Calorie Labels Continue to Have No Impact On Our Food Choices

In 2008 New York City became the first to require the posting of calorie counts in restaurants. In 2010, as part of the Affordable Care Act (Obamacare), the FDA was given control over enforcement of these same calorie labeling requirements nationwide. In this last year, New York Mayor de Blasio, upset at the slow pace of the federal response, ordered the City to begin enforcing the new calorie labeling requirements. The federal government, in turn, blocked that move as preempting their jurisdiction. And now, just a short ten years into policy we have a meta-analysis by the Cochrane group on the efficacy of these policies – shock upon shock, there is no evidence that they do any good.

The study began with an extensive literature review identifying 28 studies that met their criteria for objectivity and readily identified outcomes. Eleven of the studies were in real-world settings like restaurants or grocery stores; the remaining 17 were based on studies conducted in experimental settings. About a third had elements of reporting bias primarily associated with issues that might have resulted in the randomized studies being less than truly random. The quality of the evidence was in the best of cases “low” in their words, “our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.”

Cochrane’s findings and conclusions

• Nine studies considered labeling in restaurants, cafeterias or coffee shops showing a decrease in calories purchased in the presence of labeling by 47 calories.
• Seventeen studies considered labeling in laboratory settings and resulted in no significant change in calories consumed in the presence of labeling
• Four studies examined labeling when participants were offered a snack, like cookies, soda or a single entrée, labeling did not result in fewer calories consumed
• Five similar meta-analysis papers were also briefly reviewed, four agreed with the findings of this current Cochrane review. The fifth showed no changes for real-world situations but a decrease in calories consumed in experimental settings.

They concluded “a small body of low-quality evidence suggest[s] that nutritional labeling comprising energy information on menus may reduce energy purchased in restaurants. … Accordingly, and in the absence of observed harms, we tentatively suggest that nutritional labeling on menus in restaurants could be used as part of a wider set of measures to tackle obesity.” Italics added.

Consider for a moment those italicized words; there is not a definitive word in the bunch. The Cochrane Library, as an organization, invented meta-analysis, their work demonstrates the acknowledged, proper use of the technique. The best they could say was that labeling, in the absence of observed harms, could be used. The term, damning with faint praise, comes to mind. Note also that phrase, "absence of observed harms," surely no harm comes to consumers, they are merely ignoring the labels; but what of the cost to make these calorie determinations and the new signage; are the costs to business, not harm? Finally, as for the 47 calorie decrease, that is about 15% of a medium McDonald’s French Fries, a half an apple or a quarter cup of avocado.

While we may well try to diet our way out of the rising incidence of obesity, calorie labeling does not appear to be particularly effective. Mary Bassett, New York City health commissioner, sums up the city’s point of view, “The city is prepared to defend its right, independent of FDA action, to enforce its requirements that give New Yorkers the information they need to make informed dietary decisions.” Why let the best evidence we can find, evidence that is both poor and low-quality ruin a good regulatory narrative?