Joy Victory is deputy managing editor of HealthNewsReview.org. She tweets as @thejoyvictory.
When it comes to preventing cancer, does consuming soy help you or hurt you?
Anyone well-versed in diet studies knows that there is no final word on this question–that due to the extensive limitations of most nutrition research (including the latest study on this topic), any declaration of benefit must come with a heaping helping of caution.
Yet, if you read some of the news headlines about this most recent study, you won’t find that takeaway. Here’s a sampling:
Our favorite headlines from this batch are decidedly different:
Why do we like these better? Instead of declaring that the study has conclusive findings, these headlines nicely hint that the research in this area is uncertain and continuing to evolve. (Unfortunately, this framing didn’t carry over into the stories themselves. CNN stated that this latest study offers proof that the question is “settled” and TIME stated it “clears up confusion.”)
But as you can see in the other headline examples, for the most part, there was a tendency to frame the study as either:
1) showing that soy is beneficial (ex: “Soy tied to longer life after breast cancer“)
2) showing that soy isn’t hazardous (“Soy doesn’t worsen breast cancer and may prevent it, study finds“)
Neither style of framing is ideal, but if we had to choose the lesser of two evils, it would be the latter, which gives at least some sense to readers that the science isn’t settled.
Yet we’re not crazy about those examples, either. As with many diet studies, this one is based on observational research that’s notoriously full of limitations and caveats. Researchers can find correlations with those types of studies, but they can’t find cause and effect. Foods and nutrients linked to disease in these types of studies are often found not to have any direct impact in later, more authoritative research.
So, how would we have written a headline based on this study? As we suggest in our primer on writing better headlines, a good starting point is to back away from helping verbs like “may, might, can, could.” It’s also a good idea to ditch active verbs (“reduces,” “protects,” “worsens”), which suggest direct impact, when the research doesn’t support cause-and-effect conclusions.
And lastly, in this case–and with many diet studies–get comfortable with expressing uncertainty.
As an example, instead of “Soy doesn’t worsen cancer and may prevent it, study finds” (which is too conclusive considering the limitations of the research), try “Study shows link between soy and lower cancer risk, but questions remain.”
This way, you’ve summarized the research, but you haven’t oversold it as the final word. And when another study comes along and contradicts the findings of this one (as they so often do), your headline stands the test of time.