty-four cases gave no specific information about any aetiologic factor for epilepsy. Among the 47 cases who were included in the analysis of the study, 6 (13.3%) gave a positive birth trauma history; in this particular group of students school achievement was within normal limits. Students win epilepsy had higher febrile seizure history compared to the nonepileptic group (Table 4). A febrile seizure history was positive for 12.4% of our study sample: the difference between epilepsy cases and the total study sample is statistically significant (P < 0.05).
Table 2: Distribution of the onset of epileptic seizures according to age and gender.
DISCUSSION
Onset age of the seizure
n
Male (%)
n
Female (%)
n
Total (%)
0–5a 6–10 11+
10 9 3
46.0 41.0 13.0
14 7 4
56.0 28.0 13.0
24 16 7
51.0 34.0 15.0
Total
22
100.0
25
100.0
47
100.0
n
Total (%)
a Excluding 1st month of age. b P > 0.05.
Table 3: Positive epileptic seizures in last yeara . Epileptic seizures in the last year
n
Male (%)
n
Female (%)
No seizures 1 seizure 2 seizures 3 seizures 4+ Unknown
12 3 2 3 2 —
54.5 13.6 9.1 13.6 9.1 —
16 3 2 3 — 1
64.0 12.0 8.0 12.0 — 4.0
28 6 4 6 2 1
59.6 12.8 8.5 12.8 4.3 2.1
Total
22
100.0
25
100.0
47
100.0
a All the cases presented in this table were under medical control.
When we explored the epilepsy group’s history in terms of a etiologic factors it was found that 1 case had brain tumour, 2 had cerebrovascular diseases (1 of them was a late vitamin K deficiency), 4 had a head
In this study the cumulative epilepsy prevalence was 11.3 per thousand amongst females, 11.1 per thousand amongst males and 11.2 per thousand for the total population. The results are considerably higher than the developed countries data reported by Tsuboi 6 , Hauser 10 , Sangrador 11 and Maremmani 28 ; their results were 8.0, 5.7, 2.8 and 5.7 per thousand, respectively. This finding could be explained by the lower socioeconomic indicators of Turkey. On the other hand our findings are relatively low compared to the results reported from other developing countries 5, 29, 30 . Active epilepsy is defined in the ‘Guidelines’ as ‘positive epilepsy history within the last 5 years and continuing medication with an antiepileptic drug’ 26 . In our study the active epilepsy prevalence was found to be 4.5 per thousand among females, 7.0 per thousand among males and 5.6 per thousand for the total population in the 7–17 age group. Previously, the active epilepsy prevalence for Turkey was reported as 10.01 per thousand for females, 10.39 per thousand for males and 10.20 per thousand for the total population by Karaagac¸ et al. 22 . Similar findings to Karaagac¸ et al.’s have been reported from Turkey 21, 23 . In many
Epilepsy prevalence in Izmir
395
Table 4: Distribution of the febrile convulsion history of epileptic and nonepileptic groups. Febrile convulsion history
Male Nonepileptics n (%)
No Yes
1707 262
86.7 13.3
Total
1969
100.0
P
Female Epileptics n (%)
Nonepileptics n (%)
17 5
77.3 22.3
1933 254
88.4 11.6
22
100.0
2187
100.0
0.02
studies it has been reported that cumulative epilepsy prevalence rates for paediatric age groups are higher than the general population of people with epilepsy 1 . The differences between the results of Karaagac¸ et al. 22 , the other prevalence studies from Turkey and our findings are probably due to differences between the sample population. In those studies samples were selected from the population which covered all age groups, however our study represents only students aged 7–17. Schools for special education, which serve children with different levels of mental retardation— who were known as high risk groups for epilepsy— were not represented in our study. Thus, the actual prevalence rate could be higher than that reported in our study. In our study the active epilepsy prevalence rate was found to be half the cumulative epilepsy prevalence rate. Similar results have been reported amongst paediatric age groups previously e.g. 2/3 of the cumulative prevalence rate 31, 32 . Although prevalence rates show a variation among age and gender groups, the 95% confidence intervals were not significantly differ. There are some sources of bias, such as economic factors, that could affect our results. Economic factors force students to drop out of school and usually enter the labour force at an early age as a child worker. Therefore, we may not have found those children coming from the low economic income families, who are under higher risk for epilepsy. A positive febrile convulsion history is a recognised epilepsy risk factor 33 . A febrile convulsion history was positive in 12.4% of the cases. Febrile convulsion prevalence is reported to be between 1–4% in western countries 34–36 . Since this risk factor is more common in Turkey, this might contribute to the higher prevalence rates found in our study compared to those in developed countries 6, 10, 11, 28 . The significant association between a febrile seizure history and epilepsy is supported in the literature by the Rochester–Minnesota follow-up study which found that the risk of having epilepsy was 5 times higher for children with febrile seizures when they had reached the age of 25 and the cumulative
Epileptics n (%)
Total study population Nonepileptics Epileptics n (%) n (%)
18 7
72.0 28.0
3640 516
87.6 12.4
25
100.0
4156
100.0
0.001
35 12
74.5 25.5
47
100.0
0.003
epilepsy prevalence for children with positive febrile seizure history was reported to be 6% 35, 36 . In the developed western countries febrile seizure prevalence is reported to be between 1–4%, whilst it is 12.4% in our study. This would be another explanation for our high prevalence compared to that in developed countries. Childhood epilepsy is increased when risk factors such as febrile seizures, head trauma, central nervous system infections, mental retardation and cerebral palsy are present 37 . Special education institutions for children with mental retardation and cerebral palsy are not wide spread in our country, and such an institution does not exist in Izmir. Therefore, these factors were not investigated. On the other hand, when the aetiologic factors wer examined, it was found that 18% of the cases had had a central nervous system infection, 13% had had birth trauma, 12% had had head trauma, 6% had cerebrovascular disease, 2% had a brain tumour history. Central nervous system infection and trauma history in our cases were higher than in most of the literature 38 . Epileptic seizures were very well controlled in terms of attack rate in the previous year for the cases under treatment. A possible explanation of this result could be that resistant cases, due to pathologies such as birth defects, neurometabolic disorders or developmental abnormalities were probably under represented in our study population, since those patients usually drop out from school or attend special education institutions which were not included in our sample frame.
ACKNOWLEDGEMENT We gratefully acknowledge the help of the Bureau of Statistics of Provincial Education Administration of Izmir.
REFERENCES 1. Bharucha, N. E. and Sharvon, S. D. Epidemiology in developing countries. In: Epilepsy a Comprehensive Textbook. Vol. 1, (Eds J. Engel and T. A. Pedley). Philadelphia,
396 Lippincott-Raven, 1998: pp. 105–118. 2. Rose, S. W., Penry, J. K., Markush, R. E., Radloff, L. A. and Putnam, P. L. Prevalence of epilepsy in children. Epilepsia 1973; 14: 133–152. 3. Meighan, S. S., Queener, L. and Weitman, M. Prevalence of epilepsy in children of Multnomah County, Oregon. Epilepsia 1976; 17: 245–256. 4. Baumann, R. J., Marx, M. B. and Leonidakis, M. G. Epilepsy in rural Kentucky: prevalence in a population of school age children. Epilepsia 1978; 19: 75–80. 5. Chiofalo, N., Kirschbaum, A., Fuentes, A., Cordero, M. L. and Madsen, J. Prevalence of epilepsy in children of Melipilla, Chile. Epilepsia 1979; 20: 261–266. 6. Tsuboi, T. Prevalence and incidence of epilepsy in Tokyo. Epilepsia 1988; 29: 103–110. 7. Koul, R., Razdan, S. and Motta, A. Prevalence and pattern of epilepsy (Lath/Mirgi/Laran) in rural Kashmir, India. Epilepsia 1988; 29: 116–122. 8. Cowan, L. D., Leviton, A., Bodensteiner, J. B. and Doherty, L. Problems in estimating the prevalence of epilepsy in children: the yield from different sources of information. Paediatric Perinatal Epidemiology 1989; 3: 386–401. 9. Aziz, H., Hasan, M. and Hasan, K. Z. Prevalence of epilepsy in children (a population survery report). JPMA Journal of the Pakistan Medical Association 1991; 41: 134–136. 10. Hauser, W. A., Annegers, J. F. and Kurland, L. T. Prevalence of epilepsy in Rochester, Minnesota: 1940–1980. Epilepsia 1991; 32: 429–445. 11. Sangrador, C. O. and Palencia Luaces, R. Study of the prevalence of epilepsy among school children in Valladolid, Spain. Epilepsia 1991; 32: 791–797. 12. Sillanpaa, M. Epilepsy in children: prevalence, disability, and handicap. Epilepsia 1992; 33: 444–449. 13. Murphy, C. C., Trevathan, E. and Yeargin-Allsopp, M. Prevalence of epilepsy and epileptic seizures in 10-year-old children: results from the Metropolitan Atlanta Developmental Disabilities Study. Epilepsia 1995; 36: 866–72. 14. Sidenvall, R., Forsgren, L. and Heijbel, J. Prevalence and characteristics of epilepsy in children in northern Sweden. Seizure 1996; 5: 139–146. 15. Endziniene, M., Pauza, V. and Miseviciene, I. Prevalence of childhood epilepsy in Kaunas, Lithuania. Brain Development 1997; 19: 379–387. 16. Hackett, R. J., Hackett, L. and Bhakta, P. The prevalence and associated factors of epilepsy in children in Calicut District, Kerala, India. Acta Paediatrical 1997; 86: 1257–1260. 17. Eriksson, K. J. and Koivikko, M. J. Prevalence, classification, and severity of epilepsy and epileptic syndromes in children. Epilepsia 1997; 38: 1275–1282. 18. Bellmann, A., Napa, A., Soot, A. and Taivik, T. Prevalence of childhood epilepsy in Estonia. Epilepsia 1999; 40: 1011–1019. 19. Nicoletti, A., Reggio, A., Bartoloni, A. et al. Prevalence of epilepsy in rural Bolivia: a door-to-door survey. Neurology 1999; 53: 2064–2069. 20. Waaler, P. E., Blom, B. H., Skeidsvoll, H. and Mykletun, A. Prevalence, classification, and severity of epilepsy in children in western Norway. Epilepsia 2000; 41: 802–810. 21. Guvener, H. A., Akhtar, S. W. and Hasan, K. Z. Comparative epidemiology of epilepsy in Pakistan and Turkey: population-
A. Aydin et al.
22.
23.
24.
25.
26.
27. 28.
29. 30.
31.
32.
33.
34.
35.
36.
37.
38.
based studies using identical protocols. Epilepsia 1997; 38: 716–722. Karaagac, N., Yeni, S. N., Senocak, M., Bozluolcay, M., Savrun, F. K., Ozdemir, H. and Cagatay, P. Prevalence of epilepsy in Silivri, a rural area of Turkey. Epilepsia 1999; 40: 637–632. Topalkara, K., Aky¨uz, A., Sumer, H., Bekar, D., Topaktas, S. and Dener, S. An epilepsy prevalence study performed using stratified sampling method among urban residents of Sivas. Epilepsi 1999; 5: 24–29. Okan, N., Okan, M., Eralp, O and Aytekin, A. H. The prevalence of neurological disorders among children in Gemlik (Turkey). Developmental Medicine & Child Neurology 1995; 37: 597–603. Kleinbaum, D. G., Kupper, L. L. and Morgenstern, H. Epidemiologic Research. Principals and Quantitative Methods. New York, Van Nostrand Reinhold, 1982: pp. 54–55, 321–324, 332–333. Commision on Epidemiology and Prognosis, InternationalLeague against Epilepsy. Guidelines for epidemiologic studies onepilepsy. Epilepsia 1993; 34: 592–596. Rothman, K. J. Modern Epidemiology. Boston/Toronto, Little Brown and Company, 1986: pp. 203–208, 289–295. Maremmani, C., Rossi, G., Bonuecille, U. and Murri, L. Descriptive epidemiologic study of epilepsy syndrome in a district of northwest Tuscany, Italy. Epilepsia 1991; 32: 298–294. Gomesz, J. G., Arciniegas, E. and Torres, J. Prevalence of epilepsy in Bogota, Colombia. Neurology 1978; 28: 90–94. Garcia, F., de Lao, S. L., Castillo, L., Larreategui, M., Archbold, C., Brenes, M. M. and Reeves, W. C. Epidemiology of epilepsy in Guaymi Indians from Bocas del Torl Province, Republic of Panama. Epilepsia 1990; 31: 718–723. Blom, S., Heijbel, J. and Bergfors, P. G. Incidence of epilepsy in children: a follow-up study three years after the first seizure. Epilepsia 1978; 19: 343–350. Olafsson, E., Hauser, W. A., Ludvigsson, P. and Gudmundsson, G. Incidence of epilepsy in rural Iceland: a population-based study. Epilepsia 1996; 37: 951–955. Annegers, J. F., Hauser, W. A., Anderson, V. E. and Kurland, L. T. The risk of seizure disorder among relatives of patients with childhood epilepsy. Neurology 1982; 32: 174–179. Nelson, K. B. and Ellenberg, J. H. Antecendents of seizure disorders in early childhood. American Journal of Diseases of Children 1986; 140: 1051–1061. Rocca, W. A., Sharbroug, F. W., Hausher, W. A., Annegers, J. F. and Schoeberg, B. S. Risk factors for generalized tonic– clonic seizures: a population based-control study. Annals of Neurology 1987; 21: 22–31. Anneger, J. F., Rocca, W. A. and Hauser, W. A. Causes of epilepsy. Contributions of Rochester Epidemiology Project. Mayo Clinic Proceedings 1996; 71: 570–575. Hauser, W. A. Incidence and prevalence. In: Epilepsy, A Comprehensive Textbook. Vol. 1 (Eds J. Engel and T. A. Pedley). Philadelphia, Lippincott-Raven, 1998: pp. 47–57. Hesdorrfer, D. C. and Verity, C. M. Risk factors. In: Epilepsy, A Comprehensive Textbook. Vol. 1 (Eds J. Engel and T. A. Pedley). Philadelphia, Lippincott-Raven, 1998: pp. 59–67.
doi:10.1053/seiz.2002.0684, available online at http://www.idealibrary.com on
The prevalence of epilepsy amongst school children in Izmir, Turkey ¨ ERGOR‡ & ERAY DIRIK† ADEM AYDIN† , ALP ERGOR‡ , GUL Departments of † Pediatric Neurology and ‡ Public Health, Dokuz Eylul ¨ University Faculty of Medicine, Izmir, Turkey Correspondence to: Adem Aydin MD, 1826/3 Sok. No: 2/1, 35530 Kars¸iyaka, Izmir, Turkey. E-mail: [email protected]
Objective. The aim of this study was to assess the prevalence of epilepsy among school children between the ages of 7–17 in Izmir province, in Turkey. Methods. A cross sectional study was conducted. Sample size is calculated as 4654, from the target population of 420 054 students. A stratified random sampling technique was used to select the schools located in Izmir Metropolitan area. For the 130 (2.8%) students whose families gave a positive ‘epilepsy history’ a telephone interview was conducted in order to verify the diagnosis and to evaluate a etiological factors. Results. Response rate was 90.5% (4216 students). Out of the 130 students who had a positive epilepsy history, 47 (36.2%) were accepted as epilepsy ‘cases’. The crude prevalence rates for females, males and the total study population were found respectively to the 11.3, 11.1 and 11.2 per thousand. Prevalence of active epilepsy rates for females, males and total study population were found respectively to be 4.5, 7.0 and 5.6 per thousand. Conclusion. The prevalence of epilepsy is higher in school-age children in Izmir province compared to that in developed countries. c 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.
Key words: school-children; epilepsy; prevalance.
Epilepsy is a widespread neurological disorder with a relatively high prevalence. The prevalence of epilepsy has been reported in a very wide range, between 2.3– 44 per thousand in all age groups in different parts of the world 1 . Among the paediatric age group, the prevalence of epilepsy is 18.5 (0.8–49.0) per thousand 2–20 . This wide range is probably due to the difference between study designs and population groups. A limited number of studies have been reported from Turkey, in which all age groups were covered. Therefore the data on the prevalence and demographic characteristics of people with epilepsy in Turkey is insufficient. Prevalence rates have been reported as 7.39, 8.80, 10.20 and 6.10 per thousand 21–23 . In Okan et al.’s population based study, which was designed to examine the prevalence of neurological disorders during childhood period, the prevalence of epilepsy has been reported as 9.0 per thousand 24 . However, none of these studies gave a prevalence 1059–1311/02/060392 + 05
$35.00/0
rate specific for school aged-children. In order to define the age specific epilepsy prevalence rate among children aged 7–17, a study was conducted on a sample drawn from primary and high school students of Izmir province.
METHOD Izmir is the third biggest Turkish city located on the west coast of Turkey. Socioeconomic indicators of the province are amongst the highest in the country. Schooling until the 8th grade has been obligatory since 1997. However not all the children in this age group are attending school, even in Izmir province. A cross sectional study was conducted in school children in the metropolitan area of Izmir. A stratified random sample was drawn from a sample frame of 420 054 students in all 443 public and private primary and high schools, excluding schools for those with
c 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.
Epilepsy prevalence in Izmir
393
mental retardation and developmental disorders 25 . The information on the study population was obtained from the Bureau of Statistics of the Provincial Education Administration of Izmir. The sample frame consisted 328 888 primary and 91 116 high school students. Stratification was done according to the subdivisions of the province and the sample was selected proportional to the primary and high school distribution in the subdivisions. The sample size was estimated as 4654 students with a confidence level of 95% using EPI Info 6 27 . The school were randomly chosen among 281 primary and 162 high schools, and they were represented in the sample proportional to their size. Each primary school had 8 levels with approximately 40 students in each class, while high schools had 3 levels with a similar number of students as the primary schools. Therefore, each primary school were represented by 320 students and each high school by 120 students in the sample (Fig. 1). Target Population: 420,004 Primary Schools (PS) n: 281 Number of students: 328,888
High Schools (HS) n: 162 Number of students: 91,116
Sample size n: 4.654 (α: 0.5; d: 0.4) 4900 questionairres was given out
12 PS 8 classes of each (first through eighth)
9 HS 3 classes of each (first, second, third)
PS HS 3666 (78%) Students 1034 (22%) Students
to the families by the school administration, and were collected within 3 days. The response rate was 90.5%, and 4216 questionnaires were returned out of 4654 that were sent. The dependent variable of the study was ‘positive epilepsy history’. Socio-demographic status of the families, family history of epilepsy and febrile convulsion history, age of the onset, seizure history within the last year were also independent variables. In the questionnaire epilepsy is defined as ‘fainting of the child with involuntary movements of extremities or the body, fixed eye movements, licking, with or without loss of consciousness’ 26 . Children’s families, who answered the question of epilepsy history as ‘YES’ were accepted as ‘probable cases’. Probable cases had a telephone interview by a paediatric neurologist in order to confirm the diagnosis. During the telephone interviews the criteria for diagnosis were questioned in detail for verification of epilepsy cases. If the diagnosis had been made by a neurologist and listed in the follow-up procedure for epilepsy for a certain period of time, the child was accepted as an ‘confirmed epilepsy case’. Therapeutics given by the specialist, birth history, tumours, trauma, cerebrovascular diseases, blood diseases were also recorded. Due to the prognostic differences, cases with a single seizure or a neonatal epileptic seizure history were excluded from the study. Five out of the 130 ‘probable cases’ could not be reached by telephone or refused to answer. In order to calculate the age specific prevalence, Izmir Provincial Health Administration’s 1999 data of population, and 1990 census data were used. SPSS 8.0 software package was used for statistical analysis. Differences between groups were statistically tested using chi-square and confidence intervals were calculated by using EPI Info 6 27 .
Telephone interview for VALIDATION
RESULTS 130 cases with + seizure history
83 were excluded according to the study criteria
47 epilepsy cases
22 active epilepsy
Fig. 1: Sampling method.
Written permission was obtained by the provincial education administration of Izmir. A letter was sent to seek the consent of the families of the sample population. The questionnaire, which had 22 questions, was prepared according to the prevalence study criteria of the guideline for epidemiologic studies which was proposed by The Commission on Epidemiology and Prognosis, International League Against Epilepsy (CEP/ILAE) in 1993 26 . Questionnaires were sent out
Mean age was 11.0 (SD 3.0) for 1996 male students, and 11.4 (SD 3.1) for 2220 female students, cumulative mean of age for the study sample was 11.3 (SD 3.1). Out of 4216 respondents, 38 females and 28 males were classified as ‘cases’. 14 students out of 66 cases, who had a positive seizure history, had their first seizure during the neonatal period (before the end of the first month); 5 of them described only one seizure; therefore, they were excluded from the epilepsy group. 47 responders, 25 female (53.2%) and 22 male (46.8%) were finally accepted as confirmed cases. Age and gender specific prevalence rates for males, females and total population is shown in Table 1. Prevalence rates for females, males and the total study
394
A. Aydin et al.
Table 1: Age and gender specific epilepsy prevalence (EP) and active epilepsy prevalence (AEP) of total study populationa . Male n: 1996 Age
Female n: 2220 AEP
EP
AEP
EP
7–9 (n: 1506)
0.9 CI: 0.39–1.92
0.5 CI: 0.16–1.41
0.9 CI: 0.41–2.03
10–14 (n: 1981)
0.86 CI: 0.39–1.74
0.6 CI: 0.26–1.48
15+ (n: 669)
2.5 CI: 1.12–5.43
Total (n: 4156)
1.1 CI: 0.71–1.69
Total n: 4216 EP
AEP
0.4 CI: 0.11–1.3
0.9 CI: 0.52–1.58
0.4 CI: 0.20–0.99
1.5 CI: 0.88–2.47
0.5 CI: 0.23–1.30
1.2 CI: 0.78–1.80
0.6 CI: 0.32–1.07
1.5 CI: 0.40–4.20
0.5 CI: 0.85–1.97
0.2 CI: 0.10–1.50
1.3 CI: 0.64–2.59
0.7 CI: 0.27–1.81
0.7 CI: 0.40–1.21
1.1 CI: 0.74–1.69
0.4 CI: 0.23–0.86
1.1 CI: 0.83–1.49
0.6 CI: 0.37–0.86
a Rates and confidence intervals of prevalances were shown as percent (%).
population were found respectively to be 11.3, 11.1 and 11.2 per thousand. Prevalence of active epilepsy rates for females, males and total study population were found respectively to be 4.5, 7.0 and 5.6 per thousand population (Table 1). There were no statistical difference between male and female children with epilepsy according to the age of onset of epilepsy (Table 2). Among epileptic ‘cases’ 28 (59.6%) had not had an epileptic seizure during the previous year, while 6 cases (12.8%) had had 1, 4 had had (8.5%) 2, 6 had had (12.8%) 3 and 2 had had (4.3%) more than 4 seizures during the same period of time (Table 3).
injury history (3 of them were male), 6 cases had had a central nervous system infection in the past. Thirty-four cases gave no specific information about any aetiologic factor for epilepsy. Among the 47 cases who were included in the analysis of the study, 6 (13.3%) gave a positive birth trauma history; in this particular group of students school achievement was within normal limits. Students win epilepsy had higher febrile seizure history compared to the nonepileptic group (Table 4). A febrile seizure history was positive for 12.4% of our study sample: the difference between epilepsy cases and the total study sample is statistically significant (P < 0.05).
Table 2: Distribution of the onset of epileptic seizures according to age and gender.
DISCUSSION
Onset age of the seizure
n
Male (%)
n
Female (%)
n
Total (%)
0–5a 6–10 11+
10 9 3
46.0 41.0 13.0
14 7 4
56.0 28.0 13.0
24 16 7
51.0 34.0 15.0
Total
22
100.0
25
100.0
47
100.0
n
Total (%)
a Excluding 1st month of age. b P > 0.05.
Table 3: Positive epileptic seizures in last yeara . Epileptic seizures in the last year
n
Male (%)
n
Female (%)
No seizures 1 seizure 2 seizures 3 seizures 4+ Unknown
12 3 2 3 2 —
54.5 13.6 9.1 13.6 9.1 —
16 3 2 3 — 1
64.0 12.0 8.0 12.0 — 4.0
28 6 4 6 2 1
59.6 12.8 8.5 12.8 4.3 2.1
Total
22
100.0
25
100.0
47
100.0
a All the cases presented in this table were under medical control.
When we explored the epilepsy group’s history in terms of a etiologic factors it was found that 1 case had brain tumour, 2 had cerebrovascular diseases (1 of them was a late vitamin K deficiency), 4 had a head
In this study the cumulative epilepsy prevalence was 11.3 per thousand amongst females, 11.1 per thousand amongst males and 11.2 per thousand for the total population. The results are considerably higher than the developed countries data reported by Tsuboi 6 , Hauser 10 , Sangrador 11 and Maremmani 28 ; their results were 8.0, 5.7, 2.8 and 5.7 per thousand, respectively. This finding could be explained by the lower socioeconomic indicators of Turkey. On the other hand our findings are relatively low compared to the results reported from other developing countries 5, 29, 30 . Active epilepsy is defined in the ‘Guidelines’ as ‘positive epilepsy history within the last 5 years and continuing medication with an antiepileptic drug’ 26 . In our study the active epilepsy prevalence was found to be 4.5 per thousand among females, 7.0 per thousand among males and 5.6 per thousand for the total population in the 7–17 age group. Previously, the active epilepsy prevalence for Turkey was reported as 10.01 per thousand for females, 10.39 per thousand for males and 10.20 per thousand for the total population by Karaagac¸ et al. 22 . Similar findings to Karaagac¸ et al.’s have been reported from Turkey 21, 23 . In many
Epilepsy prevalence in Izmir
395
Table 4: Distribution of the febrile convulsion history of epileptic and nonepileptic groups. Febrile convulsion history
Male Nonepileptics n (%)
No Yes
1707 262
86.7 13.3
Total
1969
100.0
P
Female Epileptics n (%)
Nonepileptics n (%)
17 5
77.3 22.3
1933 254
88.4 11.6
22
100.0
2187
100.0
0.02
studies it has been reported that cumulative epilepsy prevalence rates for paediatric age groups are higher than the general population of people with epilepsy 1 . The differences between the results of Karaagac¸ et al. 22 , the other prevalence studies from Turkey and our findings are probably due to differences between the sample population. In those studies samples were selected from the population which covered all age groups, however our study represents only students aged 7–17. Schools for special education, which serve children with different levels of mental retardation— who were known as high risk groups for epilepsy— were not represented in our study. Thus, the actual prevalence rate could be higher than that reported in our study. In our study the active epilepsy prevalence rate was found to be half the cumulative epilepsy prevalence rate. Similar results have been reported amongst paediatric age groups previously e.g. 2/3 of the cumulative prevalence rate 31, 32 . Although prevalence rates show a variation among age and gender groups, the 95% confidence intervals were not significantly differ. There are some sources of bias, such as economic factors, that could affect our results. Economic factors force students to drop out of school and usually enter the labour force at an early age as a child worker. Therefore, we may not have found those children coming from the low economic income families, who are under higher risk for epilepsy. A positive febrile convulsion history is a recognised epilepsy risk factor 33 . A febrile convulsion history was positive in 12.4% of the cases. Febrile convulsion prevalence is reported to be between 1–4% in western countries 34–36 . Since this risk factor is more common in Turkey, this might contribute to the higher prevalence rates found in our study compared to those in developed countries 6, 10, 11, 28 . The significant association between a febrile seizure history and epilepsy is supported in the literature by the Rochester–Minnesota follow-up study which found that the risk of having epilepsy was 5 times higher for children with febrile seizures when they had reached the age of 25 and the cumulative
Epileptics n (%)
Total study population Nonepileptics Epileptics n (%) n (%)
18 7
72.0 28.0
3640 516
87.6 12.4
25
100.0
4156
100.0
0.001
35 12
74.5 25.5
47
100.0
0.003
epilepsy prevalence for children with positive febrile seizure history was reported to be 6% 35, 36 . In the developed western countries febrile seizure prevalence is reported to be between 1–4%, whilst it is 12.4% in our study. This would be another explanation for our high prevalence compared to that in developed countries. Childhood epilepsy is increased when risk factors such as febrile seizures, head trauma, central nervous system infections, mental retardation and cerebral palsy are present 37 . Special education institutions for children with mental retardation and cerebral palsy are not wide spread in our country, and such an institution does not exist in Izmir. Therefore, these factors were not investigated. On the other hand, when the aetiologic factors wer examined, it was found that 18% of the cases had had a central nervous system infection, 13% had had birth trauma, 12% had had head trauma, 6% had cerebrovascular disease, 2% had a brain tumour history. Central nervous system infection and trauma history in our cases were higher than in most of the literature 38 . Epileptic seizures were very well controlled in terms of attack rate in the previous year for the cases under treatment. A possible explanation of this result could be that resistant cases, due to pathologies such as birth defects, neurometabolic disorders or developmental abnormalities were probably under represented in our study population, since those patients usually drop out from school or attend special education institutions which were not included in our sample frame.
ACKNOWLEDGEMENT We gratefully acknowledge the help of the Bureau of Statistics of Provincial Education Administration of Izmir.
REFERENCES 1. Bharucha, N. E. and Sharvon, S. D. Epidemiology in developing countries. In: Epilepsy a Comprehensive Textbook. Vol. 1, (Eds J. Engel and T. A. Pedley). Philadelphia,
396 Lippincott-Raven, 1998: pp. 105–118. 2. Rose, S. W., Penry, J. K., Markush, R. E., Radloff, L. A. and Putnam, P. L. Prevalence of epilepsy in children. Epilepsia 1973; 14: 133–152. 3. Meighan, S. S., Queener, L. and Weitman, M. Prevalence of epilepsy in children of Multnomah County, Oregon. Epilepsia 1976; 17: 245–256. 4. Baumann, R. J., Marx, M. B. and Leonidakis, M. G. Epilepsy in rural Kentucky: prevalence in a population of school age children. Epilepsia 1978; 19: 75–80. 5. Chiofalo, N., Kirschbaum, A., Fuentes, A., Cordero, M. L. and Madsen, J. Prevalence of epilepsy in children of Melipilla, Chile. Epilepsia 1979; 20: 261–266. 6. Tsuboi, T. Prevalence and incidence of epilepsy in Tokyo. Epilepsia 1988; 29: 103–110. 7. Koul, R., Razdan, S. and Motta, A. Prevalence and pattern of epilepsy (Lath/Mirgi/Laran) in rural Kashmir, India. Epilepsia 1988; 29: 116–122. 8. Cowan, L. D., Leviton, A., Bodensteiner, J. B. and Doherty, L. Problems in estimating the prevalence of epilepsy in children: the yield from different sources of information. Paediatric Perinatal Epidemiology 1989; 3: 386–401. 9. Aziz, H., Hasan, M. and Hasan, K. Z. Prevalence of epilepsy in children (a population survery report). JPMA Journal of the Pakistan Medical Association 1991; 41: 134–136. 10. Hauser, W. A., Annegers, J. F. and Kurland, L. T. Prevalence of epilepsy in Rochester, Minnesota: 1940–1980. Epilepsia 1991; 32: 429–445. 11. Sangrador, C. O. and Palencia Luaces, R. Study of the prevalence of epilepsy among school children in Valladolid, Spain. Epilepsia 1991; 32: 791–797. 12. Sillanpaa, M. Epilepsy in children: prevalence, disability, and handicap. Epilepsia 1992; 33: 444–449. 13. Murphy, C. C., Trevathan, E. and Yeargin-Allsopp, M. Prevalence of epilepsy and epileptic seizures in 10-year-old children: results from the Metropolitan Atlanta Developmental Disabilities Study. Epilepsia 1995; 36: 866–72. 14. Sidenvall, R., Forsgren, L. and Heijbel, J. Prevalence and characteristics of epilepsy in children in northern Sweden. Seizure 1996; 5: 139–146. 15. Endziniene, M., Pauza, V. and Miseviciene, I. Prevalence of childhood epilepsy in Kaunas, Lithuania. Brain Development 1997; 19: 379–387. 16. Hackett, R. J., Hackett, L. and Bhakta, P. The prevalence and associated factors of epilepsy in children in Calicut District, Kerala, India. Acta Paediatrical 1997; 86: 1257–1260. 17. Eriksson, K. J. and Koivikko, M. J. Prevalence, classification, and severity of epilepsy and epileptic syndromes in children. Epilepsia 1997; 38: 1275–1282. 18. Bellmann, A., Napa, A., Soot, A. and Taivik, T. Prevalence of childhood epilepsy in Estonia. Epilepsia 1999; 40: 1011–1019. 19. Nicoletti, A., Reggio, A., Bartoloni, A. et al. Prevalence of epilepsy in rural Bolivia: a door-to-door survey. Neurology 1999; 53: 2064–2069. 20. Waaler, P. E., Blom, B. H., Skeidsvoll, H. and Mykletun, A. Prevalence, classification, and severity of epilepsy in children in western Norway. Epilepsia 2000; 41: 802–810. 21. Guvener, H. A., Akhtar, S. W. and Hasan, K. Z. Comparative epidemiology of epilepsy in Pakistan and Turkey: population-
A. Aydin et al.
22.
23.
24.
25.
26.
27. 28.
29. 30.
31.
32.
33.
34.
35.
36.
37.
38.
based studies using identical protocols. Epilepsia 1997; 38: 716–722. Karaagac, N., Yeni, S. N., Senocak, M., Bozluolcay, M., Savrun, F. K., Ozdemir, H. and Cagatay, P. Prevalence of epilepsy in Silivri, a rural area of Turkey. Epilepsia 1999; 40: 637–632. Topalkara, K., Aky¨uz, A., Sumer, H., Bekar, D., Topaktas, S. and Dener, S. An epilepsy prevalence study performed using stratified sampling method among urban residents of Sivas. Epilepsi 1999; 5: 24–29. Okan, N., Okan, M., Eralp, O and Aytekin, A. H. The prevalence of neurological disorders among children in Gemlik (Turkey). Developmental Medicine & Child Neurology 1995; 37: 597–603. Kleinbaum, D. G., Kupper, L. L. and Morgenstern, H. Epidemiologic Research. Principals and Quantitative Methods. New York, Van Nostrand Reinhold, 1982: pp. 54–55, 321–324, 332–333. Commision on Epidemiology and Prognosis, InternationalLeague against Epilepsy. Guidelines for epidemiologic studies onepilepsy. Epilepsia 1993; 34: 592–596. Rothman, K. J. Modern Epidemiology. Boston/Toronto, Little Brown and Company, 1986: pp. 203–208, 289–295. Maremmani, C., Rossi, G., Bonuecille, U. and Murri, L. Descriptive epidemiologic study of epilepsy syndrome in a district of northwest Tuscany, Italy. Epilepsia 1991; 32: 298–294. Gomesz, J. G., Arciniegas, E. and Torres, J. Prevalence of epilepsy in Bogota, Colombia. Neurology 1978; 28: 90–94. Garcia, F., de Lao, S. L., Castillo, L., Larreategui, M., Archbold, C., Brenes, M. M. and Reeves, W. C. Epidemiology of epilepsy in Guaymi Indians from Bocas del Torl Province, Republic of Panama. Epilepsia 1990; 31: 718–723. Blom, S., Heijbel, J. and Bergfors, P. G. Incidence of epilepsy in children: a follow-up study three years after the first seizure. Epilepsia 1978; 19: 343–350. Olafsson, E., Hauser, W. A., Ludvigsson, P. and Gudmundsson, G. Incidence of epilepsy in rural Iceland: a population-based study. Epilepsia 1996; 37: 951–955. Annegers, J. F., Hauser, W. A., Anderson, V. E. and Kurland, L. T. The risk of seizure disorder among relatives of patients with childhood epilepsy. Neurology 1982; 32: 174–179. Nelson, K. B. and Ellenberg, J. H. Antecendents of seizure disorders in early childhood. American Journal of Diseases of Children 1986; 140: 1051–1061. Rocca, W. A., Sharbroug, F. W., Hausher, W. A., Annegers, J. F. and Schoeberg, B. S. Risk factors for generalized tonic– clonic seizures: a population based-control study. Annals of Neurology 1987; 21: 22–31. Anneger, J. F., Rocca, W. A. and Hauser, W. A. Causes of epilepsy. Contributions of Rochester Epidemiology Project. Mayo Clinic Proceedings 1996; 71: 570–575. Hauser, W. A. Incidence and prevalence. In: Epilepsy, A Comprehensive Textbook. Vol. 1 (Eds J. Engel and T. A. Pedley). Philadelphia, Lippincott-Raven, 1998: pp. 47–57. Hesdorrfer, D. C. and Verity, C. M. Risk factors. In: Epilepsy, A Comprehensive Textbook. Vol. 1 (Eds J. Engel and T. A. Pedley). Philadelphia, Lippincott-Raven, 1998: pp. 59–67.