Whole grains and colorectal cancer: reading the prospective evidence.
A 2011 synthesis linked higher fiber and whole-grain intake with lower colorectal cancer incidence; it did not establish causation.

Illustrative food image, not a photograph of the study intervention.
A question about new diagnoses
The useful starting point is the outcome, not a promise about a particular breakfast. Aune and colleagues examined dietary fiber, whole grains and the incidence of colorectal cancer. Their 2011 BMJ review brought together 25 prospective studies, including cohort and nested case-control designs. These were observational comparisons, not trials assigning people to different grain intakes. The distinction matters even when the endpoint is a cancer diagnosis rather than a laboratory marker. The review concerns who developed colorectal cancer during follow-up; it is not evidence that grains treat an existing tumor or improve survival after diagnosis.
Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, Norat T.. BMJ (Clinical research ed.), 2011. PMID 22074852 · DOI
What the pooled estimates mean
For total dietary fiber, the dose-response analysis combined 16 studies and reported a relative risk of 0.90 for each additional 10 grams of daily intake, with a 95% confidence interval of 0.86 to 0.94. For whole grains, six studies contributed to an estimate of 0.83 per additional three daily servings, with a confidence interval of 0.78 to 0.89. These are relative comparisons across observed intake levels, not absolute percentage-point reductions in an individual’s risk. The intake increments describe the statistical analysis; they are not a personalized dietary prescription or a guarantee of protection.
Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, Norat T.. BMJ (Clinical research ed.), 2011. PMID 22074852 · DOI
Fiber sources were not interchangeable
The source-specific results deserve their own line in the story. For each additional 10 grams per day of cereal fiber, the summary relative risk of colorectal cancer was 0.90 (95% confidence interval 0.83 to 0.97). Estimates for fruit, vegetable and legume fiber had confidence intervals that included 1.0. That does not establish that those foods have no value, nor does it justify declaring cereal fiber the single active ingredient. It means this review did not produce equally clear statistical associations for every fiber category. Separate exposure analyses should not be rewritten as a randomized contest between foods.
Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, Norat T.. BMJ (Clinical research ed.), 2011. PMID 22074852 · DOI
A boundary around the conclusion
The authors explicitly called for better assessment of measurement error and additional analyses addressing residual confounding. That caution belongs beside the favorable estimates, not in invisible fine print. Recording diet and following later diagnoses strengthens the time sequence, but it does not randomly distribute the other characteristics of higher-fiber eaters. The defensible conclusion is that higher fiber intake, especially cereal fiber and whole grains, was associated with lower colorectal cancer incidence in this evidence base. The review cannot assign a cancer-prevention score to a recipe, prove that a fiber supplement reproduces the association, or establish treatment effects. The distinction between association and an assigned dietary effect applies to the whole-grain estimate just as much as it does to the total-fiber estimate; pooling does not remove that limitation.
Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, Norat T.. BMJ (Clinical research ed.), 2011. PMID 22074852 · DOI