A story on the MSNBC website is headlined, “Coffee habit may protect against breast cancer .” It states:
“Women who drink five or more cups of coffee a day are 57 percent less likely to develop estrogen receptor-negative breast cancer than women who drink less than a cup of coffee a day.”
The story concludes:
Because coffee seems to have different effects on different kinds of cancers, people are often confused as to what their risks are with coffee intake, said Dr. Michelle Shayne, an oncologist and assistant professor of medicine at University of Rochester Medical Center in New York, who wasn’t involved with the study.
For example, a review of studies published last year in the journal Nutrition and Cancer showed that coffee seemed to have a protective effect on endometrial and colorectal cancers, but didn’t confer any benefits for breast, pancreatic, ovarian, kidney or prostate cancers.
All of the conflicting evidence out there means “I wouldn’t necessarily tell my patients after reading an article like this to drink more than five cups of coffee a day to lower your risk of a particular type of breast cancer,” Shayne told MyHealthNewsDaily. “It’s just that if you happen to enjoy coffee consumption, you may possibly have an added benefit of protection against one subtype of breast cancer.”
People may be confused because of stories like this one, that spent a lot of time trying to find chemical reasons for the finding. But it didn’t devote one word to explaining the limitations of observational studies.
“But it’s important to realize that the study didn’t prove that coffee itself prevents cancer, only that woman who drink a lot of it are at lower risk. Given the uncertainty, researchers say it’s premature to recommend coffee to women worried about breast cancer.”
It, too, didn’t explain WHY such a study can’t prove that causal link. But at least it was in the ballpark of being more helpful. The MSNBC story just made it sound like science itself was flip-flopping and unhelpful.
Any journalist – or any consumer – who is confused about this should read our primer,
“Does The Language Fit The Evidence? – Association Versus Causation.”