Posted by Gary Schwitzer in Health care journalism
Maybe our continued messages about how journalists need to do a better job explaining observational studies are getting through.
HealthDay reported on an observational study, emphasizing the limitations:
“Research tracking more than 84,000 postmenopausal women for an average of 10 years found that those whose diets included more baked and broiled fish — defined as five or more servings per week — had a 30 percent lower risk of heart failure compared to women who ate less than one serving per month.
The main limitation of the study was its observational nature and the self-reported eating habits of participants, said Lona Sandon, an assistant professor of clinical nutrition at the University of Texas Southwestern in Dallas.
“What we don’t know is have these women been eating five servings of baked and broiled fish all of their lives, or is this something they started in their fifties?” Sandon said. “They may also have a more active lifestyle and eat less saturated fat. So there are a lot of differences, probably, in overall nutrition intake.”
Indeed, the study indicated that participants whose diets included more baked and broiled fish tended to be healthier and younger than peers who ate fried fish, as well as more physically active and fit. They were also more educated, less likely to smoke and had fewer incidences of diabetes, high blood pressure and cardiovascular disease.”
WebMD also referred to the new analysis as an observational study – although it didn’t explain what that means.
“This study shows an association between eating fish and heart failure risk, but it is not designed to show cause and effect.
Participants whose diets included more baked or broiled fish tended to be healthier and younger than those who ate fried fish. They also were more physically active and fit, more educated, and less likely to smoke, have diabetes, high blood pressure, and heart disease.
In addition, their diets included more fruits and vegetables, more beneficial fatty acids, and less unhealthy saturated and trans fatty acids.”
But on another study on another day this week, the Los Angeles Times was just wrong when it reported:
“Taking an acetaminophen tablet daily for at least five years reduces the risk of developing prostate cancer by 38%, researchers from the American Cancer Society reported Monday. Using the drug, the best-known form of which is Tylenol, also reduces the risk of the more aggressive form of prostate cancer by 51%, the team reportedin the online version of the journal Cancer Epidemiology Biomarkers and Prevention.”
But the researchers DIDN’T report that acetaminophen “reduces the risk.” Here’s what they reported:
Conclusion: These results suggest that long-term regular acetaminophen use may be associated with lower prostate cancer risk. Impact: If the association between acetaminophen use and lower risk of prostate cancer is confirmed, it could provide clues about biological mechanisms that are important in prostate carcinogenesis.
“May be associated” does not mean “reduces the risk.”
Association does not equal causation.
A statistical association does not warrant the active verb “reduces” the risk.
Comments are closed.