OK, so I made up the quote above, but it captures a certain zeitgeist that’s in the air right now. The media is all too keen to uncritically give gluten-free and clean diets a kicking at the moment, wagging fingers at all those ‘silly people’ who fell for the anti-gluten message even though they don’t have coeliac disease – what fools!
Except, as we have explained in multiple articles on this site, gluten has a far greater reach than that 1% who have classic coeliac disease. Non coeliac gluten sensitivity (NCGS) is a recognised and studied condition, with an estimated prevalence of up to 6% of the population.
And even a cursory look behind these dismissive headlines shows that the studies they are based on add almost nothing to our understanding of gluten pathology, and indeed contradict themselves.
Walter Willett, the subject of my misquote above, is professor of Epidemiology and Nutrition at Harvard T.H. Chan School of Public Health, and a staunch vegan. (Clearly a vegan of the wealthy first world kind, unaware of the fact that no vegan societies or tribes of Homo sapiens have ever existed, as they die out due to malnutrition pretty soon). It was his team who were responsible for linking a gluten-free diet to an increased risk of heart disease, yet as you will see below his conclusion was more wishful thinking thinking than hard science.
All of this would be idle intellectual banter, if it were not for the fact that I have many patients whose health is dependent on maintaining a properly gluten free diet. The last thing they need is to be unnecessarily worried by irresponsible media chasing headlines.
- Gluten-free foods could increase risk of obesity, says researchers – The Independent
- Gluten-free diet carries increased obesity risk, warn experts – The Guardian
These articles were stimulated by a report from the European Society for Paediatric Gastroenterology Hepatology and Nutrition, which assessed 654 gluten-free products and found many of them had lower nutritional value than their regular gluten containing counterparts.
Typically, gluten-free pizza, cakes, bread and biscuits contained more fat and sugar along with less protein than their standard wheat based counterparts. This isn’t surprising as the glue-like properties of gluten that have made wheat such a dominant success are unique and hard to mimic. Creating acceptable simulacra requires a lot of messing about with gums, modified starches, sugars and fats to get a mouth feel people will tolerate. I would challenge the inclusion of greater amounts of fat in these products as evidence of them having a poorer nutritional status too, as fat is actually essential, notwithstanding the quality of such fats in these products.
The article is quite right to state that a diet based around such foods may be nutritionally inferior, but what they fail to point out is that the gluten containing foods they are copying were second-class in the first place.
Shifting to a gluten-free diet can easily lead to improved nutrition: you just need to move away from regularly consuming cakes, biscuits, pasta and bread and replace these with more real foods: cheeses, fish, eggs, fruit, vegetables, fish, nuts and meats. These naturally have a lower GI and are more satiating and vastly higher in nutrients.
And for the few occasions when there are no other options (e.g. eating out, parties, Christmas), gluten-free products for all their sins are not going to do a lot of harm in most cases.
This story was widely reported with most memorable headlines linking a gluten-free diet to increased heart risks. The story is based on a retrospective study of 100,000 men and women’s diet and health over 25 years from Harvard Medical School, which was published in the BMJ (free text here).
The authors looked at coronary events comparing the 20% consuming the least gluten (~3g per day) with the 20% consuming the most (~9g per day).
Considering the headlines you may be surprised with the following results from the paper:
We found no significant association between estimated gluten intake and either fatal myocardial infarction or non-fatal myocardial infarction when considered as separate outcomes. We did not observe any significant associations between tenth categories of gluten intake with risk of coronary heart disease. Nor did we find a significant association between gluten intake and coronary heart disease when separately considering the time strata 1986-97 and 1998-2012
In their secondary analysis they showed that those eating less gluten tended to also eat less whole grains. Which was kind of obvious. And as whole grains are associated with heart health QED: “The avoidance of gluten may result in reduced consumption of beneficial whole grains, which may affect cardiovascular risk.” hence the newspaper headlines. Using ‘may’ twice in a conclusion is a pretty good indication that wishful thinking may have more of a causative role than the data. This study didn’t even ask if the participants were avoiding gluten, and with the lowest quintile eating ~3g per day it doesn’t look like they were. 3g is 3g too much in my book!
And on the matter of whole grains being ‘heart healthy’, I beg to differ. I attended a conference on cereal grain science which had the objective of promoting whole grains to the public. It was held at The Royal Society of Medicine, no less, and two of the speakers were professors from British universities who, independently, had been endeavouring to prove this heart-healthy whole grain hypothesis. When I challenged them directly they admitted that their work had been unable to establish any such heart benefits. So much for the claims! When put to the test they prove unsound.
I will write in depth about the evidence around whole grains at some point, but in brief: the benefits of whole grains are far from proven, and if they exist at all they are likely only relevant when comparing whole grains to refined grains. Said another way, whole grains may be a little less bad than refined grains.
Lets return to one of the authors, Walter Willet – the butt of my misquote at the top of this article. Prof. Willett is a staunch vegan. His nutrition team at Harvard have previously come to conclusions that might be said to support his dietary bias, and it looks like this is another example of the same. Indeed the correct AZ quote before I messed around with it was this:
So here’s an alternative interpretation of their conclusion (aka ‘alternative facts’) for which I have no hard data, just a cynical hunch: “People avoiding gluten may be following a paleo diet, which may mean they are eating more meat (than we want them to)”. Just enough ‘mays’ to cover my back I think.
So the media published Walter’s speculative assertion as their main headline, confirming, instead of challenging any potential bias and made some eye catching click-bait headlines in the vain attempt to prevent their empires sinking quite so fast.
I’m sure there’s a presidential quote that would be appropriate here if I could remember it…
In the meantime, lets just take a walk over here…
Come on…! This way…
(Keep up at the back)…
…to a study published in the Journal of Nutrition, at about the same time, titled: Paleolithic and Mediterranean Diet Pattern Scores Are Inversely Associated with All-Cause and Cause-Specific Mortality in Adults. (Whalen KA et al. Apr 2017).
Why have I brought you to this little corner of epidemiological research? Because this is a kind of inverse study to the one undertaken by the Walter Willet Harvard team. In this case, the highest vs lowest quintile of paleolithic diet score were compared. Paleolithic diets, as you probably know, eschew grains, so those in the lowest quintile would tend to have the lowest intake of gluten too.
And what did this study find? Those eating a diet closer to paleo principles had a reduced risk of cardiovascular related deaths (28% fewer, in fact, than those whose diets were the ‘least paleo’).
This study didn’t make headlines of course, and it is really no more conclusive than the Harvard study, but it should act as an antidote for anyone who was needlessly worried by those media headlines.