Elimination diets often act as weight-reducing diets, and on average, Dr Darlington’s patients lost about 4.5 kg (10 lbs) during the trial. Being overweight has a bad effect on diseased joints, especially those in the knees and ankles. So one might argue that the loss of weight could have contributed significantly to the improvement seen. To check this, Dr Darlington compared weight losses in those that had responded well and in those that had responded poorly. There was no difference, so this seemed unlikely to have been an important factor.

In spite of these striking results, Gail Darlington is in no way a propagandist for the food-intolerance idea: ‘I’m a very routine, orthodox physician and rheumatologist. If I spend the next ten years of my life helping to prove that the whole thing is a nonsense, or a placebo effect, or a non-specific manipulation of the immune system, I won’t be at all concerned -1 simply feel it’s an area that needs to be investigated in just as scientific a way as we look at Drug A versus Drug B. Before 1981, most people in the UK thought that food intolerance was rubbish in a rheumatologic context, as indeed I did at that time -one is fairly definitely trained to believe it’s rubbish. But I was impressed by the results I saw in my patients who had gone to other people to have their diets manipulated. After seven years working in this field, I’ve gradually come to believe that it is relevant to some patients. To take one example, I have a patient of 33 who has changed from being a limping acute arthritic to being a perfectly fit, normal young man – and that is an improvement he has maintained for two and a half years.’

Rheumatologists tend to attribute such results to natural remissions, because rheumatoid arthritis is a disease that comes and goes for no apparent reason. It is also notoriously susceptible to placebo effects. Gail Darlington understands these doubts, but feels they are misplaced: ‘Yes, obviously, it could be purely coincidental, but I do have quite a few patients in this bracket now, and it does seem unlikely that they all just happened to go into natural remission at the moment they began their diet. As for placebo effect, our trial was carefully designed to measure this in the control group. We showed that there was a placebo effect, but that it certainly couldn’t account for all the improvement seen in the patients on the diet. What is more, when the control group were later put on elimination diets, they responded just as well as the first group -far better than they had done on the placebo.’

*108\180\8*