Epilepsy is a common chronic neurological disorder characterized by a persistent predisposition to generate epileptic seizures and by the neurobiological, cognitive, psychological, and social consequences associated with the condition. It affects millions of people worldwide and represents an important cause of neurological morbidity, particularly in low- and middle-income countries (LMICs), where limitations in healthcare infrastructure, diagnostic services, treatment availability, and continuity of care may contribute to an increased disease burden [1]. The burden of epilepsy in Pakistan has been documented in population-based, hospital-based, and multicentre studies. Previous investigations have reported substantial variation in the demographic and clinical characteristics of affected individuals, with male predominance and a considerable proportion of patients presenting during childhood reported in several settings [2]. Studies from different regions of Pakistan have also highlighted the contribution of epilepsy to the overall burden of neurological disorders and the challenges associated with its diagnosis and management. However, available evidence is unevenly distributed geographically, and data from Balochistan, particularly from tertiary-care neurology services in Quetta, remain limited [3].  Shakir Ullah et al. conducted a study in 2018 in Khyber Pakhtunkhwa (KP) on 315 patients, reaffirming the correlation of epilepsy with male gender (64.39%), urban residence (60.1%) and poor economic status (88.6%) [4]. In 2021–2022 paediatric neurology clinic in Lahore study of 621 children (1 month to 14 years of age) identified that 31.4% of children were affected with epilepsy and was the most frequent neurological disorder in children. It was highly correlated with male gender, rural habitation, perinatal asphyxia, meningitis, malnutrition, and consanguinity (64.1%) [5]. During 2017-2019, a multicentre study in adult neurology clinics of Pakistan compared 28,845 patients and identified that epilepsy was responsible for approximately 12.5% of all the neurological diagnoses and ranked third after stroke and headache disorders. Among the epilepsy patients, around 10% had a disability that ranged from moderate to severe [6]. In 2023, a clinical study confirmed these trends.  Male predominance (58.65%) and most of the patients belonged to the 5–18 years age group. Suspected seizure disorder (78.32%) was the most common reason for clinical referral followed by subclinical seizures (7.38%) [7-9]. Saleem et al., in 2017 highlighted epilepsy as one of the contributor to the burden of neurological disorders in Quetta, Baluchistan [10]

Despite the clinical importance of age-related differences in EEG findings, limited information is available regarding the distribution of recorded EEG abnormalities among patients with epilepsy receiving neurological care in Balochistan, Pakistan. In particular, there is a need to describe whether specific EEG-pattern indicators differ across age groups within a tertiary-care hospital population. Therefore, the present study aimed to characterize the demographic profile and recorded EEG-pattern indicators among patients with epilepsy attending Bolan Medical Complex Hospital, Quetta, between January and December 2024, with particular emphasis on the association between age group and focal-onset, generalized-onset, generalized-slowing, burst-pattern, and bilateral-discharge indicators. Secondary analyses examined whether these EEG indicators were associated with sex, reported episode-frequency category, and district of residence.

.1 Study Design

A retrospective observational study was conducted to describe the demographic characteristics and recorded electroencephalographic (EEG) patterns of patients diagnosed with epilepsy, with particular emphasis on age-related differences in EEG findings. The study was based on existing hospital records and did not involve prospective patient recruitment or intervention.

2.2 Study Setting

The study was conducted at the Department of Neurology, Bolan Medical Complex Hospital (BMCH), Quetta, Balochistan, Pakistan. The department provides neurological assessment and care to patients attending the hospital as inpatients and outpatients.

2.3 Study Duration

The study covered the period from January 2024 to December 2024. Hospital records corresponding to this period were reviewed for data extraction and analysis.

2.4 Study Population

The study population comprised records of patients diagnosed with epilepsy who attended or were admitted to the Department of Neurology at BMCH, Quetta, during the study period. A total of 276 records were included in the supplied analytical dataset.

2.5 Eligibility Criteria

Patients should be considered eligible only when the original clinical record documented epilepsy according to the diagnostic practice used by the treating neurology team. Records of isolated acute symptomatic, provoked, febrile, or neonatal seizures should be excluded if these were not considered epilepsy by the treating team. The final exclusion criteria should be aligned with the original medical-record review and ethics-approved study protocol and should not be added retrospectively if they were not applied during data extraction.

Inclusion criteria

• Hospital records documenting a diagnosis of epilepsy.
• Records of patients who attended the neurology department as inpatients or outpatients during the study period.
• Records containing the demographic and EEG-related variables required for the analysis.

Exclusion criteria

No specific exclusion criteria were documented in the information provided. Therefore, exclusion criteria should be added only if they were defined and applied during the original study. Records should not be described as excluded for missing data, because no system-missing values were identified in the supplied dataset.

2.6 Sample Size and Sampling Technique

The study included 276 records available in the analytical dataset. The sample consisted of eligible hospital records identified for the study period. If all eligible records from January to December 2024 were reviewed, the sampling approach may be described as total enumeration or a census of eligible records. This terminology should be used only after confirming that all eligible records were included. No prospective sample-size calculation was reported in the information provided.

2.7 Data Collection Procedure

Data were extracted retrospectively from existing inpatient and outpatient hospital records. The recorded variables included demographic characteristics, including sex, age category, and district of residence; reported episode-frequency categories; a recorded disease-code variable; and five binary EEG-pattern indicators.

2.8 Statistical Analysis

Data were analyzed using IBM Statistical Package for the Social Sciences (SPSS), version 20, together with additional statistical procedures for permutation testing and multiple-comparison adjustment. Categorical variables were summarized using frequencies and percentages. Because age and episode frequency were recorded as categorical variables, their numerical codes were treated as nominal or ordinal categories rather than continuous measurements.

Age was grouped into four analytical strata: <5 years, 5–14 years, 15–24 years, and ≥25 years. The prevalence of each of the five binary EEG-pattern indicators was summarized within these age groups. Associations between age group and each EEG indicator were assessed using Pearson's chi-square statistic with 12,000 random permutations to obtain two-sided permutation p-values. Because five EEG indicators were tested within the age comparison, the Benjamini–Hochberg procedure was applied to control the false discovery rate, with an adjusted q-value <0.05 considered statistically significant.

Associations between EEG indicators and sex were assessed using two-sided Fisher's exact tests. For episode frequency, the recorded categories were grouped into weekly and monthly categories according to the original coding. Associations between these groups and the EEG indicators were also assessed using Fisher's exact tests. Associations between broad district-reporting groups and EEG indicators were assessed using Pearson's chi-square statistics with 12,000 permutations. Benjamini–Hochberg adjustment was applied separately within each family of five EEG-marker comparisons.

A multivariable logistic regression model was constructed using the A numerical age-related pattern was observed for generalized-onset positivity, but the association did not remain statistically significant after correction for multiple comparisons. Age group, sex, episode-frequency group, and district-reporting group were entered as explanatory variables. Adjusted odds ratios (ORs), 95% Wald confidence intervals (CIs), and two-sided p-values were reported. Overall model fit was assessed using the likelihood-ratio test, and McFadden's pseudo-R² was reported as a measure of model explanatory performance. The regression analysis was considered exploratory, and predictor-specific p-values were not adjusted for multiple comparisons.

3.9 Ethical Considerations

The study information provided indicates that approval was obtained from the Ethics Committee of Bolan Medical Complex Hospital, Quetta.

3.10 EEG Acquisition and Interpretation

EEG data were obtained retrospectively from the available clinical records and EEG reports. The EEG findings were recorded as binary indicators according to the terminology and coding available in the hospital database. The evaluated indicators included focal-onset discharges, generalized-onset activity, generalized slowing, burst-pattern activity, and bilateral discharges. For each indicator, a value of 1 represented the presence of the recorded finding and 0 represented its absence. The EEG recording parameters, including the EEG system and manufacturer, electrode montage, number of channels, recording duration, patient state during recording, and use of activation procedures such as hyperventilation and photic stimulation, should be reported according to the original EEG records or departmental protocol. Similarly, the professional qualifications of the EEG interpreters and whether EEGs were independently reviewed by more than one reader should be specified if this information is available. Where repeat EEG examinations were performed, the rule used to select the EEG included in the analysis should also be stated.

3.1 Frequency of gander distribution

Table 3.1. Distribution by recorded gander distribution

276 records was selected, 189 were male (68.5%) and 87 were female (31.5%). The male-to-female ratio was 2.17:1. This imbalance describes the composition of the available dataset and should not be interpreted as a population-level sex distribution.

3.2 Age-band distribution

Table 3. 2Age categories of the recorded participants

The distribution of EEG-pattern indicators varied across the four age groups. Focal-onset discharges showed the clearest age-related pattern, increasing from 3.6% (6/165) among participants younger than five years to 17.2% (11/64) among those aged 5–14 years, 19.4% (6/31) among those aged 15–24 years, and 37.5% (6/16) among those aged 25 years or older. In contrast, the generalized-onset indicator was most frequently recorded among participants younger than five years (59.4%) and progressively less frequent across the older age groups. However, the association between age group and generalized-onset positivity did not remain statistically significant after Benjamini–Hochberg correction (permutation p = 0.048; adjusted q = 0.081). The age-group association for focal-onset discharges remained statistically significant after correction (permutation p < 0.001; adjusted q < 0.001).

Figure 3.1. Distribution across the eleven recorded age categories.

3.3 District of residence

Table 3.3. Recorded district categories

District names reproduce the source dataset’s category labels; “not willing to disclose” is a response category, not a geographic district. Quetta accounted for 135 records (48.9%). A further 56 participants (20.3%) did not disclose their district. The other 85 records (30.8%) were distributed across the remaining twelve labelled districts, with Pashin (n = 12) being the largest of these. The geographic distribution is concentrated in Quetta and includes substantial nondisclosure, which limits geographic comparisons.

3.4 Reported episode frequency

Table 3.4. Distribution of episode-frequency categories

The SPSS variable is FREQUENC (source label: “Sizer Frequency”). The episode wording follows its available value labels. The most commonly recorded category was one episode per week (104/276; 37.7%), followed by one episode per month (63/276; 22.8%) and multiple episodes per month (34/276; 12.3%). For subgroup analysis, codes 1–4 were grouped as weekly categories (166/276; 60.1%) and codes 5–8 as monthly categories (110/276; 39.9%). These represent response categories rather than a continuous episode rate, and should not be averaged numerically. No record used the code 0 (“no episode”).

3.5 Frequency of recorded EEG-pattern indicators

Table 3.5. Prevalence of each binary EEG-pattern indicator (N = 276)

95% confidence intervals were computed by the Wilson binomial method. Variables were coded 1 = yes and 0 = no. Categories may overlap within a single record. “Burst pattern” reflects the source field BURST, labelled “Burst Suppuration”; its clinical terminology should be checked. The generalized-onset indicator was the most frequent pattern, recorded for 145 participants (52.5%; 95% CI 46.7%–58.4%). A burst-pattern indicator was present for 54 (19.6%), focal-onset discharges for 29 (10.5%), generalized slowing for 23 (8.3%) and bilateral discharges for 20 (7.2%). Because the indicators are independent binary fields, these percentages do not sum to 100%.

Figure 3.4. Percentage of records with each recorded EEG-pattern indicator.

3.6 Number of simultaneous EEG indicators

Table 3.6. Count of positive EEG indicators per record

Exactly one of the five EEG indicators was marked present in 253 records (91.7%). Nine records (3.3%) had two simultaneous positive markers, and 14 (5.1%) had no positive marker in the five fields. These categories summarize only the five indicators provided and do not establish whether the overall EEG was normal or abnormal.

3.7 Age strata and EEG findings

Table 3.7. EEG-pattern prevalence by grouped age

Cells show positive cases, n (row %). Ages were grouped by combining the original 11 coded age bands; no individual ages were inferred. The age-stratified proportions show that focal-onset discharges were recorded in 6/165 (3.6%) participants under five years and 6/16 (37.5%) aged 25 years or older. The generalized-onset indicator was present in 59.4% and 37.5% of these groups, respectively; the older group contains few records.

Table 4.8. Age-group association tests for the five EEG markers

Two-sided permutation tests (12,000 permutations per marker) use Pearson χ² as the test statistic. Benjamini–Hochberg (BH) correction was applied across these five age comparisons. Some expected cells were below five.

The association between grouped age and focal-onset discharges remained statistically significant after the five-comparison correction (permutation p <0.001; adjusted q <0.001). Neither generalized-onset positivity (q = 0.081) nor generalized slowing (q = 0.055) met the adjusted threshold, even though their unadjusted p values were below 0.05. Findings for the oldest stratum require caution because it contained only 16 records.

Figure 3.5. Focal-onset indicator prevalence across four age strata.

Figure 3.6. Generalized-onset indicator prevalence across four age strata.

3.9 Sex and EEG-pattern associations

Table 3.9. Recorded sex and EEG-pattern associations

Odds ratios compare female with male participants; p values are two-sided Fisher exact tests. BH correction was applied across five sex comparisons. Percentages use the column subgroup sizes (male n = 189, female n = 87).

Focal-onset discharges were recorded in 12.7% of males and 5.7% of females (female-versus-male OR 0.42, 95% CI 0.15–1.14), but the difference was not statistically significant (Fisher p = 0.093). Generalized-onset positivity was 49.7% among males and 58.6% among females (OR 1.43, 95% CI 0.86–2.39; p = 0.195). No sex-based comparison remained significant following BH correction (all q > 0.05).

3.10 Episode frequency and EEG-pattern associations

Table 3.10. EEG findings across weekly versus monthly episode categories

Odds ratios compare monthly categories (n = 110) against weekly categories (n = 166). The grouping is based only on the variable’s code labels; p values are Fisher exact tests with BH adjustment across five comparisons. Generalized-onset positivity was comparable between the weekly-category group (86/166; 51.8%) and the monthly-category group (59/110; 53.6%). The monthly-versus-weekly odds ratio was 1.08 (95% CI 0.66–1.74; Fisher p = 0.806). Differences for focal discharges, generalized slowing, burst pattern and bilateral discharges were also not statistically significant after correction (all adjusted q > 0.05). Thus, no association between the broad weekly/monthly response grouping and these indicators was established in this dataset.

3.11 Recorded geography and EEG findings

Table 3.11. EEG-pattern prevalence by broad district-reporting group

Generalized-onset positivity occurred in 61 of 135 Quetta cases (45.2%), 50 of 85 other-district cases (58.8%) and 34 of 56 cases without a disclosed district (60.7%). The nondisclosed category was kept separate rather than allocated to a geographic location.

Table 3.12. Permutation tests of EEG-pattern differences by district group

Three reporting groups (Quetta, other districts, not disclosed). Pearson χ² statistic with 12,000 random permutations; BH correction across the five markers. The global permutation test for generalized onset was not statistically significant (p = 0.055; adjusted q = 0.277). None of the other EEG indicators showed a statistically supported difference across the three broad reporting groups after BH correction.

3.13 Recorded disease codes across age strata

Table 3.13. Disease codes 1 and 2 by age group

Percentages are calculated within each row. Code 1 is labelled abnormal in SPSS; code 2 lacks any SPSS value label. Clinical interpretation of code 2 has intentionally not been assigned. Code 2 was recorded predominantly among younger individuals: 152 of 165 participants under five years (92.1%) had code 2, compared with none of the 16 participants aged 25 years or older. Although there is a pronounced numerical separation across age strata, the clinical meaning of this contrast cannot be determined until the disease coding is corrected; no clinical hypothesis test is presented for this variable.

3.14 Recorded disease codes and EEG indicators

Table 3.14. EEG markers by literal recorded disease code

Descriptive audit table only. The SPSS label of disease code 2 is undefined, so these columns are not labelled normal/abnormal groups.

Focal-onset discharges occurred in 21/76 (27.6%) of code-1 records and 8/200 (4.0%) of code-2 records. In contrast, the generalized-onset indicator was recorded in 22/76 (28.9%) and 123/200 (61.5%), respectively. These striking differences underscore the need to verify how disease was coded. They do not, by themselves, establish diagnostic associations because code 2 has no defined meaning in the supplied value labels.

3.15 Multivariable analysis of generalized-onset indicator

Table 3.15. Multivariable logistic regression: generalized-onset marker (yes versus no)

Outcome: GENRALIZ = 1, recorded in 145 of 276 cases. Model includes categorical age stratum, sex, weekly/monthly episode category and district reporting group. Adjusted odds ratios (OR), 95% Wald confidence intervals and two-sided model p values are shown. Model LR p = 0.023; McFadden pseudo-R² = 0.042. Exploratory regression p values were not adjusted for multiplicity.

After adjustment for the other recorded factors, participants aged 5–14 years had lower odds of a positive generalized-onset indicator than children under five years (adjusted OR 0.49, 95% CI 0.27–0.89; p = 0.019). Associations for ages 15–24 (adjusted OR 0.63; p = 0.248), ages 25 or older (adjusted OR 0.38; p = 0.082), female sex (adjusted OR 1.61; p = 0.081), monthly episode categories (adjusted OR 1.19; p = 0.501) and both non-Quetta groups were not individually significant at p < 0.05. The fitted model was globally significant (likelihood-ratio p = 0.023) but had limited explanatory power (McFadden pseudo-R² = 0.042). These exploratory associations cannot establish causation, and age-group effects may be imprecise for older participants.

The present retrospective study evaluated demographic and EEG-pattern indicators among 276 patients with epilepsy at a tertiary-care neurology service in Quetta, Pakistan. Most patients were male, and 59.8% were younger than five years. Generalized-onset EEG abnormalities were most frequently recorded, while focal-onset discharges showed a significant age-related increase that remained significant after Benjamini–Hochberg correction. These findings highlight potential age-related EEG differences, particularly in focal-onset discharges, warranting confirmation through standardized and prospective studies.

Demographic characteristics of the study population

The study population comprised 189 males (68.5%) and 87 females (31.5%), yielding a male-to-female ratio of approximately 2.17:1. This male predominance is comparable with findings from some previous epilepsy studies in Pakistan, including Shakir Ullah et al. in Khyber Pakhtunkhwa, where 58.8% of participants were male. However, differences in study populations, recruitment settings, healthcare access, and health-seeking behaviour may contribute to variations in sex distribution. The lower representation of females may also reflect socioeconomic or sociocultural barriers, although these factors were not directly assessed in the present study and should therefore be considered possible explanations rather than established causes. Age distribution showed a marked concentration in early childhood, with 106 participants (38.4%) aged 0–11 months, 59 (21.4%) aged 1–4 years, and 42 (15.2%) aged 5–9 years. Overall, 165 participants (59.8%) were younger than five years and 229 (83.0%) were younger than 15 years, indicating a predominantly paediatric hospital-based sample. Similar patterns have been reported in studies of paediatric epilepsy and seizure disorders [11,12]. However, the present findings describe the age distribution of hospital-recorded patients and cannot establish age-specific epilepsy incidence or prevalence without population-based denominators.

Distribution of reported episode frequency

The recorded episode-frequency categories showed that one episode per week was the most common response, reported by 104 participants (37.7%), followed by one episode per month in 63 (22.8%) and multiple episodes per month in 34 (12.3%). Overall, 166 participants (60.1%) were classified in the weekly-frequency categories, compared with 110 (39.9%) in the monthly categories. Thus, weekly episode-frequency responses were more common in the study population. However, because frequency was recorded using predefined categorical responses rather than continuous or longitudinal measurements, these findings cannot demonstrate changes in seizure frequency or control over time. Importantly, none of the five EEG indicators showed a statistically significant association with the weekly versus monthly frequency groups after Benjamini–Hochberg correction. For example, generalized-onset EEG abnormalities were recorded in 51.8% of the weekly group and 53.6% of the monthly group (OR = 1.08, 95% CI: 0.66–1.74; p = 0.806). Overall, episode-frequency category was not significantly associated with the recorded EEG indicators in this dataset.

Age-related distribution of EEG indicators

The generalized-onset indicator was the most frequently recorded EEG marker, occurring in 145 participants (52.5%; 95% CI: 46.7%–58.4%), followed by burst-pattern (19.6%), focal-onset discharges (10.5%), generalized slowing (8.3%), and bilateral discharges (7.2%). These indicators were recorded as separate binary variables and should not be interpreted as mutually exclusive clinical seizure classifications. Generalized-onset positivity was highest among participants younger than five years (59.4%) and decreased across older age groups; however, this association was not statistically significant after Benjamini–Hochberg correction (permutation p = 0.048; adjusted q = 0.081). In contrast, focal-onset discharges increased progressively with age, from 3.6% in participants younger than five years to 17.2%, 19.4%, and 37.5% in the 5–14, 15–24, and ≥25-year groups, respectively, with the association remaining significant after correction (permutation p < 0.001; adjusted q < 0.001). This represented the clearest statistically supported age-related EEG finding in the study. Although previous studies have reported a substantial proportion of generalized seizure presentations in paediatric populations [16–19], direct comparisons should be made cautiously because of differences in seizure classification, EEG criteria, and participant selection. Furthermore, the oldest age group included only 16 participants, limiting the precision of its estimated focal-onset prevalence. Overall, the findings suggest an age-related pattern in recorded focal-onset discharges, but the cross-sectional hospital-based design does not establish changes in EEG patterns over the course of life.

Generalized slowing and burst-pattern indicators

Generalized slowing was recorded in 23 participants (8.3%), with prevalence increasing from 4.2% among those younger than five years to 18.8% among those aged 25 years or older. However, this association was not statistically significant after multiple-comparison correction (permutation p = 0.022; adjusted q = 0.055). Similarly, the burst-pattern indicator was observed in 54 participants (19.6%), with prevalence ranging from 17.6% in children younger than five years to 31.2% in participants aged 25 years or older, but the age-group association was not significant (permutation p = 0.453; adjusted q = 0.567). These findings suggest numerical differences across age groups but do not establish statistically significant associations. Generalized slowing can occur in various neurological conditions and should not be interpreted as a specific indicator of epilepsy severity or seizure type. Furthermore, the dataset variable labelled “Burst Suppuration” requires clinical verification before it can reliably be interpreted as a burst-pattern indicator or compared with previous studies [20,22,23].

Bilateral discharges and overlapping EEG indicators

Bilateral discharges were recorded in 20 participants (7.2%), with prevalence ranging from 3.2% among those aged 15–24 years to 8.5% among children younger than five years. However, the association between age group and bilateral discharges was not statistically significant (permutation p = 0.776; adjusted q = 0.776). Analysis of the five EEG indicators showed that 253 participants (91.7%) had exactly one positive indicator, nine (3.3%) had two, and 14 (5.1%) had none; no participant had three or more positive indicators. These findings indicate limited overlap among the recorded markers. However, the absence of positive findings across these five indicators does not necessarily indicate a normal EEG, as other abnormalities may not have been captured. Furthermore, bilateral discharges alone cannot establish epilepsy severity, prognosis, or underlying cause; comparisons with previous studies should consider differences in EEG definitions and patient characteristics.

Associations between EEG indicators and sex

The most notable finding was the progressive increase in recorded focal-onset discharges across age groups, from 3.6% among participants younger than five years to 37.5% among those aged 25 years or older. This association remained significant after correction for multiple comparisons (adjusted q < 0.001). The pattern may reflect age-related differences in epilepsy presentations or referral characteristics; however, these explanations cannot be confirmed from the retrospective data. Furthermore, the oldest age group included only 16 participants, limiting the precision of its estimate. Therefore, the findings indicate an age-related association within this hospital-based cohort but do not establish that focal epilepsy increases with age in the general population.

Geographic distribution and EEG findings

Quetta accounted for 135 participants (48.9%), followed by other districts (85; 30.8%), while 56 (20.3%) did not disclose their district of residence. The generalized-onset indicator was recorded in 45.2% of participants from Quetta, 58.8% from other districts, and 60.7% of those with undisclosed residence. However, these differences were not statistically significant after correction for multiple comparisons (p = 0.055; adjusted q = 0.277), and no other EEG indicator showed a significant association with district-reporting category. These findings do not establish a relationship between residence category and EEG patterns. The predominance of Quetta records may reflect the hospital's location and referral patterns rather than the geographic distribution of epilepsy. Furthermore, undisclosed residence limits interpretation, highlighting the need for future studies with more complete demographic data and defined population-based sampling.

Multivariable analysis of the generalized-onset indicator

The generalized-onset indicator was the most frequently recorded EEG finding overall, occurring in 52.5% of records. Its prevalence was highest among participants younger than five years (59.4%) and lower in the older age groups. Although the unadjusted association between age group and generalized-onset positivity reached the conventional p < 0.05 threshold, it did not remain statistically significant after Benjamini–Hochberg correction (adjusted q = 0.081). Accordingly, the present data do not provide sufficient statistical evidence to conclude that generalized-onset EEG positivity differs significantly by age group. The observed numerical trend may nevertheless warrant further investigation in larger cohorts using standardized EEG terminology and clinically verified seizure classifications.

4.2 Limitations

This study has several limitations that should be considered when interpreting the findings. First, its retrospective, hospital-based design limits generalizability and does not permit estimation of population-level epilepsy prevalence or incidence. Second, the dataset contained categorical age bands rather than exact ages, and the number of participants in the older age groups was relatively small, particularly among those aged 25 years or older. Third, the analysis was based on five binary EEG indicators recorded in the available dataset rather than standardized, independently reviewed EEG classifications. Detailed information regarding EEG acquisition parameters, recording duration, patient state, activation procedures, and inter-rater agreement was not available in the supplied dataset and should be reported if available from the original records. Fourth, the clinical meaning of the disease-code category 2 could not be established from the available SPSS value labels, limiting interpretation of analyses involving this variable. Fifth, the source variable labelled “Burst Suppuration” requires verification before it can be interpreted as a specific EEG phenomenon. Finally, the available dataset did not contain a unique patient identifier or examination date, preventing independent verification of whether all records represented unique patients rather than repeated clinical encounters. These limitations highlight the need for prospective studies using standardized diagnostic criteria, validated EEG terminology, and more complete clinical information.

This study provides insight into the demographic characteristics and recorded EEG patterns of patients with epilepsy attending a tertiary-care hospital in Quetta, Pakistan. The study population was predominantly male and comprised mainly children younger than five years. Generalized-onset activity was the most frequently recorded EEG indicator, whereas focal-onset discharges demonstrated the strongest statistically supported association with age, remaining significant after correction for multiple comparisons. In contrast, the age-related variation in generalized-onset activity was not statistically significant after adjustment. These findings highlight potential age-related differences in EEG patterns within this hospital-based population. Further prospective studies using standardized diagnostic criteria, validated EEG classifications, detailed clinical data, and independent EEG review are needed to confirm these findings and establish their broader clinical significance.

Conflict of interest: None

1. Guekht, A., et al., The road to a World Health Organization global action plan on epilepsy and other neurological disorders. Epilepsia, 2021. 62(5): p. 1057-1063. https://doi.org/10.1111/epi.16856
2. Rubinos, C., et al. Epilepsy Care in Latin America and the Caribbean: overcoming challenges and embracing opportunities. in Seminars in Neurology. 2024. Thieme Medical Publishers, Inc. https://doi: 10.1055/s-0044-1782616
3. Usman, S., et al., Demographic profile of patients with epilepsy in a community clinic. Pakistan journal of Medical Sciences, 2007. 23(6): p. 873. https://www.researchgate.net/publication/238073110
4. Ullah, S., et al., The epidemiological characteristics of epilepsy in the province of Khyber Pakhtunkhwa, Pakistan. Frontiers in neurology, 2018. 9: p. 845. https://doi.org/10.3389/fneur.2018.00845
5. Ibrahim, A., et al., Neurological Disorders In Pakistan: Frequency, Distribution, Pattern and Related Factors: Neurological Disorders In Pakistan. Pakistan Pediatric Journal, 2023. 47(1). https://ppj.org.pk/index.php/ppj/article/view/134
6. Wasay, M., et al., Neurological disorders and disability in Pakistan: A cross-sectional multicenter study. Journal of the Neurological Sciences, 2023. 452: p. 120754. https://doi.org/10.1016/j.jns.2023.120754
7. Imtiaz, H., et al., A Comprehensive electroencephalogram audit in a tertiary care center in Pakistan: demographic distribution, referral indications and findings. Pakistan Journal of Neurological Sciences, 2023. 18(03). https://doi.org/10.56310/pjns.v18i03.291
8. Dai, M. and Y. Ouyang, Global, regional, and national burden of refraction disorders: findings from the global burden of disease study 2021 and projections to 2050. BMC Public Health, 2025. 25(1): p. 1247. https://doi.org/10.1186/s12889-025-22440-w
9. Jareonsettasin, P. and J.W. Sander, The Role of Social Determinants of Health in Childhood Epilepsy. Turkish Archives of Pediatrics, 2025. 60(1): p. 13. 10.5152/TurkArchPediatr.2025.24271
10. Saleem Barech, M., Neurological burden in Baluchistan province, Pakistan. Pakistan Journal of Neurological Sciences (PJNS), 2017. 12(4): p. 1-2. https://ecommons.aku.edu/pjns/vol12/iss4/1
11. Camfield, P. and C. Camfield, Incidence, prevalence and aetiology of seizures and epilepsy in children. Epileptic disorders, 2015. 17(2): p. 117-123. https://doi.org/10.1684/epd.2015.0736
12. Gaily, E., et al., Incidence and outcome of epilepsy syndromes with onset in the first year of life: a retrospective population‚Äêbased study. Epilepsia, 2016. 57(10): p. 1594-1601. https://doi.org/10.1111/epi.13514
13. Campos, M.A.G., et al., Continuous epileptiform discharges are associated with worse neurodevelopmental findings in a congenital Zika syndrome prospective cohort. Seizure: European Journal of Epilepsy, 2024. 118: p. 148-155. https://doi.org/10.1016/j.seizure.2024.04.022
14. Hassan, F., et al., Assessment Of Frequency of EEG Findings in Children with Epilepsy in A Tertiary Care Hospital in Karachi, Pakistan: Frequency of EEG Findings in Children With Epilepsy. Pakistan BioMedical Journal, 2022: p. 258-261. https://doi.org/10.54393/pbmj.v5i6.565
15. Feigin, V.L., et al., Global, regional, and national burden of epilepsy, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet Public Health, 2025. 10(3): p. e203-e227. https://doi.org/10.1016/ S2468-2667(24)00302-5
16. Khatri, I., et al., Epidemiology of epilepsy in Pakistan: review of literature. JPMA. The Journal of the Pakistan Medical Association, 2003. 53(12): p. 594-597. https://pubmed.ncbi.nlm.nih.gov/14765939/
17. Ullah, S., et al., The Epidemiological Characteristics of Epilepsy in the Province of Khyber Pakhtunkhwa. Neuroepidemiology in Resource-Limited Areas, 2020: p. 95265. https:// 10.3389/fneur.2018.00845
18. Zafar, N., et al., Etiology of Seizures in Children Aged 6 Months to 10 Years Presenting to Tertiary Care Hospital. Journal of Health and Rehabilitation Research, 2024. 4(1): p. 1-6. https://doi.org/10.61919/jhrr.v4i1.1731
19. Rehman, Z.U., Clinical characteristics and etiology of epilepsy in children aged below two years: perspective from a tertiary Childcare hospital in South Punjab, Pakistan. Cureus, 2022. 14(4). https://doi: 10.7759/cureus.23854
20. Kulsoom, S., et al., Electroencephalogram Spectrum in Children with Neurological Disorders an Experience of Tertiary Care Hospital. Pakistan Journal of Medical & Health Sciences, 2022. 16(06): p. 886-886. https://doi.org/10.53350/pjmhs22166886
21. Syvertsen, M.R., Epidemiology of epilepsy in Buskerud County: emphasasing clinical and psychosocial aspects of juvenile myoclonic epilepsy. 2019.
22. Zubair, A., et al., Clinical Spectrum and EEG Findings in Patients with Suspected Non-Convulsive Status Epilepticus. Global Immunological & Infectious Diseases Review. 9(III). http://dx.doi.org/10.31703/giidr.2024(IX-III).02
23. Gao, Z., et al., A retrospective study of electroencephalography burst-pattern indicator in children undergoing general anesthesia. Pediatric Investigation, 2021. 5(04): p. 271-276. http://dx.doi.10.1002/ped4.12287
24. Awan, S., et al., Spectrum of neurological disorders in neurology outpatients clinics in urban and rural Sindh, Pakistan: a cross sectional study. BMC neurology, 2019. 19: p. 1-6. https://doi.org/10.1186/s12883-019-1424-1
25. Beghi, E., The epidemiology of epilepsy. Neuroepidemiology, 2020. 54(2): p. 185-191. https://doi.org/10.1159/000503831