Skip navigation

Evidence-based eating

Most people have a sense that eating whole-grain foods, like brown bread, cereals and pasta is a good thing. Putting a number on it is something else. A new meta-analysis [1] of case-control studies does just that.


The researchers from Minneapolis used aggressive searching techniques to identify case-control studies which investigated the intake of whole grain foods and incidence of any cancer. They found 40 studies involving many thousands of patients and controls.


Looking at instances where high versus low intake of whole grains was analysed or analysable, 43 of 45 instances in high quality studies had odds ratios of below 1, suggesting a positive association between whole grain food consumption and reduced risk of cancer. The pooled odds ratio for all cancers was 0.66 (95% confidence interval 0.60 to 0.72).

The findings were quite similar across all cancers. For individual cancers, the numbers of studies in which odds ratios were below 1 were:

  • 9 of 10 in colorectal cancer or polyps
  • 7 of 7 in gastric cancer
  • 7 of 7 for hormone-related cancers
  • 4 of 4 in pancreatic cancer
  • 2 of 3 in brain cancer
  • 8 of 8 for other cancers
Where pooled odds ratios could be obtained, most were in the range of 0.5 to 0.8, with only breast and prostate cancer having odds ratios close to 1 in a small number of studies.

There was a dose-response for intake (Figure). Compared with never eating whole grains, eating up to three servings a week reduced the pooled odds ratio to 0.82, while eating more than four servings a week reduced it to 0.59.


The implication of this analysis is that eating more than four servings of whole grain food a week reduces the chance of many cancers by about 40%. The effect in breast and prostate cancer is less. Of course, what we have here is an association, but the researchers examined all sorts of possible confounding variables, with little or no effect on the overall result.

Do we make enough of the information we have on lifestyles, and particularly eating habits, and their relationship to health? Telling people to eat brown bread is one thing, easily forgotten. Tell people, on the basis of good evidence in 40 studies with many thousands of people studied, that eating whole grains four times a week will reduce their risk of cancer by 40%, and emblazon it in clever television adverts, and Bandolier's guess is that habits could change.

And it's not as if the evidence on lifestyle and health is bad news. Much of it is very good news indeed. In its pages over the years Bandolier has carried news on the lowering, dramatically, of the risk of stroke by eating a few vegetables, and on heart diease by taking a few alcoholic drinks, and on longevity by walking. Fish eating does great things as well. Many of these can be quantified, and people respond to numbers simply put, while not being lectured. It's time for a change in how we give people information about how to avoid seeing the doctor. Anyone out there want to sponsor a healthy living edition of Bandolier ?

It's all down to implementation. The landlady of a pub had diverticulitis, but the high fibre biscuits she was prescribed gave her wind. She put the biscuits on the bird table, with the result that customers' cars were covered in bird droppings and they couldn't see to go home. It brings a whole new meaning to outcomes research.


  1. 1 DR Jacobs, L Marquart, J Slavin, LH Kushi. Whole-grain intake and cancer: an expanded review and meta-analysis. Nutrition and Cancer 1998 30: 85-96.

previous story in this issue