A report just released concludes that food allergies are often over diagnosed and misdiagnosed. 30% of the population believe they have food allergies, but in a report commissioned by the Feds only about 8% of children have it and less than 5% of adults do. The study was done by Stanford University, the Rand Corp. and the VA. They reported that there is a large volume of studies but very few are of high enough quality to be of value. One of the problems is that the term, food allergy, it self has no universally accepted definition. There is also no well accepted criteria for diagnosis or any one test that is better than another.
There were 25 studies of treatment looked at. Of the elimination diets only one reported improvement in atopic dermatitis (allergic rash). Immunotherapy looked promising, but there is insufficient data to recommend use. Of the prevention strategies; delayed introduction of solids to babies was correlated to less food intolerance, but the quality of the studies are lacking.
There you have it a lot of us believe we have something wrong and we don't. Of course this lends itself to a lot of abuse by the woo practitioners.