This article originally appeared in the Summer 2011 edition of Celiac.com's Journal
Celiac.com 08/25/2011 - This is a controversial topic. Elizabeth Hasselbeck’s book, The G-Free Diet (1), has been attacked because it suggests that a gluten free diet can help some people lose weight. One celiac support group has condemned this book as misleading (2). However, I thought it was a pretty good book, and I’m grateful for the public attention that Hasselbeck has drawn to celiac disease and non-celiac gluten sensitivity. There are at least two sides to the question of whether a gluten free diet is useful for weight loss. As with much other dietary advice, each of these conflicting views is sometimes presented in very strident voices. On one side there are numerous websites and newspaper articles, with an array of “experts” weighing in on this issue, decrying the use of a gluten free diet for weight loss. I even saw a segment of a television show called “Dr. Oz” where the gluten free diet was asserted to cause only weight gain. On the same show the diet was referred to as a “fraud” with respect to weight loss. Similarly, one group of researchers claim that an important side effect of the gluten free diet is weight gain. Even some very popular advocates of the gluten free diet insist that it is inappropriate for weight loss. Yet there are some individuals who advocate this diet as an effective weight loss tool and there is some evidence to back them up. There are even a couple of research reports of weight loss on a gluten free diet. In fact there is at least one study that provides some support for each paradigm. So who are we to believe? What information supports each side of the argument? And how can we evaluate that information?
Before we get to the evidence, however, I’d like to say that I have listened to Ms. Hasselbeck express some of her political and economic opinions. I am now of the firm belief that she is one of the five people on this continent who may know even less about these issues than I do. So let’s leave out the politics and confine our discussion to the issue of the gluten free diet and whether it is suitable for weight loss.
The first and most compelling piece of evidence (for me) is a personal observation. I watched my mom try to lose weight, starting when I was in elementary school. She tried just about every diet out there, from radical fringe to mainstream. She drank protein powders mixed with water instead of eating meals. She tried eating these “rye” crackers that I thought tasted like cardboard.... very crunchy cardboard. She tried a low sodium diet, then a low fat diet, then a sugar free diet, an all fruit diet, a raw food diet, or maybe that was just a single diet of raw fruit. I’m not sure. She probably tried a host of other diets that I don’t remember, but I think you get the idea. She sometimes lost weight only to gain it back as soon as she stopped the diet. More often, she gave up because she got tired of being hungry all the time. She eventually gave up on dieting altogether and accepted being overweight.
Then, about fifteen years ago, in her early-mid 70s, she started a gluten free diet. It wasn’t aimed at weight loss. She was trying to reduce the pain caused by her arthritis. In the first year and a half or two years, she lost 66 pounds. From that time onward, her weight continued to gradually diminish to the point where she had lost about 100 pounds over about ten years of eating gluten free. She was not trying to lose weight. She had long since given up on that objective. Yet the excess pounds just melted away. If only because of its weight loss benefits, I suspect that the gluten free diet has extended her life substantially.
At about 85 years of age, she started eating gluten occasionally. Part of her gluten consumption is wilful. She sees something that she thinks she might enjoy eating, and she requests a serving. Perhaps because of mom’s lapses into gluten, the staff at the home where she now lives have also become quite cavalier about her gluten free diet. They frequently serve her dishes that contain gluten. Still, her weight has remained fairly stable. My mom is not the only example of weight loss on a gluten free diet. There are other stories on the Internet. Just Google “gluten free weight loss diet” and you will see what I mean. But I can’t vouch for those stories. I didn’t observe their weight loss. All I saw was my mom’s.
Currently, there are only a few formal studies that have explored body mass changes on a gluten free diet. One conducted in Ireland reveals that there are eight times as many overweight celiacs as underweight celiacs (Dickey & Kearney). That is quite surprising in light of the common perspective that celiac disease is one of under-nutrition, suggesting that underweight should be a more likely sign of celiac disease. For a long time, that was the dominant belief, but there is clearly a flaw in this paradigm.
Suspecting celiac disease only in underweight patients is not the only complication of this issue. Dickey and Kearney also report that after two years of dietary compliance, eighty two percent of their 143 overweight and obese patients with celiac disease had gained yet more weight on a gluten free diet. This would seem to suggest that the gluten free diet is not a good bet as a weight loss tool. However, these results do not seem to have been replicated by other investigators.
Another follow-up study, conducted in New Rochelle, NY, reports that ’’ 66% of those who were underweight gained weight, whereas 54% of overweight and 47% of obese patients lost weight’’ on a gluten free diet (Cheng et al ). Thus, on this side of the Atlantic, of the eighty one overweight and obese celiac subjects, about half lost weight following a gluten free diet. That is quite different from the findings in Ireland.
Another, much smaller study of childhood celiac disease revealed that about half of the eight overweight children they studied also experienced weight loss (Venkatasubramani et al ). This research was conducted in Milwaukee and is congruent with the findings from New Rochelle. So, on this side of the Atlantic, about half of the overweight celiac patients studied experienced weight loss on a gluten free diet.
Perhaps these differences are the result of variations between the versions of the gluten free diet in North America, as compared with the diet in the United Kingdom. The primary difference I am aware of is that gluten free in the UK includes wheat starch whereas most American organizations do not accept wheat starch as gluten free. However, the gluten free diet that includes wheat starch has been shown to reduce cancer risk and many other celiac-associated risk factors, and has therefore been deemed safe. Nonetheless, that same wheat starch may be a factor in the different body mass findings between Ireland and the USA. Or maybe the difference lies in variations in research methods. Without further research, it is difficult to guess.... and that is exactly what we would be doing.
Without solid evidence, our beliefs are no more than just guesses. For instance, my mom’s weight loss could have been the result of some factor other than her gluten free diet. Perhaps the beginning of her weight loss just happened to coincide with when she started the gluten-free diet. I’m convinced by my observations of her experience, but that doesn’t mean that you should be. After all, I could be kidding myself. Or her weight loss could have been caused by some other factor that I’m not even aware of or recognizing. That is why many of us contribute our hard-earned dollars to research. We need something more than stories about my mom’s experiences. We need solid, peer reviewed research such as what is found in medical journals. However, even there we need to be cautious about reported findings.
One good indicator that researchers are on the right track is when we see a convergence of results from very different studies. When one study produces a given result, and another study produces a similar result despite very different study designs and objectives, the results of the first study are said to have been replicated by the second study. The advantage, in the case of celiac patients experiencing weight loss following institution of a gluten free diet clearly goes to the two studies conducted in the USA. The studies looked at two different sub-populations of celiac patients yet produced approximately the same results. But both studies still have a problem with selection bias.
One of the greatest difficulties in assessing research findings is that we are really just assuming that what we see in one or two small groups will be reflected in the general population. This is why, where possible, study subjects are picked randomly from the general population. However, this cannot happen in studies of celiac patients. They are a select group. This is partly because these subjects have celiac disease and partly because they have a diagnosis of celiac disease. I’m really not splitting hairs here. Please bear with me for a moment as I try to explain this important distinction.
Unlike more than 95% of Americans with celiac disease, these study subjects have a diagnosis. And don’t be fooled. Clinicians are missing almost as many cases of celiac disease in Europe as they are in the USA. Thus, all three of these studies are looking at a sub-group (diagnosed with celiac disease) of a select group (celiac disease). And the lengthy delays to diagnosis, somewhere between five and eleven years, also occur in Europe and Canada, so the difference is probably not dependent on whether there is a socialist medical system in place, as some have suggested. The select group is formed by people with celiac disease. The sub-group is people drawn from the three to five percent of those who have been diagnosed with celiac disease. We know some of the ways that those with celiac disease differ from the general population. But we don’t know any of the ways, beyond the diagnostic criteria, that people with undiagnosed celiac disease differ from the general population or from the population of people whose celiac disease has been diagnosed.
Studying a small sub-group of celiac patients who have a diagnosis, then assuming that the features observed will be present in all those with celiac disease, whether they have a diagnosis or not, is a flawed approach. Statisticians call this mistake ‘selection bias’. It is a well recognized type of statistical error. For instance, if you wanted to predict the buying habits of people living in Pennsylvania, you would not just observe members of the Amish community. Doing so would not only induce a selection bias, it would lead to very misleading information about the general population of Pennsylvania. While many Amish live in Pennsylvania, their buying habits probably do not reflect the buying habits of most people in Pennsylvania. Similarly, the selection bias driven by extrapolating from observations of sub-groups of people with diagnosed celiac disease and applying those principles to undiagnosed celiacs, leading us to either assume that weight loss will or will not occur on a gluten free diet is mistaken and likely to produce misleading information.
In addition to selection bias, sample size is another important factor in predicting features of a larger population based on observations of a sub-population. The smaller the group, the less likely it is to reflect the variations present in the larger population of those with celiac disease. For instance, if the US population is currently about 311 million, and the rate of celiac disease is about one in every 133 people, then there should be about 2.3 million Americans with celiac disease. Only three to five percent of Americans with celiac disease are thought to be diagnosed with celiac disease. And the studies of overweight celiacs who gained or lost weight on a gluten free diet include about 89 Americans and 143 Irish people. Is it credible to imagine that we can predict the responses of 2.3 million Americans based on observations of a sub-group of 89 of their compatriots and 143 Europeans? I think that most readers will agree that leaping to such conclusions is unreasonable. Yet that is what we do if we insist on the exclusive correctness of either side of the question of whether the gluten free diet is an effective weight loss tool.
I am convinced, both by my observations of my mom, and by the results of these two small studies, that some celiacs will lose weight on a gluten free diet. However, I would not presume to insist that it is the best, or even a good tool for all overweight celiacs. Neither would I insist it was a good weight loss tool for all diagnosed overweight celiacs. Given the US studies, that is clearly not the case. Equally, denial of anecdotal reports or the two US studies claiming that the gluten free diet is not an effective weight loss tool for anyone is also unreasonable.
We can only say, with confidence, that these study results may apply to those who are diagnosed with celiac disease. Yet we have a fairly even split, with American researchers showing that about half of overweight celiacs lose weight on a gluten free diet, and Irish researchers asserting that eighty two diagnosed overweight celiacs gained even more weight on a gluten free diet.
Yet these statistical problems are not insurmountable. If a group of researchers conducted random screening blood tests for celiac disease in a variety of settings and circumstances, confirmed the celiac diagnosis in a large group of these individuals, and followed up with those who were overweight and undertook the gluten free diet, then their observations might reasonably be applied to the celiac population in general, whether diagnosed or undiagnosed. There would still be a relatively minor statistical error induced by cases of sero-negative celiac disease, but the statistical problems would not be anywhere near as problematic as asserting that any or all of these three studies tell us much about weight loss on a gluten free diet, except that it sometimes happens in small sub-groups of diagnosed celiac patients.
Since such research has not been conducted, it behooves all of us to take a moderate stance on either side of this debate.
That does not mean that we can’t or shouldn’t make use of the available information. Each of us can draw our own conclusions based on our interpretations of the available data. If you believe that, in North America, a gluten free diet can induce weight loss in about half of overweight, newly diagnosed celiac patients, it does seem reasonable to suggest that the gluten free diet may be all that is needed for some diagnosed celiacs to lose weight. However, since we are missing more than 95% of cases of celiac disease, it is difficult to say whether it will help those undiagnosed, overweight celiacs to lose weight. Nonetheless, it is possible. Thus, if it will help some, perhaps about half of them to lose weight, those individuals might well consider this information, limited though it may be, very valuable.
Anecdotal reports, such as my mother’s story, might also be considered very valuable by those who can lose weight on a gluten free diet. For those who do not lose weight on this diet, I suspect that many of them have walked the path my mother did, and it won’t be the first time that a diet failed to work for them. This, of course, raises the question of why some individuals and organizations have vigorously opposed and decried anecdotal claims that a gluten free diet may help some people lose weight. Clearly, there is hard scientific evidence to support this claim. The reverse is not the case. Nobody has, or can, prove that the gluten free diet is always ineffective at helping people lose weight.
Meanwhile, we can hope for more research that will answer some of the many questions that arise from this relatively new information that there may be many more overweight people with celiac disease than there are underweight people with celiac disease.
Several of the questions that remain include:
What causes overweight and obesity in patients with celiac disease? It is, after all, a disease that is characterized by inadequate absorption of nutrients from the food that passes through the gastrointestinal tract. I have previously suggested that specific nutrient deficiencies may induce food cravings that cause some to continue to eat despite feeling ’’full’’ because their bodies continue to demand these missing nutrients. The new field of metabonomic research may soon shed more light on this area. It has already demonstrated that subjects diagnosed with celiac disease are not as efficient at metabolizing glucose (usually derived from carbohydrates) as those without celiac disease.
Does wheat starch have any impact on nutrient absorption or appetite? If even small amounts of opioid peptides survive in wheat starch and are allowed access to the bloodstream and brain, they may well have an impact on appetite. Opioids or some other component of wheat starch might also alter ghrelin (a hormone that incites appetite) and/or leptin (a hormone that suppresses appetite). We just don’t know.
Are there other dietary differences between Ireland and the USA? We are aware of the difference in wheat starch, but what other factors might contribute to these divergent research results?
How does wheat starch compare with the 20 parts per million currently being put forward as the labelling standard for American legislation in the offing? Does wheat starch contain 20 ppm?
Will the legislation in question change conditions for celiac patients?
Just how much contamination from gluten grains is present in commercial oats? Even in the absence of contamination, how many people with diagnosed celiac disease experience cross-reactions with oats? This is where the selective antibodies are sensitized to protein segments found in oats as well as in gluten grains.
What other differences between Ireland and the USA might explain these variations in research findings? Could variations in sunlight, or water-borne minerals, or even genetics contribute to the difference in findings?
How representative are these groups of other groups of celiac patients? Do they reflect what is going on among all the other diagnosed celiacs in their region? And how do these findings apply to the undiagnosed celiacs?
Is region a genuine factor in all of this? I remember when many researchers were quite willing to believe that there was some difference that had Italy showing a rate of celiac disease of one in 250 while in the USA and Canada it was thought to afflict about one in twelve thousand. We now know that was silly, but at the time, there were a lot of apparently intelligent people who were vigorously asserting the accuracy of those variations and postulating many creative explanations for them. I remember one, now prominent celiac researcher, admonishing me not to take the Italian findings too seriously. He was very confident that they represented a large overestimation of the true incidence of celiac disease in Italy and could not reasonably be suggested as reflecting anything about Canada or the USA.
Now here is a really startling thought. Some of the overweight people with non-celiac gluten sensitivity might also be able to lose weight on a gluten free diet. If so, this could produce as much as a ten-fold increase in the number of people who might lose weight on our diet. Has anyone tested obese and overweight people for anti-gliadin antibodies? Could gliadin be a factor in some peoples’ weight problems?
I wonder how many people might be helped to lose weight if pre-conceived notions about the gluten free diet could be relinquished in favour of a more open minded view.... one that recognizes that there is some evidence that some people can and do lose weight on a gluten free diet?
The dogmatic certitude that abounds on the question of weight loss through the gluten free diet is profound and disturbing. As is pointed out by nutritionist, Brian Dean, in his article on gluten and heart disease in this issue of The Journal of Gluten Sensitivity, one long-standing dietary sacred cow has been killed. We now know that eating saturated fats is not a causal factor in heart disease.
Equally, the emerging sacred cow that a GF diet is not appropriate for weight loss is, as yet, supported only by flimsy evidence, all of which is contradicted by other research. So let’s avoid making rigid pronouncements about the gluten free diet until we have a better understanding of the complex and perplexing causes of obesity and overweight in the context of untreated celiac disease. And please, let’s remember that some people can and do lose weight on a gluten-free diet alone. My mother is an excellent but by no means unique example. Others have similar stories. My own experience on the diet was weight gain, and now I have to work at keeping from gaining any more.
Only those who know all there is to know should speak in absolutes. The rest of us should constrain ourselves to offering opinions and perspectives. Sources:
- Hasselbeck E, The G Free Diet: A Gluten-Free Survival Guide. Center Street- Hatchette Book Group, NY, 2009.
- Dickey W, Kearney N. Overweight in celiac disease: prevalence, clinical characteristics, and effect of a gluten-free diet. Am J Gastroenterol. 2006 Oct;101(10):2356-9.
- Cheng J, Brar PS, Lee AR, Green PH. Body mass index in celiac disease: beneficial effect of a gluten-free diet. J Clin Gastroenterol. 2010 Apr;44(4):267-71.
- Venkatasubramani N, Telega G, Werlin SL. Obesity in pediatric celiac disease. J Pediatr Gastroenterol Nutr. 2010 Sep;51(3):295-7.