Editors’ Note: Each day, there are 2.2 billion cups of coffee consumed worldwide. In the U.S. alone, residents average 2.7 cups of coffee per day. As a result, there are frequently studies reported about the potential benefits or harm caused by the beverage, which is why we’re presenting the following blog post which first appeared on Forbes.com.
Since 1991 coffee has been classified as a “possible carcinogen” by the International Agency for Research on Cancer (IARC), an arm of the World Health Organization. This assessment was based on early case-control studies, which showed some evidence of an association between coffee drinking and bladder cancer but not with other cancers.
With a great deal of fanfare, IARC recently updated its assessment of coffee to take account of the large number of new studies that have come out over the past 25 years. (The new report also covers tea and mate). In light of a number of controversies sparked by IARCs recent assessments of several environmental or dietary exposures (e.g., glyphosate and processed and red meat), the agency’s reassessment of coffee is of particular interest. But before examining it, it’s important to understand the agency’s schema for classifying the agents it evaluates.
Over the past forty years IARC has reviewed the scientific evidence on roughly one thousand agents and classified them according to the scheme in the table below. The number of agents in each group is shown in parenthesis.
- Group 1: Carcinogenic to humans (118)
- Group 2A: Probably carcinogenic to humans (80)
- Group 2B: Possibly carcinogenic to humans (289)
- Group 3: Unclassifiable as to carcinogenicity in humans (502)
- Group 4: Probably not carcinogenic to humans (1)
Group 1, known human carcinogens, includes agents for which there is strong evidence of human carcinogenicity, such as tobacco, arsenic, ionizing radiation, asbestos, etc. We know from a huge body of evidence that smoking causes lung and other cancers. We know about the effects of ionizing radiation from the studies of people exposed to the atomic bomb dropped on Hiroshima, as well as from studies of radiation used for medical treatment. We know about asbestos from studies of workers in shipyards and other industries who were exposed to asbestos. So, in Group 1 we are on solid ground.
At the other extreme is Group 4: agents which probably do not cause cancer in humans. The only compound assigned to this group is caprolactam, a chemical used in the production of nylon.
The vast majority of substances assessed by IARC fall into the intermediate groups. Group 2B, to which coffee was assigned for 25 years, also includes such things as pickled foods, the pesticide DDT, talc, and cell phone emissions.
Over the past 25 years a large number of epidemiologic studies have come out examining the association of coffee drinking with cancer in many different populations worldwide. Many of these are prospective studies, which are generally viewed as being less subject to bias than case-control studies because they obtain information about health habits/behaviors from a large group of healthy individuals who are then followed to see who develops disease. In addition, there are meta-analyses of individual studies. These are “statistical overviews,” which can provide more powerful assessments of the evidence on a given question.
What the large body of data on the effects of coffee drinking accumulated over the past 25 years shows is that coffee drinking is associated with reduced risk of several cancers, such as those of the liver and the body of the uterus. At the same time, there is no evidence of an association with other cancers including those of the bladder, kidney, colorectum, breast, prostate, pancreas, or ovary. (Here I am following IARC’s summary. Actually, there have been many studies of coffee and colorectal cancer, and a good proportion of them also indicate that there may be a reduced risk associated with coffee intake).
Unlike many other exposures, coffee is something that we can study because it is widely consumed all over the world, it is a habitual behavior, and, in general, people can tell you about their intake. Furthermore, there is a wide range of intake, with some people abstaining and others drinking more than ten cups of coffee per day. This range of intake makes it easier to determine whether an observed association is actually due to coffee itself or to some other correlated factor.
So what did IARC conclude? Here is the key paragraph in the two-page summary of its assessment, published in Lancet Oncology:
“The combination of evidence suggesting lack of carcinogenicity for cancers of the female breast, pancreas, prostate, uterine endometrium, and liver, with inverse associations for the later two and inadequate evidence for all the other sites reviewed led to the conclusion that there is inadequate evidence in humans for the carcinogenicity of coffee drinking.”
In its assessments of carcinogenicity, IARC places the greatest value on studies in humans, that is, evidence from epidemiologic studies. Here we have a common exposure that has been studied using standard epidemiologic methods – methods that have successfully used to identify such agents as tobacco, alcohol, arsenic, and asbestos as human carcinogens – in a large number of studies over twenty-five years.
And, by IARC’s own account, this body of evidence shows no indication of a positive association with any cancer. So, the obvious question is, Why wasn’t coffee reassigned to Group 4: “unlikely to cause cancer in humans”? If anything merits being placed in this category, would appear to be coffee!
IARC’s process appears to be afflicted with a deeply-ingrained double standard. A number of the agency’s assessments have been criticized for placing too much weight on isolated findings that appear to suggest a risk, while ignoring more solid studies that do not support the existence of risk. The agency’s assessments of cell phones and the weed-killer glyphosate are cases in point.
IARC’s reassessment of coffee can be viewed as a test case to see whether the agency can weigh the accumulated evidence on a question and come to a scientific, but also logical and common sense, conclusion. Of course, there is always more to be learned and the evidence is never as strong as we would like it to be. Nevertheless, in the case of coffee, we have as strong a signal as we are likely to get that this is not something we need to worry about.
In its press conference, the agency stated that “coffee may protect against cancer.” But then it went on to justify its designation of “unclassifiable as to its carcinogenicity in humans.” While no doubt there is relief at coffee’s being downgraded from “possible carcinogen,” this flagrant contradiction highlights the problems with IARC’s process, its classification scheme, and the messages it puts out to the public.
One more point, IARC made its announcement at a press conference, which coincided with the publication of a 2-page summary of its conclusions. However, the full report will only come out at a later date. To understand the reasoning behind the reclassification one will have to wait for the full report. So, why didn’t IARC wait to make its announcement when the report is available? It’s not as if a few more months in the 2B category would matter after 25 years.
All of this suggests an institution that is motivated more by the drive for publicity than by a concern for communicating useful information to the public.