Baseline Nutritional Status Predicts Ketogenic Diet Efficacy in Pediatric Drug-Resistant Epilepsy
核心洞察
A single-center study of 43 children with drug-resistant epilepsy (搜索) found that baseline nutritional status influences ketogenic diet (搜索) (KD) therapy outcomes, with 25.6% achieving ≥70% seizure reduction within 3 months.
Adequate pre-KD intake of magnesium, zinc, and vitamin E was significantly associated with ≥70% seizure reduction at 6 months, with robust odds ratios across adjusted models.
At 3 months, higher n-3/n-6 PUFA intake and adequate phosphorus were linked to better seizure control, suggesting a temporal shift in nutritional predictors during therapy.
A retrospective cohort study from National Taiwan University Hospital (NTUH) suggests that a child's nutritional status before starting a ketogenic diet (搜索) (KD) may predict how well the dietary therapy controls drug-resistant epilepsy (搜索). Among 43 children enrolled between April 2008 and August 2021, adequate baseline intake of magnesium, zinc, and vitamin E was significantly associated with achieving a 70% or greater reduction in seizure frequency at six months, while higher polyunsaturated fatty acid (PUFA) and phosphorus intake correlated with early response at three months.
The findings, published in Pediatric Research, address a notable gap in epilepsy research: while extensive literature documents micronutrient deficiencies that develop during ketogenic diet (搜索) therapy, systematic examination of baseline nutritional profiles as potential predictors of treatment outcomes has remained largely absent.
Study Design and Patient Cohort
Investigators enrolled 43 children (25 females, 18 males) with drug-resistant epilepsy (搜索) who received KD therapy at NTUH. The mean age at KD initiation was 5.41 ± 4.65 years. In the cohort, 25.58% had focal epilepsy, 20.93% had generalized epilepsy, and 53.49% had both focal and generalized epilepsy. A total of 79.07% of patients had comorbidities, and 25.6% had undergone prior ACTH/VNS therapy. The Modified Atkins Diet (搜索) (MAD) was the most commonly used KD type (n = 23, 53.5%).
Dietary intake prior to KD initiation was evaluated using 24-hour dietary recall and 3-day food records, with nutrient insufficiency defined as intake below 100% of the Taiwanese Dietary Reference Intakes (DRIs), 8th edition. Seizure types and epilepsy syndromes were classified according to International League Against Epilepsy (ILAE) guidelines.
Early Response and PUFA/Phosphorus Intake
Within the first 3 months of KD therapy, 11 children (25.6%) achieved a ≥70% seizure reduction. Among the 21 individuals (48.8%) who remained on the KD for over 2 years, 15 (71.4%) had seizure reduction exceeding 70%, including 3 patients with infantile spasms (搜索), 4 with Lennox-Gastaut Syndrome (搜索), and 1 with post-encephalitic epilepsy.
At the 3-month mark, subjects with ≥70% seizure reduction showed significantly lower rates of insufficient phosphorus intake (0% vs. 31.25%, p = 0.043) and lower rates of n-6 PUFA below the Acceptable Macronutrient Distribution Range (AMDR) (0.00% vs. 37.50%, p = 0.019). Mean energy intake percentages for n-3 PUFA (2.23% vs. 0.78%, p = 0.021) and n-6 PUFA (8.98% vs. 5.87%, p = 0.018) were also significantly higher in responders.
Six-Month Outcomes: Magnesium, Zinc, and Vitamin E
By six months, the nutritional predictors shifted. Among 32 patients stratified by seizure reduction, responders (≥70% reduction, n = 17) showed significantly lower rates of insufficient intake for magnesium (47.1% vs. 86.7%, p = 0.019), zinc (11.8% vs. 53.3%, p = 0.021), and vitamin E (17.6% vs. 53.3%, p = 0.034) compared with non-responders (n = 15).
Logistic regression analysis revealed strong associations between adequate micronutrient intake and ≥70% seizure reduction. Magnesium showed odds ratios (ORs) of 8.46 (95% CI: 1.33–53.67, p = 0.023) in a sex/gender-adjusted model, 9.00 (95% CI: 1.18–68.48, p = 0.034) when further adjusted for epilepsy etiology, seizure type, and KD type, and 22.24 (95% CI: 1.50–330.49, p = 0.024) in a fully adjusted model. Zinc ORs were 10.57 (95% CI: 1.58–70.89, p = 0.015), 12.96 (95% CI: 1.56–107.47, p = 0.018), and 10.73 (95% CI: 1.29–89.08, p = 0.028) across the three models. Vitamin E ORs were 5.35 (95% CI: 1.05–27.25, p = 0.043), 8.69 (95% CI: 1.21–62.53, p = 0.032), and 9.01 (95% CI: 1.11–72.92, p = 0.039).
The authors note that magnesium likely reduces seizure frequency by modulating NMDA receptors and GABA synthesis, while zinc and vitamin E, as antioxidants, may mitigate oxidative stress linked to epileptogenesis. Seizure reduction and nutritional outcomes did not differ significantly among patients receiving formula, G-tube prepared meals, or oral intake (p > 0.05).
A Temporal Shift in Nutritional Predictors
The investigators observed a distinct temporal shift in nutritional predictors. At 3 months, phosphorus and PUFA predominated, potentially driven by early-stage cellular energy (ATP) demands and membrane/inflammatory regulation. By 6 months, magnesium, zinc, and vitamin E took over, indicating a mechanistic transition from acute cellular reconstruction to long-term systemic maintenance. During this later phase, zinc and magnesium support prolonged tissue remodeling as critical enzyme cofactors, while vitamin E provides antioxidant protection against chronic lipid peroxidation.
Study Limitations
The authors caution that their findings should be interpreted carefully given several methodological limitations. The use of caregiver-reported dietary data is vulnerable to recall bias and underreporting, and the absence of serum nutrient level measurements—dictated by logistical and cost constraints—may affect interpretation of the observed associations. Dietary heterogeneity, specifically the use of both MAD (53.5%) and classical KD (46.5%), could introduce confounding metabolic effects. The non-blinded nature of outcome assessments may have introduced observer bias. The small sample size restricted statistical power, as evidenced by wide confidence intervals and model instability in some multivariate analyses. Finally, the single-center design and reliance on Taiwanese DRIs limit generalizability to ethnically or culturally diverse pediatric populations internationally.
Clinical Implications
The researchers recommend assessing magnesium, zinc, and vitamin E status using weighed food records collected over a typical week (2 weekdays and 1 weekend day) and dietary pattern questionnaires as part of pre-KD nutritional assessment. They conclude that sufficient intake of magnesium, zinc, vitamin E, phosphorus, and n-3/n-6 PUFAs may serve as a potentially important predictor for optimizing seizure control, while emphasizing that confirming these predictors requires larger multicenter cohorts and randomized controlled trials with blood micronutrient monitoring.