STATS ARTICLES 2012
Will Frequent Chocolate Consumption Make You Thinner?
Rebecca Goldin, Ph.D., April 3, 2012
A new study on the benefits of chocolate turned reporters into Oompa Loompas
“Chocolate Can Make You Thinner,” touted Discovery News; “10 Health Benefits of Chocolate,” trumpeted the Huffington Post; “A Chocolate a Day to Get Slimmer?” teased The Wall Street Journal.
It was as if Santa had decided to show up at Easter with the biggest bestest news ever for kids, cacao lovers, and the corporations that minister to their desires: chocolate wasn’t just good for you; lots of chocolate was good for you.
Drawing on a report published in the Archives of Internal Medicine, which documented a correlation between frequent chocolate consumption and lower BMI, each of these news reports added a heaped teaspoon of causality, namely that the frequent consumption of chocolates led to a lower Body Mass Index (BMI).
Reuters was one of the few news agencies to point to the problem of assuming that there was actually cause and effect at work, quoting a nutritionist from Harvard who wondered whether the causality could go the other way: people who lose weight could be rewarding themselves with chocolate.
The study would have benefited from a little skeptical reading by journalists. It consisted of an analysis of data from a survey of about 1000 people on their typical eating patterns.
Problems abound with self-reports of consumption, but in this particular case, the ways in which the data might not reflect the conclusion stand out like… a chocolate smear on a child’s face. Self-reports allow for all the weaknesses of human nature, including the one in which we underreport what we eat. The entire data set could be explained by the simple observation that some people say they eat less than they do, and particularly underreport how frequently they eat chocolate.
How? If we accept the premise that some people might underreport what they eat, and that these same people underreport how frequently they consume chocolate as well, we end up with the following situation: there is a group of people (the underreporters) who seem to weigh more per calorie consumption than others (the ones who report their consumption accurately). After collecting the data, a comparison is made: and voila, those people who eat less chocolate weigh more per calorie consumed than those who eat more chocolate. The effect could presumably be seen with cheesecake as well, but cheesecake just doesn’t have the appeal of chocolate.
Ah, says the truly devoted chocoholic. It was not the quantity of chocolate, but the frequency of consumption that correlated with lower BMI. What say you?
Again, an easy explanation comes to mind: the underreporting of frequency may itself be more frequent than the underreporting of quantity. In other words, people who underreport their overall food may feel that when he/she has a “bite” of chocolate, it’s not worth reporting; but when the same person eats a pint of chocolate ice cream, it goes on the record.
Other problems seem worth mentioning that, at the very least, call for better studies than self-reports of consumption. If the mechanism involves antioxidants, then dark chocolate and high-cacao content chocolate would be a better choice than white or milk chocolate – or ‘chocolate’ bars that are predominantly other ingredients than chocolate. This study found no relationship with the type of chocolate. It could of course be because the data wasn’t available – it may not have been collected on the survey or the number of participants was too low. However the lack of data is hardly an excuse to over-interpret the findings on the data we have. Several mechanistic effects were proposed, including that antioxidants are playing a role in fostering a ramped up metabolic system. Yet again, this could be verified by correlating the consumption of other foods containing antioxidants with lower BMI per consumed food calorie.
On the positive side of the story, there is evidence that chocolate has some redeeming health benefits. Two years ago, a very large German study found an impressive correlation between the consumption of dark chocolate and the risk of cardiovascular disease. If anything, one might wonder whether those people who eat chocolate frequently are also more likely to consume dark, rather than milk or white, chocolate. Unfortunately, little mention of this, or other, research was mentioned as the media fawned over the positive news that eating chocolate every day (well, 5 days a week) would lead to better health.
Rebecca Goldin is the Director of Research of STATS at George Mason University.