Weekly dose of chocolate lowers AFib risk? Oh how we wish it were true

Right on the heels of Tuesday’s overblown news that “even one glass of alcohol a day raises your breast cancer risk” comes another series of diet-related headlines–this time on another foodstuff guaranteed to garner publicity: chocolate.

In this case, instead of chocolate being bad, we learn it might be good: Researchers reviewing data from the medical records of people enrolled in the Danish Diet, Cancer, and Health Study found that those who self-reported at baseline eating chocolate more than once a month were less likely to have atrial fibrillation (AF) than those reporting eating chocolate less than once a month, over a 13-year timeframe.

The researchers concluded that “accumulating evidence indicates that moderate chocolate intake may be inversely associated with AF risk, although residual confounding cannot be ruled out.”

This is the kind of study that mainstream news outlets often can’t resist, and we certainly saw it make headlines.

• Reuters Health: Chocolate tied to decreased risk of irregular heart rhythm
• HealthDay: Could chocolate guard against an irregular heartbeat?
• CBS News: Chocolate linked to lower risk for heart condition AFib
• New York Times: Why Chocolate May Be Good for the Heart
• Fox News: A new health benefit of chocolate: reduced risk of AFib?

Tempered headlines, and a few words of caution

First, the good: We were pleased to see some tempered headlines–“tied to decreased risk” and “linked to lower risk” at least convey, albeit obliquely, that this was observational data and not a randomized controlled trial.

This is much better than stating flatly that chocolate “may reduce risk” of arrhythmia, which incorrectly implies a cause-and-effect relationship.

We also spotted some wise words of caution in several of the stories:

• “The study cannot say for certain that it was the chocolate that prevented atrial fibrillation, however.” — Reuters Health
• “The study wasn’t designed to prove cause and effect.” — HealthDay
• “[Researchers] note a number of limitations in the Danish study group: the participants were almost exclusively white; socioeconomic levels, which may affect health status, were not tracked; and the chocolate consumers had lower levels of other risk factors including hypertension and diabetes.” — CBS News

Yet no discussion of weakness of self-reported diet

Now for the not-so-good: All of the stories above said people “ate” chocolate.

Why is this problematic? For this study, we have no way of knowing if people actually ate chocolate–at all. Instead, all we have is their one-time self-reported responses at the start of the study, when they were asked to answer a food-frequency questionnaire, which had questions about chocolate consumption.

When it comes to diet studies, this is a big limitation, and the researchers noted it as such: “As with any study using self-reported exposure information, there is a concern of poor recall.”

None of the stories pointed this out. And this isn’t unusual when it comes to diet stories: We saw the same problem happen with another recent popular food story–wherein diet soda was linked to increased dementia risk. We also pointed it out when news stories didn’t hit hard enough on the pitfalls of self-reported data for a study on seafood consumption during pregnancy.

To me, it doesn’t seem too difficult to swap out “women who ate chocolate once a month” to “women who self-reported at the start of the study eating chocolate once a month” and then add a line or two about the notorious challenge of recall bias in diet studies.

Why? In this case, given how wildly popular chocolate is, many readers (and chocolate makers) will be restless to grab on to the hope that chocolate may be medicine. Especially when the mainstream news media so often (and usually so superficially) tells us “science has shown” that many of the other foods we love (red meat, sugar, to name just a couple) are probably harmful.

Science is messy, and we shouldn’t hide that fact

Readers deserve to know that actually, while the research may be intriguing–and possibly worthy of further exploration–this study and so many other diet studies like it just aren’t conclusive of anything. They’re good starting points, shining a light on possible trends, but fraught with limitations.

For example, buried at the bottom of HealthDay story, long after the lead claim that chocolate “might help keep a common and dangerous form of irregular heartbeat at bay,” was another crucial qualifier in the form of a quote from an independent expert:

“…it appears that people who do regularly consume chocolate are also those patients who had less health issues such as diabetes and high blood pressure.

“As these other health issues are known to predispose people to atrial fibrillation, it is hard to say whether eating chocolate was protective or if this population is generally less predisposed to irregular rhythms,” Bond said.

I wish these types of issues received more attention. Not only does this inform readers of the realities of science, it helps explain why we so often see future studies contradict the earlier ones. Like a Fudgsicle on a summer day, emerging science can be messy.

In other words, we’re far from seeing “take a weekly dose of chocolate” written on a prescription pad. Though one can dream.