A nurse and an economist walk into a coffee bar...

Does coffee really cause cancer? A nurse and an economist discuss the issue.
Written by Denise Amrich, Contributor

Hi everybody,

You may have noticed that I didn't blog last week. That's because I was moving my home office into my new house, including my trusty blogging computer. It was the last leg of a lengthy moving process which started in March, when my husband and I established our new home as our primary residence.

I couldn't have done it without coffee. It was my fuel.

I bring this up because of your thoughtful responses to my recent It’s official: your cell phone may cause cancer article about the correlation between cell phone use and brain cancer.

One of the letters I received was from an economist, Donald Robbins, PhD, who wanted to know if the mention I made in the article about coffee being a carcinogen was supportable by studies.

With his permission, I am sharing the interesting correspondence that transpired between us about the coffee/cancer connection (or lack thereof).

Donald's original letter to me

Thank you for your article.

I have a question, however: I have never seen evidence that coffee causes cancer. This concerns me greatly as a "serious" coffee drinker. I have tried to read all articles (BBC health, NY Times, and at one point searched on the Internet).

I would be very grateful if you could direct me to reports on the dangers of coffee drinking, particularly what you mention: that it causes cancer.

Thank you very much.

My first reply

I listed the coffee as carcinogenic because it was cited in so many of the reference articles I used to write my blog post, and I seemed to remember it being true (though I was saddened to see it on the list). I love coffee, too, and that's why I made the lame joke about how it seems safer than lead and exhaust fumes.

However, I am so glad you wrote, because you make a good point. The answer is more nuanced.

Caffeine seems to be relatively safe when used in moderation, and can actually be beneficial to many people (for example, a cup of coffee a day may help mild asthma symptoms because it can act as a bronchodilator).

When enjoyed in moderation, health benefits probably outweigh the risks for most people. Here's a Mayo Clinic article about it.

What I was taught in nursing school (and my current understanding of it from the research I did today in the Davis' Nursing Library) is that because of the acidic nature of coffee, people with stomach conditions such as gastritis, ulcers, etc., can have a problem with coffee in terms of cancer risk. That's thought to be because of the stomach and esophageal linings being repeatedly exposed to more and more irritation, and becoming more vulnerable to cancer.

High amounts of caffeine also are implicated in increased risk of bladder cancer because of damage to the bladder's lining.

Epithelial tissue is the lining tissue of the digestive tract (esophagus, intestines, stomach) and some other organs. It is more vulnerable to cancer than other tissue types because of its regular replacement by new cells. This increases the percentage of possibility by some degree that those new cells will be cancer cells.

Also, it's good to avoid caffeine if BPH (benign prostatic hyperplasia) is a problem. In terms of other ill effects on health, I also remember hearing that huge amounts are hard on the kidneys and adrenals. It can increase paranoia.

It can interfere with sleep (which is probably half the point of using it for most people). It can worsen iron deficiency anemia. It can increase blood pressure and increase coronary artery disease.

However, I just finished reading the Wikipedia article on the health effects of coffee.

The article (which seems to have been written by a coffee lover and doesn't have a neutral tone, in my opinion) mentions there are rat carcinogens in coffee but asserts that we can't just assume those are human carcinogens.

There are a whole pile of studies at the bottom of the article. If you want, use your browser to do a search on the page for the word "cancer" and check them out.

After a quick perusal, it looks like some of them are inconclusive, some of them seem to contradict my nursing resources and may indicate caffeine can actually have positive effects against cancer, and some of them seem to indicate there may be a risk.

I'd love to hear your conclusions after digging through these studies. Please keep in touch and let me know what you think about the verdict for our beloved java.

Also, I'd love to know, what's your favorite brewing method? What's your opinion of the popular Keurig brewers?

Donald's follow-up

I greatly appreciate your detailed and thoughtful response regarding coffee's possible carcinogenic effects.

I will peruse several of the references you've mentioned and share my personal evaluation with you (I'm a PhD economist, so can offer a critical reading, but cannot offer any medical or biological expertise).

My quick understanding from glancing at what you've written (paraphrasing you) is: there are two potential channels, the first indirect and the second direct.

First, via coffee's effect on the linings of the digestive system and intestinal-cancers (there seems to be a heightened emphasis in the medical community on these cancers and hence the recommendation of a glass of wine per day for its circulatory-system effects seem to have fallen into disrepute).

The second, direct, channel seems to be via the carcinogens contained in coffee, where the issue of dosage is probably determinant: whether the risk levels and typical consumption levels are of the similar orders of magnitude.

Certainly this is a potentially major health issue, given widespread (and yes, for techies, notably) coffee consumption. I've been surprised and relieved in the past to see no serious health risks with coffee. But if that is incorrect it means millions ought to change their habits (fortunately not that hard), as happened with cigarettes.

What I have read recently is that coffee use causes people to awaken with more grogginess than for non-coffee drinkers, and they need the coffee dosage in the morning to attain the normal alertness levels of people who don't consume coffee. So people develop a modest behavioral-biological dependency on coffee without any net increase in alertness.

A comment on medical research and recommendations: as an economist, we are carefully trained in statistical inference and its pitfalls. Great emphasis is placed upon how small or problematic samples often lead to incorrect inferences, or conclusions where none may be coherently drawn(statistically insignificant results), the conclusions are simply medical research and recommendations often seem silly.

My sense is that the radical and frequent changes in medical research "conclusions" result from this problem. When one looks at the study details, the sample sizes and controls are often clearly inadequate to be more than suggestive.

Medical researchers are often very statistically sophisticated and fully understand the problems, but the costs and inherent difficulties in design (because of moral issues with human-subject experiments) prevent larger, better studies.

Nonetheless, the result often seems that the medical community presents "conclusions" and recommendations, which should be no more than suggestions.

Some reversals: (20-30 years ago) cardio is good including running (to no, it causes bone-joint damage); moderate alcohol consumption is good; prostate cancer interventions now often seem as counterproductive because the intervention can be worse than the cure (many cancers treated would be mortal only well after death would occur); increasing "good" cholesterol is good (recently published research questions this); etc.

This leaves medical professionals and lay persons with an enormous, real, if uncomfortable, margin of doubt. And great difficulty for drawing practical conclusions. Given so much doubt, individuals might do best by taking the conservative path: avoid behavior, consumption, that might cause problems, until more is known.

But then there are cases like the recommending moderate wine consumption, which could actually (as seen today) be harmful. Or where one avoids consumption of something that is later shown to be good for us.

I try to glean what conclusions of the moment seem to be more robust. Sometimes this is made easier by the way in which the scientific evidence is presented (in scientific and journalistic publications), so that one needed go to the original studies to try to figure it all out.

Given the small sample sizes and inherent design limitations in medical research, however, this situation is not likely to improve, at least in the foreseeable future. Evaluating the evidence and recommendations is necessary and complex and often cannot provide the kind of clarity and certainty we would like.

Economics confronts enormous problems as well in reaching inferences. It shares some of the experimental design problems of medicine, because of moral problems with human subjects, and because most data is not from experiments at all; econometrics is largely an attempt to similar experiments with non-experimental data.

Astronomy also depends mostly on non-experimental data. And in economics, there are additional epistemological questions about the clarity of the concepts used by economists, and the continuing evolution of object of observation, economic systems.

Sorry for the methodological digression, but my sense is that these issues are central and not always appreciated and certainly not easy to overcome.

Thanks again, and best regards, Donald

My wrap-up

Until either Donald or I have time to wade through piles of studies (right now there is not enough coffee in the world to make me willing to contemplate that), I have no immediate plans to cut my coffee consumption. I am especially loving the Keurig Platinum B70 coffee brewer enshrined in my tiny new kitchen.

What about you? How much coffee are you drinking and how are you brewing yours? Do tell in the TalkBacks below.

Editorial standards