Abstract
Although the incidence of stomach cancer has been declining, it remains the second leading cause of cancer death worldwide. Potential protective effects of allium vegetables against cancer have been reported by a few epidemiologic studies in Chinese populations, but the sample sizes of these studies were relatively small. We examined the associations between allium vegetable consumption and stomach cancer in a large population-based case-control study in Shanghai (750 cases and 750 age- and gender-matched controls) and Qingdao (128 cases and 128 age- and gender-matched controls). Epidemiological data were collected by a standard questionnaire, and odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using conditional logistic regression in SAS. After adjusting for matching variables, education, body mass index, pack-years of smoking, alcohol drinking, salt intake, and fruit and vegetable intake, inverse relationships with dose response pattern were observed between frequency of onion intake and stomach cancer in Qingdao (P for trend=0.02) and Shanghai (P for trend=0.04) populations. In Shanghai, negative dose-response relationships were observed between monthly intake of onions (P=0.03), monthly intake of garlic stalks (P=0.04) and distal cancer (but not with cardia cancer). Negative association was also noted between intake of garlic stalks (often vs. never) and risk of stomach cancer in Qingdao (OR=0.30; 95% CI: 0.12–0.77). Our results confirm the protective effect of allium vegetables (especially garlic and onions) against stomach cancer.
Keywords: allium vegetable, garlic, onion, stomach cancer, risk factor, epidemiology, China
Introduction
Allium vegetables, which include garlic, onions, shallots, leeks, scallions and chives, contain high levels of flavanols and organosulfur compounds. From animal and in vitro studies, allium vegetables have been shown to have an anticarcinogenic potential of bioactive compounds, e.g. allyl sulfides (Nakagawa et al., 2001; Wargowich et al., 1996), and an antibacterial effect against Helicobacter pylori (H. pylori), which is known to be important in stomach carcinogenesis (Jonkers et al., 1999). The antibacterial effect, which is attributed to garlic’s thiosulfinate concentration, has recently been quantified. Studies have reported that garlic extract could inhibit H. pylori growth in vitro (Canizares et al., 2004; Garcia-Arata et al., 1999; Sivan et al., 1997). More general anticarcinogenic effects may derive from the organosulfur compounds in garlic that are responsible for its odor and flavor (Lawson et al., 1991). Previous animal studies have also shown that a compound extracted from garlic and onions, diallyl sulfide, has strong cancer inhibitory properties (Nakagawa et al., 2001; Reddy et al., 1993; Belman 1983; Wargowich 1987; Wargowich et al., 1988).
Because allium vegetables, especially garlic and onions, have been shown to have anticarcinogenic potential and an antibacterial effect against H. pylori, we hypothesized that allium vegetable intake is protective against stomach cancer development. Epidemiological studies of stomach cancer in relation to dietary habits suggest that consumption of allium vegetables, especially garlic and onions, may reduce the risk of stomach cancer (Table 1). Review and meta-analysis of garlic and stomach cancer studies have been published (Fleischauer et al., 2000; Fleischauer and Arab, 2001). In our study, we examined the relationship between allium vegetable intake and stomach cancer using a large matched case-control study conducted in Shanghai and Qingdao, China.
Table 1.
Epidemiological Studies of Allium Vegetables and Stomach Cancer
| Author, year | Country and Subjects | Vegetables | Consumption | OR (95% CI) or RR (95% CI) | Confounder Adjustment |
|---|---|---|---|---|---|
| Case-control | |||||
| Tajima, 1985 | Japan 93 cases, 186 controls |
Onion | <1 | Referent | Age and sex |
| 1–3 | 1.15 (NA) | ||||
| ≥4 | 2.13 (NA) | ||||
| You, 1989 | China 564 cases, 1131 controls |
Garlic | 0 kg/year | Referent | Age, sex income and other allium vegetables |
| 0.1–1.5 kg/year | 0.8 (0.5–1.2) | ||||
| >0.5 kg/year | 0.7 (0.4–1.0) | ||||
| Garlic Stalks | 0 kg/year | Referent | |||
| 0.1–0.5 kg/year | 1.0 (0.7–1.4) | ||||
| >0.5 kg/year | 0.7 (0.5–1.1) | ||||
| Onions | 0 kg/year | Referent | |||
| 0.1–0.5 kg/year | 0.8 (0.6–1.0) | ||||
| >0.5 kg/year | 1.0 (0.8–1.4) | ||||
| Scallions | ≤9.0 kg/year | Referent | |||
| 9.1–18.0 kg/year | 1.0 (0.8–1.3) | ||||
| >18.0 kg/year | 0.8 (0.5–1.0) | ||||
| Chinese Chives | <1.5 kg/year | Referent | |||
| 1.6–3.7kg/year | 0.7 (0.6–1.0) | ||||
| >3.7kg/year | 0.6 (0.4–0.8) | Age, sex, study are, SES, residence migration from South, family history of stomach cancer, and Quetelet index | |||
| Buiatti, 1989 | Italy 640 males, 376 females, 1159 controls |
Cooked garlic | Tertile low | Referent | |
| Medium | 0.6 (NA) | ||||
| High | 0.4 (NA) | ||||
| Raw onions | Tertile low | Referent | |||
| Medium | 0.8 (NA) | ||||
| High | 0.8 (NA) | ||||
| Onion/garlic condiments | Tertile low | Referent | |||
| Medium | 1.0 (NA) | ||||
| High | 0.8 (NA) | ||||
| Hansson, 1993 | Sweden 218 males, 120 females, 669 controls |
Garlic | 0 serv/mo | Referent | Age, sex and province |
| > 0 | 0.89 (0.64–1.24) | ||||
| Onion | 0 | Referent | |||
| 3 times/month | 0.84 (0.54–1.29) | ||||
| 7 times/month | 0.83 (0.56–1.22) | ||||
| 11 times/month | 0.84 (0.56–1.24) | ||||
| Leek | 0 | Referent | |||
| 0.9 times/month | 0.76 (0.56–1.04) | ||||
| 3 times/month | 0.63 (0.44–0.92) | ||||
| Ji, 1998 | China 770 males, 354 females, 1451 controls |
Allium vegetables | Male | Age, income, education, and smoking and alcohol drinking (males only) | |
| Q1 (low) | Referent | ||||
| Q2 | 1.3 (0.9–1.7) | ||||
| Q3 | 1.1 (0.8–1.5) | ||||
| Q4 (high) | 0.8 (0.6–1.1) | ||||
| Female | |||||
| Q1 (low) | Referent | ||||
| Q2 | 0.9 (0.6–1.3) | ||||
| Q3 | 0.8 (0.6–1.2) | ||||
| Q4 (high) | 0.8 (0.5–1.2) | ||||
| Gao, 1999 | China 110 males, 43 females, 234 controls |
Garlic Raw | <1 serv/mo | Referent | Age, sex, income, alcohol, smoking, tea intake, leftover gruel, pickled vegetables, meat, fruit, tomatoes, eggs, snap bean |
| ≥4 serv/mo | 0.31 (0.2–0.4) | ||||
| Onion | <1 serv/mo | Referent | |||
| ≥4 serv/mo | 0.17 (0.08–0.36) | ||||
| Welsh onion | <1 serv/mo | Referent | |||
| ≥4 serv/mo | 0.22 (0.15–0.31) | ||||
| Chinese chives | <1 serv/mo | Referent | |||
| ≥ 4 serv/mo | 0.40 (0.17–0.94) | ||||
| De Stefani, 2001 | 160 cases 320 controls |
Allium vegetables | 0.56 (0.34–0.92) | Age, gender, residence, total enery intake, fruit intake | |
| Cohort | |||||
| Dorant, 1996 | Netherlands 106 males, 33 females, 3123 subcohort members |
No supplements | Referent | Age, sex, alcohol, smoking, vit c, β-carotene, family history of stomach cancer and stomach disorders, other allium vegetables | |
| Garlic supplements only | 1.27 (0.6–2.6) | ||||
| Other supplements | Referent | ||||
| Garlic and other supplements Onion | 1.28 (0.5–3.7) | ||||
| 0 /day | Referent | ||||
| ≥0.5 /day | 0.5 (0.26–0.95) | ||||
Materials and Methods
Study Population. Shanghai is a municipality under the direct jurisdiction of the Chinese central government. It is the largest economic and trade center in the country and a comprehensive industrial base. Age-adjusted (world standard) incidence rate of stomach cancer was 38.2/100,000 during 1993–1994, which is the second highest cancer incidence in Shanghai after lung cancer (Jin et al., 1994). Qingdao (Tsingtao) city is located in the southeastern Shandong province in East China on the Yellow Sea. Qingdao is a major fishing and trade port of China and is the leading industrial city of Shandong. Description of the case-control study population (eligibility of study participants and response rates) and the collection of epidemiological data have been published elsewhere (Yu et al., 1995). Briefly, a population-based case-control study was conducted in Shanghai and Qingdao in 1991. Data included interview questionnaires and medical record reviews. Eligible cases were individuals with incident primary stomach cancer who were under 80 years of age, and who were diagnosed between October 1991 and December 1993 among residents of Shanghai and Qingdao. Case-finding was organized by the Health Station for Disease Control in each of the study areas. Community controls were selected from the same street or community where the respective cases resided and were matched to the cases by age (within 3 years) and gender. Among the eligible persons, the one living in the nearest residence to that of the case was selected as control. The Shanghai study included 750 stomach cancer cases (128 cardia and 622 distal), and 750 controls, and the Qingdao study included 201 stomach cancer cases and 201 controls.
Epidemiological data collection. Interviews with the study participant were conducted by trained interviewers at the participant’s home using a structured, pilot-tested questionnaire. Information collected included: (1) demographic characteristics, (2) medical history of subjects and family members, (3) tobacco use, (4) alcohol consumption, and (5) dietary information including tea drinking, vegetable and fruit intake, salt consumption, and allium vegetable consumption. The participants were asked questions about dietary factors in the 80s (i.e. approximately 10 years before diagnosis for cases). Allium vegetable data included frequency of allium vegetable intake and family monthly consumption of each vegetable in kilograms. Allium vegetables included were onion leaves, scallions, garlic stalks, garlic, onions and Chinese chives.
Statistical analysis. Minimum detectable ORs were calculated for various exposure proportions in controls, using a two-sided test, 0.05 α level and 80% power for univariate analysis. Our Shanghai study (with 750 cases and 750 controls) was adequate to detect ORs ranging from 0.52 to 0.72, and when frequency of exposure was 5–55% in controls. Because the etiology of the proximal part of stomach cancer (cardia and gastroesophageal junction) is different from that in other parts of the stomach (distal), and because there may be subsite-specific differences in stomach cancer risk associated with allium vegetable consumption (Dorant et al., 1996), we evaluated whether the role of allium vegetable consumption varies in cardia and distal cancer in the Shanghai study. Our Shanghai study (128 cardia cases and 622 distal cases) was adequate to detect ORs ranging from 0.24 to 0.44 for cardia cancer, and ORs ranging from 0.49 to 0.69 for distal cancer, and when frequency of exposure was 5–55% in controls. Two types of analyses were conducted, first for all stomach cancer, then stratified by tumor location (cardia and distal). In Qingdao, because the stomach cancer cases were not distinguished by tumor location, only all case analyses were performed. Based on the power calculation for univariate analysis, our sample size (201 cases and 201 controls) was adequate to detect OR ranging from 0.31 to 0.52, and when the frequency of exposure was between 5–55% in controls. Pack-years of smoking were computed by multiplying the number of packs of cigarettes smoked per day by the total years of smoking. Body mass index (BMI in kg/m2) was calculated by dividing the body weight by the height squared. Intake of vegetables and fruit were categorized into “never”, “occasional” d(few times/year), “less often” (few times/month), and “often” (few times/week). For each type of allium vegetable (onion leaves, scallions, garlic stalks, garlic, onions and Chinese chives), frequencies of intake were collected, using the same categories as noted above. Because of the limited numbers in these categories, “occasional”, and “less often” categories were collapsed. The categorization of monthly intake (in kg) was based on the distributions in the controls. Median distribution of the controls was used as the cutoff point. Monthly intake was categorized into “never”, “low” and “high”. The “never” category was used as a referent. Conditional logistic regression maintaining the matched-pairs was used to estimate the effect of allium vegetable intake on stomach cancer risk while adjusting for education, BMI, pack-years of smoking, alcohol drinking, salt intake, and vegetable and fruit intake. Trend test was computed by treating the categorical variable as a continuous predictor in the conditional logistic regression models. All analyses were conducted in SAS version 8.12.
Results
Table 2 shows the means and distribution of demographic and potential confounding variables by case-control status in Shanghai and Qingdao. As expected in an age- and gender-matched case-control study, similar means of age and gender distribution were observed in cases and controls (62.9 years and 62.4 years in Shanghai and 61.0 years and 60.8 years in Qingdao). There were more men (64–71%) than women (29–36%) in both places. In Shanghai, cardia cases were generally older (mean age=66 years) and more often men (69%) than distal cases (mean age=62 years; 63% men). Most people had education ≤ high school (>93% in both cases and controls). The mean of BMI was similar in cases and controls, with slightly higher BMI in Qingdao than in Shanghai. More smokers were found in the stomach cancer group (48–64%) than in the controls (43–54%), with a slightly higher proportion of smokers in distal (49%) than in cardia cancer (45%) in Shanghai. Mean of pack-years was higher in stomach cancer cases than in controls. The majority of study participants were non-drinkers (>62% in cases and in controls). Similar means of salt intake were observed in cases and controls. Most people were frequent vegetable eaters, with a higher proportion of frequent eaters in controls than in cases. Controls tended to eat fruits more frequently than stomach cancer cases, and slightly more cardia cancer cases (58%) ate fruits more frequently than distal cases (54%). The proportion of people who ate fruits more frequently was lower in Qingdao than in Shanghai (35% versus 65% among controls). Based on prior knowledge of stomach cancer risk factors and our confounding assessment, education, BMI, pack-years of smoking, alcohol drinking, salt intake, and fruit and vegetable intake were adjusted for when estimating the effects of allium vegetables on stomach cancer risk.
Table 2.
Means and Distribution of Demographic and Other Selected Variables among Stomach Cancer Cases and Controls in Shanghai and Qingdao
| Variables | SHANGHAI | QINGDAO | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| All Stomach cancer | Cardia Cancer | Distal Cancer | Control | All Stomach cancer | Control | |
| Mean ± SD | ||||||
| Age (years) | 62.9 ± 11.5 | 66.2 ± 9.4 | 62.2 ± 11.8 | 62.4 ± 11.5 | 61.0 ± 13.0 | 60.8 ± 13.2 |
| BMI (kg/m2) | 21.8 ± 2.8 | 22.3 ± 2.7 | 21.7 ± 2.8 | 22.2 ± 2.8 | 23.1 ± 2.9 | 23.6 ± 3.0 |
| Pack-years of smoking | 15.7 ± 21.5 | 15.2 ± 22.4 | 15.8 ± 21.4 | 13.6 ± 20.4 | 21.8 ± 22.8 | 17.7 ± 23.0 |
| Salt intake (kg/year) | 3.4 ± 1.6 | 3.0 ± 1.3 | 3.5 ± 1.6 | 3.5 ± 1.7 | 4.6 ± 1.3 | 4.6 ± 1.3 |
| N (%) | ||||||
| Gender | ||||||
| Male | 478 (63.7) | 88 (68.8) | 390 (62.7) | 478 (63.7) | 143 (71.1) | 143 (71.1) |
| Female | 272 (36.3) | 40 (31.3) | 232 (37.3) | 272 (36.3) | 58 (28.9) | 58 (28.9) |
| Education | ||||||
| ≤High School | 715 (95.3) | 120 (93.8) | 595 (95.7) | 714 (95.2) | 186 (92.5) | 195 (97.0) |
| > High School | 35 (4.7) | 8 (6.3) | 27 (4.3) | 36 (4.8) | 15 (7.5) | 6 (3.0) |
| Smoking | ||||||
| No | 387 (51.7) | 70 (54.7) | 317 (51.1) | 430 (57.4) | 72 (35.8) | 93 (46.3) |
| Yes | 362 (48.3) | 58 (45.3) | 304 (48.9) | 319 (42.6) | 129 (64.2) | 108 (53.7) |
| Alcohol drinking | ||||||
| No | 509 (67.9) | 88 (68.8) | 421 (67.7) | 516 (68.8) | 124 (61.7) | 124 (61.7) |
| Yes | 241 (32.1) | 40 (31.3) | 201 (32.3) | 234 (31.2) | 77 (38.3) | 77 (38.3) |
| Vegetable intake* | ||||||
| Less frequent | 16 (2.1) | 3 (2.4) | 13 (2.1) | 4 (0.5) | 20 (10.0) | 13 (6.5) |
| Frequent | 731 (97.9) | 124 (97.6) | 607 (97.9) | 739 (99.5) | 181 (90.0) | 188 (93.5) |
| Fruit intake* | ||||||
| Less frequent | 338 (45.4) | 53 (41.7) | 285 (46.1) | 264 (35.5) | 143 (71.1) | 130 (64.7) |
| Frequent | 407 (54.6) | 74 (58.3) | 333 (53.9) | 479 (64.5) | 58 (28.9) | 71 (35.3) |
Less frequent (never and occasional), frequent (less often and often).
The multivariate ORs and 95% CIs for individual allium vegetables and stomach cancer risk in Shanghai are presented in Table 3. All of the highest frequency categories of “often” for each allium vegetable showed protective ORs, except for scallions. A negative dose-response relationship was only observed between onion intake and stomach cancer risk (P for trend=0.04). The trend remains significant (P for trend=0.0461) after adjusting for the other allium vegetables in the same model (data not shown). We did not observe a clear relationship between frequency intake of allium vegetables and cardia cancer risk, which may be due partly to small number in some of the categories. For distal cancer, each allium vegetable “often” category had an OR below 1, except for scallions. A negative dose-response pattern was observed between increasing intake of onions with decreasing distal cancer risk (P for trend=0.04).
Table 3.
Frequency of Allium Vegetable Intake and Stomach Cancer Risk in Shanghai
| Vegetable | All Stomach Cancer | Cardia | Distal | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Cases | Controls | OR*&95% CI | Cases | Controls | OR*&95% CI | Cases | Controls | OR*&95% CI | |
| Onion Leaves | |||||||||
| Never | 387 (51.6) | 408 (54.4) | 1.00 | 51 (39.8) | 52 (40.6) | 1.00 | 336 (54.0) | 356 (57.2) | 1.00 |
| Occasional | 344 (45.9) | 321 (42.8) | 1.22 (0.94–1.57) | 72 (56.3) | 71 (55.5) | 0.91 (0.45–1.83) | 272 (43.7) | 250 (40.2) | 1.25 (0.95–1.66) |
| Often | 19 (2.5) | 21 (2.8) | 0.88 (0.43–1.82) | 5 (3.91) | 5 (3.91) | 0.94 (0.20–4.39) | 14 (2.3) | 16 (2.6) | 0.80 (0.35–1.84) |
| P for trend | 0.3015 | 0.8306 | 0.2947 | ||||||
| Scallions | |||||||||
| Never | 532 (70.9) | 526 (70.1) | 1.00 | 68 (53.1) | 71 (55.5) | 1.00 | 464 (74.6) | 455 (73.2) | 1.00 |
| Occasional | 197 (26.3) | 209 (27.9) | 0.87 (0.64–1.19) | 55 (43.0) | 53 (41.4) | 0.93 (0.45–1.90) | 142 (22.8) | 156 (25.1) | 0.81 (0.57–1.15) |
| Often | 21 (2.8) | 15 (2.0) | 1.48 (0.69–3.19) | 5 (3.9) | 4 (3.1) | 1.13 (0.24–5.36) | 16 (2.6) | 11 (1.8) | 1.49 (0.60–3.74) |
| P for trend | 0.9279 | 0.9615 | 0.6438 | ||||||
| Garlic Stalks | |||||||||
| Never | 269 (35.9) | 261 (34.8) | 1.00 | 39 (30.5) | 52 (40.6) | 1.00 | 230 (37.0) | 209 (33.6) | 1.00 |
| Occasional | 452 (60.3) | 457 (60.9) | 0.94 (0.74–1.21) | 85 (66.4) | 69 (53.9) | 1.56 (0.77–3.18) | 367 (59.0) | 388 (62.4) | 0.84 (0.64–1.10) |
| Often | 29 (3.9) | 32 (4.3) | 0.77 (0.43–1.40) | 4 (3.1) | 7 (5.5) | 0.81 (0.16–3.96) | 25 (4.0) | 25 (4.0) | 0.77 (0.40–1.47) |
| P for trend | 0.4413 | 0.4912 | 0.1806 | ||||||
| Garlic | |||||||||
| Never | 398 (53.1) | 411 (54.8) | 1.00 | 59 (46.1) | 73 (57.0) | 1.00 | 339 (54.5) | 338 (54.3) | 1.00 |
| Occasional | 331 (44.1) | 309 (41.2) | 1.11 (0.87–1.41) | 68 (53.1) | 50 (39.1) | 1.85 (0.96–3.58) | 263 (42.3) | 259 (41.6) | 0.99 (0.76–1.29) |
| Often | 21 (2.8) | 30 (4.0) | 0.68 (0.37–1.26) | 1 (0.8) | 5 (3.9) | 0.27 (0.03–2.76) | 20 (3.2) | 25 (4.0) | 0.72 (0.38–1.38) |
| P for trend | 0.9682 | 0.3216 | 0.5636 | ||||||
| Onions | |||||||||
| Never | 262 (35.5) | 266 (36.0) | 1.00 | 45 (35.2) | 47 (36.7) | 1.00 | 217 (34.9) | 219 (35.2) | 1.00 |
| Occasional | 305 (40.7) | 252 (33.6) | 1.20 (0.42–1.73) | 61 (47.7) | 55 (43.0) | 0.94 (0.38–2.35) | 244 (39.2) | 197 (31.7) | 1.27 (0.86–1.90) |
| Often | 183 (24.8) | 232 (31.4) | 0.66 (0.42–1.02) | 22 (17.2) | 26 (20.3) | 0.70 (0.22–2.23) | 161 (25.9) | 206 (33.1) | 0.62 (0.38–1.01) |
| P for trend | 0.0403 | 0.5175 | 0.0367 | ||||||
| Chinese Chives | |||||||||
| Never | 110 (14.7) | 116 (15.5) | 1.00 | 17 (13.3) | 17 (13.3) | 1.00 | 93 (15.0) | 99 (15.9) | 1.00 |
| Occasional | 553 (73.7) | 514 (68.5) | 1.12 (0.81–1.56) | 90 (70.3) | 83 (64.8) | 0.72 (0.23–2.29) | 463 (74.4) | 431 (69.3) | 1.15 (0.81–1.62) |
| Often | 87 (11.6) | 120 (16.0) | 0.61 (0.38–0.99) | 21 (16.4) | 28 (21.9) | 0.28 (0.07–1.19) | 66 (10.6) | 92 (14.8) | 0.66 (0.39–1.11) |
| P for trend | 0.1020 | 0.0641 | 0.2525 | ||||||
Adjusted for matching variables (age, gender), education, BMI, pack-years of smoking, alcohol drinking, salt intake, and vegetable and fruit intake
Table 4 shows corresponding ORs and 95% CIs for cardia cancer and distal cancer according to quartile ranking of monthly allium vegetable consumption in Shanghai. No clear dose-response pattern was observed between monthly allium vegetable consumption and cardia cancer. However, we observed negative dose-response patterns between increasing intake of scallions (P=0.09) and garlic stalks (P=0.04) with decreasing distal cancer risk. The trend for garlic stalks remains significant (P=0.04) even after adjusting for the other allium vegetables in the model (data not shown). The ORs of the “high” intake category when compared to the “never” category were 0.60 (95% CI: 0.37–0.99) for scallions and 0.72 (95% CI: 0.52–0.99) for garlic stalks.
Table 4.
Monthly Allium Vegetable Intake and Stomach Cancer Risk in Shanghai
| Vegetable | Cardia Cases | Controls | OR* & 95% CI | Distal Cases | Controls | OR* & 95% CI |
|---|---|---|---|---|---|---|
| Onion Leaves | ||||||
| Never | 51 (39.8) | 54 (42.2) | 1.00 | 336 (54.0) | 349 (56.1) | 1.00 |
| Low | 38 (29.7) | 35 (27.3) | 1.08 (0.49–2.41) | 187 (30.1) | 157 (25.2) | 1.28 (0.95–1.73) |
| High | 39 (30.5) | 39 (30.5) | 0.97 (0.42–2.20) | 99 (15.9) | 116 (18.7) | 0.92 (0.63–1.34) |
| P for trend | 0.9235 | 0.9225 | ||||
| Scallions | ||||||
| Never | 68 (54.4) | 71 (55.5) | 1.00 | 464 (74.6) | 455 (73.2) | 1.00 |
| Low | 19 (14.8) | 26 (20.3) | 0.52 (0.21–1.32) | 99 (15.9) | 89 (14.3) | 1.00 (0.68–1.48) |
| High | 41 (32.0) | 31 (24.2) | 1.32 (0.59–2.93) | 59 (9.5) | 78 (12.5) | 0.60 (0.37–0.99) |
| P for trend | 0.4489 | 0.0859 | ||||
| Garlic Stalks | ||||||
| Never | 39 (30.5) | 52 (40.6) | 1.00 | 230 (37.0) | 209 (33.6) | 1.00 |
| Low | 39 (30.5) | 41 (32.0) | 1.13 (0.51–2.48) | 226 (36.3) | 208 (33.4) | 0.94 (0.69–1.27) |
| High | 50 (39.1) | 35 (27.3) | 2.00 (0.89–4.52) | 166 (26.7) | 205 (33.0) | 0.72 (0.52–0.99) |
| P for trend | 0.0874 | 0.0409 | ||||
| Garlic | ||||||
| Never | 59 (46.1) | 73 (57.0) | 1.00 | 339 (54.5) | 338 (54.3) | 1.00 |
| Low | 37 (28.9) | 29 (22.7) | 1.66 (0.79–3.46) | 173 (27.8) | 154 (24.8) | 1.06 (0.79–1.42) |
| High | 32 (25.0) | 26 (20.3) | 1.71 (0.77–3.80) | 110 (17.7) | 130 (20.9) | 0.85 (0.61–1.17) |
| P for trend | 0.1486 | 0.3974 | ||||
| Onions | ||||||
| Never | 45 (35.2) | 47 (36.7) | 1.00 | 217 (34.9) | 219 (35.2) | 1.00 |
| Low | 46 (35.9) | 42 (32.8) | 0.97 (0.38–2.47) | 211 (33.9) | 198 (31.8) | 1.12 (0.75–1.67) |
| High | 37 (28.9) | 39 (30.5) | 0.74 (0.25–2.18) | 194 (31.2) | 205 (33.0) | 0.95 (0.61–1.49) |
| P for trend | 0.5336 | 0.7123 | ||||
| Chinese Chives | ||||||
| Never | 17 (13.3) | 17 (13.3) | 1.00 | 93 (15.0) | 99 (15.9) | 1.00 |
| Low | 48 (37.5) | 45 (42.2) | 0.51 (0.16–1.64) | 276 (44.4) | 263 (42.3) | 1.13 (0.78–1.62) |
| High | 63 (49.2) | 57 (44.5) | 0.76 (0.23–2.44) | 253 (40.7) | 260 (41.8) | 1.02 (0.70–1.50) |
| P for trend | 0.8579 | 0.9395 | ||||
Adjusted for matching variables (age, gender), education, BMI, pack-years of smoking, alcohol drinking, salt intake, and vegetable and fruit intake
Table 5 shows the ORs and 95% CIs for individual allium vegetables and stomach cancer risk in Qingdao. In this population, more people were frequent consumers of allium vegetables than in Shanghai. In Qingdao controls, 57% of people categorized themselves as “often” onion eaters as opposed to only 31% in Shanghai controls. In addition, 54% of Qingdao controls categorized themselves as “often” garlic consumers as opposed to only 4% of Shanghai controls. We observed a dose-response pattern between onions and stomach cancer risk in Qingdao (P for trend=0.02), although the trend became borderline significant (P=0.07) after adjusting for other allium vegetables in the model. The OR of the “occasional” category when compared to the “never” was 0.21 (95% CI: 0.06–0.79). The OR of the “often” category when compared to the “never” was 0.14 (95% CI: 0.03–0.71). Slight dose-response relationships were also observed for garlic (P for trend=0.06) and garlic stalks (P for trend=0.07). Table 5 also shows the ORs and 95% CIs of monthly allium vegetable intake in Qingdao. We observed an association between garlic stalk and onion intake with stomach cancer risk. The “never” category was used as a referent. The category of “low” intake of garlic stalks yielded an OR of 0.19 (95% CI: 0.08–0.45) and the “high” intake of garlic stalks yielded an OR of 0.30 (95% CI: 0.13–0.73). For onions, a negative dose-response pattern was observed (P for trend=0.004) and remains strongly significant after adjusting for the other allium vegetables (P=0.007). The ORs were 0.29 (95% CI: 0.09–0.94) for the “low” intake category and 0.07 (95% CI: 0.01–0.52) for the “high” intake category.
Table 5.
Frequency and Monthly Allium Vegetable Intake and Stomach Cancer Risk in Qingdao
| Frequency | Cases | Controls | OR* & 95% CI | Monthly Amount | Cases | Controls | OR* & 95% CI |
|---|---|---|---|---|---|---|---|
| Onion Leaves | Onion leaves | ||||||
| Never | 36 (17.9) | 39 (19.4) | 1.00 | Never | 36 (17.9) | 39 (19.4) | 1.00 |
| Occasional | 154 (76.6) | 146 (72.6) | 1.20 (0.67–2.15) | Low | 115 (57.2) | 105 (52.2) | 1.29 (0.70–2.37) |
| Often | 11 (5.5) | 16 (8.0) | 0.78 (0.30–2.05) | High | 50 (24.9) | 57 (28.4) | 0.96 (0.50–1.86) |
| P for trend | 0.9087 | P for trend | 0.7539 | ||||
| Scallions | Scallions | ||||||
| Never | 47 (23.4) | 34 (16.9) | 1.00 | Never | 47 (23.4) | 34 (16.9) | 1.00 |
| Occasional | 75 (37.3) | 89 (44.3) | 0.49 (0.25–0.93) | Low | 69 (34.3) | 83 (41.3) | 0.49 (0.26–0.94) |
| Often | 79 (39.3) | 78 (38.8) | 0.68 (0.33–1.42) | High | 85 (42.3) | 84 (41.8) | 0.68 (0.33–1.40) |
| P for trend | 0.4086 | 0.4204 | |||||
| Garlic Stalks | Garlic Stalks | ||||||
| Never | 39 (19.4) | 14 (7.0) | 1.00 | Never | 39 (19.4) | 14 (7.0) | 1.00 |
| Occasional | 115 (57.2) | 140 (69.7) | 0.21 (0.09–0.50) | Low | 75 (37.3) | 101 (50.3) | 0.19 (0.08–0.45) |
| Often | 47 (23.4) | 47 (23.4) | 0.30 (0.12–0.77) | High | 87 (43.3) | 86 (42.8) | 0.30 (0.13–0.73) |
| P for trend | 0.0660 | P for trend | 0.1462 | ||||
| Garlic | Garlic | ||||||
| Never | 16 (8.0) | 11 (5.5) | 1.00 | Never | 16 (8.0) | 11 (5.5) | 1.00 |
| Occasional | 96 (47.8) | 81 (40.3) | 0.71 (0.27–1.88) | Low | 100 (49.8) | 96 (47.8) | 0.66 (0.25–1.76) |
| Often | 89 (44.3) | 109 (54.2) | 0.45 (0.15–1.30) | High | 85 (42.3) | 94 (46.8) | 0.58 (0.20–1.70) |
| P for trend | 0.0606 | P for trend | 0.3812 | ||||
| Onions | Onions | ||||||
| Never | 19 (9.5) | 8 (4.0) | 1.00 | Never | 19 (9.5) | 8 (4.0) | 1.00 |
| Occasional | 72 (35.8) | 78 (38.8) | 0.21 (0.06–0.79) | Low | 83 (41.3) | 87 (43.5) | 0.29 (0.09–0.94) |
| Often | 110 (54.7) | 115 (57.2) | 0.14 (0.03–0.71) | High | 99 (49.3) | 105 (52.5) | 0.07 (0.01–0.52) |
| P for trend | 0.0204 | P for trend | 0.0042 | ||||
| Chinese Chives | Chinese Chives | ||||||
| Never | 23 (11.4) | 13 (6.5) | 1.00 | Never | 23 (11.4) | 13 (6.5) | 1.00 |
| Occasional | 147 (73.1) | 150 (74.6) | 0.51 (0.27–1.11) | Low | 93 (46.3) | 95 (47.3) | 0.48 (0.21–1.09) |
| Often | 31 (15.4) | 38 (18.9) | 0.44 (0.18–1.08) | High | 85 (42.3) | 93 (46.3) | 0.49 (0.22–1.10) |
| P for trend | 0.1060 | P for trend | 0.2139 |
Adjusted for matching variables (age, gender), education, BMI, pack-year of smoking, alcohol drinking, salt intake, and vegetable and fruit intake
Discussion
We found onion consumption to be inversely associated with stomach cancer risk in Shanghai and Qingdao. The highest category of onion intake was associated with reduced stomach cancer risk [Shanghai’s OR=0.66 (95% CI: 0.42–1.02) and Qingdao’s OR=0.14 (95% CI: 0.03–0.71)]. In addition, after adjusting for potential confounders and the other allium vegetables, a negative dose-response relationship was observed between onion consumption and stomach cancer risk in both populations. A decreased risk of distal stomach cancer was observed with increasing onion consumption. In Shanghai, based on monthly consumption in kg, we found that intake of garlic stalks was associated with a decreased risk of distal stomach cancer, and with a clear dose-response relationship. In Qingdao, we consistently observed an inverse association between intake of garlic stalks, garlic, and onions and stomach cancer. A strong dose-response relationship was observed between monthly intake of onions and stomach cancer.
Most epidemiological studies, including a cohort study, have suggested a decreased risk of stomach cancer with increasing consumption of garlic, onions, or related allium vegetables. One study conducted in a high-risk area of China found that onions were protective against stomach cancer; the OR of the highest frequency intake category was 0.17 (95% CI: 0.08–0.36), which is very similar to our observation in Qingdao (Gao et al., 1999). The only cohort study of allium vegetables (the Netherland Cohort Study) to date showed a strong inverse association between onion consumption and stomach cancer incidence (Dorant et al., 1996). The observation was restricted to distal cancer only. Our results were consistent with that study, as we observed a negative dose-response relationship between onions and distal stomach cancer only. One study in Japan found a positive association between high onion intake and stomach cancer, which may be explained by the selection of hospital patients with severe gastric diseases as controls (Tajima et al., 1985).
We did not observe a clear association between allium vegetable consumption and cardia cancer risk. None of the previous studies, aside from the cohort study, separated stomach cancer into subsites (cardia and distal). It is important to take into account tumor site, because it has been suggested that the etiology of cancer in the proximal part of the stomach (cardia and gastroesophageal junction) is different from that in the other part of the stomach (distal). Thus there may be subsite-specific differences in stomach cancer risk associated with allium vegetable consumption (Dorant et al., 1996). In addition, there is a possibility that cardia cancer has an etiology that is more similar to esophageal adenocarcinoma than to distal cancer (Wang et al., 1986). A large case-control study of esophageal cancer also conducted in Shanghai found no association between allium vegetable consumption and risk of esophageal cancer in men or women (Gao et al., 1994). However, a case-control study in a high-risk area in China found that allium vegetables (garlic, onions, Welsh onions and Chinese chives) were inversely associated with esophageal cancer risk (Gao et al., 1999). Whether the anticarcinogenic compounds in allium vegetables have a different effect on different areas of the stomach requires further investigation.
Several experimental studies have shown that extracts from garlic and onions have anticancer properties (for recent reviews, see Herman-Antosiewicz and Singh, 2004; Milner, 2001). Belman was the first to find that tumor promotion was inhibited by onion and garlic oil in a dose-response manner in the skin (Belman, 1983). Interestingly, Sparnins et al. found that organosulfur compounds from garlic and onions have inhibitory effects on benzo(a)pyrene-induced cancer of the mouse forestomach (Sparnins et al., 1988). They also found that these compounds can induce glutathione S-transferase activity in the forestomach. In vitro assays also showed that allium vegetables or their bioactive constituents could inhibit proliferation and cause cycle arrest and/or apoptosis in several cancer cells (Nakagawa et al., 2001; Xiao et al., 2004; Kwon et al., 2002; Sakamoto et al., 1997; Sundaram and Milner, 1996). In vitro, sulfur compounds in garlic and onions have been shown to depress nitrosamine formation and bioactivation (Dion et al., 1997). Studies have shown that they induce the antitumorigenic NSAID-activated gene by a p-53-dependant mechanism (Bottone et al., 2002), activate Bcl-2, Bax and p-53 genes (Hong et al., 2000), and protect against induced mammary epithelial cell DNA adduct formation (Schaffer et al., 1997). Sivan et al. have demonstrated in vitro that H. pylori is susceptible to garlic extract at a fairly moderate concentration. Even some antibiotic-resistant H. pylori strains are susceptible to garlic (Sivan et al., 1997). The results of experimental studies support epidemiological findings that higher intake of allium vegetables is associated with a reduction in the risks of some cancers.
Although our findings are consistent with most previous studies of stomach cancer, several potential limitations of our study should be discussed. Disease and exposure misclassification may exist in our study. In our study, most of the stomach cancer cases were pathologically, surgically or endoscopically diagnosed (85%) (Yu et al., 1995); therefore, disease misclassification is likely minimal. Misclassification of exposure, however, is more likely to have occurred. Differential recall of dietary intake due to awareness of disease status may be possible. In addition, recall of dietary intake can be influenced by current dietary habits, which may be different from those in the relevant reference period due to symptoms of disease. Our questionnaire asked about dietary habits in the 80’s, prior to the disease diagnoses, which may have minimized potential misclassification bias for allium vegetable intake. Data of allium vegetable consumption collected included intake frequency and monthly consumption in kg for each type of vegetable. Because it is very difficult to remember precisely how often each vegetable was consumed, misclassification of exposure is likely to have occurred. All allium vegetable consumption was categorized into three groups because of the low numbers of consumers in some of the categories. This crude categorization of consumption may have limited the ability to detect a relationship between higher levels of vegetable intake and stomach cancer risk. In this study, H. pylori infection status was not available. While H. pylori infection is an established risk factor for stomach cancer, garlic extract has been shown to have antibacterial activity against H. Pylori; thus H. pylori infection may act more as an intermediate than as a confounder. Statistical adjustment for H. pylori infection in the analyses may not be appropriate.
In conclusion, our large population based case-control studies of stomach cancer confirmed previous reports of protective effects associated with allium vegetable consumption. Future epidemiological studies of allium vegetables and stomach cancer should separate stomach cancer into cardia and non-cardia cancer. Prospective studies are needed to confirm the protectiveness of garlic or onions against stomach cancer because of their great potential as low-cost remedies to reduce stomach cancer incidence among high-risk populations.
Acknowledgments
This work is supported in part by the NIH National Cancer Institute, Department of Health and Human Services, Grants CA77954, CA09142, CA16042, and CA52560, and by the Ann Fitzpatrick Alper Program in Environmental Genomics at the Jonsson Comprehensive Cancer Center, UCLA.
References
- Belman S. Onion and garlic oils inhibit tumor promotion. Carcinogenesis. 1983;4:1063–5. doi: 10.1093/carcin/4.8.1063. [DOI] [PubMed] [Google Scholar]
- Bottone FG, Jr, Baek SJ, Nixon JB, Eling TE. Diallyl disulfide (DADS) induces the antitumorigenic NSAID-activated gene (NAG-1) by a p53-dependent mechanism in human colorectal HCT 116 cells. J Nutr. 2002;132:773–8. doi: 10.1093/jn/132.4.773. [DOI] [PubMed] [Google Scholar]
- Buiatti E, Palli D, Decarli A, et al. A case-control study of stomach cancer and diet in Italy. Int J Cancer. 1989;44:611–6. doi: 10.1002/ijc.2910440409. [DOI] [PubMed] [Google Scholar]
- Canizares P, Gracia I, Gomez LA, et al. Thermal degradation of allicin in garlic extracts and its implication on the inhibition of the in-vitro growth of Helicobacter pylori. Biotechnol Prog. 2004;20:32–7. doi: 10.1021/bp034135v. [DOI] [PubMed] [Google Scholar]
- De Stefani E, Correa P, Boffetta P, et al. Plant foods and risk of gastric cancer: a case- control study in Uruguay. Eur J Cancer Prev. 2001;10:357–64. doi: 10.1097/00008469-200108000-00009. [DOI] [PubMed] [Google Scholar]
- Dion ME, Agler M, Milner JA. S-allyl cysteine inhibits nitrosomorpholine formation and bioactivation. Nutr Cancer. 1997;28:1–6. doi: 10.1080/01635589709514545. [DOI] [PubMed] [Google Scholar]
- Dorant E, van den Brandt PA, Goldbohm RA, Sturmans F. Consumption of onions and a reduced risk of stomach carcinoma. Gastroenterology. 1996;110:12–20. doi: 10.1053/gast.1996.v110.pm8536847. [DOI] [PubMed] [Google Scholar]
- Fleischauer AT, Poole C, Arab L. Garlic consumption and cancer prevention: meta- analyses of colorectal and stomach cancers. Am J of Clin Nutr. 2000;72:1047–52. doi: 10.1093/ajcn/72.4.1047. [DOI] [PubMed] [Google Scholar]
- Fleischauer AT, Arab L. Garlic and cancer: a critical review of the epidemiologic literature. J Nutr. 2001;131:1032S–40S. doi: 10.1093/jn/131.3.1032S. [DOI] [PubMed] [Google Scholar]
- Gao YT, McLaughlin JK, Gridley G, et al. Risk factors for esophageal cancer in Shanghai, China. II. Role of diet and nutrients. Int J Cancer. 1994;58:197–202. doi: 10.1002/ijc.2910580209. [DOI] [PubMed] [Google Scholar]
- Gao CM, Takezaki T, Ding JH, Li MS, Tajima K. Protective effect of allium vegetables against both esophageal and stomach cancer: a simultaneous case-referent study of a high-epidemic area in Jiangsu Province, China. Jpn J Cancer Res. 1999;90:614–21. doi: 10.1111/j.1349-7006.1999.tb00791.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hansson LE, Nyren O, Bergstrom R, et al. Diet and risk of gastric cancer. A population- based case-control study in Sweden. Int J Cancer. 1993;55:181–9. doi: 10.1002/ijc.2910550203. [DOI] [PubMed] [Google Scholar]
- Herman-Antosiewicz A, Singh SV. Signal transduction pathways leading to cell cycle arrest and apoptosis induction in cancer cells by Allium vegetable-derived organosulfur compounds: a review. Mutat Res. 2004;555:121–31. doi: 10.1016/j.mrfmmm.2004.04.016. [DOI] [PubMed] [Google Scholar]
- Hong YS, Ham YA, Choi JH, Kim J. Effects of allyl sulfur compounds and garlic extract on the expression of Bcl-2, Bax, and p53 in non small cell lung cancer cell lines. Exp Mol Med. 2000;32:127–34. doi: 10.1038/emm.2000.22. [DOI] [PubMed] [Google Scholar]
- Ji BT, Chow WH, Yang G, et al. Dietary habits and stomach cancer in Shanghai, China. Int J Cancer. 1998;76:659–64. doi: 10.1002/(sici)1097-0215(19980529)76:5<659::aid-ijc8>3.0.co;2-p. [DOI] [PubMed] [Google Scholar]
- Jin F, Devesa SS, Chow WH, Zheng, et al. Cancer incidence trends in urban shanghai, 1972–1994: an update. Int J Cancer. 1999;83:435–40. doi: 10.1002/(sici)1097-0215(19991112)83:4<435::aid-ijc1>3.0.co;2-j. [DOI] [PubMed] [Google Scholar]
- Jonkers D, van den Broek E, van Dooren I, et al. Antibacterial effect of garlic and omeprazole on Helicobacter pylori. J Antimicrob Chemother. 1999;43:837–9. doi: 10.1093/jac/43.6.837. [DOI] [PubMed] [Google Scholar]
- Kwon KB, Yoo SJ, Ryu DG, et al. Induction of apoptosis by diallyl disulfide through activation of caspase-3 in human leukemia HL-60 cells. Biochem Pharmacol. 2002;63:41–7. doi: 10.1016/s0006-2952(01)00860-7. [DOI] [PubMed] [Google Scholar]
- Lawson LD, Wang ZJ, Hughes BG. Identification and HPLC quantitation of the sulfides and dialk(en)yl thiosulfinates in commercial garlic products. Planta Medica. 1991;57:363–70. doi: 10.1055/s-2006-960119. [DOI] [PubMed] [Google Scholar]
- Milner JA. Mechanisms by which garlic and allyl sulfur compounds suppress carcinogen bioactivation. Garlic and carcinogenesis. Adv Exp Med Biol. 2001;492:69–81. doi: 10.1007/978-1-4615-1283-7_7. [DOI] [PubMed] [Google Scholar]
- Nakagawa H, Tsuta K, Kiuchi K, et al. Growth inhibitory effects of diallyl disulfide on human breast cancer cell lines. Carcinogenesis. 2001;22:891–7. doi: 10.1093/carcin/22.6.891. [DOI] [PubMed] [Google Scholar]
- Ohta R, Yamada N, Kaneko H, et al. In vitro inhibition of the growth of Helicobacter pylori by oil-macerated garlic constituents. Antimicrob Agents Chemother. 1999;43:1811–2. doi: 10.1128/aac.43.7.1811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Reddy BS, Rao CV, Rivenson A, Kelloff G. Chemoprevention of colon carcinogenesis by organosulfur compounds. Cancer Res. 1993;53:3493–8. [PubMed] [Google Scholar]
- Sakamoto K, Lawson LD, Milner JA. Allyl sulfides from garlic suppress the in vitro proliferation of human A549 lung tumor cells. Nutr Cancer. 1997;29:152–6. doi: 10.1080/01635589709514617. [DOI] [PubMed] [Google Scholar]
- Schaffer EM, Liu JZ, Milner JA. Garlic powder and allyl sulfur compounds enhance the ability of dietary selenite to inhibit 7,12-dimethylbenz[a]anthracene-induced mammary DNA adducts. Nutr Cancer. 1997;27:162–8. doi: 10.1080/01635589709514519. [DOI] [PubMed] [Google Scholar]
- Sivan P, Lampe J, Lunes B, Swanzy S, Potter J. Helicobacter pylori in vitro susceptibility to garlic (Allium sativum) extract. Nutr Cancer. 1997;27:118–21. doi: 10.1080/01635589709514512. [DOI] [PubMed] [Google Scholar]
- Sparnins VL, Barany G, Wattenberg LW. Effects of organosulfur compounds from garlic and onions on benzo[a]pyrene-induced neoplasia and glutathione S-transferase activity in the mouse. Carcinogenesis. 1988;9:131–4. doi: 10.1093/carcin/9.1.131. [DOI] [PubMed] [Google Scholar]
- Sundaram SG, Milner JA. Diallyl disulfide induces apoptosis of human colon tumor cells. Carcinogenesis. 1996;17:669–73. doi: 10.1093/carcin/17.4.669. [DOI] [PubMed] [Google Scholar]
- Tajima K, Tominaga S. Dietary habits and gastro-intestinal cancers: a comparative case-control study of stomach and large intestinal cancers in Nagoya, Japan. Jpn J Cancer Res. 1985;76:705–16. [PubMed] [Google Scholar]
- Xiao D, Choi SA, Johnson DE, et al. Diallyl trisulfide-induced apoptosis in human prostate cancer cells is mediated by activation of c-Jun N-terminal kinase and extracellular-signal regulated kinase-mediated phosphorylation of Bcl-2. Oncogene. 2004;23:5594–606. doi: 10.1038/sj.onc.1207747. [DOI] [PubMed] [Google Scholar]
- You WC, Blot WJ, Chang YS, et al. Allium vegetables and reduced risk of stomach cancer. J Natl Cancer Inst. 1989;81:162–4. doi: 10.1093/jnci/81.2.162. [DOI] [PubMed] [Google Scholar]
- Yu GP, Hsieh CC, Wang LY, et al. Green-tea consumption and risk of stomach cancer: a population-based case-control study in Shanghai, China. Cancer Causes Control. 1995;6:532–8. doi: 10.1007/BF00054162. [DOI] [PubMed] [Google Scholar]
- Wang HH, Antonioli DA, Goldman H. Comparative features of esophageal and gastric adenocarcinomas: recent changes in type and frequency. Hum Pathol. 1986;17:482–7. doi: 10.1016/s0046-8177(86)80038-7. [DOI] [PubMed] [Google Scholar]
- Wargovich MJ. Diallyl sulfide, a flavor component of garlic (Allium sativum), inhibits dimethylhydrazine-induced colon cancer. Carcinogenesis. 1987;8:487–9. doi: 10.1093/carcin/8.3.487. [DOI] [PubMed] [Google Scholar]
- Wargovich MJ, Woods C, Eng VW, Stephens LC, Gray K. Chemoprevention of N- nitrosomethylbenzylamine-induced esophageal cancer in rats by the naturally occurring thioether, diallyl sulfide. Cancer Res. 1988;48:6872–5. [PubMed] [Google Scholar]
- Wargovich MJ, Uda N, Woods C, Velasco M, McKee K. Allium vegetables: their role in the prevention of cancer. Biochem Soc Trans. 1996;24:811–4. doi: 10.1042/bst0240811. [DOI] [PubMed] [Google Scholar]