Evidence of Cherry Picking Data on Sweeteners

Still Life with a Bowl of CherriesIs it still true that seeing is believing? Or in this age of truth decay, is it more likely that people see what they’ve already chosen to believe? We see a lot of this in politics and religion. But it also seems to creep into nutrition research. A case in point would be highly polarized views on low-energy sweeteners. In a new citation network analysis, Mie Normand and colleagues find evidence of cherry picking data on sweeteners and body weight. Even in systematic reviews, researchers tend to cite studies that favor their conclusions. Data that supports a different view drops out of sight.

Normand et al use a more delicate term for this pattern – citation bias.

Polarized Beliefs

Beliefs about low-energy sweeteners and body weight vary dramatically. One set of believers holds a conviction that they are good. They can satisfy a desire for something sweet, without adding a lot of calories to a person’s diet. Representing this view, Peter Rogers writes that intervention studies comparing low-energy sweeteners to sugar are definitive. Body weight goes down when swapping these sweeteners for sugar. None of the evidence to the contrary stands up to close examination. Case closed.

On the other side of this polarized debate, scientists rely on observational data to prove that low-energy sweeteners have “real risks.” People who consume them tend to be heavier and have more health problems. Plus, controlled studies show that the sweeteners are not metabolically inert. They dismiss controlled studies of the effects on body weight as having a duration too short to predict long-term outcomes. Water is better than diet drinks. End of story.

Citation Bias

Normand et al found this pattern – even in systematic reviews:

“Recent systematic reviews have differed markedly in, for example, arbitrary restrictions on the types of LES exposures included, duration of RCTs or follow-up of prospective cohort studies, differences in comparators (water or sugar, for example) and so on, all leading to conclusions being based on quite different evidence sets.”

Thus, on both sides of this divide, scientists find ways to dismiss evidence that does not conform to their beliefs. No doubt, much of this cherry picking of data on sweeteners is not overt or intentional. So it is fair to more gently label it as “citation bias.” But we do tire of one side beating correlations to death while the other side dismisses experimental evidence that sweeteners can have metabolic effects.

Scientific rigor demands both objectivity and curiosity. This is sadly lacking on the subject of low-energy sweeteners.

Click here for the study by Normand et al.

Still Life with a Bowl of Cherries, painting by Aleksandra Ekster / WikiArt

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


May 29, 2021

One Response to “Evidence of Cherry Picking Data on Sweeteners”

  1. May 29, 2021 at 11:33 am, Michael Jones said:

    In my obesity medicine practice I try to encourage little to no LES. This is not because I’ve been convinced by evidence either way, as it’s mostly unconvincing. However, in my experience people who regularly consume “keto” ice cream, diet sodas, LES energy bars, etc., in the context of what appears to be a relatively clean diet, consistently do not progress as well. Also, consistently, when I am able to convince them to “fast” from these types of things they seems to progress better. This admittedly does not prove causation, but in the exam room, on an individual patient level, I don’t really care what we can blame our successes on. That said, I would like to see some “cleaner” data emerge on this topic.