His report appears in the May 19 issue of the New England Journal of Medicine.

"This is the first large-scale study in women to assess the efficacy of colonoscopy as a colon cancer screening tool," Schoenfeld added.

"While fecal occult blood testing and flexible sigmoidoscopy are less expensive, faster and require no sedation, 65 percent of women with advanced precancerous polyps in our study would have lesions missed if these were the only screening tests performed because precancerous polyps are found deeper in the colon in women," the researcher said in a statement.

Schoenfeld's team evaluated 1,463 women, aged 50 to 79, including 230 with a family history of colon cancer. They performed colonoscopy on each one and found it detected advanced polyps -- considered a precursor to cancer -- in 72 of the women, or 4.9 percent. They also noted the location of the polyps.

Then the "diagnostic yield" of sigmoidoscopy was calculated by estimating the proportion of patients whose lesions would have been identified if they had that test alone. The lesions were considered detectable if they were found in the parts examined by a sigmoidoscopy.

Then the team compared the findings with results from another study called the Veterans Administration Cooperative Study 380, which evaluated more than 3,000 patients, 97 percent of them men. The VA study had found that fecal occult blood testing (which detects microscopic amounts of blood in the stool) and sigmoidoscopy would identify more than 70 percent of men with advanced polyps.

Not so for women, however, Schoenfeld's team found. While colonoscopy found advanced polyps in 4.9 percent of the women, sigmoidoscopy alone would have found polyps in only 1.7 percent of the women, and missed them in 3.2 percent. Only 35.2 percent of women with advanced polyps would have had their lesions identified if they had undergone just sigmoidoscopy, as compared with 66.3 percent of the men matched for age from the VA study.

Schoenfeld acknowledged that his recommendation that men and women of average risk have a screening colonoscopy at age 50 is at odds with official guidelines recommended by such organizations as the American Cancer Society.

It recommends men and women at average risk begin to be screened at age 50, using either blood stool tests annually, sigmoidoscopy every five years, annual blood stool tests plus sigmoidoscopy every five years, double contrast barium enema every five years, or colonoscopy every 10 years.

"These data reinforce two key points -- that colonoscopy is clearly the preferred colorectal cancer screening tool in women," he said. "And we need to remember that we can't take medical research studies that have been done in men and apply those to women."

Schoenfeld said both men and women should undergo colonoscopy at age 50 if they are of average risk.

Colorectal cancer is the third most common cancer in the United States, according to the American Cancer Society. About 104,950 new cases of colon cancer and 40,340 new cases of rectal cancer will be diagnosed in 2005, the society estimates. About 56,290 deaths in total are expected from both malignancies this year.

Dr. Donald David, director of the department of gastroenterology at the City of Hope National Medical Center, in Duarte, Calif., said the study results are not surprising. "We always thought colonoscopy was the best test for colon cancer screening," he said. "What we do see yet again is that colonoscopy is the only test that makes sense."

David predicted a trend toward using colonoscopy and dropping sigmoidoscopy. He added he is currently reviewing screening guidelines for a national cancer network, and will recommend that change.

More information: To learn more about colon cancer, visit the American Cancer Society.