Abstract
Compared with the abundant data from Western countries, evidence regarding meat consumption and colorectal cancer is limited in the Japanese population. We evaluated colorectal cancer risk in relation to meat consumption in a population‐based prospective cohort study in Japan. Participants were 13 957 men and 16 374 women aged ≥35 years in September 1992. Meat intake, assessed with a validated food frequency questionnaire, was controlled for the total energy intake. The incidence of colorectal cancer was confirmed through regional population‐based cancer registries and histological identification from colonoscopy in two main hospitals in the study area. From September 1992 to March 2008, 429 men and 343 women developed colorectal cancer. After adjustments for multiple confounders, a significantly increased relative risk of colorectal cancer was observed in the highest versus lowest quartile of the intake of total and red meat among men; the estimated hazard ratios were 1.36 (95% CI: 1.03, 1.79) for total meat (P for trend = 0.022), and 1.44 (95% CI: 1.10, 1.89) for red meat (P for trend = 0.009). A positive association between processed meat intake and colon cancer risk was also observed in men. There was no significant association between colorectal cancer and meat consumption in women. These results suggest that the intake of red and processed meat increases the risk of colorectal or colon cancer among Japanese men. Abstaining from excessive consumption of meat might be protective against developing colorectal cancer.
Keywords: Cohort studies, colorectal cancer, epidemiology, processed meat, red meat
Colorectal cancer has high incidence and mortality worldwide.1 Among dietary factors related to the risk of colorectal cancer, the role of meat consumption has been widely examined by researchers. The World Cancer Research Fund and American Institute for Cancer Research have judged the intake of red and processed meat to be a “convincing” risk factor for colorectal cancer.2 In October 2015, based on published literature mainly on colorectal cancer, the International Agency for Research on Cancer (IARC) classified the consumption of red meat as probably carcinogenic to humans (Group 2A), and that of processed meat as carcinogenic to humans (Group 1).3 Thus, meat consumption is a source of increasing concern in public health.
In Japan, colorectal cancer is the second leading cause of cancer among both males and females.4, 5 Compared with the abundant data from Western countries, evidence on meat consumption and colorectal cancer is limited among the Japanese population.6, 7 Six prospective cohort studies have been conducted and obtained different results.8, 9, 10, 11, 12, 13 Among them, only three studies on the incidence of colorectal cancer, including our previous report,8 have estimated the quantity of meat consumption using a validated food frequency questionnaire (FFQ), and considered several possible lifestyle confounders.8, 9, 10
In 2006, in the Takayama study, we reported an association between high consumption of processed meat and an increased risk of colon cancer identified by hospital records of colonoscopy among men after 8 years of follow‐up.8 Since then, we have collected additional information from the cancer registry in the study area. In the present study, we evaluated colorectal cancer risk, including rectal cancer, in relation to meat consumption, using updated data files of colorectal cancer and an extended period of follow‐up.
Materials and Methods
Participants and design
In September 1992, 36 990 residents of Takayama City, Gifu, Japan, aged ≥35 years who were not hospitalized were eligible to participate in the Takayama study. A total of 31 552 residents (85.3%) participated in the baseline survey and completed a self‐administered questionnaire including an FFQ. The details of this population‐based cohort study have been described elsewhere.14
Anthropometric characteristics, sociodemographic status, medical history, physical activity, smoking status, alcohol consumption, and regular diet were asked about in the baseline questionnaire. Reproductive characteristics including menopausal status and parity were included for women. Smokers were defined as people who had smoked a total of at least 20 packs of cigarettes in their life. We asked former and current smokers how long they had smoked. To assess physical activity both at work and at leisure, participants were asked the average time they spent on the following listed activities during the past year: strenuous sports, vigorous work, and moderate sports or work. The number of hours per week spent engaging in each activity was multiplied by the corresponding energy expenditure, expressed as metabolic equivalent of task (MET), and the product was taken as the physical activity score expressed as MET‐h/week. The details of this approach including its validity are described elsewhere.15
Diet consumption
Diet, including meat intake, was assessed using a semi‐quantitative FFQ. Data were collected on the average frequency of consumption and the usual serving size for 169 food items and dishes during the past year. We defined red meat as beef and pork. Food items for red meat products included beef steak, pork steak, pork cutlet, grilled meat, grilled offal, and liver. Ham, sausage, bacon, and yakibuta (Chinese style roasted pork) were defined as processed meat. These items and some other dishes including meat products used as cooking ingredients were accounted for to obtain the estimates for meat intake. Total meat was defined as the sum of any kind of meat, including red meat, poultry, and processed meat. Each intake of nutrients was estimated using the Japanese Standard Table of Food Composition (5th revised and enlarged edition), published by the Japan Science and Technology Agency.16 Both the validity and reproducibility of the questionnaire were previously reported to be reliable.17 Spearman correlation coefficients between the FFQ and 12‐day diet records kept over a 1‐year period in men and women were 0.18 and 0.62 for total meat, 0.21 and 0.54 for red meat, and 0.58 and 0.69 for processed meat, respectively. Corresponding values were 0.44 and 0.53 for total energy, 0.78 and 0.73 for calcium, 0.55 and 0.36 for vitamin D, 0.63 and 0.60 for dietary fiber, and 0.72 and 0.64 for alcohol, respectively.
Colorectal cancer and follow‐up
Participants were followed until the end of March 2008. At the baseline, 753 who were diagnosed with colorectal cancer before the baseline, and/or reported a positive history of any cancer and 535 who had a history of colorectal adenoma were excluded (there was some overlap). Consequently, 13 957 men and 16 374 women were included for analyses.
The cancer incidence was confirmed mainly through two regional population‐based cancer registries in Gifu. We also gained some information on histological identification using colonoscopy from two main hospitals in Takayama city.8, 18 The causes of cancer were coded according to the International Classification of Diseases and Health Related Problems, 10th Revision (ICD‐10). Colorectal cancer was defined as the sum of code C18 (colon cancer) and code C19 and C20 (rectal cancer). When participants developed both colon and rectal cancer, the preceding cancer was assigned.
Migration data during the study period were obtained from the city residential registers or family registers. Colorectal cancer death was identified from death certificates provided by the Legal Affairs Bureau.
The endpoint of follow‐up was determined by the earliest occurrence of one of the following events: a diagnosis of colorectal cancer, emigration from the study area, death, or the end of the study period. During the study period, 1767 persons (5.8%) moved away from the study area. For participants with who moved away on an unknown date (n = 238), their last date of residence that we could confirm was assigned as the endpoint of follow‐up. Colorectal cancer developed in 429 men and 343 women, and 115 men and 107 women died of colorectal cancer. The mortality‐to‐incidence ratio for colorectal cancer was 0.29, and patients who were ascertained by death certificate‐only registration were 6.2%, indicating satisfactory completeness of cancer information in this cohort. The ethical board of the Gifu University Graduate School of Medicine approved this study.
Statistical analyses
The values of the nutrients and foods consumed were controlled for the total energy intake using the residual method developed by Willett.19 Participants were categorized into quartile groups (Q1, Q2, Q3, or Q4) according to the distribution of their energy‐adjusted intake of total meat, red meat, and processed meat.
Relative risks and 95% confidence intervals (CIs) for colorectal cancer were estimated for the quartile groups of each category of meat intake using a Cox proportional hazard model. The reference group was set as the lowest quartile (Q1) of each meat intake. The following covariates were included as potential confounders in the models: for men, age (years, continuous), height (quartiles), body mass index (quartiles), physical activity score (continuous), smoking status (never, past, current smoker who had smoked for ≤30 years, current smoker who had smoked for ≥31 years), years of education (≤8, 9–11, 12–14, ≥15 years), history of aspirin use (yes, no), alcohol consumption (g/day), and energy‐adjusted intake of dietary fiber (g/day), calcium (mg/day), and vitamin D (μg/day); for women, age, height, body mass index, physical activity score, smoking status (never, past, current smoker), years of education, history of aspirin use, alcohol consumption, menopausal status (premenopausal, postmenopausal), and energy‐adjusted intake of dietary fiber, calcium, and vitamin D. Indicator terms were specifically created for missing data of categorical covariates. Tests for linear trend were conducted in the Cox model by treating meat intake as a continuous variable.
All analyses were conducted using the SAS program, version 9.4 (SAS Institute, Cary, NC). P‐values were calculated by a two‐sided test. A P‐value less than 0.05 was considered statistically significant in all analyses.
Results
The characteristics of participants are shown in Table 1 as the mean (standard deviation) or the percentage of each category, according to the quartile groups of energy‐adjusted total meat intake for each gender. Male and female participants in the higher quartile of total meat intake were younger and taller, and had attended school for a longer period at baseline. Women with a higher total meat intake tended to be premenopausal. Men and women among the lowest quartile of total meat intake had a higher level of alcohol consumption. Men in the lowest quartile and women in the highest quartile of total meat intake had a higher physical activity score.
Table 1.
Characteristics of study subjects at baseline
| Total meat consumption† | Men (quartile) | Women (quartile) | ||||||
|---|---|---|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |
| n | 3490 | 3489 | 3489 | 3489 | 4094 | 4093 | 4094 | 4093 |
| Age (years) | 56.5 (11.8) | 56.6 (12.5) | 55.0 (12.8) | 52.0 (11.7) | 58.9 (12.7) | 59.0 (13.6) | 55.5 (13.4) | 51.4 (11.5) |
| Height (cm) | 164.1 (6.7) | 163.9 (7.0) | 164.7 (7.0) | 165.6 (6.8) | 151.3 (6.5) | 151.1 (6.7) | 152.2 (6.4) | 153.2 (5.9) |
| Body mass index (kg/m2) | 22.6 (2.7) | 22.3 (2.7) | 22.4 (2.9) | 22.7 (2.9) | 22.0 (3.0) | 21.9 (3.0) | 22.0 (2.9) | 22.1 (2.9) |
| Smoking status (never) | 16.8% | 17.6% | 16.0% | 16.1% | 82.2% | 83.6% | 83.4% | 80.8% |
| Physical activity score (METs‐h/week) | 29.9 (43.7) | 26.5 (40.3) | 25.5 (39.5) | 27.6 (40.5) | 17.7 (28.8) | 17.9 (28.4) | 18.3 (27.5) | 22.0 (31.7) |
| Education years (≥12 years) | 33.9% | 37.3% | 44.5% | 51.9% | 24.4% | 26.8% | 35.6% | 44.9% |
| History of aspirin use (yes, no) | 6.3% | 7.4% | 6.9% | 7.8% | 9.8% | 8.3% | 9.9% | 12.0% |
| Alcohol consumption (g/day) | 56.0 (51.5) | 37.3 (37.0) | 33.8 (33.2) | 37.7 (37.0) | 9.9 (22.2) | 6.4 (13.7) | 6.2 (13.4) | 8.2 (16.2) |
| Total energy intake (kcal/day) | 2968 (851) | 2374 (726) | 2300 (760) | 2772 (960) | 2460 (829) | 1873 (641) | 1852 (660) | 2320 (814) |
| Total meat (g/day) | 44.6 (27.1) | 51.5 (28.9) | 68.4 (31.2) | 134.0 (66.8) | 33.0 (24.4) | 38.0 (24.3) | 54.1 (25.8) | 111.7 (55.6) |
| Red meat (g/day) | 25.8 (16.5) | 29.3 (16.9) | 38.0 (19.0) | 71.7 (40.3) | 17.3 (13.4) | 19.8 (13.4) | 28.1 (14.5) | 55.4 (30.3) |
| Processed meat (g/day) | 8.2 (7.1) | 9.0 (8.0) | 11.9 (9.5) | 24.4 (21.4) | 6.6 (6.6) | 7.1 (7.1) | 9.7 (8.1) | 20.9 (18.5) |
| Dietary fiber (g/day) | 18.9 (10.7) | 14.6 (7.7) | 14.1 (6.8) | 17.1 (8.3) | 20.3 (11.1) | 14.7 (7.5) | 14.3 (7.0) | 17.5 (8.4) |
| Calcium (mg/day) | 883 (508) | 654 (329) | 633 (321) | 752 (360) | 899 (514) | 641 (321) | 613 (307) | 756 (361) |
| Vitamin D (μg/day) | 10.5 (7.5) | 7.9 (5.5) | 7.8 (5.0) | 10.1 (6.4) | 8.4 (5.9) | 6.4 (4.2) | 6.5 (4.2) | 8.4 (5.0) |
| Menopausal status (pre) | 29.1% | 31.2% | 42.6% | 55.6% | ||||
Mean (standard deviation) or percentage. †Quartile groups according to the energy‐adjusted intakes of total meat by Willet method.
In Table 2, after adjustments for multiple confounders, a significantly increased relative risk of colorectal cancer was observed in the highest intake group (Q4) of total meat and red meat in men; the estimated hazard ratios (HRs) were 1.36 (95% CI: 1.03, 1.79) for total meat, and 1.44 (95% CI: 1.10, 1.89) for red meat. The linear trends in these associations were statistically significant. A higher intake of total meat was significantly associated with a higher risk of colon cancer. Participants in the highest group of red meat intake had a significantly increased risk of rectal cancer. In addition, participants in the highest group of processed meat intake had a significantly increased risk of colon cancer. In women, there was no significant association between colorectal cancer and total meat, red meat, or processed meat (Table 3).
Table 2.
Associations between meat consumption and colorectal cancer incidence among men
| Median intake (g) | No. of subjects | Person years | Colorectal cancer | Colon cancer | Rectal cancer | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. of cases | RR† | 95% CI | No. of cases | RR† | 95% CI | No. of cases | RR† | 95% CI | ||||
| Total meat‡ | ||||||||||||
| Q1 | 35 | 3490 | 46 270 | 110 | 1.00 | Reference | 72 | 1.00 | Reference | 38 | 1.00 | Reference |
| Q2 | 61 | 3489 | 45 253 | 107 | 1.14 | 0.87–1.50 | 65 | 1.08 | 0.76–1.52 | 42 | 1.26 | 0.80–1.97 |
| Q3 | 80 | 3489 | 45 223 | 105 | 1.22 | 0.92–1.61 | 71 | 1.29 | 0.92–1.82 | 34 | 1.09 | 0.67–1.76 |
| Q4 | 114 | 3489 | 46 239 | 107 | 1.36 | 1.03–1.79 | 68 | 1.36 | 0.96–1.93 | 39 | 1.34 | 0.84–2.14 |
| Trend P | 0.022 | 0.032 | 0.34 | |||||||||
| Red meat‡ | ||||||||||||
| Q1 | 19 | 3490 | 45 951 | 105 | 1.00 | Reference | 68 | 1.00 | Reference | 37 | 1.00 | Reference |
| Q2 | 33 | 3489 | 45 454 | 105 | 1.17 | 0.89–1.55 | 70 | 1.23 | 0.88–1.74 | 35 | 1.07 | 0.66–1.71 |
| Q3 | 44 | 3489 | 45 400 | 104 | 1.24 | 0.94–1.65 | 72 | 1.37 | 0.97–1.93 | 32 | 1.03 | 0.63–1.68 |
| Q4 | 64 | 3489 | 46 180 | 115 | 1.44 | 1.10–1.89 | 66 | 1.31 | 0.92–1.86 | 49 | 1.65 | 1.06–2.58 |
| Trend P | 0.009 | 0.12 | 0.023 | |||||||||
| Processed meat‡ | ||||||||||||
| Q1 | 4 | 3490 | 46 197 | 127 | 1.00 | Reference | 78 | 1.00 | Reference | 49 | 1.00 | Reference |
| Q2 | 9 | 3489 | 45 752 | 106 | 0.98 | 0.75–1.28 | 68 | 1.05 | 0.75–1.47 | 38 | 0.87 | 0.56–1.35 |
| Q3 | 13 | 3489 | 44 833 | 92 | 0.94 | 0.71–1.25 | 58 | 1.00 | 0.70–1.42 | 34 | 0.85 | 0.54–1.35 |
| Q4 | 23 | 3489 | 46 203 | 104 | 1.22 | 0.93–1.60 | 72 | 1.43 | 1.02–2.01 | 32 | 0.91 | 0.57–1.45 |
| Trend P | 0.32 | 0.17 | 0.86 | |||||||||
| Excluding cases within 2 years | ||||||||||||
| Total meat‡ | ||||||||||||
| Q1 | 35 | 3481 | 46 272 | 99 | 1.00 | Reference | 63 | 1.00 | Reference | 36 | 1.00 | Reference |
| Q2 | 61 | 3480 | 45 273 | 97 | 1.15 | 0.86–1.54 | 59 | 1.13 | 0.79–1.63 | 38 | 1.18 | 0.74–1.89 |
| Q3 | 80 | 3480 | 45 221 | 95 | 1.23 | 0.92–1.65 | 64 | 1.35 | 0.94–1.95 | 31 | 1.03 | 0.62–1.69 |
| Q4 | 114 | 3480 | 46 175 | 102 | 1.43 | 1.07–1.91 | 65 | 1.50 | 1.05–2.16 | 37 | 1.31 | 0.81–2.12 |
| Trend P | 0.012 | 0.014 | 0.35 | |||||||||
| Red meat‡ | ||||||||||||
| Q1 | 19 | 3481 | 45 967 | 94 | 1.00 | Reference | 60 | 1.00 | Reference | 34 | 1.00 | Reference |
| Q2 | 33 | 3480 | 45 416 | 98 | 1.24 | 0.92–1.64 | 64 | 1.29 | 0.90–1.85 | 34 | 1.11 | 0.69–1.81 |
| Q3 | 44 | 3480 | 45 404 | 94 | 1.25 | 0.93–1.68 | 64 | 1.39 | 0.96–2.00 | 30 | 1.03 | 0.62–1.71 |
| Q4 | 64 | 3480 | 46 154 | 107 | 1.49 | 1.12–1.98 | 63 | 1.43 | 0.99–2.05 | 44 | 1.57 | 0.99–2.50 |
| Trend P | 0.006 | 0.047 | 0.058 | |||||||||
| Processed meat‡ | ||||||||||||
| Q1 | 4 | 3481 | 46 237 | 112 | 1.00 | Reference | 68 | 1.00 | Reference | 44 | 1.00 | Reference |
| Q2 | 9 | 3481 | 45 728 | 102 | 1.07 | 0.81–1.42 | 64 | 1.15 | 0.81–1.64 | 38 | 0.96 | 0.61–1.50 |
| Q3 | 13 | 3479 | 44 784 | 84 | 0.98 | 0.73–1.33 | 55 | 1.11 | 0.76–1.62 | 29 | 0.80 | 0.49–1.31 |
| Q4 | 23 | 3480 | 46 193 | 95 | 1.26 | 0.94–1.68 | 64 | 1.47 | 1.03–2.11 | 31 | 0.95 | 0.59–1.55 |
| Trend P | 0.24 | 0.14 | 0.98 | |||||||||
†Estimated hazard ratio after adjustments for age, height (quartiles), body mass index (quartiles), physical activity score, smoking status (never, past, current smoker for 30 years or less, current smoker for 31 years or more), education years (≤8, 9–11, 12–14, ≥15 years), history of aspirin use (yes, no), alcohol consumption (g/day), and the intakes (quartiles) of total fiber, calcium, and vitamin D. ‡Meat consumption was adjusted for total energy intake by Willet method.
Table 3.
Associations between meat consumption and colorectal cancer incidence among women
| Median intake (g) | No. of subjects | Person years | Colorectal cancer | Colon cancer | Rectal cancer | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. of cases | RR† | 95% CI | No. of cases | RR† | 95% CI | No. of cases | RR† | 95% CI | ||||
| Total meat‡ | ||||||||||||
| Q1 | 26 | 4094 | 56 142 | 97 | 1.00 | Reference | 73 | 1.00 | Reference | 24 | 1.00 | Reference |
| Q2 | 48 | 4093 | 55 387 | 101 | 1.12 | 0.84–1.49 | 72 | 1.04 | 0.74–1.45 | 29 | 1.35 | 0.78–2.34 |
| Q3 | 64 | 4094 | 56 520 | 75 | 0.94 | 0.69–1.29 | 54 | 0.88 | 0.61–1.28 | 21 | 1.13 | 0.61–2.07 |
| Q4 | 93 | 4093 | 58 632 | 70 | 1.01 | 0.74–1.40 | 45 | 0.88 | 0.60–1.30 | 25 | 1.43 | 0.80–2.56 |
| Trend P | 0.48 | 0.33 | 0.80 | |||||||||
| Red meat‡ | ||||||||||||
| Q1 | 12 | 4094 | 55 685 | 98 | 1.00 | Reference | 68 | 1.00 | Reference | 30 | 1.00 | Reference |
| Q2 | 23 | 4093 | 55 271 | 97 | 1.07 | 0.80–1.42 | 77 | 1.21 | 0.86–1.69 | 20 | 0.75 | 0.42–1.33 |
| Q3 | 32 | 4094 | 57 089 | 69 | 0.86 | 0.62–1.18 | 42 | 0.75 | 0.50–1.12 | 27 | 1.10 | 0.64–1.89 |
| Q4 | 49 | 4093 | 58 636 | 79 | 1.07 | 0.79–1.46 | 57 | 1.13 | 0.79–1.64 | 22 | 0.95 | 0.54–1.68 |
| Trend P | 0.98 | 0.99 | 0.97 | |||||||||
| Processed meat‡ | ||||||||||||
| Q1 | 3 | 4094 | 56 692 | 100 | 1.00 | Reference | 74 | 1.00 | Reference | 26 | 1.00 | Reference |
| Q2 | 8 | 4093 | 56 261 | 88 | 0.96 | 0.71–1.28 | 61 | 0.88 | 0.62–1.25 | 27 | 1.17 | 0.68–2.03 |
| Q3 | 11 | 4094 | 55 763 | 81 | 0.98 | 0.72–1.34 | 63 | 1.00 | 0.70–1.44 | 18 | 0.90 | 0.48–1.69 |
| Q4 | 19 | 4093 | 57 965 | 74 | 1.12 | 0.82–1.55 | 46 | 0.95 | 0.64–1.40 | 28 | 1.62 | 0.92–2.84 |
| Trend P | 0.95 | 0.25 | 0.10 | |||||||||
| Excluding cases within 2 years | ||||||||||||
| Total meat‡ | ||||||||||||
| Q1 | 26 | 4087 | 56 176 | 87 | 1.00 | Reference | 66 | 1.00 | Reference | 21 | 1.00 | Reference |
| Q2 | 48 | 4087 | 55 452 | 89 | 1.09 | 0.80–1.47 | 61 | 0.97 | 0.68–1.39 | 28 | 1.46 | 0.82–2.59 |
| Q3 | 64 | 4087 | 56 479 | 71 | 0.97 | 0.70–1.35 | 51 | 0.91 | 0.62–1.33 | 20 | 1.19 | 0.63–2.25 |
| Q4 | 93 | 4086 | 58 540 | 69 | 1.09 | 0.78–1.51 | 44 | 0.93 | 0.62–1.38 | 25 | 1.59 | 0.87–2.91 |
| Trend P | 0.73 | 0.46 | 0.60 | |||||||||
| Red meat‡ | ||||||||||||
| Q1 | 12 | 4087 | 55 707 | 89 | 1.00 | Reference | 62 | 1.00 | Reference | 27 | 1.00 | Reference |
| Q2 | 23 | 4087 | 55 326 | 86 | 1.03 | 0.76–1.39 | 66 | 1.13 | 0.79–1.61 | 20 | 0.81 | 0.45–1.46 |
| Q3 | 32 | 4087 | 57 044 | 65 | 0.87 | 0.62–1.21 | 40 | 0.77 | 0.51–1.16 | 25 | 1.10 | 0.62–1.93 |
| Q4 | 49 | 4086 | 58 569 | 76 | 1.10 | 0.80–1.51 | 54 | 1.14 | 0.78–1.67 | 22 | 1.02 | 0.57–1.84 |
| Trend P | 0.83 | 0.90 | 0.79 | |||||||||
| Processed meat‡ | ||||||||||||
| Q1 | 3 | 4087 | 56 698 | 92 | 1.00 | Reference | 68 | 1.00 | Reference | 24 | 1.00 | Reference |
| Q2 | 8 | 4087 | 56 310 | 78 | 0.91 | 0.67–1.25 | 54 | 0.85 | 0.59–1.23 | 24 | 1.10 | 0.62–1.96 |
| Q3 | 11 | 4087 | 55 750 | 74 | 0.95 | 0.69–1.32 | 56 | 0.96 | 0.65–1.40 | 18 | 0.94 | 0.50–1.79 |
| Q4 | 19 | 4086 | 57 888 | 72 | 1.15 | 0.83–1.60 | 44 | 0.96 | 0.64–1.44 | 28 | 1.70 | 0.95–3.03 |
| Trend P | 0.85 | 0.31 | 0.062 | |||||||||
†Estimated hazard ratio after adjustments for age, height (quartiles), body mass index (quartiles), physical activity score, smoking status (never, past, current smoker), education years (≤8, 9–11, 12–14, ≥15 years), history of aspirin use (yes, no), alcohol consumption (g/day), menopausal status (premenopausal, postmenopausal), the intakes (quartiles) of total fiber, calcium, and vitamin D. ‡Meat consumption was adjusted for total energy intake by Willet method.
To eliminate those who might have had colorectal cancer but who had not yet been diagnosed at baseline, we excluded 63 patients who were diagnosed with colorectal cancer in the first 2 years of follow‐up and then re‐analyzed the data. Although none of the results were substantially altered, the positive association between meat consumption and colon cancer in men seemed to be slightly strengthened (Table 2). Additionally, we observed a significant association between higher intake of red meat and increased risk of colon cancer (P for linear trend = 0.047).
When the association between seafood (fish and shellfish) consumption and colorectal cancer was evaluated similarly, the estimated HRs (95% CI) for colorectal cancer in the Q2, Q3, and Q4 vs the Q1 of seafood consumption were, respectively, 1.36 (0.99, 1.87), 1.16 (0.80, 1.67), and 1.29 (0.86, 1.94) for men, and 0.88 (0.62, 1.26), 1.17 (0.80, 1.72), and 1.48 (0.96, 2.27) for women.
Discussion
After the addition of cancer registry data and the extension of follow‐up to our previous report,8 this study revealed significant positive associations of total and red meat consumption with colorectal cancer risk in men. Higher intake of processed meat was associated with a higher risk of colon cancer in men.
Several supporting mechanisms have been proposed for the carcinogenic effects of meat.20 When meat is cooked at high temperatures by pan‐frying, grilling, or barbecuing, heterocyclic aromatic amines and polycyclic aromatic hydrocarbons (PAHs), which are carcinogenic and mutagenic, are formed.21, 22 Meat processing such as curing and smoking also produces PAHs and another carcinogenic N‐nitroso compounds (NOCs).22, 23 In the digestive tract, heme iron from red meat mediates the formation of NOCs,24, 25 and the high fat content of red meat could also be a candidate for promoting tumorigenesis by enhancing the production of secondary bile acids by gut bacteria.26 In our study however, dietary intake of iron, animal fat, or saturated fat overall were not directly associated with colorectal cancer risk (data not shown).
The Japanese population is reported to consume a much lower amount of meat compared to Western populations.27 Furthermore, dietary habits, lifestyle, and genetic background in Japan are also different from those in Western countries. In a 2014 meta‐analysis of six Japanese cohort studies and eight Japanese case‐control studies on meat consumption and colorectal cancer incidence or mortality, Pham et al. reported that colorectal cancer risk had a statistically significant positive association with the consumption of red and processed meat, but not with total meat consumption.7 However, individual prospective studies have failed to demonstrate a clear association between meat consumption and colorectal cancer. Three Japanese cohort studies on meat consumption and colorectal cancer mortality revealed no significant association.11, 12, 13 Three other Japanese cohort studies have estimated the incidence of colorectal cancer.8, 9, 10 Firstly, the Miyagi cohort study revealed no significant association with total, red, or processed meat consumption.10 Secondly, the Japan Public Health Center‐based prospective study reported that the HR of colon cancer for the highest versus lowest quintiles of red meat intake was 1.48 (95% CI: 1.01, 2.17, P for linear trend = 0.03) in women. Also, in men, the corresponding HR for total meat intake was 1.44 (95% CI: 1.06, 1.98, P for linear trend = 0.07).9 Thirdly, using updated data of the Takayama study, here we found significant positive associations of total and red meat consumption with colorectal cancer risk in men, in addition to an increased risk of colon cancer among men with high consumption of processed meat as observed previously. In women, there was no significant association of colorectal cancer with the intake of total, red, or processed meat, although the link between meat and colorectal cancer was not modified by sex. The effects of heme iron in red meat, which may be involved in carcinogenesis, might be weakened in women because of blood loss by menstruation, although the associations between meat consumption and colorectal cancer were not substantially different between premenopausal and postmenopausal women in our study (data not shown). Also, the lower intake of meat in women than men might have made it difficult to detect any association with colorectal cancer (mean in men and women: 74.3 and 59.0 g/day for total meat, 41.0 and 30.1 g/day for red meat, and 13.3 and 11.0 g/day for processed meat, respectively).
We conducted a meta‐analysis of these six prospective studies and found that the pooled RRs of colorectal cancer in the highest category of consumption compared with the lowest category were recalculated as 1.22 (95% CI: 1.07, 1.39) for red meat, 1.09 (95% CI: 0.95, 1.26) for processed meat, and 1.07 (95% CI: 0.91, 1.26) for total meat, respectively, among all subjects. The increased risk of higher consumption of red meat was stronger in men (1.31 [95% CI: 1.10, 1.56]) than women (1.06 [95% CI: 0.81, 1.39]), and more prominent for colon cancer (1.29 [95% CI: 1.10, 1.52]) than rectal cancer (1.06 [95% CI: 0.81, 1.40]). Thus, the present study reinforced the positive association between red meat consumption and colorectal cancer in Japan, and also provided possible evidence for the difference in sex or subsites. In addition, the pooled RR did not show a significant association with processed meat, which has more robust evidence of carcinogenicity based on the international classification, but high consumption of processed meat is suggested to be associated with an increased risk of colon or colorectal cancer as observed in the Takayama study. Thus, more prospective studies are warranted to further assess these associations in Japan.
The strengths of our study include the prospective design, which minimalized recall bias, as well as the robust representation of the general population, good participation rate, long follow‐up duration, and information on several confounders.
Several potential limitations should be mentioned. The FFQ was designed to measure an individual's relative intake of foods and nutrients, rather than absolute values. Some of values for dietary intakes we presented in the table may have been overestimated by the FFQ. In addition, the validity of the FFQ for total and red meat was low in men, which might have affected the results with colorectal cancer risk in this study. However, the observed association of total and red meat with colorectal cancer might have been underestimated because a misclassification on dietary intake would likely have occurred non‐differentially between participants with colorectal cancer and those without colorectal cancer. The exposure evaluation was performed only at baseline and changes in lifestyle patterns during the follow‐up period were unknown. Although underlying diseases or preclinical signs may have affected lifestyle at baseline, the exclusion of cases during the first 2 years of follow‐up did not substantially change the results. Despite considering several lifestyle and reproductive factors in the analyses, we could not fully exclude the possibility of residual confounders.
In conclusion, this prospective study in Japan demonstrated increased risk of colorectal cancer among men with higher intake of total and red meat. Higher intake of processed meat was associated with a higher risk of colon cancer in men. These results suggest that the intake of red and processed meat increases the risk of colorectal or colon cancer especially among men. In accordance with the assessment of the IARC, abstaining from overeating meat might be protective against developing colorectal cancer among the Japanese population.
Disclosure Statement
The authors have no conflict of interest.
Abbreviations
- CI
confidence intervals
- FFQ
food frequency questionnaire
- IARC
the international agency for research on cancer
- MET
metabolic equivalent
- NOC
N‐nitroso compound
- PAH
polycyclic aromatic hydrocarbon
Acknowledgments
This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labour and Welfare of Japan.
Cancer Sci 108 (2017) 1065–1070
Funding Information
This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labour and Welfare of Japan.
References
- 1. Ferlay J, Soerjomataram I, Ervik M et al GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon, France: International Agency for Research on Cancer, 2013. Available from: http://globocan.iarc.fr, last accessed on 17 April 2017. [Google Scholar]
- 2. World Cancer Research Fund / American Institute for Cancer Research . Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. Washington DC: AICR, 2007. [Google Scholar]
- 3. Bouvard V, Loomis D, Guyton KZ et al Carcinogenicity of consumption of red and processed meat. Lancet Oncol 2015; 16: 1599–600. [DOI] [PubMed] [Google Scholar]
- 4. Katanoda K, Hori M, Matsuda T et al An updated report on the trends in cancer incidence and mortality in Japan, 1958–2013. Jpn J Clin Oncol 2015; 45: 390–401. [DOI] [PubMed] [Google Scholar]
- 5. Hori M, Matsuda T, Shibata A, Katanoda K, Sobue T, Nishimoto H. Cancer incidence and incidence rates in Japan in 2009: A study of 32 population‐based cancer registries for the Monitoring of Cancer Incidence in Japan (MCIJ) project. Jpn J Clin Oncol 2015; 45: 884–91. [DOI] [PubMed] [Google Scholar]
- 6. Alexander DD, Weed DL, Miller PE, Mohamed MA. Red meat and colorectal cancer: A quantitative update on the state of the epidemiologic science. J Am Coll Nutr 2015; 34: 521–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Pham NM, Mizoue T, Tanaka K et al Meat consumption and colorectal cancer risk: An evaluation based on a systematic review of epidemiologic evidence among the Japanese population. Jpn J Clin Oncol 2014; 44: 641–50. [DOI] [PubMed] [Google Scholar]
- 8. Oba S, Shimizu N, Nagata C et al The relationship between the consumption of meat, fat, and coffee and the risk of colon cancer: A prospective study in Japan. Cancer Lett 2006; 244: 260–7. [DOI] [PubMed] [Google Scholar]
- 9. Takachi R, Tsubono Y, Baba K et al Red meat intake may increase the risk of colon cancer in Japanese, a population with relatively low red meat consumption. Asia Pac J Clin Nutr 2011; 20: 603–12. [PubMed] [Google Scholar]
- 10. Sato Y, Nakaya N, Kuriyama S, Nishino Y, Tsubono Y, Tsuji I. Meat consumption and risk of colorectal cancer in Japan: The Miyagi Cohort Study. Eur J Cancer Prev 2006; 15: 211–8. [DOI] [PubMed] [Google Scholar]
- 11. Kojima M, Wakai K, Tamakoshi K et al Diet and colorectal cancer mortality: Results from the Japan Collaborative Cohort Study. Nutr Cancer 2004; 50: 23–32. [DOI] [PubMed] [Google Scholar]
- 12. Khan MM, Goto R, Kobayashi K et al Dietary habits and cancer mortality among middle aged and older Japanese living in hokkaido, Japan by cancer site and sex. Asian Pac J Cancer Prev 2004; 5: 58–65. [PubMed] [Google Scholar]
- 13. Hirayama T, ed. Life‐Style and Mortality. A Large‐Scale Census‐Based Cohort Study in Japan. Basal: S. Karger, 1990. [Google Scholar]
- 14. Shimizu H. The Basic Report on Takayama Study. Gifu: Department of Public Health, Gifu University School of Medicine, 1996. [Google Scholar]
- 15. Suzuki I, Kawakami N, Shimizu H. Reliability and validity of a questionnaire for assessment of energy expenditure and physical activity in epidemiological studies. J Epidemiol 1998; 8: 152–9. [DOI] [PubMed] [Google Scholar]
- 16. Council for Science and Technology; Ministry of Education C, Sports, Science and Technology, Japan , ed. Standard Tables of Food Composition in Japan (5th revised and enlarged edition) (in Japanese). Tokyo: National Printing Bureau, 2005. [Google Scholar]
- 17. Shimizu H, Ohwaki A, Kurisu Y et al Validity and reproducibility of a quantitative food frequency questionnaire for a cohort study in Japan. Jpn J Clin Oncol 1999; 29: 38–44. [DOI] [PubMed] [Google Scholar]
- 18. Oba S, Nagata C, Shimizu N et al Soy product consumption and the risk of colon cancer: A prospective study in Takayama, Japan. Nutr Cancer. 2007; 57: 151–7. [DOI] [PubMed] [Google Scholar]
- 19. Willett W. Implication of total energy intake for epidemiological analyses In: Willett W, ed. Nutritional Epidemiolgy. Oxford: Oxford University Press, 1990; 245–71. [Google Scholar]
- 20. Santarelli RL, Pierre F, Corpet DE. Processed meat and colorectal cancer: A review of epidemiologic and experimental evidence. Nutr Cancer 2008; 60: 131–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Sinha R, Rothman N, Brown ED et al Pan‐fried meat containing high levels of heterocyclic aromatic amines but low levels of polycyclic aromatic hydrocarbons induces cytochrome P4501A2 activity in humans. Cancer Res 1994; 54: 6154–9. [PubMed] [Google Scholar]
- 22. Alomirah H, Al‐Zenki S, Al‐Hooti S et al Concentrations and dietary exposure to polycyclic aromatic hydrocarbons (PAHs) from grilled and smoked foods. Food Control 2011; 22: 2028–35. [Google Scholar]
- 23. Lijinsky W. N‐Nitroso compounds in the diet. Mutat Res 1999; 443: 129–38. [DOI] [PubMed] [Google Scholar]
- 24. Cross AJ, Pollock JR, Bingham SA. Haem, not protein or inorganic iron, is responsible for endogenous intestinal N‐nitrosation arising from red meat. Cancer Res 2003; 63: 2358–60. [PubMed] [Google Scholar]
- 25. Santarelli RL, Vendeuvre JL, Naud N et al Meat processing and colon carcinogenesis: Cooked, nitrite‐treated, and oxidized high‐heme cured meat promotes mucin‐depleted foci in rats. Cancer Prev Res (Phila) 2010; 3: 852–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Nagengast FM, Grubben MJ, van Munster IP. Role of bile acids in colorectal carcinogenesis. Eur J Cancer 1995; 31A: 1067–70. [DOI] [PubMed] [Google Scholar]
- 27. Speedy AW. Global production and consumption of animal source foods. J Nutr 2003; 133: 4048S–53S. [DOI] [PubMed] [Google Scholar]