Hyperinsulinemia suspected risk factor: large study links colorectal cancer with diabetes

OB/GYN News, Sept 15, 2004 by Mitchel L. Zoler

NEW ORLEANS -- Colorectal cancer was linked with diabetes in a case-control study with data on about 50,000 people.

People with diabetes were 32% more likely to have been diagnosed with colorectal cancer than were those without diabetes, a statistically significant difference, Dr. Vikas Khurana reported in a poster presentation at the annual Digestive Disease Week.

Because the average age of the people in the study was 60 years, it's likely that most of those with diabetes had the type 2 form, and the hyperinsulinemia in patients with type 2 diabetes may be what links the disorder with colorectal cancer, said Dr. Khurana, who is a gastroenterologist at the Overton Brooks Veterans Affairs Medical Center in Shreveport, La.

The finding is consistent with growing clinical and experimental evidence suggesting that patients with diabetes, and specifically with hyperinsulinemia, have an increased risk of developing colorectal cancer. Hyperinsulinemia leads to upregulation of growth hormone secretion, including insulin-like growth factor. Higher systemic level of insulin-like growth factor triggers cell cycle progression and inhibits cellular apoptosis, two mechanisms that might boost the appearance and growth of cancerous cells in the colon, said Dr. Rambabu Chalasani, another physician at Overton Brooks who collaborated on the study.

Eventually, the link between hyperinsulinemia and colon cancer may become strong enough to warrant changes in the surveillance schedule for colon cancer among people with type 2 diabetes or insulin resistance. For the time being, the findings are compelling enough to provide physicians with an extra incentive to start surveillance for colorectal cancer in patients with type 2 diabetes as soon as they reach 50 years of age, Dr. Khurana said in an interview with this newspaper.

The researchers reviewed medical records for 50,715 patients at Overton Brooks during October 1998 through June 2003; 91% were men. In this group. 8,976 had diabetes mellitus, and 770 had a diagnosis of colorectal cancer. In a univariate analysis, patients with diabetes were about twice as likely to also have colorectal cancer compared with people without diabetes.

A multivariate analysis was performed to adjust for the covariates of obesity, smoking, alcohol consumption, and use of aspirin.

After these possible confounders were taken into account, the increased risk of colon cancer among the patients with diabetes was 32%.

Alcohol use and obesity were both independent risk factors for an increased risk of colon cancer.

Because the effect of diabetes remained statistically significant after adjustment for the contribution of alcohol and obesity, the epidemiologic link between diabetes and colon cancer is even more believable. Dr. Khurana said at the meeting.

BY MITCHEL L. ZOLER

Philadelphia Bureau