Coffee and Prostate Cancer — Evaluating a Health Study
Driving to work the other day I heard a brief report on National Public Radio (NPR) about coffee and prostate cancer. It said that men who drink large amounts of coffee had a 60% lower risk of developing “lethal or aggressive” prostate cancer. 60%! It was enough to make me want to pull into a Starbuck’s and order a triple venti- extra humongous coffee-type drink RIGHT THEN. And I don’t even have a prostate!
Seriously though, I also thought about my Dad. He drank coffee all day long, at least a pot a day for most of his life. He also died from prostate cancer. By the time he was diagnosed the cancer has spread beyond the prostate (Stage IV). He died 15 months after diagnosis, from a disease that — according to the American Cancer Society — has a nearly 100% 5-year survival rate.
So apparently Dad and his prostate did not get the memo. . .
Everyone knows an exception to the rule when it comes to health claims. The cigar smoker who lives to 100, the woman who never smoked yet dies of lung cancer. How can health claims have any merit if there are all these people who do everything right (or wrong) and die at a seemingly inappropriate age anyway?
It makes a person want to throw up their hands and just do whatever the heck they want.
I realize its hard to make life decisions with this kind of information. I assure you, the scientists making these claims are not some smoky backroom cabal plotting to make your life difficult. Here are a few things you need to know about science before we go on:
- Scientific progress is incremental and definitive conclusions take years to reach. One study does not make something true. But many studies make for a stronger conclusion.
- Some scientific conclusions are stronger than others. How a study is done can effect the strength of the conclusions we draw from it. I would like to expand on this in a separate post, because it deserves careful treatment.
- The media does not tell you any of these things. Often they don’t even tell you the facts you need to decide these things for yourself. I’m pretty much convinced that the only criterion most media outlets have for science stories is whether it will make a good headline.
So keep these things in mind as we talk about the coffee/prostate study.
To be fair to NPR, later in the day they broadcast an in-depth interview with Dr. Kathryn Wilson of the Harvard School of Public Health and Harvard Medical School. Dr. Wilson is the post-doctoral researcher who actually did the study. In the interest of bringing you the raw data connected to the scientific research discussed here, I looked for a publication on this study in the National Institutes of Health PubMed database, but did not find a publication on this data. I did find this press release from the American Association for Cancer Research (AACR). The press release links to this program for the December 2009 Frontiers in Cancer Prevention Research meeting. My guess is that this work is close to being complete and was presented at the meeting, but has not been written up for publication yet. It’s standard practice to share pre-publication data at meetings. It’s a way to share information more quickly — publications can take up to a year to get through the peer review and editing process — and to get feedback on the best way to present the work in the official writeup. It’s also common for universities and organizations like the AACR to issue press releases about science presented at conferences.
All the other newspaper and radio articles I found about this study link back to the AACR press release as their primary source. I will have to consider the press release and the interview with the researcher who did the work as the primary sources for now, and I will bring you the publication and talk about it here when it becomes available.
Let’s look at the information we have about the coffee study in terms of the criteria I outlined above.
- Scientific progress is incremental and definitive conclusions take years to reach. One study does not make something true. But many studies make for a strong conclusion.
Notice that during the interview Dr. Wilson stresses over and over again that this is just one study. She says, correctly, that her group needs to do more studies and get information from more people to try to see if their conclusion holds up. It is possible that there is something else common to the 10% of men in the study who developed prostate cancer that made their cancers less likely to be lethal or aggressive. The collaboration Dr. Wilson describes in the interview will bring more data into the study, allow for more detailed questions, and begin to firm up the conclusions from this study.
- Some scientific conclusions are stronger than others. How a study is done can affect the strength of the conclusions we draw from it.
This study appears well done, but we are limited in the facts that we have. The researchers followed a large number of men (50,000) over a substantial period (20 years). They also collected data regularly during the course of the study (every four years, so 5-6 times over the course of the study) instead of just once at the beginning and once at the end. The men were asked to self-report their coffee drinking, which many criticize as being an unreliable way of collecting data. The men could lie about their coffee drinking habits, or simply not remember how much they drink. However, coffee drinking is not something a person is likely to lie about, and if the men were sampled at four-year intervals it would be easy to see if their answers were consistent from survey to survey. And again, this is another reason to do the study again with a different group of men.
Over the course of the study, 4,975 men developed prostate cancer. We do not know from our sources how Dr. Wilson and her colleagues defined “aggressive and lethal,” but I think it’s safe to say that we can define it as a prostate cancer that spreads beyond the prostate or cancer that causes death. Now, remember what I said above about the 5 year survival rate of prostate cancer being almost 100%. The 10-year survival rate is 91%. So prostate cancer is not likely to be a deadly or aggressive cancer anyway, even in men of the age group studied (mid-50’s to mid-70’s).
However, when the men who got prostate cancer are divided into groups based on their coffee drinking habits, the group that did not drink any coffee had the highest number of cases where the cancer left the prostate and/or caused the person’s death. The group who drank 1 or 2 6 oz. cups of coffee had the next highest number of cases of lethal or aggressive cancer, and so on down to the group who drank the most coffee (6 or more 6 oz. cups per day). This group had the lowest number of cases of lethal or aggressive prostate cancer. This is what is known as a correlation, where two or more data sets vary together. Correlations can be strong, meaning the data points always vary together and to the same degree, or they can be weak. There are many ways to figure out the strength of a correlation. We have not been provided with data on the strength of the correlation between amount of coffee consumed and the number of cases of aggressive or lethal prostate cancer. Harvard is not an institution known for sloppy science, I’m sure they have these numbers. Unfortunately, we will have to wait for the publication to see them. I think this says more about how the media than it does about science. But I digress.
The important thing to know is this: correlation is not the same as causation. As Dr. Wilson points out in her interview, we do not know whether the coffee is actually having an effect on the cancer. Certain compounds in coffee could have a direct effect on cancer cells. This should be a topic for future cell biology research. It could be something the coffee is doing in another part of the body (the hormone and insulin effects talked about in the interview) that then feeds into cancer biology. Or it could be something that goes along with both coffee drinking and lower risk of aggressive or lethal prostate cancer. Maybe the men who drink more coffee also spend more time relaxing in coffee shops, and relaxation is key to preventing lethal or aggressive prostate cancer. Perhaps more data was collected from the men and we could find out this answer. This study alone is not reason to think that coffee is a medication for prostate cancer. Dr. Wilson says this is a premature conclusion to draw and she is right. However, this study is interesting and important because it has pointed out avenues for future study.
- The media does not tell you any of these things. Often they don’t even tell you the facts you need to decide these things for yourself.
As I said above, unless you were present at the meeting, there is not yet a way to see the raw data that backs up these conclusions. The AACR issued the press release before the data was available to the public as a peer-reviewed publication. Would it matter to most people/news networks if the work was available, or would they still go by the press release and not look for the paper? The media’s presentation of science is just as important as the science itself. It’s important to hold science news to a high standard if we want quality science information.
So what can we take away from this story? A study done at the Harvard School of Public Health appears to show a (good??) correlation between increased coffee consumption and lower risk of aggressive or lethal prostate cancer in men in their mid-50’s to mid-70’s. This does not mean that drinking large amounts of coffee will keep you from dying of prostate cancer. It does mean that coffee could have effects on our bodies that are beneficial, particularly with prostate cancer, and scientists need to do more work to figure out exactly what those benefits are and what causes them.