Coffee protects from oral cancer?
Er, no, not really. It has not been shown conclusively yet. Research findings are currently conflicting. Nevertheless, that was the reassuring headline in many “newspapers”, acting fast to be the bearers of such a happy and relieving message: drink more coffee -it’s good for you! It’s approved by science!


I took my cue from a colleague who mentioned that researchers have found that coffee significantly reduces the risk of oral cancer. After reading Ben Goldacre’s blog for several years now, his “Bad Science” book last year, and witnessing myself first hand the extremely bad misrepresentation of scientific findings in the media, I have become much more alert and cautious in accepting the conclusions of the media articles. Instead, I much prefer to go to the source -the original scientific article, and find out on my own. Which is what I did this time as well. And no, there can absolutely be no claims that “coffee protects from oral cancer” yet.

The study comes from Japan [1], and is a prospective cohort study where the researchers followed a large number of people (around 40,000) for about 14 years. You can find the study here, and if you have an Athens account you can view the whole article.

The authors found that drinking one or more cups of coffee a day halves the risk of developing oral, pharyngeal, or esophageal cancers. This sounds like a very impressive decrease. Well, lets see. The absolute number of people that developed this kind of cancer was 157 (out of around 40,000), which means this is a rare kind of cancer. Therefore, a relative risk reduction will sound much more impressive than the absolute risk reduction: instead of about 4 in 1000 it will go down to about 2 in 1000 (a reduction of 2 in 1000 cases). Is that enough to start recommending more coffee?

The answer is a resounding no, and for a variety of reasons.
  • First, this was an observational study where many other factors may be affecting the results. And despite the authors trying to handle such confounding variables, it is very difficult to be sure that all environmental factors have been eliminated. Better controlled studies, as well as molecular studies on the identity of the coffee compound that might have such an effect, would help shine some more light into this issue.

  • Second, as the authors themselves acknowledge, results from other studies have been very inconsistent:
    Although many studies have assessed the relation between coffee consumption and oral, pharyngeal, or esophageal cancers, their results have been inconsistent. Of 12 case-control studies, 4 supported an inverse association (9, 13, 18, 19), 2 showed an increased risk (especially for hot coffee) (16, 17), and the other 6 suggested no association (8, 10–12, 14, 15). To our knowledge, only 2 prospective studies have investigated this issue. One, conducted in Norway, found no association (20). In that study, among 16,555 subjects during 11.5 years of follow-up, only 53 cases of cancer were ascertained, whereas we documented 157 cases of oral, pharyngeal, or esophageal cancer. The other prospective study reported an inverse association between coffee consumption and the risk of cancer of the buccal cavity and pharynx. Compared with those for the lowest coffee consumption category, the relative risks for the highest category were 0.5 for men and 0.7 for women (21). This result was consistent with that of our study, although the risk of esophageal cancer was not investigated.
    [numbers in brackets are original reference numbers in the paper]

  • Third, as again the authors discuss, they only collected information on coffee consumption at the beginning! So in the remaining 14 years they have no idea if and how the coffee drinking habits of the study population might have changed! To me, this seems like a significant limitation.

  • Finally, this is a study about Japan, so we cannot extrapolate to the rest of the world due to significant dietary, environmental, and cultural differences that can easily affect the results. And the fact that the method the coffee was consumed was not investigated by the authors, makes any such generalizations even more risky!
To conclude, while it seems that some component in the coffee might play a role in oral cancer prevention, one cannot jump to the conclusion that more coffee should be part of our dietary habits. As the authors mention:
cessation of alcohol consumption and cigarette smoking is currently the best known way to help reduce the risk of developing these cancers
and I could not agree more.

  1. Naganuma T, Kuriyama S, Kakizaki M, et al., "Coffee Consumption and the Risk of Oral, Pharyngeal, and Esophageal Cancers in Japan," American journal of Epidemiology 2008, vol. 168, pp. 1425-1432 []

One Response to this post
1. sciencebitches
So now we have to go to Japan to drink our coffee? :-P
Add a comment
Leave a comment
(Email required for authentication purposes only and will not be published. Please also note that if your comment contains links, profanity, or other sensitive content it might be held for moderation. In that case it will not appear until accepted by the administrator. Please do not post the same comment twice.)
Name *:

Email (not published) *:


Comment: (you may use HTML tags for styling)