点击显示 收起
1 From the Istituto di Ricerche Farmacologiche "Mario Negri," Milan, Italy (CG, CP, EN, and CLV); the Registre vaudois des tumeurs, Institut universitarie de médicine sociale et préventive, CHUV-Falaises 1, Lausanne, Switzerland (FL); the International Agency for Research on Cancer, Lyon, France (SF); the Servizio di Epidemiologia e Biostatistica, Centro di Riferimento Oncologico, Aviano (Pordenone), Italy (RT); the Policlinico di Monza, Monza (Mi), Italy (AG); and the Istituto di Statistica Medica e Biometria, Università degli Studi di Milano, Milan, Italy (CLV)
2 Supported by contributions from the Italian Association for Cancer Research, the Italian and Swiss (KLS-01633-02-2005) Leagues against Cancer, and the Italian Ministry of Education (PRIN 2005). 3 Address reprint requests to C Galeone, Istituto di Ricerche Farmacologiche "Mario Negri," Via Eritrea, 62 - 20157 Milano, Italy. E-mail: galeone{at}marionegri.it.
ABSTRACT
Background: Interest in the potential benefits of allium vegetables, in particular, onion (Allium cepa) and garlic (Allium sativum), has its origin in antiquity, but the details of these benefits are still open to discussion.
Objective: We investigated the role of allium vegetables in the etiology of various neoplasms. Previous data are scanty and are based mainly on Chinese studies.
Design: Using data from an integrated network of Italian and Swiss case-control studies, we analyzed the relation between frequency of onion and garlic use and cancer at several sites. We calculated odds ratios (ORs) by using multivariate logistic regression models that were adjusted for energy intake and other major covariates.
Results: Consumption of onions varied between 0–14 and 0–22 portions/wk among cases and controls, respectively. The multivariate ORs for the highest category of onion and garlic intake were, respectively, 0.16 and 0.61 for cancer of the oral cavity and pharynx, 0.12 and 0.43 for esophageal cancer, 0.44 and 0.74 for colorectal cancer, 0.17 and 0.56 for laryngeal cancer, 0.75 and 0.90 for breast cancer, 0.27 and 0.78 for ovarian cancer, 0.29 and 0.81 for prostate cancer, and 0.62 and 0.69 for renal cell cancer.
Conclusions: This uniquely large data set from southern European populations shows an inverse association between the frequency of use of allium vegetables and the risk of several common cancers. Allium vegetables are a favorable correlate of cancer risk in Europe.
Key Words: Allium vegetables • onion • Allium cepa • garlic • Allium sativum • tomato • diet • risk factors • oral cancer • pharyngeal cancer • esophageal cancer • colorectal cancer • laryngeal cancer • breast cancer • ovarian cancer • prostate cancer • renal cell cancer • case-control study
INTRODUCTION
Interest in the potential health benefits of allium vegetables, in particular, onion (Allium cepa) and garlic (Allium sativum), has its origin in antiquity, but the details of these benefits are still open to discussion. Several epidemiologic investigations have suggested an inverse relation between intake of allium vegetables and stomach, colorectal, and prostate cancers (1, 2), but the results for other cancer sites are scanty, especially for Western countries. In fact, most studies were conducted in China, and the results of those studies are not easily applicable to Western populations, whose dietary habits are largely different; in particular, garlic intake is far lower (3-6).
One of the early epidemiologic studies considering this issue was a case-control study of diet and gastric cancer conducted in Italy (7) that found an odds ratio (OR) for the highest consumption of garlic of 0.40. Later, 3 Chinese case-control studies confirmed the favorable role of allium vegetables (8-10), and the relative risk (RR) of stomach cancer for garlic consumption estimated by a meta-analysis (5) was 0.54 (95% CI: 0.44, 0.66). That study also found a protective role of garlic against colorectal cancer, with a RR of 0.67 (95% CI: 0.56, 0.80), on the basis of 2 cohort and 4 case-control studies (5). Allium vegetables were also inversely related to breast cancer in 2 European investigations. In a French case-control study of breast cancer (11), the OR was 0.30 (95% CI: 0.17, 0.52) for the highest intake of allium vegetables, and a Swiss study found similar results (12). Also, the role of allium vegetables in the etiology of prostate cancer has been rarely studied. One study found an OR for the highest intake versus nonuse of 0.64 (95% CI: 0.38, 1.09) (13), whereas in a recent Chinese study, the OR was 0.51 (95% CI: 0.34, 0.76) (3). The studies that examined the issue on upper aerodigestive tract cancers were conducted in China and found a protective role of allium vegetables, with ORs varying between 0.3 and 1.1 (4, 9, 14, 15). Furthermore, a recent US study reported an inverse relation between allium vegetable intake and the risk of pancreatic cancer (OR = 0.46; 95% CI: 0.33, 0.63) (6). To provide further information on the issue, we have analyzed the relation between the frequency of onion and garlic use and the risk of cancer at various sites, using data from an integrated network of case-control studies conducted in southern Europe.
SUBJECTS AND METHODS
Between 1991 and 2004, we conducted a series of case-control studies on several neoplasms in various areas of northern Italy (greater Milan; the provinces of Pordenone, Padua, Udine, and Gorizia; the urban area of Genoa; and the province of Forlì), central Italy (the provinces of Rome and Latina), and southern Italy (the urban area of Naples). Studies on cancers of the oral cavity and pharynx, esophagus, large bowel, larynx, and breast were also conducted in the Swiss Canton of Vaud. Data were obtained with the same design, questionnaire, and inclusion criteria (12, 16-24). All studies included only incident, histologically confirmed cases. The study protocol was revised and approved by the ethical committees and directors of the hospitals involved, and all participants gave informed consent.
The first study, on cancer of the oral cavity and pharynx, included 749 cases (median age: 57 y) and 1772 controls (median age: 57 y). The second study, on cancer of the esophagus, included 395 cases (median age: 60 y) and 1066 controls (median age: 60 y). The third study, on the large bowel, included 1394 cases of colon cancer (median age: 62 y), 886 cases of rectal cancer (median age: 63 y), and 4765 controls (median age: 58 y). The fourth study, on laryngeal cancer, included 527 cases (median age: 61 y) and 1297 controls (median age: 61 y). The fifth study, on breast cancer, included 2900 cases (median age: 55 y) and 3122 controls (median age: 56 y). The sixth study, on cancer of the ovary, included 1031 cases (median age: 56 y) and 2411 controls (median age: 57 y). The seventh study, on prostate cancer, included 1294 cases (median age: 66 y) and 1451 controls (median age: 63 y). The eighth study, on renal cell cancer, included 767 cases (median age: 62 y) and 1534 controls (median age: 62 y).
Controls were admitted to the same network of hospitals as cases for a wide spectrum of acute, nonneoplastic conditions unrelated to known or potential risk factors for the corresponding cancer site considered or to long-term diet modification (overall, 24% had traumas, 29% had nontraumatic orthopedic disorders, 21% had acute surgical conditions, and 26% had miscellaneous other illnesses). All cases and controls were <80 y old and were identified and questioned by trained interviewers during their hospital stay in a network of teaching and general hospitals in the areas under surveillance. The proportion of refusals was <5% among cases and controls in all Italian centers and 15% in Switzerland.
Interviewers used a structured questionnaire that included information on sociodemographic factors, anthropometric variables, smoking, alcohol and other lifestyle habits, a problem-oriented medical history, physical activity, and history of cancer in relatives. Information on diet referred to the 2 y preceding diagnosis and was based on a food-frequency questionnaire (FFQ) that was validated for nutrient intake and tested for reproducibility for specific nutrients and food items. The FFQ included 78 foods, food groups, or recipes and allowed the estimation of energy intake (25, 26). The average weekly frequency of consumption of foods was elicited, and therefore the FFQ looked more like a menu from a restaurant than a shopping list. Frequencies of less than once per week but at least once per month were coded as 0.5 per week, whereas never or less than once per month was coded as 0. The FFQ was divided into 6 sections: 1) bread, cereals, and first courses; 2) second courses (ie, meat, fish, and other main dishes); 3) side dishes (ie, vegetables and fried or baked potatoes); 4) fruit; 5) sweets, desserts, and soft drinks; and 6) milk, hot beverages, and sweeteners. At the end of each section, 1 or 2 open questions were used to include foods that were not in the questionnaire but were eaten at least once per week.
Among the items in the FFQ, 2 questions referred specifically to consumption of onion and garlic. For frequency of onion use, we asked for the weekly frequency of consumption and usual portion size (small, medium, large), for which an intermediate portion corresponded to 80 g onion. For frequency of garlic use, we asked for the common consumption as a qualitative variable, scored as 1 for nonuse or low use, 2 for intermediate use, and 3 for high use. No additional information on the type of garlic and onion and on manner of using (fresh, powders, or garlic supplements) was available.
Statistical analyses
ORs and the corresponding 95% CIs of selected cancers according to the different frequencies of onion use (never, 1, >1 to <7, and 7 portions/wk) and garlic use (nonuse-low, moderate, and high) were derived by using unconditional multiple logistic regression (in unmatched studies, ie, breast, ovarian, colorectal, and prostate cancers) or conditional multiple logistic regression (in studies matched for age, sex, and study center, ie, upper aerodigestive tract and renal cell cancers) (27). All regression models included terms for age, sex (where appropriate), study center, education, body mass index, and energy intake. According to the cancer site analyzed, further adjustments were made for alcohol drinking, smoking habit, physical activity, parity, and family history of cancer at the same site.
RESULTS
The distribution of cases and controls according to sex, age, and information on selected dietary aspects in the studies analyzed are shown in Table 1. For each cancer study, cases consumed both vegetables and onions less frequently than did controls, except for vegetable intake in the prostate cancer study. In our data, consumption of onions varied between 0–14 and 0–22 portions/wk among cases and controls, respectively. Garlic use was lower in cases in all studies, except for the breast, ovary, and prostate cancer studies, for which mean values were not significantly different.
View this table:
TABLE 1. Sex and age distribution and information on selected dietary aspects in the studies investigated, Italy and Switzerland, 1991–2004
Shown in Table 2 for each cancer site considered are the distribution of cases and controls and the ORs and 95% CIs for subsequent frequencies or score of use of onion and garlic. Significant inverse associations were obtained for several cancer sites: the ORs for the highest versus the lowest category of use of onion were 0.44 (95% CI: 0.28, 0.67) for cancer of the large bowel, 0.17 (95% CI: 0.05, 0.61) for cancer of the larynx, and 0.27 (95% CI: 0.08, 0.85) for cancer of the ovary. For the frequency of use of garlic, the ORs for the highest versus the lowest category were 0.61 (95% CI: 0.44, 0.85) for cancer of the oral cavity or pharynx, 0.43 (95% CI: 0.28, 0.67) for cancer of the esophagus, 0.74 (95% CI: 0.63, 0.86) for cancer of the large bowel, 0.56 (95% CI: 0.38, 0.82) for cancer of the larynx, 0.81 (95% CI: 0.64, 1.00) for cancer of the prostate, and 0.69 (95% CI: 0.53, 0.92) for renal cell cancer. These results did not substantially change when we adjusted the multivariate models for total vegetable intake. None of the ORs was above unity. No significant direct relations were observed.
View this table:
TABLE 2. Distribution of cases and controls and odds ratios (ORs) and 95% CIs of selected cancers according to frequency of onion use (portions/wk) and garlic use, Italy and Switzerland, 1991–20041
We considered consumption of onions and garlic in strata of age and sex (data not shown). The estimates were not significantly different across strata of the considered covariates for each cancer site, except for onion consumption in strata of age in the laryngeal study (P = 0.01) and in strata of sex in the esophageal study (P = 0.04) and for garlic consumption in strata of age in the ovarian study (P = 0.01). The 3-factor interaction (sex, age, and frequency of onion or garlic use) was not significant for any cancer site, with the exception of larynx cancer (P = 0.005) and onion use.
DISCUSSION
Studies of experimental carcinogenesis in animal models and in cell culture systems indicate that several allium-derived compounds exhibit inhibitory effects and that the underlying mechanisms may involve both the initiation and the promotion phases of carcinogenesis (28). However, the specific components of allium vegetables involved in the specific cellular and molecular events that govern these anticancer properties are not known (29). Garlic contains several different organosulfur compounds in addition to amino acids, vitamins, and micronutrients. Its anticarcinogenic actions may be explained by particular organosulfur compounds. Diallyl sulfide, for example, which is responsible in part for its strong taste and odor, has been shown to selectively inhibit as well as induce certain P-450 enzymes (30). The potential anticarcinogenic action of onions may also be related to their high content of organosulfur compounds or to their high antioxidant activity, which is principally due to their wide content of flavonoids. However, there are important varietal differences in the composition, concentration, and beneficial activities of these bioactive compounds, which also result by modalities of cooking (31).
In our study, we found a protective role of a moderate frequency of onion consumption against colorectal, laryngeal, and ovarian cancers. The inverse relation was even more evident for high frequency of use, when it was also significant for oral cavity and esophageal cancers, but not for prostate, breast, or renal cell cancers. A moderate frequency of garlic consumption was inversely related to colorectal and renal cell cancers, and a high frequency of garlic consumption was significantly and inversely related to all cancer sites, except again for breast and prostate. These latter cancers are mainly associated with hormonal and reproductive factors, whereas their relation with dietary factors is inconsistent and still open to discussion (32).
Onion and garlic intakes could be simply considered markers of a healthier lifestyle, which may include complex aspects of quantity and quality of diet. In fact, several of the cancer sites considered have been inversely related to intake of vegetables in several studies (12, 16, 18-24), and the evidence is more convincing for cancers of the oral cavity and pharynx, esophagus, and large bowel (33). For this reason, we adjusted all the models for total vegetable intake, but the results did not substantially change. We were also able to adjust for selected major lifestyle and nutritional factors in the analysis, including alcohol use, tobacco use, and physical activity and were unable to explain the association observed. However, it is still conceivable that a complex combination of favorable correlates of allium vegetable intake may have contributed to the apparent protection at several cancer sites. In fact, in the Italian diet, onion and garlic are often eaten or cooked in combination with foods such as tomatoes and olive oil in salads and sauces for pasta, and some studies reported that the synergistic action of garlic and tomato could have a preventive effect against the carcinogenic process (34, 35).
We tried to minimize selection bias by excluding all control patients with diagnoses linked to long-term changes in diet or admitted for chronic conditions. Furthermore, the findings of our study cannot be due to a selectively higher response rate of health-conscious control subjects (36), because participation was practically complete for both cases and controls. An important limitation of our study is that we collected no information on the variety of onions and type of garlic used and on modalities of cooking. Among the strengths of these studies are the large number of subjects included; the similar interview setting of cases and controls, which provides reassurance against potential information bias (37); the ability to adjust for major selected potential confounding factors; and the use of a validated and reproducible FFQ (25, 26). In particular, the Spearman correlation coefficient for reproducibility of frequency of onion use was 0.48, and the concordance of the subjective question on garlic use was well reproducible in >70–80% of the subjects (25). The Pearson correlation coefficient for the validity of total energy intake was 0.63 (26).
Because the role of allium vegetables on cancer risk has been rarely studied, especially in Western countries, our study is original and of particular interest. Our findings confirm the indications of mainly Chinese studies of a protective role of allium vegetables on the risk of several common cancers.
ACKNOWLEDGMENTS
We thank C Pasche and F Lucchini for the Swiss data collection and validation and I Garimoldi for editorial assistance.
CG and CLV had the original idea for the study and wrote the report with contributions from EN and CP. CG and CP analyzed the data. CLV, FL, SF, RT, and AG conceived and coordinated the studies. The authors had no conflicts of interest.
REFERENCES