Friday, May 16, 2014

Real Science Works Like This

I happened upon this article in a tweet, and I like it as much for what is shows about a rigorous researcher's methods as the actual content of the study: I actually read a brief article a while ago that said, simply, "if you don't have Celiac's, gluten won't mess you up." As I've said this to a few people, it's nice to have a bit more research to back up that statement!

My non-specific-to-the-gluten-conversation take-aways are these:

Experimental design is vital to getting good results, and good researchers are obsessive about eliminating variables. This is actually a big problem I have with most research about climate change; you can't eliminate the variables to the extent that you can attribute the cause to human activity. The first study was actually very well done, but the guy wasn't satisfied. Wow.

The 'Nocebo' effect is real.

Good research deliberately sets out to disprove it's prior findings rather than confirm them, because it's failing to disprove them when you are rigorously trying to do so that confirms your previous findings....


In 2011, Peter Gibson, a professor of gastroenterology at Monash University and director of the GI Unit at The Alfred Hospital in Melbourne, Australia, published a study that found gluten, a protein found in grains like wheat, rye, and barley, to cause gastrointestinal distress in patients without celiac disease, an autoimmune disorder unequivocally triggered by gluten. Double-blinded, randomized, and placebo-controlled, the experiment was one of the strongest pieces of evidence to date that non-celiac gluten sensitivity (NCGS), more commonly known as gluten intolerance, is a genuine condition.
By extension, the study also lent credibility to the meteoric rise of the gluten-free diet. Surveys now show that 30% of Americans would like to eat less gluten, and sales of gluten-free products are estimated to hit $15 billion by 2016 -- that's a 50% jump over 2013's numbers!
But like any meticulous scientist, Gibson wasn't satisfied with his first study. His research turned up no clues to what actually might be causing subjects' adverse reactions to gluten. Moreover, there were many more variables to control! What if some hidden confounder was mucking up the results? He resolved to repeat the trial with a level of rigor lacking in most nutritional research. Subjects would be provided with every single meal for the duration of the trial. Any and all potential dietary triggers for gastrointestinal symptoms would be removed, including lactose (from milk products), certain preservatives like benzoates, propionate, sulfites, and nitrites, and fermentable, poorly absorbed short-chain carbohydrates, also known as FODMAPs. And last, but not least, nine days worth of urine and fecal matter would be collected. With this new study, Gibson wasn't messing around.

***Update/addition*** This is fascinating: http://www.newyorker.com/reporting/2010/12/13/101213fa_fact_lehrer

1 comment: