Hypertriglyceridemia is an independent risk factor

for only colorectal tubular adenoma in male



Masafumi Tabuchi 1) 2),  Joji Kitayama 2)  Kazushige Kawai 2),  Kenji Hatano and 2) and Hirokazu Nagawa2)


1) Nakameguro Digestive Disease Clinic

2) Department of Surgical Oncology, University of Tokyo



         Many studies have shown the positive correlation between lipid especially cholesterol and the incidence of colorectal adenoma or cancer.  Most of these results suggests that high intake of dietary fat tends to promote colon carcinogenesis 1, 2 3.  The fecal excretion of cholesterol was elevated in patients with colon cancer and adenomatous polyps, suggesting a possibility that the direct contact of cholesterol or its derivatives on colonic epithelia may be important for colonic carcinogenesis 4  On the serum lipid level, early studies showed an inverse association between cholesterol level and colon cancer, while contradictory results have also been reported 5, 6  However, another study has shown the small but significant positive association between serum cholesterol levels and the risk of colorectal adenomas 7  In addition, positive correlation with colonic adenomas with atherosclerotic diseases has been reported 8.  Those facts suggest that low serum cholesterol associated with malignancies may be a consequence of the impaired nutritional status rather than a cause of cancer.

         As compared with cholesterol, there has been less information on the relationship between serum triglyceride and colorectal neoplasm.  The striking similarity in lifestyle risk factors for colorectal cancer and insulin resistance, suggested that the hypertriglyceridemia as well as hyperinsulinemia and hyperglycemia associated with insulin resistance can support the development of colon cancer 9-12  Animal study also indicates that development of aberrant crypt foci is associated with circulating triglyceride as well as glucose, suggesting a possible role of triglyceride on colon cancer development 13.  On the contrary, other reports failed to detect significant association between serum levels of triglyceride as well as cholesterol and risk of colorectal cancer in human 14, 15 and animal studies16 . 

Thus, the contribution of circulating lipid level on colonic neoplasm development has not been satisfactorily determined.  In our clinic, we have performed large number of total colonoscopy for the screening of colon cancer and intensively tried to find the colonic adenomatous polyps, since it is generally accepted that most colorectal cancers arise from adenomatous polyps.  From the experience, we retrospectively analyzed the relationship between serum level of cholesterol or triglycerides and the presence of colonic adenoma.

                                          Methods and Patients

         A total of 4887 patients underwent total colonoscopy more than three times for the screening of colon cancer and precancerous lesions from 1995 to 2003.  Total colon was carefully examined with high resolution magnified chromo-endoscope.  More than 20 minutes were spent for observation in each colonoscopy in order to avoid the failure to detect small lesions.  All the adenomatous polyps including mucosal cancers detected in colonoscopy were simultaneously resected by endoscopic polypectomy technique and assessed for pathological examination.  A total of 110 invasive cancers were found and surgically resected during the periods.

       The fasten serum level of two major lipids, TC (Total cholesterol), Low density lipoprotein (LDL), and/or Triglycerides (TG), were measured before the colonoscopy.  The data for TC and TG were available in 4883 and 4874 cases, respectively.  Then, the relationship between the incidence of colonic neoplasm and fasten serum levels of these lipids was retrospectively evaluated.  The statistical analysis was performed with JUMP software.  The serum lipid levels in patients were analyzed with ANOVA and Fisherfs exact test.  The odds ratios were evaluated with multiple logistic regression tests.  P values less than 0.05 were determined as statistically significant.



         In our careful endoscopic observation, invasive cancer was found in 110 (2.3%) in 4887 patients.  All the patients also had the 1-77 adenomatous polyps in other part of the colon.  In the remaining 4777 patients, no adenomas were found only in 956 (19.6%) patients, whereas at least one adenoma can be detected in other 3821 (78.2%) patients, and all of them were endoscopically removed simultaneously.  In these 3931 patients with adenomas (including patients with invasive cancer), only one adenoma was detected in 660 (13.5%) patients, 2 and 3 adenomas were found in 549 (11.2%) and 458 (9.4%) patients, respectively.  In other patients, 1934 (39.6%) patients had from 4 to 9 while other 788 (16.1%) patients had more than 10 adenomas in total colon.  Pathological examinations showed that mucosal cancer were observed in 386 (9.9%) patients.  Most of the resected adenomas, whether associated with cancer or not, were histologically diagnosed as tubular adenoma, and thus tubular adenoma was detected in 3928 (80%) cases in total patients.  Villous adenoma, and serrated adenoma were observed in 53 (1.4%) and 36 (0.9%) cases, respectively (Fig.1).

         At the first step, we analyzed if the serum levels of TC and TG of patients with cancer or adenoma were different from those in patients without these lesions (Fig.2).
As examined with ANOVA, both values were significantly elevated in patients with adenomas as compared with patients without any neoplastic lesions. (TC
207.6}29.5 vs 199.5}34.3, p<0.001; TG 135.0}82.2 vs 108.7}71.5 P<0.001).  Patients with invasive or mucosal cancer showed the same levels of TG (142.0}75.1, 139.3}91.6, respectively) as those with adenomas that were significantly higher than that in patients without adenomas as well as cancers (p<0.001).  In TC, however, patients with mucosal cancer also showed same level as those with adenomas (206.2}28.7), while patients with invasive cancer showed relatively reduced level (200.8}28.3), that was not significantly different from that in patients without neoplastic lesions.  LDL cholesterol was examined in 1483 patients and showed the similar profile as TC .

         Next, we examined the pathological component of these differences of lipid levels in patients with adenomas.

Table 1. Serum levels of Total cholesterol (TC) and Triglyceride (TG) in patients








with tubulatr, villous and serrated adenomas
















Pathological finding



p value




p value

tubular adenoma
























villous adenoma
























serrated adenoma
























As shown in Table 1, the serum levels of these lipids were significantly different in patients with tubular adenomas, whereas not in patients with villous or serrated adenomas.  This indicates that hyperlipidemia has the specific association with the development of tubular adenoma, but not the development of carcinomas as well as other types of adenomas.

         Therefore, we evaluated the association between serum lipid levels and tubular adenomas.  Since TC and TG are known to be closely related with age and gender, we analysed the odds ratios adjusted with these factors. 


Table 2. Univariate analysis of risk for tubular adenoma















Tubular adenoma



negative (%)


positive (%)


p value




















765 (30.9)


1710 (69.1)






194 (8.0)


2218 (92.0)














402 (13.4)


2595 (86.6)






557 (29.5)


1333 (70.5)












Negative: <220mg/dl


695 (21.0)


2623 (79.0)




Positive: =>220mg/dl


263 (16.8)


1302 (83.2)












Negative: <150mg/dl


784 (22.5)


2704 (77.5)




Positive: =>150mg/dl


170 (12.3)


1216 (87.7)



As shown in Table 2, univariate analysis showed that male and old patients had colonic adenomas more frequently than their counterparts.  Then, we separately calculated the odds ratios for the change of 1standard deviation (SD) adjusted with age in male and female patients (Table 3). 

Table 3. Adjusted Odds ratios and 95%CIs for tubular adenoma by 1SD change








in male or female patients

















Total Cholesterol







Total (3787)








Gross OR








Age adjusted OR








Age, Gender adjusted OR








Age, Gender and TC or TG adjusted OR
















Male (2997)








Gross OR








Age adjusted OR








Age and TC or TG adjusted OR
















Female (1890)








Gross OR








Age adjusted OR








Age and TC or TG adjusted OR








Multiple logistic regression analysis showed the clear contrast between TC and TG.  When age and gender were considered, TC lost the relationship with adenoma risk for either male or female subpopulations.  On the contrary, TG was remained as a candidate of an independent risk factor for tubular adenoma in total (p<0.05) and male (p<0.01) patients, but not in female patients. 

         Finally, we evaluated the relationship between the serum lipid level and the number of tubular adenoma (Fig 3).

Interestingly, when the serum levels of TG was elevated, the average of the number of tubular adenoma were gradually increased, and the correlation was statistically significant (p<0.0001). 


Conclusions and Discussions

         Our study clearly indicates that hyperlipidemia is frequently associated with colonic adenoma and carcinoma.  These facts were often reported in previous papers. In this retrospective large scale study, some new interesting facts were discovered additionally.

1) Total cholesterol level was not a specific risk factor for colonic tumors from multivariate analysis, possibly because of the strong association with age.

2) Serum triglyceride level is remained as a candidate of an independent risk factor for colonic tubular adenoma even after multivariate analysis. 

3) And moreover there was a gender difference in the theory 2), which was observed only in male but not in female patients.

4) Villous tumor is not associated with serum lipid level.

5) Serrated tumor is not associated with serum lipid level neither.

6) Total cholesterol level in patients with invasive cancer was rather decreased to the same level as patients without adenoma, while TG levels in patients with invasive cancer kept at the same level as those in patients with adenoma. 

Serum cholesterol was not so linked to colonic tumor as serum triglyceride. The link point of serum TG to colorectal tumor genesis was that of tubular adenoma genesis. And the feature of this linkage was sex dependent. These facts remind us of PPAR-ƒΑ, which diminishes both tumor genesis and serum TG level, and interrupted by estrogen. PPAR-ƒΑmight be playing an important role in colorectal tubular adenoma genesis. To elucidate this speculation, further study will be needed.        

Previous studies have shown the same association between cholesterol level and colonic villous tumors7. But the other paper17 showed the inverse result. In the later paper the number of tumors was so small compared with our study and the odds were just on the threshold. So we think that there is no relationship between villous component and serum lipid level.

About relationship between serum lipids and serrated histological component, there were no previous reports as far as we researched.



         1. Reddy BS. Dietary fat and its relationship to large bowel cancer. Cancer Res 1981;41(9 Pt 2):3700-5. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=6266660.

         2. Vogel VG, McPherson RS. Dietary epidemiology of colon cancer. Hematol Oncol Clin North Am 1989;3(1):35-63. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=2537285.

         3. Potter JD. Risk factors for colon neoplasia--epidemiology and biology. Eur J Cancer 1995;31A(7-8):1033-8. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=7576987.

         4. Reddy BS, Wynder EL. Metabolic epidemiology of colon cancer. Fecal bile acids and neutral sterols in colon cancer patients and patients with adenomatous polyps. Cancer 1977;39(6):2533-9. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=872053.

         5. Neugut AI, Johnsen CM, Fink DJ. Serum cholesterol levels in adenomatous polyps and cancer of the colon. A case-control study. Jama 1986;255(3):365-7. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3941515.

         6. Stemmermann GN, Heilbrun LK, Nomura AM. Association of diet and other factors with adenomatous polyps of the large bowel: a prospective autopsy study. Am J Clin Nutr 1988;47(2):312-7. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3341261.

         7. Mannes GA, Maier A, Thieme C, Wiebecke B, Paumgartner G. Relation between the frequency of colorectal adenoma and the serum cholesterol level. N Engl J Med 1986;315(26):1634-8. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3785334.

         8. Correa P, Strong JP, Johnson WD, Pizzolato P, Haenszel W. Atherosclerosis and polyps of the colon. Quantification of precursors of coronary heart disease and colon cancer. J Chronic Dis 1982;35(5):313-20. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=7068807.

         9. McKeown-Eyssen G. Epidemiology of colorectal cancer revisited: are serum triglycerides and/or plasma glucose associated with risk? Cancer Epidemiol Biomarkers Prev 1994;3(8):687-95. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=7881343.

         10. Giovannucci E. Insulin and colon cancer. Cancer Causes Control 1995;6(2):164-79. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=7749056.

         11. Bruce WR, Wolever TM, Giacca A. Mechanisms linking diet and colorectal cancer: the possible role of insulin resistance. Nutr Cancer 2000;37(1):19-26. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=10965515.

         12. Yamada K, Araki S, Tamura M, Sakai I, Takahashi Y, Kashihara H, et al. Relation of serum total cholesterol, serum triglycerides and fasting plasma glucose to colorectal carcinoma in situ. Int J Epidemiol 1998;27(5):794-8. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=9839735.

         13. Koohestani N, Chia MC, Pham NA, Tran TT, Minkin S, Wolever TM, et al. Aberrant crypt focus promotion and glucose intolerance: correlation in the rat across diets differing in fat, n-3 fatty acids and energy. Carcinogenesis 1998;19(9):1679-84. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=9771941.

         14. Saydah SH, Platz EA, Rifai N, Pollak MN, Brancati FL, Helzlsouer KJ. Association of markers of insulin and glucose control with subsequent colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 2003;12(5):412-8. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=12750235.

         15. Gaard M, Tretli S, Urdal P. Blood lipid and lipoprotein levels and the risk of cancer of the colon and rectum. A prospective study of 62,173 Norwegian men and women. Scand J Gastroenterol 1997;32(2):162-8. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=9051877.

         16. Barton TP, Cruse JP, Lewin MR. Changes in serum lipids related to the presence of experimental colon cancer. Br J Cancer 1987;56(4):451-4. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3689662.

         17. Houghton J, Lardieri GG, Zauber NP, Kim KH, Cable G. Effect of cholesterol levels on villous histology in colonic adenomas. Dig Dis Sci 2000;45(5):896-9. Available from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=10795751.