Meta-analysis: dietary interventions have small effect in ASD

  • Fraguas D & al.
  • Pediatrics
  • 04/10/2019

  • Susan London
  • Clinical Essentials
L'accesso ai contenuti di questo sito è riservato agli operatori del settore sanitario italiano L'accesso ai contenuti di questo sito è riservato agli operatori del settore sanitario italiano


  • Dietary supplementation overall and of various types netted significant but small improvements in symptoms, function, and clinical domains among predominantly young patients with autism spectrum disorder (ASD).

Why this matters

  • Lack of effective pharmacologic treatment, combined with noninvasiveness and popularity of dietary interventions.

Key results

  • Mean participant-weighted intervention duration: 10.6 weeks.
  • With placebo as comparator, improvements were seen in selected symptoms, functions, and clinical domains with:
    • Dietary supplementation overall.
    • Vitamin supplementation alone.
    • Omega-3 supplementation alone.
  • Greatest effect sizes for:
    • Dietary supplementation overall on:
      • Anxiety and/or affect (Hedges’ g, 0.482; P=.007);
      • Behavioral problems and impulsivity (Hedges’ g, 0.482; P=.001).
    • Vitamin supplementation alone on:
      • Stereotyped, restricted, repetitive behaviors (Hedges' g, 0.531; P=.009);
      • Global severity (Hedges’ g, 0.464; P=.038);
      • Hyperactivity, irritability (Hedges’ g, 0.426; P=.003).
  • Most significant effect sizes were small, with a mean Hedges’ g of 0.31.

Expert comment

  • In a commentary, Amy S. Weitlauf, PhD, and coauthors write, “Although high-quality trials of specific interventions are available, we should be cautious when interpreting preliminary results or risk continuing to support intervention choices, some of which hold potential for harm, on the basis of anecdotal evidence.”

Study design

  • Meta-analysis of 27 randomized controlled trials of dietary interventions among 1028 mainly pediatric patients with ASD (mean age, 7.1 years).
  • Main outcomes: 17 symptom and/or function groups; 4 clinical domains.
  • Funding: Spanish Ministry of Science, Innovation, and Universities, Instituto de Salud Carlos III.


  • Heterogeneity of interventions, measures, patients across trials.
  • Scant data for some interventions.
  • Majority of studies did not assess baseline nutritional deficits.