If you buy it at the grocery store….

Scientists know that asking people to report on their own diet is a flawed means of finding out what people eat. People tend to minimize eating things that are bad for them (the "self-report bias") and over-report eating what they perceive to be good for them (the "social desirability bias"), because they don't want to appear to be unhealthy. It's understandable, but it makes it difficult to accurately assess the effects of diet on health.

Researchers at Morehead State University in Kentucky thought of a novel way to assess people's diets: the grocery store receipt (Nutr J 2006;5:10-17). They stopped 50 people outside of a local grocery store and asked them to give up their grocery receipt and answer a few questions in exchange for $1. The respondents all appeared to be between 20 and 40 years of age, had full grocery baskets, and most admitted that they were, in fact, shopping for themselves and their immediate families. They were asked if what was in their cart was representative of what they ate; only two people said it was not and were eliminated from the study.

Each person was asked to look at a series of pictures of body silhouettes depicting various body sizes and was asked to match their own body silhouette and that of their spouse and children (if any) to one or more of the silhouettes. These body silhouettes have been shown by other researchers to be fairly reliable in terms of estimating Body Mass Index, and were used that way in this study. They were also asked how often by day, week, or month, they ate fast food; whether they or anyone in their household smoked; and whether they exercised.

The grocery receipts were analyzed and most of their contents put into three categories: fats, oils and sweets (such as sugared soft drinks, regular potato chips, cookies, and butter); processed foods (such as hot dogs, bacon, and non-diet frozen meals); and low-fat/low-calorie substitutions (such as skim milk, diet soft drinks, and baked potato chips).

Not too surprisingly, a comparison of the body silhouettes for each person and their family correlated fairly closely with what was in their shopping cart: in households where no one was perceived to be overweight, a smaller proportion of their grocery bill came from the fats, oils, and sweets category and a larger proportion from the low-fat/low-calorie category than those respondents where at least one person in the household was perceived to be overweight. It is interesting to note that 10 of the 48 respondents (almost 21%!) said that they ate fast food 30 times per month.

What this means for you:

Although this is a small study, the results appeal to common sense: If you buy it at the grocery store, chances are that's what you eat. How much of your food dollar is being spent in the frozen foods and packaged foods aisles as opposed to the produce and fresh foods aisles?

First posted: August 25, 2006