Beating Charles Sumner in the U.S. Senate

Diet Soda: Beating a Correlation to Death

How many times have we documented a correlation between diet sodas and cardiovascular disease? Who cares, say the editors of Stroke. Apparently, the click bait is irresistable. Thus, we have the upteenth correlation study, unsupported suggestions of causality, and a tidal wave of sensational headlines about diet soda, strokes, heart attacks, and death.

We have a word for this and that word is wrong.

Data from the Women’s Health Initiative

The study used data from 81,714 individuals in the Women’s Health Study. The results hinge upon only one data point about low-calorie beverages. It’s a self-report, not a measurement. Three years after these subjects enrolled in the study, they estimated how much they routinely drank artificially sweetened beverages. They gave no information on exactly what they drank. Plus, the study has no information on usage over time.

Just a single self-report for a study that followed these women over more than a decade.

From that, the investigators concluded that artificially sweetened beverages are linked to an increased risk of stroke (23%), coronary disease (29%), and death (16%). Only in passing do they mention the problem of residual confounders and reverse causality. These things are called diet drinks for a reason. And the reason is because the consumers who prefer them are concerned about their body weight.

Our president captured it in a tweet: “I have never seen a thin person drinking Diet Coke.”

Sensational Headlines Powered by AHA

Nonetheless, the American Heart Association issued a press release about diet drinks and the risk of stroke in postmenopausal women. It touched off a storm of headlines.

Those headlines shout about “Real Risks,” not to mention dire predictions of strokes, heart attacks, and death with just two cans of diet soda per day. But the fine print at the bottom says nevermind. It’s “only an association.” Too late. Everyone passes out from fright before they get to that part. So no one reads it.

Confirmation and Familiarity Bias

In sum, we have confirmation and familiarity bias run amok. People who believe that diet drinks are wrong love this study. Even people who should know better give into the bias of familiarity. When they hear an assertion over and over again, they start believing it. Thus, American Academy of Neurology President Dr. Ralph Sacco says:

This is another confirmatory study showing a relationship between artificially sweetened beverages and vascular risks. While we cannot show causation, this is a yellow flag to pay attention to these findings

The more studies there are coming up with the same associations, the more you begin to question. The more you begin to feel strongly about the association being real.

Strong Feelings

Sacco put his finger on it. This is all about feeling strongly. Not science. Not facts. Just feelings. In a similar vein, we know people who have strong feelings about going to the hospital. It’s associated with death. But an association is not adequate evidence for causality.

No matter how many times you repeat the observation.

Click here for the study and here for the companion editorial. For a sample of the reporting, click here .

Beating Charles Sumner in the U.S. Senate, lithograph by John L. Magee / Wikipedia

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


February 18, 2019

5 Responses to “Diet Soda: Beating a Correlation to Death”

  1. February 18, 2019 at 7:27 am, Joe Gitchell said:

    Thank you for this, Ted. You’ve made some strong and trenchant points.

    In a similar vein but dealing with nicotine and vaping, check out the most recent three posts from BU’s Dr Michael Siegel, below.

    Akin to your closing point about never going to the hospital, you might want to drop health insurance, and quickly!


    • February 18, 2019 at 8:53 am, Ted said:

      Ha. Health insurance will definitely kill you!

  2. February 18, 2019 at 9:18 am, John DiTraglia said:

    Drinking diet soda is probably harmless but still stupid.

    • February 18, 2019 at 9:31 am, Ted said:

      I am very close to a brilliant person with a PhD. She drinks Diet Coke. She is definitely not stupid, nor is her choice to drink it. She enjoys it. It brings her pleasure and almost certainly will not harm her health.

  3. February 18, 2019 at 10:24 am, David Brown said:

    From about 1978 onward I collected newspaper articles about the connection between heart disease and saturated fat, cholesterol, and sugar intake. In 25 years only two articles about sugar made their way into a file folder. Meanwhile, I accumulated dozens of articles yearly warning consumers to reduce cholesterol and saturated fat intake. Note that the American Heart Association (AHA) did not even begin to warn consumers to reduce intake of foods and beverages containing added sugars until 2006. In 2009 the AHA finally published a statement proposing a specific upper limit for sugar intake. One wonders why it took so long for AHA scientists to notice that excessive sugar consumption is problematic for health.

    Detecting a link between artificially sweetened beverage (ASB) consumption and vascular diseases is probably an exercise in futility. The magnitude of the effect as compared to other factors is too small.

    Actually, the same can be said of saturated fats. I’ve been eating a diet rich in saturated (mostly dairy) fat for decades. Unfortunately, my consumption of the omega-6s linoleic acid and arachidonic acid has been excessive. By steps, as the consequences of excessive omega-6 intake manifested themselves in the form of a skin ulcer, leg pain due to varicose veins, and shoulder pain. I have reduced intake of soybean oil, mayonnaise, peanut butter, and 99% fat-free turkey. My weight and blood pressure remain steady, I no longer develop a chronic cough during the cold months, and I seem to be immune to colds.

    Meanwhile, during American Heart Month, health experts and journalists churn out articles admonishing consumers to reduce saturated fat consumption or swap saturated fats for vegetable oils. And the American Heart Association continues to support such advice based on studies that indicate a 30% reduction in heart attack events and deaths if AHA advice is followed.

    By way of contrast, The Lyon Diet Heart study produced a 70% reduction in heart attacks and deaths by reducing linoleic acid intake.