This story is about a preliminary report to the European Society of Cardiology. The story asserts that UK scientists analyzed a large health information dataset and found a correlation between taking statin drugs and a reduced likelihood of dying from certain cancers.
However, the story got this correlation wrong. As the news release from the ESC states, the correlation is between “high cholesterol”–not statin use–and a reduced likelihood of dying from certain cancers. This disconnect seems to also stem from the news release, which offers up a misleading headline, and hints later that it’s probably the use of statins in individuals who have high cholesterol that creates this relationship between cholesterol levels and cancer risk. But, crucially, it offers no evidence of that, just conjecture.
Other than this major problem, the story does some things right: It wisely urges caution about the data in its opening sentence, repeats that caution several times, and tried to help readers see the difference between correlation and causation– something we hold dear here. The story used phrases like “the study found an association — but no proof — of prolonged survival” and quoted a scientist saying: “We cannot, however, recommend statins for cancer prevention without a positive clinical trial.”
News information about statins is of value to many people, since high cholesterol is very common. News stories like this are noticed widely by both the lay audience and the medical community, as evidenced by this study in the British Medical Journal, (and examined by us in a recent blog post). Considering this history, it is even more crucial that reporting on this news is careful.
The story does not discuss the cost of statins. While their use for cancer is only hypothetical, the story could have touched on how widely they are prescribed now and some cost information.
The story clearly states several times that there is no clinical benefit known, but only a correlation between people taking the medications and lowered incidence of certain cancers. Of course, based on the actual release, this is wrongly stated. The correlation is between high cholesterol levels and a lowered incidence of certain cancers.
Also, importantly, we’re only given relative risk numbers here, not absolute risks, and that should have been explained. Also, we’re told the “statin” group lived longer overall, but that doesn’t mean much without numbers.
Statins have been identified with some serious side effects, including memory loss, muscle pain, and a condition known as rhabdomyolysis, and the story should have made some reference to this. In this article, the FDA explains risks.
The story fails to accurately explain that the finding is correlation between high cholesterol and not the use of statins. If that mistake had been absent, we would have praised an expert saying it is “too soon” for anyone to accept the correlation as significant for clinical use.
Here is an important excerpt from the news release, with italics from editors:
“Dr Carter said: “Our research suggests that there’s something about having a high cholesterol diagnosis that improves survival and the extent to which it did that was quite striking in the four cancers studied. Based on previous research we think there’s a very strong possibility that statins are producing this effect.”
There is an enormous gap between “a strong possibility” and stating in a story that the correlation shows statins are lowering risk.
We do praise the story for including a cautionary quote from an outside expert:
“Regardless of whether or not a person has cancer, statin use should be discussed with a health care provider,” said Eric Jacobs, strategic director of pharmacoepidemiology at the American Cancer Society.”
There is no disease mongering, though the jump to attributing this possible side benefit of statins could lead to people using statins for reasons not fully backed up by science.
The story does not identify whether there are any conflicts for the study authors, but they do not appear to have any when we researched them ourselves. We are glad the story included a quote from an independent source at the American Cancer Society.
The story is only about the concept of prevention, and not about a specific therapy regimen that is being suggested. In light of that, we can call it “not applicable.”
It’s obvious from the story that statins are widely available. If research shows they can lower risk, they are available already for that purpose.
The story does not establish the novelty for this particular association of statins with decreased cancer risk. There have been a number of large meta-analyses and retrospective studies about statins and cancer risk reduction (all showing different findings), and this latest result should have been placed in context within the body of research.
It is implied (but not stated) that perhaps the use of 1 million patient records is the largest study searching for the correlation – but the story does not state it.
Because the story includes quotes not found in the news release, this is satisfactory. But it is ironic that a closer reading of the news release might have helped the authors avoid a mistake.