Study on Sucralose Not So Sweet


New Study is Very Bad News–For Splenda-Addicted Mice

I don’t care about sucralose, don’t actually use it myself and have never spent much time thinking about it. But I do care about bad science and bad reporting on science and this sucralose study and associated media reporting is hard to ignore.  It’s one of the worst examples of hype and ‘jumping on the bandwagon’  bad science reporting seen in recent times.

Here is what we can definitively say (the only thing we can definitively say) based on data from this study: Sucralose appears to be associated with an increased risk of cancer (particularly leukemia) in mice who ingested 100X – 3000X the recommended daily intake. That is all we know. This is, of course, very bad news for Splenda®-addicted mice, but unless that describes you, there probably is not much for you to worry about. Contrary to what many health reporters appear to understand, it require leaps of logic over many massive hurdles to apply this data to actual risk of cancer in humans who are ingesting normal amounts of sucralose. I’ll go out on a limb and suggest that if you are ingesting 100x – 3000x the recommended daily dose of Splenda®every day, leukemia is probably the least of your concerns.

Massive over and under-dosing are tricks used by advocacy researchers (people who do research with a political agenda, not to get a truly useful result).  By force-feeding toxic-level doses, they basically ensure the negative result they desire. This is marketing, not science, and it should not be entirely surprising given the scientists involved in the sucralose study. The Ramizzini Institute in Italy performed the study. This is a group whose scientific bona fides have been questioned in the past due to biased and sloppy studies*. Additionally, while they brag about not taking money from industry special interests, they have no such problem taking money from non-industry special interest groups, aligning themselves with ‘clean eating/clean living’ advocacy groups. (Special interest influence is not limited to industry and the Center for Science in the Public Interest (CSPI), who actually changed a policy recommendation based on this single study feeling it had more ‘weight’ because there were no industry special interests involved, would be well-advised to remember that).

In reality, giving 100x – 3000x the dose of any substance (even a ‘clean’ one like bananas, for instance), will likely produce a negative biological result. This will not tell you that bananas are dangerous or cause cancer. But it will tell you that the researchers involved are astonishingly bad at planning studies that will produce truly useful information.

Is sucralose dangerous to humans who are using it in normal amounts for its intended purpose? I don’t know–the bulk of prior studies indicate it is safe, and this particular study does nothing to actually answer that question. Which is exactly the problem with all the media hype. It legitimizes a very flawed study–both in methodology and in inherent bias due to sponsorship. This does not illuminate, educate or help consumers to make decisions in any practical way (and it may, in fact, contribute to growing concern over orthorexia nervosa, an eating disorder that causes sufferers to be excessively concerned with eating the ‘right’ or ‘clean’ foods). It also brings to mind a cautionary appeal about (mis)applying laboratory or mouse data to humans. From Dr. Vinay Prasad of the Cancer Network: 

If a drug only has mouse or cell culture or laboratory data, it does not deserve any press (emphasis added). We know for a fact that the majority of the most promising compounds at this stage will fail in human trials, and covering these drugs is akin to doing a nightly news segment on someone who bought a lottery ticket, asking them what they will do if they win.” 

So the chance that this data will actually prove to be relevant to humans is roughly the same as the chance of winning the lottery, yet mouse studies continue to be hyped by the press and gobbled up as relevant by consumers all the time. It is really time for that to stop.

Then there is the ‘common sense’ perspective which is entirely missing in the reporting on this study. If there is indeed a link between sucralose and leukemia in humans  we would expect a corresponding increase in the incidence of leukemia since the introduction of sucralose. Fortunately, this data is readily available from a number of credible sources, including the SEER database maintained by the National Cancer Institute. Here is the incidence of leukemia from 1992 to 2012. Splenda®hit the American consumer market in 1999. There has been no alarming bump in new cases since that time. In fact, the incidence curve is about as close to a flat line as it is possible to get and, in fact, deaths from leukemia have decreased since 1999. So if there is actually an association between normal use of sucralose and leukemia it appears to be so vanishingly small that it has virtually no impact on disease incidence.  This is quite a stark difference to what the headlines suggest and should give longtime users of sucralose some peace of mind.

And that is the real problem with bad science and bad reporting on science. Thanks to a hyped and distorted sucralose scare, we have a population that will worry needlessly that they are at risk for leukemia, when in fact, unless they are mouse-sized with a mouse metabolism and eating massive quantities of the stuff, it is unlikely to be true. What a different health reporting landscape we would have if only media outlets would use a tiny bit of critical thinking before going all in on alarming stories like this.


* Oh the tangled web: Ramazinni is notorious for badly designed–by accident or intent, (not sure it really matters much if quality studies are important to you) studies– see the fallout from their 2006 anti-aspartame study, which was roundly criticized by regulatory authorities around the world for using sick rodents who were at increased risk for developing cancer in the first place. Their Splenda study had been reportedly rejected by six publishers since they first hyped it via press release (no data) in 2012. So kudos International Journal of Occupational and Environmental Health! You just started a Wakefield-esque sucralose panic based on information other journals deemed unfit for publication.  As an aside, the single purchase price for this article is $130, making it virtually impossible for most people to actually assess the content.