Better Have Another Cup of Joe
Coffee, especially caffeinated coffee, has been the subject of numerous epidemiological investigations; and the endless cycles of "this week scientists said it causes cancer, last week they said it doesn't and next week they'll probably say it prevents it" did as much as anything to make even epidemiologists ask whether it was time for the science of black box epidemiology "to call it a day". Of course they didn't call it a day and many continue to try to discover whether everyday "stuff" like coffee and tea are in fact deadly carcinogens. They do it by attempting to discern an association between exposure and disease beyond what chance would predict. Here's the newest one on coffee: "Coffee and Tea Intake and Risk of Head and Neck Cancer: Pooled Analysis in the International Head and Neck Cancer Epidemiology Consortium".
The study is large and looks to be very well done. Overall, neither coffee nor tea seems to be associated with head and neck cancers but what's interesting is the unexpected and dose dependent decrease in cancer associated with drinking lots of caffeinated coffee each day. Those downing three or four cups each day reduced their risk by 30% and those drinking more than four cups per day lowered their risk by almost 40%. What to make of it?
Well, as we regularly argue here, it's pretty much impossible to know what to make of it. We could jump to the conclusion that coffee battles cancer but we'd be ignoring, among a host of other factors, the fact that a lot of head and neck cancers are caused by human papillomavirus (HPV). The authors write that "there is no reason to think that coffee intake is associated with human papillomavirus infection" but just because they can't think of a reason doesn't mean there isn't one. Maybe heavy coffee drinkers have fewer partners (all work and no play ... etc.). Maybe some people have a rare gene that simultaneously protects against head and neck cancer and makes them crave caffeinated coffee. Who knows?
None of which is to say that these sorts of studies are useless. In fact they're often invaluable. The point is that such studies ought to be put in perspective. They aren't valuable because they answer the question of causation. They can't. They're valuable because they generate insightful clues that can be shaped into testable hypotheses that if borne out lead to real answers about causation and from there to further clues about prevention and treatment.
