My local NBC station (KARE-11) reported the “coffee may protect against heart failure” study 9 minutes deep into its local newscast last evening. It wasn’t a local study. It wasn’t a local story. In our community, we had a 5 year old shot and killed while he was sleeping in his home. We had major flooding in the state this week. We had news of a longrunning battle between 3M and local citizens over burning hazardous material. We had lots of big local news. But this TV station grabbed a story off the wire and put it up high in their newscast.
Another observational study about supposed benefits of coffee with no discussion of the limitations of observational studies and how they can’t establish cause-and-effect.
In other media, tired old coffee clichés and the same tired old trend of failing to explain the limitations of such research played out over and over.
“Hey java drinkers, that coffee buzz you love so much may also help prevent heart failure
…They found that those who drank a moderate amount of coffee daily, defined as the equivalent of two 8-ounce American cups per day, may experience protective benefits against heart failure by as much as 11 percent.”
Protective benefit? Wrong. Can’t be proven by such a study. Strong statistical assocation? That would be OK.
“Could two cups of coffee a day keep the heart doctor away? A new study shows that your daily cup of joe might provide that health benefit.”
“If you need an excuse to pour yourself that second cup of coffee, read on. Moderate, daily coffee drinking may be good for your heart — to a point, a new study suggests.”
”That morning cup o’ joe or mid-afternoon coffee pick-me-up may play a role in keeping your heart healthy, depending on how much you drink.”
I don’t know which is worse: the tired old writing that isn’t cute (cup o’ joe…if you need an excuse…java drinkers…coffee buzz…) or the tired old failure of most of these stories to explain how observational studies are done, what they can show and what they can’t.
Here are two examples of a better job of journalism:
However, he stresses that “the study does not prove benefit [of coffee in reducing heart failure risk].”
Researchers caution, however, that they can’t be sure whether these associations mean that drinking coffee actually makes people live longer.
As tired as we are of this continuing trend of journalistic failure to evaluate the quality of the evidence, we continue to try to help journalists and it is in that spirit that we once again remind people of our primer on the proper language to use in describing observational studies.