Efficacy and Safety of Dietary Therapies for Childhood Drug-Resistant Epilepsy: A Systematic Review and Network Meta-analysis

JAMA Pediatr. 2023 Mar 1;177(3):258-266. doi: 10.1001/jamapediatrics.2022.5648.


Importance: Despite advances in the understanding of dietary therapies in children with drug-resistant epilepsy, no quantitative comparison exists between different dietary interventions.

Objective: To evaluate the comparative efficacy and safety of various dietary therapies in childhood drug-resistant epilepsy.

Data sources: Systematic review and network meta-analysis (frequentist) of studies in PubMed, Embase, Cochrane, and Ovid published from inception to April 2022 using the search terms ketogenic diet, medium chain triglyceride diet, modified Atkins diet, low glycemic index therapy, and refractory epilepsy.

Study selection: Randomized clinical trials comparing different dietary therapies (ketogenic diet, modified Atkins diet, and low glycemic index therapy) with each other or care as usual in childhood drug-resistant epilepsy were included. Abstract, title, and full text were screened independently by 2 reviewers.

Data extraction and synthesis: Data extraction was conducted following Preferred Reporting Items for Systematic Reviews and Meta-analyses reporting guideline. Cochrane risk-of-bias tool was used to assess the study quality. Effect sizes were calculated as odds ratio with 95% CI using random-effects model. The hierarchy of competing interventions was defined using the surface under the cumulative ranking curve.

Main outcomes and measures: Short-term (≤3 months) 50% or higher and 90% or higher reduction in seizure frequency and treatment withdrawal due to adverse events were the primary efficacy and safety outcomes.

Results: Of 2158 citations, 12 randomized clinical trials (907 patients) qualified for inclusion. In the short term, all dietary interventions were more efficacious than care as usual for 50% or higher seizure reduction (low glycemic index therapy: odds ratio [OR], 24.7 [95% CI, 5.3-115.4]; modified Atkins diet: OR, 11.3 [95% CI, 5.1-25.1]; ketogenic diet: OR, 8.6 [95% CI, 3.7-20.0]), while ketogenic diet (OR, 6.5 [95% CI, 2.3-18.0]) and modified Atkins diet (OR, 5.1 [95% CI, 2.2-12.0]) were better than care as usual for seizure reduction of 90% or higher. However, adverse event-related discontinuation rates were significantly higher for ketogenic diet (OR, 8.6 [95% CI, 1.8-40.6]) and modified Atkins diet (OR, 6.5 [95% CI, 1.4-31.2]) compared with care as usual. Indirectly, there was no significant difference between dietary therapies in efficacy and safety outcomes.

Conclusions and relevance: This study found that all dietary therapies are effective in the short term. However, modified Atkins diet had better tolerability, higher probability for 50% or higher seizure reduction, and comparable probability for 90% or higher seizure reduction and may be a sounder option than ketogenic diet. Direct head-to-head comparison studies are needed to confirm these findings.

Publication types

  • Meta-Analysis
  • Systematic Review

MeSH terms

  • Child
  • Diet, High-Protein Low-Carbohydrate*
  • Diet, Ketogenic* / adverse effects
  • Drug Resistant Epilepsy* / therapy
  • Humans
  • Network Meta-Analysis
  • Seizures
  • Treatment Outcome