Diet Soda Still Under Attack

People using social media too often read headlines and don’t look at the rest of the story or, perish the thought, do a little research to find out the truth. We see articles posted where an event happened in another country but the comments read as if it happened right next door. (I highlighted an example of some of the dangers of this kind of behavior, blindly reposting stuff, earlier this month.)

That is why I wasn’t surprised to run across another condemnation of diet soda being spread on Facebook a few weeks ago when the original news story came out last February 14. (Does the American Heart Association (AHA) always release a story on Valentine’s Day or was that just a coincidence?) This story presented a good example of how a headline can be very misleading, and how 15 minutes of research leads to a different conclusion.

Multiple news outlets reported it, but the Fortune headline was typical: “Drinking Two or More Diet Sodas a Day Increases Likelihood of Strokes, Heart-Attacks, American Heart Association Says.” Although they have been seen as a healthier alternative to sugary drinks, the study shows that “immoderate drinking of the low-calorie drinks is associated with a greater risk of strokes or heart attacks.” 

They go on to say that over 82,000 women participated, which is a very large sample, but it means the findings can apply only to women. Furthermore, and not mentioned in the Fortune article, they were post-menopausal women between the ages of 50 and 79.

Part of the findings correctly reported was that the women “who drank two or more artificially sweetened drinks had a 23% higher risk of strokes in general.”

A more thorough report came from Science Daily. They also report the 23% figure, but add near the beginning of the piece an important reminder for non-statisticians. “While this study identifies an association between diet drinks and stroke, it does not prove cause and effect because it was an observational study based on self-reported information about diet drink consumption.” Also researchers can determine no mechanism by which the ingredients in diet drinks could cause the associated ailments.

A further question arises: How bad is a 23% increase in risk? That depends not on the size of the sample, but on the size of the risk. If, for example, the overall risk of something bad happening is 60%, a 23% increase brings the risk up to almost 75% - a big difference. On the other hand, if the risk is small, say 2% such an increase brings the risk to less than 2.5%. Without specifying the original size of the risk, the unanswered question is: how much behavior should I change to try to avoid it?

The percentage increase says nothing about the real size of the increased risk. To find out how serious it was I went to the AHA Journal article cited in those news stories. I found the data in Table 2 of the study.

The data showed that the risk of a stroke over the twelve years of follow up was about 3% for the group that drank almost no diet drinks and 3.7% for those who regularly drank two or more. Remember, this very small risk applies only to post-menopausal women. And also mentioned in the conclusion of the AHA study, those women in the higher-risk group “were more likely to be obese, had lower levels of exercise, had … lower diet quality… [and] were more likely to be a current smoker.”

A different study with a smaller sample asking the same question was published two years earlier by Harvard Health. It concluded that the risk from diet soft drinks is small. They also referred to a larger study that “detected a slightly higher risk of stroke in people who drank more than one soda per day, regardless of whether it contained sugar or an artificial sweetener.” (That flies in the face of the aspartame urban myth that won't seem to go away.) And agreeing with the conclusion in the AHA study, for people drinking diet sodas, “heightened stroke risk may result from their health problems rather than their beverage choice.”

An honest headline should have read: "Postmenopausal women who drink two or more diet drinks a day seem to have a slightly higher risk of strokes but what they drink might have nothing at all to do with it" - not exactly click-bait.

It takes just a little critical thinking and research to uncover the misrepresentations in so many headlines. If you don’t care enough about the issue to do the research, ignore the headlines. If you do, don’t repost them on social media until you've looked into the matter.