When it comes to judging the healthiness of packaged foods, nutrition content, not health claims, is the best way to go. But the health claims on the front of the package often get in the way.

The results show just how powerful front of package health claims can be, and how misleading they often are when taken out of context.

A study recently conducted by the FDA tested the effect of a front-of-package "low- carbohydrate" claim on people's perception of how healthy packaged bread and frozen dinners were. Consumers perceived these foods to be considerably healthier when their nutritional content (the nutrition facts panel from the back of the package) was missing.

Participants who saw only front-of-package claims rated products labeled low-carbohydrate as more helpful for weight management and lower in calories than the same products without a low-carbohydrate claim. Bread with a low-carbohydrate claim was also rated as more healthful than bread without the claim. But when a nutrition facts label was also present, the presence or absence of a low-carbohydrate claim did not affect a product's rating.

In other words, the low carbohydrate claim made a big impression on anyone who didn't read the nutrition facts panel. But once people saw what was actually in the food, that low-carbohydrate claim seemed unimportant.

The study was performed over the Internet on 4,320 volunteers. The volunteers were randomly shown 1 of 12 different package labels for bread or a frozen dinner. Seven of the twelve labels included nutrition facts information. The volunteers then answered questions about the food's perceived healthfulness, helpfulness for weight management and calorie content.

The results show just how powerful front of package health claims can be, and how misleading they often are when taken out of context.

The researchers note that the study conditions made reading the nutrition information much easier than it would be to most shoppers. It's one thing to read a label off a computer screen while sitting at home. Reading the label in a crowded supermarket aisle as shopping carts, strollers, wheelchairs and shelf stockers all go racing by can be much more difficult. But the study shows that it might be worth the effort to do so.

The study dealt with a true label claim: that a food was low in carbohydrates. An editorial in Scientific American deals with a more sinister type of food labeling claim: one that isn't even based in reality.

Until May 2009, boxes of Cheerios claimed that eating the cereal could lower cholesterol by 4% in six weeks. That's when the FDA sent the manufacturer a cease-and-desist letter, since there's no good scientific evidence that Cheerios can do so. One can only speculate on what else a person had to do, besides eating Cheerios, to lower their cholesterol by 4%.

In March 2010, the FDA sent warning letters to 17 food and beverage manufacturers about false or misleading health claims on their products. One of these was to a manufacturer of decaffeinated green tea that claimed to be protective against Alzheimer's disease, rheumatism and cancer.

As the editorial notes, this is an unusually activist approach by the agency. But even so, these foods with their dubious health claims were sold to unwary shoppers for months before the health claims were removed from the packages. While applauding the FDA for taking a step in the right direction, the editorial's authors think that the FDA's actions "still do not go far enough to protect consumers from dubious marketing."

They think that the FDA should hold food manufacturers to the same standards drug manufacturers are held to and require them to provide convincing evidence of all alleged health claims before releasing products for sale. They point to a similar system, currently in effect in Europe, where 80% of such health claims have been rejected.

Industry representatives say that having to prove their health claims would cost too much and take too long. Apparently, truth is an expensive commodity.

In the absence of such a system, it might be best to ignore all front of panel health claims and skip straight to the nutrition information to judge how healthy a food is.

An article detailing the FDA study appears in the September 2010 issue of the Journal of Nutrition Education and Behavior. The editorial, Snake Oil in the Supermarket, appears in the September 2010 print issue of Scientific American and is freely available online at the journal's website.