You probably have heard of a new study that links sucralose (brand name: Splenda) to increased chance of leukemia in mice. My facebook feed is a buzz with people who “knew” that stuff is poison.
But, the question is, is it really?
I’m going to link back to a similar report that I did on GMO corn causing cancer in mice. A research paper that was subsequently retracted, then republished (with no changes) in a different journal. This particular study is very similar in that, I don’t buy it. I don’t know if the papers are just getting it wrong, but the data really doesn’t support the claims made about it. There’s something even more interesting, but I’ll get to that in a second.
The paper I refer to is here (PDF). Sucralose administered in feed, beginning prenatally through lifespan, induces hematopoietic neoplasias in male swiss mice by M. Soffritti.
I need to mention some things about this particular author. Not unlike the GMO study I looked at, the author of this paper is not exactly held up as a paragon of scientific integrity. There has been a fair amount of controversy surrounding this author.
First, in 2007, this author produced another paper. That paper found that cancer rates also increased, specifically, leukemia, just like this study. However, that study was about aspartame… another artificial sweetener. The irregularities around that study were large including
- Using Sprague-Dawley rats, which, are engineered to get cancers at the end of their two-year life span. I mentioned that this is actually a critical flaw in the GMO paper as well.
- comparing cancer rates of older aspartame-consuming rats to younger control rats
- unspecified composition of the “Corticella” diet and method of adding aspartame, leading to possible nutritional deficiencies
- unspecified aspartame storage conditions
- lack of animal randomization
- overcrowding and a high incidence of possibly carcinogenic infections
Further, it went against all the other relevant research on the topic. With the problems mentioned and the conclusions at odds with properly conducted research, the Soffritti paper on aspartame was ignored.
Now, Soffritti has taken on sucralose, the artificial sweetener in Splenda. Again, a increased chance of leukemia is reported as compared to the control animals. But there are some oddities with this study too.
First, the report the amount of sucralose added to the mouse food in parts per million. That’s kind of a strange way of doing it. I’ve read a fair number of these reports and this (and the other Soffritti paper) are the only ones that use ppm. That’s not a flaw, just odd. It make it difficult to compare values accurately.
Here’s the conclusion from the paper
Results: We found a significant dose-related increased incidence of males bearing malignant tumors (p < 0.05) and a significant dose-related increased incidence (p < 0.01) of hematopoietic neoplasias in males, in particular at the dose levels of 2,000 ppm (p < 0.01) and 16,000 ppm (p < 0.01).
Conclusions: These findings do not support previous data that sucralose is biologically inert. More studies are necessary to show the safety of sucralose, including new and more adequate carcinogenic bioassay on rats. Considering that millions of people are likely exposed, follow-up studies are urgent.
It’s kind of odd that the minimum dosage level of 2,000 ppm and the maximum dosage lvel of 16,000 ppm both have in increased rate of tumors, while the dosages of 500 ppm and 8,000 ppm don’t.
Why is it odd? If you have a gradually increasing amount of a substance and that substance is the direct cause (not that direct cause is important here) of a condition. Then you would expect to see, as the amount of substance increases, then the rate of that condition also increases. But that’s not the case here. The 8,000 ppm group doesn’t have the same rate of tumors as the other groups. It should have more than the 2,000 ppm group and less than the 16,000 ppm group.
But it gets words and this is the really interesting part. The authors own data doesn’t support the claim. They said “We found a significant dose-related increase incidence of males bearing malignant tumors”
But when you go to their data (page 7), you really don’t see that. There’s a lot of data here, but I’ll try to summarize.
For male mice, with benign tumors, the rates (as a percentage of animals in that group) is as follows
- CONTROL 35.9%
- 500 ppm 36.8%
- 2000 ppm 35%
- 8000 ppm 37.9%
- 16000 ppm 35.7%
For male mice with malignant tumors
- CONTROL 56.4%
- 500 ppm 58.8%
- 2000 ppm 58.8%
- 8000 ppm 53%
- 16000 ppm 62.9%
So, that’s a slight increase of almost 7%.
For all mice (male and female) with malignant tumors
- CONTROL 61.6%
- 500 ppm 61.6%
- 2000 ppm 60.7%
- 8000 ppm 54.2%
- 16000 ppm 61.2%
The rate actually decreases and we get more sucralose until the 16000 ppm group, but it’s still not higher than the control group. That’s a little weird. No, it’s a lot weird.
Lymphomas actually decrease in all groups. Sarcomas actually decrease for all female groups and increase only in the male groups. Leukemias are the only tumors that increase in the higher concentration groups. However, the 500 ppm and 2000 ppm groups show a decrease in leukemias.
So, it seems that while eating too much sucralose could increase the chance of leukemia, a smaller amount would decrease the chances. The authors don’t mention that…
It gets worse.
For male mice: all skin carcinomas decreased in all test categories as compared to the control. Basically, the control mice had more skin tumors than any of the test groups. All ear carcinomas decreased in all groups compared to the control. All lung carcinomas decreased though adenomas increased in the 500ppm and 8000ppm groups, they decreased in the other groups.
All liver adenomas, carcinomas, and fibroangiomas decreased in all test groups compared to the control groups. The testes had fewer adenomas in all test groups compared to controls. Thyroid and parathyroid adenomas also decreased in all test groups.
For females, the results were largely the same. All skin tumors, mammary tumors (except fibrosarcoma), lung carcinoma, and liver sarcomas all decreased in the test animals over the controls.
Most of the other cases had such few results that I don’t think it’s reasonable to make a case one way or another.
Of all the tests in females, 24 different kinds of tumors had 3 or fewer instances across all groups (including the controls). In the males, it was 26 different kinds of tumors.
Even further, the actual counts of cases are all over the place. Just as one example, the lung adenoma for male mice had the following counts.
- CONTROL 6.8%
- 500 ppm 16.7%
- 2000 ppm 5%
- 8000 ppm 19.7%
- 16000 ppm 10%
That’s a strong hint that this substance has no effect on the condition. Again, if the condition was directly caused by the substance, as the amount of substance increased, the percent of the population in that test group with the condition should also increase. Not bounce around in an almost random fashion.
This is a clear case of something that has no effect on the condition.
Now, back to the leukemia.
Male mice Leukemias
- CONTROL 1.7%
- 500 ppm 5.3%
- 2000 ppm 17.5%
- 8000 ppm 10.6%
- 16000 ppm 15.7%
Again, we see the same thing. The numbers are bouncing all over the place. You can’t just look at the control and at the highest dose group and say… “ZOMG, it’s a huge increase in the condition”.
It is, but what is the cause of that condition?
If it was the substance we’re testing, we SHOULD see a slight increase in each dosage, but we don’t. The highest dose is still less than the second lowest dose. The second highest dose is the second lowest percent of cases.
In other words, these numbers are meaningless.
Does this prove that sucralose is safe? No, it doesn’t.
Does this prove that sucralose is toxic? No. it doesn’t. Not at all.
In fact, this test doesn’t show anything at all. While it’s an important study to do. The results are not what the authors (and all the health and nutrition mags) claim. Sucralose, more than likely, has no effect on any cancers or tumors at all.