-References
|
|
SYSTEMATIC REVIEW
OBJECTIVE: To review all prospective studies which published information about:
a) total dairy,
b) eggs,
c) or specific dairy items,
and their relationship with colorectal, colon, and rectum cancer.
DATA SOURCE: I searched the Pubmed database (January 1966 to December 14, 2007) for relevant studies using
any of the following keywords: dairy, milk, yoghurt, cream, egg, eggs, cheese, butter, margarine, diet, and dietary
combined with the word cancer.
Prospective studies published in the English language were included.
RESULTS: Data about 25 cohorts met the inclusion criteria:
1) Total disease risk.
- Ten studies (11 cohorts) combining 5,321 cases provided information about total dairy in relation to colorectal cancer risk.
Though most RRs for total dairy were < 1, significance was reached in only 2 studies combining 1,379 cases. No evidence was found that total dairy fat
increases colorectal cancer risk among 4 studies.
- A pooled analysis of 10 studies and an additional 9 studies (8.980 cases) provided information about total milk in relation to colorectal
cancer risk. No study found any significantly or nonsignificantly increased risk. Data of studies which stratified risk by cancer site (colon vs rectum) is presented:
The pooled analysis and 4 out of 5 studies found a protective effect against colon cancer risk. The pooled analysis found a significant protective effect against
rectal cancer risk, but the remaining 3 studies did not.
- A pooled analysis of 9 studies and an additional 6 studies (5,198 cases) provided information about cheese in relation to colorectal cancer risk.
Of total cheese, the pooled analysis found a weak increased colorectal cancer risk which tended to be stronger for colon than for rectal cancer. Two out of the
remaining five studies found an increased colon cancer risk. But of cottage cheese, the pooled analysis and both remaining studies found a decreased colorectal cancer risk.
- A pooled analysis of 8 studies and an additional 3 studies (4,681 cases) provided information about yoghurt in relation to colorectal cancer risk. No study found
any significant or nonsignificant association with risk.
- Data about butter and cream in relation to colorectal cancer is limited (A pooled analysis for butter/A pooled analysis and 1 additional study for cream).
No significant relations were found in the pooled analysis.
- Four studies provided information about eggs in relation to colorectal. Though all available RRs were above 1, no study found a significantly
increased risk.
2) Advanced-stage disease risk.
No prospective study provided information about any of the variables in relation to advanced-stage or metastatic cancer risk.
3) Mortality risk.
- One study provided information about dairy products (defined as milk + ice cream) in relation to colorectal cancer mortality. It found a nonsignificantly decreased risk.
- Three studies provided information about milk in relation to colorectal cancer mortality. No study found a significant association.
- Two studies privided information about cheese in relation to colorectal cancer mortality. One of the studies found an increased female rectal cancer mortality
risk.
- One study provided information about yoghurt in relation to colorectal cancer mortality. It found a decreased male rectal cancer mortality risk.
- Three studies provided information about eggs in relation to colorectal cancer mortality. Two studies found an increased trend in total colorectal
cancer mortality. The remaining study found an increased male colon cancer mortality risk only.
CONCLUSION: Prospective studies suggest that total dairy may decrease colorectal cancer risk. A probable protective effect of total milk against rectal
cancer - and even more consistently - against colon cancer risk is suggested. Total cheese may increase colon cancer risk, while there is consistent evidence
that cottage/ricotta cheese decreases colorectal cancer risk. No association between yoghurt and colorectal cancer risk is found consistently. Data of the relation
between butter,cream or eggs and colorectal cancer risk is to limited to come to a conclusion.
Little data is available about colorectal cancer mortality. It is possible that eggs increase mortality risk.
AUTHORS NOTE: This conclusion is in line with findings of the WCRF report (october 2007):
"The WCRF found a probable protective effect of milk against colorectal cancer, and limited evidence suggesting that cheese is a cause of colorectal cancer.
They generally stated about eggs: "For eggs, the evidence was too limited in amount, consistency, or quality to draw any conclusions".
The WCRF did not provide information about cottage cheese, yoghurt, butter and cream. And also, not about colorectal cancer mortality.
SPECIFIC FINDINGS:
A) Total dairy and colorectal cancer risk
Eleven cohorts (5,321 cases) examined total dairy consumption in relation to total colorectal (10, 21, 22, 24), colon (1a, 3, 5b, 16, 17, 23), and
rectal (3, 17, 23) cancer risk.
- Only two RRs were above 1 (1a, 17), and only once significance was reached: One small cohort (1a) using 41 cases found a significantly
increased colon cancer risk, but the association was for dairy products without milk.
- Two studies (10, 24) using 2,210 cases found a (non) significant protective effect against total colorectal cancer, and an additional two studies
using 690 cases found a (non)significant protective effect against colon (5b, 23) and rectal (23) cancer.
- Though most RRs for total dairy were < 1, significance was reached in only 2 studies (5b, 24) combining 1,379 cases.
- Two studies stratified risk by sex (16, 24). Results were identical in both cohorts.
Four cohort provided data about high- or low-fat dairy in relation to colorectal (14b, 21) or colon (5b, 6a) cancer risk.
- One study (5b) found no significant difference between high- and low-fat dairy in relation to cancer risk, and one (21) found no
significant association with disease risk of high- and low-fat dairy (no data shown).
- One study (14b) found a significant protective effect of high-fat dairy against colorectal cancer risk, and one (6a) found a slight
protective effect of fat from dairy products.
Prospective studies of total dairy and colorectal cancer risk:
| Author | Cohort | Cases | Cancer site | Relative Risk (RR) |
| 24) Park SY (2007) | The Multiethnic Cohort Study | 2,110 (1,138 men and 927 women) | Colorectal | Men: R = 0.80 (0.64-0.99; P = 0.02).
Women: RR = 0.81 (0.65-1.00; P = 0.09) |
| 23) Larsson SC (2006) | The Cohort of Swedish Men | 449 (276 colon and 173 rectum) | Colon, rectal | Colon: RR = 0.44 (0.25-0.76; P = 0.07).
Rectal: RR = 0.48 (0.23-0.99; P = 0.07) |
| 22) Kesse E (2005) | The E3N-EPIC Study | 172 | Colorectal | RR = 0.78 (0.49-1.22; P = 0.28) |
| 21) Lin J (2005) | The Women's Health Study | 223 | Colorectal | RR = 0.89 (0.54-1.47; P = 0.42) |
| 17) McCullough (2003) | The Cancer Prevention Study II Nutrition Cohort | 683 (522 colon and 180 rectum) | Colon, rectal | Colon: RR = 0.84 (0.54-1.29; P = 0.32).
Rectal: RR = 1.22 (0.64-2.33; P = 0.93) |
| 16) Wu K (2002) | The Nurses' Health Study and the Health Professionals Follow-up Study | 1,025 (626 women, and 399 men) | Colon | Women: RR = 0.78 (0.50-1.21; P = 0.26).
Men: RR = 0.78 (0.53-1.16; P = 0.33).
NOTE: Associations are with dairy calcium instead of dairy! |
| 10) Kato I (1997) | The New York University Women's Health Study | 100 | Colorectal | An inverse association. |
| 5b) Sellers TA (1998) | The Iowa Women's Health Study | 241 | Colon | RR = 0.7 (0.4-1.0; P = 0.03) |
| 3) Stemmermann GN (1990) | The Japan-Hawaii Cancer Study | 277 (189 colon and 88 rectum) | Colon, rectal | Colon: RR = 1.2 (0.9-1.8; P = 0.27).
Rectal: No significant association.
NOTE: RR = for low consumption vs high, and associations are with dairy calcium instead of dairy! |
| 1a) Philips RL (1975) | The Adventist Health Study | 41 | Colon | RR = 2.5 (P = 0.05).
NOTE: Association is for dairy products without milk! |
| Total number of cases: | 5,321 | | |
Open extended table in a new window.
B) Milk and colorectal cancer risk
One pooled analysis of 10 studies and an additional 9 studies (8,980 cases) provided information about milk in relation to total colorectal (18, 21, 22, 24),
colon (1a, 9, 13b, 17, 19, 23), and rectal (13b, 17, 19, 23) cancer risk.
- No study found a significant or nonsignificantly increased risk. Three studies (9, 13b, 21) found any RR > 1.
- Two out of four studies found a nonsignificant protective effect against total colorectal cancer risk (22, 24), but the latter for men only.
- The pooled analysis of 10 studies found a significant protective effect against colon cancer risk (3,482 cases), in addition 4 out of 5 studies found a
significant (1a, 23) or nonsignificant (13b, 17) protective effect against colon cancer risk.
- The pooled analysis of 10 studies found a significant protective effect against rectal cancer risk (1,437 cases), but none of the 3 remaining studies
(387 cases) found a significant or nonsignificant (P = < or = 0.10) protective effect against rectal cancer risk.
Only 3 studies provided any information about whole milk only (6a, 14b), or Relative Risks based on milk type stratified by fat content (23).
Prospective studies of total milk and colorectal cancer risk:
| Author | Cohort | Cases | Cancer site | Relative Risk (RR) |
| 24) Park SY (2007) | The Multiethnic Cohort Study | 2,110 (1,138 men and 927 women) | Colorectal | Men: RR = 0.78 (0.63-0.96; P = 0.08).
Women: RR = 0.85 (0.68-1.06; P = 0.15) |
| 23) Larsson SC (2006) | The Cohort of Swedish Men | 449 (276 colon and 173 rectum) | colon, rectal | Colon: RR = 0.65 (0.46-0.91; P = 0.02).
Rectal: RR = 0.69 (0.45-1.06; P = 0.11) |
| 22) Kesse E (2005) | The E3N-EPIC Study | 172 | Colorectal | RR = 0.54 (0.33-0.89; P = 0.09) |
| 21) Lin J (2005) | The Women's Health Study | 223 | Colorectal | RR = 1.12 (0.72-1.74; P = 0.79) |
| 19) Cho E (2004) | Pooled analysis of 10 studies | 4,992 (3,482 colon, and 1,437 rectum) | Colon, rectal | Colon: RR = 0.88 (0.79-0.99; P = 0.01).
Rectal: RR = 0.80 (0.66-0.96; P = 0.02) |
| 18) Sanjoaquin MA`(2004) | The Oxford Vegetarian Study | 95 | Colorectal | RR = 1.0 (0.65-1.87; P = 0.78) |
| 17) McCullough (2003) | The Cancer Prevention Study II Nutrition Cohort | 683 (522 colon and 180 rectum) | Colon, rectal | Colon: RR = 0.81 (0.60-1.10; P = 0.10).
Rectal: RR = 0.89 (0.54-1.47; P = 0.61) |
| 13b) Jarvinen R (2001) | Study name = ? | 72 (38 colon and 34 rectum) | Colon, rectal | Colon: RR = 0.46 (0.14-1.46; P = 0.09).
Rectal: RR = 1.13 (0.39-3.31) |
| 9) Gaard M (1996) | Study name = ? | 143 | Colon | Men: RR = 0.72 (0.25-2.07).
Women: RR = 1.24 (0.35-4.40) |
| 1a) Philips RL (1975) | The Adventist Health Study | 41 | Colon | RR = 0.3 (P = 0.05) |
| Total number of cases: | 8,980 | | |
Open extended table in a new window.
C) Cheese and colorectal cancer risk
A pooled analysis of 9 studies and an additional 6 studies (5,198 cases) provided information about cheese in relation to total colorectal, colon, and rectal cancer risk.
Total cheese: The pooled analysis (4,146 cases) found a weak increased colorectal cancer risk which tended to be stronger for colon (RR = 1.14), than for rectal
(RR = 1.08) cancer risk. Two other studies (1a, 13b) combining 79 cases found an increased colon cancer risk.
Cottage and/or ricotta cheese: The pooled analysis (4,146 cases) found a decreased colorectal cancer risk of cottage/ricotta cheese. Both other studies
(22, 23) combining 621 cases found a nonsignificantly decreased colorectal cancer risk of cottage cheese.
Prospective studies of cheese and colorectal cancer risk:
| Author | Cohort | Cases | Cancer site | Total cheese
(Relative Risk) | Cottage/ricotta cheese
(Relative Risk) |
| 23) Larsson SC (2006) | The Cohort of Swedish Men | 449 | Colorectal | RR = 0.79 (0.56-1.12; P = 0.15).
NOTE: This RR = for hard cheese, not for total cheese. | RR = 0.68 (0.40-1.16) |
| 22) Kesse E (2005) | The E3N-EPIC Study | 172 | Colorectal | RR = 0.97 (0.61-1.54; P = 0.74) | RR = 0.71 (0.46-1.08; P = 0.09) |
| 21) Lin J (2005) | The Women's Health Study | 223 | Colorectal | RR = 1.38 (0.87-2.19; P = 0.53) | |
| 19) Cho E (2004) | Pooled analysis of 9 studies | 4,146 (2,912 colon [1,505 proximal, and 1,238 distal], and 1,208 rectal) | Colorectal | RR = 1.10 (0.98-1.24; P = 0.21) | RR = 0.83 (0.72-0.96; No P-value) |
| 18) Sanjoaquin MA`(2004) | The Oxford Vegetarian Study | 95 | Colorectal | RR = 0.98 (0.48-2.03; P = 0.63) | |
| 13b) Jarvinen R (2001) | Study name = ? | 72 (38 colon, and 34 rectum | Colon, rectal | Colon: RR = 2.42 (0.91-6.43).
Rectum: RR = 1.12 (0.43-2.91) | |
| 1a) Philips RL (1975) | The Adventist Health Study | 41 | Colon | RR = 3.8 (P = > 0.05) | |
| Total number of cases: | 5,198 | | |
Open extended table in a new window.
D) Yoghurt and colorectal cancer risk
A pooled analysis of 8 studies and an additional 3 studies (4,681 cases) provided information about yoghurt in relation to total colorectal, colon, and rectal cancer risk.
No significant or nonsignificant association was found. And no consistent trend is suggested.
Prospective studies of yoghurt and colorectal cancer risk:
| Author | cohort | Cases | Cancer site | Relative Risk (RR) |
| 23) Larsson SC (2006) | The Cohort of Swedish Men | 449 (276 colon and 173 rectum) | Colon, rectal | Colon: RR = 1.17 (0.88-1.56; P = 0.22).
Rectal: RR = 0.94 (0.66-1.33; P = 0.29) |
| 22) Kesse E (2005) | The E3N-EPIC Study | 172 | Colorectal | RR = 0.82 (0.54-1.25; P = 0.42) |
| 21) Lin J (2005) | The Women's Health Study | 223 | Colorectal | RR = 1.23 (0.79-1.92; P = 0.66) |
| 19) Cho E (2004) | Pooled analysis of 8 studies | 3,837 (2,724 colon and 1,129 rectum) | Colon, rectal | Colon: RR = 0.89 (0.78-1.01; P = 0.22).
Rectal: RR = 1.03 (0.86-1.25; P = 0.73) |
| Total number of cases: | 4,681 | | |
Open extended table in a new window.
E) Eggs and colorectal cancer risk
Four small studies provided information about eggs in relation to total colorectal, colon, and rectal cancer risk. And only 1 study stratified risk by cancer site
(colon vs rectum).
- Two studies (1a, 1c) found no association with colon cancer risk, but did not provide Relative Risks.
- The other two studies found nonsignificantly increased trends: Once for total colorectal cancer risk (18), and once for rectal cancer risk only (13a).
Only one study (13a) provided Relative Risk of eggs stratified by prepariation method. It found an increased total colorectal cancer risk of fried eggs
(RR = 1.45; 95% CI = 0.82-2.55), but not of boiled eggs (RR = 1.01; 95% CI = 0.60-1.70).
Prospective studies of eggs and colorectal cancer risk:
| Author | Cohort | Cases | Cancer site | Relative Risk (RR) |
| 18) Sanjoaquin MA`(2004) | The Oxford Vegetarian Study | 95 | Colorectal | RR = 1.29 (0.66-2.52; P = 0.43) |
| 13a) Jarvinen R (2001) | Study name = ? | 109 (63 colon and 46 rectum) | Colon, rectal | Colon: RR = 1.02 (0.48-2.14).
Rectal: RR = 1.54 (0.66-3.64). |
| 1c) Singh PN (1998) | The Adventist Health Study | 157 | Colon | Little evidence of an important association (no data shown). |
| 1a) Philips RL (1975) | The Adventist Health Study | 41 | Colon | No association. |
| Total number of cases: | 402 | | |
Open extended table in a new window.
|
|