Dietary total antioxidant capacity and colorectal cancer: A large case–control study in Italy

Authors

  • Carlo La Vecchia,

    1. Department of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri—IRCCS, Milan, Italy
    2. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
    Search for more papers by this author
  • Adriano Decarli,

    1. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
    2. Unit of Medical Statistics, Biometry and Bioinformatics, Fondazione IRCSS Istituto Nazionale Tumori di Milano, Milan, Italy
    Search for more papers by this author
  • Mauro Serafini,

    1. Agricultural Research Council, CRA Ex-INRAN, Rome, Italy
    Search for more papers by this author
  • Maria Parpinel,

    1. Unit of Hygiene and Epidemiology, Department of Biological and Medical Sciences, University of Udine, Udine, Italy
    Search for more papers by this author
  • Rino Bellocco,

    1. Department of Statistics, University of Milano-Bicocca, Milan, Italy
    2. Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Solna, Sweden
    Search for more papers by this author
  • Carlotta Galeone,

    1. Department of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri—IRCCS, Milan, Italy
    2. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
    Search for more papers by this author
  • Cristina Bosetti,

    1. Department of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri—IRCCS, Milan, Italy
    Search for more papers by this author
  • Antonella Zucchetto,

    1. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
    2. Unit of Epidemiology and Biostatistics, Centro di Riferimento Oncologico, Aviano, Italy
    Search for more papers by this author
  • Jerry Polesel,

    1. Unit of Epidemiology and Biostatistics, Centro di Riferimento Oncologico, Aviano, Italy
    Search for more papers by this author
  • Pagona Lagiou,

    1. Department of Hygiene, Epidemiology and Medical Statistics, University of Athens Medical School, Athens, Greece
    2. Department of Epidemiology, Harvard School of Public Health, Boston, MA
    Search for more papers by this author
  • Eva Negri,

    1. Department of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri—IRCCS, Milan, Italy
    Search for more papers by this author
  • Marta Rossi

    Corresponding author
    1. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
    • Department of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri—IRCCS, Milan, Italy
    Search for more papers by this author

Correspondence to: Dr. Marta Rossi, Department of Epidemiology, Istituto di Ricerche Farmacologiche “Mario Negri”, Via La Masa, 19, 20156 Milano, Italy, Tel: [+39–0239014541], Fax: [+39–0233200231], E-mail marta.rossi@marionegri.it

Abstract

A favorable role of fruit and vegetables on colorectal cancer risk has been related to the antioxidant properties of their components. We used data from an Italian case–control study including 1,953 patients with incident, histologically confirmed colorectal cancer (1,225 colon and 728 rectal cancers). Controls were 4,154 patients admitted to hospital for acute, non-neoplastic conditions. A reproducible and valid food frequency questionnaire was used to assess subjects' usual diet. Total antioxidant capacity (TAC) was measured using Italian food composition tables in terms of ferric reducing-antioxidant power (FRAP), Trolox equivalent antioxidant capacity (TEAC) and total radical-trapping antioxidant parameter (TRAP). We estimated the odds ratios (ORs) and the corresponding 95% confidence intervals (CIs) through multiple logistic regression models, including terms for potential confounding factors, and energy intake. TAC was inversely related with colorectal cancer risk: the OR for the highest versus the lowest quintile was 0.68 (95% CI, 0.57–0.82) for FRAP, 0.69 (95% CI, 0.57–0.83) for TEAC and 0.69 (95% CI, 0.57–0.83) for TRAP. Corresponding values, excluding TAC deriving from coffee, were 0.75 (95% CI, 0.61–0.93) for FRAP, 0.76 (95% CI, 0.61–0.93) for TEAC and 0.71 (95% CI, 0.57–0.89) for TRAP. The inverse association was apparently—though not significantly—stronger for rectal than for colon cancer. This is the first case–control study indicating consistent inverse relations between dietary TAC and colorectal cancer risk.

A diet rich in vegetables and fruit has been associated with a reduced risk of various common cancers, particularly of the respiratory and digestive tracts.[1-6] It is unclear whether such a favorable effect may be attributed to specific micronutrients or bioactive compounds contained in plant foods. Among them, various antioxidants, such as vitamin C and E and flavonoids have been inversely associated to cancer risk.[7-10] Total antioxidant capacity (TAC) rather than individual antioxidants has been suggested as a relevant factor for cancer risk.[11-13] TAC has been defined as the moles of oxidants neutralized by one litre of plasma, food extracts or single molecules, and represents a biomarker of antioxidant potential, including redox synergistic interactions.[14, 15]

With reference to colorectal cancer, an inverse association between dietary TAC and rectal cancer risk has been reported in the Health Professionals Follow-up Study.[16] In that study, TAC was evaluated by ferric reducing-antioxidant power (FRAP) assay, which measures in vitro the reduction of the Fe3+ (ferric ion) to Fe2+ (ferrous ion) in the presence of antioxidants.

Since antioxidants may act in vivo through different mechanisms, other assays are also used to evaluate TAC.[17, 18] These include Trolox equivalent antioxidant capacity (TEAC), which measures the ability of antioxidant molecules to quench the long-lived ABTS+ compared with that of 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, Trolox, and total radical-trapping antioxidant parameter (TRAP), which measures the protection provided by antioxidants on the fluorescence decay of R-phycoerythrin (lag-phase) during a controlled peroxidation reaction.

The aim of this study is to investigate the relation between dietary TAC and colorectal cancer risk—through FRAP, TEAC and TRAP—using data from a large Italian study.

Material and Methods

Between 1992 and 1996 in six Italian area, we conducted a multicentric case–control study including 1,953 cases with colorectal cancer and 4,154 controls.[19]

Cases were subjects with histologically confirmed colorectal cancer diagnosed no longer than 1 year prior to the interview and no previous diagnoses of cancer at other sites. Overall, 1,225 subjects with cancer of the colon and 728 with cancer of the rectum and recto-sigmoid junction were included. Controls were patients admitted to the same hospitals as cases for acute, non-neoplastic conditions unrelated to digestive tract diseases: 23% were admitted for traumas, 28% for other orthopedic disorders, 20% for acute surgical conditions, 19% for eye diseases and 10% for miscellaneous other illnesses, such as ear-nose-and-throat, skin or dental conditions. This allowed to exclude controls with digestive/abdominal, non-neoplastic problems that may be predictors of colorectal cancer and influence food habits. About 5% of subjects approached for interview (cases and controls) refused to participate.

Centrally trained interviewers administered a standard questionnaire to cases and controls during their hospital stay. The questionnaire included personal and socio-demographic characteristics, anthropometric measures and lifestyle habits, including tobacco smoking and alcohol consumption. A reproducible[20] and valid[21] food frequency questionnaire (FFQ) was used to assess the patients' usual diet in the 2 years preceding cancer diagnosis (for cases) or hospital admission (for controls). The FFQ included the average weekly consumption of 78 food items or food groups and of 5 alcoholic beverages. Intakes lower than once a week, but at least once per month were coded as 0.5 per week.

We developed a food composition database for our FFQ using TAC measurements, which were assessed in terms of FRAP, TEAC and TRAP by Italian food tables.[17, 22] We then translated the frequency of consumption of each food item of the FFQ into average cumulative daily TAC, also taking portion size into account. To this purpose, we standardized the units of measures as mmol of Fe2+ equivalents per 100 g (solid foods) or 100 ml (beverages) for FRAP and mmol of Trolox per 100 g (solid foods) or 100 ml (beverages) for TEAC and TRAP, and obtained average mmol per day (mmol/day) of TAC for each subject.[23]

We computed both overall TAC and TAC excluding the contribution of coffee in order to estimate the risk for TAC that are not influenced by coffee consumption. In fact, coffee has been inversely related to colorectal cancer risk[24, 25] and the contribution of coffee represented about half of the total amount of TAC in our controls (56% for FRAP, 48% for TEAC and 62% for TRAP) with a high correlation between coffee and TAC. TAC without the contribution of coffee mainly derived from the consumption of fruit and vegetables (40% for FRAP, 42% for TEAC and 35% for TRAP), including citrus fruit (11% for FRAP, 12% for TEAC and 8% for TRAP) and apple and pears (5% for FRAP, 6% for TEAC and 8% for TRAP), and the consumption of wine (38% for FRAP, 36% for TEAC and 47% for TRAP). The three indexes were strongly correlated, with correlation coefficients between 0.97 and 0.99.

We also computed the intake of flavonoids by using food composition data published by the US Department of Agriculture (USDA).[26-28] Nutrient and energy intakes were computed using an Italian food composition database, supplemented with other published data.[29]

We computed energy-adjusted FRAP, TEAC and TRAP using the residual method.[30] The energy-adjusted FRAP, TEAC and TRAP were categorized into quintiles based on the control distribution, and the corresponding odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using unconditional multiple logistic regression models. All models included terms for age (quinquennia, categorically), sex, study centre, years of education (<7, 7–11, > 11, categorically), alcohol consumption (quartiles, categorically), body mass index (BMI) (quintiles, categorically), family history of colorectal cancer (yes/no), occupational physical activity (low, medium, and high, categorically) and FRAP, TEAC or TRAP from coffee alone (when considering TAC without the contribution of coffee). We also examined additional models including terms for the intakes of fruit, vegetables, red meat, flavonoids, vitamin C, beta-carotene, vitamin E, vitamin D, calcium and folates.

ORs per an increment of intake equal to the difference between the upper cut-off points of the IV and the I quintiles were also computed.

Results

Table 1 gives the correlation coefficients between FRAP, TEAC and TRAP—without the contribution of coffee—and other selected dietary covariates, including fruit, vegetables and flavonoids, among controls. A strong positive correlation was found between the three TAC indexes and total flavonoids (r ∼0.7), especially anthocyanidins (r ∼0.9).

Table 1. Correlationa of three non-coffee total antioxidant capacity indices with selected dietary covariates among 4,154 controls. Italy, 1992–1996
 FRAPTEACTRAP
  1. a

    Partial correlation coefficients adjusted for sex. The top value is for crude covariates and the bottom value is for energy-adjusted covariates.FRAP: Ferric reducing antioxidant-power; TEAC: Trolox equivalent antioxidant capacity; TRAP: Total radical-trapping antioxidant parameter.

Fruit0.270.290.19
 0.110.120.03
Vegetables0.210.240.16
 0.000.02−0.03
Flavonoids0.730.720.66
 0.650.640.56
Anthocyanidins0.880.880.91
 0.860.860.89
Flavones0.270.280.19
 0.050.04−0.02
Flavonols0.490.500.48
 0.300.300.27
Proanthocyanidins ≥ 10 mers0.520.550.48
 0.320.350.30
Vitamin C0.330.350.21
 0.090.10−0.03
Carotene0.190.210.13
 −0.04−0.03−0.09
Vitamin E0.320.390.25
 −0.22−0.15−0.24
Vitamin D0.140.170.10
 −0.16−0.14−0.17
Total energy0.590.630.53
 −0.10−0.09−0.08

Table 2 gives the mean daily TAC without the contribution of coffee among controls, and the ORs of colorectal cancer according to quintiles of the three TAC indices. The mean was 11.45 mmol for FRAP, 4.47 mmol for TEAC, 4.56 mmol for TRAP. We found inverse associations between TAC and colorectal cancer risk. The OR estimates were very similar for the three TAC estimates: the OR for the highest versus the lowest quintile was 0.75 (95% CI, 0.61–0.93; p-trend, 0.002) for FRAP, 0.76 (95% CI, 0.61–0.93; p-trend, 0.001) for TEAC and 0.71 (95% CI, 0.57–0.89; p-trend, <0.001) for TRAP. Corresponding values for overall TAC (considering both diet and coffee) were 0.68 (95% CI, 0.57–0.82), 0.69 (95% CI, 0.57–0.83) and 0.69 (95% CI, 0.57–0.83). TAC deriving from coffee alone was also inversely related to colorectal cancer risk: the OR was 0.71 (95% CI, 0.59–0.86) for FRAP, 0.72 (95% CI, 0.59–0.87) for TEAC and 0.71 (95% CI, 0.58–0.86) for TRAP.

Table 2. Odds ratios (ORs)a of colorectal cancer among 1,953 cases with colorectal cancer and 4,154 controls, and corresponding 95% confidence intervals (CIs) according to quintilesb (I–V) of three energy-adjusted non-coffee total antioxidant capacity indices. Italy, 1992–1996.
  OR1 (95% CI), Quintiles  
 Mean (SD)bIdIIIIIIVVp for trendOR continuousc
  1. aEstimated using multiple logistic regression models adjusted for sex, age, study centre, education, alcohol consumption, body mass index, family history, physical activity, TAC from coffee, and energy intake, according to the residual method.
  2. bMean and standard deviation (SD) among controls.
  3. cEstimated for an increment of intake equal to the difference between the upper cut-off points of the IV and the I quintiles.
  4. dReference category.
  5. eComputed as the sum of the upper cut off-points of energy-adjusted TAC quintiles plus the mean of TAC.FRAP: Ferric reducing-antioxidant-power; TEAC: Trolox equivalent antioxidant capacity; TRAP: Total radical-trapping antioxidant parameter.
FRAP (mmol/d)        
Upper cut off-pointse 7.939.9311.7514.34   
 11.45 (6.65)0.86 (0.72–1.03)0.77 (0.64–0.93)0.73 (0.60–0.88)0.75 (0.61–0.93)0.0020.88 (0.81–0.97)
TEAC (mmol/d)        
Upper cut off-pointse 3.173.934.595.54   
 4.47 (2.55)0.93 (0.78–1.11)0.77 (0.64–0.92)0.75 (0.62–0.91)0.76 (0.61–0.93)0.0010.88 (0.81–0.96)
TRAP (mmol/d)        
Upper cut off-pointse 2.883.784.615.94   
 4.56 (3.09)0.82 (0.68–0.98)0.75 (0.63–0.90)0.68 (0.56–0.83)0.71 (0.57–0.89)<0.0010.89 (0.82–0.97)

When we studied cancers of the colon and rectum separately, the OR for colon cancer was 0.81 (95% CI, 0.63–1.05; p-trend, 0.082) for FRAP, 0.86 (95% CI, 0.67–1.10; p-trend, 0.094) for TEAC and 0.77 (95% CI, 0.62–1.00; p-trend, 0.018) for TRAP. Corresponding values for rectal cancer were 0.69 (95% CI, 0.51–0.94; p-trend, 0.002) for FRAP, 0.64 (95% CI, 0.48–0.87; p-trend, 0.001) for TEAC and 0.65 (95% CI, 0.48–0.89; p-trend, 0.002) for TRAP. Tests for heterogeneity were not significant.

ORs changed little toward the null after adjustment for fruit, vegetables and other dietary factors, but more so after adjustment for some classes of flavonoids, particularly anthocyanidins. The OR for FRAP became 1.06 (95% CI, 0.80–1.42) adjusting for anthocyanidins, 0.88 (95% CI, 0.70–1.10) adjusting for flavonols and 0.84 (95% CI, 0.67–1.05) adjusting for proanthocyandins with more than 10 mers.

Discussion

In our study, dietary TAC was inversely related with the risk of colorectal cancer. Associations were somewhat stronger for rectal cancer, though the results were not significantly heterogeneous. Our results are in line with those from the Health Professional Follow-up Study that found an inverse association between FRAP and rectal cancer (relative risk, RR, 0.58, 95% CI, 0.35–0.96) on a cohort of 47,399 men including 201 rectal cancers.[16] However, in that study, the association was essentially explained by TAC from coffee, decaffeinated coffee and tea combined (RR: 0.62; 95% CI: 0.33–1.14, p-trend = 0.02), and dietary TAC from other food sources was not associated with the risk of rectal cancer (RR: 0.92; 95% CI: 0.53–1.60, p-trend = 0.84). In our study, not only TAC from coffee, but also TAC from other food sources was associated to colorectal cancer risk. We did not consider tea separately because the contribution of tea to total dietary TAC was very low in our data (∼1%).

With reference to the few available data on other cancer sites, dietary TRAP was inversely associated to the risk of gastric cancer in a Spanish[13] and a Swedish study.[12]

TAC mainly derives from vegetables and fruit. A favorable effect of fruit and vegetables on cancer risk was reported by the Greek EPIC Cohort study,[1] which found a 33% reduction in cancer incidence for subjects in the highest compared to the lowest quintile of fruit and vegetable consumption, as well as by a network of Italian case–control studies.[2, 31] High consumption of fruit and vegetables in Mediterranean populations, where seasonal and fresh vegetables and fruit are widely available, may facilitate the documentation of an inverse relation in those populations and explain, at least in part, the apparent discrepancy with American results.[1] Mediterranean diet has been inversely associated to the risk of selected cancers,[4, 32, 33] and a recent intervention study found that a Mediterranean diet was associated with high plasma antioxidant capacity.[34]

Results from observational studies and clinical trials on the influence of single antioxidants on cancer risk are inconclusive,[3] but compatible with the hypothesis that multiple antioxidants jointly act for preventing adverse consequences of oxidative stress, including carcinogenesis. A nested case–control study in the EPIC cohort[35] on biomarkers of oxidative stress and colorectal cancer found a positive association between prediagnostic serum levels of oxidative stress indicators (i.e., reactive oxygen metabolites) and colorectal cancer. The estimation of the overall antioxidant activity of foods in vitro allows to take into account the activity of all antioxidants and their potential synergistic and redox interactions.[36]

With reference to possible sources of bias, the interview setting and catchment areas were the same for cases and controls, and the participation rate was almost complete, thus limiting the size of selection bias. We excluded from the control group all admission diagnoses that might have involved long-term changes in diet. With reference to information bias, diagnosis of colorectal cancer is usually made after a period of symptoms that may influence food habits. However, our FFQ refers to the 2 years previous to disease diagnosis, thus limiting this possible source of bias. Moreover, our FFQ was satisfactorily reproducible[20] and valid,[21] though information on vitamin supplements was not available. Further, our dietary FRAP, TEAC and TRAP were comparable with those presented in a previous Italian study.[23] The main food contributors of TAC were wine, citrus fruits, apples and pears, as well as coffee, in our data. Given the favorable effect of coffee on colorectal cancer risk[24, 25] and the high contribution of coffee consumption to TAC, we also examined dietary TAC not deriving from coffee consumption in order to control for the possible confounding effect of coffee on the risk estimates.

Among the strengths of the study are the large sample size, and the use of a reproducible and valid FFQ in terms of specific food items and nutrients.[20, 21] We were able to adjust for major recognized risk factors for colorectal cancer, as well as for total energy intake, and the study has generated results on other colorectal cancer risk factors that were in line with other investigations,[19, 37] providing reassurance that major biases were not operating.

In previous investigations,[27, 28, 38] we found significant inverse associations between various micronutrients and flavonoids and colorectal cancer risk. Adjustment for flavonoids and particularly anthocyanidins reduced the strength of the inverse relations between TAC and colorectal cancer risk, but the high correlation between TAC and anthocyanidin intake made the models difficult to interpret. Anthocyanidins are poorly absorbed antioxidants. This suggests that not absorbed antioxidants may contribute to the protective effect, but the data were too limited for any additional inference. Likewise, although TAC assays may not adequately reflect antioxidant activity in vivo because the bioavailability of antioxidants is highly variable, TAC may exert important local activity in the gastrointestinal tract.[39]

This is to our knowledge the first report from a case–control study indicating consistent inverse relations between dietary TAC and colorectal cancer. Further studies, also based on plasma TAC assessments, are needed to provide stronger evidence on the role of antioxidants on the risk of colorectal and other digestive tract cancers.

Acknowledgement

M.R. was supported by a fellowship from Fondazione Umberto Veronesi

Ancillary