Mmmmm. Chocolate cake. As in “cake” with “chocolate” in it. It’s a straightforward, yet somehow radical, concept. Except it’s not that radical. It’s just everyday, common-or-garden, run-of-the-mill chocolate cake. Tastes nice, but is not good for you; eating lots of it makes you fat, etc., etc., etc. Everybody knows this. I mean, if you were to stop eating, say, fruit for breakfast every morning, and started to eat chocolate cake for breakfast instead, that would just be crazy, wouldn’t it? For one thing, it would make you put on weight, that would. Wouldn’t it?
But have a look at these headlines: “Want to lose weight? Eat chocolate cake for breakfast, say researchers” (Daily Mirror), “Chocolate cake breakfast could help you lose weight” (Daily Telegraph), “Happy Chocolate Day! Chocolate cake for breakfast could help you lose weight” (Huffington Post), “Dessert with breakfast can help weight loss” (MSN New Zealand), “Desserts at Breakfast Could Lead to Weight Loss (Really!)” (Allure Magazine), and many more.
It’s those darned “researchers” again, turning the breakfast tables on us and grabbing the media’s attention. According to these guys, it seems as though eating chocolate cake at breakfast actually helps you to lose weight, rather than put it on. Whodathunkit? I guess that’s why people who eat lots of cake are just so thin.
And that’s that, I suppose. End of. Nothing to see here.
But hang on a minute…
…SERIOUSLY!? Eating cake for breakfast HELPS YOU LOSE WEIGHT? Are you MAD? Well, no not mad. Just mistaken. You see, while the various headlines circulating on this story are kind of right, they are actually only half-right. Which makes them half-wrong. So, in other words, they’re wrong.
Let’s do some parsing. The basic shape of these headlines is something like “X could help Y“, with X signifying chocolate cake for breakfast and Y signifying weight-loss. So “X could help Y” becomes “Chocolate cake for breakfast could help you lose weight“.
If you want to use technical language, we would say that Chocolate cake for breakfast (i.e., X) is the independent variable, while weight-loss (i.e., Y) is the dependent variable. The first is called “independent” because you can choose to do it or not to do it; the second is called “dependent” because it depends on the first.
However, the original research was a little more subtle than that. First of all, the researchers were looking at a specific group of people, who were (a) already overweight and (b) in the habit of eating cake anyway. In other words, they were not looking at average-weight people who simply might be thinking about adopting a cake-for-breakfast lifestyle. Secondly, the researchers were not testing whether chocolate cake affected weight, but instead were looking at the impact of the time-of-day at which people ate their cake.
Following a controlled study, the researchers found that their sample of already-overweight habitual cake-consumers were more likely to lose weight if they moved their carbohydrate consumption to the mornings. (This was most likely because their metabolisms were more active at that time, meaning that they burned off more calories than they would have in the evenings. In addition, they had the rest of the day to move about a bit and burn off some more, rather than simply going to bed shortly after eating, thereby allowing all that fat stuff to be absorbed into their bodies unimpeded.)
So to revert to technical language, you could say that the headline writers were failing to realise that the independent variable — chocolate cake consumption — was being specifically qualified by a moderating variable — time of day. In reality, it was two variables rather than one.
In other words, the correct format for the headlines should have been “X for Y could help Z“, with X signifying chocolate cake, Y signifying breakfast, and Z signifying weight-loss. And the headlines should only apply to people who eat cake on a daily basis anyway.
So headlines like “Chocolate cake for breakfast could help you lose weight” are missing some critical information. First of all “you” doesn’t mean “you” unless “you” are an overweight person who eats cake on a daily basis. And secondly, it should be clarified that the key variable here is “breakfast” and not “chocolate cake”. So the proper headline should be something like this:
Chocolate cake for breakfast — in contrast to chocolate cake at other times of the day — could help you lose the weight you put on in the first place by eating cake in the afternoons and evenings
Now I admit that isn’t particularly catchy. That’s often the problem with accuracy. However, something like this would have been both catchy and accurate:
Want to lose weight? Then LAY OFF THE CAKE!
But that wouldn’t have been news. And that’s something of a dilemma if you’re a newspaper (or a person who wants to eat cake every day and would like an evidence-base for doing so).
I guess we just can’t have our cake and eat it…
* * *
Related on The Science Bit
What women want…to be exploited by capitalists
Stop the spread of the ecological fallacy
A study not worth its salt?
Brian Hughes is an academic psychologist and university professor in Galway, Ireland, specialising in stress, health, and the application of psychology to social issues. He writes widely on the psychology of empiricism and of empirically disputable claims, especially as they pertain to science, health, medicine, and politics.