Back to Menu

Diabetes and bowel cancer
(March 2006)

These days, a popular approach to studies of risk factors for disease is the meta-analysis which combines multiple studies. This provides much larger numbers and increases the likelihood of finding statistical significance. In doing so, it combines positive and negative studies. This has now been done with diabetes as a risk factor for bowel cancer. A Swedish group from the famed Karolinska Institute analyzed a total of 15 studies with a total of 2,593,935 participants and concludes there is a modest, but statistically significant, 30 percent increased risk of bowel cancer among diabetics. They note "our findings strongly support a relationship between diabetes and increased risk of colon and rectal cancer in both women and men".

Commentary: The results do give some support to the argument that diabetes is a risk factor for bowel cancer, but it is "some" support, not "strong" support, and the increased risk of 30 percent is quite modest. There are two important issues:

- the meta-analysis glosses over the fact that about one-half the studies show a statistically significant increase and one-half do not (and are, therefore, negative)

- some of the studies take into account (adjust for) only, at most, a few of the other factors thought to influence bowel cancer (age, weight, physical activity, family history, diet, etc). Others adjust for many variables, but there is no mention of screening for bowel cancer by sigmoidoscopy or colonoscopy. If the diabetics undertook such screening less often than the non-diabetics that could easily account for a 30 percent increased risk. To this reviewer, that seems to be a serious flaw.

That does not mean the study should be ignored. The prudent conclusion is that every diabetic starting at age 45 should be sure to have regular bowel examinations for the detection of and removal of benign growths (polyps) that could turn malignant. Healthful Life recommends such screening for every adult. Our recommendation is:

Starting at age 45,

either full colonoscopy every 10 years

or

Left-sided lower bowel examination (sigmoidoscopy) every 5 years plus annual testing of bowel movements for subtle bleeding (hemoccult test).

The question is whether diabetics should be considered high risk. Until the issue is settled, the conservative approach is to designate diabetics as at possibly increased risk and recommend a full colonoscopy every ten years (rather than an option for the less extensive sigmoidoscopy every five years).

Larsson, S.C., et al. Diabetes mellitus and risk of colorectal cancer. Journal of the National Cancer Institute. Vol 97 (November 16) Pgs 1679-1687. 2005.

Supported by