Epilepsy as a consequence of infantile vitamin B12 deficiency: observations from a prospective cohort study and a review of the literature

Introduction

Infantile vitamin B12 deficiency is commonly encountered in predominantly vegetarian regions of the Indian subcontinent (1). The deficiency typically affects an exclusively or predominantly breastfed child of a vegetarian mother. The typical cutaneous and hair manifestations, pallor, with developmental delay or regression and variable presence of movement disorders, helps in clinical identification of this syndrome, also called the Neurocutaneous Infantile B12 deficiency Syndrome (NIBS) (1, 2). There has been a growing body of evidence suggesting the long-term neurodevelopmental consequences of B12 deficiency in infancy despite treatment and normalization of blood levels of B12 and homocysteine (3–5). The commonly encountered consequences are poor head growth leading to microcephaly, developmental delays of variable degrees, autistic traits, epilepsy, and, rarely, death (1, 3–5). The long-term presence of these neurodevelopmental morbidities in children with B12 deficiency underscores the critical role of vitamin B12 in brain growth, maturation, and developmental progression (6).

Epilepsy, if seen as a sequel, may add to the developmental morbidities seen in these children. The description of epilepsy in children treated for infantile B12 deficiency is limited to case reports, case series, and small retrospective observational studies (7–9). Hence, we aimed to determine the incidence of epilepsy by following up a previously published prospective cohort of children with symptomatic vitamin B12 deficiency. The objectives were to evaluate the seizure and epilepsy types, neuroimaging and electroencephalography findings, and response to treatment in children who developed epilepsy.

Materials and methods

This hospital-based cohort study was a follow-up of the prospective cohort study, conducted from January 2021 to June 2023 at a tertiary care hospital for children in India, aimed at studying the neurological consequences of Infantile vitamin B12 deficiency (3). The study was approved by the Institute’s ethics committee (Ref.no. INT/IEC/2021/SPL-707, INT/IEC/2022/SPL-127, IEC-INT/2026/Study-4159). There were 141 children enrolled in the study. This cohort consisted of consecutive children with proven Infantile vitamin B12 deficiency. All children in the parent study (3) had a baseline assessment of clinical symptoms and signs, including anthropometric indices, skin and hair changes, and neurological findings. Also, a baseline development assessment using CAPUTE scales: Cognitive Adaptive Test (CAT)/ Clinical Linguistic & Auditory Milestone (CLAMS) was available for all children. Baseline hemogram (measured by an automated cell Counter - HmX500, Beckman Coulter, USA/XN-1000, Sysmex, Japan), serum vitamin B12 levels (measured by competitive immunoassay using ADVIA Centaur and ADVIA Centaur XP systems), and plasma homocysteine levels (measured by competitive immunoassay using ADVIA Centaur system) were available in all children. The lower normal limit of serum vitamin B12 was 211 pg/ml, and the higher normal limit of plasma homocysteine levels was 15 mmol/L. Data on baseline Magnetic Resonance Imaging (MRI) of the brain, consisting of the following sequences: 3D T1-weighted axial spoiled gradient recalled, T2-weighted axial, Fluid Attenuated Inversion Recovery (FLAIR) axial, diffusion weighted imaging (DWI) sequences, done in a 3.0 Tesla MR Scanner using a standard 32-channel head coil (Ingenia, Philips Healthcare, Netherlands), was available for all children, and a follow-up imaging done after 3 months or more after the baseline MRI.

All surviving children were followed up till March 2026. During the follow-up period, children who developed epilepsy were evaluated. The diagnosis of epilepsy was made in those who fulfilled the 2014 International League Against Epilepsy criteria for diagnosis of Epilepsy [(i) at least two unprovoked (or reflex) seizures that occur more than 24 h apart; (ii) one unprovoked (or reflex) seizure and a probability of additional seizures that is comparable to the general recurrence risk (at least 60%) following two unprovoked seizures, occurring over the next 10 years; and (iii) an epilepsy syndrome diagnosis] (10). Age at onset of epilepsy, epilepsy type/syndrome were recorded. At least one electroencephalogram was performed for 30 min or more on all children who developed epilepsy in follow-up. Records of the electroencephalogram, therapy details, and response to treatment were reviewed. Those who were seizure-free for at least 12 months with or without anti-seizure medications were labeled as having controlled epilepsy. Those with seizure recurrence despite treatment with two or more anti-seizure medications, excluding hormonal therapy, were labeled as drug-refractory. The collected data were entered into the Microsoft Access database.

Statistical analyses were performed using IBM SPSS Statistics for Windows, version 29.0.1.0(171) (IBM Corporation, Armonk, New York, USA). Descriptive data were expressed as numbers and proportions. Numerical data were expressed as Mean ± Standard Deviation (SD) or Median and Interquartile ranges (IQR). Paired t-test was employed to analyze paired dependent quantitative variables, unpaired t-test was used to analyze independent quantitative variables, and the chi-square test or Fisher’s Exact test was used to analyze two or more qualitative variables. A p-value of less than 0.05 was considered significant. Missing follow-up developmental quotient (DQ) values were addressed using Multiple Imputation with 20 fully conditional specifications using predictive mean matching. Time-to-epilepsy was evaluated using Kaplan–Meier curves and multivariable Cox proportional hazards regression, with variables with univariate p ≤ 0.20 considered for model entry. Analyses were pooled using Rubin’s rules, and proportional hazards assumptions were confirmed through log–minus–log plots and time-dependent covariates. Time-to-event analysis was conducted to evaluate cumulative epilepsy-free survival across stratified patient cohorts. The primary outcome was the time to an epileptic event, with patients remaining event-free at the conclusion of observation being right-censored at their last recorded follow-up date. Non-parametric Kaplan–Meier curves were constructed to estimate epilepsy-free survival across the group stratifications that emerged after the multivariable Cox proportional hazards regression, namely, Acute-phase seizures present or not, and groups dichotomized based on the median of the observed DQ.

For the literature review, a standard search strategy was adopted. A thorough search of electronic databases, including PubMed and Embase, was performed for studies on epilepsy in Infantile vitamin B12 deficiency. The MeSH terms used for the search strategy were Infantile vitamin B12 deficiency [OR] Infantile tremor syndrome [OR] Vitamin B12 deficiency in infancy [OR] Vitamin B12 deficiency in children [AND] epilepsy [OR] seizures [OR] epileptic spasms [OR] spasms [OR] West syndrome [OR] Infantile Epileptic Spasms Syndrome [OR] IESS. The search returned a total of 757 articles, including case reports, case series, and observational studies. After removing duplicates, articles were screened to exclude the following: (i) published abstracts presented in conferences, (ii) non-English articles, (iii) articles that reported seizures or epileptic spasms at the time of diagnosis of infantile vitamin B12 deficiency and not during follow-up, (iv) those that reported characteristics of epilepsy associated with inherited cobalamin pathway defects, and other genetic causes related to vitamin B12 absorption and metabolism, and (v) unrelated articles. After exclusion, eight articles reporting on the development of epilepsy and/or outcomes in children treated for infantile vitamin B12 deficiency were included in the final review (Supplementary Figure 2). Supplementary Table 1 summarizes available information on the onset and characteristics of epilepsy in follow-up of treated children from previous studies.

Results

The cohort of 141 children who were treated for Infantile B12 deficiency was followed through March 2026 to assess epilepsy incidence and outcomes. Four died during the initial phase of follow-up, and 25 were lost to follow-up. Of the remaining cohort of 112 children, two died due to epilepsy. The remaining 110 children had a median follow-up of 51 months (IQR: 45–55), with similar medians in those with epilepsy [50 (IQR: 45–54)] and those without epilepsy [52 (IQR: 45–56)]. Thirty-one children [28%; (95%CI: 19–36%)] developed epilepsy, of whom two had died due to epilepsy-related complications (Supplementary Figure 1).

The baseline characteristics, including age, gender, growth indices, pertinent clinical features, serum B12 and plasma homocysteine levels, brain MRI findings at the time of B12 deficiency diagnosis, and baseline neurodevelopmental assessment, are summarized in Table 1. In univariate analysis, factors significantly associated with epilepsy occurrence were the presence of seizures at diagnosis and a basal ganglia signal change on MRI at diagnosis (Supplementary Table 2). The children with a diagnosis of epilepsy had a significantly lower mean developmental quotient at follow-up, even prior to the onset of epilepsy.

Table 1. Summary of characteristics of children with infantile B12 deficiency at diagnosis who developed epilepsy on follow-up (N = 31).

* Done median 120 days (IQR 98–200) after the first MRI.

The mean age at the onset of epilepsy was 19.0 ± 7.1 months. The mean interval to epilepsy onset from diagnosis of Infantile B12 deficiency was 4.7 ± 3.2 months. Generalized seizures were seen in 24 of the 31 children (77.4%) and focal seizures in 7 children (22.6%). The most common generalized seizure type was epileptic spasms, seen in 15 children (48.4%) who developed epilepsy. Generalized clonic seizures were seen in 6 children (19.4%), and generalized tonic–clonic seizures were seen in 3 children (9.7%). All the 7 children with focal seizures had focal clonic type of seizures. A single type of seizure was seen in all 31 children at the onset of epilepsy (Table 2).

Table 2. Summary of epilepsy characteristics and outcomes in children treated for Infantile B12 deficiency.

All 15 (48.4%) with IESS had hypsarrhythmia as the interictal finding, with abnormal background and poor sleep architecture for age. Generalized interictal epileptiform discharges (IEDs) were seen in 9 children (29%), and focal IEDs were seen in 7 children (22.6%) (Table 3).

Table 3. Interictal EEG findings in children treated for Infantile B12 deficiency developing Epilepsy in follow-up (N = 31).

The brain MRI done at the time of diagnosis showed abnormalities in 30 (96.8%) of the children. A follow-up brain MRI was performed in 30 children with abnormalities on baseline imaging, after a median interval of 120 days from the first MRI, showing residual abnormalities in a sizeable proportion (Table 1).

The 15 children with IESS were treated with intramuscular injections of Adrenocorticotropic hormone (ACTH), and one out of the 15 with IESS required the addition of vigabatrin in view of persistent epileptic spasms despite ACTH therapy. All of those with IESS had a documented electro-clinical resolution of epileptic spasms after therapy. Of those with IESS, 4 did not require any other therapy apart from ACTH and remain seizure-free. Among the rest, 5 developed drug-refractory epilepsy with other seizure types (Supplementary Table 3). Among those with generalized epilepsy, two of the nine developed drug-refractory epilepsy, while one of the seven with focal epilepsy developed drug-refractory epilepsy (Supplementary Figure 1). Overall, 31 [28%; (95% CI: 19–36%)] of 112 children developed epilepsy, 21 (68%) had either resolved or well-controlled epilepsy, 8 (26%) had drug-refractory epilepsy, and 2 (6%) died.

In the multivariable Cox proportional hazards model, acute-phase seizures were strongly associated with increased epilepsy risk (pooled HR 4.27, 95% CI: 1.6–11.5, p = 0.004), indicating that children with seizures at the time of diagnosis had more than a fourfold higher hazard of developing epilepsy compared with those without seizures. Higher follow-up developmental quotient (DQ) was independently protective (pooled HR 0.94, 95% CI: 0.89–0.99, p = 0.015), with each 1-point increase in DQ reducing epilepsy hazard by approximately 6%. MRI signal abnormalities, cortical atrophy, sex, socioeconomic status, tremors, and age were not significant predictors in the imputed model and were excluded from the final adjusted model. The proportional hazards assumption was satisfied for both retained predictors, supported by parallel log–minus–log curves and non-significant time-interaction terms. Kaplan–Meier analysis showed significantly lower epilepsy-free survival in children with acute-phase seizures and in those with a developmental quotient below the median value of 46, with both comparisons demonstrating clear early separation of survival curves (log-rank p < 0.001 for seizures; p = 0.021 for DQ). These patterns were consistent with the multivariable Cox findings, where seizures and follow-up DQ remained the only independent predictors of epilepsy (Figure 1). The Kaplan–Meier curves also demonstrated clear early clustering of epilepsy events within the first 10 months after diagnosis.

Discussion

This study shows that epilepsy is an important long-term consequence of symptomatic infantile vitamin B12 deficiency. Among those who survived the initial illness, 28% later developed epilepsy, and two died from epilepsy-related complications. Acute-phase seizures were the most powerful predictor of subsequent epilepsy, conferring more than a fourfold increase in risk. The Kaplan–Meier curves demonstrated clear early clustering of epilepsy events within the first 10 months, indicating that the period immediately after diagnosis and treatment represents the highest-risk window for epileptogenesis. Similarly, children with higher DQ values showed consistently better epilepsy-free survival than those with lower DQ, possibly indirectly indicating the relationship of severity and duration of deficiency with long-term epilepsy risk.

The median age of onset of epilepsy, with only Epileptic spasms being the only reported type of epilepsy in the available literature in the previously reported patients, was 10.5 months, around 3 months after the diagnosis of vitamin B12 deficiency. In our cohort, epilepsy began at a mean age of 19.9 months, approximately 4.7 months after the diagnosis of B12 deficiency. The mean age of onset of epilepsy was later in our cohort compared to 6.5 months in the largest retrospective cohort by Sharawat et al. (9). The major caveat in the interpretation of this slightly later age at onset of epilepsy in our cohort with the previously published patients is that the previously available patient data are mostly case reports or series, which are subject to reporting and publication bias. This timing suggests that epilepsy may be a delayed consequence of metabolic brain injury caused by early deficiency. Only 5 of the 31 children had seizures during the acute, untreated phase. The interval between nutritional correction and seizure onset suggests that early B12 deficiency triggers neurodevelopmental vulnerabilities such as delayed myelination, impaired synaptic maturation, and increased cortical excitability that become clinically apparent only months later (11, 12). This interpretation is consistent with neuroimaging and neurophysiological evidence showing that the effects of early cobalamin deficiency can persist well beyond hematologic recovery (3). The mechanisms underlying epilepsy and epileptic spasms in these children remain uncertain because this subgroup has been poorly studied. In rat models, pretreatment with vitamin B12 reduces seizure frequency in PTZ-kindled animals, suggesting a possible antiepileptogenic effect of adequate B12 supplementation (13). Evidence from animal studies and adult reports also suggests that B12 deficiency-related myelin dysfunction may enhance glutamate signaling, the brain’s main excitatory neurotransmitter, thereby promoting epileptogenesis (11, 12). Chronic methylcobalamin use has also been shown to protect cortical neurons from NMDA receptor-mediated glutamate toxicity (14). Taken together, these observations support the possibility that vitamin B12 deficiency increases susceptibility to epilepsy in children. Seizure presentation in this cohort was predominantly generalized: 77% of affected children had generalized seizures, and nearly half presented with epileptic spasms or infantile epileptic spasms syndrome (IESS). This prominence of spasms is notable because IESS is usually linked to structural, metabolic, or genetic causes. Early B12 deficiency may act as a global encephalopathic insult, predisposing children to epileptic spasms in a way similar to other early brain injuries. Encouragingly, all 15 children with spasms achieved electroclinical resolution, suggesting that despite presentation with spasms, this phenotype may respond better to treatment than spasms caused by classic structural etiologies. Current evidence on epilepsy after treatment of infantile B12 deficiency is limited largely to case reports, small case series, and retrospective studies (7, 15–19) (Supplementary Table 1). In our cohort, epileptic spasms were the most common form of epilepsy, similar to patterns seen after other early brain insults and in keeping with the scarce literature on the subject reviewed by us (Supplementary Table 1). In addition, lower vitamin B12 levels have been observed in children with epileptic spasms compared with children who have developmental delay without spasms, suggesting a possible link (20). In univariate analysis, epilepsy was significantly associated with seizures at diagnosis and basal ganglia signal changes on MRI at diagnosis (Supplementary Table 2). From a clinical standpoint, seizures during the acute phase are particularly important because affected children may remain at risk of developing epilepsy even after apparent recovery from the vitamin deficiency, warranting closer follow-up in these children. Notably, children who later developed epilepsy already had significantly lower developmental quotients at follow-up before epilepsy onset. This, along with the involvement of the basal ganglia, which indicates a final energy failure in severe vitamin B12 deficiency, suggests that epilepsy may be a marker of the severity of the deficiency. The presence of drug-refractory epilepsy and epilepsy-related deaths underscores the severity of this condition and highlights the importance of early diagnosis, timely nutritional rehabilitation, and close neurological follow-up.

The findings of this study have important clinical implications. Firstly, children treated for infantile vitamin B12 deficiency require ongoing neurological monitoring, particularly during the first few months after treatment, because epilepsy may develop after a delay. Secondly, the high rate of epileptic spasms suggests that clinicians should have a low threshold for performing an EEG in children with developmental regression or new-onset seizures following B12 deficiency. Those who had seizures at the time of diagnosis appear to be at risk for later epilepsy. In addition, although seizure control is often possible in most, the presence of drug-refractory epilepsy and epilepsy-related deaths underscores the severity of this condition. Finally, despite all care, developmental outcomes remain a concern, underscoring the need for early developmental support and long-term neurocognitive follow-up.

Acute-phase seizures were the most powerful predictor of subsequent epilepsy, conferring more than a fourfold increase in risk. This strong association reinforces the concept that early seizures reflect greater underlying brain vulnerability and a heightened propensity for epileptogenesis, as well as a longer period of deficiency impairing protective mechanisms. Conversely, a higher follow-up DQ was independently protective, with each incremental improvement reducing epilepsy hazard. The lack of significant associations with MRI signal abnormalities, cortical atrophy, tremors, or age indicates that, within this population, acute-phase seizures and developmental status exert a dominant influence over structural or demographic factors. The early and persistent separation of Kaplan–Meier curves complements the adjusted hazard estimates and underscores the clustering of epilepsy events among clinically vulnerable subgroups.

Several limitations should be considered when interpreting the findings of this study. Although the cohort initially included 141 children, only 112 had evaluable outcomes. With only 31 children developing the outcome of interest, the study has limited statistical power to estimate incidence with high precision; the resulting confidence intervals are wide, and small shifts in the number of events could meaningfully alter the incidence estimate. This also limited the number of covariates that could be included in the Cox model and may reduce the stability of estimates for less common predictors. The outcomes of the children lost to follow-up remain unknown, and even a modest number of additional events among them could substantially change the observed incidence and its interpretation. This may have influenced time-to-event estimates despite censoring. Multiple imputation was used to address missing developmental data, but imputed values may not fully capture the true variability of neurodevelopmental trajectories. Secondly, some associations we observed may be chance findings, particularly given the multiple comparisons performed. Third, epilepsy in early childhood is dynamic: diagnoses may evolve, seizure patterns may change, and treatment responses may vary over time. Children who were seizure-free at assessment may later develop epilepsy, whereas those in remission may relapse. These changes could influence outcome patterns and limit the stability of current inferences. Overall, these limitations highlight the need for a well-designed, adequately powered cohort with systematic long-term follow-up and minimal attrition to define more clearly the course of epilepsy and neurodevelopmental outcomes after infantile vitamin B12 deficiency.

In summary, epilepsy developing after infantile vitamin B12 deficiency appears to be an underrecognized and potentially preventable cause of childhood epilepsy. This is especially important in regions where nutritional deficiency remains common. From a public health perspective, prevention of infantile B12 deficiency through maternal supplementation, dietary counseling, and early screening of high-risk groups may be the most effective way to reduce the long-term neurological sequelae observed in this cohort.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. During the preparation of this work the authors used Chat GPT, and MS copilot in order to improve the readability and language of the manuscript and to create figures. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2026.1903748/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Jagdish Chandra, ESIC Medical College, Faridabad, India

Reviewed by: Haider Mohammed Majeed, University of Baghdad, Iraq

Jatinder Goraya, CLIO Mother and Child Institute, India