Alzheimer’s disease (AD) is the most common cause of dementia and one of the most prevalent diseases in the elderly, affecting 10–20 million people worldwide [1]. The disease is characterized by the formation of neurofibrillary tangles and amyloid plaques in the brain, manifesting as progressive cognitive impairment. Annual direct and indirect costs associated with this disease amount to $100 billion, making the search for new diagnostic and therapeutic methods critically important. It is predicted that delaying disease progression by five years could reduce healthcare costs related to AD by half [2–5].

Despite significant advances in laboratory and genetic diagnostics, as well as modern neuroimaging methods (magnetic resonance imaging (MRI) of the brain with morphometry, positron emission tomography), diagnosing dementia-related diseases remains challenging and, in many cases, inaccessible due to the cost of examinations. Currently, electroencephalography (EEG) is not part of the standard examination protocol for patients with dementia, including those with AD. However, numerous literature sources report that pathological brain activity (e.g., slowing of the background rhythm or epileptiform activity) may be recorded during EEG in AD patients, which could exacerbate the progression of cognitive impairments and increase the risk of epileptic seizures, further disadapting patients with AD [6][7].

EEG with functional tests is a simple diagnostic method that allows assessment of the bioelectrical activity of the brain. Non-epileptiform pathological activity, such as theta or delta slowing (regional/diffuse) of bioelectrical brain activity, is a common finding in AD patients during routine EEG [6]. There is evidence suggesting that increased relative power of theta oscillations may be an early sign preceding dementia, thus being an important biomarker of disease progression [8][9].

The detection of epileptiform activity in AD patients holds a great significance. However, routine EEG is often insufficient for capturing epileptiform activity, since approximately 85% of standard EEG recordings fail to detect epileptiform activity even in AD patients with epileptic seizures [10]. This highlights the need for more sensitive methods, such as prolonged video-EEG monitoring including sleep, magnetoencephalography (MEG), or invasive electrode placement through the foramen ovale to identify this pathological activity [10]. Some authors emphasize the higher prevalence of epileptiform activity in AD patients compared to the healthy population, as well as its significance in the progression of cognitive impairments in neurodegenerative disease. Thus, this pathological activity may represent a promising target for intervention in treating cognitive impairments in AD.

The aim of the study is to detect the frequency of epileptiform activity in patients with AD and verify its clinical significance.

The search for scientific literature was conducted in electronic bibliographic databases in the Russian (eLibrary, CyberLeninka) and English (PubMed, Google Scholar) languages. The search queries included the following keywords or their combinations: Alzheimer’s disease, electroencephalography, video-EEG monitoring, subclinical epileptiform activity.

The inclusion criteria for publications were literature systematic reviews and meta-analyses with data on the prevalence of subclinical epileptiform activity in AD, the pathophysiology of its occurrence, and the relationship between the neurodegenerative process, epileptiform activity, and cognitive impairments. The exclusion criteria were publications covering theoretical models, abstracts, and conference materials. In total, 52 articles published from 1998 to 2024 were reviewed.

Subclinical epileptiform activity (SEA) in patients with Alzheimer’s disease (AD)

Subclinical epileptiform activity (SEA) typically refers to epileptiform activity detected in EEG in patients without a history of epileptic seizures. According to scientific research, data on the prevalence and diagnostic significance of epileptiform activity in patients with AD are limited and contradictory. The reviewed publications show a significant variability (ranging from 2% to 54%) in the reported prevalence of SEA among patients diagnosed with AD, likely due to substantial methodological differences across the studies (Table) [11].

The presence of epileptiform activity may exacerbate the progression of cognitive impairments in AD patients. Moreover, its detection may serve as a marker for the potential development of epileptic seizures. For example, Kang and Mendez et al. observed the development of epilepsy in approximately 10–22% of AD patients [7][12].

When investigating the prevalence and significance of epileptiform discharges in patients with various types of dementia, Liedorp et al. [15] found that routine 30-min EEG detected epileptiform activity (predominantly regional in the temporal areas) only in 2% of patients with AD, mild cognitive impairment (MCI) and in 1% of patients with other types of dementia. These rates are similar to those in the general population. Only in 10% of patients with dementia, whose EEG showed epileptiform activity, developed seizures later in the course of the disease [15].

The low detection rate of epileptiform activity in patients with AD prompted researchers to use additional foramen ovale EEG electrodes for its identification. Thus, Lam et al. applied foramen ovale electrodes and showed that subclinical epileptiform activity, predominant during sleep (affecting memory consolidation), can be detected in the early stages of AD, in the absence of changes in the routine scalp EEG [21]. This highlights the need for larger EEG studies using additional techniques, including foramen ovale recording, to determine the diagnostic value of EEG in clinical practice.

According to a number of studies, a significant prevalence of SEA has been detected in AD patients, which is likely associated with the use of prolonged video-EEG monitoring including during sleep. Recently, increasing attention has been paid to the presence of SEA in AD patients due to evidence of a more pronounced cognitive decline and a faster disease progression in patients with SEA compared to those without it [16][18][22][23].

Thus, Horvath et al. analyzed SEA in 52 AD patients and detected significantly more frequent subclinical epileptiform discharges (54%) among this group compared to healthy elderly people of corresponding age (25%) [18]. SEA was detected predominantly in the temporal regions, mostly on the left side, with bitemporal and right-temporal epileptiform activity being less common. The vast majority of SEA episodes occurred during sleep, most frequently recorded during stages 2 and 3 of sleep, while fewer spikes were detected in stage 1 sleep. Moreover, the presence of SEA was associated with more severe cognitive impairments. Horvath et al. showed that in patients with AD combined with SEA, cognitive decline over the observation period (3 years) occurred 1.5 times faster than in patients without epileptiform activity [18]. According to Vossel et al., epileptiform activity was detected in 42.4% of patients with AD and only in 10.5% of individuals in the control group of corresponding age without cognitive impairments [16]. Patients with SEA showed a faster decline in executive functions and global cognition, as measured by the instrument of Mini-Mental State Examination (MMSE), averaging 3.9 points/year compared to 1.6 points/year in patients without SEA [16].

Nous et al. studied patients with different stages of AD (preclinical, MCI, dementia) using such various methods as prolonged EEG, 50-min MEG, and high-density EEG [11]. The prevalence of SEA in these patients was 31% compared to the control group (8%) without cognitive dysfunction. The frequency was increasing along with the disease progressed, i.e., in 50% of cases with developed dementia, in 27% with MCI, and in 25% at the preclinical stage of AD. Although the use of MEG did not lead to a more frequent detection of SEA in AD compared to prolonged EEG and high-density EEG, MEG significantly outperformed other methods in terms of spike detection rate per 50 min (epileptiform activity representation index). Furthermore, in AD patients, the presence of SEA was associated with more pronounced impairments in visuospatial functions and attention, as well as with a relatively larger volume of the left frontal, left temporal, and entorhinal cortex compared to patients without epileptiform activity [7].

Pathophysiological mechanisms of the relationship between epileptiform activity, neurodegenerative process, and cognitive impairment in Alzheimer’s disease

A number of authors consider epileptiform activity to be part of the pathophysiological mechanisms leading to cognitive impairment in AD. The proposed mechanisms include compromised glutamatergic system, excitotoxicity-induced neurodegeneration, accelerated amyloid and tau protein deposition under the influence of epileptiform discharges, remodeling due to increased excitability leading to functional network disconnection, and sleep architecture changes [23].

One hypothesis describes a vicious cycle where molecular changes in AD promote neuronal hyperexcitability [24], which in turn exacerbates the neurodegenerative process in AD [25]. It was reported that in AD, soluble oligomeric Aβ (amyloid-beta), rather than Aβ plaques, is the primary cause of neuronal hyperexcitability [24]. For instance, Aβ1-42 (the most toxic form of soluble Aβ peptides) was to increase neuronal excitability through selective inhibition of K+ currents [26]. It was described that under the influence of Aβ, AD patients experience impaired neuronal and glial glutamate reuptake, leading to excitotoxicity. Similarly, glutamate excitotoxicity is also exacerbated by the effect of Aβ on the function of the N-methyl-D-aspartate receptor (NMDA-R) [27]. It was suggested that activation of cholinergic receptors and Ca2+ channels under the influence of Aβ may cause early subclinical epileptic activity preceding the development of clinical Alzheimer’s disease [28]. Furthermore, it was shown that beta-secretase 1 (BACE1 is a key protein involved in Aβ formation) cleaves the β2 and β4 subunits of the voltage-gated Na+ channel [24]. Cleavage of β2 alters transcription and receptor expression on the cell surface [21]; cleavage of β4 significantly increases intracellular Na+ levels [26]. Both processes contribute to overall neuronal hyperexcitability, which may promote the development of epileptic seizures.

Both epilepsy and Alzheimer’s disease involve neuroinflammation induced by Aβ, characterized by the induction of an immune response in the central nervous system (CNS) in reaction to the pathological process [29]. Inflammation in the CNS is primarily mediated by microglia, astrocytes, and oligodendrocytes [30]. Aβ-induced glial activation leads to the release of numerous pro-inflammatory cytokines (e.g., TNF-α, IL-6, or IL-1β), triggering generalized neuroinflammation. This process, in turn, promotes neurotoxic effects that ultimately result in neuronal hyperexcitability, exacerbating the neurodegenerative process [24]. It was also described that pro-inflammatory cytokines, such as IL-1β, increase neuronal hyperexcitability either by enhancing glutamate release from astrocytes and reducing its reuptake [31], or by upregulating NMDA-R, which increases intracellular Ca2+ influx [32].

Tau protein plays a distinct role in epileptogenesis during AD, given that this protein is one of the key mediators of Aβ-induced epileptogenic mechanisms [33]. Tau protein contributes to neuronal excitotoxicity by increasing extracellular glutamate and causing NMDA-R dysfunction [34]. Tau protein is also associated with abnormal neuronal migration in the hippocampus—a brain structure closely linked to the development of epilepsy [35, 36]. Furthermore, animal models of epileptogenesis confirmed reduced activity of phosphatase 2A, leading to increased p-tau in epileptogenic brain regions [37].

The neurosteroid allopregnanolone was also linked to the development of Alzheimer’s disease [38]. Some authors reported decreased levels of allopregnanolone in the plasma and brain, particularly in the prefrontal cortex, of patients with AD. Reduced allopregnanolone levels lead to diminished neuroprotection, activation of astrocytes and microglia, which in turn promotes the production of neurotoxic cytokines, chemokines, and reactive oxygen and nitrogen species. These mechanisms contribute to the progression of neurodegenerative disease and neuronal hyperexcitability [38].

The key components of the pathogenetic relationship between epileptiform activity and the neurodegenerative process are presented in the Figure below. Increased activity of the glutamatergic system in AD leads to elevated expression of AMPA receptors and mobilization of intracellular calcium. The rise in intracellular calcium levels results in the release of amyloid oligomers into the extracellular space and enhanced phosphorylation of tau oligomers (red arrows). Increased neuronal excitation, manifested as epileptic discharges, is also a consequence of glutamate-mediated hyperexcitability. On the other hand, the accumulation of amyloid plaques and tau neurofibrils alters the expression of glutamate receptors and triggers excessive glutamate release from microglial cells and astrocytes (green arrows). This bidirectional pathological interaction can lead to progressive neurodegeneration (black arrows), which is typically accompanied by cognitive impairment [23].

Pathological remodeling of hippocampo-cortical connections is also considered to play a major role in the presence of epileptiform activity. As a result of epileptiform activity, local intrahippocampal connections are strengthened, while the strength and number of long-range connections are reduced. This remodeling of neural networks leads to relative isolation of the hippocampus from the cortex, impairing the functioning of hippocampo-cortical circuits [23].

Furthermore, the presence of epileptiform activity disrupts physiological sleep patterns and impairs the memory consolidation process. Thalamic sleep spindles at a frequency of 12–16 Hz are crucial for memory formation, synchronizing hippocampal activity with cortical neurons. Slow waves associated with cortical sleep provide the highest degree of synchronization, promoting the activation of hippocampal activity and thalamic sleep spindles. Epileptiform activity contributes to:

In combination, these changes reduce the efficiency of memory consolidation [23].

Treatment of subclinical epileptiform activity as an alternative approach to AD therapy

Given the existing concept of SEA potentiating the pathophysiological mechanisms that contribute to the progression of cognitive impairment in AD, some authors propose therapeutic approaches for treating AD patients with SEA, such as prescribing antiseizure medication (ASM). There is a wide range of ASMs available; however, due to the negative effects of most of them on cognitive functions and memory, the choice of ASM in such patients is limited.

According to numerous studies on the effects of ASMs on cognitive functions in patients with epilepsy, some drugs exhibit a so-called “pro-cognitive” effect. Levetiracetam is one example of such drugs. Due to the potentially beneficial effects of levetiracetam on cognitive functions, most studies aimed at treating SEA and epilepsy in AD patients focus on this particular medication [14][39–44]. Experiments showed that levetiracetam modulates neuronal hyperexcitability, reduces the number of amyloid plaques, and regulates neurotrophic factors [39, 45]. It is known that in AD patients with epileptiform activity, cognitive functions deteriorate faster than in those without such activity. For instance, Vossel et al. studied the effects of levetiracetam on various domains of cognitive function in a group of 34 participants with AD. The analysis showed that in the group of patients with seizures or SEA, the use of levetiracetam led to positive dynamics in tests of executive function and visuospatial memory [46].

Lamotrigine, which has no negative effect on cognitive functions, may also be considered for use in patients with AD and SEA [12][45][47–50]. Lamotrigine prevents the accumulation of extracellular β-amyloid, suppresses glutamate excitotoxicity, thereby exerting neuroprotective properties [51, 52]. A study by Tekin et al. of AD patients without epilepsy showed that the use of lamotrigine at a dose of 300 mg/day for eight weeks had a positive effect on cognitive indicators (in performing tasks on recognition and naming of objects and matching names with objects) and mood [52]. However, there are currently no clear clinical guidelines for prescribing antiseizure therapy to AD patients with SEA having no seizures, which requires further study.

The conducted review indicates the high clinical significance of performing electroencephalography and detecting epileptiform activity in patients with Alzheimer’s disease due to its potential negative impact on the progression of cognitive impairment and increased risk of epileptic seizures in such patients. The frequency of SEA in AD patients can vary (2–54%) depending on the duration of EEG recording, sleep inclusion, and the use of additional techniques (MEG, foramen ovale electrodes). Most literature data emphasize a higher incidence of SEA in AD patients compared to those with other types of dementia or healthy individuals of the corresponding age. Pathophysiological mechanisms highlight common etiopathogenetic links in the progression of AD and the formation of neuronal hyperexcitability, which is associated with the appearance of epileptiform activity on EEG. The use of ASM for SEA therapy may become a new treatment strategy for AD patients, not only as a means of preventing epileptic seizures but also in the treatment of cognitive impairment. However, the advisability of treating subclinical epileptiform activity in Alzheimer’s disease patients remains a subject for further investigation.