It might also be because many people are bad observers of their own situation, subject to confirmation bias, and do nothing to rule out placebo effects. Given

It might also be because many people are bad observers of their own situation, subject to confirmation bias, and do nothing to rule out placebo effects. Given what we know about people, this hypothesis is a priori far more plausible than gluten acting in some mysterious way that can't be identified, and somewhat more plausible than gluten being associated with other things that people are sensitive to. The suggestion that gluten is bad in ways that can't be detected is made less plausible by generic blind testing. Put people self-identified gluten-sensitive people on blinded diets with and without gluten and see if there is a difference. This has been done, and the test is negative. This is a black-box test that is completely independent of the mechanism of action. Look: http://www.realclearscience.com/blog/2014/05/gluten_sensitivity_may_not_exist.html http://www.realclearscience.com/blog/2014/05/gluten_sensitiv... As that same study shows, the suggestion that gluten is associated with something else bad is not totally implausible, but anyone aware of the difficulty of teasing out confounding factors in nutritional research will tend to strongly discount any form of anecdotal self-report as unreliable, because self-report about a subject where people are known to very easily fool themselves simply does not increase the posterior plausibility of the hypothesis very much. The potential for confirmation and reporting bias is far too high. This is not close-mindedness, but rather a realistic assessment of the situation based on the long history of dietary fads, all of which were strongly supported by self-report, and later turned out to be ineffective, sometimes dangerous, nonsense.