There is emerging evidence of a bidirectional association between dementia, particularly Alzheimer’s dementia (AD), and epilepsy based on observations that support an overlap in patterns of neuronal degeneration in the hippocampal circuit that may explain the progressive worsening of epilepsy. cognition and memory in both conditions.
Previous studies show a 2- to 10-fold increased risk of seizures in patients with dementia and an increased risk of dementia in patients with epilepsy.
However, the role of epileptogenesis remains unclear. Given the increasing life expectancy of the population, understanding the epidemiological link between these 2 common conditions is of great importance for public health.
The authors evaluated the incidence of dementia and epilepsy in the Framingham Heart Study (FHS), a large population-based cohort, to determine the influence of each condition on the risk of developing the other and to evaluate potential modifying factors. The FHS provides prospective dementia surveillance and conducts a rigorous review of available studies and medical records to retrospectively identify cases of epilepsy.
| Methods |
The FHS is an ongoing longitudinal study that began in 1948 with the enrollment of 5,209 participants (Gen 1 cohort) to prospectively investigate cardiovascular risk factors. In 1971, the cohort’s offspring and their spouses (n = 5124) were enrolled in the offspring cohort (Gen 2). Surveillance of both cohorts is based on controls every 2 and 4 years, respectively.
The FHS administers surveillance forms at each checkup, and has access to medical records from outpatient, hospital, and emergency visits. To identify epilepsy cases, both cohorts were screened using (1) record review for any neurological condition performed as part of the Stroke and Dementia Study or history of conditions with a high risk of seizures (brain injury traumatic brain injury [TBI] and brain tumors), (2) self-report of seizures or syncope on each control’s questionnaire, (3) International Classification of Diseases (ICD-9) codes associated with epilepsy, myoclonus, seizures/seizures, syncope or loss of consciousness, and (4) use of antiepileptics (AE).
After consensus review, cases were adjudicated as (1) definite, (2) probable, or (3) suspected epilepsy; (4) definite/probable/suspected single unprovoked seizure; (5) acute symptomatic seizure; (6) no epilepsy or seizures; or (7) cases with insufficient data to reach a diagnosis based on the International League Against Epilepsy definition of epilepsy. Dementia cases meet DSM-IV criteria.
Screening for dementia began at control 17 (1981–1984) and control 5 (1991–1995) in Gen 1 and Gen 2, respectively, and screening for epilepsy at control 21 (1988–1992) in Gen 1 and in control 5 in Gen 2. A case-control design was used to investigate cases of epilepsy after diagnosis of dementia. Each dementia case was matched in the year of diagnosis (match year) with 3 controls from the same cohort (Gen 1 or Gen 2) and of the same sex, who did not have dementia in that year and were aged within 2 years after the age of the corresponding case.
Epilepsy follow-up was until 2016, with cases followed until the year of diagnosis and controls followed until the last year without known epilepsy. Each epilepsy case was compared in the year of diagnosis (match year) with 3 controls from the same cohort (Gen 1 or Gen 2) and of the same sex, who did not have epilepsy in that year and were aged within 2 years after the age of the case.
Dementia cases were followed until the year of diagnosis, and controls until at least age 65 years and until the last known dementia-free year. Epilepsy cases were compared with non-epilepsy controls with respect to incident dementia, and dementia cases with non-demented controls with respect to incident epilepsy. Models were adjusted for education and presence of the APOE e4 (APOE4) allele, and effect modification by age, sex, APOE4, and educational level was investigated.
| Results |
A total of 4906 participants were followed up to at least age 65 and had information on epilepsy and dementia. There were 888 cases of dementia (815 with screening for epilepsy) and 90 cases of epilepsy (43 with screening for dementia beyond age 65).
Among the 815 patients with dementia and follow-up for epilepsy, there were 3 controls matched by cohort, sex, and age for each of the 660 dementia cases (78% with AD, 10% dementia with Lewy bodies [LCD], 5 % vascular dementia [VaD], 1% fronto-temporal dementia [FTD], and 6% others) for a total of 2,640 participants.
There were 58 cases of epilepsy during follow-up, 19 (2.9%) in those with prevalent dementia, and 39 (2%) among dementia-free controls. No significant interactions were observed between dementia and age, sex, APOE4 or education level in their effect on epilepsy. Exclusion of participants with dementia and stroke and their controls showed a slightly increased risk of subsequent epilepsy.
The 43 patients with epilepsy and available cognitive evaluation were matched to 3 epilepsy-free controls by sex, age, and cohort, for a total of 172 participants. There were 51 incident cases of dementia at follow-up (84% with AD type), 18 (41.9%) among patients with epilepsy and 33 (25.6%) among controls. No significant interactions were observed between epilepsy and age, sex or APOE4 in their effect on dementia. Sensitivity analysis limited to cases of epilepsy of unknown cause yielded similar results.
A significant interaction was found between prevalent epilepsy and education in its effect on dementia.
Almost half of the participants (n = 80) had no education beyond high school; of these, 29 developed dementia, and epilepsy did not significantly increase the risk of dementia compared with controls without epilepsy. Among the 90 participants with a higher education level, 22 developed dementia, and epilepsy was associated with an almost 5 times greater risk of developing dementia vs. controls with the same educational attainment.
In an additional analysis, when excluding suspected epilepsy and single unprovoked seizures, the association between epilepsy and subsequent dementia showed an identical effect size, but with a wider confidence interval. The same was true among participants with prevalent dementia when suspected cases of epilepsy/seizures were excluded.
| Discussion |
This study provides strong evidence that epilepsy and dementia are interrelated, and that patients have twice the risk of developing one condition in the presence of the other.
Epileptic seizures are more common in patients with dementia, with a 2- to 10-fold increased risk in AD, but also in DVaD, DLB, dementia associated with Down syndrome, and FTD. Worsening dementia has been linked to an increased risk of seizures, but the duration and age of onset have yielded conflicting data. The authors’ findings support a nearly doubled risk of developing epilepsy after diagnosis of dementia (mainly AD).
The proportion of patients with AD who have had at least one unprovoked seizure varies between 1.5% and 64% in different prospective and retrospective hospital studies. The lower rates observed in this study may reflect differences in epilepsy/seizure case adjudication methods, study design, dementia surveillance methods, and racial composition of the study populations. FHS participants are mostly Caucasian, but a higher rate of seizures has been found among black patients.
The mechanism driving epileptogenesis in dementia is not well defined. In mice, excessive deposition of β-amyloid peptide (Aβ) is associated with neuronal loss and hippocampal hyperexcitability, impaired synaptic plasticity, and generation of interictal epileptiform activity that may precede memory decline.
In humans, direct amyloid epileptogenicity is supported by observations of very high risk of seizures in patients with early-onset AD secondary to autosomal dominant mutations in the amyloid precursor protein (APP), presenilin 1 (PSEN1) and, to a lesser extent, presenilin 2 (PSEN2).
The APOE4 allele has also been linked to late-onset epilepsy, even in people without dementia, as well as post-traumatic epilepsy. In addition to the modulating effect of tau protein on amyloid excitotoxicity, there appears to be a direct proepileptic toxicity, as seizures are also observed in tauopathies lacking amyloid deposition (FTD).
Although the risk of seizures in dementia has been widely studied, very few studies have reported on the risk of dementia in patients with epilepsy. In a Dutch study, a history of epilepsy was shown to carry a moderate risk of developing dementia. This result is similar to the 2-fold increased risk observed in the present study, but the interaction with education level is reported for the first time. Among more educated participants, the risk of dementia was almost 5 times higher in those with epilepsy vs. controls of the same educational level.
The Rotterdam study showed that the prevalence of AD was higher among less educated participants, despite adjustment for cardiovascular disease, and it was argued that early dementia may be missed in highly educated people.
A recent study by Horvath et al. showed that higher education was associated with a higher risk of comorbid seizures, a finding that may be associated with more advanced disease at the time of dementia diagnosis among the more educated. One possible explanation is that more educated patients have a high degree of AD or neurodegenerative pathology before showing obvious cognitive symptoms, leading to seizures rather than clinical dementia. New studies will be required to examine the dose effect of education on the risk of dementia in patients with epilepsy.
There is increasing evidence that epilepsy can coexist in certain individuals in the early stages of dementia and accelerate its process.
In a retrospective study examining the prevalence of adult-onset cryptogenic seizures prior to the clinical manifestation of AD, seizures were found to begin ~4.6 years before the onset of cognitive symptoms and cognitive decline 3, 6 years earlier compared to those with AD without seizures. Evidence also suggests that this harmful effect can be partially reversed by certain AEs, such as levetiracetam, which was associated with better cognitive performance in patients with AD and seizures vs. lamotrigine and phenobarbital, despite similar seizure control.
This study is one of the few large, prospective, population-based studies to demonstrate a bidirectional link between dementia and epilepsy. A rigorous method was used to define cases of new-onset epilepsy by excluding acute symptomatic seizures, which are believed to result from acute metabolic disorders, as well as other seizure mimics. The sensitivity analysis after exclusion of suspected cases showed a risk identical to that seen in the primary model in both directions of association.
When investigating the association between epilepsy and dementia, a possible confounding factor is the presence of stroke, which is a risk factor for both conditions. In this study, despite the higher number of strokes between participants with dementia and their controls, only 3 had subsequent epilepsy. Therefore, further mediation analyzes could not be completed, but sensitivity analyzes excluding those with dementia and stroke showed similar results. Another limitation is that possible effects of antiepileptic drugs in the diagnosis of dementia were not taken into account. Previous studies showed an association between EC use and dementia, but without a proven causal link.
Possible confounding factors include the underlying indication for EC use, which may be the determinant of dementia, or that some of these indications actually reflect symptoms of early dementia. Seizure frequency, which is associated with worse cognitive outcomes, was not collected in epilepsy cases and therefore could not be assessed if there is an effect of seizure burden on the positive association between epilepsy and later dementia. Finally, the FHS focuses on a primarily Caucasian population, which limits the applicability of these results to other populations.
Understanding the epidemiology of epilepsy and dementia can help shape healthcare policies and reduce the burden of disease. Further studies are needed to define whether there is a bidirectional causal association between these 2 entities or whether shared underlying pathophysiological mechanisms cause both.
