NEW YORK — A study of restaurant items promoted as gluten-free detected gluten in 32% of the items, but the study was unable to quantify the percentage of items over 20 parts per million of gluten. A Food and Drug Administration rule states that packaged foods labeled as gluten-free must contain less than 20 p.p.m. of gluten. No similar rule applies to restaurant foods.

The study appeared online March 26 in The American Journal of Gastroenterology and involved researchers from the Columbia University Medical Center in New York and Nima Labs in San Francisco. Items with the highest levels of gluten detection were pizza at 53% and pasta at 51%. A higher percentage of gluten detection was found at dinner (34%) than at lunch (30%) or breakfast (27%).

People with celiac disease must avoid gluten.

“Patients have long suspected that gluten contamination in restaurant foods is a frequent occurrence, and these results support that,” said Benjamin Lebwohl, M.D., who works in the Celiac Disease Center at New York’s Presbyterian Hospital and is assistant professor of medicine and epidemiology at Columbia Mailman School of Public Health in New York. “Our findings suggest that pizza, pasta and foods served at dinner were more likely to have a problem.”

The study involved 804 people performing 5,624 tests by using Nima, a commercially available gluten detection device that has been shown to detect gluten at levels as low as 5 p.p.m. and 10 p.p.m.  The researchers then uploaded the crowd-sourced data. Funding from the National Institutes of Health developed Nima.

“The accuracy of our reported rates of gluten contamination is critically dependent on the underlying distribution of gluten concentrations (i.e. the percentage of positive tests containing 10, 30 or 300 p.p.m.), data that are unknown in restaurant foods,” the study said. “However, even if we were to assume that only half of the reported positive tests contained more than 20 p.p.m., cutting our estimated rate of contamination from 32% to 16%, that would still be triple the rate quantified for packaged (gluten-free) labeled foods.”

Dr. Lebwohl added, “The device can detect levels as low as 5 to 10 p.p.m., which most do not consider clinically significant. So a ‘gluten found’ result does not necessarily mean ‘unsafe for celiac disease.’ The device also does not detect certain forms of gluten, such as fermented gluten. So both false positives and false negatives will affect this estimate.”

The study said gluten-free pizza may become contaminated by aerosolized wheat flour, which contains gluten, during various stages of production. Sharing of ovens with gluten-containing pizza is another concern. It’s possible gluten-free pasta could become contaminated by being cooked in the same water as conventional pasta. Sauces also could be responsible for gluten detection on pizza or pasta. The higher percentage of gluten detection during dinner hours may reflect collective cumulative contamination of gluten-free ingredients and/or equipment over the course of the day, according to the researchers.