For people with celiac disease, even tiny amounts of gluten in foods can cause trouble, and restaurants may be the hardest places to avoid the hidden protein, a US study suggests.
More than half of gluten-free pizza and pasta dishes tested in restaurants were positive for the presence of gluten, and overall, about one third of supposedly gluten-free foods had some gluten, researchers report in the American Journal of Gastroenterology.
"The long-suspected problem of gluten contamination in restaurant foods that has been reported by patients likely has some truth behind it," said senior study author Dr. Benjamin Lebwohl, director of clinical research at the Celiac Disease Center at New York Presbyterian Hospital and Columbia University Medical Center in New York City.
"We don't know how common it is. But our findings suggest that pizza and pasta and foods served during dinner time are more likely to have this problem."
Lebwohl often heard anecdotally from patients that they suspected they had consumed gluten at restaurants, even though the foods they ate were labelled gluten-free.
When he learned that the manufacturer of the portable Nima Gluten Sensor had a stash of data submitted by restaurant patrons who used the device to test foods they suspected, Lebwohl asked if he could use the information in a study.
The company supplied what they had: 5,624 food tests performed by 804 users during an 18-month period. When the researchers analysed the data, they found that 32 per cent of tests revealed gluten contamination in dishes that were supposed to be gluten-free.
Gluten-free pasta samples were positive for the protein in 50.8 per cent of tests, while gluten-free pizza turned out to contain gluten in 53.2 per cent of tests. Gluten was detected in 27.2 per cent of breakfasts, 29 per cent of lunches and 34 per cent of dinners.
Lebwohl admits there are a lot of limitations to the dataset. "The people tested what they wanted to test," he said. "And the users chose which results to upload to the company. They may have uploaded the results that surprised them the most. So, our findings don't mean that 32 per cent of foods are unsafe."
Moreover, Lebwohl said, the Nima is very sensitive. To be labelled gluten free in the US, a product must contain less than 20 parts per million (ppm). But the device can detect levels as low as 5 to 10 ppm, which "most people would not consider to be clinically significant. So, it's possible for a food to have less than 20 ppm and still indicate that gluten is present."
The Nima gluten detector, which sells for about $229 (S$310) and uses disposable $6 test strip capsules, displays a wheat symbol when it detects gluten and a smiley face when it doesn't, the authors note. It can upload results to the company via a smartphone app.
Lebwohl suspects that gluten-free foods are being inadvertently contaminated. "If a gluten-free pizza is put in an oven with a gluten-containing pizza, aerosolized particles could come in contact with the gluten-free pizza," he said. "And it's possible that cooking gluten-free pasta in a pot of water that had just been used for pasta that contained gluten might result in contamination."
The solution, Lebwhol said, may be better education for food preparers.
About 1 per cent of Americans have celiac disease, Lebwohl said. The amount of gluten needed to cause intestinal damage in these patients is tiny, he added. "It would be barely visible, like tiny crumbs of bread," he explained.
While the limitations of the data mean it's not possible to determine what percentage of restaurant foods labelled gluten-free actually contain the protein, "it's interesting to know that so many foods marked as gluten-free aren't," said Therezia Alchoufete, a clinical dietician in the University of Pittsburgh Medical Center's Digestive Disorders Center.