Abstract
Objective
The aging Danish population is associated with increasing morbidity and healthcare utilization. This study examines age‐related differences in resource utilization and diagnostic outcomes between older (65–95 years) and younger adults (18–64 years) evaluated for suspected epilepsy, including clinical management, diagnostic yield, and final diagnoses.
Methods
This single‐center retrospective cohort study used electronic health record registry data from adults referred for epilepsy evaluation at Aalborg University Hospital between April 1, 2022, and January 8, 2024, with 1‐year follow‐up. Patients were stratified into different age groups and the relative risk (RR) of receiving an epilepsy diagnosis was modeled using a modified Poisson regression method, with age as the primary predictor. The diagnostic yield of magnetic resonance/computed tomography (MR/CT) and the sensitivity and specificity of electroencephalography (EEG) were compared using the chi‐square test and Fisher's exact test.
Results
Of 530 patients evaluated, 30% received an epilepsy diagnosis and 17% were diagnosed with a first unprovoked seizure. International Classification of Diseases, 10th Revision (ICD‐10) codes varied with age, and the RR for epilepsy was significantly higher in older adults 65–95 years of age compared with both adults 18–64 years of age and those 40–64 years of age. This association remained unchanged after adjustment for sex. Furthermore, 97.7% underwent neuroimaging, and 92.6% underwent an EEG. The diagnostic yield of EEG and neuroimaging (magnetic resonance imaging [MRI] and CT) did not differ significantly between the two age groups.
Significance
Older adults referred for epilepsy evaluation were more likely to receive an epilepsy diagnosis. More than half of patients initially thought to have epilepsy, or a first unprovoked seizure, were ultimately assigned a diagnosis other than epilepsy after comprehensive clinical assessment. These findings underscore the importance of careful, age‐sensitive diagnostic evaluation, and highlight the need for further research to elucidate age‐related diagnostic patterns and refine inclusion criteria and workup strategies in first‐seizure clinics.
Keywords: EEG, epilepsy, first seizure clinic, neuroimaging, older adults, seizure age
Key points.
In a first seizure clinic, approximately half of all evaluated patients were ultimately assigned a diagnosis other than epilepsy.
Older adults of 65–95 years of age were more likely to receive an epilepsy diagnosis than adults 18–64, with a relative risk of 1.32.
Neuroimaging and electroencephalography (EEG) were applied widely; among patients diagnosed with epilepsy, 97.7% underwent imaging and 92.6% had at least one EEG study.
The diagnostic yield of magnetic resonance imaging (MRI), computed tomography (CT), and EEG was comparable across age groups.
Implementing early specialist assessments and optimizing referral pathways for EEG and neuroimaging could enhance resource utilization.
1. INTRODUCTION
Epilepsy affects ~65 million people worldwide, with a prevalence of 0.6%–0.7% in Denmark. 1 According to the International League Against Epilepsy's (ILAE) definition of epilepsy, a diagnosis can be made when a person has had either two or more unprovoked seizures occurring more than 24 h apart; one unprovoked seizure with a probability of seizure reoccurrence over 60% during the next 10 years, typically based on findings from neuroimaging or interictal electroencephalography (EEG); or an epilepsy syndrome. 2 , 3 Seizures are classified by onset as focal, generalized, or unknown 4 and usually last 15–120 s, 5 and can progress from focal to bilateral tonic–clonic seizures.
Depending on epilepsy type, onset, and progression, clinical manifestations vary from subtle non‐motor seizures with preserved consciousness to absence seizures, focal seizures with impaired consciousness, and motor seizures involving parts of the body or, in generalized seizures, presenting as bilateral tonic–clonic seizures. 4 , 6 These diverse presentations pose diagnostic challenges for clinicians globally.
Epilepsy imposes high societal and healthcare costs, including hospitalizations, diagnostics, and treatment. 7 Yet, diagnosis remains challenging, with misdiagnosis rates of 16%–26% in population‐based studies. 8 , 9 Optimizing diagnostic pathways and patient selection is essential for reducing unnecessary investigations, minimizing waiting times, and improving resource use.
To address these challenges, many centers have implemented specialized First Seizure Clinics (FSCs). These clinics follow diagnostic protocols that support systematic evaluation, aiming to reduce misdiagnosis, expedite appropriate treatment, and minimize unnecessary investigations. Similarly, the Epilepsy Clinic at Aalborg University Hospital, Denmark, serves as a regional referral center for seizure assessment, primarily receiving patients from general practitioners, private neurologists, and emergency departments.
An epilepsy diagnosis relies heavily on clinical and history assessment, supported by EEG to detect epileptiform abnormalities and neuroimaging (computed tomography [CT] and magnetic resonance imaging [MRI]) to identify epileptogenic lesions. Thus, in FSCs, diagnostic uncertainty remains high, with previous studies reporting that 39%–50% of referred patients ultimately receive a non‐epilepsy diagnosis. 10 , 11 , 12
Demographic shifts toward older age groups in Denmark have increased diagnostic complexity in clinical practice, 13 , 14 including the diagnostic work‐up of elderly patients presenting with suspected seizures. This trend raises concerns about diagnostic efficiency, as older individuals often present with less specific symptoms, such as confusion or syncope, which may be accompanied by convulsions. Age‐related factors may further complicate evaluation, as the etiology of seizures and the diagnostic yield of EEG and imaging vary with age. Notably, interictal epileptiform abnormalities have also been reported in older adults without epilepsy. 15 These factors contribute to diagnostic complexity and may lead to unnecessary or inconclusive investigations. Given the resource‐demanding nature of EEG and neuroimaging, there is a need for optimized diagnostic strategies for patients referred for evaluation of suspected seizures.
Although the diagnostic utility of EEG and imaging is well established in epilepsy, fewer studies have evaluated their performance in undifferentiated patient populations referred for suspected seizures. 12 FSCs often manage patients with unclear clinical presentations, and it remains unclear whether diagnostic strategies perform consistently across age groups in this context.
To address these diagnostic challenges, we investigated the epidemiology and diagnostic evaluation of adults accepted for assessment following a suspected first seizure, focusing on the diagnostic performance of MRI, CT, and EEG across age groups. We hypothesize that both the likelihood of an epilepsy diagnosis and the diagnostic yield of these modalities vary by age.
2. METHODS
2.1. Clinical setup
The Epilepsy Clinic at Aalborg University Hospital serves ~600 000 individuals. As a public teaching hospital, it is staffed by certified neurologists, neurology residents, and interns under supervision. The mean waiting time for the first clinical evaluation at the epilepsy clinic was 73.2 days (ranging from 8 to 497 days). Those requiring advanced diagnostics or surgical intervention are referred to other hospitals with higher specialization levels, including epilepsy surgery centers.
2.2. Study design and participants
A single‐center retrospective cohort study was conducted at the adult Epilepsy Clinic at Aalborg University Hospital between April 1, 2022, and January 8, 2024.
Patients 18 years of age or older scheduled for their first clinical evaluation at the epilepsy clinic due to one or more suspected seizures were eligible for inclusion. Patients who did not attend the first visit scheduled, as well as adolescents referred from pediatric care with a pre‐established epilepsy diagnosis, were excluded. Adults receiving anti‐seizure medication before the visit were included, if anti‐seizure medication was initiated in an emergency department.
Patients were referred to the clinic by general practitioners, private neurology practices, hospital departments, or other qualified healthcare providers when epileptic seizures were suspected. All referrals were reviewed by a senior neurologist, and patients were accepted for evaluation if the referral was considered appropriate. During the study period, a routine EEG recording was ordered for all accepted referrals prior to the first outpatient visit.
All data for the first visit refer to the patient's first in‐person assessment at the Epilepsy Clinic. During this visit, a neurologist obtained a detailed clinical history, including witness descriptions when available, reviewed prior investigations such as EEG and neuroimaging, and performed a neurological examination to guide the subsequent diagnostic evaluation.
2.3. Data resources
Data were retrieved from the electronic medical record system at Aalborg University Hospital called Nord Elektronisk Patient Journal "NordEPJ" and subsequently entered into the secure electronic data capture tool Research Electronic Data Capture (REDCap).
2.4. Patient characteristics
Data retrieved included referral date, clinical evaluation, gender, age at first visit, and diagnosis International Classification of Diseases, 10th Revision (ICD‐10) code at the first visit and 3‐ and 12‐month follow‐ups. The timing and results of diagnostic procedures (CT, MRI, and routine EEG recordings) were also collected, as was evaluation in an epilepsy monitoring unit (EMU).
2.5. Diagnosis codes
Diagnosis codes were assigned according to the ICD‐10, and documented at the first visit and at the 3‐ and 12‐month follow‐ups.
If the clinician at the epilepsy clinic did not assign an updated diagnostic code; the referral diagnosis from the referring provider was recorded instead. In cases where the medical record indicated an epilepsy diagnosis and the initiation of anti‐seizure medication, but a correct ICD‐10 (DG40.X) code of epilepsy had not been registered, the appropriate code was applied, retrospectively, to both the study dataset and the medical record. The categorization of diagnosis codes is found in Appendix 1.
2.6. Follow‐up and study outcome
The index date was the first visit to the epilepsy clinic. Patients were followed for 1 year or until death, whichever occurred first. The primary study outcome was the probability of receiving an epilepsy diagnosis in relation to age. Secondary outcomes included the use of MRI, CT, and EEG, as well as their diagnostic yield and timing of EEG testing.
Brain imaging (MRI and/or CT) findings were categorized into three groups: (1) no abnormalities, (2) abnormalities of uncertain or no relevance to an epilepsy diagnosis, and (3) epileptogenic lesion(s). Epileptogenic lesions included scarring after stroke or trauma, tumors, vascular malformations, gliosis, hippocampal or mesial temporal sclerosis, congenital malformations, inflammatory, infectious, or immune‐mediated changes, and metabolic or genetic abnormalities—provided that these were considered epileptogenic by the treating physician and documented in the patient's medical record. 16
The timing of EEG examinations was assessed by recording the interval between EEG and (I) referral date, (II) first outpatient assessment, and (III) the most recent documented seizure prior to EEG, when available. The corresponding dates were obtained from a review of all medical records.
EEG results were categorized as normal, abnormal, or with epileptiform discharges, defined as EEG summaries indicating spikes, sharp waves, spike‐and‐slow‐wave complexes or sharp‐wave‐and‐slow‐wave complexes, polyspike‐wave complexes, and generalized polyspikes. 17
EEG studies were generally performed as routine EEG recordings, and patients were instructed to arrive sleep deprived. When appropriate, hyperventilation and photic stimulation were performed as activation procedures.
EEG results were classified as follows: true positives were defined as EEG recordings showing epileptiform discharges in patients ultimately diagnosed with epilepsy, whereas false positives referred to epileptiform discharges in patients without epilepsy. Normal EEG recordings or those with non‐epileptiform abnormalities were categorized as false negatives in patients with epilepsy and as true negatives in patients without epilepsy.
2.7. Statistical analyses
Statistical analyses were conducted using STATA and R v.4.4.2 (R Foundation for Statistical Computing).
Patient characteristics were described using counts and proportions with corresponding 95% confidence intervals (CIs). Proportions were calculated within each age and diagnostic subgroup. Comparisons between groups were performed using the chi‐square test or Fisher's exact test, depending on group size.
To estimate the relative risk (RR) of receiving an epilepsy diagnosis according to age group, we used a modified Poisson regression with a log link and robust variance estimation. In the primary analysis, patients 65–95 years of age were compared with those 18–64 years of age. In an additional analysis, patients were stratified into three age groups (18–39, 40–64, and ≥65 years), and RRs with 95% CIs were estimated using the same approach. All analyses were further adjusted for sex.
Age was also modeled as a continuous variable using restricted cubic splines with knots placed at the 10th, 50th, and 90th percentiles of the age distribution. For graphical presentation, the spline curve was centered at age 40 years (RR = 1).
The diagnostic pathway and test outcomes were analyzed as counts and proportions of CT and MRI scans performed with corresponding 95% CIs. The study population was stratified into two age groups (18–64 years and ≥65 years), and differences in diagnostic outcomes were assessed based on epilepsy vs non‐epilepsy diagnoses. Statistical comparisons were conducted using the chi‐square or Fisher's exact test, depending on group sizes.
The timing of EEG testing was determined by calculating the mean intervals, with standard deviations (SDs) computed for each time measure. EEG sensitivity was defined as true positives / (true positives + false negatives), and specificity as true negatives / (true negatives + false positives). Confidence intervals were estimated using the Clopper–Pearson exact method. The sensitivity and specificity of routine EEG recordings were compared between the two age groups using the chi‐square test.
The reporting of this study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. 18
3. RESULTS
3.1. Cohort characteristics
A total of 530 patients were included. The mean age was 51.8 years (range: 18–95), and 300 (56.6%, 95% CI 52.4–60.8) were male. A total of 158 patients (29.8%, 95% CI 26.1–33.8) were diagnosed with epilepsy within a year, whereas 90 patients (17%) were considered to have had a first‐time unprovoked seizure.
Of the 158 patients diagnosed with epilepsy within a year, most (111, 70.3%) were diagnosed at the first visit. In addition, 33 (20.9%) were diagnosed within 3 months, and 14 (8.9%) between 3 and 12 months. In only three cases, the epilepsy diagnosis was withdrawn at the 1‐year follow‐up. Basic characteristics and diagnostic tests performed for each group are listed in Table 1.
TABLE 1.
Basic characteristics and diagnostic tests performed grouped by diagnosis of epilepsy.
| Patients with epilepsy | Patients without epilepsy | All patients | |
|---|---|---|---|
| Number of patients |
158 29.8% (95% CI 26.1–33.8%) |
372 (70.2%, 95% CI 66.2–73.9) |
530 (100%) |
| Mean age (range) | 53.9 (19–90) | 50.9 (18–95) | 51.8 (18–95) |
| Age groups |
18–64: 94 59.5% (95% CI 51.7–66.8) |
18–64: 256 68.8% (95% CI 63.9–73.3) |
18–64: 350 66% (95% CI 61.9–69.9) |
|
65–95: 64 40.5% (95% CI 33.2–48.3) |
65–95: 116 31.2% (95% CI 26.7–36.1) |
65–95: 180 34% (95% CI 30.1–38.1) |
|
| Gender |
Male: 98 62% (95% CI 54.3–69.2) |
Male: 202 54.3% (95% CI 49.2–59.3) |
Male: 300 56.6% (95% CI 52.4–60.8) |
|
Female: 60 38% (95% CI 30.8–45.7) |
Female: 170 45.7% (95% CI 40.7–50.8) |
Female: 230 43.4% (95% CI 39.2–47.6) |
|
|
Magnetic Resonance Imaging within follow‐up |
131 82.9% (95% CI 76.3–88.0) |
277 74.5% (95% CI 69.8–78.6) |
408 77% (95% CI 73.2–80.4) |
|
Computed Tomography within follow‐up |
97 61.4% (95% CI 53.6–68.6) |
173 46.5% (95% CI 41.5–51.6) |
270 51.1% (95% CI 46.7–55.2) |
| Neuroimaging not performed within follow‐up |
2 1.3% (95% CI 0.2–4.5) |
10 2.7% (95% CI 1.3–4.9) |
12 2.3% (95% CI 1.2–3.9) |
|
Electroencephalography within follow‐up |
150 94.9% (95% CI 90.3–97.4) |
340 91.4% (95% CI 88.1–93.8) |
491 92.6% (95% CI 90.1–94.6) |
| Long‐term electroencephalography monitoring |
2 1.3% (95% CI 0.15–4.5) |
10 2.7% (95% CI 1.3–4.9) |
12 2.3% (95% CI 1.2–3.9) |
3.2. Diagnosis distribution by age
The distribution of epilepsy and non‐epilepsy diagnoses varied by age (Figure 1). More substance abuse‐related seizures were observed among patients 18–64 years of age (5.7%, 95% CI 3.7–8.7; p < .05), whereas epilepsy diagnoses were more common in those 65–95 years of age (35.6%, 95% CI 28.9–42.8; p < .05) (Table 2).
FIGURE 1.

Relative risk of epilepsy diagnosis according to age modeled using Poisson regression with restricted cubic splines. Knots were placed at the 10th, 50th, and 90th percentiles of the age distribution. The curve is centered at age 40 (relative risk [RR] = 1), and the shaded area represents the 95% confidence interval.
TABLE 2.
Age‐based distribution of diagnoses.
| Age 18–64 | Age 65–95 | p‐value | |
|---|---|---|---|
| Epilepsy | 94 (36.9%, 95% CI 22.5–31.7) | 64 (35.6%, 95% CI 28.9–42.8) | <.05 |
|
Substance abuse |
20 (5.7%, 95% CI 3.7–8.7) | 1 (0.6%, 95% CI 0.1–3.1) | <.05 |
| Functional seizure | 12 (3.4%, 95% CI 2.0–5.9) | 4 (2.2%, 95% CI 0.9–5.6) | NS |
| Syncope | 86 (24.6%, 95% CI 20.4–29.3) | 34 (18.9%, 95% CI 13.8–25.2) | NS |
| Unprovoked seizure | 57 (16.3%, 95% CI 12.8–20.5) | 33 (18.3%, 95% CI 13.4–24.6) | NS |
| Unspecified | 63 (18%, 95% CI 14.3–22.4) | 30 (16.7%, 95% CI 11.9–22.8) | NS |
| Other | 18 (5.1%, 95% CI 3.3–8.0) | 14 (7.8%, 95% CI 4.7–12.6) | NS |
Note: The p‐values are calculated for differences between age groups.
Abbreviations: CI: Confidence Interval. NS: non‐significant difference as compared by chi‐square or Fisher's exact tests.
Among younger adults, structural etiologies were dominated by malformations of cortical development, post‐traumatic lesions, and post‐stroke changes. In contrast, older adults showed a markedly higher proportion of post‐stroke etiologies. Tumors and vascular malformations occurred infrequently in both age groups. An unknown etiology was observed in 38.3% of younger adults and 43.8% of older adults, and this difference was not statistically significant (odds ratio [OR] 0.8; p = .51).
The majority of patients had 1‐year follow‐up data available; 18 (3.4%) died within the study period. Regarding re‐admission during follow‐up, we found no significant difference in the risk of seizure‐related readmission within the first 3 months after the index event between younger and older adults (12.8% vs 12.5%; OR 1.02, 95% CI 0.36–3.09; p = 1.0). In contrast, patients 65 years of age or older had a higher risk of readmission in the 3‐ to 12‐month period (OR 2.22, 95% CI 0.96–5.20; p < .049). There was no difference in the distribution of admission etiologies between younger and older adults (Fisher's exact test, p = .15). Although younger patients more often presented with Idopathic Generalized Epilepsy (IGE), developmental abnormalities, or temporal lobe–related etiologies, and older adults more frequently with post‐acute lesions, these differences did not reach statistical significance.
Using modified Poisson regression, patients 65–95 years of age were more likely to receive an epilepsy diagnosis than those aged 18–64 years (RR 1.32, 95% CI 1.02–1.72; p = .036).
When stratifying patients further by age into three groups (18–39, 40–64, and ≥65 years), those ≥65 years of age were more likely to receive an epilepsy diagnosis than those aged 40–64 years of age (RR 1.40, 95% CI 1.01–1.93; p = .042). No significant difference was observed between younger adults 18–39 years of age and the middle‐aged group (RR 1.11, 95% CI 0.79–1.57; p = .553). These age‐related increases in RR remained unchanged after adjustment for sex (Table 3).
TABLE 3.
Relative risks (RRs) for epilepsy diagnosis across age groups estimated using modified Poisson regression.
| Compared age groups | Relative risk (95% Confidence Intervals) | p‐value |
|---|---|---|
| Age ≥65 vs age 18–64 | 1.32 (1.02–1.72) | .036 |
| Age ≥65 vs age 40–64 | 1.4 (1.01–1.93) | .042 |
| Age 18–39 vs age 40–64 | 1.11 (0.79–1.57) | .553 |
| Age ≥65 vs age 18–39 | 1.26 (0.93–1.71) | .141 |
| Adjusted relative risk (95% Confidence Intervals) | Adjusted p‐value | |
|---|---|---|
| Age ≥65 vs age 18–64 | 1.3 (1.0–1.69) | .047 |
| Age ≥65 vs age 40–64 | 1.38 (1.0–1.92) | .048 |
| Age 18–39 vs age 40–64 | 1.12 (0.79–1.59) | .513 |
| Age ≥65 vs age 18–39 | 1.23 (0.91–1.67) | .181 |
Note: Unadjusted estimates and sex‐adjusted estimates are presented. Relative risks represent the probability of receiving an epilepsy diagnosis in the compared age groups.
When age was modeled as a continuous variable using restricted cubic splines, the RR of epilepsy diagnosis remained relatively stable across younger and middle adulthood and increased progressively with advancing age (Figure 1). CIs widened at the extremes of the age distribution, reflecting fewer observations in these age groups.
3.3. Utilization of neuroimaging
For diagnostic imaging, 518 (97.7%) of the cohort underwent either a CT scan or MRI of the brain (or both) within 2 years prior to the first visit or during follow‐up (77% MRI, 51.1% CT). Among epilepsy patients, 131 (82.9%, 95% CI 76.3–88.0) received an MRI, and 97 (61.4%, 95% CI 53.6–68.6) had a CT scan during the same timeframe. Neuroradiological lesions were considered epileptogenic in 22.9% (95% CI 15.5–32.0) of CT images and 31.3% (95% CI 24.0–39.7) of MRI images of patients with epilepsy who were scanned. In 13 patients, the epileptogenic lesion was not detected on the initial CT scan but was subsequently identified on MRI, confirming the structural etiology (6 younger and 7 older adults).
Diagnostic test outcomes varied by age between the epilepsy and non‐epilepsy groups. A higher proportion of patients 65–95 years of age, regardless of epilepsy status, had abnormal CT and MRI findings compared to those 18–64 years of age (p < .05). Younger participants underwent CT scans less frequently (p < .05). In contrast, MRI use did not differ significantly between the age groups. In addition, the detection of epileptogenic lesions was similar in both age groups for both CT and MRI (Figure 2).
FIGURE 2.

Computed tomography (CT) and Magnetric Resonance Imaging (MRI) results of patients referred for epilepsy evaluation, categorized by age and the presence of an epilepsy diagnosis.
3.4. The diagnostic utility of EEG
Most patients (n = 491, 92.6%; 95% CI 90.1–94.6) underwent EEG during the follow‐up period, including 150 patients (94.9%; 95% CI 90.3–97.4) diagnosed with epilepsy. Among 45 patients with epileptogenic lesions on imaging, 42 (93%) underwent EEG testing, which did not contribute further to the diagnostic process in most cases. Within the follow‐up period, 12 patients (2.3%, 95% CI 1.2–3.9) were admitted for EMU evaluation.
In the total cohort, the mean interval from referral to EEG was 49.3 days (SD 54.2). The first EEG was performed with a mean of 19.7 days before the initial clinical evaluation (SD 65.8). Information on the timing of the most recent seizure prior to EEG was available for 403 patients, with a mean interval of 48.1 days (SD 53.3) between the last suspected seizure and EEG.
Among patients with epilepsy, the first EEG was performed in most (111 patients, 70.3%) prior to the first visit, whereas 39 (24.7%) underwent their first EEG between the baseline visit and the 3‐month follow‐up. Eight patients (5.1%) did not undergo EEG evaluation during the study period.
EEG was usually performed before the epilepsy diagnosis. This was the case in 80.2% of patients diagnosed at baseline, in all patients diagnosed within 3 months, and in 92.9% of those diagnosed between 3 and 12 months.
Epileptiform discharges on routine EEG recording supported the diagnosis in 33 patients (29.7%) diagnosed at baseline, 20 patients (60.6%) diagnosed within 3 months, and 4 patients (28.6%) diagnosed between 3 and 12 months.
EEG performed after diagnosis was uncommon, occurring in 16.2% of patients diagnosed at baseline and 9.1% of those diagnosed within 3 months, but none of those diagnosed between 3 and 12 months underwent EEG after diagnosis. Four patients (2.5%) presented initially with a negative EEG followed by a later EEG demonstrating epileptiform discharges.
Interictal epileptiform discharges were identified in 38.7% patients (n = 58) with epilepsy who underwent EEG testing during the whole 1‐year follow‐up period, and the overall specificity was 94.4%. Among patients with epilepsy 18–64 years of age, the EEG studies of 39 patients exhibited epileptiform discharges, resulting in a sensitivity of 45.3% (95% CI 34.6–56.5). In contrast, 19 patients 65–95 years of age showed epileptiform discharges on their EEG recordings, yielding a sensitivity of 31.4% (95% CI 19.9–44.3). The specificity of EEG was 96.1% (95% CI 92.6–98.2) for patients 18–64 years of age and 90.8% (95% CI 83.8–95.5) for patients 65–95 years of age. EEG sensitivity and specificity appeared to vary by age, but the differences were not statistically significant (chi‐square test: p = .12, p = .09).
4. DISCUSSION
Our study demonstrates an age‐dependent increase in the risk of receiving an epilepsy diagnosis in an adult cohort accepted for epilepsy evaluation. This is relevant because the increasing proportion of elderly individuals presents a challenge to healthcare systems, particularly in ensuring appropriate referrals and diagnostic assessments. This applies not only at the point of referral to university hospitals but throughout the diagnostic pathway.
Overall, our findings align with other studies at FSCs showing that 39%–50% of patients were ultimately diagnosed with non‐epileptic conditions and seizure mimics. 10 , 11 , 12 , 19 This means that more than half of the referred patients ultimately received a diagnosis other than epilepsy, such as syncope, substance‐related, or functional seizures. In fact, previous studies indicated that up to 13% presented with provoked seizures, 12 but our cohort had only 4%, likely due to rigorous screening before evaluation.
A notable proportion remained undiagnosed after a 1‐year follow‐up. Besides the clinical implications, diagnostic uncertainty and delayed clarification may increase the patient's and relatives' anxiety and delay appropriate care.
Older patients may present with nonspecific symptoms, multiple comorbidities, and seizure mimics such as transient ischemic attack, amyloid spells, cerebral hypoperfusion, and syncope, 14 , 15 , 16 , 17 which probably influence the diagnostic process. However, a recent study found that first seizure diagnoses were more likely than seizure mimics in older age groups, 17 consistent with our findings, where patients 65–95 years of age had a significantly higher risk of epilepsy than those 18–64 years of age.
Additional analyses stratifying patients into three age groups indicated that the association between age and epilepsy diagnosis was primarily driven by patients aged ≥65 years. In multivariate analyses adjusting for sex, this association remained largely unchanged, suggesting that the observed age‐related difference was not confounded by the sex distribution in our cohort. No significant differences were observed between younger and middle‐aged adults.
This finding aligns with the well‐established bimodal distribution of epilepsy incidence, with a marked increase in older adults reported in epidemiological studies, including recent Danish population‐based data. 20 , 21 Referral patterns may also contribute to the observed age gradient. Younger adults referred to FSCs may more frequently present with transient events or functional episodes that are difficult to distinguish from epileptic seizures at the initial clinical evaluation in emergency departments or primary care, whereas older adults often have a higher pretest probability of epilepsy due to underlying structural brain disease. 20
In our cohort, older subjects also showed a markedly higher risk of readmission due to seizures during Months 3–12 following evaluation. This suggests that older patients may experience a more unstable clinical course after the initial presentation, potentially driven by comorbidities, evolving cerebrovascular pathology, or delayed recognition of recurrent seizures. These findings underscore the need for heightened clinical vigilance in the first year after a new‐onset seizure, especially in older adults, including timely diagnostic clarification and expedited follow‐up in a FSC to prevent avoidable readmissions and optimize long‐term management.
We tested whether etiologies differed between age groups. The trends toward age‐related differences in structural etiologies in our cohort may reflect an established epidemiological pattern, with developmental malformations and traumatic injuries being more common among younger adults, and cerebrovascular disease predominating in older adults. These trends, however, were not significant. Of interest, the proportion of patients with an unknown etiology was similar across age groups.
Neuroimaging showed that epileptogenic lesions were detected in 31.3% of MRI scans and 22.9% of CT scans. Our findings match with prior research, including a study of 1013 adults undergoing neuroimaging after their first unprovoked seizure, in which MRI identified epileptogenic lesions in 29% of cases. 22 Similarly, a more recent study reported structural lesions in 30.7% of patients presenting with a first seizure. 23 A study from an FSC in Chicago reported epileptogenic lesions in 24.1% of patients who underwent MRI or CT, 12 which is slightly lower than the 31.3% we observed. However, the effect of age was not analyzed in these studies.
Of note, MRI contributed uniquely to the diagnostic clarification in 13 patients, in whom the epileptogenic lesion was not visible on initial CT but was subsequently identified on MRI. Although these numbers are small, younger adults primarily had cortical dysplasia missed on CT, whereas all seven older adults had small chronic cortical infarcts not detectable on CT. This highlights the importance of high‐quality (3 T) MRI in standardized protocols and consistent radiological assessment in the evaluation of new‐onset seizures. 24
Among patients with MRI‐positive lesions, older adults were more frequently diagnosed with epilepsy at their first visit in the epilepsy clinic. By comparison, some younger adults required a period of observation before a definitive diagnosis was established.
Routine EEG is a key component of the diagnostic evaluation of suspected epilepsy, and most patients in our cohort underwent at least one EEG recording during follow‐up. The sensitivity and specificity observed in our study were comparable to those reported in the ILAE meta‐analysis of EEG as a diagnostic tool in epilepsy. 25
The diagnostic yield of EEG is influenced by several factors, including the interval between seizure and recording, 16 , 26 , 27 , 28 sleep deprivation, 27 , 29 repeated EEG examinations, 25 , 29 and patient age. 15 , 25 , 30 , 31 Although incidental epileptiform discharges have been reported more frequently in older adults, 15 , 30 the diagnostic performance of EEG testing was comparable across age groups in our cohort. A recent population‐based study reported a considerably lower sensitivity (19.2% vs 38.7%) and slightly higher specificity (98.0% vs 94.4%) in patients with suspected new‐onset seizures. 32 The higher sensitivity observed in our cohort may reflect several methodological differences. Unlike Joelsson et al., our analyses were based on the final clinical epilepsy diagnosis following specialist assessment and longitudinal follow‐up rather than seizure recurrence. Furthermore, all patients in our cohort were instructed to arrive sleep deprived, and EEG findings obtained throughout the 1‐year follow‐up period were included in the analysis. However, repeated EEG recordings had limited impact on our estimates, as only four patients had a later positive EEG following an initially negative recording. In addition, inclusion of patients with remote childhood epilepsy or previous epilepsy investigations may have increased the pretest probability of epileptiform EEG abnormalities in our cohort.
Although current guidelines recommend EEG within 24–48 h of a suspected first seizure, 16 , 27 , 28 the mean interval between the most recent suspected seizure and EEG was ~7 weeks in our cohort, reflecting routine outpatient practice. Nonetheless, interictal epileptiform discharges were identified in 38.7% of patients with epilepsy, supporting the continued clinical utility of routine EEG, which aligns with the study by Naganur et al., who reported no significant differences in EEG yield between recordings performed within the first week and those obtained later. 33
During the study period, EEG was performed frequently before the initial outpatient consultation because investigations were often initiated upon referral, while waiting times for specialist evaluation remained relatively long. Although this approach may have accelerated the diagnostic process, it likely also resulted in EEG examinations in patients later found to have seizure mimics or other non‐epileptic conditions. Similarly, 93% of patients with epileptogenic lesions on neuroimaging underwent EEG despite limited additional diagnostic value. Following completion of the study and due to shorter outpatient wait times, our diagnostic workflow has been revised, and EEG is now requested on an individual basis after specialist evaluation of seizure semiology, clinical history, and neuroimaging findings.
Although neuroimaging was widely applied, even in cases where the suspected events were believed to be non‐epileptic, it is more challenging to determine whether too many unnecessary scans were performed. Furthermore, neuroimaging remains vital for adult seizure diagnosis, most specifically to identify structural issues like cortical malformations, cerebrovascular disease, and neoplasia. Clear guidelines for neuroimaging may help ensure timely and consistent diagnostic evaluation. In our region, this has prompted the development of a guideline to improve referral practices and standardize the diagnostic work‐up after a first suspected seizure.
The study's retrospective, observational design introduces certain limitations. It is essential to acknowledge that misclassification and interobserver variability may be present, as physicians with varying levels of epilepsy expertise evaluated the patients. Nonetheless, all residents and interns were supervised by specialists. The risk of misclassification from ICD‐10 coding was minimized through researcher‐led chart reviews and corrections when epilepsy was clearly documented. A potential selection bias related to referral source was not accounted for; prior studies suggest a higher diagnostic yield when patients are referred from emergency departments compared to primary care. 12 Furthermore, the study cohort comprised patients with referrals accepted for neurological evaluation. Data on referrals not accepted for evaluation were not available, as only accepted referrals are registered in a format that permits retrospective data extraction and research use.
As a regional epilepsy clinic, we refer patients to tertiary centers for advanced diagnostics or surgical evaluation. Outcomes of patients referred elsewhere were not included. These limitations may introduce verification bias. Finally, stratifying all patients into age groups reduces statistical power in some of the analyses.
5. CONCLUSION
Age was a significant determinant of diagnostic outcome, with older adults more likely to receive an epilepsy diagnosis. The diagnostic performance of EEG and neuroimaging was comparable across age groups, supporting their continued use irrespective of patient age. These findings underscore the need for sustained diagnostic vigilance in older patients. Although EEG and neuroimaging remain central to the diagnostic work‐up, early specialist assessment and optimized referral pathways may enhance resource utilization and facilitate timely identification of epilepsy and other clinically relevant conditions.
AUTHOR CONTRIBUTIONS
Study conception and design: All authors. Material preparation and data collection: Anne Line Birkmose and Corinna Lackmann. Interpretation of the results: all authors. Statistical analysis: Anne Line Birkmose and the Department of Statistics, Aalborg University Hospital. Preparation of the first draft of the manuscript: Anne Line Birkmose. All authors commented on previous versions of the manuscript and critically revised the paper. All authors have read and approved the final manuscript and agree to be accountable for all aspects of the work.
FUNDING INFORMATION
This research project did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
ETHICS STATEMENT
According to the Danish Health Care Act, the project is classified as a quality assurance study, which does not require any additional ethical assessment according to Danish legislation. The study is conducted in compliance with the General Data Protection Regulation and is a part of North Denmark Region's record of processing activities (identification number: K2024‐148). Due to the observational design of this study, informed consent forms were not needed from participants. Substitute consent for accessing patient records was given by the hospital management, and the study was reported to the Danish Data Protection Agency (“Datatilsynet”) through registration at the North Denmark Region.
ETHICAL PUBLICATION STATEMENT
We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
DESCRIPTION FOR SOCIAL MEDIA AND ARTICLE PROMOTION
Should patient age influence the diagnostic work‐up in first seizure clinics? We examined age‐related diagnoses and diagnostic pathways.
ACKNOWLEDGMENTS
The authors have nothing to report.
APPENDIX 1. Categorization of ICD‐10 Diagnoses Codes
Epilepsy:
DG40, DG400, DG401, DG402, DG402D, DG403, DG403F, DG406, DG408, DG409.
Substance abuse:
DF1, DF11, DF100, DF1006, DF101, DF102, DF1026, DF103, DF1031, DF110, DF1106, DF141, DF1920, DF1931.
Functional Seizures:
DF445, DF449, DR568G, DR688A9, DR688A9B5, DS063.
Syncope:
DR55, DR559, DR559A, DR559B.
Unprovoked Seizures:
DR252A, DR258, DR258A, DR568, DR568D, DR568E.
Unspecified:
DZ, DZ03, DZ033, DZ033A, DZ033C, DZ039.
Other:
DD137A, DE162, DG454, DG459, DG4744, DI350, DI471L, DI495, DI672, DI678A, DI690, DI694, DI951, DI952, DR119, DR202, DR251, DR296, DR418, DR429, DR478, DR539, DR410.
DATA AVAILABILITY STATEMENT
The dataset used in this study cannot be made publicly available due to personally identifiable information, according to the Danish law on personal data. All data in this project are original, and no reproduced data were used.
REFERENCES
- 1. Christensen J, Vestergaard M, Pedersen MG, Pedersen CB, Olsen J, Sidenius P. Incidence and prevalence of epilepsy in Denmark. Epilepsy Res. 2007;76:60–65. [DOI] [PubMed] [Google Scholar]
- 2. Fisher RS, Acevedo C, Arzimanoglou A, Bogacz A, Cross JH, Elger CE, et al. ILAE Official Report: A practical clinical definition of epilepsy. Epilepsia. 2014;55:475–482. [DOI] [PubMed] [Google Scholar]
- 3. International league against epilepsy . The 2014 definition of epilepsy: a perspective for patients and caregivers. International League Against Epilepsy, 2024.
- 4. Scheffer IE, Berkovic S, Capovilla G, Connolly MB, French J, Guilhoto L, et al. ILAE classification of the epilepsies position paper of the ILAE Commission for Classification and Terminology. Epilepsia. 2017;58:512–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Meritam Larsen P, Wüstenhagen S, Terney D, Gardella E, Aurlien H, Beniczky S. Duration of epileptic seizure types: a data‐driven approach. Epilepsia. 2023;64:469–478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Fisher RS, Cross JH, D'Souza C, French JA, Haut SR, Higurashi N, et al. Instruction manual for the ILAE 2017 operational classification of seizure types. Epilepsia. 2017;58:531–542. [DOI] [PubMed] [Google Scholar]
- 7. Christensen J, Dreier JW, Sun Y, Linehan C, Tomson T, Marson A, et al. Estimates of epilepsy prevalence, psychiatric co‐morbidity and cost. Seizure. 2023;107:162–171. [DOI] [PubMed] [Google Scholar]
- 8. Smith D, Defalla BA, Chadwick DW. The misdiagnosis of epilepsy and the management of refractory epilepsy in a specialist clinic. QJM. 1999;92:15–23. [DOI] [PubMed] [Google Scholar]
- 9. Leach JP, Lauder R, Nicolson A, Smith DF. Epilepsy in the UK: misdiagnosis, mistreatment, and undertreatment?: the Wrexham area epilepsy project. Seizure. 2005;14:514–520. [DOI] [PubMed] [Google Scholar]
- 10. Mc Manus E, Gilbertson L, Timmings P, Lynch C, Asztely F. Long‐term outcome of 200 patients referred to a first seizure clinic. Acta Neurol Scand. 2021;143:140–145. [DOI] [PubMed] [Google Scholar]
- 11. Jackson A, Teo L, Seneviratne U. Challenges in the first seizure clinic for adult patients with epilepsy. Epileptic Disord. 2016;18:305–314. [DOI] [PubMed] [Google Scholar]
- 12. Nathan CL, Kinsinger M, Giammarco‐Hock A, Cunningham E, Gerard E, Schuele S, et al. Clinical characteristics of a first seizure evaluation clinic in an academic center. Epilepsy Behav. 2024;161:110054. [DOI] [PubMed] [Google Scholar]
- 13. Oksuzyan A, Höhn A, Krabbe Pedersen J, Rau R, Lindahl‐Jacobsen R, Christensen K. Preparing for the future: the changing demographic composition of hospital patients in Denmark between 2013 and 2050. PLoS One. 2020;15:e0238912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kallestrup‐Lamb M, Marin AOK, Menon S, Søgaard J. Aging populations and expenditures on health. The Journal of the Economics of Ageing. 2024;29:100518. [Google Scholar]
- 15. Pirgit ML, Beniczky S. EEG and semiology in the elderly: a systematic review. Seizure. 2024;128:90–121. [DOI] [PubMed] [Google Scholar]
- 16. King MA, Newton MR, Jackson GD, Fitt GJ, Mitchell LA, Silvapulle MJ, et al. Epileptology of the first‐seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients. Lancet. 1998;352:1007–1011. [DOI] [PubMed] [Google Scholar]
- 17. Chen H, Koubeissi MZ. Electroencephalography in epilepsy evaluation. CONTINUUM: Lifelong Learning in Neurology. 2019;25:431–453. [DOI] [PubMed] [Google Scholar]
- 18. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335:806–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Maloney EM, Chaila E, O'Reilly ÉJ, Costello DJ. Incidence of first seizures, epilepsy, and seizure mimics in a geographically defined area. Neurology. 2020;95:e576–e590. [DOI] [PubMed] [Google Scholar]
- 20. Christensen J, Trabjerg BB, Wagner RG, Newton CR, Kwon C, Aaberg KM, et al. Prevalence of epilepsy: a population‐based cohort study in Denmark with comparison to global burden of disease (GBD) prevalence estimates. J Neurol Neurosurg Psychiatry. 2025;96:480–488. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Beghi E, Giussani G. Aging and the epidemiology of epilepsy. Neuroepidemiology. 2018;51:216–223. [DOI] [PubMed] [Google Scholar]
- 22. Ho K, Lawn N, Bynevelt M, Lee J, Dunne J. Neuroimaging of first‐ever seizure. Neurol Clin Pract. 2013;3:398–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Carpio A, Salgado C, DiCapua D, Fleury A, Suastegui R, Giagante B, et al. Causes and prognosis of adults experiencing a first seizure in adulthood: a pilot cohort study conducted in five countries in Latin America. Epilepsia Open. 2024;9:776–784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Scheepers B, Clough P, Pickles C. The misdiagnosis of epilepsy: findings of a population study. Seizure. 1998;7:403–406. [DOI] [PubMed] [Google Scholar]
- 25. Baldin E, Hauser WA, Buchhalter JR, Hesdorffer DC, Ottman R. Yield of epileptiform electroencephalogram abnormalities in incident unprovoked seizures: a population‐based study. Epilepsia. 2014;55:1389–1398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Llauradó A, Santamarina E, Fonseca E, Olivé M, Requena M, Sueiras M, et al. How soon should urgent EEG be performed following a first epileptic seizure? Epilepsy Behav. 2020;111:107315. [DOI] [PubMed] [Google Scholar]
- 27. Schreiner A, Pohlmann‐Eden B. Value of the early electroencephalogram after a first unprovoked seizure. Clin Electroencephalogr. 2003;34:140–144. [DOI] [PubMed] [Google Scholar]
- 28. Sofat P, Teter B, Kavak KS, Gupta R, Li P. Time interval providing highest yield for initial EEG in patients with new onset seizures. Epilepsy Res. 2016;127:229–232. [DOI] [PubMed] [Google Scholar]
- 29. Geut I, Weenink S, Knottnerus ILH, van Putten MJM. Detecting interictal discharges in first seizure patients: ambulatory EEG or EEG after sleep deprivation? Seizure. 2017;51:52–54. [DOI] [PubMed] [Google Scholar]
- 30. Hughes JR, Zialcita ML. EEG in the elderly: seizures vs. syncope. Clin Electroencephalogr. 2000;31:131–137. [DOI] [PubMed] [Google Scholar]
- 31. Neufeld MY, Chistik V, Vishne TH, Korczyn AD. The diagnostic aid of routine EEG findings in patients presenting with a presumed first‐ever unprovoked seizure. Epilepsy Res. 2000;42:197–202. [DOI] [PubMed] [Google Scholar]
- 32. Joelsson S, Andersson K, Brannefors P, Klemetz S, Gärdesmed L, Wennberg E, et al. Diagnostic value of EEG after a first unprovoked seizure in adults—a population‐based study. Epilepsy Behav. 2025;162:110151. [DOI] [PubMed] [Google Scholar]
- 33. Naganur V, Tse J, Muthusamy J, Snell S, Wang S, Seneviratne U. Does the interval between the first unprovoked seizure and EEG influence the diagnostic yield? Epilepsy Behav. 2025;165:110311. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The dataset used in this study cannot be made publicly available due to personally identifiable information, according to the Danish law on personal data. All data in this project are original, and no reproduced data were used.
