This article was published on 29 July 2012, and has had over 70,000 views. This is the number one article I’ve ever written, I enjoyed writing it, but I never thought it would be such a big hit. It basically arose from a meme I saw on Facebook that claimed that bananas with dark spots had anti-cancer compounds in it. And it was all based on a misreading of a published article, a lack of knowledge about tumor necrosis factor, and a complete misunderstanding of human physiology and immunology. And this is my number 1 favorite, and number 1 most popular article for 2013.
Last year, I wrote an article about how to critically analyze pseudoscience and misinformation to get at the scientific evidence which may help you accept or reject something you might read on the internet, even if it appeared to be accurate. On Facebook, Twitter and many internet sites (including Wikipedia), there is an amazing tendency of individuals to accept what is written as “the truth” without spending the effort to determine if what is written is based on accurate science. Twitter, of course, limits itself to 140 characters, which means you either have to click on a link to get more information, or just accept that the 140 characters are factual. And if you can make a complex scientific argument in 140 characters, I’m impressed.
Facebook is filled with false memes on just about everything from politics to medicine. The anti-vaccination crowd fills Facebook with their amusing and highly inaccurate memes. For more than a year, there have been dozens of photos of bananas with a few words that some Japanese scientists claim that ripe bananas have high levels of “tumor necrosis factor“, so eat bananas to cure cancer and maintain a healthy immune system. Facebook is famous for these things, little pictures with a few words, no sources of the information, and broad conclusions. Eat bananas. Cure cancer. And people share them with a click of the button and move on to the next cute cat picture. It’s really the lazy person’s way of learning. Although who doesn’t enjoy the cute cat pictures?
The mysterious Japanese Study
The actual study was published here (pdf) (or here, also pdf) in Food Science and Technology Research in 2009. I have several issues with the article with respect to the banana meme, especially in trying to make any conclusion that bananas have anything to do with “curing cancer” in humans:
- The authors do not make one single claim (as best as I can tell) that there is tumor necrosis factor in a banana. No, this article does not say anything about bananas having TNF. Really, nothing else I write about this article really matters, because they just don’t make any claim whatsoever about TNF and bananas.
- The study is in a mouse model, and many animal model experiments just don’t transfer to human clinical research. It’s not that it is impossible, but until it does, it’s difficult to make an appropriate prediction from such an experiment. So, even if the authors claimed that bananas had TNF (and that would be Nobel Prize winning research), we have no clue as to whether it has any clinical impact. But let’s not forget that the authors make not a single claim that TNF is in bananas.
- The article is published in a low impact journal. This journal has an impact factor of less than 1.0, which indicates a very low quality journal and it’s not even indexed in PubMed, which further indicates its low value.
- The bananas are not fed to the mice, the are injected into the peritoneal cavity. I am not sure why the authors did that, and I am unclear what it is supposed to prove. That it induces TNF activity might be expected since the body would react to any foreign substance (an apple, viruses, bacteria, anything) injected into the peritoneum, and production of TNF might be expected. But the TNF does not flow out of the banana, it is just the immune system’s reaction to a banana injected into the body.
- What is the clinical significance of what was induced? In other words, is there actually TNF isolated from the bananas in a form that actually can do something? Or is there some other effect just because there’s a blob of intraperitoneal banana extract which induces some other immune response.
- Bananas produce small amounts of serotonin (5-hydroxy tryptamine) and dopamine, depending on their stage of ripeness. These can have a stimulatory effect on neutrophils and macrophages in a living organism, and these can in turn produce the touted TNF-a, Interleukin-12 (IL-12) and other cytokines. In this role serotonin and dopamine are said to serve as a “biological response modifier” (BRM). In plain English, the stuff in ripe bananas can (but not necessarily will) stimulate a subset of white blood cells to produce chemical signals to deal with a variety of threats. However, this is a small effect, and it will not help you fight off cancer.
- No one has repeated this experiment, and certainly no one has shown this effect in humans. There is not a single double-blind clinical trial to show that bananas induce a TNF-alpha response in humans. And one more point to remember: bananas do not produce TNF. It is simply an evolutionary impossibility, unless through some amazing instance of evolutionary convergence, the banana plant evolved the ability to produce the TNF molecule for a completely different purpose for the banana plant. This would violate several principles of evolution, since there are reasons why the TNF molecule evolved in mammals and not in plants.
Let’s review–I dug up the original “Japanese scientific paper”, and what I found was essentially simple–bananas do not produce TNF, but that wasn’t even the point of the article. Whether the article has some future relevance into the study of the immune system is unknown, but through a completely ignorant reading of the article, people keep pushing the point that bananas cure “cancer.”
So let’s look at this pseudoscience even more carefully:
- The “Japanese scientists” made no claim that there’s TNF in a banana, but the junk medicine pushers continue to make the claim, facts be damned. The problem is that anyone with a basic comprehension of biochemistry would understand that tumor necrosis factor (TNF) is a fairly complex protein, with a highly specific role in the human immune system (something notably lacking in a plant), so the chances that a banana would have some substance that exactly mimics or copies TNF is so tiny as to be close to impossible. A banana has no need for TNF, since it lacks an immune system of a vertebrate, so evolving a complex protein like TNF would be crazy; in fact, if it did, we’d have to rewrite our understanding of evolution. Let’s make this clear–we don’t have to rewrite evolution, because there is no TNF in bananas.
- Even if we could assume that a banana makes TNF, the digestive tract would break down the complex proteins and substances, such as TNF, into its constituent components, such as amino acids, simple sugars, and fats, before being absorbed into the bloodstream. The TNF probably wouldn’t survive intact within the digestive tract. The true scientific skeptic would, even if they thought that maybe a banana evolved a TNF molecule by some strange mechanism, know that it could not enter the body.
- Since TNF has a very specific action on the immune response, and not directly on cancer cells, how do we eat sufficient bananas (even if it did have TNF and it could be absorbed into the bloodstream) to increase the level locally to actually cause the appropriate immune response? And which one of the 200 or so cancers would it effect? It is clear that the name, tumor necrosis factor, leads one to believe that a drop of it on a cancer immediately kills the cancer. I would have to write a 20 page paper just to describe how TNF is regulated and disregulated within the immune system both locally and generally in response to a wide variety of immune challenges, including cancers. It is incredibly complex, and the name is simply one given without consideration to future alternative medicine pushers who jump on it as the the “cure” to all cancers. It isn’t.
- Which leads us to how TNF causes many of the clinical problems associated with autoimmune disorders such as rheumatoid arthritis, ankylosing spondylitis, inflammatory bowel disease, psoriasis, hidradenitis suppurativa and refractory asthma. So, if a banana had TNF, if it could pass into the bloodstream from the digestive system, and if it could reach high enough levels to actually do something, it probably wouldn’t have any effect on any of the cancers, but it could have serious side effects. However, I wouldn’t worry about it, since there are so many “if’s” that it’s just not possible.
- And this “boost the immune system” myth? The immune system is an incredibly complex system, with an almost infinite number of interactions between various proteins, compounds, organs, factors, and cells. As long as you’re healthy, so is your immune system, there is nothing you can do to make it stronger, better, or “boostier.” Well, vaccinations help, but they rely upon an appropriate immune response. So even if you won the Nobel Prize by finding some miracle compound that “boosts the immune system,” it would work on one tiny part of that system, and it would have a zero net effect.
So here we go again. Someone reads that some scientists discovered bananas cure cancer. They didn’t dig up the actual paper to find out that the scientists didn’t actually say that it did. They didn’t think through the problem that bananas wouldn’t actually contain tumor necrosis factor or that the digestive tract wouldn’t actually absorb it. Or that you couldn’t possibly eat enough bananas to get enough TNF to do any good. Or that if you could that it would have some very bad effects.
If you want to eat a banana, go ahead. It does have some benefits I’m sure. But it isn’t going to prevent cancer. It’s not going to cure cancer. It’s not going to do make you live longer. It’s just going to provide nutrition.
- Iwasawa, Haruyo, and Masatoshi Yamazaki. “Differences in biological response modifier-like activities according to the strain and maturity of bananas.” Food science and technology research 15.3 (2009): 275-282. doi:10.3136/fstr.15.275