Int. J.
Cancer: 73, 525–530 (1997)
r 1997 Wiley-Liss, Inc.
INTAKE OF SELECTED MICRONUTRIENTS AND RISK
OF COLORECTAL CANCER
Carlo LA VECCHIA1,2*, Claudia BRAGA1, Eva NEGRI1, Silvia FRANCESCHI3, Antonio RUSSO3, Ettore CONTI4, Fabio FALCINI5,
Attilio GIACOSA 6, Marizio MONTELLA7 and Adriano DECARLI2,8
1Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’, Milan, Italy
2Istituto di Biometria e Statistica Medica, Università degli Studi di Milano, Milan, Italy
3 Servizio di Epidemiologia, Centro di Riferimento Oncologico, Aviano, PN, Italy
4Servizio di Epidemiologia e Oncogenesi, Istituto Regina Elena, Rome, Italy
5Divisione di Oncologia Medica, Ospedale Pierantoni, Forli, Italy
6Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy
7Istituto Tumori ‘‘Fondazione Pascale’’, Naples, Italy
8Istituto Nazionale per lo Studio e la Cura dei Tumori, Milan, Italy
The relationship between various micronutrients and colo-
rectal cancer risk was investigated using data from a case-
control study conducted between January 1992 and June 1996
in Italy. Cases were 1,953 incident, histologically confirmed
colorectal cancers (1,225 of the colon and 728 of the rectum),
admitted to the major teaching and general hospitals in the
study areas, and 4,154 controls with no history of cancer,
admitted to hospitals in the same catchment areas for acute,
non-neoplastic diseases unrelated to the digestive tract and
requiring no long-term modifications of the diet. Dietary
habits were investigated using a validated food-frequency
questionnaire. Odds ratio (ORs) were computed after allow-
ance for age, sex and other potential confounding factors,
including physical activity, total energy and fibre intake. For
most micronutrients, ORs were below unity with increasing
quintile of intake. The most consistent protective effects
were for carotene, riboflavin and vitamin C (Multivariate ORs
from the continuous model, with unit set as the difference
between the upper cut-point of the 4th quintile and that of
the 1st one, were 0.65, 0.73 and 0.80, respectively). Inverse
relationships were observed also for calcium and vitamin D
(ORs of 0.85 and 0.93, respectively). When the combined
effect of calcium and vitamin D and selected anti-oxidants was
considered, the OR reached 0.46 in subjects reporting high
calcium/vitamin D and high anti-oxidant intake compared to
those reporting low intake of both groups of micronutrients.
Most results were apparently stronger for colon cancer and
among females. Our results provide further support for a protec-
tive effect of several micronutrients on colorectal cancer risk
and some indications for a specific and stronger effect of
selected anti-oxidants. Int. J. Cancer 73:525–530, 1997.
r 1997 Wiley-Liss, Inc.
There are indications that various aspects of diet, including
several micronutrients, may influence the risk of colorectal cancer,
in the absence, however, of a clear and coherent pattern (Schatzkin
et al., 1995; Potter, 1996a). Micronutrients of potential relevance
may be subdivided into 4 major groups with different mechanisms.
Calcium and vitamin D, which may bind bile acids and free fatty
acids to form insoluble soaps and, hence, inhibit cell proliferation
in the intestinal mucosa (Newmark et al., 1984); ascorbic acid,
dietary carotenoids and vitamin E, which may act through an
anti-oxidant effect (Iscovich et al., 1992; Longnecker et al., 1992;
Maiani et al., 1995); folate, which may exert a protective effect on
colorectal carcinogenesis by increasing DNA methylation and the
production of S-adenosylmethionine, the primary methyl donor in
the body (Giovannucci et al., 1993; Glynn et al., 1996); and iron,
which may increase the risk by increasing free-radical generation
(Wurzelmann et al., 1996).
The epidemiological evidence on the issue, however, is unclear.
For instance, only about half of the published case-control and
cohort studies have found an inverse association between calcium
and colorectal cancer (Bergsma-Kadijk et al., 1996), and the
Nurses’ Health Study (Martinez et al., 1996), based on over 500
cases, found no relationship with calcium but an inverse one with
Publication of the International Union Against Cancer
Publication de l’Union Internationale Contre le Cancer
vitamin D, in agreement with previous suggestions from ecological
studies (Garland and Garland, 1980).
Most studies have found inverse relationships between anti-
oxidant vitamin intake and colorectal cancer (or its precursor,
colorectal adenoma [Enger et al., 1996]), though the most impor-
tant vitamins responsible for the apparent protection remain
unclear. For instance, in an Italian case-control study, beta-carotene
and ascorbic acid emerged as independent factors (Ferraroni et al.,
1994), while in the prospective Iowa Women’s Health Study
(Bostick et al., 1993), low vitamin E intake was significantly
related to increased colorectal cancer risk. In a meta-analysis of 13
case-control studies of colorectal cancer (Howe et al., 1992),
inverse associations were observed for vitamin C and beta-
carotene, which, however, were explained largely by allowance for
fibres. In contrast, in a case-control study of adenomatous polyps
(Witte et al., 1996), the most consistent protection was given by
vegetables and related micronutrients rather than fibres.
In general, most studies have been unable to include adequate
allowance for various micronutrients, as well as for other dietary
factors of specific interest. More important, it is unclear whether
the apparent protection of selected micronutrients can be explained
partly or largely by fruit and vegetable consumption, which have
been shown consistently to be protective against colorectal cancer
(Potter, 1996a). In some studies, moreover, the inverse associations
were more consistent and apparently stronger in females than in
males (Potter, 1996b). If not due to more accurate recall by females,
this may suggest a differential sex effect of selected micronutrients
on colorectal carcinogenesis.
To provide further information on the issue, we analysed the
relationship between selected micronutrients and colorectal cancer
risk using data from one of the largest studies of diet and colorectal
cancer to date, conducted in various Italian regions and based on a
validated food-frequency questionnaire (FFQ).
SUBJECTS AND METHODS
The data were derived from a case-control study of colorectal
cancer conducted between January 1992 and June 1996 in 6 Italian
areas: greater Milan, the provinces of Pordenone and Gorizia, the
urban area of Genoa, and the province of Forlì, in northern Italy;
the province of Latina, in central Italy; and the urban area of
Contract grant sponsor: Italian National Research Council (CNR);
Contract grant numbers 96.00548.PF39, 96.00701.PF39, 96.00759.PF39;
Contract grant sponsor: Italian Association For Cancer Research; Contract
grant sponsor: Commission of the European Communities.
*Correspondence to: Istituto di Ricerche Farmacologiche ‘‘Mario Ne-
gri’’, Via Eritrea, 62, 20157 Milano, Italy. Fax: (39)02-33200231.
Received 28 April 1997; Revised 18 June 1997
526 LA VECCHIA ET AL.

Naples, in southern Italy (Franceschi et al., 1997). The same without a linear term for the micronutrient’s quintile. Nutrients
structured questionnaire and coding manual were used by centrally were also introduced into the model as continuous variables. The
trained interviewers in all study centres. Data checking for unit for every micronutrient was set as the difference between the
consistency and reliability also was conducted centrally. On upper cut-point of the 4th quintile and that of the 1st one.
average, less than 4% of cases or controls refused to participate. To analyse the joint effects of various micronutrients and to
Cases were incident (i.e., diagnosed within 1 year before overcome the problem of different ranges of consumption, calcium,
interview), histologically confirmed patients with colorectal can- carotene and vitamins C, D and E were transformed (cubic root
cer, admitted to the major teaching and general hospitals in the function) and standardised so that they became approximately
areas under surveillance. Overall, 1,225 subjects with cancer of the normal deviates. Two variables were defined, one of which was the
colon (ICD-9: 153.0–153.9) and 728 with cancer of the rectum sum of calcium and vitamin D and the other that of carotene and
(ICD-9: 154.0–154.1) aged 23–74 years (median age 62) were vitamins C and E. These variables then were introduced into a
included. logistic regression model to analyse the combined role of intake of
Controls were patients with no history of cancer, from the same calcium and vitamin D and major identified anti-oxidants.
catchment areas as the cases and admitted to the same hospitals for
acute, non-neoplastic conditions unrelated to digestive tract dis- RESULTS
eases and requiring no long-term modifications of diet. A total of
4,154 control subjects, aged 20–74 years (median age 58), were Table I gives the distribution of cases and controls according to
interviewed. Of these, 23% were admitted for traumas (mostly sex, age, education, physical activity and quintiles of total energy
fractures and sprains), 28% for other orthopaedic disorders, 20% and fibre intake. Both colon and rectal cancer cases were older than
for acute surgical conditions, 19% for eye diseases and 10% for controls and reported a slightly higher total energy intake, while
miscellaneous other illnesses, such as ear, nose and throat, skin and they had levels of fibre intake comparable to those from controls.
dental conditions. Colon, but not rectal, cancer cases were more educated than
The questionnaire included information on socio-demographic controls and reported more frequently a low level of physical
characteristics, such as education and occupation; lifetime smoking activity. Thus, allowance for these variables was made in subse-
and alcohol-drinking habits; physical activity; anthropometric quent analyses.
measures at various ages; a problem-oriented personal medical Table II gives the ORs of colorectal cancer according to
history; and cancer family history. subsequent intake quintiles of various micronutrients compared to
An interviewer-administered FFQ was developed to assess the the lowest one, together with the tests for linear trend in risk and the
usual diet during the 2 years preceding diagnosis or hospital estimates from the continuous models. For all micronutrients
admission for the controls and, therefore, the intake of total energy except zinc, retinol, lycopene and folic acid, the ORs tended to
as well as macro- and micronutrients. The FFQ included 78 foods, decrease with increasing quintile of intake, with significant trends
groups of food or recipes divided into 7 sections: (i) bread, cereals
and first courses; (ii) second courses (e.g., meat and other main
dishes); (iii) side dishes (i.e., vegetables); (iv) fruits; (v) sweets, TABLE I – DISTRIBUTION OF 1,225 CASES OF COLON CANCER, 728 OF RECTAL
CANCER AND 4,154 CONTROLS ACCORDING TO SEX AND SELECTED
desserts and soft drinks; (vi) milk, hot beverages and sweeteners; CO-VARIATES: ITALY, 1992–1996
(vii) alcoholic beverages.
Cancer cases
For a few vegetables and fruits subject to seasonal variation,
Characteristic Colon Rectum Controls
consumption in season and the corresponding duration were
elicited. At the end of each section, 1 or 2 open questions were used Number (%) Number (%) Number (%)
to report foods not included in the questionnaire but eaten at least Sex
once a week. For 40 food items, the portion was defined in Male 688 (56.2) 437 (60.0) 2,073 (49.9)
‘‘natural’’ units (e.g., 1 teaspoon of sugar, 1 egg), while for the Female 537 (43.8) 291 (40.0) 2,081 (50.1)
remainder it was defined as small, average or large with the help of Age (years)
pictures. Dietary supplements were not considered, given their low ,40 55 (4.5) 26 (3.6) 347 (8.4)
levels of consumption by this population. 40–49 114 (9.3) 67 (9.2) 732 (17.6)
50–59 321 (26.2) 197 (27.1) 1,244 (30.0)
To compute energy and nutrient intake, Italian food-composition 60–69 518 (42.3) 306 (42.0) 1,356 (32.6)
data bases were used for about 80% of food items. These sources, $70 217 (17.7) 132 (18.1) 475 (11.4)
however, had to be integrated with other sources and with Education (years)1
information from manufacturers, especially as regards carotene, ,7 621 (50.9) 422 (58.2) 2,276 (55.2)
retinol and vitamin E (Salvini et al., 1996). Reproducibility and 7–11 331 (27.2) 181 (25.0) 1,156 (28.0)
validity of the FFQ were satisfactory (Decarli et al., 1996; $12 267 (21.9) 122 (16.8) 693 (16.8)
Franceschi et al., 1995). Physical activity
(score)1
Data analysis Low 534 (44.4) 291 (40.4) 1,625 (40.4)
Medium 415 (34.5) 232 (32.2) 1,462 (36.3)
Multiple logistic regression models were used to obtain odds High 255 (21.2) 197 (27.4) 940 (23.3)
ratios (ORs) and the corresponding 95% confidence intervals (CIs) Total energy intake
of colorectal cancer (Breslow and Day, 1980). All regression (quintiles)
equations included terms for age in quinquennia, area of residence, 1st (low) 211 (17.2) 148 (20.3) 863 (20.8)
sex, education, level of physical activity, total energy and fibre 2nd 246 (20.1) 128 (17.6) 846 (20.4)
intake. Further inclusion in the models of terms for the intake of the 3rd 260 (21.2) 151 (20.7) 812 (19.6)
major energy sources (i.e., starch, protein and fat), as well as of 4th 252 (20.6) 140 (19.2) 828 (19.9)
5th (high) 256 (20.9) 161 (22.1) 805 (19.4)
alcohol, which was not related materially to colorectal cancer risk Fibre intake (quin-
in this data set (OR 5 0.96, 95% CI 0.75–1.18 for the highest tiles)
quintile of intake), did not modify appreciably any of the estimates. 1st (low) 238 (19.4) 166 (22.8) 818 (19.7)
Quintiles were computed (i) directly on the micronutrients and 2nd 243 (19.8) 147 (20.2) 830 (20.0)
(ii) on the residuals of the regression of the micronutrients on 3rd 254 (20.7) 126 (17.3) 843 (20.3)
energy, following the method suggested by Willett (1990). Both 4th 246 (20.1) 135 (18.5) 840 (20.2)
5th (high) 244 (19.9) 154 (21.2) 823 (19.8)
analyses yielded similar results, and the former were chosen for
presentation. Tests for trends in the quintiles of nutrients were 1The sum does not add up to the total because of some missing

based on the likelihood ratio test between the models with and values.
 MICRONUTRIENTS AND COLORTECTAL CANCER 527
TABLE II – ODDS RATIOS1
(ORs) AND CORRESPONDING 95% CONFIDENCE INTERVALS (CIs) ACCORDING TO INTAKE
QUINTILE OF SELECTED MICRONUTRIENTS AMONG 1,953 CASES OF COLORECTAL CANCER AND 4,154
CONTROLS: ITALY, 1992–1996

Intake quintile,2 OR (95% CI) OR3

Nutrient
2 3 4 5 (high) x2trend (continuous)

Calcium (mg/day) 799–1,012 1,013–1,214 1,215–1,494 $1,495

0.94 0.71 0.77 0.72 10.77 0.91
(0.8–1.1) (0.6–0.9) (0.6–0.9) (0.6–0.9) p , 0.01 (0.8–1.0)
Iron (mg/day) 13–14 15–17 18–20 $21
0.84 0.70 0.64 0.57 13.27 0.83
(0.7–1.0) (0.5–0.9) (0.5–0.8) (0.4–0.8) p , 0.001 (0.7–1.0)
Phosphorus (mg/day) 1,230–1,466 1,467–1,691 1,692–2,003 $2,004
0.75 0.63 0.59 0.55 13.33 0.90
(0.6–0.9) (0.5–0.8) (0.4–0.8) (0.4–0.8) p , 0.001 (0.8–1.1)
Potassium (mg/day) 3,641–4,400 4,401–5,095 5,096–5,978 $5,979
0.72 0.63 0.51 0.45 23.28 0.66
(0.6–0.9) (0.5–0.8) (0.4–0.7) (0.3–0.6) p , 0.001 (0.5–0.8)
Zinc (mg/day) 10–11 12–13 14–16 $17
0.84 0.69 0.80 0.79 1.79 1.02
(0.7–1.0) (0.5–0.9) (0.6–1.0) (0.6–1.1) p 5 0.18 (0.9–1.2)
Retinol (µg/day) 156–233 234–380 381–1,654 $1,655
0.97 0.93 1.13 1.07 2.19 1.05
(0.8–1.2) (0.8–1.1) (0.9–1.4) (0.9–1.3) p 5 0.14 (1.0–1.1)
Carotene (µg/day) 3,595–4,853 4,854–6,077 6,078–7,692 $7,693
0.82 0.63 0.61 0.45 43.69 0.65
(0.7–1.0) (0.5–0.8) (0.5–0.8) (0.4–0.6) p , 0.001 (0.6–0.7)
Lycopene (µg/day) 5,411–7,377 7,378–9,181 9,182–11,720 $11,721
1.06 0.96 1.03 1.02 0.001 1.04
(0.9–1.3) (0.8–1.2) (0.8–1.2) (0.8–1.2) p 5 0.98 (0.9–1.1)
Thiamine (mg/day) 0.77–0.92 0.93–1.07 1.08–1.27 $1.28
0.76 0.74 0.67 0.65 6.42 0.96
(0.6–0.9) (0.6–0.9) (0.5–0.9) (0.5–0.9) p 5 0.01 (0.8–1.2)
Riboflavin (mg/day) 1.29–1.57 1.58–1.85 1.86–2.22 $2.23
0.92 0.70 0.69 0.72 10.02 0.86
(0.8–1.1) (0.6–0.9) (0.6–0.9) (0.6–0.9) p , 0.01 (0.8–1.0)
Niacin (mg/day) 14.9–17.6 17.7–20.1 20.2–23.8 $23.9
0.95 0.71 0.71 0.73 9.68 0.87
(0.8–1.1) (0.6–0.9) (0.6–0.9) (0.6–0.9) p , 0.01 (0.8–1.0)
Vitamin B6 (mg/day) 1.70–2.04 2.05–2.37 2.38–2.77 $2.78
0.75 0.60 0.58 0.53 15.26 0.88
(0.6–0.9) (0.5–0.8) (0.4–0.8) (0.4–0.7) p , 0.001 (0.7–1.1)
Folic acid (µg/day) 246–301 302–352 353–421 $422
0.88 0.74 0.72 0.83 3.54 0.95
(0.7–1.1) (0.6–0.9) (0.6–0.9) (0.6–1.1) p 5 0.06 (0.8–1.1)
Vitamin C (mg/day) 140–188 189–239 240–312 $313
0.92 0.72 0.76 0.73 8.78 0.86
(0.8–1.1) (0.6–0.9) (0.6–0.9) (0.6–0.9) p , 0.01 (0.8–1.0)
Vitamin D (mg/day) 2.02–2.66 2.67–3.34 3.35–4.27 $4.28
0.98 0.92 0.80 0.77 10.34 0.89
(0.8–1.2) (0.8–1.1) (0.7–1.0) (0.6–0.9) p , 0.01 (0.8–1.0)
Vitamin E (mg/day) 9.72–12.31 12.32–14.80 14.81–18.41 $18.42
0.90 0.70 0.64 0.63 17.24 0.87
(0.7–1.1) (0.6–0.9) (0.5–0.8) (0.5–0.8) p , 0.001 (0.8–1.0)
1Estimates from multiple logistic regression models including terms for age, centre, sex, education,
physical activity, total energy and fibre intake.–2Quintile 1 (lowest) is the reference category.–3Odds ratio
for a difference in intake equal to the difference between the upper cut-point of the 4th quintile and that of
the 1st.

in risk, and the ORs estimated using the continuous terms to be
below unity. Inclusion of a quadratic term to account for deviations
from linearity improved the fitting of the models for most
micronutrients, the only exceptions being carotene, for which the
linear model appeared to be preferable, and lycopene.
When we considered a model including a continuous term for all
nutrients significantly related to colorectal cancer, besides other
potential confounding factors, the only persisting significant protec-
tive effects were those of carotene (OR 5 0.65, 95% CI 0.6–0.8),
riboflavin (OR 5 0.73, 95% CI 0.6–0.9) and vitamin C (OR 5 0.80,
95% CI 0.6–1.0). Calcium and vitamin D were still below unity and
of borderline significance, with ORs of 0.85 (95% CI 0.6–1.1) and
0.93 (95% CI 0.8–1.1), respectively.
The ORs from the continuous models are shown in Table III for
colon and rectal cancers separately. Iron, phosphorus, niacin and
vitamins B6, D and E seemed to confer a stronger protection
against colon than rectal cancer, but no differential action was
observed for other micronutrients.
The relationship between various micronutrients and colorectal
cancer risk is examined further in Table IV in separate strata of sex
and age. No major differences emerged across age categories, but
several associations seemed stronger in females than in males, with
the exception of carotene. Interaction terms were significant for
calcium (x21 5 3.89), carotene (x21 5 5.96) and vitamin D
(x21 5 10.56).
The combined effect of a variable including calcium and vitamin
D and another including the major anti-oxidants identified (caro-
tene, vitamins C and E) also was analysed (Table V). The ORs
decreased across increasing levels of calcium/vitamin D and
anti-oxidants, reaching 0.46 (95% CI 0.35–0.60) in subjects
reporting high intake compared to those reporting low intake of
both groups of micronutrients.
528 LA VECCHIA ET AL.

TABLE III – ODDS RATIOS1,2 (ORs) AND CORRESPONDING 95% CONFIDENCE At variance with previous suggestions, no indication emerged of
INTERVALS (CIs) ACCORDING TO INTAKE OF SELECTED MICRONUTRIENTS
AMONG 1,225 CASES OF COLON CANCER, 728 OF RECTAL CANCER AND 4,154 an increased risk with high levels of iron intake (Wurzelmann et al.,
CONTROLS: ITALY, 1992–1996 1996), and no meaningful association emerged between folic acid
OR (95% CI) and colorectal cancer risk (Giovannucci et al., 1993; Glynn et al.,
Micronutrient 1996). Apparent differences between various studies may be
Colon Rectum
related to levels of intake and to different dietary and non-dietary
Calcium (mg/day) 0.92 0.90 sources of various micronutrients, including supplementation,
(0.8–1.0) (0.8–1.1) which, however, is relatively uncommon in Italy. Moreover, there
Iron (mg/day) 0.75 0.95
(0.6–0.9) (0.8–1.2) have been few systematic efforts to consider a large number of
Phosphorus (mg/day) 0.84 1.01 micronutrients simultaneously in the same data set or to allow for
(0.7–1.0) (0.8–1.3) the potential effect of one nutrient on the others.
Potassium (mg/day) 0.65 0.66 Although several micronutrients were considered in the present
(0.5–0.8) (0.5–0.9)
Zinc (mg/day) 0.99 1.08 study, which are protective against colorectal cancer remains an
(0.8–1.2) (0.9–1.3) open question, at least partly because of the co-linearity between
Retinol (µg/day) 1.08 1.00 various micronutrients and between selected foods (such as fruits
(1.0–1.2) (0.9–1.1) and vegetables; Franceschi et al., 1997) and various micronutrients.
Carotene (µg/day) 0.66 0.63 When we attempted to control simultaneously for the effect of
(0.6–0.8) (0.5–0.8) several micronutrients, results seemed to favour independent
Lycopene (µg/day) 0.97 1.15
(0.9–1.1) (1.0–1.3) actions of carotene, riboflavin and vitamin C, as well as calcium
Thiamine (mg/day) 0.91 1.04 and vitamin D. Even in the presence of several limitations in the
(0.7–1.1) (0.8–1.4) model, these results are consistent with the findings of a previous
Riboflavin (mg/day) 0.86 0.83 study on the same population, which found strong and independent
(0.7–1.0) (0.7–1.0) effects for carotene and vitamin C (Ferraroni et al., 1994).
Niacin (mg/day) 0.75 1.08
(0.6–0.9) (0.9–1.3) This study was sufficiently large to obtain reasonably precise risk
Vitamin B6 (mg/day) 0.80 0.99 estimates and significant associations for several micronutrients. Its
(0.6–1.0) (0.7–1.3) limitations and strengths are common to most hospital-based
Folic acid (µg/day) 0.95 0.95 case-control studies and have been discussed widely (Breslow and
(0.8–1.2) (0.7–1.2) Day, 1980). Hospital controls may differ from the general popula-
Vitamin C (mg/day) 0.91 0.79 tion in several respects, but we excluded from the control group
(0.8–1.1) (0.6–1.0)
Vitamin D (mg/day) 0.81 1.03 diagnoses related to known or likely risk factors for colorectal
(0.7–0.9) (0.9–1.2) cancer or to long-term modification of diet. That cases and controls
Vitamin E (mg/day) 0.73 1.14 came from comparable catchment areas and the almost complete
(0.6–0.9) (1.0–1.4) participation rate are reassuring against selection and recall bias.
1Estimates from multiple logistic regression models including terms Moreover, the hospital setting is likely to have improved the
for age, centre, sex, education, physical activity, total energy and fibre comparability of diet recall by cases and controls (D’Avanzo et al.,
intake.–2Odds ratio for a difference in intake equal to the difference 1997), and estimated intakes of various micronutrient sources were
between the upper cut-point of the 4th quintile and that of the 1st. satisfactorily reproducible and valid (Franceschi et al., 1993;
Decarli et al., 1996). Still, it is possible that diet in the more remote
past has an influence on colorectal cancer risk. With reference to
DISCUSSION confounding, all ORs were adjusted for age and other major
confounding factors, including education, physical activity, total
Our study, one of the largest case-control investigations on diet energy and fibre intake.
and colorectal cancer to date, suggests that colorectal cancer risk is
inversely related to consumption of several micronutrients. Most In conclusion, our large case-control study provides further
results were similar when colon and rectal cancers were analysed support for a protective effect of several micronutrients on colorec-
separately and when strata of age and sex were considered. tal cancer risk and some indications for a more relevant effect of
However, several associations seemed stronger for colon cancer selected anti-oxidants. Whether this reflects a specific effect of any
and among females (Potter, 1996b). such micronutrient(s) or the general composition of a diet richer in
various sources of micronutrients (Franceschi et al., 1997) remains,
Our data are in agreement with the hypothesis of a protective however, open to discussion and, given the complexity of the
effect of dietary carotenoids and vitamins C and E, possibly sources of micronutrients, possibly beyond the scope of observa-
through anti-oxidant mechanisms (Iscovich et al., 1992; Long- tional epidemiological studies. It is, nonetheless, of interest that the
necker et al., 1992; Maiani et al., 1995). This study also provides simple combination of high intakes for a few micronutrients led to
support for a protection by calcium and vitamin D against a more than 50% reduction in colorectal cancer risk in this
colorectal cancer, which has been explained in terms of binding of population.
bile and free fatty acids (Newmark et al., 1984) but is open to
discussion (Bergsma-Kadijk et al., 1996). More interesting, the
data suggest that the influence of each group of micronutrient
(anti-oxidants and calcium/vitamin D) is independent, and appar- ACKNOWLEDGEMENTS
ently multiplicative, for protection from colorectal cancer. Thus, This work was conducted within the framework of the CNR
subjects in the upper tertile of both groups of micronutrients had an (Italian National Research Council) Applied Project Clinical Appli-
.50% reduced colorectal cancer risk. cations of Oncological Research’’ (contracts 96.00548.PF39,
The apparent protective effect of riboflavin, if not due to bias, 96.00701.PF39 and 96.00759.PF39) and with the contributions of
could be explained in terms of induction of DNA-repair enzymes, the Italian Association for Cancer Research and the Europe against
as suggested by experiments on rodents (Webster et al., 1996). Cancer Programme of the Commission of the European Communi-
However, this association may be a chance finding, and, in the ties. The authors thank Ms. J. Baggott, Ms. I. Garimoldi, Ms. M.P.
absence of independent confirmation, any inference remains specu- Bonifacino and the G.A. Pfeiffer Memorial Library staff for
lative. editorial assistance.
 MICRONUTRIENTS AND COLORTECTAL CANCER 529
TABLE IV – ODDS RATIOS1,2
(ORs) AND CORRESPONDING 95% CONFIDENCE INTERVALS (CIs) ACCORDING TO INTAKE
OF SELECTED MICRONUTRIENTS AMONG 1,953 CASES OF COLORECTAL CANCER AND 4,154 CONTROLS IN STRATA OF
SEX AND AGE: ITALY, 1992–1996

Sex, OR (95% CI) Age (years), OR (95% CI)

Micronutrient
Males Females ,60 $60

Calcium (mg/day) 1.06 0.71 0.93 0.88

(0.9–1.2) (0.6–0.8) (0.8–1.1) (0.8–1.0)
Iron (mg/day) 0.90 0.76 0.80 0.87
(0.7–1.1) (0.5–1.1) (0.6–1.0) (0.7–1.1)
Phosphorus (mg/day) 1.12 0.62 0.95 0.85
(0.9–1.4) (0.5–0.8) (0.8–1.2) (0.7–1.1)
Potassium (mg/day) 0.72 0.58 0.59 0.74
(0.5–1.0) (0.4–0.8) (0.4–0.8) (0.6–1.0)
Zinc (mg/day) 1.13 0.79 0.99 1.05
(0.9–1.4) (0.6–1.0) (0.8–1.2) (0.9–1.3)
Retinol (µg/day) 1.02 1.11 1.05 1.07
(0.9–1.1) (1.0–1.3) (0.9–1.2) (1.0–1.2)
Carotene (µg/day) 0.57 0.76 0.61 0.69
(0.5–0.7) (0.6–0.9) (0.5–0.7) (0.6–0.8)
Lycopene (µg/day) 0.99 1.07 1.06 1.02
(0.9–1.1) (0.9–1.3) (0.9–1.2) (0.9–1.2)
Thiamine (mg/day) 0.99 0.86 0.96 0.96
(0.8–1.3) (0.6–1.2) (0.7–1.3) (0.7–1.2)
Riboflavin (mg/day) 0.95 0.74 0.85 0.86
(0.8–1.1) (0.6–0.9) (0.7–1.0) (0.7–1.0)
Niacin (mg/day) 0.92 0.75 0.84 0.91
(0.8–1.1) (0.6–0.9) (0.7–1.0) (0.7–1.1)
Vitamin B6 (mg/day) 1.00 0.67 0.82 0.95
(0.8–1.3) (0.5–0.9) (0.6–1.1) (0.7–1.2)
Folic acid (µg/day) 1.01 0.90 0.85 1.05
(0.8–1.3) (0.7–1.2) (0.7–1.1) (0.8–1.3)
Vitamin C (mg/day) 0.86 0.87 0.76 0.98
(0.7–1.1) (0.7–1.1) (0.6–0.9) (0.8–1.2)
Vitamin D (mg/day) 1.00 0.73 0.82 0.96
(0.9–1.1) (0.6–0.9) (0.7–1.0) (0.8–1.1)
Vitamin E (mg/day) 0.92 0.75 0.83 0.92
(0.8–1.1) (0.6–0.9) (0.7–1.0) (0.8–1.1)
1Estimates from multiple logistic regression models including terms for age, centre, sex, education,
physical activity, total energy and fibre intake.–2Odds ratio for a difference in intake equal to the difference
between the upper cut-point of the 4th quintile and that of the 1st.

TABLE V – ODDS RATIOS (ORs) AND CORRESPONDING 95% CONFIDENCE INTERVALS (CIs)1 OF COLORECTAL CANCER
ACCORDING TO THE COMBINED EFFECT OF CALCIUM/VITAMIN D AND SELECTED ANTI-OXIDANT (CAROTENE,
VITAMIN C AND E) INTAKE: ITALY, 1992–1996

Tertile of calcium/vitamin D,
Tertile of carotene/vitamins OR (95% CI) Total
C and E
1 (low) 2 3 (high)

1 (low) 12 0.94 0.66 12

(0.8–1.2) (0.5–0.9)
2 0.79 0.72 0.61 0.80
(0.6–1.0) (0.6–0.9) (0.5–0.8) (0.7–0.9)
3 (high) 0.61 0.54 0.46 0.61
(0.4–0.9) (0.4–0.7) (0.4–0.6) (0.5–0.8)
Total 12 0.91 0.74
(0.8–1.1) (0.6–0.9)
1Estimates from multiple logistic regression models including terms for age, centre, sex, education,
physical activity, total energy and fibre intake.–2Reference category.

REFERENCES
BERGSMA-KADIJK, J.A., VAN’T VEER, P., KAMPMAN, E., and BUREMA, J.,
Calcium does not protect against colorectal neoplasia. Epidemiology, 7,
590–597 (1996).
BOSTICK, R.M., POTTER, J.D., MCKENZIE, D.R., SELLERS, T.A., KUSHI, L.H.,
STEINMETZ, K.A. and FOLSOM, A.R., Reduced risk of colon cancer with high
intake of vitamin E: the Iowa Women’s Health Study. Cancer Res., 53,
4230–4237 (1993).
BRESLOW, N.E. and DAY, N.E., Statistical methods in cancer research, vol.
1, The analysis of case-control studies, IARC Sci. Publ., 32, IARC, Lyon
(1980).
D’AVANZO, B., LA VECCHIA, C., KATSOUYANNI, K., NEGRI, E. and TRICHO-
POULOS, D., An assessment and reproducibility of food frequency data
provided by hospital controls. Europ. J. Cancer Prev. 6, 288–296 (1997).
DECARLI, A., FRANCESCHI, S., FERRARONI, M., GNAGNARELLA, P., PARPINEL,
M.T., LA VECCHIA, C., NEGRI, E., SALVINI, S., FALCINI, F. and GIACOSA, A.,
Validation of a food frequency questionnaire to assess dietary intakes in
cancer studies in Italy: results for specific nutrients. Ann. Epidemiol. 6,
110–118 (1996).
ENGER, S.M., LONGNECKER, M.P., CHEN, M.-J., HARPER, J.M., LEE, E.R.,
FRANKL, H.D. and HAILE, R.W., Dietary intake of specific carotenoids and
vitamins A, C, and E, and prevalence of colorectal adenomas. Cancer
Epidemiol. Biomarkers Prev., 5, 147–153 (1996).
FERRARONI, M., LA VECCHIA, C., D’AVANZO, B., NEGRI, E., FRANCESCHI, S.
530 LA VECCHIA ET AL.
and DECARLI, A., Selected micronutrient intake and the risk of colorectal
cancer. Brit. J. Cancer, 70, 1150–1155 (1994).
FRANCESCHI, S., BARBONE, F., NEGRI, E., DECARLI, A., FERRARONI, M.,
FILIBERTI, E., GIACOSA, A., GNAGNARELLA, P., NANNI, O., SALVINI, S. and LA
VECCHIA, C., Reproducibility of an Italian food frequency questionnaire for
cancer studies. Results for specific nutrients. Ann. Epidemiol., 5, 69–75
(1995).
FRANCESCHI, S., FAVERO, A., LA VECCHIA, C., NEGRI, E., CONTI, E.,
MONTELLA, M., GIACOSA, A., NANNI, O. and DECARLI, A., Food groups and
risk of colorectal cancer in Italy. Int. J. Cancer, 72, 56–61 (1997).
FRANCESCHI, S., NEGRI, E., SALVINI, S., DECARLI, A., FERRARONI, M.,
FILIBERTI, R., GIACOSA, A., TALAMINI, R., NANNI, O., PANARELLO, G. and LA
VECCHIA, C., Reproducibility of an Italian food frequency questionnaire for
cancer studies: results for specific food items. Europ. J. Cancer, 29A,
2298–2305 (1993).
GARLAND, C.F. and GARLAND, F. C., Do sunlight and vitamin D reduce the
likelihood of colon cancer? Int. J. Epidemiol., 9, 227–231 (1980).
GIOVANNUCCI, E., STAMPER, M.J., COLDITZ, G.A., RIMM, E.B., TRICHOPOU-
LOS, D., ROSNER, B.A., SPEIZER, F.E. and WILLETT, W.C., Folate, methio-
nine, and alcohol intake and risk of colorectal adenoma. J. nat. Cancer Inst.,
85, 875–884 (1993).
GLYNN, S.A., ALBANES, D., PIETINEN, P., BROWN, C.C., RAUTALAHTI, M.,
TANGREA, J.A., GUNTER, E.W., BARRETT, M.J., VIRTAMO, J. and TAYLOR,
P.R., Colorectal cancer and folate status: a nested case-control study among
male smokers. Cancer Epidemiol Biomarkers Prev., 5, 487–494 (1996).
HOWE, G.R. and 23 OTHERS, Dietary intake of fiber and decreased risk of
cancers of the colon and rectum: evidence from the combined analysis of 13
case-control studies. J. nat. Cancer Inst., 84, 1887–1896 (1992).
ISCOVICH, J.M., L’ABBÉ, K.A., CASTELLETO, R., BERNEDO, A., CHOPITA,
N.A., JMELNITZSKY, A.C., KALDOR, J. and HOWE, G.R., Colon cancer in
Argentina. II. Risk from fibre, fat and nutrients. Int. J. Cancer, 51, 858–861
(1992).
LONGNECKER, M.P., MARTIN MORENO, J.M., KENEKT, P., NOMURA, A.M.Y.,
SCHOBER, S.E., STAHELIN, H.B., WALD, N.J., GEY, K.F. and WILLETT, W.C.,
Serum alpha-tocopherol concentration in relation to subsequent colorectal
cancer: pooled data from five cohorts. J. nat. Cancer Inst., 84, 430–435
(1992).
MAIANI, G., PAPPALARDO, G., FERRO-LUZZI, A., RAGUZZINI, A., AZZINI, E.,
GUADALAXARA, A., TRIFERO, M., FROMMEL, T. and MOBARHAN, S., Accumu-
lation of [67d] b-carotene in normal colorectal mucosa and colonic
neoplastic lesions in humans. Nutr. Cancer, 24, 23–31 (1995).
MARTINEZ, M.E., GIOVANNUCCI, E.L., COLDITZ, G.A., STAMPFER, M.J.,
HUNTER, D.J., SPEIZER, F.E., WING, A. and WILLETT, W.C., Calcium, vitamin
D, and the occurrence of colorectal cancer among women. J. nat. Cancer
Inst., 88, 1375–1382 (1996).
NEWMARK, H.L., WARGOWICH, M.J. and BRUCE, W.R., Colon cancer and
dietary fat, phosphate, and calcium: a hypothesis. J. nat. Cancer Inst., 72,
1323–1325 (1984).
POTTER, J.D., Nutrition and colorectal cancer. Cancer Causes Control., 7,
127–146 (1996a).
POTTER, J.D., Food and phytochemicals, magic bullets and measurement
error: a commentary. Amer. J. Epidemiol., 144, 1026–1027 (1996b).
SALVINI, S., GNAGNARELLA, P., PARPINEL, M.T., BOYLE, P., DECARLI, A.,
FERRARONI, M., GIACOSA, A., LA VECCHIA, C., NEGRI, E. and FRANCESCHI,
S., The food composition database for an Italian food frequency question-
naire. J. Food Compos. Anal., 9, 57–71 (1996).
SCHATZKIN, A., FREEDMAN, L.S., LANZA, E. and TANGREA, J., Diet and
colorectal cancer: still an open question. J. nat. Cancer Inst., 87, 1733–1735
(1995).
WEBSTER, R.P., GAWDE, M.D. and BHATTACHARYA, R.K., Modulation of
carcinogen-induced DNA damage and repair enzyme activity by dietary
riboflavin. Cancer Lett., 98, 129–135 (1996).
WILLETT, W., Nutritional epidemiology, Oxford University Press, Oxford
(1990).
WITTE, J.S., LONGNECKER, M.P., BIRD, C.L., LEE, E.R., FRANKL, H.D. and
HAILE, R.W., Relation of vegetable, fruit, and grain consumption to
colorectal adenomatous polyps. Amer. J. Epidemiol., 144, 1015–1025
(1996).
WURZELMANN, J.I., SILVER, A., SCHREINEMACHERS, D.M., SANDLER, R.S. and
EVERSON, R.B., Iron intake and the risk of colorectal cancer. Cancer
Epidemiol. Biomarkers Prev., 5, 503–507 (1996).