Mental health

ADHD children 'should have diets assessed'

Scientists have suggested, “children with attention deficit hyperactivity disorder (ADHD)… should be put on a restricted diet for several weeks to establish whether particular foods are the cause,” reported The Guardian.

This news report is based on a trial that examined 100 young children (average age 6.9 years) with ADHD. The children were randomly allocated to either a five-week restricted diet consisting mainly of hypoallergenic (non-allergy producing) foods, or a control diet where parents were given advice to follow a healthy diet. On scores of ADHD symptoms, the restricted diet group showed reductions in symptoms of 53.4%, while there was little difference in the control group (2.7% reduction).

The causes of ADHD are not established but genetic and environmental factors are thought to play a role. This study suggests that restricting certain food substances could improve symptoms in some children. Importantly, if parents of children with ADHD want to investigate if this is the case for their child, they are advised to consult their doctor first, rather than choosing which foods to eliminate themselves.

Where did the story come from?

The study was carried out by researchers from the ADHD Research Centre and other institutions in the Netherlands. Funding was given by the Foundation of Child and Behaviour, the Foundation Nuts Ohra, the Foundation for Children’s Welfare Stamps Netherlands, and the KF Hein Foundation. The authors report affiliations with several pharmaceutical companies including Janssen Cilag, Eli Lilly, Bristol-Myers Squibb, Schering Plough, UCB, Shire, Medice, and Servier. The study was published in The Lancet , a peer-reviewed medical journal.

The research has generally been well represented in the news.

What kind of research was this?

This is a randomised controlled trial designed to investigate the effects of a ‘restricted elimination diet’ in children with ADHD. A restricted elimination diet is when the diet is reduced to a few basic foods and then gradually broadened to include other foods, to see which foods have an effect on a person. The researchers wanted to see whether there was any association between diet and behaviour.

A randomised controlled trial is the best way of investigating the effect of an intervention on an outcome such as behaviour change. However, this sort of study also benefits from ‘blinding’, when the participants are unaware of what intervention they are receiving.

In this study it was not possible to blind the children and their parents as they would have known which foods the children were restricted from eating. These results also cannot tell us what the longer-term effects of a restricted diet would have on ADHD. In addition, although this study can tell us about the effect of diet on existing ADHD symptoms, it cannot conclude that diet is the only cause of these children’s ADHD.

What did the research involve?

This study, called the Impact of Nutrition on Children with ADHD (INCA), recruited 100 children from health centres across the Netherlands and Belgium. The children recruited met diagnostic criteria for ADHD, were aged four to eight years, and had families that were willing to follow a five-week restricted diet. Children who were already receiving drug, behavioural or diet treatment for ADHD were excluded. The trial was conducted in two phases.

In the first phase, 50 children were randomly assigned to what was described as ‘an individually tailored’ restricted diet, and 50 were given advice to follow a healthy and balanced control diet. The children and parents knew which diet they were assigned to but in some cases, the researcher assessing the outcomes of the study was blinded to which diet the children had been on.

The restricted diet included only a small number of foods such as rice, turkey, lamb, a range of vegetables (lettuce, carrots, cauliflower, cabbage and beet), pears and water. These foods were selected because they were hypoallergenic (non-allergy producing).

The children’s diets were also supplemented with specific foods such as potatoes, fruits and wheat on an individual basis to make it easier for them and their parents to comply with the restrictive diet. These additional foods were removed if the child showed no improvement after two weeks of the diet. The children were also given a non-dairy rice drink with added calcium to ensure that they did not develop calcium deficiency.

In the second four-week phase, the children who had responded to the restricted diet (as demonstrated by at least a 40% decrease in symptoms) were randomised to one of two dietary challenges. Each challenge involved a different set of foods: one group was given foods expected to induce a hypersensitivity/allergic response if the child was susceptible, and the other group were given foods that are not expected to induce allergic response. This section of the study was a crossover trial, where both groups started one set of foods then after two weeks they swapped over and received the other set of foods for the remaining two weeks. During this phase of the trial both parents, children and assessors did not know whether the food challenges being given were likely to induce an allergic response.

The children’s ADHD symptoms were assessed using two main questionnaires: the 18-item ADHD Rating Scale (ARS, score range 0 to 54) and the 10-item Abbreviated Conners’ Scale (ACS, score range 0 to 30). These tests were conducted at study start (baseline), after the eight-week diet, after the first two weeks of the food challenge phase, and then after the remaining two weeks of the food challenge (when they had switched over to the other diet).

Due to the children’s young age, the ARS and ACS tests were completed by both parents and teachers (rather than the child). The ARS test was conducted by a researcher who was blinded to the dietary groups, whereas the ACS was only conducted by a researcher who was aware which diet the child had been given.

The main outcomes of interest to the researchers were changes in ADHD symptoms from study start to the end of the first restricted diet phase, and differences between the end of the first phase and the end of the second phase. They also assessed the effect of the food hypersensitivity challenge on the child’s immune system. This was assessed by measuring the children’s blood levels of antibodies (IgG) to specific foods at the start of the study and comparing them to measurements taken after the challenge phase when they could have been expected to have produced antibodies to any allergens.

What were the basic results?

Most of the children in the study were male (86%) with a mean (average) age of 6.9 years. Of the 50 children in the restricted diet group, 41 completed the first phase (82%). Of these, 32 (78%) were found to respond to the restrictive diet (show a reduction in their ADHD symptoms of at least 40%).

Between study start and the end of the first phase, the difference between the diet group and the control group in the mean ARS total symptom score was 23.7 points (95% confidence interval [CI] 18.6 to 28.8). There was a 53.4% reduction in score in the diet group (from mean score 45.3 at study start to 21.1 post-diet), and a 2.7% reduction in the control group (from mean score 47.6 at study start to 46.2 post-diet).

There was also a significant difference between the groups in mean ACS symptom score from study start to the end of the diet phase (11.8 point score difference between groups, 95% CI 9.2 to 14.5), with a 50.7% score reduction in the diet group compared with 0.3% reduction in the control group.

Thirty children who responded to the restrictive diet took part in the food challenge phase, 29 of whom completed it. After the food challenge, which involved foods expected to induce an immune response or not, the ARS total score increased by an average of 20.8 points (95% CI 14.3 to 27.3) and the ACS score increased by an average of 11.6 points (95% CI 7.7 to 15.4). In the challenge phase, relapse of ADHD symptoms occurred in 18 of 29 children tested (62%), but this was not related to whether they received foods expected to induce an immune response or their IgG blood levels.

How did the researchers interpret the results?

The researchers conclude that ‘a strictly supervised restricted elimination diet is a valuable instrument to assess whether ADHD is induced by food’.

The second part of the study tested ADHD scores after foods that might be expected to produce a sensitivity/allergic reaction were introduced to the diet. This found some foods were associated with a significant rise in symptom scores. However, the extent to which symptoms returned was independent of blood antibody (IgG) levels. As such, the researchers say that prescribing diets according to the child’s specific IgG blood results (i.e. being guided by which substances the child has blood antibodies against) should be discouraged.


This well-designed trial aimed to examine the effect of a five-week food restriction diet on ADHD symptoms. The study had several strengths, including that all children met diagnostic criteria for ADHD and that their symptoms at the start of the study and afterwards were assessed using validated and widely used assessment tools. The children and parents could not be blinded to the diet given, but some blinding was achieved because in some instances the assessors did not know which diet the children had been given.

However, as the assessments were largely based on parental reports of symptoms, the researchers acknowledge that ‘expectations of the parents cannot be fully ruled out as a possible cause of the behavioural improvements’. In addition, though the study was larger than previous similar studies, it was still relatively small. Ideally, larger studies would confirm its findings.

The study found a clear reduction in ADHD symptom scores in children following the five-week restriction diet. In comparison, the control group assigned to being given healthy dietary advice showed no reduction.

It is difficult to answer from this study questions such as the longer-term effects of dietary change, (for example, whether the foods would have to be permanently withdrawn and what effect this would have, or whether they could be gradually reintroduced). Although no adverse effects were seen during this eight-week study, longer-term use of a restrictive diet would need to be carefully monitored by dietitians and other health professionals to ensure that no nutrient deficiencies were occurring.

The causes of ADHD are not established and both genetic and environmental factors are thought to play a role. Although this study can tell us about the effect of diet on existing ADHD symptoms, it cannot tell us for certain whether diet alone contributed to causing these children’s ADHD or if other causative factors were also having an effect.

It should also be noted that this study examined only young children with ADHD (average age 6.9), so the findings may not apply to adolescents or young adults with ADHD. It also excluded those who were being treated with medication or behavioural therapies for ADHD, who may have had different results.

The diets were described as individually tailored and full details of the individual diets followed are not provided in the main publication. As such, this study has not ‘implicated’ particular foods or food substances. For now, it remains advisable for parents and carers of children with ADHD to be guided by the advice of their health-care providers, rather than trying to try to eliminate foods themselves.

Regarding the restriction of food, NICE 2008 guidance on ADHD recommends:

  • if there is a clear link, healthcare professionals should advise parents or carers to keep a diary of food and drinks taken and ADHD behaviour
  • if the diary supports a relationship between specific foods and drinks and behaviour, then referral to a dietitian should be offered
  • further management (for example, specific dietary elimination) should be jointly undertaken by the dietitian, mental health specialist or paediatrician, and the parent or carer and child or young person

NHS Attribution