Gluten-free diets may increase the risk of heart disease, according to a new study.
There has been increasing interest in the possible health benefits of avoiding gluten among people who do not have celiac disease, though the long term evidence about its effects in this group is currently limited. Despite this, the gluten-free food market is reported to have made $3.5bn worth of global sales in 2016.
Gluten is a protein found in grains such as wheat, barley and rye. In people with celiac disease, it damages the intestines and triggers digestive symptoms such as diarrhea, meaning they need to follow a gluten-free diet.
The study was conducted by researchers from Columbia University in New York, and Massachusetts General Hospital, Harvard Medical School, Brigham and Women’s Hospital, and the Harvard T. H. Chan School of Public Health in Boston.
Prospective Cohort Study
The team followed more than 100,000 people from 1986 to 2012, assessing their diets and whether they had heart attacks during that time. These people did not have heart disease at the start of the study, and importantly did not have celiac disease.
While a randomised controlled trial is generally the best way to test whether a particular factor causes a specific outcome, it would not be feasible to randomly allocate thousands of people to eat gluten or not for a long period of time. Therefore, a large cohort study such as this is the best way to look at this question.
The main challenge with this study type is to try and single out the effect of gluten as opposed to any other factor. Researchers do this by using statistical techniques to try and “remove” the impact of these other factors.
The 110,017 participants without celiac disease who did not have heart disease in 1986 filled out detailed questionnaires about their diet at the start of the study and every four years after that, up to 2010. The researchers followed them up to see who developed heart disease over this period, and whether different levels of gluten consumption affected the likelihood of developing the condition.
The standard diet questionnaire included more than 130 questions about how often a person consumed specified portions of certain foods and drinks. The researchers used the participants’ responses to estimate how much gluten they were consuming on average over the study period using a database of nutritional contents of the foods and drinks.
They included gluten from wheat, rye and barley, but did not include the small amounts of gluten which are present in oats or condiments such as soy sauce as they felt these would be negligible. People were then split into five groups with increasing levels of gluten consumption for comparison.
Because people might change their diet as a result of illness, for people who developed diabetes, cancer, or certain heart disease events such as stroke or had surgery to treat heart disease, the researchers only considered their diet before they developed these conditions.
Participants filled out questionnaires about their health every two years, and if they reported having a heart attack their medical records were checked.
Deaths from heart attack were identified from state and national records, or reports from next of kin. Medical and post-mortem records and death certificates were also checked for these individuals. If these record checks confirmed the reported diagnosis, these people were considered as having developed heart disease.
The researchers analysed whether participants who consumed more gluten were any more or less likely to develop heart disease. They took into account many potential confounders that could be related to heart disease risk, including:
- body mass index
- history of diabetes, high blood pressure or high cholesterol
- regular use of aspirin and non-steroidal anti-inflammatory drugs
- current use of statins
- current use of a multivitamin
- smoking history
- physical activity
- parental history of heart attack
- menopausal status and menopausal hormone use
- other dietary factors such as alcohol, red and processed meats, polyunsaturated and trans fats, and fruit and vegetables
In addition, the researchers also looked at what happened if they took into account consumption of whole and refined grains, as these contain gluten, and have been linked to level of heart disease risk.
The mean daily intake of gluten at the start of the study was:
- 7.5g among women and 10.0g among men in the highest consumption group
- 2.6g among women and 3.3g among men in the lowest consumption group
People with higher gluten intake tended to:
- have lower alcohol intake
- smoke less
- consume less fat overall
- eat less unprocessed red meat
- consume more whole grains and refined grains
During the study 6,529 participants (5.9%) experienced a heart attack.
Before taking into account potential confounders, heart attacks were more common in the group with the lowest gluten consumption than in those with the highest consumption. However, after taking into account known risk factors for heart disease, the difference between the groups was not statistically significant.
When the researchers looked at the impact of consumption of gluten from just refined grains they also found the difference between groups was not statistically significant.
But when they considered the impact of consumption of gluten in whole grains – they found those with highest gluten consumption were 15% less likely to develop heart attacks over follow up (hazard ratio 0.85, 95% confidence interval 0.77 to 0.93).
The researchers concluded that differences in long term dietary intake of gluten were not associated with risk of heart disease. However, their results suggested that avoiding gluten may reduce consumption of whole grains, and this may lead to increased risk of heart disease.
They recommended that “promotion of gluten-free diets among people without [coeliac] disease should not be encouraged”.
This study has found that while overall gluten consumption in people without celiac disease may not be related to heart disease risk, avoiding whole grains (wheat, barley and rye) in order to avoid gluten may be associated with increased heart disease risk.
This study has several strengths, including its large size, the fact that data was collected prospectively and diet assessed at several time-points, the long period of follow up, and that it took into account a wide range of potential confounders.
As with all studies of this type, it is possible that other factors may affect the results. However, the researchers took into account as many potential confounding factors as they could in their analyses. This increases confidence in the results, but it is still possible that these or other unmeasured confounding factors are having an effect.
The researchers noted that they did not specifically ask participants whether they were intentionally following a “gluten free” diet or consumption of gluten-free substitute foods.
It is important to emphasise that this study was only in people who did not have celiac disease. People with celiac disease need to eat a gluten free diet to control their symptoms, and it is thought that this diet may actually contribute to the reduction in risk of heart disease seen after diagnosis in this group.
So people eating a gluten free diet for this purpose should not be concerned by the findings in this study.
The study collected data from 1986 to 2012. Diets over this period have changed, and avoidance of gluten is likely to be more common nowadays. It would be interesting to repeat the study now to see if the same results are found. While it would be good to have these findings confirmed by other studies, carrying out similarly large scale and long term research will take time.
Ideally, if you don’t need to avoid gluten for medical reasons, then this study suggests it may be beneficial to continue including whole grains in your diet for their cardiovascular benefits.
Lebwohl Benjamin, Cao Yin, Zong Geng, Hu Frank B, Green Peter H R, Neugut Alfred I et al.
Long term gluten consumption in adults without celiac disease and risk of coronary heart disease: prospective cohort study
BMJ 2017; 357 :j1892