Incidence density per 100 000 person-years of epilepsy by calendar year from 1973 to 2013 among the total population in Finland.
Incidence density per 100 000 person-years of epilepsy by age group and calendar year from 1973 to 2013 among the total population in Finland.
Incidence density per 100 000 person-years of epilepsy by sex and calendar year from 1973 to 2013 among the total population in Finland.
Incidence density per 100 000 person-years for infants younger than 1 year by etiology from 1973 to 2013 among the total population in Finland; 3-year moving averages.
Temporal changes in the incidence density of focal epilepsy vs generalized epilepsies per 100 000 person-years from 1973 to 2013 in the total population in Finland.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Sillanpää M, Gissler M, Schmidt D. Efforts in Epilepsy Prevention in the Last 40 YearsLessons From a Large Nationwide Study. JAMA Neurol. 2016;73(4):390–395. doi:10.1001/jamaneurol.2015.4515
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Prevention of new-onset epilepsy is an important public health issue and presents a pressing unmet need. It is unclear whether progress has been made in preventing new-onset epilepsy.
To determine whether progress has been made in the prevention of epilepsy in Finland during the last 40 years.
Design, Setting, and Participants
Using a long-term national register study of 5.04 million Finnish individuals, we looked at first-time inpatient admissions in Finland for a diagnosis of epilepsy from 1973 to 2013. Patients with epilepsy were defined by the occurrence of 2 or more unprovoked seizures. This study was conducted on July 29, 2015.
Main Outcomes and Measures
In Finland, patients with epilepsy are routinely hospitalized at time of diagnosis, thus providing evidence for the incidence of epilepsy.
Of the mean 5.04 million Finnish individuals followed up for the development of epilepsy from 1973 to 2013, 100 792 people were identified as having epilepsy. Of these, 46 995 (47%) had focal epilepsy. The mean age for those included in the study was 45 years for men (interquartile range, 24-65 years) and 46 years for women (interquartile range, 23-71 years). We found no change in the incidence of epilepsy in the age range of those younger than 65 years (60 per 100 000 in 1973 and 64 per 100 000 in 2013). However, there was a significant increase in epilepsy among those older than 65 years (from 57 per 100 000 to 217 per 100 000).
Conclusions and Relevance
We found no evidence that progress has been made in preventing new-onset epilepsy in those younger than 65 years in the last 40 years; in fact, there was a nearly 5-fold rise of new-onset epilepsy among the elderly population.
Studies on the natural history of new-onset epilepsy indicate that seizures can be controlled in up to 80% of patients taking antiepileptic drugs.1,2 Despite their ability to symptomatically block seizures in most patients, antiepileptic drugs have consistently failed to prevent new-onset epilepsy in persons at risk.3,4
Although agents are being developed to prevent epilepsy before it becomes clinically apparent, none of these so-called antiepileptogenic drugs is currently on the market.4-6 Primary prevention of epilepsy by reducing the risk for traumatic brain injury, stroke, and dementia is, by default, the preferred current strategy.7
Earlier attempts to assess temporal changes of the incidence of new-onset epilepsy over several decades were either limited to regions of countries, subgroups of patients, or based on drug reimbursement data and had various methodological limitations.8-11 In an earlier Finnish study, which was based on nationwide antiepileptic drug reimbursement data, the incidence of new-onset epilepsy had significantly declined from 1986 to 2002. In this study, the age-related incidence had significantly decreased over the years in children and adults but had increased in the elderly population, primarily in women.8
In a nationwide study of inpatients and outpatients in Denmark, the incidence of new-onset epilepsy remained rather constant between 1995 and 2002 for most age groups.9 However, the incidence rate increased over time for persons older than 60 years in general and persons older than 80 years in particular9,11,12 and decreased for children, especially for infants younger than 1 year.9 These trends were also seen in the Danish study when the authors restricted the data set to inpatients, which were not shown in the publication.9
Prevention of new-onset epilepsy is an important public health issue and presents a pressing unmet need of current drug treatment of epilepsy. The uncertainties of earlier results and the relevance of temporal changes in the incidence of epilepsy to assess the success of prevention efforts prompted us to examine, in a large nationwide study of inpatients presenting with new-onset epilepsy, whether the incidence of new-onset epilepsy has changed in the last 40 years.
Question: Has prevention of new-onset epilepsy improved during the last 40 years?
Findings: The incidence of epilepsy in those younger than 65 years has not changed, but epilepsy in the elderly population has increased 5-fold.
Meaning: Failure to improve epilepsy prevention is a serious public health problem. The development of drugs to improve the prevention of epilepsy is needed.
The target population consisted of the total population permanently residing in Finland during 1973 to 2013 that were included in the national Hospital Discharge Register (HDR), which was used for data collection. It was established in 1969 and is currently maintained by the National Institute of Health and Welfare. The inclusion criteria for our study were the following: Finnish resident; ascertained as having epilepsy (defined by a history of ≥2 unprovoked seizures); and hospitalized for the first time ever between 1973 and 2013. Excluded were individuals who were not listed in the HDR as having epilepsy; had a second or later hospitalization for epilepsy between 1973 and 2013; or, while having epilepsy, were not hospitalized between 1973 and 2013. The HDR data on hospital outpatient visits, collected since 1998, were not included in our study. Based on data from 1998 to 2013, the number of new-onset epilepsies would have doubled after adding outpatient care (unpublished data from the HDR; November 2015).
Institutional review board approval is not required for register-based research, according to the Medical Research Act, given by Finland’s Ministry of Social Affairs and Health; therefore, approval of this study was not obtained. Patient consent was not obtained because data were from the HDR, which can be used without informed consent for statistical and research purposes.
In this study, we used the year of the first hospitalization for the diagnosis of epilepsy as evidence for the incidence of epilepsy. In Finland, it is national medical practice to routinely hospitalize children and adults with epileptic seizures.13,14 Based on clinical experience on admissions to Finnish hospitals, in all probability, the first admission was prompted by a first seizure, either an episode of status epilepticus or any other type of seizure. While it is true that video electroencephalogram recordings and presurgical evaluations have become more common in recent decades, their number is minimal with regard to the total annual number of admitted patients and, furthermore, most of them are severe cases who have been hospitalized for the assessment prior to those more advanced investigations. The Finnish health care system, particularly concerning epilepsy care, has been very hospital-centered. Among children, 97% of incident cases were derived from hospitals and institutions (after preceding hospital investigations) (Sillanpää15 and unpublished data; M.S.; September 1972). We excluded hospitalizations of nonseizure-related causes. No particular reasons can be seen to assume that the indications for admitting these patients to hospital for investigations and treatment would have changed during the study.
The national administrative registers are well accepted by the Finnish population and considered valid in the Nordic countries.16-21 Furthermore, 2 large Finnish register data sets, collected for different purposes, on 7-year-old children born in 1985-1986 vs 1987 confirmed the HDR data on overall long-term medication use.20
The clinical diagnosis of epilepsy was based on the International Classification of Diseases (ICD): ICD, Eighth Revision (ICD-8) (World Health Organization, 1967); ICD, Ninth Revision (ICD-9) (1975); and ICD, Tenth Revision (ICD-10) (1996). For the diagnosis of epilepsy, the category of 345 was for the years 1973 to 1995 and G40 for the years 1996 to 2013. For focal epilepsy, the categories were 345.30-345.39 (World Health Organization, 1967) for the years 1973 to 1986, 3454-3457A for the years 1987 to 1995, and G40.0-G40.2 for the years 1996 to 2013. Epileptic seizures and syndromes were defined according to the contemporary guidelines of the International League Against Epilepsy.22-25
All incidences were calculated per 100 000 midyear population, separately for men and women and for each age group. In the Figures, age groups 0 to 14 years, 15 to 64 years, and 65 years or older are given. Population figures were taken from the Population Statistics compiled by Statistics Finland (http://www.tilastokeskus.fi/til/vaerak/index_en.html). The trend was calculated by fitting a polynomial function. The R2 value was used as a test of goodness of fit.26
The mean total Finnish population of all ages followed up for the development of epilepsy in 1973 to 2013 was 5.04 million. During the study, 100 792 people with epilepsy were identified (mean age, men: 45 years [interquartile range, 24-65 years] and women: 46 years [interquartile range, 23-71 years]), giving an overall incidence density of 50 new patients with epilepsy per 100 000 person-years. Of all 100 792 people with any type of epilepsy, 46 995 (47%) had focal epilepsy. The overall incidence density was virtually unchanged until 2000; thereafter, a substantial increase was found (Figure 1).
The incidence density remained rather constant between 1973 and 2013 for children (Figure 2A) and middle-aged individuals (Figure 2B). The incidence density increased over time for persons older than 65 years (up to 217 per 100 000; Figure 2C). The incidence rate did not decrease in any age group. The age-specific incidence densities were similar in men and women (Figure 3).
The incidence density was 147 in children younger than 1 year of age (up to 270 new cases per 100 000 person-years). It gradually increased throughout the observation period (Figure 4). While the incidence of genetic generalized epilepsy was virtually unchanged, that of mostly symptomatic focal epilepsy showed a slight increase.
The incidence density of focal epilepsy was substantially increased from 1973 to 2013. The lowest incidence density was seen in 1978, with 7 new cases per 100 000 person-years, and the highest in 2013, with 39 per 100 000 (Figure 5).
The incidence density of focal epilepsy differed by sex in various age groups. The incidence density was higher in boys (25 per 100 000) than in girls (22 per 100 000) in the age group of 0 to 14 years and higher in women (13 per 100 000) than in men (11 per 100 000) in the age group of 15 to 34 years. After age 35 years, the incidence rate was again higher for men (31 per 100 000) than for women (23 per 100 000).
Except for infants of younger than 1 year, the incidence rate of genetic generalized epilepsy was similar to that seen for all patients with epilepsy, with the lowest incidence rate of about 43 new cases per 100 000 person-years between 1973 and 2013 (Figure 5). The incidence rate of genetic generalized epilepsy showed a gradual decline with age for both sexes. A notable exception was identified between 12 and 17 years of age, the incidence rate here being one-fourth higher for women than for men.
In our large nationwide study of inpatients presenting with new-onset epilepsy, we found that the incidence of new-onset epilepsy has not changed in the last 40 years in those younger than 65 years. However, a nearly 5-fold increase over 40 years was noted in male and female patients aged 65 years and older.
Interestingly, focal epilepsy showed an increasing trend in the present study. The increase might be, at least in part, due to increased survival of premature infants. In Finland, the percentage of premature births before 37 gestational weeks (5.6 per 1000 liveborn) and infant mortality rate, including premature born (2.1 per 1000 liveborn in 2010), are lowest worldwide,27 leading to an increased risk for brain injuries and epilepsy.28 The prevalence of cerebral palsy increased from 1.6 per 1000 live births in 1968 to 1972 to 2.5 per 1000 in 1978 to 1982.28,29
In previous studies, which had various methodological limitations outlined in the introduction, the incidence of new-onset epilepsy overall and the age-related incidence had significantly declined over time, except among the elderly population,8-11 where it had increased less than 2-fold, primarily in women, according to a previous report.8 In a nationwide study of inpatients and outpatients in Denmark, the incidence of new-onset epilepsy remained rather constant between 1995 and 2002 for most age groups, except for children, particularly for infants younger than 1 year of age, where it had declined.9
Our main finding was 2-fold; the incidence of new-onset epilepsy has not substantially changed over the last 40 years and the incidence of epilepsy among elderly individuals has increased significantly over the same time. We concluded that the findings suggest no progress has been made in preventing new-onset epilepsy in those younger than 65 years in the last 40 years, and that a substantial rise of new-onset epilepsy among the elderly population was not prevented. Three questions need to be addressed. Why did we fail to achieve a net gain in prevention of new-onset epilepsy in the last 40 years? Why did epilepsy increase significantly in the elderly population over the same time? What are the implications of our findings for management of epilepsy and epilepsy research?
The finding that the incidence of new-onset epilepsy has not changed over the last 40 years (except in those ≥65 years) is unexpected given that age-standardized traffic crashes have decreased in Finland (from 21 per 100 000 in 1970 to 4 per 100 000 in 201230) and military-related injuries have slightly declined31 and despite efforts in primary prevention of traumatic brain injury, stroke, and infections of the brain. However, antiseizure drugs on the market for the last 40 years have not been shown to be able to prevent acquired epilepsy32 and acquired causes have been estimated to be responsible for 40% of new-onset epilepsy.33 Prevention of genetic causes has not made progress during the last 40 years. Our second finding that the incidence of new-onset epilepsy has increased substantially is not unexpected given the change in demographics, with more elderly individuals aged 65 years or older in the Finnish population, from 10.1% to 19.4% in 1973 to 2013,34 and the increase in the incidence of stroke and dementias in the elderly population,35,36 which are the leading causes of new-onset epilepsy in elderly individuals. Longevity in the general population has increased in Finland by 9.6 years, from 71.3 years in 1973 to 80.9 years in 2013.37 The increase in longevity may have contributed to an increased incidence of epilepsy in our sample. However, increased mortality in our incident cohort would reduce longevity, although our population is not biased toward refractory epilepsy, which is well known to be associated with increased mortality.37,38
While, to our knowledge, our investigation is the first population-based study of new-onset epilepsy with a wide range of seizure control and very long median follow-up, limitations existed. Underreporting of the incidence may occur in those with rare seizures, while those with a single status epilepticus or a single cluster, also rare, may have led to the overreporting of new-onset epilepsy. Also, precipitating factors, or lack thereof, were not always documented, which may have led to overreporting of new-onset epilepsy when in fact seizures were provoked. There might have been a selection bias toward more severe cases of epilepsy at least in the first years of follow-up. One possible reason for an increase in the incidence of epilepsy in elderly individuals is that our diagnostic acumen for more subtle epilepsy cases may have improved, particularly among the elderly population. However, assessing the diagnostic acumen was beyond the scope of our study. The study may be somewhat biased toward more severe epilepsy, as patients who were never hospitalized were not included. This study used the previous 2-seizure definition of epilepsy as we could not include single-seizure cases.
Data were extracted using 3 versions of ICD: ICD-8, ICD-9, and ICD-10. The different ICD versions do not fully match. One concern is that with new-onset epilepsy, it may not be possible to determine whether a patient has focal or generalized epilepsy; therefore, an unspecific ICD code would be used for epilepsy. We decided on a focal or generalized classification for all cases. Also, with the use of computed tomography and magnetic resonance imaging, cases would be more likely to have an identified focal lesion even if the seizure semiology was generalized; thus, more recent cases would possibly be classified as focal. We agree that this was a limitation. However, data collection from a very long follow-up period inevitably produces problems including, in this case, changing classifications and developing new diagnostic tools and technologies. A rising trend in focal epilepsies may argue for a real increase, but it can hardly be attributed to any one explanation, because the increase started to occur already in the beginning of the 1980s, when the new imaging technology was not in full use.
Our findings of no net gain in preventing new-onset epilepsy at any age group in the last 40 years nor preventing a significant rise in the elderly population has several implications for the management of epilepsy and epilepsy research. We need to develop antiepileptogenic agents for secondary prevention of acquired new-onset epilepsy.39 Several promising agents are in target-related preclinical development.4 New clinical trial designs to test prevention in new onset-epilepsy have been developed39 and a first clinical trial with statins to prevent stroke-related epilepsy has been published.40 This may be a window of opportunity to lower the incidence of stroke-related epilepsy among elderly individuals.
In conclusion, primary and secondary preventions of epilepsy are continuing to be pressing unmet needs as no progress has been made in preventing new-onset epilepsy in those younger than 65 years in the last 4 decades, and a significant rise of new-onset epilepsy in the elderly population was not prevented.
Corresponding Author: Dieter Schmidt, MD, Epilepsy Research Group, Goethestr 5, D-14163 Berlin, Germany (firstname.lastname@example.org).
Accepted for Publication: November 20, 2015.
Published Online: February 15, 2016. doi:10.1001/jamaneurol.2015.4515.
Author Contributions: Drs Sillanpää and Schmidt had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Sillanpää, Schmidt.
Acquisition, analysis, or interpretation of data: Sillanpää, Gissler.
Drafting of the manuscript: Sillanpää, Schmidt.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Gissler.
Administrative, technical, or material support: Sillanpää.
Study supervision: Sillanpää, Schmidt.
Conflict of Interest Disclosures: None reported.