First-Generation Antihistamines and Seizures in Young Children

Key Points Question What is the association of prescriptions of first-generation antihistamines with seizure events in young children? Findings In this cohort study of 3178 children in Korea, prescriptions of a first-generation antihistamine were associated with a 22.0% higher seizure risk in children, especially in those aged 6 to 24 months. Meaning These findings reinforce the importance of cautious first-generation antihistamine prescription in young children and underline the need for further research to fully understand associations between antihistamine use and seizure risk.


Introduction
First-generation H1 antihistamines, developed in the 1940s and 1950s, initially served as tranquilizers and antipsychotics by crossing the blood-brain barrier (BBB) and suppressing histamine neurotransmission in the central nervous system (CNS). 1 Despite reduced therapeutic use owing to their poor selectivity and interactions with other receptors, these drugs are still widely used for managing rhinorrhea in the common cold 2,3 or to control an itching sensation 4,5 in children.Because first-generation antihistamines can cross the BBB, their effects may extend beyond somnolence and drowsiness 1 to markedly influence brain wave activity. 6Thus, caution is advised when prescribing these antihistamines to children younger than 2 years, an age group for whom drug safety data are lacking and first-generation antihistamines are generally not recommended. 7merous studies 6,[8][9][10] have indicated that first-generation antihistamines can affect brain waves.In particular, they can induce symptomatic seizures, affect electroencephalographic (EEG) activity and seizure thresholds in adults with inherent seizure susceptibility, and alter resting EEG activity. 6,8Animal studies have indicated that H1 antihistamines increase seizure susceptibility in rodents, 9 and studies using genetically modified animal models have suggested a link with epileptic seizures. 10Clinically, antihistamines have been identified as a common trigger of acute symptomatic seizures, 11 and there have been some instances of seizure induction by second-generation antihistamines. 12,13Altered seizure patterns were noted in children with febrile seizures who were exposed to antihistamines. 14,15However, the impact of first-generation antihistamines on brain waves and the heightened sensitivity to them among vulnerable age groups remains less explored in clinical practice. 11,16 hypothesized that acute prescription of first-generation antihistamines increases the risk of seizures, especially in young children.Therefore, this study investigated the association between prescription of first-generation antihistamines and seizures in young children by using a nationwide population-based claims dataset.In this study, children served as their own controls, eliminating time-invariant confounding factors and focusing on the age groups vulnerable to seizures owing to the transient nature of CNS development in children.

Study Design and Data Source
This retrospective cohort study used representative data from the National Health Insurance Service

Case-Crossover Design
A case-crossover approach was adopted to assess the risk of seizure events associated with firstgeneration antihistamine prescription (eFigure in Supplement 1).It analyzes short-term exposure effects on acute events by comparing transient exposure histories within the same individual. 4,17,18 chose this design because antihistamines are typically used for short periods, and seizures are considered their immediate health impacts.In the case-crossover design, each participant acts as their own control.This approach inherently eliminates confounding factors that remain constant among individuals.Because the same participants' data were considered for both the case and control periods, issues related to differential recall were potentially reduced.However, one must be cautious about bias arising from temporal changes in disease severity.
The date of occurrence of a seizure event was marked as the index date.Because the 95th percentile duration of antihistamine prescriptions ranged from 13 to 18 days, the main observational window for first-generation antihistamine prescription was assessed within the 15 days preceding this index date, referred to as the hazard period.This was compared with prescriptions 31 to 45 days and 61 to 75 days before the index date, which served as control periods 1 and 2, respectively.These control periods were presumed to be unrelated to seizures.The findings are reported in accordance with the recommended guidelines for observational studies that use routinely collected health data. 19

Study Population
The study population is shown in

Seizure Event
The seizure events of interest were defined as emergency department visits with a principal diagnosis of epilepsy (ICD-10 code G40.X), status epilepticus (ICD-10 code G41.X), or convulsions (ICD-10 code R56.8), with each individual's first event being designed as the index date.Due to the complexities associated with analyzing multiple events from the same individual in a case-crossover study, we focused our analysis on the initial event for each participant.This approach was chosen to prevent overlapping exposure windows and to ensure the independence of observations, essential for the reliability of our statistical analysis.

Prescription of First-Generation Antihistamines
The focus of this study was on prescription of first-generation antihistamines, that is, pure firstgeneration antihistamines and compound medications containing first-generation antihistamines.
These prescriptions were identified using prescription codes from the NHIS database.The firstgeneration antihistamines included chlorpheniramine maleate, mequitazine, oxatomide, piprinhydrinate, and hydroxyzine hydrochloride, as listed in eTable 1 in Supplement 1.All drugs were approved by the Korean Food and Drug Administration.Prescription of first-generation antihistamines was defined as having been prescribed these drugs during the case or control period for at least 1 day.Additionally, the participants were classified as new users of first-generation antihistamines if they had not been exposed to these drugs in the 16 to 30 days preceding the index date.

Covariates
Participants were classified into 3 age groups at the index date: 6 to 24 months, 25 months to 6 years, and 7 years or older.Participants' birth regions were classified into 4 categories: Seoul, metropolitan areas (Busan, Daegu, Incheon, Gwangju, Daejeon, and Ulsan), urban areas, and rural areas.Economic status was estimated based on insurance payments.
To assess clinical conditions, cases of hospitalization in an intensive care unit before the age of 6 months were identified.Additionally, perinatal conditions were determined using ICD-10 codes, including length of gestation and fetal growth (ICD-10 codes P05.X-P08.X), convulsions and CNS disturbances (ICD-10 codes P90.X-P91.X), birth trauma (ICD-10 codes P10.X-P15.X), respiratory and cardiovascular disorders specific to the perinatal period (ICD-10 codes P20.X-P29.X), infections specific to the perinatal period (ICD-10 codes P35.X-P39.X), congenital malformations (ICD-10 codes Q00.X-Q89.X), and chromosomal anomalies (ICD-10 codes Q90.X-Q99.X).If a participant was diagnosed with 1 or more of these conditions, they were considered to have a perinatal condition.

Statistical Analysis
Data were analyzed from June 3, 2023, to January 30, 2024.In the case-crossover design analysis, the focus was on participants who were exposed to antihistamines during either the hazard period or a control period but not both (ie, discordant participants).These individuals contributed to the odds ratio (OR) estimation.Odds ratios were calculated to assess the association between antihistamine prescription and seizures.This was achieved by comparing the exposure status of the participants during the hazard period with their exposure status during the control periods.
The risks of seizure event with antihistamine prescriptions were analyzed using multivariable conditional logistic regression, with results estimated by ORs and 95% CIs.The analysis was adjusted for various factors, including age, sex, residential area, economic status, season of the index date, and perinatal conditions. 20Since the missing data were less than 2% only for region of birth and economic status, the analysis was conducted without imputing the missing data.
For sensitivity analysis, the risk of antihistamine prescription for seizure events was determined using alternative time windows of 10 or 5 days.Hence, the hazard period was defined as 1 to 10 days before the seizure event, with the corresponding control periods set at 31 to 40 or 61 to 70 days before the event, or as 1 to 5 days before the seizure event, with the corresponding control periods

JAMA Network Open | Neurology
First-Generation Antihistamines and Seizures in Young Children Additionally, to match the seasons of the hazard and control periods, the risk was analyzed by considering 1 to 15 days before the index date as the hazard period and 15 days before the start of a 365-day period before the index date as the control period.In addition, to adjust for clinical conditions that could potentially be associated with the occurrence of seizures, the 9 most common concomitant diseases during the risk and control periods were defined as time-dependent covariates for each participant.These diseases included acute nasopharyngitis, acute pharyngitis, acute tonsillitis, acute sinusitis, acute purulent otitis media, acute upper respiratory tract infection, acute bronchitis, acute bronchiolitis, and asthma.The risk of seizure with antihistamine prescriptions was further analyzed by adjusting for these time-varying covariates.
We also evaluated whether comparable results were obtained when focusing solely on firstgeneration single-formulation antihistamines and excluding compound medications with firstgeneration antihistamines.Subgroup analyses were performed by stratifying participants' characteristics, including age group at the index date, sex, region of birth, economic status, calendar year of birth, season of the index date, and perinatal conditions.For intrasubgroup OR comparisons, the log-OR difference between strata was used to calculate z scores and P values.Two-sided P < .05indicated statistical significance.This analysis was conducted using the LOGISTIC procedure in SAS software, version 9.4 (SAS Institute Inc).

Baseline Characteristics of the Study Population
Baseline demographic and clinical characteristics of the children are presented in

Association Between Antihistamine Prescription and Seizure Events
Table 2 presents the data on antihistamine prescription during different periods before the index date and its association with seizure events.Focusing on the time window of 1 to 15 days, the number of participants who were prescribed antihistamines was 1476 (46.4%) during the hazard period compared with 1239 (39.0%) during control period 1 and 1278 (40.2%) during control period 2.
Consequently, the risk of seizure events within 15 days of antihistamine prescription was found to be elevated (adjusted OR [AOR], 1.22 [95% CI, 1.13-1.31]).Notably, new users of antihistamines during the hazard period also demonstrated a higher risk of seizures (AOR, 1.25 [95% CI, 1.14-1.35])compared with those in the control periods.
In the sensitivity analyses in which the time window was set to either 1 to 10 days or 1 to 5 days, the elevated risk of seizures associated with antihistamine prescription showed similar patterns.
Specifically, AORs were 1.25 (95% CI, 1.15-1.36)for the 10-day window and 1.36 (95% CI, 1.23-1.51)for the 5-day window.Moreover, even when further adjusted for concomitant diseases as a timevarying covariate, antihistamine prescriptions significantly increased the risk of seizure events in the

JAMA Network Open | Neurology
First-Generation Antihistamines and Seizures in Young Children

Association Between Antihistamine Prescription and Seizure Events According to Participant Characteristics
Figure 2 illustrates the risk of seizure events associated with antihistamine prescription, broken down by participant characteristics.The analysis revealed interaction effects between subgroups based on age groups at the index date (P = .04for interaction).Notably, children aged 6 to 24 months with an index date had a significantly increased risk, with an AOR of 1.49 (95% CI, 1.31-1.70).

Discussion
In this cohort study, we used large-scale national data and a self-controlled case-crossover design to analyze the association between first-generation antihistamine prescription and seizure events.
Interestingly, we found that first-generation antihistamine prescription was linked to a 22.0% increase in seizures among children aged 6 to 24 months.This increased risk was consistently observed across stratified analyses and remained consistent when considering different time-control points and medication types.Notably, the study underscores a substantial increase in seizure risk associated with antihistamine prescription among children aged 6 to 24 months.We are not aware of any other studies that have pointed out the increased risk of seizures with first-generation antihistamines in this particular age group, in contrast to older children, for whom we found no such association.
Previous studies 11,12,14,15 have raised concerns about the potential of antihistamines to induce seizures, especially in acute symptomatic cases and patients with epilepsy and new-onset seizures, and EEG analyses support these findings. 8This effect might be mediated by histamine receptors distributed in the brain's cortex and hippocampus 21,22 and the ability of antihistamines to cross the BBB, affecting CNS activity. 23,24Studies in H1 receptor knockout mice and genetically modified animal models of temporal lobe epilepsy showing increased histidine decarboxylase activity, 10,25 as well as long-term administration of antihistamines resulting in glutamine synthetase dysfunction, 9,26 indicated that the role of histamine in modulating brain seizure susceptibility, particularly affecting neuronal excitability in the hippocampus.The possible mechanism by which antihistamines may increase the risk of seizures is complex and involves multiple pathways: inhibition of the role of antihistamines in impeding the antiseizure effect 27 ; depletion of hypothalamic neural histamine causing neuronal excitability 16 ; impairment of glutamine synthase, a key enzyme in glutamate metabolism and γ-aminobutyric acid synthesis 9 ; and direct inhibition of neuronal channels. 28nsequently, antihistamines might increase seizure susceptibility by enhancing neural activity and a Calculated by multivariable conditional logistic regression to assess the association between first-generation antihistamine use and seizure event.
b Adjusted for various factors, including age, sex, residential area, economic status, season at the index date, and perinatal conditions.
c The hazard period was the time window just before the index date, while control periods 1 and 2 were the same duration, starting at 31 and 61 days before the index date, respectively.
d The hazard period was the time window just before the index date, while the control periods were the same duration, starting 1 year before the index date.

JAMA Network Open | Neurology
First-Generation Antihistamines and Seizures in Young Children Interestingly, children exposed to antihistamines between the ages of 6 and 24 months showed an increased risk of seizures compared with other young children.This period is critical for brain development and is characterized by a primitive brain structure and rapid developmental processes, making it a susceptible phase for future brain development, including cognition and fine motor functions. 29,30The vulnerability of the infant brain to antihistamines is partly due to the developing BBB, which continues to evolve in this age group. 31In infants, incomplete formation of the BBB leads to increased permeability and a higher risk of drug penetration into the brain tissue.Therefore, antihistamines, which are relatively harmless to adults and older children, might markedly affect infants in a negative manner. 17Additionally, underdeveloped metabolic pathways in infants influence drug metabolism and excretion. 32Moreover, incomplete neural myelination of the brain in young children, which develops rapidly during early childhood, contributes to seizure susceptibility. 33These factors contribute to an increased risk of seizures due to antihistamine use during infancy.Therefore, our finding that antihistamines are linked to seizure risk in children 24 months and younger suggests a potential developmental impact.Our results may suggest the need for careful consideration when prescribing antihistamines to infants or patients prone to seizures.In other words, these findings reinforce the importance of prudent and reasonable administration of first-generation antihistamines to vulnerable infants.
The variation in the risk of seizures associated with antihistamines among different age groups of children is likely related to brain development.We found that the older the children, the less pronounced the risk of seizures with the use of antihistamines.This is probably due to improved metabolism and a more mature brain, which are less affected by first-generation antihistamines. 32,34wever, antihistamines are available over the counter for children older than 24 months but require

JAMA Network Open | Neurology
First-Generation Antihistamines and Seizures in Young Children

(
NHIS) database in Korea, which covers the country's entire population.Data for individuals born in Korea between January 1, 2002, and December 31, 2005, and followed up until December 31, 2019, or until participants became ineligible for health care insurance were considered.The NHIS database offers essential demographic details, such as birth date, sex, insurance premium, and region of residence.It also includes information on use of health care services, hospital visit types, International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), diagnosis codes, prescribed medication codes, and procedure codes.Only deidentified individual data were used, adhering to the National Health Insurance Act 17 and ethical guidelines.The study protocol was reviewed and approved by the Institutional Review Board of the Korea National Institute for Bioethics Policy, which waived the need for informed consent because of the use of JAMA Network Open | Neurology First-Generation Antihistamines and Seizures in Young Children

Figure 1 .
The data of 1 893 314 children born in Korea between 2002 and 2005 were analyzed.From this cohort, children whose birthdate records were missing and those who died were excluded.The focus of the study was on children who had a seizure event, totaling 12 923 individuals.Additionally, those who had a seizure event at younger than 6 months (n = 678) and those who had not been prescribed first-generation antihistamines before the seizure event (n = 516) were excluded.The final analysis included only those children who were prescribed first-generation antihistamines prior to their seizure event, amounting to 11 729 individuals.

Figure 1 .
Figure 1.Flowchart of Participant Selection

Figure 2 .
Figure 2. Subgroup Analysis of the Risk for Seizure Events Associated With First-Generation Antihistamines

94 052 Excluded 85 394 No record of birth date 8658 Death
Discordant participants by time window 15 d: 3178 Participants 10 d: 2546 Participants 5 d: 1651 Participants 1194 Excluded 678 Aged <6 mo at seizure event 516 Not prescribed first-generation antihistamines prior to seizure event Children were born in Korea from January 1, 2002, to December 31, 2005.A seizure event of interest in this study was defined as an emergency department visit with a principal diagnosis of epilepsy, status epilepticus, or convulsion.

Table 1 .
Baseline Characteristics of the Study Population at the Index Date With 15 Days of Window Period a X), or convulsion (ICD-10 code R56.8).cOwing to missing data for 44 participants (1.4%), percentages do not total 100.d Owing to missing data for 61 participants (1.9%), percentages do not total 100.