ARTICLES Relationship of Diet to Risk of Colorectal Adenoma in Men
Background: Rates of colorectal cancer in various countries are strongly correlated with per-capita consumption of red meat and animal fat and inversely associated with fiber consumption. There have been few studies, however, of dietary risk factors for colorectal adenomas, which are precursors of cancer. Purpose: Our purpose was to determine prospectively the relationship between dietary factors and risk of colorectal adenomas. Methods: Using data from the Health Professionals Follow-up Study, we documented 170 cases of adenomas of the left colon or rectum in 7284 male health professionals who completed a food-frequency questionnaire in 1986 and who had a colonoscopy or sigmoidoscopy between 1986 and 1988. Relative risk (RR) of adenoma was determined according to quintiles of nutrient intakes. Results: After adjustment for total energy intake, saturated fat was positively associated with risk of colorectal adenoma (P for trend = .006); RR for the highest versus the lowest quintile of intake was 2.0 (95% confidence interval [CI] = 1.2-3.2). Dietary fiber was inversely associated with risk of adenoma (P for trend Indication for endoscopy, % Blood in stool Occult fecal blood Abdominal pain Change in bowel habit Routine screening
•Men in both highest quintile for saturated fat intake and lowest quintile for fiber intake versus those in both lowest quintile for saturated fat intake and highest quintile for fiber intake.
Vol. 84, No. 2, January 15, 1992
ARTICLES
Downloaded from http://jnci.oxfordjournals.org/ at The University of British Colombia Library on July 11, 2015
The prevalence of adenoma of the distal colon or rectum increased approximately linearly with age. Prevalence of adenoma in the subpopulation we studied was defined as the proportion of men who had a diagnosis of adenoma in the colon or rectum of the total number who had a colonoscopy or a sigmoidoscopy. Subjects who reported a history of colon or rectal cancer in either parent had an elevated risk of developing an adenoma (age-adjusted relative risk [RR] = 1.8; 95% CI = 1.2-2.8). We adjusted for age and family history in all the subsequent analyses. Men on a low-fat, high-fiber diet tended to be older, thinner, and more physically active, smoked fewer cigarettes, and drank less alcohol than men on a high-fat, low-fiber diet (Table 1). A higher proportion of men on a low-fat, high-fiber diet had endoscopic procedures for routine screening whereas more men on a high-fat, low-fiber diet had procedures to evaluate fecal blood. The relationship of total energy intake to the risk of colorectal adenoma was not statistically significant (Table 2). For total fat intake adjusted for total energy intake (energy-adjusted), a nonlinear relationship was observed; RRs for men in each of the four upper quintiles were elevated compared with those for men in the lowest quintile. Analysis by the primary sources and types of fat showed a statistically significant positive trend for association of risk for colorectal adenomas with intake of animal fat (P = .03) and saturated fat (P = .006). The RR for men in the highest quintile for
saturated fat was 1.97 (95% CI = 1.2-3.2) compared with that for men in the lowest quintile. When we included energy-adjusted saturated, monounsaturated, and polyunsaturated fats in a multivariate model with age, total energy intake, and family history of colorectal cancer, the effect of saturated fat persisted (RR = 2.4, 95% CI between highest and lowest quintiles = 1.1-5.5); the effect of polyunsaturated fat remained null; the effect of monounsaturated fat disappeared (RR = 0.85). Total protein, animal protein, and cholesterol were not related to the risk of adenoma. Energy-adjusted carbohydrate intake was inversely related to the risk of adenoma; this relationship was statistically significant but did not persist after adjustment for fiber intake. We examined different sources of animal fat in relation to the risk of adenoma. Increased intake of red meat and increased intake of fat from dairy sources were both associated with an increased risk of adenoma (Table 3) (P - .03 for red meat and P = .04 for dairy fat). Nonsignificant inverse trends were present for chicken (P = .28), fish (P = .23), and combined intake of chicken and fish (P = .13). The association of specific food groups with risk of adenoma may be difficult to interpret because of their intercorrelations. For example, chicken and fish are recommended as substitutes for red meats. Not surprisingly, red meat intake was inversely correlated with total chicken and fish consumption (r - -.20). Thus, to represent this pattern of substitution, we computed a ratio of total consumption of red meat (beef, pork, and lamb from all sources) to consumption of chicken and fish. This ratio showed a positive association with the risk of adenoma that was statistically significant (test for trend, P - .02). The association of energy-adjusted intake of crude and total dietary fiber with risk of adenoma was both inverse and statistically significant (P65 2.29 >39 1.97 >31.3 1.42 >23.5 1.97 >24.7 1.49 >13.1 1.49
1.35-3.88
.003
1.21-3.21
.03
0.90-2.24
.23
1.20-3.22
.006
0.91-2.45
.04
0.91-2.45
.12
63 1.02
0.42-1.07
.43
0.65-1.60
.94
Carbohydrates, g/d RR
< 162.6 1.0
162.3-181.2 1.09
181.3-197.5 0.60
197.6-217.2 0.66
>217.2 0.47
0.28-0.77
Background: Rates of colorectal cancer in various countries are strongly correlated with per-capita consumption of red meat and animal fat and inversely associated with fiber consumption. There have been few studies, however, of dietary risk factors for colorectal adenomas, which are precursors of cancer. Purpose: Our purpose was to determine prospectively the relationship between dietary factors and risk of colorectal adenomas. Methods: Using data from the Health Professionals Follow-up Study, we documented 170 cases of adenomas of the left colon or rectum in 7284 male health professionals who completed a food-frequency questionnaire in 1986 and who had a colonoscopy or sigmoidoscopy between 1986 and 1988. Relative risk (RR) of adenoma was determined according to quintiles of nutrient intakes. Results: After adjustment for total energy intake, saturated fat was positively associated with risk of colorectal adenoma (P for trend = .006); RR for the highest versus the lowest quintile of intake was 2.0 (95% confidence interval [CI] = 1.2-3.2). Dietary fiber was inversely associated with risk of adenoma (P for trend Indication for endoscopy, % Blood in stool Occult fecal blood Abdominal pain Change in bowel habit Routine screening
•Men in both highest quintile for saturated fat intake and lowest quintile for fiber intake versus those in both lowest quintile for saturated fat intake and highest quintile for fiber intake.
Vol. 84, No. 2, January 15, 1992
ARTICLES
Downloaded from http://jnci.oxfordjournals.org/ at The University of British Colombia Library on July 11, 2015
The prevalence of adenoma of the distal colon or rectum increased approximately linearly with age. Prevalence of adenoma in the subpopulation we studied was defined as the proportion of men who had a diagnosis of adenoma in the colon or rectum of the total number who had a colonoscopy or a sigmoidoscopy. Subjects who reported a history of colon or rectal cancer in either parent had an elevated risk of developing an adenoma (age-adjusted relative risk [RR] = 1.8; 95% CI = 1.2-2.8). We adjusted for age and family history in all the subsequent analyses. Men on a low-fat, high-fiber diet tended to be older, thinner, and more physically active, smoked fewer cigarettes, and drank less alcohol than men on a high-fat, low-fiber diet (Table 1). A higher proportion of men on a low-fat, high-fiber diet had endoscopic procedures for routine screening whereas more men on a high-fat, low-fiber diet had procedures to evaluate fecal blood. The relationship of total energy intake to the risk of colorectal adenoma was not statistically significant (Table 2). For total fat intake adjusted for total energy intake (energy-adjusted), a nonlinear relationship was observed; RRs for men in each of the four upper quintiles were elevated compared with those for men in the lowest quintile. Analysis by the primary sources and types of fat showed a statistically significant positive trend for association of risk for colorectal adenomas with intake of animal fat (P = .03) and saturated fat (P = .006). The RR for men in the highest quintile for
saturated fat was 1.97 (95% CI = 1.2-3.2) compared with that for men in the lowest quintile. When we included energy-adjusted saturated, monounsaturated, and polyunsaturated fats in a multivariate model with age, total energy intake, and family history of colorectal cancer, the effect of saturated fat persisted (RR = 2.4, 95% CI between highest and lowest quintiles = 1.1-5.5); the effect of polyunsaturated fat remained null; the effect of monounsaturated fat disappeared (RR = 0.85). Total protein, animal protein, and cholesterol were not related to the risk of adenoma. Energy-adjusted carbohydrate intake was inversely related to the risk of adenoma; this relationship was statistically significant but did not persist after adjustment for fiber intake. We examined different sources of animal fat in relation to the risk of adenoma. Increased intake of red meat and increased intake of fat from dairy sources were both associated with an increased risk of adenoma (Table 3) (P - .03 for red meat and P = .04 for dairy fat). Nonsignificant inverse trends were present for chicken (P = .28), fish (P = .23), and combined intake of chicken and fish (P = .13). The association of specific food groups with risk of adenoma may be difficult to interpret because of their intercorrelations. For example, chicken and fish are recommended as substitutes for red meats. Not surprisingly, red meat intake was inversely correlated with total chicken and fish consumption (r - -.20). Thus, to represent this pattern of substitution, we computed a ratio of total consumption of red meat (beef, pork, and lamb from all sources) to consumption of chicken and fish. This ratio showed a positive association with the risk of adenoma that was statistically significant (test for trend, P - .02). The association of energy-adjusted intake of crude and total dietary fiber with risk of adenoma was both inverse and statistically significant (P65 2.29 >39 1.97 >31.3 1.42 >23.5 1.97 >24.7 1.49 >13.1 1.49
1.35-3.88
.003
1.21-3.21
.03
0.90-2.24
.23
1.20-3.22
.006
0.91-2.45
.04
0.91-2.45
.12
63 1.02
0.42-1.07
.43
0.65-1.60
.94
Carbohydrates, g/d RR
< 162.6 1.0
162.3-181.2 1.09
181.3-197.5 0.60
197.6-217.2 0.66
>217.2 0.47
0.28-0.77
