There have been so many stories about the possible health benefits of cocoa flavanol-derived chocolate over the years and so many studies purporting to document these benefits that it is useful to have a meta-analysis of the available clinical trial evidence. This release is based on a review of 19 randomized controlled trials that studied the effects of cocoa flavanols on biomarkers of heart health.
We expected to see a lot of good data in this release but found very little. Instead we found strong statements about the findings being “significant” or about how the “benefits they were evident for both women and men.” There were a few caveats sprinkled in too, helpfully. The study authors are quoted saying the review and meta-analysis point to the need for larger, longer clinical trials to judge whether these changes in biomarkers translate into actual improved outcomes for people. And there was an acknowledgment that some of the funding for the work came from one of the world’s biggest chocolate makers.
What’s the harm with another story or another news release about chocolate, or cocoa flavanols? There isn’t any — if the information is balanced and complete. The problem with so much of the information about chocolate is that it’s usually imbalanced and incomplete, superficially shallow, and treating chocolate as a no-brainer superfood.
The lack of specificity in this release, for example, would not allow a reporter to know whether this story was worth pursuing. And if reporters did report on this unquestioningly, the general public would not be getting all the information they need.
There is no discussion of costs in the release. We didn’t expect to see chocolate bar prices. But there are existing dietary cocoa flavanol products on the market that could easily have been referenced.
Only 7% of the releases we’ve reviewed so far adequately address cost. It has to start somewhere. Each news release writer can say to him/herself, ‘Not in this case.’ And in 93% of cases so far, cost is inadequately discussed.
There is no quantification of the benefits.
The release would have been improved with an example of one or two biomarkers that were different in the higher flavanol vs less intake groups. When the release notes that study participants “saw significant declines in blood glucose and insulin,” adding a number or range would have been helpful.
The authors used the “weighted mean difference” as a measure which may be cumbersome to describe in a news release. However, some quantification was needed here and there was no attempt made to add any.
We think that you can’t have a release that talks about how “tempting” the results of a new chocolate-is-good-for-you story without talking about the countervailing wind that is the high calorie count of so many chocolate products. If everyone starts having a chocolate milkshake at lunch after reading this study, our obesity epidemic is going to be even worse.
The release included a quote from the study author who noted that “the findings from the current study apparently shouldn’t be generalized to different sorts of chocolate candies or white chocolates, of which the content of sugar/food additives could be substantially higher than that of the dark chocolate.”
But that didn’t go far enough. We think there was an opportunity here to discuss the real health implications of adding cocoa flavanols to the diet in whatever form it might take.
Only 37% of all releases we’ve reviewed are graded satisfactory on harms. The discussion on an intervention’s harms needs to start somewhere, and a news release is often the headwaters of the news stream.
We liked how the release explained the specifics of the metanalysis and also talked about how small most of the studies were and that many of them were short in duration.
The release overall was very clear about the goal of the review and meta-analysis — both to quantify the differences in biomarkers and to determine whether the body of evidence is adequate to support additional research.
There is no disease mongering in this release.
The release makes the funding sources clear.
In addition to Lin and Liu, the study’s other authors are Isabel Zhang, Alina Li from Brown University, JoAnn Manson, Howard Sesso and Lu Wang from Brigham and Women’s Hospital.
The authors acknowledge funding from the National Institutes of Health including the National Heart, Lung and Blood Institute, the American Heart Association, and Pfizer. Co-authors at Brigham and Women’s also noted receiving funding from Mars Symbioscience, a research segment of Mars, Inc., which makes chocolate products.
There is no comparison of alternatives in the release. Are statins a much bigger factor, for example, in cardio health? What about omega-3 fatty acids? What about just eating a balanced diet overall?
The release would have been improved with a mention that modest weight loss and exercise are standard ways to improve metabolic markers for cardiovascular disease.
The release notes that cocoa flavanols can be obtained from various sources including drinks (cocoa) and dark chocolate. It’s common knowledge that these are widely available. It’s less widely known that dietary cocoa flavanol products are available and it would have been useful to point that out.
The release did not establish the novelty of the meta-nalysis itself nor the underlying studies.
There was no unjustifiable language in the piece. In fact, we thought it did a pretty good job bringing in the appropriate caveats and using language like: “There were small-to-modest but statistically significant improvements among those who ate flavanol-rich cocoa product vs. those who did not.”
The release was very measured and like the scientific review, did not overstate the findings.