Send to

Choose Destination
Eur J Nutr. 2011 Apr;50(3):173-84. doi: 10.1007/s00394-010-0128-5. Epub 2010 Aug 10.

Is dietary fat associated with the risk of colorectal cancer? A meta-analysis of 13 prospective cohort studies.

Author information

Department of Gastrointestinal Surgery, The Second Affiliated Hospital of Nan Chang University, 330006, Nanchang City, Jiangxi Province, China.



The results of animal studies suggest there is a significant role for dietary fat in the development of colorectal cancer (CRC). However, inconsistent results have been reported by epidemiological studies.


To evaluate the association between total dietary fat and risk of colorectal cancer development using a meta-analysis based on prospective cohort studies.


Published literature was retrieved from Medline, Embase and CNKI (China Knowledge Resource Integrated Database) databases updated to 1st May, 2009. Overall, thirteen prospective cohort studies with 3,635 cases and 459,910 participants were included.


The combined relative risk (RR) [95% confidence interval (95%CI)] for the risk of CRC was 0.99 (0.89,1.09) when the highest level of total dietary fat was versus (vs.) the lowest level. Stratified analyses according to gender, ethnicity, country and age showed that the highest level of total dietary fat did not increase the risk of CRC [RR (95%CI): 0.89 (0.77,1.03) for males; 1.09 (0.94,1.26) for females; 1.08 (0.94,1.25) for Caucasians; 0.90 (0.77,1.04) for Asians; 1.13 (0.94,1.36) for Americans; 0.92 (0.81,1.04) for individuals older than 40]. Besides those, the highest level of total fat diet also did not increase the risk of neither colon cancer [RR (95%CI): 0.96 (0.82,1.13)] nor rectal cancer [RR (95%CI):1.07 (0.63,1.82)]. Furthermore, neither animal fat nor plant fat were associated with the risk of CRC [RR (95%CI): 1.05 (0.91-1.22) for animal fat and 0.96 (0.82-1.11) for plant fat].


This meta-analysis suggests that dietary fat may not be associated with the increased risk of CRC. More well-designed studies with larger population performed among Asians are needed to further evaluate the associations. In addition, probable bias caused by measurement error should be noticed in this meta-analysis, and measurement error needs to be adjusted in the future studies.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center