How "gluten-free" does a gluten-sensitive person need to be and is it important to purchase products explicitly gluten-free?

Over the past decade, there has been a growing literature that recognizes a set of disorders associated with gluten intolerance in addition to celiac disease.  There is little to do for the roughly 1- and 10-percent of the population thought to have celiac disease or gluten intolerance other than eliminate gluten from their diet.  However, following a gluten free diet poses it's challenges since gluten is either a direct component -- often from wheat -- or enters as a contaminent.  In fact, even "gluten-free" products actually contain small amounts of gluten.  This raises two questions: One, how much gluten can a gluten sensitive individual tolerate and, two, how contaminated are traditionally gluten free products?

Research on the gluten limits are understandably directed at celiacs since it has long been recognized as a disorder and there are observable symptoms.  The studies generally take the form of giving people regular doses of gluten and then evaluating their response to them.  Consider the two literature summaries by the FDA and Akobeng and Thomas.  Even a casual reader will immediately recognize two broad results.  One, within each study the scientists recognize a wide range of individual tolerance.  Two, the recommended thresholds across studies also have a wide range in part dues to whatever metric(s) they find important, the duration of the experiment, and what percentage of celiacs the recommendation is targeting. With just a small sample of studies, I've seen recommendations for less than 1 mg, 10 mg, and 50 mg of gluten a day.  Broadly speaking, the recommended threshold I've seen most often 10 mg/d.  It's pretty clear that this should be considered a starting point for gluten sensitive individuals.    

The gluten microchallenge disclosed large interpatient variability in the sensitivity to gluten traces. Some CD patients showed a clear-cut worsening of the small-intestinal architecture after ingesting only 10 mg gluten/d, whereas others had an apparent improvement in mucosal histology after the 3-mo challenge with 50 mg gluten/d. Furthermore, one patient challenged with 10 mg gluten/d experienced clinical symptoms after a few weeks, whereas none of the 13 subjects receiving 50 mg gluten/d had clinical evidence of relapse. (reference)

In case it is unclear, 10 mg/d is a teeny tiny amount of gluten.  A typical diet is estimated to contain 10 to 40 grams of gluten.  Taking the lower bound of 10 grams, a person would be limited to one-thousandth of that amount.  A single slice of wheat bread contains orders of magnitude more gluten!

Pursuing a gluten-free diet requires dedication and moderate effort.  It quickly becomes apparent that products explicitly labeled gluten free are often quite expensive even though they are theoretically gluten free.  How worried should one be about possible wheat/gluten contamination of say something like oats which should naturally be gluten-free?  There are studies that randomly sampled and test commercial products for gluten and concluded that one should indeed be worried if one is targeting some commonly accepted thresholds -- 20 parts-per-million is what one generally sees in the US -- for a product to be gluten free.

Thompson, 2004

A broader study that included the same author concluded ...

Twenty-two inherently gluten-free grains, seeds, and flours not labeled gluten-free were purchased in June 2009 and sent unopened to a company who specializes in gluten analysis. All samples were homogenized and tested in duplicate using the Ridascreen Gliadin sandwich R5 enzyme-linked immunosorbent assay with cocktail extraction. Thirteen of 22 (59%) samples contained less than the limit of quantification of 5 parts per million (ppm) for gluten. Nine of 22 (41%) samples contained more than the limit of quantification, with mean gluten levels ranging from 8.5 to 2,925.0 ppm. Seven of 22 samples (32%) contained mean gluten levels >/=20 ppm and would not be considered gluten-free under the proposed FDA rule for gluten-free labeling.

Consequently, if one is maintaining a gluten-free diet one must use uncertified products with some caution.  Although as a quick aside, it does matter how much of the product one consumes.  The 20 ppm recommendation is based on an assumption of how much gluten-free flour one is consuming -- the rough assumption is half a kilogram -- and the aforementioned 10 mg daily limit.  If one is less sensitive or eats less than those assumptions, than one might be willing to take greater risks.

An obvious assumption here is that studies on celiac disease are applicable to other forms of gluten sensitivity.  Since other gluten disorders have only recently become broadly recognized the literature is very sparse and the assumption is unavoidable, in my opinion.  The important point is that if one is trying a gluten-free diet in response to some symptoms, one should be proactive with one's diet and pay close attention to its effects.  Naturally, this brings up a discussion about placebo effects ...