The latest report on coffee and prostate cancer may mislead those who cannot tell the difference between causation and correlation. Beware of observational studies that propose “preventions” of any disorder.
Comment on “Coffee May Lower Risk Of Deadliest Prostate Cancer” on Morning Edition 18 May 2011. This study shows there is a link between people who drink coffee and not dying from prostate cancer. I discuss how news media reports have distorted the findings to suggest that drinking coffee may prevent dying from prostate cancer. I discuss possible controlled experiments that could be done to show the latter conclusion, as well as the kind of thinking that must be done before doing a study like this to make sure that the methods are based upon well-grounded physiological theory.
Updated: 2 Sept 2014
NPR reporter John Knox talked with Harvard researchers who took information from almost 50,000 male health professionals about their coffee-drinking habits and found that those who drank at least six cups of coffee a day formed a statistical group that developed prostate cancer at a level of 60% below that in a statistical group of men who drank less. This is a very clear restatement of what John Knox said. He, and the researcher he interviewed, used the term “risk”, which implies that to avoid the likelihood (“risk”) of developing prostate cancer, you could drink six cups of coffee per day, thus reducing the “risk” of prostate cancer. They cite a study done by Wilson et. al., (2011). The study did not find any difference in the occurrence of or presence of sub-lethal stages of prostate cancer.
My Comment Posted at NPR
Not only do reporters on medical studies get it wrong, so do many doctors. Nothing like this conclusion can be made. A correlation does not show cause and effect. No study has shown this, certainly not this one. A serious problem with this study is that it is “observational”. It just happened that people who drank 6 or more cups a day were in a group of people who did not get prostate cancer as often as people in the other group. Absolutely no study was done which was experimental. The researchers did not take a group of people who were found likely to develop prostate cancer (a subgroup of the above large group who had developed low-grade prostate cancer or pre-cancer), subjected them to a regimen of drinking at least 6 cups of coffee per day, and compared their outcomes years later with a matched group in age and health status who were not subjected to this regimen. It is just as likely that people who are not prone to getting prostate cancer will choose to drink that much coffee–in other words, the opposite conclusion can be drawn from this study, especially when you look at the nutritional status of coffee.
Interpretation of Observational Studies Is Wrong
All coffee studies have been “observational”. The conclusion in past studies that the antioxidants in coffee were of health benefit to the people studied was also incorrect, for another reason. All biochemists know that non-polyphenol antioxidants are degraded to non-antioxidant form when exposed to heat. It is difficult to brew a cup of coffee, hot or cold, without exposing the beans to heat. Freezing does the same thing, so if the coffee beans were frozen (“to keep them fresh”), the coffee brewed from them will not have antioxidants in them. Many people make claims that certain foods are good for you because they have antioxidants in them, but no one can eat these foods without cooking them because they may be poisonous in that state (e.g. rhubarb), or some of these foods are frozen first, e.g. ice cream or freezer pops made from fruit juice.
Researchers also claim that the “coffee effect” persisted even when the researchers corrected the data for the effects of exercise. Again, earlier studies show that, in a correct conclusion from observational studies, people who were in a statistical group that exercised as little as two or three brisk walks per week, died from prostate cancer at a rate 50% lower than those in a group who did not do this exercise. Nothing in that study suggests a conclusion that exercise reduces your likelihood of getting prostate cancer. Not that exercise is not good for you, but there was no experimental study controlling for all other factors and imposing an exercise regimen on people who were not doing it, and making no change in a control group matched in age and health status. Furthermore, statistical controls are not the same as experimental controls. Statistical tests cannot control for variables used to determine cause and effect, if no experimental test was done (Sloan et al., 2004).
Instead of drawing these overstated and incorrect conclusions, the researchers might use the results to ask questions about why they saw this difference and then construct controlled experiments. There are tons of questions we could ask. Does exercise actually reduce the chance of dying of prostate cancer even when we discover that the person has the cancer early on? Does removing the exercise regimen make the person worse off? (To those who object to a study testing this for ethical reasons, a sample can be drawn up of many people, and those who become bed-ridden or stop doing exercise can then be tallied as a separate, “experimental” group who were by chance put into this category). What about all the symptoms preceding the imposition of the exercise change–are there any that seem to be apparent in all patients that have prostate cancer? Calling it “symptomless” is not enough. Sometimes you have to be as observant as a doctor who practices Traditional Chinese Medicine (TCM).
Other questions involve nutrition. We choose to drink coffee for some reason and the reason does not have to stem from the energy we get from it. It is extremely high in calcium and boron. Our needs for calcium and boron vary during the day and with our health status. We are motivated to get calcium or boron in various forms that set up cravings. Coffee tends to be very available at most jobs, but not ice cream, which is also very high in calcium (if it is truly ice cream and not ice milk. See my comments at “On Healthy Ice Cream“). Apples are very high in boron but do not tend to be readily available at work. Some people need more calcium than others because they
- have a higher metabolism,
- great amounts of glandular activity,
- move around a lot more,
- excrete more toxins in their feces, or
- are doing a lot more tissue repair (often as a result of No. 3 or 4).
The causes for each of the above reasons might suggest to most people why everyone’s physiology is different (and many will say “genetic” when genes do not cause anything; they are only associated with conditions and are activated by environmental influences).
People’s physiology shouldn’t be that different. If one person has a higher need for calcium than another person, it may be due to a lot of reasons:
- Fat cells are not capable of metabolizing fat in the second person,
- The pancreas, thyroid, adrenal glands do not work as well in the second person due to damage
- Muscles do not work as well in the second person, probably also due to damage there or in the nervous system
- The first person is getting rid of toxins much more efficiently, possibly due to non-damaged connective tissue in the hypodermis, liver, or blood vascular system,
- The nervous system has been damaged in ways that prevent calcium delivery when needed in the second person.
All of the other diseases whose risk may be decreased by drinking coffee (as some suggest) may actually be more associated with calcium intake and uptake within the body. The problem that many scientists and doctors seem to miss is that nutrition studies cannot assume that the nutrient is being taken up by needy cells, once the person has eaten the substance. Even simple tests like measuring calcium intake and the amount excreted are not satisfactory, since calcium may never reach the cells that need it but, instead, only some cells, or it might be “stored” inappropriately in the form of stones, cysts or inside damaged cells that cannot use it or lose it.
Therefore, news reporters need to be very careful when they present advice to readers/listeners, and not attribute cause and effect when none is demonstrated. There is no evidence reported here for saying that drinking coffee reduces the risk of prostate cancer, unless you also say that having prostate cancer reduces the risk of drinking coffee.
Sloan, R. P., Carroll, D., & Greenland, S. (2004). Risk factors, confounding, and the illlusion of statistical control. Psychosomatic Medicine 66(6), 868-875. [Freely Available].
Wilson, K. M., Kasperzyk, J. L., Rider, J. R., Kenfield, S., van Dam, R. M., Stampfer, M. J. … Mucci, L. A. (2011). Coffee consumption and prostate cancer risk and progression in the health professionals follow-up study. Journal of the National Cancer Institute, 103(11), 876-884. [Freely Available].
Traditional Chinese Medicine: An Introduction. National Institutes of Health, National Center for Complementary and Alternative Medicine.
For other reports on nutrition, see Nutrition Reviews.
Keep up with new posts I make by subscribing to this blog: go to the top and click on “Subscribe” in the gray WordPress Choice Bar (if you are already registered in WordPress.com and have logged in) or when you comment on this blog, click on the “notify” check boxes.
© Copyright 2014 by Martha L. Hyde and https://marthalhyde.wordpress.com.