ASPPH logo


Member Research & Reports

Member Research & Reports

Columbia Study Finds One-third of ‘Gluten-free’ Restaurant Foods in U.S. Test Positive For Gluten

One-third of the “gluten-free” foods sold in U.S. restaurants tested positive for trace levels of the substance, according to new research conducted at Columbia University Irving Medical Center and Columbia University Mailman School of Public Health.

The finding will be of particular concern to the 1 percent of Americans with celiac disease for whom gluten — the protein in wheat and other grains — can damage the intestinal lining.

“As awareness of celiac disease and the gluten-free diet have increased in recent years, restaurants have sought to offer selections that are compatible with these restrictions,” said study author Dr. Benjamin Lebwohl,  assistant professor of medicine and epidemiology. “But some establishments do a better job than others at preventing cross-contamination.”

For the study, more than 800 users of a portable gluten sensor uploaded test results on gluten content of dishes listed as gluten-free on menus. Based on more than 5,600 gluten tests over 18 months, this crowd-sourced analysis found that 27 percent of gluten-free breakfast meals tested positive for gluten. At dinner time, this figure increased to 34 percent. The rise could reflect a steady increase in gluten contamination risk throughout the day, Dr. Lebwohl and colleagues said.

According to Dr. Lebwohl, some gluten-free foods are riskier than others. For example, more than half of all purportedly gluten-free pastas and pizzas tested positive for gluten.

“The fact that gluten was so often found in pizza suggests that sharing an oven with gluten-containing pizza is a prime setting for cross-contamination,” explained Dr. Lebwohl, who leads the Columbia University’s Celiac Disease Center. “Gluten-free pasta can be contaminated if prepared in a pot of water that was used to prepare gluten-containing pasta.”

Although the U.S. Food and Drug Administration regulates packaged foods with gluten-free labeling, there is no federal oversight of gluten-free claims in restaurants, noted Dr. Lebwohl.

“This estimate of gluten contamination rates in restaurant food is just that: an estimate,” cautioned Dr. Lebwohl. “The estimate may be affected by multiple factors, including the possibility that the device sometimes detects gluten at concentrations under 20 parts per million, which is the clinically relevant cutoff for considering a food gluten-free. And most importantly, these results were voluntarily uploaded by users, who may be more likely to share results that show gluten contamination.”

“These results underscore the need for education in food preparation at restaurants, and the need for diners to inquire about these precautions,” Dr. Lebwohl said.

The study was presented at the American College of Gastroenterology annual meeting in Philadelphia, October 8-10.