Abstract
Several studies on pancreatic cancer have reported significant positive associations for intake of red meat but null associations for heme iron. We assessed total, red, white, and processed meat intake, meat cooking methods and doneness, and heme iron and mutagen intake in relation to pancreatic cancer in the NIH-AARP Diet and Health Study cohort. 322,846 participants (187,265 men; 135,581 women) successfully completed and returned the food frequency questionnaire between 1995–1996. After a mean follow-up of 9.2 years (up to 10.17 years), 1,417 individuals (895 men, 522 women) developed exocrine pancreatic cancer. Cox proportional hazard models were used to calculate hazard ratios (HR) and 95% confidence intervals (CI), and trends were calculated using the median value of each quantile. Models incorporated age as the time metric and were adjusted for smoking history, BMI, self-reported diabetes, and energy-adjusted saturated fat. Pancreatic cancer risk significantly increased with intake of total meat (Q5 vs. Q1 HR=1.20, 95% CI 1.02–1.42, p-trend=0.03), red meat (HR=1.22, 95% CI 1.01–1.48, p-trend=0.02), high-temperature cooked meat (HR=1.21, 95% CI 1.00–1.45, p-trend=0.02), grilled/barbequed meat (HR=1.24, 95% CI 1.03–1.50, p-trend=0.007), well/very well done meat (HR=1.32, 95% CI 1.10–1.58, p-trend = 0.005), and heme iron from red meat (Q4 vs. Q1 HR=1.21, 95% CI 1.01–1.45, p-trend=0.04). When stratified by sex, these associations remained significant in men but not women except for white meat intake in women (HR = 1.33, 95% CI 1.02–1.74, p-trend = 0.04). Additional studies should confirm our findings that consuming heme iron from red meat increases pancreatic cancer risk.
Keywords: pancreatic cancer, heme iron, meat
INTRODUCTION
Pancreatic cancer is the fourth leading cause of cancer-related mortality in men and women of all ages in the United States (1). It is rapidly fatal and has a 5-year survival rate of <7% (1). Cigarette smoking, history of diabetes mellitus, and overweight and obesity are consistent potentially modifiable risk factors for pancreatic cancer (2). The association between diet and pancreatic cancer risk is unclear mostly because of inconsistent study findings.
In 2012, a panel from the World Cancer Research Fund and American Institute for Cancer Research (WCRF/AICR) concluded that the evidence that red and processed meats contribute to pancreatic cancer was suggestive. Dose-response meta-analyses of eight prospective cohort studies revealed that higher intake of red meat was positively, but insignificantly, associated with pancreatic cancer risk with some heterogeneity between studies (3). In addition, two meta-analyses on 5 case-control studies and 11 prospective cohort studies similarly found significant positive associations for red meat which were stronger in case-control studies and among men, respectively (4, 5). Processed meat was also positively associated with pancreatic cancer; however, the association was only significant in men (3, 5). Further, these associations could potentially be explained by other factors in meat particularly confounded by compounds that are generated with meat cooking methods and doneness levels (6–10).
Our prior examination of the NIH-AARP Diet and Health Study cohort revealed significant positive associations between total, red, and high-temperature cooked meats and pancreatic cancer in men but not women (8). The study also noted that associated meat mutagens specifically, overall mutagenic activity in men and the heterocyclic amine (HCA) 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (DiMeIQx) in both men and women may contribute to pancreatic carcinogenesis. Therefore, we examined the association between meat, meat cooking methods, and compounds in meat in the large NIH-AARP Diet and Health Study cohort. In contrast to our previous examination of these hypotheses, the present study has longer follow-up time (6 additional years) and more than three times as many pancreatic cancer cases (n > 1400 cases) with meat cooking methods and related mutagens. The greater number of cases increases the power of our study and also enables us to look at interactions by other exposures. In addition to the meat mutagens, our current study also examines heme iron intake as a risk factor for pancreatic cancer.
SUBJECTS AND METHODS
Study Population
The NIH-AARP Diet and Health Study is a large prospective study of AARP members established in 1995–1996 (8, 11). In total, 567,169 AARP members aged 50–71 years living in six US states (California, Florida, Louisiana, New Jersey, North Carolina, and Pennsylvania) and in two metropolitan areas (Atlanta, GA and Detroit, MI) successfully completed and returned the self-administered questionnaires (8, 11) that assessed demographic characteristics, dietary intake over the previous year, and health-related factors (8). Six months after this baseline questionnaire was sent, participants who responded to the baseline questionnaire received a second risk factor questionnaire (RFQ) eliciting information on meat cooking methods (8, 11, 12). In total, 332,913 participants completed and returned the second questionnaire (8). Informed consent was obtained from all study participants and the study was approved by the National Cancer Institute Special Studies Institutional Review Board (8).
For the present study, we only included participants who completed the meat module portion of the second questionnaire. We excluded subjects who had questionnaires filled out by proxy respondents (n=6,959), who had prevalent cancers as determined by the cancer registry data (n=2,361) and whose energy consumption lay outside the normal sex-specific distribution for energy intake by two interquartile ranges above the 75th or below the 25th percentile on the logarithmic scale (n=2,701; ref. 8). We also excluded those with ≤0 years of follow-up (n=38). Our final analytic cohort consisted of 322,846 individuals (187,265 men, 135,581 women; ref. 8).
Cohort Follow-up and Case Ascertainment
Pancreatic cancer cases were ascertained by linking cohort members to state cancer registries where the study participants reside, as well as Arizona, Texas and Nevada and to the U.S. National Death Index from 1995 to 2006 (8, 12, 13). Vital status of cohort participants was also ascertained by linkage to the Social Security Administration Death Master File. We included incident adenocarcinoma of the exocrine pancreas as the primary outcome in our analysis [International Classification of Disease for Oncology, Third Edition (code C250-C259); ref. 8, 12]. Our case definition excluded pancreatic endocrine tumors, sarcomas, and lymphomas (histology types, 8150, 8151, 8153, 8155, 8240) as the etiologies may differ (8, 12).
Dietary Assessment and Meat Variable
The baseline questionnaire gathered information on demographic characteristics, medical history, and health-related behaviors and also contained a grid based food frequency questionnaire (FFQ) that assessed the frequency and portion size of 124 food items consumed over the past year (8, 11). These line items were constructed using over 5,000 individual food codes found in the United States Department of Agriculture (USDA)’s Continuing Survey of Food Intake by Individuals database (14). Intake of total, red, white, processed, and high-temperature cooked meats, heme iron, as well as energy and other nutrients was determined from the baseline FFQ. While the specific items in each category have been described in detail elsewhere (8), a few categories have been altered. We included all forms of poultry products in the white meat category. Processed meat from red meat and poultry contributed to the processed meat category. The validity of both questionnaires was discussed in an earlier study (8).
The RFQ included a meat module that elicited meat-cooking methods (barbequed/grilled, pan-fried, oven-broiled, and microwaved) and doneness levels (rare/medium and well/very well done; ref. 8, 15, 16). Meat samples cooked by different methods and to varying levels of doneness were analyzed for levels of HCAs, polycyclic aromatic hydrocarbons (PAHs), and overall mutagenicity. Overall mutagenic activity was quantified using the Ames assay (8, 9, 15, 17). These meat mutagens were entered into the CHARRED database (8, 15, 18–21). The RFQ was linked to the CHARRED database to estimate the daily intake of meat-mutagens including HCAs DiMeIQx, 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), and benzo(a)pyrene (BaP), a marker for PAHs (8, 15, 18–21).
Statistical Analysis
We used generalized linear models to calculate the means within each total meat intake quintile for the continuous variables and frequency proportions for the categorical variables in Table 1.
Table 1.
Selected baseline characteristics of NIH-AARP Diet and Health Study cohort participants by total meat intake 1
| Characteristics | Quintile of daily total meat intake (g/1,000 kcal) in men | ||||
|---|---|---|---|---|---|
|
| |||||
| Men (n=187,265) | ≤43.6 | >43.6 and ≤58.5 | >58.5 and ≤72.9 | >72.9 and ≤92.3 | >92.3 |
|
| |||||
| Meat Intake (g/1,000 kcal) | |||||
| Total Meat | 30.9 | 51.3 | 65.6 | 81.8 | 115.9 |
| Red Meat | 17.0 | 28.8 | 36.8 | 45.1 | 59.0 |
| White Meat | 13.9 | 22.5 | 28.8 | 36.7 | 56.9 |
|
| |||||
| Age (y) | 63.2 | 63.0 | 62.7 | 62.2 | 61.6 |
|
| |||||
| Education, college graduate or postgraduate (%) | 46.5 | 47.0 | 47.6 | 48.0 | 47.6 |
|
| |||||
| Race | |||||
| African American (%) | 3.7 | 3.1 | 3.0 | 3.0 | 3.5 |
| Non-Hispanic white (%) | 90.8 | 93.0 | 93.3 | 93.5 | 92.5 |
|
| |||||
| Smoking history2 | |||||
| Never (%) | 30.4 | 29.5 | 29.7 | 29.6 | 29.0 |
| Former smoker (%) | 57.0 | 58.0 | 57.9 | 57.4 | 57.5 |
| Current smoker (%) | 8.9 | 9.0 | 9.1 | 9.6 | 10.0 |
|
| |||||
| BMI (kg/m2) | 26.2 | 26.7 | 27.1 | 27.5 | 28.1 |
| <25 kg/m2 (%) | 39.5 | 33.4 | 30.2 | 27.0 | 22.12 |
| 25–<30 kg/m2 (%) | 44.9 | 48.3 | 49.2 | 49.3 | 49.0 |
| ≥30 kg/m2 (%) | 13.8 | 16.7 | 19.2 | 22.3 | 27.4 |
|
| |||||
| Heavy physical activity, ≥5 times/wk (%) | 26.9 | 23.2 | 21.2 | 20.4 | 19.4 |
|
| |||||
| Self-reported diabetes (%) | 6.6 | 7.8 | 9.2 | 11.2 | 14.2 |
|
| |||||
| Dietary Intake | |||||
| Energy (kcal) | 2,032 | 1,985 | 1,987 | 2,004 | 2,024 |
| Total fat (g/1,000 kcal/day) | 28.5 | 32.5 | 34.0 | 35.4 | 37.3 |
| Saturated fat (g/1,000 kcal/day) | 8.9 | 10.2 | 10.6 | 11.0 | 11.6 |
| Heme iron (mg/1,000 kcal/day) | 86.1 | 153.5 | 202.5 | 255.8 | 358.9 |
| Advanced glycation end products | 333.2 | 1274.6 | 1290.5 | 1444.8 | 1985.1 |
|
| |||||
| Alcohol intake (# drinks/day) | 2.0 | 1.3 | 1.1 | 1.0 | 0.8 |
| Alcohol intake ≥ 3 drinks/day (%) | 15.9 | 12.3 | 10.8 | 9.2 | 6.8 |
| Characteristics | Quintile of daily total teat intake (g/1,000 kcal) in women | ||||
|---|---|---|---|---|---|
|
| |||||
| Women (n=135,581) | ≤38.0 | >38.0 and ≤53.4 | >53.4 and ≤68.4 | >68.4 and ≤88.3 | >88.3 |
|
| |||||
| Meat Intake (g/1,000 kcal) | |||||
| Total Meat | 25.7 | 45.9 | 60.7 | 77.5 | 113.7 |
| Red Meat | 12.6 | 22.6 | 29.4 | 35.7 | 45.0 |
| White Meat | 13.2 | 23.3 | 31.4 | 41.8 | 68.6 |
|
| |||||
| Age (y) | 62.5 | 62.4 | 62.2 | 61.9 | 61.4 |
|
| |||||
| Education, college graduate or postgraduate (%) | 33.2 | 30.6 | 31.1 | 32.1 | 31.8 |
|
| |||||
| Race | |||||
| African American (%) | 2.7 | 2.2 | 2.0 | 2.0 | 2.2 |
| Non-Hispanic white (%) | 91.5 | 94.1 | 94.3 | 94.6 | 94.0 |
|
| |||||
| Smoking history2 | |||||
| Never (%) | 46.3 | 44.8 | 44.5 | 43.4 | 40.9 |
| Former smoker (%) | 38.5 | 38.4 | 39.1 | 39.8 | 41.8 |
| Current smoker or having quit <1 y ago (%) | 11.8 | 13.6 | 13.6 | 13.8 | 14.0 |
|
| |||||
| BMI (kg/m2) | 25.5 | 26.3 | 26.8 | 27.2 | 27.8 |
| <25 kg/m2 (%) | 53.8 | 46.8 | 43.8 | 40.3 | 36.4 |
| 25–<30 kg/m2 (%) | 27.6 | 31.1 | 31.7 | 33.0 | 33.1 |
| ≥30 kg/m2 (%) | 15.2 | 19.3 | 21.9 | 24.3 | 27.8 |
|
| |||||
| Heavy physical activity, ≥5 times/wk (%) | 21.1 | 17.2 | 15.8 | 15.3 | 15.3 |
|
| |||||
| Self-reported diabetes (%) | 4.8 | 5.7 | 6.5 | 7.8 | 10.0 |
|
| |||||
| Dietary Intake | |||||
| Energy (kcal) | 1,553 | 1,557 | 1,563 | 1,573 | 1,567 |
| Total fat (g/1,000 kcal/day) | 28.9 | 32.1 | 33.7 | 34.8 | 36.1 |
| Saturated fat (g/1,000 kcal/day) | 9.1 | 10.0 | 10.4 | 10.7 | 10.9 |
| Heme iron (mg/1,000 kcal/day) | 59.5 | 114.6 | 154.5 | 196.5 | 270.7 |
| Advanced glycation end products | 372.7 | 1364.4 | 1382.8 | 1684.4 | 1698.5 |
|
| |||||
| Alcohol intake (drinks/day) | 0.5 | 0.5 | 0.4 | 0.4 | 0.4 |
| Alcohol intake ≥ 3 drinks/day (%) | 3.4 | 3.4 | 3.0 | 2.4 | 1.7 |
Generalized linear models were used to estimate mean values for the continuous variables and frequencies for dichotomous proportions within each total meat intake quintile.
6,599 (3.52%) men and 4,214 (3.11%) women having missing smoking history data. 2,878 (1.54%) men and 3,762 (2.77%) women have missing BMI data.
Person-years were determined by the date of receipt of the RFQ and date of pancreatic cancer diagnosis, death, emigration out of the registry area, or December 2006, whichever occurred first (8, 13, 16).
Cox proportional hazard models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CI). Trends were estimated using the median value of each quantile. As the dietary variables were correlated with energy, we adjusted for energy using the density method. The distribution of the meat related variables differed by sex (e.g. men consumed higher levels of red meat, women consumed higher levels of white meat); therefore, the quintiles for the meat related variables were based on the distribution of each variable in the cohort by sex. For foods that were consumed by less than 20% of the cohort (pan-fried, oven-broiled, and microwaved meats, DiMeIQx), sex specific categories were created using zero intake as the referent and remaining subjects categorized as quartiles or tertiles. Subjects were merged in sex-specific quantile categories for sex-combined risk estimates. Relevant meat groups were controlled simultaneously in our models (red and white; processed and non-processed; high- and low-temperature cooked; barbequed/grilled, pan-fried, oven-broiled, and microwaved; rare/medium and well/very well done).
Our crude models included age at study entry, energy (kcal) and sex (in sex combined models). Additional variables were included in our models if they were confounders that altered the risk estimate ≥10% or were putative risk factors for pancreatic cancer. Variables tested as potential confounders included smoking history, BMI, self-reported diabetes, saturated fat intake, and alcohol use. Smoking was a confounder in most of the models and saturated fat in our model for total meat. Our final model included age, smoking (never, quit ≥10 y ago, quit 5–9 y ago, quit 1–4 y ago, quit <1 y ago or current and smokes ≤20 cigarettes/day, quit <1 y ago or current and smokes >20 cigarettes/day, and missing), BMI (kg/m2, <18.5, ≥18.5 and <25, ≥25 and <30, ≥30 and <35, ≥35, and missing), self-reported diabetes (yes, no), and energy-adjusted saturated fat (continuous) as well as sex (in sex combined models). The meat mutagens, total heme iron, heme iron from red meat, and advanced glycation end-products were all correlated with each other as they are all present in meat and mutual adjustment of all attenuated associations; therefore we show associations without mutual adjustment. We evaluated interactions by sex, diabetes, BMI and smoking status (never, former, and current) on the association between meat exposure and pancreatic cancer by including cross-product terms for the meat trend variable with the respective effect modifier in the multivariable model. Significant interactions were further stratified. We additionally performed sensitivity analyses of the baseline FFQ meat variables (total, red, white, and high-temperature cooked meats, and heme iron) on the full baseline cohort to assess internal consistency with associations observed for the RFQ cohort. Statistical Analysis Systems (SAS) software was used to perform all statistical analyses, and P values for all tests were 2-tailed (8, 12).
RESULTS
During follow-up up to 10.17 years (median 10.07 y, 2,974,128 person-years of observation), 1,417 individuals (895 men, 522 women) in the risk factor cohort developed exocrine pancreatic cancer (12). The characteristics of our study population according to total meat intake quintile are shown in Table 1. Men and women with higher total meat intake were more likely to be younger, current smokers, of higher BMI, less physically active, and diabetic; they also consumed more red and white meat, total and saturated fat, and heme iron but less alcohol (12). A greater proportion of men consumed more red than white meat while women consumed more white than red meat.
In men and women combined, total, red, and high-temperature cooked meat intake were significantly associated with pancreatic cancer with significant trends across quintiles (Table 2, Q5 vs. Q1 HR=1.20, 95% CI 1.02–1.42, p-trend=0.03; HR=1.22, 95% CI 1.01–1.48, p-trend=0.02; HR=1.21, 95% CI 1.00–1.45, p-trend=0.02, respectively). These associations tended to be stronger in men than in women; however, the interaction by sex was only significant for red meat (p-interaction=0.03; Q5 vs. Q1 men HR=1.36, 95% CI 1.07–1.73, p-trend=0.004; women HR=1.01, 95% CI 0.74–1.38, p-trend=0.91). White meat intake showed a significant positive association in women (Q5 vs. Q1 HR = 1.33, 95% CI 1.02–1.74, p-trend = 0.04) but not in men (HR = 1.05, 95% CI 0.85–1.30, p-trend = 0.63, p-interaction=0.26). Men and women consuming moderate amounts of processed meat (quintiles 2–4 compared to quintile 1) had significantly elevated risk for pancreatic cancer, although extreme intake was not associated. We also conducted substitution model analyses to examine the associations of sub-groups of meat while holding total meat constant and observed associations similar to those obtained by the partition method. For example, holding total meat constant the association between red meat and pancreatic cancer still remained significant. Therefore, our associations were robust with both the partition and substitution methods.
Table 2.
Hazard ratios (HR) and 95% confidence intervals (CI) for baseline meat intake in the NIH-AARP Diet and Health Cohort1
| Variable | Quintile of daily meat intake (g/1,000 kcal) | P trend3 | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| 1 | 2 | 3 | 4 | 5 | ||
|
| ||||||
| Total meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 265/593,355 | 286/595,543 | 271/595,177 | 282/595,934 | 313/594,119 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.09 (0.92–1.29) | 1.06 (0.89–1.25) | 1.14 (0.96–1.34) | 1.32 (1.12–1.56) | 0.0007 |
| Multivariable HR (95% CI)2 | 1.00 (reference) | 1.05 (0.89–1.24) | 1.01 (0.85–1.19) | 1.06 (0.89–1.25) | 1.20 (1.02–1.42) | 0.03 |
| Men | ||||||
| Cases/person-years | 162/340,658 | 188/342,044 | 173/342,094 | 165/342,570 | 207/341,577 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.18 (0.95–1.45) | 1.11 (0.89–1.37) | 1.09 (0.88–1.36) | 1.44 (1.17–1.77) | 0.002 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.12 (0.91–1.39) | 1.04 (0.83–1.29) | 1.00 (0.80–1.24) | 1.27 (1.02–1.57) | 0.08 |
| Women | ||||||
| Cases/person-years | 103/252,697 | 98/253,499 | 98/253,082 | 117/253,369 | 106/252,543 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.96 (0.73–1.26) | 0.98 (0.74–1.29) | 1.20 (0.92–1.56) | 1.14 (0.87–1.49) | 0.13 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.93 (0.71–1.23) | 0.95 (0.72–1.26) | 1.15 (0.88–1.51) | 1.08 (0.82–1.42) | 0.27 |
|
| ||||||
| Red Meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 243/598,789 | 268/596,590 | 284/595,847 | 306/593,567 | 316/589,339 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.11 (0.94–1.33) | 1.21 (1.02–1.44) | 1.35 (1.14–1.60) | 1.47 (1.24–1.74) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.06 (0.88–1.26) | 1.11 (0.92–1.32) | 1.18 (0.99–1.42) | 1.22 (1.01–1.48) | 0.02 |
| Men | ||||||
| Cases/person-years | 148/344,177 | 159/342,803 | 180/342,009 | 191/341,440 | 217/338,513 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.09 (0.87–1.37) | 1.27 (1.02–1.58) | 1.38 (1.11–1.71) | 1.66 (1.34–2.05) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.02 (0.81–1.28) | 1.14 (0.91–1.44) | 1.20 (0.95–1.51) | 1.36 (1.07–1.73) | 0.004 |
| Women | ||||||
| Cases/person-years | 95/254,612 | 109/253,788 | 104/253,838 | 115/252,127 | 99/250,826 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.16 (0.88–1.52) | 1.12 (0.85–1.48) | 1.29 (0.98–1.69) | 1.16 (0.87–1.54) | 0.3224 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.11 (0.84–1.47) | 1.05 (0.78–1.40) | 1.17 (0.87–1.56) | 1.01 (0.74–1.38) | 0.91 |
|
| ||||||
| White Meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 284/587,575 | 294/593,947 | 272/595,811 | 272/597,199 | 295/599,601 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.01 (0.86–1.19) | 0.94 (0.79–1.11) | 0.96 (0.81–1.13) | 1.09 (0.93–1.29) | 0.25 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.04 (0.88–1.23) | 0.98 (0.83–1.16) | 1.02 (0.86–1.20) | 1.15 (0.98–1.36) | 0.08 |
| Men | ||||||
| Cases/person-years | 185/336,636 | 182/340,946 | 183/342,711 | 171/343,262 | 174/345,387 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.96 (0.78–1.17) | 0.96 (0.78–1.18) | 0.92 (0.75–1.14) | 0.99 (0.81–1.22) | 0.96 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.99 (0.81–1.22) | 1.01 (0.82–1.24) | 0.99 (0.80–1.22) | 1.05 (0.85–1.30) | 0.63 |
| Women | ||||||
| Cases/person-years | 99/250,939 | 112/253,001 | 89/253,100 | 101/253,936 | 121/254,214 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.11 (0.85–1.46) | 0.89 (0.67–1.19) | 1.03 (0.78–1.36) | 1.28 (0.98–1.67) | 0.07 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.13 (0.86–1.49) | 0.92 (0.69–1.23) | 1.07 (0.81–1.42) | 1.33 (1.02–1.74) | 0.04 |
|
| ||||||
| Processed Meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 228/597,449 | 308/597,036 | 301/595,084 | 313/593,677 | 267/590,886 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.36 (1.14–1.62) | 1.33 (1.12–1.59) | 1.38 (1.15–1.64) | 1.17 (0.98–1.40) | 0.73 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.29 (1.08–1.53) | 1.23 (1.02–1.46) | 1.24 (1.04–1.49) | 1.02 (0.85–1.24) | 0.23 |
| Men | ||||||
| Cases/person-years | 141/343,575 | 200/342,861 | 181/341,653 | 200/341,213 | 173/339,640 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.43 (1.15–1.77) | 1.30 (1.04–1.62) | 1.42 (1.14–1.77) | 1.23 (0.98–1.55) | 0.49 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.33 (1.07–1.66) | 1.17 (0.93–1.47) | 1.26 (1.00–1.58) | 1.05 (0.83–1.33) | 0.44 |
| Women | ||||||
| Cases/person-years | 87/253,874 | 108/254,176 | 120/253,431 | 113/252,464 | 94/251,246 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.25 (0.94–1.66) | 1.39 (1.05–1.84) | 1.30 (0.98–1.74) | 1.06 (0.79–1.43) | 0.60 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.22 (0.91–1.62) | 1.33 (1.00–1.76) | 1.23 (0.91–1.65) | 0.98 (0.72–1.33) | 0.25 |
|
| ||||||
| Meat cooked at high temperatures | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 245/598,389 | 264/596,728 | 293/595,287 | 301/593,045 | 314/590,684 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.08 (0.91–1.29) | 1.23 (1.03–1.46) | 1.29 (1.09–1.54) | 1.43 (1.20–1.70) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.02 (0.86–1.22) | 1.12 (0.94–1.34) | 1.15 (0.96–1.37) | 1.21 (1.00–1.46) | 0.02 |
| Men | ||||||
| Cases/person-years | 146/344,080 | 161/343,002 | 174/341,836 | 210/340,482 | 204/339,542 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.12 (0.89–1.40) | 1.24 (0.99–1.55) | 1.54 (1.24–1.91) | 1.58 (1.27–1.97) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.04 (0.83–1.31) | 1.11 (0.88–1.40) | 1.33 (1.06–1.68) | 1.30 (1.03–1.66) | 0.009 |
| Women | ||||||
| Cases/person-years | 99/254,309 | 103/253,725 | 119/253,451 | 91/252,563 | 110/251,142 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.04 (0.78–1.37) | 1.20 (0.92–1.57) | 0.94 (0.70–1.26) | 1.20 (0.91–1.59) | 0.30 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.99 (0.75–1.32) | 1.13 (0.86–1.49) | 0.87 (0.64–1.17) | 1.07 (0.79–1.45) | 0.85 |
Cohort that successfully completed the meat module on the RFQ: n=322,846 subjects (1,417 cases), 187,265 men (895 cases), and 135,581 women (522 cases).
Cox proportional hazard models used to calculate hazard ratios with age as the time metric. All nutrients are adjusted for energy by the density method with energy also in the model. Models are additionally adjusted for smoking (never, quit ≥10 y, quit 5 to 9 y ago, quit 1 to 4 y ago, quit <1 y ago or current and smoked ≤20 cigarettes/day, quit <1 y ago or current and smoked >20 cigarettes/day, and missing), BMI (kg/m2, <18.5, ≥18.5 and <25, ≥25 and <30, ≥30 and <35, ≥35, and missing), self-reported diabetes (yes, no), and energy-adjusted saturated fat (continuous) P value for interaction by sex: total meat = 0.46, red meat = 0.03, white meat = 0.26, processed meat = 0.34, high temperature cooked meat = 0.12.
P trend calculated using median values for each quintile
Sex combined models additionally adjusted for sex.
Certain meat preparation methods were associated with pancreatic cancer (Table 3). In sex combined models, those consuming the most barbequed meat compared to the least had a significant 24% (95% CI 1.03–1.50) increased pancreatic cancer risk with significant trends across the quintiles (p-trend = 0.007). In sex-stratified analysis, men in the highest quintile of barbequed meat intake and quartile of broiled meat intake had a 33% (95% CI 1.05–1.68) and 34% (95% CI 1.09–1.64) greater risk, respectively, than those with the lowest intake. This risk increased across the quintiles and quartile (p-trends = 0.004 and 0.01), respectively. There was a borderline non-significant interaction for broiled meat by sex (p-interaction = 0.07). Consumption of pan-fried and microwaved meats was not associated with pancreatic cancer. With regards to doneness levels, well/very-well done meat was associated with a significantly elevated risk for pancreatic cancer in sex combined and stratified models (Q5 vs. Q1 men and women HR=1.32, 95% CI 1.10–1.58, p-trend = 0.005; men HR=1.33, 95% CI 1.05–1.67, p-trend=0.01). Rare/medium done cooked meat showed non-significant positive associations with a significant trend in sex combined models (p-trend=0.03).
Table 3.
Hazard ratios and 95% confidence intervals (CI) for meat-cooking methods and doneness levels in the NIH-AARP Diet and Health Study 1
| Variable | Quintile of daily meat intake (g/1,000 kcal) | P trend3 | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| 1 | 2 | 3 | 4 | 5 | ||
|
| ||||||
| Grilled or barbecued meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 280/590,431 | 287/594,951 | 266/595,472 | 278/596,239 | 306/597,039 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.07 (0.90–1.26) | 1.02 (0.85–1.21) | 1.11 (0.93–1.33) | 1.32 (1.10–1.59) | 0.0005 |
| Multivariable HR (95% CI)2 | 1.00 (reference) | 1.07 (0.90–1.26) | 0.99 (0.83–1.18) | 1.07 (0.89–1.28) | 1.24 (1.03–1.50) | 0.007 |
| Men | ||||||
| Cases/person-years | 178/338,896 | 165/341,761 | 170/342,324 | 179/342,876 | 203/343,086 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.99 (0.80–1.24) | 1.06 (0.85–1.33) | 1.18 (0.94–1.47) | 1.43 (1.14–1.80) | 0.0002 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.99 (0.80–1.23) | 1.03 (0.83–1.29) | 1.12 (0.89–1.41) | 1.33 (1.05–1.68) | 0.004 |
| Women | ||||||
| Cases/person-years | 102/251,536 | 122/253,191 | 96/253,148 | 99/253,363 | 103/253,953 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.18 (0.90–1.54) | 0.94 (0.70–1.25) | 1.00 (0.74–1.34) | 1.12 (0.82–1.53) | 0.57 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.18 (0.90–1.55) | 0.93 (0.70–1.24) | 0.98 (0.72–1.31) | 1.09 (0.80–1.49) | 0.71 |
|
| ||||||
| Pan-fried meat (quartiles) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 526/1,102,279 | 207/472,711 | 215/469,921 | 233/465,750 | 236/463,472 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.98 (0.83–1.15) | 1.01 (0.86–1.19) | 1.08 (0.92–1.27) | 1.14 (0.96–1.34) | 0.10 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.98 (0.83–1.16) | 0.98 (0.83–1.16) | 1.01 (0.86–1.19) | 1.02 (0.86–1.21) | 0.71 |
| Men | ||||||
| Cases/person-years | 319/594,905 | 124/281,626 | 135/279,941 | 161/277,080 | 156/275,390 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.90 (0.73–1.11) | 0.97 (0.78–1.19) | 1.14 (0.94–1.39) | 1.18 (0.96–1.44) | 0.06 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.91 (0.73–1.13) | 0.93 (0.75–1.15) | 1.05 (0.86–1.29) | 1.04 (0.84–1.29) | 0.51 |
| Women | ||||||
| Cases/person-years | 207/507,374 | 83/191,085 | 80/189,980 | 72/186,670 | 80/188,082 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.12 (0.86–1.45) | 1.11 (0.85–1.44) | 0.98 (0.74–1.28) | 1.07 (0.81–1.40) | 0.86 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.13 (0.87–1.46) | 1.08 (0.83–1.41) | 0.92 (0.70–1.22) | 0.99 (0.74–1.31) | 0.74 |
|
| ||||||
| Oven-broiled meat (tertiles) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 838/1,818,804 | 205/384,723 | 170/385,123 | 204/385,483 | -- | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.15 (0.99–1.35) | 0.96 (0.81–1.14) | 1.19 (1.01–1.41) | -- | 0.09 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.13 (0.96–1.32) | 0.93 (0.78–1.11) | 1.15 (0.97–1.36) | -- | 0.16 |
| Men | ||||||
| Cases/person-years | 507/1,013,280 | 132/231,299 | 107/232,031 | 149/232,331 | -- | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.18 (0.97–1.44) | 0.96 (0.77–1.19) | 1.39 (1.14–1.70) | -- | 0.006 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.15 (0.94–1.40) | 0.93 (0.75–1.16) | 1.34 (1.09–1.64) | -- | 0.01 |
| Women | ||||||
| Cases/person-years | 331/805,524 | 73/153,423 | 63/153,091 | 55/153,152 | -- | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.10 (0.85–1.43) | 0.96 (0.73–1.27) | 0.86 (0.63–1.17) | -- | 0.35 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.08 (0.83–1.40) | 0.93 (0.71–1.24) | 0.84 (0.61–1.14) | -- | 0.28 |
|
| ||||||
| Sautéed, baked or microwaved meat (quartiles) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 589/1,291,368 | 208/420,332 | 215/419,049 | 194/420,675 | 211/422,710 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.12 (0.94–1.32) | 1.17 (0.99–1.37) | 1.06 (0.89–1.25) | 1.18 (1.00–1.40) | 0.10 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.09 (0.92–1.29) | 1.13 (0.96–1.33) | 1.03 (0.87–1.23) | 1.17 (0.99–1.39) | 0.08 |
| Men | ||||||
| Cases/person-years | 430/865,872 | 119/210,873 | 130/209,788 | 106/210,391 | 110/212,018 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.13 (0.91–1.40) | 1.28 (1.04–1.56) | 1.06 (0.85–1.33) | 1.13 (0.91–1.42) | 0.46 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.10 (0.89–1.37) | 1.22 (1.00–1.50) | 1.03 (0.83–1.29) | 1.12 (0.90–1.40) | 0.45 |
| Women | ||||||
| Cases/person-years | 159/425,495 | 89/209,458 | 85/209,261 | 88/210,284 | 101/210,692 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.09 (0.84–1.43) | 0.98 (0.75–1.30) | 1.01 (0.77–1.33) | 1.18 (0.90–1.55) | 0.21 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.07 (0.82–1.41) | 0.97 (0.73–1.27) | 1.00 (0.76–1.32) | 1.19 (0.91–1.55) | 0.17 |
|
| ||||||
| Rare/medium done cooked meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 374/793,313 | 237/544,683 | 271/545,249 | 265/544,831 | 270/546,056 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.95 (0.80–1.12) | 1.11 (0.94–1.30) | 1.12 (0.95–1.33) | 1.21 (1.01–1.44) | 0.01 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.96 (0.81–1.13) | 1.10 (0.94–1.29) | 1.11 (0.94–1.31) | 1.17 (0.99–1.40) | 0.03 |
| Men | ||||||
| Cases/person-years | 227/439,578 | 154/316,768 | 168/317,124 | 168/317,341 | 178/318,132 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.00 (0.81–1.23) | 1.12 (0.91–1.38) | 1.16 (0.94–1.44) | 1.31 (1.05–1.63) | 0.01 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.01 (0.82–1.24) | 1.10 (0.90–1.36) | 1.13 (0.91–1.40) | 1.23 (0.99–1.54) | 0.05 |
| Women | ||||||
| Cases/person-years | 147/353,736 | 83/227,915 | 103/228,125 | 97/227,490 | 92/227,924 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.88 (0.67–1.15) | 1.09 (0.84–1.42) | 1.05 (0.80–1.39) | 1.04 (0.77–1.40) | 0.37 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.89 (0.67–1.16) | 1.11 (0.85–1.43) | 1.08 (0.82–1.41) | 1.07 (0.80–1.42) | 0.35 |
|
| ||||||
| Well/very well-done cooked meat | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 254/595,971 | 302/593,133 | 264/593,450 | 277/595,427 | 320/596,152 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.20 (1.01–1.43) | 1.07 (0.90–1.29) | 1.14 (0.95–1.37) | 1.37 (1.14–1.65) | 0.003 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.16 (0.97–1.37) | 1.03 (0.86–1.23) | 1.11 (0.92–1.33) | 1.32 (1.10–1.58) | 0.005 |
| Men | ||||||
| Cases/person-years | 159/342,670 | 180/340,844 | 165/341,142 | 191/342,311 | 200/341,975 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.16 (0.93–1.44) | 1.10 (0.87–1.38) | 1.30 (1.04–1.63) | 1.42 (1.13–1.79) | 0.003 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.11 (0.89–1.37) | 1.04 (0.82–1.30) | 1.23 (0.98–1.55) | 1.33 (1.05–1.67) | 0.01 |
| Women | ||||||
| Cases/person-years | 95/253,300 | 122/252,289 | 99/252,308 | 86/253,115 | 120/254,178 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.26 (0.96–1.66) | 1.03 (0.76–1.38) | 0.89 (0.65–1.22) | 1.27 (0.94–1.73) | 0.29 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.23 (0.94–1.62) | 1.02 (0.76–1.37) | 0.90 (0.66–1.23) | 1.30 (0.97–1.75) | 0.23 |
Cohort that successfully completed the meat module questions on the RFQ: n=322,846 subjects (1,417 cases), 187,265 men (895 cases), and 135,581 women (522 cases).
Cox proportional hazard models used to calculate hazard ratios with age as the time metric. All nutrients are adjusted for energy by the density method with energy also in the model. Models are additionally adjusted for smoking (never, quit ≥10 y, quit 5 to 9 y ago, quit 1 to 4 y ago, quit <1 y ago or current and smoked ≤20 cigarettes/day, quit <1 y ago or current and smoked >20 cigarettes/day, and missing), BMI (kg/m2, <18.5, ≥18.5 and <25, ≥25 and <30, ≥30 and <35, ≥35, and missing), self-reported diabetes (yes, no), and energy-adjusted saturated fat (continuous). P value for interaction by sex: grilled/barbequed meat = 0.31, pan-fried meat = 0.45, broiled meat = 0.07, microwaved meat = 0.23, rare/medium done cooked meat = 0.66, well/very well-done cooked meat = 0.54.
P trend calculated using median values for each quintile
Sex combined models additionally adjusted for sex.
Heme iron showed borderline non-significant positive associations with pancreatic cancer in men and sex combined models (Table 4). In multivariable models, before energy-adjusting for saturated fats, the association between heme iron and pancreatic cancer was significant (Q5 vs. Q1 HR=1.21, 95% CI 1.00–1.45, p-trend=0.03). This association was significant in men (Q5 vs. Q1 HR=1.31, 95% CI 1.04–1.65, p-trend=0.02) but not in women (Q5 vs. Q1 HR=1.05, 95% CI 0.79–1.42, p-trend=0.83). Consumption of heme iron from red meat was significantly associated with pancreatic cancer in sex combined models and among men (Q4 vs. Q1 men and women HR=1.21, 95% CI 1.01–1.45, p-trend=0.04; men HR=1.34, 95% CI 1.06–1.69, p-trend=0.02). Overall mutagenic activity intake was significantly associated with increased pancreatic cancer in men (Q5 vs. Q1 HR=1.30, 95% CI 1.04–1.63, p-trend=0.13). None of the other meat mutagens showed significant associations with pancreatic cancer.
Table 4.
Hazard ratios and 95% confidence intervals (CI) for heme iron and meat mutagen intake in the NIH-AARP Diet and Health Study 1
| Variable | Quintile of heme intake (mg/1,000 kcal) or daily meat mutagen intake (ng/1,000 kcal) | P trend3 | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| 1 | 2 | 3 | 4 | 5 | ||
|
| ||||||
| Heme iron intake (mg/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 241/296,684 | 274/596,346 | 302/596,125 | 301/593,768 | 299/591,211 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.14 (0.96–1.35) | 1.28 (1.08–1.52) | 1.32 (1.11–1.56) | 1.38 (1.17–1.64) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.07 (0.90–1.28) | 1.18 (0.99–1.40) | 1.17 (0.98–1.39) | 1.17 (0.97–1.41) | 0.08 |
| Men | ||||||
| Cases/person-years | 146/342,879 | 166/342,461 | 193/342,647 | 192/341,077 | 198/339,879 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.14 (0.91–1.43) | 1.36 (1.10–1.69) | 1.39 (1.12–1.73) | 1.52 (1.22–1.88) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.06 (0.85–1.33) | 1.22 (0.98–1.53) | 1.21 (0.96–1.51) | 1.24 (0.98–1.57) | 0.07 |
| Women | ||||||
| Cases/person-years | 95/253,805 | 108/253,884 | 109/253,478 | 109/252,690 | 101/251,332 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.13 (0.86–1.49) | 1.16 (0.88–1.53) | 1.20 (0.91–1.58) | 1.18 (0.89–1.56) | 0.28 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.09 (0.83–1.44) | 1.10 (0.83–1.46) | 1.11 (0.84–1.48) | 1.06 (0.78–1.43) | 0.82 |
|
| ||||||
| Heme from red meat (mg/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 238/597,849 | 282/596,472 | 278/595,592 | 312/593,486 | 307/590,734 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.19 (1.00–1.41) | 1.20 (1.01–1.42) | 1.38 (1.16–1.63) | 1.43 (1.20–1.70) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.12 (0.94–1.33) | 1.09 (0.91–1.30) | 1.21 (1.01–1.45) | 1.20 (0.99–1.44) | 0.04 |
| Men | ||||||
| Cases/person-years | 140/343,676 | 173/342,480 | 176/342,084 | 204/341,016 | 202/339,687 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.25 (1.00–1.56) | 1.30 (1.04–1.63) | 1.55 (1.25–1.92) | 1.62 (1.30–2.01) | <0.0001 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.16 (0.93–1.46) | 1.17 (0.93–1.47) | 1.34 (1.06–1.69) | 1.32 (1.04–1.69) | 0.02 |
| Women | ||||||
| Cases/person-years | 98/254,173 | 109/253,993 | 102/253,508 | 108/252,470 | 105/251,047 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.10 (0.84–1.45) | 1.04 (0.79–1.38) | 1.13 (0.86–1.49) | 1.16 (0.88–1.53) | 0.30 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.06 (0.80–1.39) | 0.97 (0.73–1.30) | 1.03 (0.77–1.38) | 1.02 (0.75–1.38) | 0.94 |
|
| ||||||
| Overall mutagenic activity (revertant colonies/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 258/595,113 | 297/593,480 | 290/593,917 | 281/594,746 | 291/596,878 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.18 (1.00–1.39) | 1.18 (1.00–1.40) | 1.18 (0.99–1.39) | 1.26 (1.06–1.49) | 0.03 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.11 (0.94–1.31) | 1.09 (0.92–1.29) | 1.07 (0.90–1.27) | 1.13 (0.95–1.34) | 0.28 |
| Men | ||||||
| Cases/person-years | 143/341,599 | 197/340,877 | 175/341,981 | 190/341,495 | 190/342,990 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.41 (1.13–1.75) | 1.28 (1.03–1.60) | 1.43 (1.15–1.77) | 1.47 (1.18–1.83) | 0.008 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.32 (1.06–1.64) | 1.17 (0.94–1.47) | 1.28 (1.03–1.60) | 1.30 (1.04–1.63) | 0.13 |
| Women | ||||||
| Cases/person-years | 115/253,513 | 100/252,603 | 115/251,937 | 91/253,251 | 101/253,887 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.89 (0.68–1.17) | 1.06 (0.82–1.37) | 0.86 (0.65–1.13) | 0.99 (0.75–1.29) | 0.98 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.85 (0.65–1.12) | 1.00 (0.76–1.30) | 0.80 (0.60–1.06) | 0.92 (0.70–1.21) | 0.71 |
|
| ||||||
| DiMeIQx** (ng/1,000 kcal) (quartiles) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 494/1,076,064 | 217/475,182 | 218/474,216 | 240/474,017 | 244/474,653 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.95 (0.81–1.12) | 0.98 (0.83–1.14) | 1.10 (0.95–1.29) | 1.16 (0.99–1.35) | 0.02 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.97 (0.82–1.13) | 0.96 (0.82–1.13) | 1.06 (0.91–1.24) | 1.09 (0.94–1.27) | 0.13 |
| Men | ||||||
| Cases/person-years | 306/602,034 | 134/277,205 | 132/276,672 | 158/276,640 | 165/276,392 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.93 (0.76–1.13) | 0.93 (0.75–1.12) | 1.14 (0.94–1.38) | 1.22 (1.01–1.48) | 0.007 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.94 (0.77–1.16) | 0.91 (0.74–1.12) | 1.09 (0.90–1.33) | 1.14 (0.95–1.38) | 0.07 |
| Women | ||||||
| Cases/person-years | 192/474,030 | 83/197,978 | 86/197,544 | 82/197,377 | 79/198,261 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.01 (0.78–1.30) | 1.06 (0.83–1.37) | 1.04 (0.81–1.35) | 1.04 (0.80–1.35) | 0.80 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.01 (0.78–1.31) | 1.05 (0.81–1.36) | 1.02 (0.78–1.32) | 1.00 (0.77–1.31) | 0.99 |
|
| ||||||
| MeIQx (ng/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 266/595,605 | 269/595,642 | 287/595,857 | 309/593,701 | 286/592,328 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.03 (0.87–1.22) | 1.11 (0.94–1.31) | 1.22 (1.04–1.44) | 1.16 (0.98–1.37) | 0.04 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.00 (0.84–1.18) | 1.04 (0.88–1.24) | 1.12 (0.95–1.32) | 1.02 (0.86–1.21) | 0.73 |
| Men | ||||||
| Cases/person-years | 162/342,183 | 162/343,158 | 181/342,592 | 205/341,273 | 185/339,735 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.02 (0.82–1.26) | 1.15 (0.93–1.42) | 1.32 (1.08–1.63) | 1.23 (1.00–1.52) | 0.03 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.98 (0.79–1.22) | 1.06 (0.86–1.32) | 1.19 (0.96–1.47) | 1.05 (0.84–1.31) | 0.54 |
| Women | ||||||
| Cases/person-years | 104/253,421 | 107/253,483 | 106/253,365 | 104/252,428 | 101/252,593 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.06 (0.81–1.38) | 1.06 (0.81–1.38) | 1.06 (0.81–1.39) | 1.05 (0.80–1.38) | 0.84 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.04 (0.79–1.36) | 1.02 (0.78–1.34) | 1.01 (0.76–1.33) | 0.97 (0.73–1.29) | 0.69 |
|
| ||||||
| PhIP (ng/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 282/592,538 | 286/594,293 | 289/594,297 | 268/595,216 | 292/597,790 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.03 (0.88–1.22) | 1.08 (0.91–1.27) | 1.02 (0.86–1.21) | 1.14 (0.97–1.35) | 0.10 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.99 (0.84–1.16) | 1.01 (0.85–1.19) | 0.95 (0.80–1.12) | 1.06 (0.89–1.25) | 0.39 |
| Men | ||||||
| Cases/person-years | 164/340,098 | 185/341,296 | 187/341,818 | 160/342,443 | 199/343,287 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 1.14 (0.93–1.41) | 1.19 (0.96–1.46) | 1.03 (0.83–1.29) | 1.32 (1.07–1.62) | 0.03 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 1.08 (0.87–1.34) | 1.10 (0.89–1.36) | 0.94 (0.76–1.18) | 1.20 (0.97–1.48) | 0.16 |
| Women | ||||||
| Cases/person-years | 118/252,440 | 101/252,996 | 102/252,479 | 108/252,772 | 93/254,504 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.88 (0.68–1.15) | 0.92 (0.71–1.20) | 1.01 (0.78–1.31) | 0.90 (0.68–1.18) | 0.72 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.85 (0.65–1.11) | 0.88 (0.67–1.15) | 0.95 (0.73–1.24) | 0.85 (0.64–1.12) | 0.52 |
|
| ||||||
| BaP (ng/1,000 kcal) | ||||||
| Men and women combined4 | ||||||
| Cases/person-years | 308/590,261 | 282/593,461 | 282/595,836 | 260/597,161 | 285/597,413 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.93 (0.79–1.09) | 0.94 (0.80–1.11) | 0.89 (0.76–1.05) | 1.04 (0.88–1.22) | 0.37 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.92 (0.78–1.08) | 0.95 (0.81–1.12) | 0.88 (0.74–1.04) | 0.98 (0.83–1.16) | 0.86 |
| Men | ||||||
| Cases/person-years | 199/338,320 | 178/341,441 | 172/342,464 | 162/343,290 | 184/343,427 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.90 (0.74–1.11) | 0.89 (0.72–1.09) | 0.85 (0.69–1.05) | 1.02 (0.83–1.25) | 0.46 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.90 (0.74–1.11) | 0.90 (0.73–1.11) | 0.84 (0.68–1.03) | 0.95 (0.78–1.17) | 0.94 |
| Women | ||||||
| Cases/person-years | 109/251,941 | 104/252,020 | 110/253,372 | 98/253,871 | 101/253,986 | |
| Age-adjusted HR (95% CI) | 1.00 (reference) | 0.98 (0.75–1.28) | 1.04 (0.80–1.36) | 0.97 (0.74–1.27) | 1.07 (0.81–1.40) | 0.27 |
| Multivariable HR (95% CI) 2 | 1.00 (reference) | 0.96 (0.73–1.25) | 1.05 (0.80–1.37) | 0.96 (0.73–1.26) | 1.04 (0.79–1.37) | 0.74 |
Cohort that successfully completed the meat module questions on the RFQ: n=322,846 subjects (1,417 cases), 187,265 men (895 cases), and 135,581 women (522 cases).
Cox proportional hazard models used to calculate hazard ratios. All nutrients are adjusted for energy by the density method with energy also in the model. Models are additionally adjusted for smoking (never, quit ≥10 y, quit 5 to 9 y ago, quit 1 to 4 y ago, quit <1 y ago or current and smoked ≤20 cigarettes/day, quit <1 y ago or current and smoked >20 cigarettes/day, and missing), BMI (kg/m2, <18.5, ≥18.5 and <25, ≥25 and <30, ≥30 and <35, ≥35, and missing), self-reported diabetes (yes, no), and energy-adjusted saturated fat (continuous). P value for interaction by sex: heme iron = 0.17, heme iron from red meat = 0.11, overall mutagenic activity = 0.15, DiMeIQx = 0.25, MeIQx = 0.32, PhIP = 0.12, BaP = 0.96
P trend calculated using median values for each quintile
Sex combined models additionally adjusted for sex.
There was a significant interaction for total meat intake by smoking status in men (p-interaction = 0.01) such that in stratified and joint effects models positive associations were present among former and current smokers but not never smokers (Table 5). BMI and diabetes did not significantly modify any of the associations. None of the risk estimates for the various meat exposures significantly changed in lag analyses excluding the first two years of follow-up.
Table 5.
Multivariable adjusted hazard ratios (HR) and 95% confidence intervals (CI) for total meat intake by smoking and joint effects in the NIH-AARP Diet and Health Study 1,2
| Smoking status | Quintile of daily meat intake (g/1,000 kcal) | P trend3 | |||||
|---|---|---|---|---|---|---|---|
|
| |||||||
| 1 | 2 | 3 | 4 | 5 | |||
|
| |||||||
| Men and Women Combined4 | Stratified by smoking | ||||||
| Never (n=411) | 1.00 (reference) | 0.96 (0.72–1.29) | 0.79 (0.58–1.08) | 0.94 (0.69–1.27) | 0.90 (0.66–1.23) | 0.53 | |
| Former (n=736) | 1.00 (reference) | 1.19 (0.93–1.52) | 1.20 (0.94–1.53) | 1.22 (0.95–1.56) | 1.45 (1.14–1.83) | 0.004 | |
| Current (n=224) | 1.00 (reference) | 0.80 (0.52–1.24) | 0.94 (0.61–1.43) | 0.89 (0.58–1.36) | 1.15 (0.76–1.73) | 0.33 | |
| Joint Effects | |||||||
| Never | 1.00 (reference) | 0.97 (0.72–1.29) | 0.79 (0.58–1.08) | 0.94 (0.69–1.26) | 0.90 (0.66–1.22) | ||
| Former | 1.58 (0.99–2.53) | 1.90 (1.20–3.01) | 1.91 (1.20–3.03) | 1.94 (1.22–3.09) | 2.32 (1.46–3.67) | ||
| Current | 2.43 (1.64–3.60) | 1.94 (1.29–2.92) | 2.26 (1.53–3.35) | 2.14 (1.44–3.17) | 2.77 (1.91–4.01) | ||
|
| |||||||
| Men | Stratified by smoking | ||||||
| Never (n=215) | 1.00 (reference) | 0.91 (0.61–1.35) | 0.69 (0.45–1.06) | 0.75 (0.49–1.14) | 0.75 (0.49–1.15) | 0.14 | |
| Former (n=522) | 1.00 (reference) | 1.25 (0.94–1.66) | 1.19 (0.89–1.59) | 1.10 (0.81–1.49) | 1.44 (1.08–1.92) | 0.04 | |
| Current (n=127) | 1.00 (reference) | 0.98 (0.53–1.79) | 1.12 (0.62–2.03) | 1.11 (0.61–2.01) | 1.55 (0.88–2.73) | 0.08 | |
| Joint Effects | |||||||
| Never | 1.00 (reference) | 0.94 (0.63–1.39) | 0.72 (0.47–1.10) | 0.79 (0.52–1.20) | 0.81 (0.53–1.23) | ||
| Former | 1.73 (0.96–3.11) | 2.14 (1.20–3.84) | 2.04 (1.14–3.67) | 1.89 (1.05–3.41) | 2.47 (1.38–4.42) | ||
| Current | 2.38 (1.41–4.03) | 2.20 (1.29–3.75) | 2.50 (1.50–4.18) | 2.42 (1.46–4.02) | 3.33 (2.10–5.28) | ||
|
| |||||||
| Women | Stratified by smoking | ||||||
| Never (n=196) | 1.00 (reference) | 1.03 (0.67–1.59) | 0.92 (0.58–1.45) | 1.18 (0.77–1.83) | 1.09 (0.69–1.72) | 0.57 | |
| Former (n=214) | 1.00 (reference) | 1.04 (0.65–1.68) | 1.21 (0.77–1.92) | 1.50 (0.96–2.33) | 1.45 (0.93–2.27) | 0.03 | |
| Current (n=97) | 1.00 (reference) | 0.66 (0.35–1.24) | 0.79 (0.43–1.44) | 0.71 (0.38–1.33) | 0.81 (0.44–1.50) | 0.67 | |
| Joint Effects | |||||||
| Never | 1.00 (reference) | 1.01 (0.65–1.55) | 0.89 (0.57–1.39) | 1.13 (0.74–1.73) | 1.02 (0.65–1.61) | ||
| Former | 1.10 (0.49–2.45) | 1.16 (0.52–2.58) | 1.36 (0.62–3.00) | 1.68 (0.77–3.67) | 1.63 (0.75–3.55) | ||
| Current | 1.94 (1.02–3.66) | 1.32 (0.67–2.59) | 1.58 (0.83–3.03) | 1.44 (0.74–2.79) | 1.65 (0.86–3.17) | ||
Cohort that successfully completed the meat module and smoking questions on the RFQ: n=312,033 subjects (1,371 cases), 180,666 men (864 cases), and 131,367 women (507 cases).
Cox proportional hazard models used to calculate hazard ratios with age as the time metric. All nutrients are adjusted for energy by the density method with energy also in the model. Models are additionally adjusted for smoking (never, quit ≥10 y, quit 5 to 9 y ago, quit 1 to 4 y ago, quit <1 y ago or current and smoked ≤20 cigarettes/day, quit <1 y ago or current and smoked >20 cigarettes/day, and missing), BMI (kg/m2, <18.5, ≥18.5 and <25, ≥25 and <30, ≥30 and <35, ≥35, and missing), self-reported diabetes (yes, no), and energy-adjusted saturated fat (continuous). P value for interaction by smoking: men and women combined = 0.07, men = 0.01, women = 0.99.
P trend calculated using median values for each quintile
Sex combined models additionally adjusted for sex.
DISCUSSION
In this large prospective cohort of AARP members (8), there were significant associations overall between intake of total, red, high-temperature cooked, grilled/barbequed, and well/very well done meats, heme iron from red meat and pancreatic cancer. When analyses were stratified by sex, white meat was the only dietary exposure that was significantly associated with increased risk for pancreatic cancer in women. Consumption of total, red, high-temperature cooked, grilled/barbequed, oven-broiled, and well/very well done meats and heme iron from red meat appreciably was associated with elevated risk more strongly in men compared to that in women, although the interaction by sex was only significant for red meat. None of the meat mutagens were appreciably associated with pancreatic cancer, although a significant threshold association was observed for overall mutagenic activity in men only. Smoking tended to modify the association for total meat such that the positive association was strongest among former smokers, particularly among men. In the joint analysis, associations were significant in former and current smokers, but not in never smokers. To the best of our knowledge, this is the largest prospective study to examine meat cooking methods, related mutagens, and heme iron in relation to pancreatic cancer.
Our results add to the evidence that meat intake may play a role in pancreatic cancer development (3). The WCRF/AICR’s Continuous Update Project concluded that there was suggestive evidence that red and processed meat increased the risk for pancreatic cancer. The panel found a positive, borderline non-significant, association between red meat intake and pancreatic cancer (per 100 g compared to 20 g RR=1.19, 95% CI 0.98–1.86), and in the meta-analysis the increased risk was most apparent in men (RR=1.43, 95% CI 1.10–1.86) but not women (RR=1.06, 95% CI 0.86–1.30). Previous cohort studies have found significant findings in men (8, 22) but not in women (23). Two meta-analyses on 10 (4) and 11 (5) prospective studies found no significant relation between red meat intake and pancreatic cancer. However, both meta-analyses reported a significant P-heterogeneity among the cohort studies due to differing red meat definitions. Overall the WCRF showed a borderline significant increased risk for processed meat (per 50 g/day increase, RR=1.17, 95% CI 1.01–1.34) which was significant in men (per 50 g/day increase, RR=1.21, 95% CI 1.01–1.45) but not women (per 50 g/day increase, RR=1.09, 95% CI 0.69–1.73) (3). A meta-analysis on 11 prospective studies also found a significant association between consumption of processed meat and pancreatic cancer (5). In the present analysis, we observed significant positive associations at intermediate quantiles of processed meat intake for both men and women but no associations comparing intake extremes. We also observed significant positive associations with a significant trend across quintiles for white meat intake among women. To the best of our knowledge, no studies have found a significant positive association between pancreatic cancer and white meat consumption. A large meta-analysis on poultry consumption and pancreatic cancer found a non-significant inverse association (combined OR=0.84, 95% CI 0.68–1.06) in four case control studies and a non-significant association (total RR=1.02 95% CI 0.84–1.24) in six cohort studies (4). The positive associations that we observe for red, processed, and white meat intake were independent of saturated fat intake and other potential confounders. The present results contrast earlier analyses conducted in NIH-AARP with less follow-up time and fewer cancer cases. Previously, we observed positive associations for total meat for sexes combined and red meat among men but no associations for processed meat or poultry (8). The positive association for total meat seen in the highest quintile may be explained by high red meat intake among these individuals.
In the present study we observe significant associations with higher intake of high temperature cooked, grilled/barbequed, and well/very well done meats with stronger associations among men although the interaction by sex was not significant. Three retrospective case-controls studies and two cohort studies have examined meat cooking methods, doneness levels, or mutagens and pancreatic cancer risk with most showing positive associations (6–10). Among prospective studies, in the PLCO cohort, well to very well done red and barbequed red meat were associated with significantly elevated pancreatic cancer risk, and in an earlier analysis in the NIH-AARP study grilled/barbequed and broiled meats were positively associated with pancreatic cancer but only among men (8). HCAs (DiMeIQx, MeIQx, and PhIP) and PAHs (BaP) present in barbequed and high temperature cooked meat could explain these associations and have been variably associated with pancreatic cancer (7–9). Experimental studies show that human and animal pancreatic tissues are susceptible to HCAs and PAHs (7). Animals exposed to HCAs are found to have higher levels of HCA-induced DNA adducts in their pancreas than any other organ. Specifically, PhIP DNA adducts have been detected more frequently in the pancreatic tissue of cancer patients than in non-cancer patients (8). PAHs and all HCAs other than DiMeIQx have been found to be carcinogenic in animals (7). In contrast to our previous study in NIH-AARP that showed significant elevated pancreatic cancer risks for DiMeIQx overall and PhIP in men (8), we do not observe any associations with these meat mutagens. It is possible that these previous associations were chance. Our present analysis has an additional 6 years of follow-up and over three times as many cases, therefore more power to observe associations if they exist. Recent literature suggests that advanced end glycation products (AGEs) may mediate the association between red meat cooked at high temperatures and pancreatic cancer in men (24). AGEs have been hypothesized to promote chronic inflammation and may be another mechanism in addition to mutagens that might explain the relationship between meat intake consumption and pancreatic cancer. Yet, biomarker studies of AGEs measured in the peripheral blood have been inconsistently associated with pancreatic cancer (24, 25).
Given the presence of heme iron in red meat and heme iron’s role in promoting oxidative stress, heme iron has also been hypothesized as a dietary risk factor for pancreatic cancer (26). Heme iron may help catalyze the formation of N-nitroso compounds (NOCs) and malondialdehyde, both of which are carcinogens and promote DNA damage and inflammation (26, 27). Studies on colorectal carcinogenesis in mouse models suggest that heme iron also promotes epithelial cell proliferation by inducing lipid oxidation (28, 29). Heme iron from meat is more efficiency absorbed in the small intestine in comparison to non-heme iron from vegetable food sources. Though prediagnostic serum iron concentrations have been linked with pancreatic cancer (30), the results of a few case-control (31–36) and prospective cohort studies (37, 38, 39) assessing dietary intake of iron have been null (26). Neither a prospective study of male health professionals nor participants men or women in the Netherlands Cohort Study observed significant association between total or heme iron intake and pancreatic cancer (37). The European Prospective Investigation into Cancer and Nutrition Cohort (EPIC) did not show an association between heme-iron and pancreatic cancer overall; however heme iron was significantly associated with greater pancreatic cancer risk in both men and women with abdominal obesity (26). Further, non-significant positive associations for heme iron intake were observed in women, and were statistically significant among women who were current smokers (26).
We observed interactions by smoking status such that in joint effects models for men and women, associations between higher total meat intake and pancreatic cancer were strongest among former and current smokers. These interactions along with those observed in the EPIC study may suggest that the effects of smoking and meat consumption on pancreatic cancer may be synergistic particularly as smoking is a source of nitrosamines, and heme iron catalyzes NOC production. However, residual confounding from smoking should also be considered as an alternative explanation for these interactions. In contrast to the EPIC findings, however, we observed significant positive associations between heme iron from red meat in men but not women. Although speculative this sex-specific difference could possibly be explained by men having higher iron status in comparison to women. Men may have higher iron stores due to greater lifetime meat consumption as well as lack of menstrual blood loss (40).
An important strength of our study is its large prospective design with detailed information about diet being assessed prior to cancer diagnosis, thereby reducing the influence of reverse causation, and differential biases such as recall and selection (8, 26). We have a relatively large number of cancer cases compared to earlier reports, as well as a wide distribution of dietary intake (11). This increases the power of our study to observe associations if they exist, as well as enables us to stratify our analyses by sex and smoking status. Our meat exposures were also based on a unique meat questionnaire and mutagen and heme iron database (36). Our study also has limitations. As with other dietary intake studies, measurement error related to both dietary assessment and the meat-related nutrients and mutagens database is likely present and could result in inaccurate risk estimates. Dietary intake and other exposures were only assessed for the year before baseline. Repeated assessments would better reflect long-term exposure, as well as changes over time. There may be residual confounding from the exposures controlled in our models. Although we controlled for confounders and known risk factors for pancreatic cancer, other exposures correlated with meat intake could potentially explain our associations. Finally, our population is primarily Caucasian and may not be generalizable to other ethnic groups.
In conclusion, we observed that higher consumption of total, red, high-temperature cooked, grilled/barbequed, and well/very well done meats and heme iron from red meat was significantly associated with increased pancreatic cancer risk. Further research is required to confirm our study findings and possibly evaluate biomarkers of iron status in relation to pancreatic cancer.
Novelty and impact of paper.
To the best of our knowledge, this is the largest prospective study to examine meat cooking methods, related mutagens, and heme iron in relation to pancreatic cancer.
Acknowledgments
Cancer incidence data from Arizona was collected by the Arizona Cancer Registry; from Georgia by the Georgia Center for Cancer Statistics; from California by the California Department of Health Services, Cancer Surveillance Section; from Michigan by the Michigan Cancer Surveillance Program; from Florida by the Florida Cancer Data System under contract to the Department of Health; from Louisiana by the Louisiana Tumor Registry; from New Jersey by the New Jersey State Cancer Registry; from North Carolina by the North Carolina Central Cancer Registry; from Pennsylvania by the Division of Health Statistics and Research, Pennsylvania Department of Health; from Texas by the Texas Cancer Registry and from Nevada by the Nevada Central Cancer Registry of the Division of Public and Behavioral Health for the State of Nevada. The views expressed herein are solely those of the authors and do not necessarily reflect those of the Cancer registries or contractors. The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.
We are indebted to the participants of the NIH-AARP Diet and Health Study for their outstanding cooperation.
This research was supported by the Intramural Research Program of the National Institutes of Health, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services.
Abbreviations used
- AARP
American Association of Retired Persons
- BaP
benzo(a)pyrene
- CI
confidence interval
- DiMeIQx
2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline
- FFQ
Food Frequency Questionnaire
- HCAs
heterocyclic amines
- HR
hazard ratio
- MeIQx
2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline
- NIH
National Institutes of Health
- PAHs
polycyclic aromatic hydrocarbons
- PhIP
2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
- RFQ
Risk Factor Questionnaire
- SAS
Statistical Analysis Systems
- WCFR/AICR
World Cancer Research Fund and American Institute for Cancer Research
Footnotes
The authors have no conflicts of interest.
References
- 1.Siegel R, Jiemin M, Zou Z, Jemal A. Cancer Statistics. CA Cancer J Clin. 2014;64:9–29. doi: 10.3322/caac.21208. [DOI] [PubMed] [Google Scholar]
- 2.Preziosi G, Oben JA, Fusai G. Obesity and pancreatic cancer. Surg Oncol. 2014;23:61–71. doi: 10.1016/j.suronc.2014.02.003. [DOI] [PubMed] [Google Scholar]
- 3.World Cancer Research Fund, American Institute for Cancer Research. Food, nutrition, physical activity, and the prevention of pancreatic cancer. Washington DC: AICR; 2012. Continuous Update Report. [Google Scholar]
- 4.Paluszkiewicz P, Smolinska K, Debinska I, Turski WA. Main dietary compounds and pancreatic cancer risk. The quantitative analysis of case-control and cohort studies. Cancer Epidemiol. 2011;36:60–7. doi: 10.1016/j.canep.2011.05.004. [DOI] [PubMed] [Google Scholar]
- 5.Larson SC, Wolk A. Red and processed meat consumption and risk of pancreatic cancer: meta-analysis of prospective studies. Br J Cancer. 2012;106:603–7. doi: 10.1038/bjc.2011.585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Anderson KE, Sinha R, Kulldorff M, Gross M, Lang NP, Barber C, Harnack L, DiMagno E, Bliss R, Kadlubar FF. Meat intake and cooking techniques: associations with pancreatic cancer. Mutat Res. 2002;506–507:225–31. doi: 10.1016/s0027-5107(02)00169-0. [DOI] [PubMed] [Google Scholar]
- 7.Li D, Day RS, Bondy ML, Sinha R, Nguyen NT, Evans DB, Abbruzzese JK, Hassan MM. Dietary mutagen exposure and risk of pancreatic cancer. Cancer Epidemiol Biomarkers Prev. 2007;16:655–61. doi: 10.1158/1055-9965.EPI-06-0993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Stolzenberg-Solomon RZ, Cross AJ, Silverman DT, Schairer C, Thompson FE, Kipnis V, Subar AF, Hollenbeck A, Schatzkin A, Sinha R. Meat and meat-mutagen intake and pancreatic cancer risk in the NIH-AARP cohort. Cancer Epidemiol Biomarkers Prev. 2007;16:2664–75. doi: 10.1158/1055-9965.EPI-07-0378. [DOI] [PubMed] [Google Scholar]
- 9.Anderson KE, Mongin S, Sinha R, Stolzenberg-Solomon R, Gross MD, Ziegler RG, Mabie JE, Risch A, Kazin SS, Church TR. Pancreatic cancer risk: Associations with meat-derived carcinogen intake in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) Cohort. Mol Carcinog. 2012;51:128–37. doi: 10.1002/mc.20794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jansen R, Robinson D, Frank R, Stolzenberg-Solomon R, Bamlet WR, Oberg AL, Rabe KG, Olson JE, Petersen GM, Sinha R, Anderson KE. Meat-related mutagens and pancreatic cancer: null results from a clinic-based case-control study. Cancer Epidemiol Biomarkers Prev. 2013;22:1336–39. doi: 10.1158/1055-9965.EPI-13-0343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schatzkin A, Subar AF, Thompson FE, Harlan LC, Tangrea J, Hollenbeck AR, Hurwitz PE, Coyle L, Schussler N, Michaud DS, Freedman LS, Brown CC, et al. Design and serendipity in establishing a large cohort with wide dietary intake distributions: the National Institutes of Health-American Association of Retired Persons Diet and Health Study. Am J Epidemiol. 2001;154:1119–25. doi: 10.1093/aje/154.12.1119. [DOI] [PubMed] [Google Scholar]
- 12.Stolzenberg-Solomon RZ, Schairer C, Moore S, Hollenbeck A, Silverman DT. Lifetime adiposity and risk of pancreatic cancer in the NIH-AARP Diet and Health Study cohort. Am J Clin Nutr. 2013;98:1057–65. doi: 10.3945/ajcn.113.058123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Michaud DS, Midthune D, Hermansen S, Leitzmann M, Harlan LC, Kipnis V, Schatzkin A. Comparison of cancer registry case ascertainment with SEER estimates and self-reporting in a subset of the NIH-AARP Diet and Health Study. J Registr Manag. 2005;32:70–5. [Google Scholar]
- 14.Subar AF, Midthune D, Kulldorff M, Brown CC, Thompson FE, Kipnis V, Schatzkin A. Evaluation of alternative approaches to assign nutrient values to food groups in food frequency questionnaires. Am J Epidemiol. 2000;152:279–86. doi: 10.1093/aje/152.3.279. [DOI] [PubMed] [Google Scholar]
- 15.Sinha R, Cross A, Curtin J, Zimmerman T, McNutt S, Risch A, Holden J. Development of a food frequency questionnaire module and databases for compounds in cooked and processed meats. Mol Nutr Food Res. 2005;49:648–55. doi: 10.1002/mnfr.200500018. [DOI] [PubMed] [Google Scholar]
- 16.Arem H, Gunter MJ, Cross AJ, Hollenbeck AR, Sinha R. A prospective investigation of fish, meat and cooking-related carcinogens with endometrial cancer incidence. Br J Cancer. 2013;109:756–60. doi: 10.1038/bjc.2013.252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ames BN, Mccann J, Yamasaki E. Methods for detecting carcinogens and mutagens with the Salmonella/mammalian-microsome mutagenicity test. Mutat Res. 1975;31:347–64. doi: 10.1016/0165-1161(75)90046-1. [DOI] [PubMed] [Google Scholar]
- 18.National Cancer Institute. CHARRED: computerized heterocyclic amines database resource for research in the epidemiologic of disease. 2006 Available from: http://dceg.cancer.gov/tools/design/charred.
- 19.Sinha R, Knize MG, Salmon CP, Brown ED, Rhodes D, Felton JS, Levander OA, Rothman N. Heterocyclic amine content of pork products cooked by different methods and to varying degrees of doneness. Food Chem Toxicol. 1998;36:289–97. doi: 10.1016/s0278-6915(97)00159-2. [DOI] [PubMed] [Google Scholar]
- 20.Sinha R, Rothman N, Salmon CP, Knize MG, Brown ED, Swanson CA, Rhodes D, Rossi S, Felton JS, Levander OA. Heterocyclic amine content in beef cooked by different methods to varying doneness and gravy made from meat drippings. Food Chem Toxicol. 1998;36:279–87. doi: 10.1016/s0278-6915(97)00162-2. [DOI] [PubMed] [Google Scholar]
- 21.Sinha R, Rothman N, Brown ED, Salmon CP, Knize MG, Swanson CA, Rossi SC, Mark SD, Levander OA, Felton JS. High concentrations of the carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) occur in chicken but are dependent on the cooking method. Cancer Res. 1995;55:4516–9. [PubMed] [Google Scholar]
- 22.Cross A, Leitzmann M, Gail M, Hollenbeck AR, Schatzkin A, Sinha R. A prospective study of red and processed meat intake in relation to cancer risk. PLoS Med. 2007;4:e325. doi: 10.1371/journal.pmed.0040325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Inoue-Choi M, Flood A, Robien K, Anderson K. Nutrients, food groups, dietary patterns, and risk of pancreatic cancer in postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2011;20:711–4. doi: 10.1158/1055-9965.EPI-11-0026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jiao L, Stolzenberg-Solomon R, Zimmerman TP, Duan Z, Chen L, Kahle L, Risch A, Subar AF, Cross AJ, Hollenbeck A, Vlassara H, Striker G, et al. Dietary consumption of advanced glycation end products and pancreatic cancer in the prospective NIH-AARP Diet and Health Study. Am J Clin Nutr. 2015;101:126–34. doi: 10.3945/ajcn.114.098061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Grote VA, Nieters A, Kaaks R, Tjonneland A, Roswall N, Overvad K, Nielsen MR, Clavel-Chapelon F, Boutron-Ruault MC, Racine A, Teucher B, Lukanova A, et al. The associations of advanced glycation end products and its soluable receptor with pancreatic cancer risk: a case-control study within the prospective EPIC Cohort. Cancer Epidemiol Biomarkers Prev. 2012;21:619–28. doi: 10.1158/1055-9965.EPI-11-1139. [DOI] [PubMed] [Google Scholar]
- 26.Molina-Montes E, Wark PA, Sanchez MJ, Norat T, Jakszyn P, Lujan-Barroso L, Michaud DS, Crowe F, Allen N, Khaw KT, Wareham N, Trichopoulou A, et al. Dietary intake of iron, heme-iron and magnesium and pancreatic cancer risk in the European prospective investigation into cancer and nutrition cohort. Int J Cancer. 2012;131:1134–47. doi: 10.1002/ijc.27547. [DOI] [PubMed] [Google Scholar]
- 27.Risch H. Etiology of pancreatic cancer, with a hypothesis concerning the role of N-nitroso compounds and excess gastric acidity. J Natl Cancer Inst. 2003;95:948–60. doi: 10.1093/jnci/95.13.948. [DOI] [PubMed] [Google Scholar]
- 28.Oostindjer M, Alexander J, Amdam GV, Anderson G, Bryan NS, Chen D, Corpet DE, De Smet S, Ove Dragsted L, Huag A, Karlsson AH, Kleter G, et al. The role of red and processed meat in colorectal cancer development: a perspective. Meat Science. 2014;97:583–96. doi: 10.1016/j.meatsci.2014.02.011. [DOI] [PubMed] [Google Scholar]
- 29.Ijssennagger N, Rijnierse A, de Wit NJ, Boekschoten MV, Dekker J, Schonewille A, Muller M, van der Meer R. Dietary heme induces acute oxidative stress, but delayed cytotoxicity and compensatory hyperproliferation in mouse colon. Carcinogenesis. 2013;34:1628–35. doi: 10.1093/carcin/bgt084. [DOI] [PubMed] [Google Scholar]
- 30.Friedman GD, van den Eeden SK. Risk factors for pancreatic cancer: an exploratory study. Int J Epidemiol. 1993;22:30–37. doi: 10.1093/ije/22.1.30. [DOI] [PubMed] [Google Scholar]
- 31.Baghurst PA, McMichael AJ, Slavotinek AH, Baghurst KI, Boyle P, Walker AM. A case-control study of diet and cancer of the pancreas. Am J Epidemiol. 1991;134:167–79. doi: 10.1093/oxfordjournals.aje.a116069. [DOI] [PubMed] [Google Scholar]
- 32.Lin Y, Tamakoshi A, Hayakawa T, Naruse S, Kitagawa M, Ohno Y. Nutritional factors and risk of pancreatic cancer: a population-based case-control study based on direct interview in Japan. J Gastroenterol. 2005;40:297–301. doi: 10.1007/s00535-004-1537-0. [DOI] [PubMed] [Google Scholar]
- 33.Olsen GW, Mandel JS, Gibson RW, Wattenberg LW, Schuman LM. Nutrients and pancreatic cancer: a population-based case-control study. Cancer Cause Control. 1991;2:291–7. doi: 10.1007/BF00051668. [DOI] [PubMed] [Google Scholar]
- 34.Silverman DT, Swanson CA, Gridley G, Wacholder S, Greenberg RS, Brown LM, Hayes RB, Swanson GM, Schoeberg JB, Pottern LM, Schwartz AG, Fraumeni JF, Jr, et al. Dietary and nutritional factors and pancreatic cancer: a case-control study based on direct interviews. J Natl Cancer Inst. 1998;90:1710–9. doi: 10.1093/jnci/90.22.1710. [DOI] [PubMed] [Google Scholar]
- 35.Ghadirian P, Simard A, Baillargeon J, Maisonneuve P, Boyle P. Nutritional factors and pancreatic cancer in the francophone community in Montreal, Canada. Int J Cancer. 1991;47:1–6. doi: 10.1002/ijc.2910470102. [DOI] [PubMed] [Google Scholar]
- 36.Bravi F, Polesel J, Bosetti C, Talamini R, Negri E, Dal Maso L, Serraino D, La Vecchia C. Dietary intake of selected micronutrients and the risk of pancreatic cancer: an Italian case-control study. Ann Oncol. 2011;22:202–6. doi: 10.1093/annonc/mdq302. [DOI] [PubMed] [Google Scholar]
- 37.Kesavan Y, Giovannucci E, Fuchs S, Michaud DS. A prospective study of magnesium and iron intake and pancreatic cancer in men. Am J Epidemiol. 2010;171:233–41. doi: 10.1093/aje/kwp373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Michaud DS, Giovannucci E, Willett WC, Colditz GA, Fuchs CS. Dietary meat, dairy products, fat, and cholesterol and pancreatic cancer risk in a prospective study. Am J Epidemiol. 2003;157:1115–25. doi: 10.1093/aje/kwg098. [DOI] [PubMed] [Google Scholar]
- 39.Heinen MM, Verhage BA, Goldbohm RA, van den Brandt PA. Meat and fat intake and pancreatic cancer risk in the Netherlands Cohort Study. Int J Cancer. 2009;125:1118–26. doi: 10.1002/ijc.24387. [DOI] [PubMed] [Google Scholar]
- 40.Romeu M, Aranda N, Giralt M, Ribot B, Nogues MR, Arija V. Diet, iron biomarkers and oxidative stress in a representative sample of Mediterranean population. Nutr J. 2013;12:102–111. doi: 10.1186/1475-2891-12-102. [DOI] [PMC free article] [PubMed] [Google Scholar]