That was an actual headline from a story the other day about a paper published online first by Mayo Clinic Proceedings pointing to an association – not proof of cause – between heavy coffee consumption (>28 cups a week) and higher death rate in people younger than 55.
Drinking More Than Four Cups of Coffee A Day Raises the Risk of an Earlier Death (Business Insider story on San Francisco Chronicle website)
Four Cups of Coffee a Day Means Higher Death Risk – UPI story on HispanicBusiness.com
No it can’t…no it doesn’t…and no it doesn’t. At least there is no proof of “raising the risk of an earlier death” from the study in question. The researchers reported a “positive (statistical) association.” That’s not causation. So causal phrases such as “raises risk” or “can kill you” are unfounded. I always have to remind first-time readers that I’m not knocking the potential importance of observational studies. They are what they are and they ain’t what they ain’t.
At last check there were more than 170 stories about this study that turned up on a simple web search.
One of the study authors, in this video, explains that another of the team’s papers in another journal points to benefits or protective effects of coffee consumption. But that was a point that didn’t make it into many stories.
“Gregg Fonarow, co-chief of clinical cardiology at the David Geffen School of Medicine at UCLA, says, “Differences in other dietary factors, marital status and other socioeconomic factors that were not adjusted for in this study may account for some or all of these observations.”
Fonarow, who was not involved in this research, says observational studies that survey people about their coffee intake and tie that to how many died from any cause have yielded mixed results.
Consider a 2012 study that found that coffee drinkers ages 50-71 had a lower risk of death than their peers who did not consume coffee. In that study, researchers from the National Cancer Institute, part of the National Institutes of Health, and AARP found that the more coffee consumed, the more a person’s death risk declined.”
Look what happens when you make another phone call and involve another – independent – expert voice to scrutinize the evidence.
Coffee causation stories are getting stale with us. Just check out some past posts:
Anyone who wants to learn more about why the language matters in such stories should read our primer, “Does The Language Fit the Evidence? Association Versus Causation.”
ADDENDUM ON AUGUST 22: See Evelyn Lamb’s post, “Sex Makes You Rich? Why We Keep Saying ‘Correlation Is Not Causation.’ ”
Follow us on Facebook, and on Twitter: