Alcohol and Exercise and Association and Causation
Many times before I have written about the association between alcohol and breast cancer risk and alcohol and recurrence risk and exercise and recurrence risk. Those associations go both ways--meaning, the alcohol part is associated with increased risk, and the exercise part with diminished risk. To point out the obvious: association does not equal causation and, if you read enough studies, you are sure to find contradictions. I remember two studies about caffeine and cancer risk many years ago. Literally published a few days apart, one screamed that X amount of coffee would increase pancreatic and other cancer risks, and the second screamed that the same amount of coffee could have a protective feature.
It is all enough to make your head spin. And usually leaves me in the same spot of trying to exercise common sense and remember the old mantra about moderation in all things.
From Health News Review comes this helpful in untangling all this article:
Exercise can cancel out the booze? Association ≠ causation
It has been suggested to us that we offer a weekly “Association ≠ causation” feature. Not sure we can promise weekly delivery, but we’ve certainly written on this theme a lot in 10 years. And here’s the latest installment.
CNN reports: “Exercise can cancel out the booze, says study.”
No, it did not. What the study, in the British Journal of Sports Medicine, stated was “We found a direct association between alcohol consumption and cancer mortality risk.” And – repeat after me – association ≠ causation. So a statement of causation – such as
“exercise can cancel out the booze” – is an overstatement.
Read our toolkit primer, “Observational Studies – Does The Language Fit The
Evidence? – Association Versus Causation.”
The journal – and the authors – cannot be absolved of all responsibility for
misleading statements, since their conclusion read: “Meeting the current physical
activity public health recommendations offsets some of the cancer and all-cause
mortality risk associated with alcohol drinking.” The term “offsets…risk” implies
But, in the end, we’re reviewing the journalism. And if you’re going to cover studies,
you need to independently vet the evidence, and what researchers claim. And that
didn’t happen here.
Not when CNN went on to report:
“You might want to chase that next beer with a little exercise.”
“Exercising the recommended amount “cancels out” the higher risk of cancer
death brought about by drinking.”
“Similarly, physical activity lessened any greater risk of death resulting from any cause due to alcohol.”
They allowed the researcher to discuss a “moderating effect of physical activity” when, in fact, cause and effect had not been proven.
They did throw in, 3/4 of the way into the story, an independent expert’s comment that “Because it is an observational study, the results only “suggest a relationship” between exercise, drinking and health benefits.” Too little, too late, after rampant use of causal language was embedded in readers’ minds by then.