Epidemiology of Colorectal Cancer Revisited: Are Serum … · served between exercise and the risk...
Transcript of Epidemiology of Colorectal Cancer Revisited: Are Serum … · served between exercise and the risk...
Gail McKeown-Eyssen’
Department of Preventive Medicine and Biostatistics, University of
Toronto, 4th Floor, McMurrich Building, 1 2 Queen’s Park Crescent West,
Toronto, Ontario, Canada M5S 1A8
Received 5/27/94; revised 9/14/94; accepted 9/14/94.
1 To whom requests for reprints should be addressed, at University of
Toronto, Department of Preventive Medicine and Biostatistics, 4th Floor,McMurrich Building, 1 2 Queen’s Crescent West, Toronto, Ontario, CanadaM5S 1A8.
Vol. 3, 687- 695, December 1994 Cancer Epidemiology, Biomarkers & Prevention 687
Review
Epidemiology of Colorectal Cancer Revisited: Are SerumTriglycerides and/or Plasma Glucose Associated with Risk?
Abstract
Several aspeds of Western diets, alcohol use, andexercise patterns which are related to the risk ofcoloredal cancer have systemic effeds in common.Those which increase the risk of coloredal cancer arepositively associated with serum triglycerides and plasmaglucose; those which decrease risk are negativelyassociated with serum triglycerides and plasma glucose.These observations suggest the hypothesis that serumtriglycerides and/or plasma glucose may themselves beassociated with coloredal cancer risk. Evidence forassociations between colorectal neoplasia andtriglycerides and glucose comes from two recent studiesof adenomatous polyps, presumed precursors forcolorectal cancer, and from previous studies of diabetesand cancer. In addition, three randomized trials, one inhumans and two in animal models, suggest that dietswhich would be expeded to increase serum triglyceridesand plasma glucose increase the levels of cellularindicators of coloredal cancer risk.
Biological mechanisms explaining associationsbetween colorectal neoplasia and serum triglyceridesand/or plasma glucose might involve luminal orcirculatory effeds: (a) triglycerides and/or glucose maybe associated with fecal bile acids, acids which arepositively associated with coloredal cancer risk inepidemiological studies and which promote colorectalcancer in animal models; (b) serum triglycerides and/orplasma glucose might influence circulating hormones,such as insulin, which might themselves be involved incancer development; (C) serum triglycerides and/orplasma glucose might be indicators of energy availablethrough the circulation for neoplastic cells. Futureresearch should examine the associations between therisk of coloredal neoplasia and serum triglycerides andplasma glucose, explore possible biological mechanisms,and develop an understanding of whether biologicalprocesses involving triglycerides, glucose, and insulinmay be common to a number of sites of cancer.
Introdudion
Evidence from epiderniological and animal studies has sug-
gested that dietary factors and related aspects of lifestyle areinvolved in the etiology of colorectal cancer. Western diets,typically dense in energy and fat and low in fiber, have been
associated with increased risk, and vegetarian diets, or in-
creased intake in a mixed diet of vegetables, fruits or otherplant foods, especially those high in fiber, have been asso-ciated with reduced risk (1 -3). In addition, there is evidence(3) of positive associations between risk and consumptionof alcohol, (especially beer) and body size, although asso-ciations are less well established than for Western eatingpatterns. In contrast, exercise is consistently associated withreduced risk (3), and there is some evidence that consump-tion of fish oil may reduce risk (1 , 2). These dietary and
lifestyle risk factors for colorectal cancer have two systemiceffects in common. Those which increase the risk of cob-
rectal cancer are positively associated with serum triglyc-erides and plasma glucose; those which decrease risk are
negatively associated with serum triglycerides and plasmaglucose. Is it possible, then, that serum triglycerides and/orplasma glucose may themselves be associated with the riskof colorectal cancer? This paper presents epidemiobogical
evidence for such associations. It also considers possiblebiological mechanisms. Study of biological mechanisms forcoborectal cancer (1 , 2) has concentrated on luminal factorssuch as fecal bile acid concentrations, stool bulk, and tran-sit time. This paper suggests that circulatory effects mightalso be important.
Risk Fadors for Coloredal Cancer
The extensive data from epidemiologic studies and fromanimal models suggesting that factors in Western diets may
be related to the risk of colorectal cancer have been re-viewed by expert committees of the National Academy ofScience (1 , 2). Positive associations with total energy, andwith each of the energy-containing nutrients, fat, protein,and carbohydrate, have been reported in case-control andcohort studies (1-5). Negative associations have been ob-served with consumption of vegetables, fruits, and cereals(1-5), with possible protective mechanisms involving fiber,
vitamins, and other components of these plant foods (3, 6).The accumulated evidence is of sufficient strength that sci-
entific bodies and governments in North America and else-where (1 , 2, 7-1 0) have advocated reduced consumption of
fat and increased consumption of fruits, vegetables, and
cereals. Evidence for other risk factors for coborectal canceris accumulating. A positive association with consumptionof alcohol (especially beer) has been observed in geo-graphic correlation studies, two of four population cohortstudies, and seven of fourteen case-control studies (3). As-
sociations with body size and coborectal cancer risk havealso not been entirely consistent and appear to differ some-
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
688 Review: Triglycerides, Glucose, and Colorectal Cancer
2 The abbreviations used are: VLDL, very low density lipoprotein; RR,
relative risk; SMR, standardized mortality ratio; OR, odds ratio.
what with the measure of body size examined. A positive
association with body mass has been suggested by twoprospective studies (1 1 , 1 2) and by the findings of themajority, but not all, ofthe case-control studies reviewed by
Potter et a!. (3). In addition, height, weight, and Quetelet’sIndex were reported to be positively associated with cob-
rectal cancer recurrence (1 3). The single investigationwhich examined the distribution of body fat (1 1 ) reported asmall but not statistically significant increase in risk only inthe highest qumntile of measured waist to hip ratio. Three
studies have reported positive associations with increasedstature, independent of weight (1 1 , 1 4, 1 5). A protective
effect of fish oil has been suggested by the observation thatEskimos with diets rich in fish oils have low rates of cancer,including colorectal cancer. This possibility is supported in
animal models where administration of fish oil not only
failed to stimulate tumour growth as would be expectedwith high fat diets, but might even have an inhibiting effect
(2, 16). A negative association has been consistently ob-served between exercise and the risk of colon cancer (3),when activity has been assessed in association with both
occupation and leisure, but findings have been less consis-
tent for colorectal cancer or for rectal cancer alone (3).
Common Risk Factors for Serum Triglycerides, PlasmaGlucose, and Coloredal Cancer
The relationship between lifestyle risk factors and serumtriglycerides and plasma glucose has been studied in the
context of research into heart disease and diabetes, ratherthan in studies of cancer. However, the findings suggest that
a number of factors which are positively related to the riskof colorectal cancer are also positively related to serumtriglycerides and plasma glucose; other factors negativelyrelated to colorectal cancer risk are negatively related to
these physiological characteristics.Comparison of persons on mixed Western diets with
those on vegetarian diets has shown that those on Westerndiets have increased serum triglycerides and plasma glu-cose. Higher levels of serum triglycerides and blood glu-cose have been observed in populations that consume dietshigh in fat and low in complex carbohydrates (such as those
in the United Kingdom and among African whites) than arefound in populations which habitually consume dietswhich are low in fat and high in complex carbohydrates(such as those in Southern Africa and in Japan) (1 7-20).
Diets high in fat and energy density have been shown infeeding trials to increase triglycerides when compared withlow fat, low energy dense diets (21-23). High simple car-bohydrate diets have long been recognized as associatedwith hypertriglyceridemia but diets which are high in fiber-rich carbohydrates do not induce hypertniglycenidemia
when fed over prolonged periods (1 7, 24, 25). Indeed,feeding trials have shown that the more complex the car-bohydrate source, the less the rise in triglycerides associ-
ated with the inclusion of high levels of simple carbohy-
drates, and that fiber is often, though not always, associatedwith reduced serum triglycerides (26-35). The effects ofindividual foods on plasma glucose have also been exam-med in feeding studies. A glycemic index has been devel-
oped which describes the postprandial glycemic responseof individual foods compared to that of a standard glucose
stimulus (1 7, 26). Plant foods, especially legumes, tend toproduce low glycemic responses (26, 29, 33, 36, 37).
Alcohol, like carbohydrate, has long been recognized
as increasing serum triglycerides (38), and beer is a sourceof both alcohol and carbohydrates (about 30 g/liter) (38).Alcohol intake has also been positively associated withhyperglycaemia (39), although the mechanism is unclean
(38). Alcohol after meals, but not in a fasting state, results inhyperglycaemia in humans and animals (38). Thus, alcoholincreases serum triglycerides and plasma glucose, espe-
cially when consumed in a nonfasting state.Positive associations between obesity, serum tniglycer-
ides, and blood glucose are well established (40-42). Clin-
ical studies have shown that the degree of obesity is asso-
ciated with production of VLDL2 triglyceride and elevatedlevels of blood glucose, and that increases in weight areassociated with increases in triglycerides and glucose (42).Further, the regional distribution of body fat can influence
both serum triglycerides and blood glucose (40-46).In normal subjects, dietary supplementation with fish
oil reduces serum triglycerides (47-49) without adverseeffects on glucose, and in patients with hypertrigbyceni-
demia it may also reduce glucose (50). However, in patientswith noninsulin-dependent diabetes, fish oil supplementa-tion does not always reduce triglycerides (51 ) and mayincrease blood glucose levels (52-56). In nondiabetics, di-
etary enrichment with w-3 fatty acids may increase theirincorporation into pancreatic f3 cell phospholipid mem-
branes, enhancing insulin secretion and reducing carbohy-drate intolerance (52).
Evidence suggests that exercise reduces serum triglyc-erides and improves glucose tolerance (57, 58). Studies
comparing lipids in persons with different levels of physicalactivity have found higher serum triglycerides in inactivepeople (59-64). Studies of physical training have foundreductions in triglycerides (65-67), together with increasesin glucose tolerance (58).
Thus, factors which may be positively associated withcolorectal cancer, Western eating patterns, alcohol use, andobesity are also positively associated with serum tniglycer-ides and plasma glucose, and factors which may be nega-tively associated with colorectal cancer, diets which
emphasize fruits and vegetables, fish oil consumption, andexercise, are associated both with reduced colorectal can-
cer risk and with reduced serum triglycerides and plasmaglucose. Such associations suggest the possibility of directassociations between colorectal cancer and serum tniglyc-erides and/or plasma glucose.
Associations of Serum Triglycerides and GlucoseMetabolism with Colorectal Cancer
The relationship between colorectal cancer and serum trig-lycenides or plasma glucose has not been studied directly.Rates of colorectal cancer and coronary heart disease are
correlated internationally (r = 0.77) (68), so one mightexpect to find international correlations between colorectal
cancer and serum lipids and/or glucose as they are riskfactors for heart disease (69-72). However, no studies have
examined these relationships, perhaps because reliable fast-ing blood data are not available at the national level (73).Information beaning on the relationship between cobonectalneoplasia and triglyceride and glucose metabolism comes
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 689
3 W. R. Bruce, personal communication.
only indirectly from studies which were conducted withother primary objectives.
Diets high in simple carbohydrates and low in fiber
increase serum triglycerides and plasma glucose, so it might
be expected that such diets would be associated with in-creased risk of colorectal neoplasia. Associations between
colorectal cancer and sucrose have been in the direction ofincreased risk, though not always statistically significant.Two cohort studies (1 1 , 74) reported close to a 2-fold
increase in risk comparing groups with highest to lowestintake of all nondairy sucrose-containing foods (RR = 2.0, Pfor trend = 0.03) (1 1 ), or with pie and cake (RR = 1 .7, Pfortrend > 0.05) (74). Nine of 1 0 case-control studies reviewed
by Bostick et a!. (1 1 ) reported relative risks between 1 .0 and2.3 with intake of sucrose or sucrose-containing foods,
although only 3 reached statistical significance at the 5%level; a single study reported a nonsignificant reduction in
risk (RR = 0.7, P for trend > 0.05) (75). Three randomizedtrials of diets high in sucrose and low in dietary fiber, one inhuman volunteers (76) and two in animal models (77, 78),have supported a positive association by observing associ-
ations with cobonic cell proliferation and/or rates of growthof dysplastic crypt foci, indicators of risk of neoplasia (79-
87). Slavin eta!. (76) found that consumption ofa liquid diethigh in simple carbohydrate was associated with higherrates of rectal cell proliferation in each of 5 crypt compart-ments than were seen when healthy volunteers consumed
their normal diets, or when the liquid diet was supple-mented with breads containing cereal or vegetable fiber.The finding that total cell proliferation was highest on the
fiber-free diet (P < 0.03) is compatible with the ability,
discussed above, of dietary fiber to counteract the increasein serum triglycerides produced by intake of simple carbo-hydrates (27, 28). Using the azoxymethane (AOM) rat
model, Stamp et a!. (77) observed increased colonectal cellproliferation and aberrant crypt foci after gastric gavage
with sucrose ard fructose compared to controls who re-ceived a water gavage (4.1 rnitotic cells/crypt with sucrose
compared to 0.4 with water, P< 0.001 ; 31 .4 aberrant cryptfoci/colon with sucrose compared to 7.2 with water, P <
0.001 ). Finally, Caderni et a!. (78) have compared rats fedfor 1 05 days with a simple carbohydrate (sucrose) with rats
fed a complex carbohydrate (starch). Colonic cell prolifer-ation and size of dysplastic crypt foci were greatest in the
sucrose fed rats (mean labeling index: 1 0.4 with sucrosecompared to 4.4 with starch, P < 0.001 ; size of dysplasticcrypt foci 2.9 dysplastic crypts/focus with sucrose corn-
pared to 2.6 with starch, P < 0.05). These studies demon-strate that diets expected to increase serum triglycerides andplasma glucose are associated with indicators of risk ofneoplasia in the colon but, because serum triglycerides andglucose were not assessed in the investigations, direct as-sociations have still to be demonstrated. Further investiga-tions would be needed to determine whether the dietsaffected the colon through systemic effects involving the
circulation and/on through luminal effects.Direct evidence of an association between serum trig-
lycenides and presumed precursors of colorectal cancer,adenomatous cobonectal polyps (88, 89), has been observedin studies conducted in Toronto (90-92). As part of a ran-
domized trial designed to examine the effects of manipu-bating colonic fermentation on markers of colorectal cancerrisk, Kashtan eta!. (90) measured pretrial blood lipids in 45
patients with a history of colorectal adenomatous polypsand in 49 controls who were free of polyps on colonoscopy.
Average serum triglyceride levels were 1 7% higher (P =
0.06) in polyp patients than in controls.3 In patients withadenomatous polyps who participated in a randomized trial
examining the effect on polyp recurrence of a low fat, high,
fiber diet (91 ), associations were observed between serum
triglycerides assessed after 1 year ofdiet counseling and therate of polyp recurrence (92, 93). Associations were
i-shaped, with polyp recurrence increasing steadily fromthe second to the fourth quartiles of serum triglycerides
(recurrence rates in quartiles 1 to 4 were 31.6, 15.0, 20.0,and 25.0%, P> 0.2 for men; and 8.3, 7.1 , 1 4.3, and 30.8%,
P = 0.1 5 for women). These findings suggest associationsbut lack power because of small sample sizes (n = 79 for
men; n = 58 for women); larger studies would therefore beneeded to confirm them and to examine the possibility of
nonlinear associations. Clues to the biology of associationsare provided by the observation in the same trial (91 ) of
positive correlations (r= 0.42, P< 0.001 in men; r= 0.21,P> 0.05 in women) between serum triglycerides measured
after 12 months on study and fecal bile acids, acids whichare related to the risk of coborectal cancer in human studies
and which promote colorectal neoplasia in animal models
(94-100). In addition, it is of interest that fecal bile acids
after one year of diet counselling exhibited i-shaped pat-terns of association with the rate of recurrence of adeno-matous polyps similar to those observed for triglycerides(recurrence rates in quartiles 1 to 4 were 1 6.7, 1 1 .1 , 22.2,
and 33.3%, P> 0.2 for men; and 7.7, 7.7, 21 .4, and 30.8%,P = 0.06 for women). These findings suggest that factors inthe circulation and in the lumen may both be associatedwith polyp recurrence but the biological mechanisms for
these associations have still to be determined. That serumtriglycerides and fecal bile acids may themselves be bio-logically associated has recently been postulated by Ange-lin and Einarson (101) as described below.
Direct evidence of an association between plasma glu-cose and cancer comes from studies early this century when
patients with various sites of cancer were observed to havehigher plasma glucose than were healthy individuals (102-1 07). Indeed, in 1 91 9 it was even proposed that blood sugar
be used to screen for cancer (107). Recently, two cohortstudies designed to study heart disease have reported asso-ciations between glucose assessed after glucose challengeand mortality from colorectal cancer 1 2-20 years later(108, 109), but both were able to observe only small num-bers of colorectal cancer deaths. The Chicago Heart Asso-
ciation Detection Project in Industry (108) concluded thatwomen who died of colon, but not rectal, cancers had
higher mean plasma glucose than did survivors (meanplasma glucose, 1 56.1 in colon cancer decedents and
140.0 in survivors). They also reported that mean plasmaglucose levels were higher among men who died of colon
(plasma glucose, 1 45.9) and rectal cancers (plasma glucose,1 46.9) than among survivors (plasma glucose, 1 44.5). How-
ever, no statistical tests were reported for individual cancersites. In the Whitehall Study (109), nondiabetic men expe-rienced a small but nonsignificant increase in risk associ-ated with blood glucose (RR = 1 .03 for an increase of 1 SDin glucose; 95% confidence interval, 0.88-i .24) but exhib-
ited no increase in risk of rectal cancer (RR = 0.94 for
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
690 Review: Triglycerides, Glucose, and Colorectal Cancer
Common EnvironmentalRisk Factors and/orBiologic Pathways
Faecal Bile Acids >. Colorectal
� Serum Triglycenides/ CancerPlasma Glucose
Fig. 1. A model illustrating factors in the colonic lumen, which may link environmental factors to colorectal cancer risk.
an increase of 1 SD in glucose; 95% confidence interval,0.7-1 .27).
Diabetics have high levels of plasma glucose and se-
rum triglycerides so studies of the association of diabetesand colonectal cancer might provide relevant information.A study of death certificates in Massachusetts (1 1 0) reported
that diabetes and colorectal cancer occurred together ondeath certificates more frequently than expected (SMR =
1 .23 for men and 1 .08 for women) but reported no tests ofsignificance. Two case-control studies (1 1 1 , 1 1 2) reported
associations between colorectal cancer and diabetes.Clicksrnan and Rawson (1 1 1 ) observed an excess of abnon-mal glucose tolerance tests in coborectal cancer patientscompared with controls with benign diseases (OR = 2.4) butprovided no statistical evaluation. O’Mara et a!. (1 1 2) ob-served elevated but nonsignificant (P> 0.05) odds ratios forhistory of diabetes among 27 cases with colon cancer (OR= 1 .2 for men and 1 .4 for women) but reported lower odds
ratios for 26 cases of rectal cancer (OR = 1 .1 for men and1 .0 for women). Cohort studies of diabetics which haveattempted analysis of individual cancers have been able to
observe only small numbers of cobonectal cancers. Adami eta!. (1 1 3) observed nonsignificant (P > 0.05) increases formale colon (RR = 1 .2) and rectal cancers (RR = 1 .3) basedon 94 and 73 cases, respectively, but observed no excess forfemale bowel cancers. Ragozzino et a!. (1 1 4) reported arelative risk of 1 .3 (95% confidence interval, 0.8-i .9) forcoborectal cancer based on 22 cases. Kessler (1 1 5) observedan SMR of 1 .09 (P > 0.3) for colon cancer based on 147cases but found no increase for rectal cancer (SMR = 0.81;
P > 0.1) based on 45 cases. Smith et a!. (109) found no
excess of colon cancer (RR = 0.62; 95% confidence inter-val, 0.09-4.47) among diabetics based on 1 case of colon
cancer and no excess among men with impaired glucosetolerance (RR = 0.81 ; 95% confidence interval, 0.35-1 .83)
based on 6 cases. Considering the high cardiovascular mor-tality and reduced expectation of life in diabetics (1 1 6), it is
perhaps surprising that any excess of cancer mortality wasreported in these studies. However, the small numbers of
cases and other methodological problems associated withstudies of individual cancer sites (102) suggest that furtherdata will be needed before firm conclusions can be drawn.
In summary, data which have bearing on the relation-
ship between colorectal neoplasia and serum triglyceridesor plasma glucose are generally compatible with the hy-pothesis of a positive association. However, studies areneeded to establish associations directly and to examinetheir biological basis.
Possible Biological Mechanisms
The risk of colorectal neoplasia has usually been thought to
be associated with factors affecting the colonic lumen. Inparticular, fecal bile acids have been shown to promotecolorectal neoplasia in animal models and have been pos-itively related to the risk of coborectal cancer in a number of
epidemiobogical studies (94-100). An association between
serum triglycerides and colorectal neoplasia might there-
fore be observed if serum triglycerides were biologically
associated with fecal bile acids. Such an association has
recently been suggested by Angelin and Einarson (101),
who have proposed a link between perturbations of bile
acid enterohepatic circulation and VLDL triglyceride pro-
duction. Because VLDL triglycerides constitute the majority
of total fasting triglycerides, such an association could ac-
count for the positive correlations between total serum
triglycerides and fecal bile acids, and between serum trig-
lycerides and polyps, described above. Further, an obser-
vation of higher bile acid concentrations in healthy volun-teens fed 165 g of refined sugar daily compared with
volunteers who consumed 60 g of sugar (1 1 7) suggests that
the possibility of an association between bile acids andplasma glucose should also be explored, since a high sugar
diet should increase plasma glucose as well as triglycerides.
The relationships of serum triglycerides and plasma glucose
to bile acid metabolism therefore warrants further study.
However, it is also possible that the risk of colorectal
neoplasia is directly related to factors in the circulation.
Two biological mechanisms might explain associations be-
tween coborectal neoplasia and serum triglycerides and
plasma glucose.
First, hormones may play a key role. Increased blood
glucose increases insulin secretion and increased tniglycer-ides are secondary to hypeninsulinemia and insulin nesis-
tance (1 1 8). Thus, increased triglycerides and glucose could
be indicators of increased insulin levels. The possible role
of insulin in colorectal neoplasia is suggested by severallines ofevidence. Insulin is considered to be a growth factor
(1 1 9), and insulin receptors are observed in both normal
and malignant colorectal cells (1 20). Insulin receptors can
be bound by insulin-like growth factors, factors which areexpressed by colorectal cancers (121). Insulin has been
shown to increase insulin-like growth factor I in rats (1 22),and a binding protein for this growth factor inhibits the
growth of colon cancer cells in vitro (123). In addition, a
recent case-control study reported a strong direct associa-tion between insulin and breast cancer (1 24), a site which
shares common epidemiobogical risk factors with coloncancer (1 , 2). An effect of insulin in carcinogenesis is further
suggested by the observation that exogenous insulin in-
duces liver carcinogenesis in hypophysectomized rats fed
an otherwise ineffective dose of carcinogen (1 25), as well asby the finding that insulin is capable of damaging DNA in
endothelial cells (126). Further, insulin is negatively asso-ciated with somatostatin (1 27), a hormone with atrophic
effects (1 28) which has been shown to inhibit mouse colonadenocarcinoma (1 29) and human pancreatic tumors (1 30).
Because fiber reduces the insulin response (1 31 , 1 32), and
long-term ingestion of fructose can affect cellular insulin
binding and sensitivity (1 33), it is possible that dietary as-
sociations with colonectal cancer could arise through effectson insulin. The interrelationships among insulin and related
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 691
Common Environmental Faecal Bile AcidsRisk Factors and/or < Serum Triglycenides/ � Metabolism of � ColorectalBiologic Pathways
Plasma Glucose Neoplastic Cells Cancer
lnsulin/other Hormones
/Fig. 2. A model illustrating factors in the circulation, which may link environmental factors to colorectal cancer risk.
hormones, serum triglycerides, plasma glucose, foods, and
colorectal neoplasia require further study.
Second, an association between coborectal neoplasia
and circulating levels of serum triglycerides and/or plasmaglucose could arise if triglycerides or glucose was an mdi-
cator of energy available for neoplastic cells.Normal cells of the colonic epitheliurn are capable of
both aerobic and anaerobic metabolism, using short chainfatty acids, amino acids, and glucose as fuel (1 34). How-ever, short chain fatty acids are the predominant energysource (135, 136). lndeed, availability of the short chainfatty acid, butyrate, as a fuel suppresses glucose utilization,
especially in the left side of the bowel (1 36). Butynate maybe an especially important source of energy because it hasregulatory effects on nucleic acid metabolism (137, 138)and acts as a differentiating agent (1 39), so it may assist inmaintaining the health of the large bowel epitheliurn (135).
In contrast to the normal colonic epithelium, neoplas-
tic cells switch from aerobic to anaerobic metabolism (140,141), and anaerobic metabolism uses glucose rather thanfatty acids for fuel. Thus, plasma glucose might be associ-
ated with colorectal neoplasia by acting as direct source ofenergy for neoplastic cells. In contrast to plasma glucose,the fats transported by serum triglycerides could not provide
energy for neoplastic cobonic cells. However, fasting serumtriglycerides may be a general marker of energy availability
because VLDL triglycerides (the major component of serumtriglycerides) represent the triglycerides which are being
sent from the liver to the adipose tissues for long-termstorage, as well as to peripheral tissues for energy.
Because diets which increase plasma glucose are thosewhich tend to be low in plant food, the high levels ofcolorectal cancer in the Western world might anise from thelow levels of plant food in the Western diet and fromcorresponding increases in availability of plasma glucose asan energy source for neoplastic cells. Further, there is apreponderance of left-sided colon cancer in Western pop-ulations. This might be explained by the increase of glucoseavailability to the area of the colon which is the mostvasculanized (the left side), and by depletion of butyrate
formed from fermentation of fiber from plant foods, a fuelon which the left side ofthe colon is especially reliant (1 36).
Thus, several biological mechanisms involving eitherluminal or circulatory paths could account for associations
between the risk of colonectal neoplasia and serum tniglyc-
erides and plasma glucose.
Research Implications
A first step in examining the hypothesis that serum tniglyc-erides and/or plasma glucose may be associated with therisk of colorectal cancer is to establish directly that an
association exists between these blood characteristics and
colonectal neoplasia. Observational epidemiobogical inves-
tigations could establish such associations. However, to
ensure that either triglycerides or glucose was not merelycorrelated with some ultimate biological causes of malig-
nant disease which had not yet been recognized, human
trials would need to demonstrate that reduction in levels of
serum triglycerides or plasma glucose could reduce the risk
of colorectal neoplasia, just as experimental studies were
needed to examine the possibility of a direct association
between serum cholesterol and coronary heart disease.
Understanding the biology of associations between
triglycerides, glucose, and coborectal neoplasia would re-
quire studies to determine the relative importance of lumi-
nal and circulatory factors. If fecal bile acids or other lumi-
nal factors were the major determinant of colorectal
neoplasia (Fig. 1), associations between coborectal neopla-
sia and circulatory factors would occur only because thecirculatory and Iuminal factors shared common environ-
mental risk factors or common biological pathways; no
direct biological association would exist between coborec-tab neoplasia and serum triglycerides and/or plasma glu-
cose. In contrast, serum triglycerides and/or plasma glucose
could be the major determinant of colorectal neoplasia,
acting directly on the metabolism of neoplastic cells or
through insulin or other hormones. In this case, lurninalfactors would have no direct biological association with
colonectal neoplasia; associations would occur only be-cause fecal bile acids shared common environmental or
biological determinants with serum triglycerides and/orplasma glucose (Fig. 2). Distinguishing the relative contri-butions of these models of colorectal neoplasia will be thenext major challenge.
Understanding the development of colorectal cancer,and its relationship to pobyps, will ultimately require under-standing the relationship between physiological mecha-nisms and the genetics of neoplasia. Fearon and Vogelstein(1 42) have postulated a sequence of genetic changes whichoccur as polyps develop through the proposed adenoma-
carcinoma sequence. It would be of considerable interest todetermine whether these genetic changes are associatedwith the changes in enzyme profiles involved in the switchfrom aerobic to anaerobic metabolism described by Jass as
associated with increasing dysplasia in polyps (1 40, 1 41 ). Itwould also be of interest to examine the relationship be-
tween the genetic changes and the ability of the cell torespond to insulin and related growth factors.
The possibility that factors in the circulation may beassociated with the risk of colorectal cancer is appealingbecause it could bead to research which might explain thecorrelations between coborectal cancer and cancers of the
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
692 Review: Triglycerides, Glucose, and Coloredal Cancer
breast, endometrium, ovary, and prostate, sites which share
some risk factors (1 , 2) and which might be influenced bybiological mechanisms affecting proliferation such as thosedescribed for coborectal cancer. Because serum tniglycer-ides and plasma glucose have also been positively associ-
ated with cardiovascular disease (69-72), these factorsmight account for associations between colorectal cancer
and heart disease (68). However, much remains to be un-derstood about the relationship between colorectal cancerand cardiovascular disease. Because circulatory risk factorsfor cardiovascular disease, cholesterol, and other blood
components share some dietary and lifestyle risk factorswith serum triglycerides and plasma glucose, a positive
association might be expected between serum cholesteroland the risk of coborectal cancer. In the 1 980s, however,
some studies reported increased colorectal cancer risk as-sociated with low cholesterol levels (143-146). In 1990, aconference was held to evaluate the evidence on associa-tions between cholesterol and a number of diseases, includ-ing large bowel cancer, from a meta-analysis of 19 of 20cohort studies with relevant data (147). Because it wasrecognized that a negative association would be observed if
malignancy caused a drop in cholesterol in the years beforediagnosis, this analysis considered only deaths occurring atleast 5 years after baseline cholesterol measurement. Theanalysis found that there was little evidence of a gradient of
risk of colon cancer across cholesterol classes, althoughthere was some evidence of variability among studies. Theauthors recognized that information was needed about thetime course of cholesterol change in disease, and there is
still a need for studies which can distinguish cleanly be-tween the biological effects of early disease and the effectsof factors which influence tumor growth in the period be-fore diagnosis. While the meta-analysis of the relationshipbetween cholesterol and colon cancer took account of anumber of demographic and lifestyle-confounding van-ables, it was not able to consider the effects of other bloodlipids or measures of glucose metabolism. Reduction in the
risk of colonectal and other cancers will require determina-
tion of optimal levels of blood lipids, glucose, and insulin,in combination as well as singly. In this context, the syn-
drome of insulin resistance is of interest because it is amultifaceted syndrome with a combination of dyslipidemia
and hypeninsulinemia and it has been associated with ath-erosclerotic cardiovascular disease, noninsulin-dependentdiabetes rnellitus, hypertension, and obesity (148). Furtherresearch appears justified to explore the relationship ofcancers to the abnormal lipids and increased levels of in-
sulin in the insulin-resistance syndrome.Epiderniological research has made great progress in
identifying dietary and other lifestyle risk factors associated
with colonectal cancer and other malignancies. Future re-search will require collaboration between basic scientists
and epidemiologists in order to determine the relative im-portance of luminal and circulatory factors in colorectal
neoplasia, to develop an understanding of the effects of
dietary and lifestyle risk factors on the relevant biologicaland metabolic processes, to link these models with anunderstanding of the genetics of neoplastic cells, and todevelop an understanding of biological processes whichmay be common to a number of sites of cancer.
AcknowledgmentsI am most grateful to the colleagues who have provided useful comments onthe ideas expressed in this papei: Drs. Andrew Baines; Cornelia Baines;
Robert Bruce; Gordon Greenberg; David Jenkins; Gary Kakis; John Potter;
and Hartley Stern.
References1 . Committee on Diet and Health. Diet and Health: Implications for Reduc-ing Chronic Disease Risk. Washington DC: National Academy Press, 1989.
2. Committee on Diet, Nutrition and Cancer. Diet, Nutrition and Cancer.
Washington DC: National Academy Press, 1982.
3. Potter, J. D., Slattery, M. L., Bostick, R. M., and Gapstur, S. M. Colon
cancer: a review of the epidemiology. Epidemiol. Rev., 15: 499-545, 1993.
4. WiIIett, W. C., Stampfer, M. J., Colditz, G. A., Rosner, B. A., and Speizer,F. E. Relation of meat, fat and fibre intake to the risk of colon cancer in aprospective study among women. N. EngI. J. Med., 323: 1 664-1 672, 1990.
5. Giovannucci, E., Stampfer, M. J., Colditz, G., Rimm, E. B., and WiIlett,
W. C. Relation of diet to the risk of colorectal adenoma in men. J. Natl.Cancer Inst., 15: 74-75, 1992.
6. Steinmetz, K. A., and Potter, J. D. Vegetables, fruit and cancer. II Mech-
anisms. Cancer Causes Control, 2: 427-442, 1991.
7. Policy Statement on Prevention of Colorectal Cancer by Diet. Toronto:
Canadian Cancer Society, 1983.
8. American Cancer Society. New Dietary Guidelines. New York: American
Cancer Society News Service, 1984.
9. European Organization for Cooperation on Cancer Prevention Studies
and the International Union of Nutritional Sciences. Consensus Statement onProvisional Dietary Guidelines. Nutr. Cancer, 8: 39-40, 1986.
1 0. Statement of the National Cancer Institute. Office of Cancer Communi-
cations, Diet and Cancer, 1986.
1 1 . Bostick, R. M., Potter, J. D., Kushi, L. Fl., Sellers, T. A., Steinmetz, K. A.,McKenzie, D. R., G�pstur, S. M., and Folsom, A. R. Sugar, meat and fatintake, and nondietary risk factors for colon cancer incidence in Iowa
women (United States). Cancer Causes Control, 5: 38-52, 1994.
1 2. Lew, E. A., and Garfinkel, L. Variations in mortality by weight among
750,000 men and women. J. Chron. Dis., 32: 563-576, 1979.
1 3. Tartter, P. I., Slater, C., Paptestas, G., and Aufses, A. H. Cholesterol,weight, height, Quetelet’s index and colon cancer recurrence. J. Surg.
Oncol., 27:232-235, 1984.
14. Albanes, D., and Taylor, P. R. International differences in body heightand weight and their relationship to cancer incidence. NuIr. Cancer, 14:
69-77, 1990.
1 5. Albanes, D., Jones, D. Y., Schatzkin, A., and Micozzi, M. S. Adult stature
and risk of cancer. Cancer Res., 48: 1658-1662, 1988.
1 6. Carroll, K. K. Biologic effects of fish oils in relation to chronic diseases.
Lipids, 21: 731-732, 1986.
1 7. Mann, J. Complex carbohydrates: replacement of energy for fat or useful
in their own right? Am. J. Clin. Nutr., 45: 1202-1206, 1987.
1 8. Truswell, A. S., and Mann, J. I. Epidemiology of serum lipids in SouthernAfrica. Artheriosclerosis, 16: 15-29, 1972.
1 9. Mann, J. Diabetes Melitus: some aspects of aetiology and managementof noninsulin-dependent diabetes. In: H. Trowell, D. Burkitt, and K. Heaton
(eds.), Dietary Fibre, Fibre Dependent Foods and Disease, pp. 263-287.London: Academic Press, 1985.
20. Brown, W. V., and Karmally, W. Coronary heart disease and consump-tion of diets high in wheat and other grains. Am. J. Clin. NuIr., 41: 1163-1171, 1985.
21 . Antonis, A., and Bersohn, I. Influence of diet on serum triglycerides.Lancet, 3-9, 1961.
22. Hatch, J. T., Abell, L. L., and Kendall, F. E. Effects of restriction of fat andcholesterol upon serum lipids and lipoproteins in patients with hypertension.Am. J. Med., 19:48-60, 1955.
23. Grundy, S. M. Comparison of monounsaturated fatty acids and carbo-
hydrates for lowering plasma cholesterol. N. EngI. J. Med., 314: 745-749,
1986.
24. Coulston, A., Lui, G. C., and Reaven, G. M. Plasma glucose, insulin andlipid responses to high-carbohydrate low-fat diets in normal humans.Metabolism, 32: 52-56, 1983.
25. Lewis, B., Hammett, F., and Katan, M. Towards an improved lipidlowering diet: additive effects of changes in nutrient intake. Lancet, 2:1310-1313, 1981.
26. Jenkins, D. J. A. Lente carbohydrate: a newer approach to dietary man-
agement of diabetes. Diabetes Care, 5: 634-639, 1982.
27. Anderson, J. W., Chen, W. J. L., and Sieling, B. Hypolipidemic effects ofhigh carbohydrate high fibre diets. Metabolism, 29: 551-558, 1980.
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 693
28. Albrink, M. J., and Ullrich, I. H. Interaction of dietary sucrose and fiberon serum lipids in healthy young men fed high carbohydrate diets. Am. J.Clin. Nutr., 43:419-428, 1986.
29. Karlstrom, B., Vessby, B., Asp, N. G., Boberg, M., Gustaffson, I. B.,
Lithell, H., and Werner, I. Effects of an increased content of cereal fibre in the
diet of Type 2 (non-insulin dependent) diabetic patients. Diabetologia, 26:272-277, 1984.
30. Aderson, J. W., Story, L., Sieling, R. D., Chen, W. j. L., Petro, M. S., andStory, J. Hypocholesterolemic effects of oat bran or bean intake for hyper-
cholesterolemic men. Am. J. Clin. Nutr., 40: 1146-1155, 1984.
31. Bosello, 0., Cominacini, L., Zocca, I., Garbin, U., Ferrari, F., and
Davoli, A. Effects of guar gum on plasma lipoproteins and apolipoproteins
C-Il and C-Ill in patients affected by familial hyperlipoproteinemia. Am. J.Clin. Nutr., 40: 1 1 65-1 1 74, 1984.
32. Angelico, F., Clemente, P., Menoti, A., Ricci, G., and Urbinati, G. Bran
and changes in serum lipids: observations during a project of primary pre-
vention of coronary heart disease. In: L. A. Carlson (ed), International Con-ference on Artherosclerosis, pp. 205-207. New York: Raven Press, 1978.
33. Riccardi, G., Rivellese, A., Pacioni, D., Genovese, S., Mastranzo, P., andMancini, M. Separate influence of dietary carbohydrate and fibre on themetabolic control in diabetes. Diabetologia, 26: 116-121, 1984.
34. Hollenbeck, C. B., Coulston, A. M., and Reaven, G. M. To what extentdoes increased dietary fibre improve glucose and lipid metabolism in pa-tients with noninsulin-dependent diabetes mellitus (NIDDM)? Am. J. Clin.
Nutr., 43: 16-24, 1986.
35. Salvioli, G., Lugli, R., and Pradelli, J. M. Cholesterol absorption andsterol balance in normal subjects receiving dietary fibre or ursodeoxycholic
acid. Dig. Dis. Sci., 30: 301-307, 1985.
36. Wahlquist, M. L. Dietary fibre and carbohydrate metabolism. Am. J.Clin. Nutr., 45: 1232-1236, 1987.
37. Miettinen, T. A. Dietary fibre and lipids. Am. J. Clin. Nutr., 45: 1237-
1242, 1987.
38. Committee on Diet and Health. Diet and Health: Implications for Re-
ducing Chronic Disease Risk, pp. 431-464. Washington DC: National Acad-emy Press, 1989.
39. Gerard, M. J., Klatsky, A. L., Siegelaub, A. B., Friedman, G. D., and
Feldman, R. Serum glucose levels and alcohol-consumption habits in a largepopulation. Diabetes, 26: 780-785, 1977.
40. Ashley, F. W., and Kannel, W. B. Relation of weight change to changesin artherogenic traits: the Framingham Study. J. Chron. Dis., 27: 103-104,1974.
41 . Committee on Diet and Health. Diet and Health: Implications for Re-
ducing Chronic Disease Risk, pp. 529-547. Washington DC: National Acad-emy Press, 1989.
42. Committee on Diet and Health. Diet and Health: Implications for Re-ducing Chronic Disease Risk, pp. 563-592. Washington DC: National Acad-emy Press, 1989.
43. Lapidas, L., Bengtsson, C., Larsson, B., Ponnert, K., Rypo, E., and Sjos-
trom, L. Distribution of adipose tissue and risk of cardiovascular disease and
death: a 1 2 year follow-up of participants in the population study of womenin Gothenburg, Sweden. Br. Med. J., 289: 1 257-1 261 , 1984.
44. Bjorntorp, P. Adipose tissue in obesity (Willendorf lecture). In:J. Hirsch
and T. B. Van ltallie (eds.), Recent Advances in Obesity Research. Vol 4,pp. 163-170. London: John Libbey, 1985.
45. Krotkiewski, M., Bjorntorp, P., Sjostrom, L., and Smith, U. Impact ofobesity on metabolism in men and women. Importance of regional adiposetissue distribution. J. Clin. Invest., 72: 1150-1162, 1983.
46. Hartz, R. J., Rupley, D. C., and Rimm, A. A. The association of girth
measurements with disease in 32,856 women. Am. J. Epidemiol., 119:71-80, 1984.
47. Beynen, A. C., and Katan, M. B. Why do polyunsaturated fatty acidslower serum cholesterol? Am. J. Clin. Nutr., 42: 560-563, 1985.
48. Nestel, P. J., Connor, W. E., Reardon, M. F., Connor, S., Wong, S., andBoston, R. Suppression by diets rich in fish oil of very low density lipoprotein
production in men. J. Clin. Invest., 74: 82-89, 1984.
49. Bronsgeest-Schoute, H. C., von Gent, C. M., Luten, J. B., and Ruiter, A.The effect of various intakes of w3 fatty acids on the blood lipid compositionin healthy human subjects. Am. J. Clin. Nutr., 34: 1752-1757, 1981.
50. Zak, A., Zeman, M., Hrabak, P. 0., Svarcova, H., and Mares, P. Changesin the glucose tolerance and insulin secretion in hypertriglyceridemia: effects
of dietary n-3 fatty acids. Nutr. Rep. Int., 39: 235-242, 1989.
51 . Kasim, S. E., Stern, B., Khilnani, S., McLin, P., Baciorowski, S., and Jen,K. L. C. Effects of (0-3 fish oils on lipid metabolism, glycemic control, and
blood pressure in type-Il diabetic patients. J. Clin. Endocrinol. Metab., 67:1-5, 1988.
52. Lardinois, C. K. The role of w 3 fatty acids on insulin secretion and
insulin sensitivity. Med. Hypotheses, 24: 243-248, 1984.
53. Lauber, H., Wallace, P., Griver, K., and Brechtel, G. Adverse metabolic
effect of co-3 fatty acids in non-insulin dependent diabetes mellitus. Ann.Intern. Med., 108:663-668, 1988.
54. Schectman, G., Kaul, S., and Kissebah, A. H. Effect of fish oil concen-
tration on lipoprotein composition in NIDDM. Diabetes, 37: 1567-1573,1988.
55. Friday, K. E., Childs, M. T., Tsunehara, C. H., Fujimoto, W. Y., Bierman,
E. L., and Ensinck, J. W. Elevated plasma glucose and lowered triglyceridelevels from w-3 fatty acid supplementation in type II diabetes. Diabetes Care,
12:276-281, 1989.
56. Hendra, T. J., Britton, M. E., Roper, D. R., Wagaine-Twabwe, D., Jer-
emy, J. Y., Dandona, P., Haines, A. P., and Yudkin, J. S. Effects of fish oil
supplements with NIDDM subjects. Diabetes Care, 13: 821-829, 1990.
57. Huttunen, J. K., Lansimies, E., Voutilainen, E., Ehnholm, E., Hietanen, E.,Pentilla, I., Siltonen, B., and Rauramaa, R. Effect of moderate physical
exercise on serum lipoproteins: a controlled clinical trial with special refer-
ence to serum high-density lipoproteins. Circulation, 60: 1 220-1 229, 1979.
58. Committee on Diet and Health. Diet and Health: Implications for Re-
ducing Chronic Disease Risk, pp. 1 39-1 58. Washington DC: National Acad-emy Press, 1989.
59. Cooper, K. H., Pollock, M. L., Martin, R. P., White, S. R., Linnerud,
A. C., and Jackson, A. Physical fitness levels vs selected coronary risk factors.
IAMA, 236: 166-169, 1976.
60. Montoge, H. J., Block, W. D., Metzner, H. L., and Keller, J. B. Habitual
physical activity and serum lipids: males age 16-64 in a total community.J. Chron. Dis., 29: 697-709, 1976.
61 . Nikkila, E. A., Kussi, T., and Myllynen, P. High density lipoprotein and
apolipoprotein A-i during physical inactivity. Artherosclerosis, 37: 457-462, 1980.
62. Kussi, T., Nikkila, E. A., Saarinen, P., Varjo, P., and Laitinen, L. A.Plasma high density lipoproteins HDL2, HDL, and postheparin plasmalipases in relation to parameters of physical fitness. Artherosclerosis, 41:
209-219, 1982.
63. Hartung, G. H., Foreyt, J. P., Mitchell, R. E., Vlasek, I., and Gotto, A. M.
Relation of diet to high-density-lipoprotein cholesterol in middle-aged mar-athon runners, joggers and inactive men. N. EngI. J. Med., 302: 357-361,1980.
64. Wood, P. D., and Haskell, W. L. The effect of exercise on plasma highdensity lipoproteins. Lipids, 14:417-427, 1979.
65. Huttunen, J. K., Lansimies, E., Voutilainen, E., Ehuholm, C., Hietanen,E., Pentilla, I., Siitonen, 0., and Rauramaa, R. Effect of moderate physicalexercise on serum lipoproteins. Circulation, 60: 1220-1229, 1979.
66. Blairs, N., Cooper, K. H., Gibbons, L. W., Gettman, L. R., Lewis, S., and
Goodyear, N. Changes in coronary heart disease risk factors associated withincreased treadmill time in 753 men. Am. J. Epidemiol., 1 18: 352-359,
1983.
67. Lipson, L. C., Bonow, R. 0., Schafer, E. J., Brewer, H. B., and Lindgren,F. T. Effect of exercise conditioning on plasma high density Iipoproteins and
other lipoproteins. Artherosclerosis, 37: 529-538, 1980.
68. Rose, G., Blackburn, H., Keys, A., Kannel, W. B., Paul, 0., Reid, D. D.,and Stamler, J. Colon cancer and blood-cholesterol. Lancet, 1: 181-183,
1974.
69. Austin, M. A. Plasma triglyceride and coronary heart disease. Arterio-scler.Thromb., 11:2-14, 1991.
70. Pyoraia, K. Relationship of glucose tolerance and plasma insulin to theincidence of coronary heart disease: results from two population studies inFinland. Diabetes Care, 2: 131-141, 1979.
71 . Ducimetiere, P., Eschwege, E., Papoz, L., Richard, J. L., Claude, J.
and Rosselin, G. Relationship of plasma insulin levels to the incidence of
myocardial infarction and coronary heart disease mortality in a middle-aged
population. Diabetologia, 1 9: 205-2 1 0, 1980.
72. Welborn, T. A., and Wearne, K. Coronary heart disease incidence andcardiovascular disease mortality in Busselton with reference to glucose andinsulin concentrations. Diabetes Care, 2: 1 54-1 60, 1979.
73. Committee on Diet and Health. Diet and Health: Implications for Re-
ducing Chronic Disease Risk, p. 1 70. Washington DC: National AcademyPress, 1989.
74. Phillips, R. Role of life-style and dietary habits in risk of cancer among
Seventh-day Adventists. Cancer Res., 35: 3215-3222, 1975.
75. Manousos, 0., Day, N. E., Trichopoulos, D., Gerovassilis, G., and Tzo-nou, A. Diet and colorectal cancer: a case-control study in Greece. nt. J.
Cancer, 32: 1 55-1 61 , 1983.
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
694 Review: Triglycerides, Glucose, and Colorectal Cancer
76. Slavin, J. L., Lampe, J. W., Potter, J. D., and Lipkin, M. Cell proliferationin rectal biopsies collected from normal subjects consuming defined diets
differing in dietary fiber content. J. Nutr., in press, 1994.
77. Stamp, D., Zhang, X. M., Medline, A., Bruce, W. R., and Archer, M. C.Sucrose enhancement of early steps of colon cancer carcinogenesis in mice.
Carcinogenesis (Lond.), 14: 777-779, 1993.
78. Caderni, G., Bianchini, F., Mancina, A., Spagnesi, M. T., and Dolara, P.Effect of dietary carbohydrates on the growth of dysplastic crypt foci in thecolon of rats treated with 1,2-dimethylhydrazine. Cancer (Phila.), 51: 3721-
3725, 1991.
79. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinal can-cer: new applications to studies of cancer prevention in human subjects.
Cancer Res., 48: 235-245, 1988.
80. Deschner, E. E., and Lipkin, M. Proliferative patterns in colonic mucosain familial polyposis. Cancer (Phila.), 35:413-418, 1975.
81 . Romagnoli, P., Filipponi, F., Bandettini, L., and Brugnols, D. Increase ofmitotic activity in the colonic mucosa of patients with colorectal cancer. Dis.Colon Rect., 27: 305-308, 1984.
82. Deschner, E. E., Winawer, S. J., Katz, S., and Kahn, E. Proliferativedefects in ulcerative colitis patients. Cancer Invest., 1: 41-47, 1983.
83. Tepstra, 0. T., van Blaukenstein, M., Decs, J., and Eilers, C. A. M.Abnormal patterns of cell proliferation in the entire colonic mucosa ofpatients with colon adenoma or cancer. Gastroenterology, 92: 704-708,
1987.
84. Lipkin, M., Enker, W. E., and Winawer, S. J. Tritiated-thymidine label-ling of rectal epithetical cells in “non-prep” biopsies of individuals at in-creased risk for colonic neoplasia. Cancer Lett., 37: 153-167, 1987.
85. Bird, R. P. Observation and quantification of aberrant crypts in themurine colon treated with a colon carcinogen: preliminary findings. Cancer
Lett., 37: 147-151, 1987.
86. McLennan, E. A., and Bird, R. P. Aberrant crypts: potential preneoplasticlesions in the murine colon. Cancer Res., 48: 6187-6192, 1988.
87. Tudek, B., Bird, R. P., and Bruce, W. R. Foci of aberrant crypts in thecolons of mice and rats exposed to carcinogens associated with foods.
Cancer Res., 49: 1 236-1 240, 1989.
88. Day, D. W., and Morson, B. C. The adenoma-carcinoma sequence. In:
B. C. Morson (ed), The Pathogenesis of Colorectal Cancer, pp. 58-71.Philadelphia: W. B. Saunders, 1978.
89. Correa, P. Epidemiology of polyps and cancer. In: B. C. Morson, (ed),
The PathogenesisofColorectalCancer, pp. 126-152. Philadelphia:W. B.
Saunders, 1978.
90. Kashtan, H., Stern, H. S., Jenkens, D. J. A., Jenkens, A. L., Thompson,L. U., Hay, K., Marcon, N., Minkin, S., and Bruce, W. R. Effect of manipu-
lating colonic fermentation on markers of colorectal cancer risk. Am. J. Clin.Nutr., 50: 723-728, 1992.
91 . Eyssen, G. E. M., Bright-See, E., Bruce, W. R., and Jazmaji, V. A ran-domized trial of a low fat high fibre diet in the recurrence of colorectal
polyps. J. Clin. Epidemiol., 47: 525-536, 1994.
92. Eyssen, G. E. M. The epidemiology of colorectal cancer revisited: areserum triglycerides and plasma glucose associated with risk? The Application
of Molecular Biology to Cancer Control: New Avenues for EpidemiologicResearch. Montreal: Health Canada, 1994.
93. Eyssen, G. E. M. Are serum triglycerides and plasma glucose associatedwith the risk of colorectal cancer? Washington DC: American Society of
Preventive Oncology, 1994.
94. Committee on Diet and Health. Diet and Health: Implications for Re-
ducing Chronic Disease Risk, pp. 1 59-258. Washington DC: National Acad-emy Press, 1989.
95. Stadler, J., Yeung, K. S., Furrer, R., Marcon, N., Himal, H. S., and Bruce,W. R. Proliferative activity of rectal mucosa and soluble fecal bile acids in
patients with normal colons and in patients with colonic polyps or cancer.
Cancer Lett., 38: 315-320, 1988.
96. Yueng, K. S., McKeown-Eyssen, G. E., Li, G. F., Glazer, E., Hay, K.,Child, P., Gurgin, V., Zhu, S. L., Baptista, J., Aloe, M., Mee, D., Jazmaji, V.,Austin, D. F., Li, C. C., and Bruce, W. R. Comparisons of diet and biochem-
ical characteristics of stool and urine between chinese populations with lowand high colorectal cancer rates. J. NatI. Cancer Inst., 83: 46-50, 1991.
97. Rafter, J. J., Child, P., Anderson, A. M., Alder, R., Eng, V., and Bruce,W. R. Cellular toxicity of fecal water depends on diet. Am. J. Clin. Nutr., 45:559-563, 1987.
98. Rafter, J. I., Eng, V. W. S., Furrer, R., Medline, A., and Bruce, W. R. Effectof calcium and pH on the mucosal damage produced by deoxycholic acidin the rat colon. Gut, 27: 1320-1329, 1986.
99. Narisawa, T., Magadia, N. E., Weisburger, J. H., and Wynder, E. L.
Promoting effects of bile acids on colon carcinogenesis after intra-rectal
instillation of N-methyl-N-nitrosoguanine in rats. J. NatI. Cancer Inst., 53:1093-1095, 1974.
1 00. Sarwal, A. N., Cohen, B. I., Raicht, R. F., Takahashi, M., and Fazzini, E.Effects of dietary administration of chenodeoxycholate on N-methyl-N-ni-trosourea induced colon cancer in rats. Biochem. Biophys. Acta., 574:423-432, 1979.
101. Angelin, B., and Einarsson, K. Bile acids and lipoprotein metabolism.In: Grundy, S. M. (ed), Bile Acids and Artherosclerosis, pp. 41-66. NewYork: Raven Press, 1986.
102. Kessler, I. I. Cancer and diabetes mellitus a review of the literature.J. Chron. Dis., 23: 579-600, 1971.
1 03. Marks, P. A., and Bishop, J. S. The glucose metabolism of patients with
malignant disease and of normal subjects as studied by means of an intra-
venous glucose tolerance test. J. Clin. Invest., 36: 254-264, 1957.
104. Bishop, J. S., and Marks, P. A. Studies on carbohydrate metabolism in
patients with neoplastic disease. II. Response to insulin administration.J. Clin. Invest., 38: 668-672, 1959.
105. Weisenfeld, S., Hecht, A., and Goldner, M. G. Tests of carbohydratemetabolism in carcinomatosis. Cancer (Phila.), 15: 18-27, 1962.
106. MacBeth, R. A. L., and Bekesi, J. G. Plasma glycoproteins in various
disease states including carcinoma. Cancer Res., 22: 1 1 70-1 1 76, 1962.
107. Rohdenberg, G. L., Bernhard, A., and Krehbiel, 0. Sugar tolerance in
cancer. JAMA, 72: 1528-1530, 1919.
108. Levine, W., Dryer, A. R., Shekelle, R. B., Schoenberger, J., and
Stamler, J. Post-load glucose and cancer mortality in middle-aged men andwomen. Am. J. Epidemiol., 131: 254-262, 1990.
109. Smith, G. D., Egger, M., Shipley, M. J., and Marmot, M. G. Post-chal-
lenge glucose concentration, impaired glucose tolerance, diabetes and can-
cer mortality in men. Am. J. Epidemiol., 136: 1110-1114, 1992.
110. Wilson, E. B., and Maher, H. C. Cancer and tuberculosis with some
comments on cancer and other diseases. Am. J. Cancer, 16: 227-250, 1932.
1 1 1 . Glickman, A. S., and Rawson, R. W. Diabetes and altered carbohy-
drate metabolism in patients with cancer. Cancer (Phila.), 9: 1 1 27-i 134,1956.
1 1 2. O’Mara, B. A., Byers, T., and Schoenfeld, E. Diabetes mellitus and
cancer risk: a multisite case-control study. J. Chron. Dis., 38: 435-441,1985.
1 1 3. Adami, H. 0., McLaughlin, J., Ekbom, A., Berne, C., Silverman, D.,
Hacker, D., and Persson, I. Cancer risk in patients with diabetes mellitus.
Cancer Causes Control, 2: 307-31 4, 1 991.
114. Ragozzino, M., Melton, L. J., Chu, C. P., and Palumbo, P. J. Subsequent
cancer risk in the incidence cohort of Rochester, Minnesota, residents withdiabetes mellitus. J. Chron. Dis., 35: 13-19, 1982.
1 1 5. Kessler, I. I. Cancer mortality among diabetics. J. NatI. Cancer Inst., 44:
673-686, 1970.
1 1 6. Kessler, I. I. Mortality experience of diabetic patients, a twenty-six yearfollow-up. Am. J. Med., 51: 715-724, 1971.
1 1 7. Kruis, W., Forstmaier, G., Scheurlen, C., and Stellaard, F. Effect of diets
low and high in refined sugars on gut transit, bile acid metabolism and
bacterial fermentation. Gut, 32: 367-371 , 1 991.
1 1 8. Reaven, G. M. The role of insulin resistance in human disease. TheBanting Lecture 1 988. Diabetes, 37: 1 595-1 607, 1988.
1 1 9. Korc, M. Growth factors and oncogenes. Regul. Pept. Lett., 2: 7-1 1,1990.
1 20. Wong, M., and Holdaway, I. M. Insulin binding by normal and neo-plastic colonic tissues. Int. J. Cancer, 35: 335-341, 1985.
1 21 . Cullen, K. J., Yee, D., and Rosen, N. Insulin-like growth factors inhuman malignancy. Cancer Invest., 9: 443-454, 1991.
1 22. Salamon, E. A., Luo, J., and Murphy, L. J. The effect of acute and
chronic insulin administration on insulin-like growth factor-I expression in
the pituitary-intact and hypophysectomised rat. Diabetologia, 32: 348-353,1989.
1 23. Culouscou, J. M., and Shoyah, M. Purification of a colon cancer cellgrowth inhibitor and its identification as an insulin-like growth factor binding
protein. Cancer Res., 51: 2813-2819, 1991.
124. Bruning, P. F., Bonfrer, J. M. G., Van Noord, P. A. H., Hart, A. A. M.,
Jing-Bakker, M. D. E., and Nooyen, W. J. Insulin resistance and breast cancerrisk. Int. J. Cancer, 52: 51 1-516, 1992.
1 25. Dodge, B. G., O’Neal, M. A., Chang, J. P., and Griffin, A. G. Effects ofadrenocorticotropic and insulin on carcinogenesis in hypophysectomizedrats. J. NatI. Cancer Inst., 27:817-825, 1961.
126. Lorenzi, M., Montisano, D. F., Toledo, S., and Barvieux, A. Highglucose induces DNA damage in cultured human endothelial cells. J. Clin.
Invest., 77:322-325, 1986.
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 695
127. Schade, D. S., Santigao, J. V., Skyler, J. S., and Rizza, R. A. IntensiveInsulin Therapy. Excerpta Med., 23-35, 1983.
1 28. Bristol, J. B., and Williamson, R. C. N. Large bowel growth. In: J. M.Pollak, S. R. Bloom, N. A. Wright, and A. G. Butler, (eds.), Basic Science in
Gastroenterology, pp. 307-31 6. Ware, Herts, UK: Glaxo Group ResearchLtd., 1984.
1 29. Townsend, C. M. Peptides and gastrointestinal cancer growth. Regul.
Pept. Left., 2: 12-15, 1990.
1 30. Upp, J. R., Olson, D., and Poston, G. J. Inhibition of growth of twohuman pancreatic adenocarcinomas in vitro by somatostatin analog SMS
201-995. Am. J. Surg., 155: 29-35, 1988.
1 31 . Jenkins, D. J. A., Leeds, A. R., Gassull, M. A., Cochet, B., and Alberti,K. G. M. M. Decrease in postprandial insulin and glucose concentrations by
guar and pectin. Am. Intern. Med., 86: 20-23, 1977.
1 32. Jenkins, D. J. A., Leeds, A. R., Gassull, M. A., Wolever, T. M. S., Goff,
D. V., Alberti, K. G. M. M., and Hockaday, T. D. R. Unabsorbable carbohy-drate and diabetes: decreased post-prandial hyperglycaemia. Lancet, 2: 1 72-
174, 1976.
133. Beck-Nielsen, H., Pederson, 0., and Lindskov, 0. Impaired cellular
insulin binding and insulin sensitivity by high-fructose feeding in normalsubjects. Am. J. Clin. Nutr., 33: 273-278, 1980.
1 34. Roediger, W. E. W. Role of anaerobic bacteria in the metabolic welfare
ofthe colonic mucosa in man. Gut, 21: 793-798, 1980.
135. Cummings, J. H., and Englyst, H. N. Fermentation in the human large
intestine and available substrates. Am. J. Clin. Nutr., 45: 1243-1255, 1987.
1 36. Roediger, W. E. W. The effect of bacterial metabolites on nutrition and
function of the colonic mucosa: symbiosis between bacteria and man. In: H.
Kasper and H. Goebell (eds.), Colon and Nutrition, pp. 1 i-i 3. Lancaster, PA:MTP Press Ltd., 1982.
1 37. Kruh, J. Effects of sodium butyrate, a new pharmacological agent on
cells in culture. Mol. Cell. Biochem., 42: 65-82, 1982.
1 38. Smith, P. J. n-Butyrate alters chromatin accessibility to DNA repair
enzymes. Carcinogenesis (Lond.), 7: 423-429, 1986.
1 39. Kim, Y. S., Tsao, D., Morita, A., and Bella, A. Effect of sodium butyrateon three human colorectal adenocarcinoma cell lines in culture. In: R. A.
Malt and R. C. N. Williamson, (eds.), Colonic Carcinogenesis, pp. 31 7-323.
Falk Symposium 31, Lancaster: MTP Press Ltd., 1982.
1 40. Jass, J. R., Strudley, I., and Faludy, J. Histochemistry of epithelial meta-
plasia and dysplasia in human stomach and colorectum. Scand. J. Gastro-
enterol., 104 (suppl.): 109-130, 1984.
1 41 . Jass, J. R. Diet, butyric acid and differentiation of gastrointestinal tract
tumours. Med. Hypotheses, 18: 1 1 3-1 1 8, 1985.
142. Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumor-
igenesis. Cell, 61: 759-767, 1990.
143. McMichael, A. J., Jensen, 0. M., Parkin, D. M., and Zaridze, D. G.
Dietary and endogenous cholesterol and human cancer. Epidemiol. Rev., 6:
192-216, 1984.
144. Sorlie, P. D., and Feinleib, M. The serum cholesterol-cancer relation-
ship: an analysis oftime trends in the Framingham study. J. NatI. Cancer Inst.,
69: 989-996, 1982.
145. Miller, S. R., Tariter, P. I., Papatestas, A. E., Slater, G., and Aufses, A. H.
Serum cholesterol and colon cancer. J. NatI. Cancer Inst., 67: 297-300,
1981.
146. Stemmerman, G. N., Normura, A. M. Y., Heibrun, L. K., Pollack, E. S.,
and Kagan, A. Serum cholesterol and colon cancer incidence in Hawaiian
Japanese men. J. NatI. Cancer Inst., 67: 1 1 79-1 1 82, 1 981.
147. Jacobs, D., Blackburn, H., Higgins, M., Reed, D., Iso, H., McMiIIan, G.,
Neaton, J., Nelson, J., Potter, J., Rifkind, B., Rossouw, J., Shekelle, R., and
Yusuf, S. Report of the conference on low blood cholesterol: mortality
associations. Circulation, 86: 1046-1060, 1992.
148. Dc Fronzo, R. D., and Ferrannini, E. Insulin resistance: a multifaceted
syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and
artherosclerotic cardiovascular disease. Diabetes Care, 14: 1 73-1 94, 1 991.
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
1994;3:687-695. Cancer Epidemiol Biomarkers Prev G McKeown-Eyssen triglycerides and/or plasma glucose associated with risk?Epidemiology of colorectal cancer revisited: are serum
Updated version
http://cebp.aacrjournals.org/content/3/8/687
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cebp.aacrjournals.org/content/3/8/687To request permission to re-use all or part of this article, use this link
on April 21, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from