Review by Joan Breakey, specialist dietitian working with diet and ADHD for 30 years
An important new study on Diet and ADHD was published in a key medical journal: The Lancet, in February this year: Effects of a restricted elimination diet on the behaviour of children with attention-deficit hyperactivity disorder (INCA study): a randomised controlled study. It was conducted by Pelsser of the ADHD Research Centre in the Netherlands and her team.
Overall the INCA study is another important contribution to the work available on diet and ADHD. It shows that diet changed unselected ADHD children to a statistically significant degree even though any family who were already using diet to help their child was excluded from the trial.
In this area of research there are important issues to be dealt with: the diet itself, the method, the possible mechanism, and the method of assessment. It is important to realise that each of these are more difficult to manage in diet studies than in usual drug research.
Diet is often seen as the ‘item being researched’ as if it was similar to one drug being tested. But diet means different things to different researchers. It can mean the Feingold Diet, or the more researched RPAH elimination diet, the more liberal Diet Detective Family Elimination Diet, the much stricter Few Foods Diet, a diet free of all additives or a diet just excluding six artificial colours and benzoate. Whatever the choice here, it is crucial that the diet used for the research lowers the intake of suspect foods and additives sufficiently to show an effect if there is one. These researchers wisely used the Few Foods Diet so it excluded many whole foods, particularly dairy and wheat, reported by others to cause reactions. By using only a few foods the diet also reduced foods high in the suspect food chemicals: salicylates, amines and glutamates, reported as causing reactions by other researchers. As well it excluded most suspect additives, as foods allowed were so limited that home cooking was necessary, and added flavours, spices or chocolate were not specifically allowed. Strict limitation of foods ensured that the amount of change in ADHD features satisfied necessary statistical methodology.
Here we note something important. Such strict diets are useful for research as they exclude so many foods, suspect food chemicals, and additives, that they will see a diet effect if there is one. However they are not a useful basis for long term diet management. This is where clinical research becomes important. In clinical research each clinician, preferably a dietitian, begins with one of the research diets and has patients trial the various exclusions to see what is tolerated and what is not, in the real world. In such research what is happening for the individual is emphasised rather than the change within a group in a research study. This has been ongoing work in Australia [including RCH Melbourne, RPAH Sydney, and my work in Brisbane] so evidence has accumulated about what additives, food chemicals, and whole foods most often cause reactions. In fact it has been the basis of my Diet Detective Process outlined in Are You Food Sensitive? so that the diet only excludes the chemicals, and foods that most often cause reactions. This means this diet is minimally strict and more workable in ADHD families. If individuals react to some additional food or additive this shows up as Diet Detective Work with individual food trials progresses.
The method is important as the diet effect is not additional in the way that a new medicine is in a drug trial. Change is measured with challenges from a baseline which is only reached after the exclusion of suspect foods in a screening diet for four weeks. Researchers may also use the initial screening diet to select those who improve at this time as they are the more likely responders. Time matters. The initial diet needs to be for 4 weeks in children over 5 years, and one week is necessary as washout after challenge days to ensure that the effect has cleared. These issues were attended to in the INCA study. It is interesting to note that where researchers only exclude some additives, but allow additive flavour, all whole foods and chocolate, they get sufficient change if the challenge dose is high enough to produce change. However the outcome is that the food industry minimise the significance of the results by reporting that the doses of additives in the trial are above the usual intake, while not reporting that additive flavours are used in ten times the dose of colours.
Method of assessment
Assessment is difficult as ADHD is a complex condition and can be confounded by comorbid oppositional defiant disorder [ODD]. These researchers rightly noted and included this. The parents and teachers were blinded to a useful degree to the initial elimination diet as the good nutrition diet option is seen by many as perhaps being useful, as it would have greatly reduced high sugar foods, take-away foods, and foods seen by the public as ‘junk foods’, thus reducing high additive foods. Those who think that some of the problems that present in ADHD may be due to poor nutrition generally would expect improvement with healthy diet advice, but this was not found. The challenges still minimised placebo effects by using high IgG foods considered suspect for each family. This team used a questionnaire validated for ADHD and only claimed a diet effect if there was a 50% reduction in scores. This is a large change. May parents will continue using what is seen as a difficult diet if the change is as low as 25%. This is important clinically. As well the parents may also be seeing change in more than just the factors in the ADHD questionnaire. These include improvement in physical symptoms, mood, sleep and other areas such as coordination and speech.
Possible mechanism and susceptible group
One of the most confusing aspects of the role of diet in ADHD is what the mechanism is. Research is much easier when a mechanism can be investigated physiologically. Because the mechanism is not known various researchers attend to a possible aspect, as well as usefully assessing outcomes as mentioned above. Here the researchers went to great effort to assess the possible usefulness of using IgG blood test results. The children showed ADHD relapses when challenged with both high-IgG or low IgG foods independent of IgG blood levels. Past research has reported that the likelihood of allergy is not increased in diet responders and this additional information clarifies the IgG aspect further. The researchers wisely advise that the prescription of diets based on IgG blood tests should be discouraged. At this time all is not clear with regard to what the physiological mechanism is or just who is likely to respond. From the literature and my clinical research the likelihood of a positive response is increased if there is a family history of the presence of allergic symptoms, as well as another group of symptoms that occur in food sensitive families, not just the patient. These include migraine, chronic headaches, irritable bowel syndrome, mouth ulcers, bad breath, car sickness, or limb pains. These families also often have a family member who has adverse reaction to aspirin, and some have an increased sensitivity to smell both in food and in the environment.
Overall the INCA study into the role of diet in ADHD children was well conducted. However the critics commenting on it in the professional journals are still not satisfied. They hope that all the above issues of just what diet factors matter, just which changes should be measured, and just what the mechanism is, should somehow become simple and be neatly researched. This study has achieved its aim and will be quoted as an important contribution to the work in this area.
The researchers who use a few foods diet (as distinct from a liberal one) achieve their aim of showing that diet has a role in some ADHD children. They also recognise that the few foods diet is difficult to maintain. The issue of just what are the key exclusions is in fact a different aim attended to with some of the diet variations in the work of many researchers over the last 30 years, and reviewed in my paper in the Journal of Paediatric and Child Health in 1997. Diet management needs to be seen as a different aspect or type of research better done using clinical research methodology. Data can be collected on the number of subjects who report reactions to specific foods and the various additives, in large numbers of children, presuming individual variation. Over the last 30 years this has largely been achieved when all the various findings from all the research are added together. I have collected applied, and refined the diet issues in over 1000 families, investigated over another 100 in detail for my thesis, and reported in the book Food Additives. The diet was applied in another 1000. The diet therapy resulting using the Diet Detective Method which incorporates information from the family, and allows for much individual variation, gives the best chance of clarifying whether diet has a role, with minimal restriction.