by Marion Nestle
Feb 10 2011

Do diet sodas really cause stroke? I’m dubious.

I’ve been asked repeatedly this week to comment on the huge press outcry about a study that links diet sodas to an increased risk of stroke and heart disease.

I have not seen the study and neither has anyone else. It is not yet published.

It was presented at the American Stroke Association’s International Stroke Conference 2011.  The American Heart Association has a short summary on its website.  And Rosie Mestel has an excellent account in the Los Angeles Times.

Here’s what I can glean from the limited information available:

  • The study started in 2003.  It was designed to determine risk factors for heart disease and stroke in a multi-ethnic New York City population.
  • It used a food frequency questionnaire to ask about 2,500 people how often they drank diet sodas (among many other questions).
  • Nine years later, it assessed rates of stroke and heart disease.
  • The result: people who said they habitually drank diet sodas had a 60% higher rate of stroke and heart attacks.
  • They had a 48% higher rate when the data were controlled for contributing factors: age, sex, race, smoking, exercise, alcohol, daily calories, and metabolic syndrome.

That is all we know.

Does this study really mean that “diet soda may not be the optimal substitute for sugar-sweetened beverages for protection against vascular outcomes,” as the lead author is quoted as saying?

As Rosie Mestel puts it:

It’s worth noting, as some scientists did, that this is a link, not proof of cause and effect. After all, there are many things that people who slurp diet sodas every day are apt to do – like eat a lousy diet — and not all of these can be adjusted for, no matter how hard researchers try. Maybe those other factors are responsible for the stroke and heart attack risk, not the diet drinks. (Those who drink daily soda of any stripe, diet or otherwise, are probably not the most healthful among us.)

Leaving questions about the accuracy of dietary information obtained by questionnaire, the study raises more important questions:

  1. Could this finding simply be a statistical result of a “fishing expedition?”  The food frequency questionnaire undoubtedly asked hundreds of questions about diet and other matters.  Just by chance, some of them are going to give results that look meaningful.  The increase in stroke risk seems astonishingly high and that also suggests a need for skepticism.
  2. What is the mechanism by which diet sodas lead to stroke or heart disease?  I can’t think of any particular reason why they would unless they are a marker for some known risk factor for those conditions.

Please understand that I am no fan of diet sodas.  I don’t like the metallic taste of artificial sweeteners and they are excluded by  my “don’t eat” rule: never eat anything artificial.

But before I believe that this study means that artificial sweeteners cause cardiovascular problems, I want to see a study designed to test this particular hypothesis and a plausible biological reason for how diet sodas might cause such problems.

  • Michael

    Marion’s blog normally attracts a mixture comments: some fluff, some obvious antiphonies from the nutrition community choir songbook, and some serious, nuanced discussion. I am amazed by how weighted the responses to this post are by people who condemn any serious analysis of the study in favor of the belief that any study that seems to show a problem with diet sodas either *must be* true, or that any doubts should be ignored or suppressed in favor of getting rid of the stuff, because the writer thinks they’re somehow bad for one for other reasons even if not for this one.

    Marion’s comments are sensible, and in particular, as she says, there is no plausible mechanism whereby diet sodas could raise stroke risk. The worries about artificial sweeteners based on anecdotes and small, uncontrolled, or otherwise weakly-designed pilot studies have never held up to scientific scrutiny; phosphoric acid may be a little bit bad for your bones, but (a) only if your calcium intake is also marginal, and (b) the mechanism doesn’t in any way relate to stroke risk; caffeine may raise stroke risk in a small minority of people, but studies consistently showw that coffee and other caffeinated beverages have no effect on, or *lower* , cardiovascular events in the population *as a whole* .

    I would add another clear weakness in the study: from the AHA website report, and from the researchers’ own institution’s website:

    … the study does NOT appear to have been controlled for the subjects’ BMI, raising the problem of reverse causation already raised by a couple of commenters: many overweight people drink diet soda, often as a kind of “halo” defense against the high Calorie diets they continue to consume, so if you don’t control for BMI, “drinks diet soda” often is a surrogate marker for “has an unhealthy diet and/or is overweight or obese.”

  • James Bond Bond

    I was searching for decisive information on this topic. The information you have provided in the blog is really important.
    stroke recovery time

  • Barbara amsel

    I believe this is actually the most useful blog I’ve been through this entire day.