How do they calculate those calories?
In the wake of the study, these companies have reexamined their products' nutritional information and updated the calorie numbers. A prepared statement from Denny's also noted that food is prepared fresh by individual cooks and variations can occur. The company added, "All restaurants are granted a 20 percent leeway in reporting calorie counts because food portioning and ingredients may vary … when a dish is made." A spokesman for Lean Cuisine made a similar point: "Since there is natural variability in ingredients, our label values may vary up to 20 percent."
Margin of error
That 20 percent margin of error is acceptable, according to the U.S. Food and Drug Administration's food labeling regulations. A food is only considered misbranded if the nutrient content "is greater than 20 percent in excess of the value … on the label," says Siobhan DeLancey of FDA's Office of Public Affairs.
The reasons for the 20 percent allowance is because food samples are not always exactly the same. Take a chocolate chip cookie with nuts, for example. "One cookie may contain more chips or walnuts than another, which may change the nutritional profile for that particular cookie," DeLancey wrote in an e-mail. Most importantly, she added, testing methods "are not necessarily 100 percent sensitive."
Lean Cuisine and Weight Watchers respond
Both Lean Cuisine and Weight Watchers also noted that they base their products' nutritional information on "an average value taken from multiple samples in accordance with government regulations," as Weight Watchers' spokesman Tracey Parsons wrote in an e-mail. The Tufts study tested only one sample of food from each company and restaurant.
To test the calorie counts of the sampled foods, Roberts and her researchers blended the food, freeze-dried it and burned the resulting powder to see how much heat it produced — a measure of its energy content, or calories.
What they found, she says, was that many "mixed" foods — such as packaged meals like chicken and pasta or steak and vegetables — contained more calories than their manufacturers stated. She attributes the discrepancies to the fact that manufacturers don't typically have the equipment to test calories; instead, they rely on calorie counts for individual ingredient foods as listed in the government's database.
In mixed foods, she says, the ingredient levels might not be precise, meaning the calorie count numbers are off, too. For example, Roberts and her team tested several frozen meals by Lean Cuisine, Healthy Choice and Weight Watchers and found that some contained more calories than the manufacturers stated on the nutrition label. A few foods were underestimated. Lean Cuisine roasted garlic chicken, for example, had 4 percent fewer calories, the researchers say.
Roberts acknowledges that by testing only one sample, researchers couldn't conclude that these meals are consistently off the mark. And although the calorie differences were slight in many cases, for someone using a daily calorie goal to lose or maintain weight, even slight differences can add up to big effects.