Pelsser · Lancet (London, England) 2011 · randomized controlled trial with double-blind crossover phase · n=100

Effects of a restricted elimination diet on the behaviour of children with attention-deficit hyperactivity disorder (INCA study): a randomised controlled trial.

Cited 214 times in the scientific literature.

Level 2 - randomized trial

Individual randomized controlled trial with a double-blind crossover phase

PubMed 21296237 · doi:10.1016/S0140-6736(10)62227-1 · record verified 2026-08-29

What was done

Children aged 4–8 years with ADHD in the Netherlands and Belgium were randomized to a 5-week restricted elimination diet (n=50) or healthy diet instructions (n=50) in an open-label phase with masked pediatrician assessments. Diet responders (≥40% improvement on the ADHD rating scale [ARS]) then entered a 4-week double-blind crossover food challenge phase adding high-IgG or low-IgG foods based on baseline testing. Primary endpoints were masked ARS score and unmasked abbreviated Conners' scale (ACS) score changes.

What was found

In phase 1, the mean difference between diet and control groups was 23.7 points on the ARS total score (95% CI 18.6–28.8; p<0.0001) and 11.8 points on the ACS score (95% CI 9.2–14.5; p<0.0001). Following the double-blind food challenge in responders, ARS score increased by 20.8 points (95% CI 14.3–27.3; p<0.0001) and ACS score increased by 11.6 points (95% CI 7.7–15.4; p<0.0001). Relapse occurred in 19 of 30 children (63%), independent of IgG blood test levels. No harms or adverse events were reported.

Why it matters

This trial provides evidence that an elimination diet can substantially improve behavior in a subset of children with ADHD, but demonstrates that IgG blood tests do not predict which foods trigger symptom relapse.

Limits

The initial 5-week dietary phase was open-label to participants and parents, introducing potential behavioral or reporting bias. The double-blind challenge phase had a small sample size (n=30), long-term feasibility and nutritional sustainability were not evaluated, and findings may not generalize to older age groups.

Cited by