Previous Month | RSS/XML | Current
Most of the previous entries in this series on the supposed health effects of coffee have been about claims that coffee is good for you, but rarely a study claims that drinking coffee is the opposite1. Here's a recent one.
A study claiming that drinking very hot coffee increases the likelihood of cancer of the esophagus has recently been reported by several news outlets2. However, according to the study in question, the problem is not with coffee itself, but with drinking any very hot liquid, so that hot tea and hot chocolate are included.
The headline claim is that drinking "scalding" hot beverages can triple the risk of esophageal cancer. Tripling the risk certainly sounds bad, but thrice what? It's a meaningless claim if we don't know what the risk of getting the cancer is for those who don't drink very hot beverages: three times zero is still zero.
The tripling claim is a relative risk or risk ratio (RR), that is, the ratio between the people in the study who drank "very hot" drinks and ended up with cancer of the esophagus to those who didn't drank such hot drinks and still got it. The RR tells us that there were three times as many in the former group as in the latter, but how many were in the latter? Before we can evaluate the RR claim, we need to know the absolute risk (AR) of esophageal cancer, that is, the risk of getting it without drinking superheated beverages3.
Cancer of the esophagus appears to be rare, but how rare? I've known more than one person who had cancer, including a couple who died from it, but I've never met anyone with cancer of the esophagus. The only public figure I've ever heard of who had it was Christopher Hitchens, who died all too young from complications of the cancer4. Did Hitchens drink extremely hot coffee or other beverages? I don't know, but he was notorious for smoking cigarettes and drinking copious quantities of alcoholic beverages, presumably cold5. As many of the news articles point out, cigarette smoking and drinking alcohol are higher risks for esophageal cancer than drinking very hot beverages.
Now, I'm not saying there's no risk of getting cancer of the esophagus―obviously, there must be some risk―but if the risk is very low then tripling it will produce a risk that is still low, maybe very low. This is important to know because life is inherently risky, and we need to know just how risky an activity is to rationally decide whether to do it. For instance, skydiving is risky, but apparently it's enough of a thrill for some people that it's worth the risk. Similarly, there may be some people who enjoy drinking scalding hot coffee so much that a tripled risk of cancer is worth it for them; Hitchens, for instance, claimed that smoking and drinking were worth not just risking it, but actually getting cancer.
In these entries on how health claims are reported, I'm usually quite critical of the reporting, but most of the reports on this study were better than usual. Of course, that's a low standard since so many news reports of new studies simply rewrite the press release without asking questions. However, some of these reporters interviewed an expert who was not connected to the study itself, who then pointed out that the absolute risk is less than 1%, and that smoking cigarettes and drinking alcohol are greater risks than drinking very hot drinks. So, the reporters, to their credit, did some actual reporting.
That said, "less than 1%" still sounded rather high to me, since that would mean one out of a hundred people gets esophageal cancer. Actually, "less than 1%" refers to the percentage of cases of cancer that are of the esophagus, but not everybody gets cancer. What we really need to know is the prevalence of esophageal cancer in the population. It turns out that the rates are quite different for men and women, though both rates are below 1%. For men, it's about .76% and, for women, only around .23%6. Putting these two figures together, the prevalence is approximately .5%7, which is indeed less than 1%. Tripling this rate by drinking searing-hot coffee leads to a risk of 1.5% for people in general.
Whether this is enough of a risk to justify letting your coffee cool off a bit longer before drinking it, I'll leave to your judgment.
Notes: