Tufts University researchers discovered inaccuracies in calorie information voluntarily provided by chain restaurants in a study published in the July 20 issue of the Journal of the American Medical Association.
While 81 percent of food items tested from restaurants such as Boston Market, Burger King, Domino’s Pizza, Olive Garden, McAlister’s Deli and Arby’s contained the number of calories the chains’ websites said they did within a 20 percent margin of error, some items were major outliers.
The big discrepancies occurred largely with low-calorie items and side dishes.
The Food and Drug Administration allows for packaged goods to be off by as much as 20 percent.
“Overall, we were pleased that the study showed that, on average, the stated calorie count was accurate,” said Joy Dubost, director of nutrition and healthy living for the National Restaurant Association.
“I was frankly surprised at the blaring headlines related to the study,” said Scott Vinson, vice president of the National Council of Chain Restaurants, adding that any handmade item will have some variation, certainly more than manufactured packaged goods.
The researchers used a bomb calorimetry technique — which measures calories released when the food is burned — to test 269 items from 242 restaurants in Massachusetts, Arkansas and Indiana between January 2010 and June 2010.
Most of the items tested were within 10 calories of what restaurants said, but some were way off, particularly at casual dining restaurants.
Some of the biggest errors in calorie counts belonged to soups, salads and side dishes.
Among them was a minestrone at Olive Garden that had 165 calories instead of 111, a Zuppa Toscana from the same restaurant that had 391 calories instead of 191, a wedge salad from Outback Steakhouse that had 1,035 calories instead of 376, white cheddar mashed potatoes from Ruby Tuesday that had 277 calories instead of 136, and white rice from P.F. Chang’s China Bistro that had 560 calories instead of 226.
Dubost noted that most of those items would be served in larger portions in takeout orders just to fill the containers.
“If you think about it, they might not have the appropriate package size to reflect the portion that would be on the plate. So they might have provided more to fill the container,” she said.
“The researchers were aware of that,” she added, and they stated that as part of their findings.
Vinson said the study results indicated a high level of accuracy in calorie declaration. “That’s a good thing, and it shows the process is currently working.”
“But the restaurant industry is not the packaged food industry,” he added.
Bruegger’s did fairly well in the study, but the researchers found its blueberry muffin to have 601 calories, not the 455 its website reports.
Bruegger’s executive chef Philip Smith responded to the study: “We strive for accuracy, but because our items are baked and prepared by hand in each of our bakeries, there can be some variations, though the calories per ounce remain the same. We understand that the predictability in calorie count is important, and Bruegger’s is revisiting its baking procedures with our staff to make sure baking guidelines are followed and portion sizes are accurate.”
Au Bon Pain’s biggest outlier was its steakhouse on ciabatta sandwich, which had 851 calories, 17 percent more than the 740 calories it stated.
“The risk you take when you make items by hand is that the person making the sandwich or salad can over-portion or under-portion,” Au Bon Pain spokesman Ed Frechette said.
“It can also happen with our hand-cut pastries,” he noted. “We think guests would understand that possibility and would have a tolerance for a slight deviation from the listed calories.
“At Au Bon Pain, we have controls and measures in place to prevent this from happening … If we were to be off by a large amount, say 100 calories, it would undermine, understandably, our guests’ confidence in our brand. For a brand that prides itself on transparency, Au Bon Pain would find this unacceptable.”
The Food and Drug Administration is still wrangling details of a national menu labeling law, but Vinson noted that Congress’s statute explicitly stated that the “reasonable basis standard” be used to measure calories and enforce compliance.
That standard, which Vinson said has been in effect for about 20 years, dictates that restaurants use a standard and objective third-party source to analyze menu items, whether through lab analysis or standardized references recognized by the FDA to determine calorie content, and then to make reasonable efforts to have procedures in place and staff trained to consistently make menu items.
However, the FDA has stated that it plans to use the 80-120 rule, or the 20-percent variance allowed for packaged goods, when enforcing the law.
Vinson said the NCCR wrote in its comments to the FDA that such enforcement violated the statue, and that the bill’s sponsors also wrote saying they did not intend for the law to veer from the reasonable basis standard.