Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Statins and colorectal cancer

Study
Results
Comment
Statin and other relevant stuff on Bandolier online

Associations between statins, low cholesterol, and cancer have been sought on many occasions in retrospective analyses of randomised trials, and in observational studies, with mixed results. Bandolier reviewed these in an on-line article in 2003. There has also been a suggestion that statins reduce colorectal cancer, but with conflicting results. A new observational study from Israel [1] provides more information, and highlights the differences between relative and absolute effects.

Study

This case-control study was performed in Israel in a managed care organisation. Cases were people with a diagnosis of colorectal cancer between 1998 and 2004, identified from a database. Controls without colorectal cancer were individually matched according to age, sex, location, and ethnicity.

Participants were interviewed to collect demographic information, and information about family history of cancer over three generations, and personal medical, dietary, and physical history. Medication use over the previous five years was by recall, and specific questions were asked about use of aspirin and NSAIDs, as well as statins. Statin use was verified through prescription records.

Results

After exclusions, there were 1953 patients and 2015 controls, with 1651 matched pairs. The average age was 70 years, split equally between men and women. For the matched pairs, the main differences were that cases had lower rates of hypercholesterolaemia (21 vs 26%) and sports participation (32 vs 41%), and higher rates of family history of colorectal cancer (6.5 vs 4.1%) and lower vegetable consumption (40 vs 32%). Use of statins for at least five years was lower in cases than in controls (Figure 1).



Figure 1: Percentage of cases and controls taking statins for five years or more





After adjusting for possible confounding factors, use of statins was associated with a significant reduction of risk of colorectal cancer, with an odds ratio of 0.53 (0.38 to 0.74). Similar reductions were found for colon cancer and rectal cancer separately, and for patients with hypercholesterolaemia, ischaemic heart disease, and irritable bowel disease. Use of aspirin or NSAIDs for five years or more was also associated with reduced risk of colorectal cancer, with an adjusted odds ratio of 0.70 (0.55 to 0.90).

Comment

This is a high quality study showing a large association between higher statin use and lower colorectal cancer rate. There is biological plausibility from experimental studies, though other observational studies with smaller numbers of cases have had mixed results, not all producing significant reductions, though tending to lower colorectal cancer rates with higher statin use.

At one level this is really interesting, but the authors also took the trouble to evaluate what the absolute results would look like if these observational results translated into clinical effect of treatment. In 2002 the incidence of colorectal cancer was 42 per 100,000 Jewish men in Israel. If 100,000 men took a statin for five years, the implication is that there would be 21 fewer colorectal cancers, using an odds ratio of 0.5.

That would be equivalent to an NNT of 100,000 divided by 21, or 4,800. In high-risk cases with double the incidence, the NNT would still be 2,400. For comparison, the five-year NNT to prevent heart attack or stroke for statins in people with an annual risk of 3% is about 20. Hundreds of thousands of people would have to be enrolled in randomised trials to prove the point.

Thus while the result is interesting, and probably important, the immediate practical use is that of knowing that long-term use of statins has a small additional benefit.

Reference:

  1. JN Poynter et al. Statins and risk of colorectal cancer. New England Journal of Medicine 2005 352: 2184-2192.

Statin and other relevant stuff on Bandolier online

Statin use continues to be a major component of interventions to reduce cardiovascular disease, along with healthy living exhortation. Readers should know that Bandolier online has been collecting and reviewing good quality evidence for both.

The Bandolier statin pages have collected systematic reviews and meta-analyses of efficacy data, in broad populations and in special groups. They also have reviewed the evidence on the major possible adverse events occasionally attributed to statins. All of this is collected together in a 30-page downloadable PDF.

The Bandolier healthy living pages have recently been overhauled, with material collected together by age. Evidence on eating, drinking, exercise, and weight, plus much more. The pages are written for anyone, not just professionals, and may be useful for anyone wanting to improve their chances of a healthier, and perhaps better and longer life, but who wants to avoid drugs or fads.

previous or next story