Background:

Several studies have investigated the association between alcohol consumption and pancreatic cancer risk. However, these studies reported inconsistent results.

Methods:

This study included 95,812 participants from the Japan Public Health Center-Based Prospective Study, whose alcohol drinking habit was collected through the questionnaire of the baseline survey in 1990 to 1994 for Cohort I and in 1993 to 1995 for Cohort II, and followed-up until December 2013. The hazard ratios (HR) and 95% confidence intervals (CI) were calculated using multivariable-adjusted Cox proportional hazard regression models.

Results:

During 1,969,101 person-years of follow-up, 598 patients (315 men and 283 women) were newly diagnosed with pancreatic cancer. No association was observed between alcohol consumption at baseline and pancreatic cancer risk in either men or women, even in the analyses conducted among men stratified by facial flushing responses or smoking status. When the population was limited to men whose alcohol drinking habit remained unchanged from the baseline survey to the 5-year follow-up survey, a significant association was observed in drinkers with alcohol consumption of 1 to 299 g/week compared with non/occasional drinkers (multivariable-adjusted HR, 1.73; 95% CI, 1.15–2.60).

Conclusions:

A significant association between alcohol consumption and pancreatic cancer risk was observed when the population was limited to men with a 5-year unchanged alcohol drinking habit and was particularly strong in never smokers.

Impact:

This study suggested an association between continuous alcohol consumption and pancreatic cancer risk in the Japanese population. However, further investigations using other cohort studies are required.

Pancreatic cancer is the fourth leading cause of cancer-related deaths in Japan and the United States, and its incidence continues to increase particularly in countries with higher socioeconomic status (1–3). Pancreatic cancer is an aggressive tumor type with a poor prognosis, mainly because it is usually diagnosed at an advanced stage owing to the lack of effective screening, and the available treatments have limited effectiveness. In Japan, the 5-year survival rate of patients with pancreatic cancer was only 8.5% among those who were diagnosed between 2009 and 2011 (4, 5).

Several specific risk factors for pancreatic cancer have been reported in previous epidemiologic studies, including smoking status, obesity, history of diabetes mellitus, family history of pancreatic cancer, and history of chronic pancreatitis (6–8). However, the association between alcohol consumption and risk of pancreatic cancer remains unclear. Although several previous studies have suggested an increased risk of pancreatic cancer associated with heavy drinking, no association was observed in some studies (9). Thus, previous studies have reported inconsistent results.

Alcohol dehydrogenase (ADH1B) and aldehyde dehydrogenase 2 (ALDH2) are key enzymes in the conversion of ethanol to acetaldehyde and acetaldehyde to acetic acid, respectively. Active ADH1B and inactive ALDH2 are associated with increased accumulation of acetaldehyde, which is attributed to the facial flushing response to alcohol consumption (10, 11). Previous studies have suggested that acetaldehyde is a risk factor for pancreatic cancer development (12).

We conducted a population-based cohort study in Japan to investigate the association between alcohol consumption and risk of pancreatic cancer. A previous study conducted using data from the same cohort analyzed the risk factors for pancreatic cancer and suggested that pancreatic cancer risk had no association with alcohol consumption (13). We used the same cohort study, but our study had a follow-up period which was 10-year longer than that of the previous study; moreover, the categories for the amount of alcohol consumed in the analysis was finely divided compared with that in the previous study to investigate the association between pancreatic cancer risk and the amount of alcohol consumed. In addition, we aimed to analyze the association between alcohol intake and the risk of pancreatic cancer by including information on the presence or absence of facial flushing response.

Study population

The Japan Public Health Center-Based Prospective Study (JPHC study) is a nationwide, population-based cohort study that assessed the risk factors for lifestyle-related diseases in the Japanese population. This cohort study consisted of two cohorts, and the participants were registered Japanese inhabitants in 11 public health center (PHC) areas. The study design has been reported in detail elsewhere (14).

Cohort I (participants ages 40–59 years) and Cohort II (participants ages 40–69 years) enrolled 140,420 participants during 1990 to 1994 and 1993 to 1995, respectively. In both cohorts, 113,461 participants responded to the baseline survey. Participants were excluded for the following reasons: non-Japanese nationality (n = 19), late report of relocation out of the study area prior to the start of follow-up (n = 37), incorrect date of birth (n = 3), rejection for follow-up (n = 26), and duplicate registration (n = 2). All participants from the Katsushika PHC area (n = 7,077) in Cohort I were excluded as the incidence data were not available, whereas those from the Suita PHC area (n = 6,540) in Cohort II were excluded as the definition of the study population was different from that of others. Participants who were diagnosed with cancer at baseline (n = 2,135), who had missing information on drinking frequency (n = 1,605), or who had missing information on the date of follow-up completion (n = 205) were also excluded. Consequently, a population-based cohort of 95,812 participants was analyzed in this study.

Exposure assessment and variables

Alcohol drinking habits were assessed at baseline using a questionnaire. In Cohort I, the questionnaire comprised questions related to the frequency of alcohol consumption, type of beverage, and average volume of daily consumption if the participants drank alcoholic beverages for at least 1 day/week. In Cohort II, the questionnaire contained questions about alcohol drinking status (never, former, or current), frequency and type of beverages, and average volume of daily consumption for participants who were former and current alcohol drinkers.

The amount of alcohol consumed was calculated in grams of pure ethanol consumed per week. The participants were divided into six groups based on the average frequency and volume of alcohol consumed per week: “nondrinkers” (consumption of <1 day/month in Cohorts I and II or former drinkers in Cohort II), “occasional drinkers” (1–3 days/month), and four categories of “regular drinkers” (“1–149,” “150–299,” “300–449,” and “≥450” g of ethanol per week) to investigate the association between pancreatic cancer risk and the amount of alcohol consumed at baseline. When the analysis was stratified by facial flushing responses or smoking status, the categories “300–449” and “≥450” g of ethanol per week were combined into the category of “≥300” g of ethanol per week. The participants were categorized as “non/occasional drinkers,” “1–299 g of ethanol per week,” and “≥300 g of ethanol per week” to investigate the association of pancreatic cancer with the alcohol consumption habit unchanged from the baseline survey to the 5- or 10-year follow-up. The amount of ethanol was calculated in grams of ethanol: 180 mL of rice wine as 23 g of ethanol, 180 mL of shochu and awamori (white spirits) as 36 g of ethanol, 30 mL of whiskey or brandy as 10 g of ethanol, 60 mL of wine as 6 g of ethanol, and 633 mL of beer as 23 g of ethanol, as described in previous studies (15, 16).

Smoking status was classified as “never,” “former,” or “current smokers.” In Cohort II, participants who answered questions related to smoking cessation were regarded as former smokers. The cumulative exposure to cigarette smoking among former and current smokers was assessed in pack-years (PY), which was calculated as the number of packs of cigarettes smoked per day multiplied by the number of years of smoking. PY was categorized into 0, 1 to 19, 20 to 29, 30 to 39, or ≥40.

Facial flushing responses were divided into three categories: “positive,” “negative,” and “unknown.” A response of “occasionally” in Cohort I was regarded as “positive,” based on the previous report of the JPHC study (17).

The frequency of alcohol consumption was classified into four categories: “non/occasional drinkers,” “1–2,” “3–4,” and “≥5” days/week. Binge drinking was assessed using the questionnaire in Cohort II, which comprised questions related to the frequency of alcohol drinking for the purpose of maintaining a good relationship and the amount of alcohol consumed at the time, as Cohort I had no related data. The amount of binge drinking was calculated using the method adopted to measure the total ethanol consumption and was classified into four categories: “<70,” “70–<140,” “140–<210,” and “≥210” g/time. The frequency of binge drinking was classified into four categories: “no binge drinking,” “1,” “2–3,” and “≥4” times/month.

Follow-up

The participants were followed up from the baseline survey until December 31, 2013, using the data on the residential status and survival collected from the residential registers of each study area, or through the municipal office of the area to which they moved out.

The incidence of cancer was identified mainly using two sources: active patient notification from major local hospitals and population-based cancer registries. Information from death certificates were used as supplementary information. In accordance with the International Classification of Diseases for Oncology, third edition (18), pancreatic cancer cases were identified using the codes C25.0 to C25.9; however, participants with endocrine tumors (C25.4) were excluded due to the difference in etiology.

Statistical analysis

The number of person-years of follow-up was calculated from the date of the baseline survey until the date of cancer diagnosis, relocation outside of the study area, loss to follow-up, death, or the end of the study period (December 31, 2013), whichever occurred first.

The HRs and 95% confidence intervals (CI) calculated using multivariable-adjusted Cox proportional hazards regression models were estimated with adjustment for potential confounders to measure the association between alcohol consumption and pancreatic cancer risk. The HRs for alcohol consumption were calculated using two models. One model (HRa) was adjusted for age at baseline (continuous variable) and study area (10 PHC areas), whereas the other model (HRb) was further adjusted for history of diabetes (no, yes, or unknown), body mass index (14–<18.5, 18.5–<25, 25–<30, 30–<40 kg/m2, or missing), family history of cancer (no or yes), coffee intake (≥1 and <1 cup/week), cumulative exposure to cigarette smoking (PY category), facial flushing response (negative, positive, or unknown), and participation in sports (≥1 and <1 time/week). The proportionality-hazards assumption was statistically tested on the basis of the Schoenfeld residuals for each population analyzed.

Tests for identifying the linear trends in the risk of pancreatic cancer were performed by assigning ordinal values to each alcohol consumption category as continuous variables in each model. Analysis was also performed with stratification by facial flushing responses or smoking status to evaluate the effect modification and investigate the confounding factors. All statistical analyses were performed using the STATA statistical software, version 16 (StataCorp LP). All P values were two-sided, and a P value of <0.05 was considered significant.

Data availability

For information on how to submit an application for gaining access to JPHC data and/or biospecimens, please follow the instructions at http://epi.ncc.go.jp/en/jphc/805/8155.html.

A total of 95,812 participants (45,577 men and 50,235 women) were included in the analysis and followed-up for 1,969,101 person-years. During the follow-up period, 598 patients (315 men and 283 women) were newly diagnosed with pancreatic cancer. The baseline characteristics of the participants according to the alcohol drinking category are shown in Table 1. At baseline, 23.1%, 9.2%, and 67.7% of the men were nondrinkers, occasional drinkers, and regular drinkers, respectively. The proportion of nondrinkers among women (78.8%) was much higher than that among men. The proportion of never smokers was high in non/occasional drinkers, whereas that of current smokers was high in the group with higher alcohol consumption in both men and women. The proportion of male participants with a positive facial flush response tended to decrease as they consumed more alcohol, but the trend was different in women.

Table 1.

Baseline characteristics of the study participants by alcohol drinking category.

 Alcohol drinking
NondrinkerOccasional drinkerRegular drinker
<1 day/month1–3 days/month1–149 g/week150–299 g/week300–449 g/week≥450 g/week
Men (n = 45,577) 
Number of subjects, n (%) 10,518 (23.1) 4,190 (9.2) 10,367 (22.7) 9,121 (20.0) 5,979 (13.1) 5,402 (11.9) 
Age (years, mean) ± SD 53.9 ± 8.4 50.3 ± 7.4 51.5 ± 8.0 51.8 ± 7.9 51.4 ± 7.4 50.4 ± 7.0 
Weekly ethanol consumption (g), mean ± SD — — 81.5 ± 45.8 227.6 ± 45.6 371.9 ± 45.2 713.7 ± 480.0 
History of DM, % 
 No 61.5 66.1 58.4 63.3 67.5 69.7 
 Yes 7.8 6.1 5.9 5.9 5.9 7.4 
 Unknown 30.7 27.8 35.7 30.7 26.6 23.0 
BMI (kg/m2), %             
 14–<18.5 4.1 1.6 2.5 2.3 2.4 2.1 
 18.5–<25 67.7 63.4 70.2 70.1 69.9 68.6 
 25–<30 24.6 30.4 24.8 25.0 24.8 25.9 
 30–<40 2.2 3.6 1.6 1.6 2.1 2.5 
 Missing 1.5 1.1 0.9 0.9 0.8 0.9 
Family history of cancer, % 
 No 79.7 81.1 78.6 77.8 78.1 79.5 
 Yes 20.3 18.9 21.4 22.2 21.9 20.5 
Coffee intake (≥1 cup/week), % 
 No 58.5 56.5 56.5 62.7 66.0 66.8 
 Yes 40.3 42.5 42.6 36.4 33.0 32.2 
 Unknown 1.2 1.1 0.9 0.9 0.9 1.0 
Smoking, % 
 Never 27.8 32.5 28.6 19.8 16.3 15.7 
 Former 24.7 20.4 25.3 24.5 22.8 21.1 
 Current 47.0 46.7 45.7 55.3 60.3 62.7 
 Unknown 0.5 0.5 0.4 0.4 0.7 0.5 
Facial flushing response, % 
 Negative 14.6 30.8 41.1 51.1 53.6 55.3 
 Positive 65.6 63.4 52.8 43.0 40.2 38.4 
 Unknown 19.9 5.8 6.1 6.0 6.2 6.3 
Sports (≥1 time/week), % 
 No 81.7 80.0 77.2 79.8 80.3 83.3 
 Yes 17.0 19.3 21.9 19.4 18.8 15.8 
 Unknown 1.3 0.8 0.9 0.9 0.9 0.9 
Women (n = 50,235) 
Number of subjects, n (%) 39,605 (78.8) 4,848 (9.7) 4,589 (9.1) 728 (1.4) 226 (4.5) 239 (4.8) 
Age (years, mean) ± SD 53.1 ± 8.1 48.9 ± 6.9 50.1 ± 7.5 49.3 ± 7.2 48.7 ± 6.9 48.0 ± 6.4 
Weekly ethanol consumption (g), mean ± SD — — 52.3 ± 39.4 209.7 ± 44.9 371.0 ± 46.2 845.4 ± 515.3 
History of DM, % 
 No 60.0 65.3 57.6 61.8 65.0 72.0 
 Yes 3.3 2.0 2.1 2.1 2.7 3.3 
 Unknown 36.7 32.7 40.3 36.1 32.3 24.7 
BMI (kg/m2), % 
 14–<18.5 3.6 2.5 3.7 4.4 4.0 4.6 
 18.5–<25 66.2 69.7 73.3 69.0 68.1 64.4 
 25–<30 25.7 24.1 20.5 23.4 22.1 26.4 
 30–<40 3.4 2.9 2.0 2.2 4.0 2.9 
 Missing 1.2 0.6 0.6 1.1 1.8 1.7 
Family history of cancer, % 
 No 79.7 76.2 73.7 76.1 80.1 75.7 
 Yes 20.3 23.8 26.3 23.9 19.9 24.3 
Coffee intake (≥1 cup/week), % 
 No 64.8 54.0 49.9 53.8 61.9 66.9 
 Yes 34.2 45.2 49.5 44.9 37.6 31.4 
 Unknown 1.0 0.8 0.6 1.2 0.4 1.7 
Smoking, %             
 Never 94.0 89.5 84.6 61.0 52.2 48.1 
 Former 1.3 2.4 3.1 6.2 7.5 4.2 
 Current 4.2 7.8 12.0 32.6 39.8 47.7 
 Unknown 0.4 0.3 0.4 0.3 0.4 0.0 
Facial flushing response, % 
 Negative 16.7 48.0 57.2 62.2 60.6 54.0 
 Positive 34.8 40.6 34.5 32.1 32.3 40.2 
 Unknown 48.5 11.4 8.3 5.6 7.1 5.9 
Sports (≥1 time/week), % 
 No 81.9 80.5 77.4 80.2 84.1 84.9 
 Yes 16.7 18.6 21.7 18.5 15.5 14.2 
 Unknown 1.4 0.9 1.0 1.2 0.4 0.8 
 Alcohol drinking
NondrinkerOccasional drinkerRegular drinker
<1 day/month1–3 days/month1–149 g/week150–299 g/week300–449 g/week≥450 g/week
Men (n = 45,577) 
Number of subjects, n (%) 10,518 (23.1) 4,190 (9.2) 10,367 (22.7) 9,121 (20.0) 5,979 (13.1) 5,402 (11.9) 
Age (years, mean) ± SD 53.9 ± 8.4 50.3 ± 7.4 51.5 ± 8.0 51.8 ± 7.9 51.4 ± 7.4 50.4 ± 7.0 
Weekly ethanol consumption (g), mean ± SD — — 81.5 ± 45.8 227.6 ± 45.6 371.9 ± 45.2 713.7 ± 480.0 
History of DM, % 
 No 61.5 66.1 58.4 63.3 67.5 69.7 
 Yes 7.8 6.1 5.9 5.9 5.9 7.4 
 Unknown 30.7 27.8 35.7 30.7 26.6 23.0 
BMI (kg/m2), %             
 14–<18.5 4.1 1.6 2.5 2.3 2.4 2.1 
 18.5–<25 67.7 63.4 70.2 70.1 69.9 68.6 
 25–<30 24.6 30.4 24.8 25.0 24.8 25.9 
 30–<40 2.2 3.6 1.6 1.6 2.1 2.5 
 Missing 1.5 1.1 0.9 0.9 0.8 0.9 
Family history of cancer, % 
 No 79.7 81.1 78.6 77.8 78.1 79.5 
 Yes 20.3 18.9 21.4 22.2 21.9 20.5 
Coffee intake (≥1 cup/week), % 
 No 58.5 56.5 56.5 62.7 66.0 66.8 
 Yes 40.3 42.5 42.6 36.4 33.0 32.2 
 Unknown 1.2 1.1 0.9 0.9 0.9 1.0 
Smoking, % 
 Never 27.8 32.5 28.6 19.8 16.3 15.7 
 Former 24.7 20.4 25.3 24.5 22.8 21.1 
 Current 47.0 46.7 45.7 55.3 60.3 62.7 
 Unknown 0.5 0.5 0.4 0.4 0.7 0.5 
Facial flushing response, % 
 Negative 14.6 30.8 41.1 51.1 53.6 55.3 
 Positive 65.6 63.4 52.8 43.0 40.2 38.4 
 Unknown 19.9 5.8 6.1 6.0 6.2 6.3 
Sports (≥1 time/week), % 
 No 81.7 80.0 77.2 79.8 80.3 83.3 
 Yes 17.0 19.3 21.9 19.4 18.8 15.8 
 Unknown 1.3 0.8 0.9 0.9 0.9 0.9 
Women (n = 50,235) 
Number of subjects, n (%) 39,605 (78.8) 4,848 (9.7) 4,589 (9.1) 728 (1.4) 226 (4.5) 239 (4.8) 
Age (years, mean) ± SD 53.1 ± 8.1 48.9 ± 6.9 50.1 ± 7.5 49.3 ± 7.2 48.7 ± 6.9 48.0 ± 6.4 
Weekly ethanol consumption (g), mean ± SD — — 52.3 ± 39.4 209.7 ± 44.9 371.0 ± 46.2 845.4 ± 515.3 
History of DM, % 
 No 60.0 65.3 57.6 61.8 65.0 72.0 
 Yes 3.3 2.0 2.1 2.1 2.7 3.3 
 Unknown 36.7 32.7 40.3 36.1 32.3 24.7 
BMI (kg/m2), % 
 14–<18.5 3.6 2.5 3.7 4.4 4.0 4.6 
 18.5–<25 66.2 69.7 73.3 69.0 68.1 64.4 
 25–<30 25.7 24.1 20.5 23.4 22.1 26.4 
 30–<40 3.4 2.9 2.0 2.2 4.0 2.9 
 Missing 1.2 0.6 0.6 1.1 1.8 1.7 
Family history of cancer, % 
 No 79.7 76.2 73.7 76.1 80.1 75.7 
 Yes 20.3 23.8 26.3 23.9 19.9 24.3 
Coffee intake (≥1 cup/week), % 
 No 64.8 54.0 49.9 53.8 61.9 66.9 
 Yes 34.2 45.2 49.5 44.9 37.6 31.4 
 Unknown 1.0 0.8 0.6 1.2 0.4 1.7 
Smoking, %             
 Never 94.0 89.5 84.6 61.0 52.2 48.1 
 Former 1.3 2.4 3.1 6.2 7.5 4.2 
 Current 4.2 7.8 12.0 32.6 39.8 47.7 
 Unknown 0.4 0.3 0.4 0.3 0.4 0.0 
Facial flushing response, % 
 Negative 16.7 48.0 57.2 62.2 60.6 54.0 
 Positive 34.8 40.6 34.5 32.1 32.3 40.2 
 Unknown 48.5 11.4 8.3 5.6 7.1 5.9 
Sports (≥1 time/week), % 
 No 81.9 80.5 77.4 80.2 84.1 84.9 
 Yes 16.7 18.6 21.7 18.5 15.5 14.2 
 Unknown 1.4 0.9 1.0 1.2 0.4 0.8 

Abbreviation: DM, diabetes mellitus.

The association between alcohol consumption and risk of pancreatic cancer is shown in Table 2. The proportions of alcohol drinkers (occasional and regular) among patients with pancreatic cancer were 77.1% in men and 17.3% in women. The multivariable-adjusted HR for regular drinkers with alcohol consumption of ≥450 g/week was 0.86 (95% CI, 0.57–1.30; Ptrend = 0.678) compared with that in nondrinkers in the total population. No significant association was observed between alcohol consumption and pancreatic cancer risk in either men or women.

Table 2.

Association between alcohol drinking and pancreatic cancer risk.

 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Men 
Nondrinker 72 208,320 Ref Ref 
Occasional drinker 26 87,971 1.09 (0.69–1.71) 1.16 (0.73–1.84) 
1–149 g/week 76 208,419 1.20 (0.87–1.66) 1.27 (0.91–1.78) 
150–299 g/week 67 182,830 1.17 (0.84–1.65) 1.19 (0.84–1.70) 
300–449 g/week 43 119,784 1.16 (0.79–1.71) 1.15 (0.77–1.72) 
≥450 g/week 31 109,300 0.97 (0.63–1.50) 0.94 (0.60–1.47) 
Ptrend     0.702 0.897 
Women 
Nondrinker 234 827,323 Ref Ref 
Occasional drinker 30 103,555 1.38 (0.94–2.04) 1.30 (0.87–1.95) 
Regular drinker 19 121,599 0.71 (0.44–1.14) 0.61 (0.37–1.00) 
Ptrend     0.463 0.163 
 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Men 
Nondrinker 72 208,320 Ref Ref 
Occasional drinker 26 87,971 1.09 (0.69–1.71) 1.16 (0.73–1.84) 
1–149 g/week 76 208,419 1.20 (0.87–1.66) 1.27 (0.91–1.78) 
150–299 g/week 67 182,830 1.17 (0.84–1.65) 1.19 (0.84–1.70) 
300–449 g/week 43 119,784 1.16 (0.79–1.71) 1.15 (0.77–1.72) 
≥450 g/week 31 109,300 0.97 (0.63–1.50) 0.94 (0.60–1.47) 
Ptrend     0.702 0.897 
Women 
Nondrinker 234 827,323 Ref Ref 
Occasional drinker 30 103,555 1.38 (0.94–2.04) 1.30 (0.87–1.95) 
Regular drinker 19 121,599 0.71 (0.44–1.14) 0.61 (0.37–1.00) 
Ptrend     0.463 0.163 

Abbreviation: Ref, reference.

aAdjusted for age (continuous), and area.

bAdjusted for age (continuous), area, history of diabetes, BMI, family history of cancer, coffee intake, smoking (pack-year), facial flushing, and status of sports.

The HRs and 95% CIs of the incidence of pancreatic cancer associated with alcohol consumption stratified by facial flushing response among men are shown in Table 3. No association was observed between alcohol consumption and pancreatic cancer risk in patients with and without a facial flushing response. Among men without a facial flushing response, occasional drinkers showed the highest HRs in the further adjusted model compared with nondrinkers (HR, 2.35; 95% CI, 0.88–6.26). Table 4 shows the HRs and 95% CIs of the incidence of pancreatic cancer associated with alcohol consumption stratified by smoking status among men. No association was observed between alcohol consumption and pancreatic cancer risk for any smoking status.

Table 3.

Association between alcohol drinking and pancreatic cancer risk among men by alcohol facial flushing response.

 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Positive 
Nondrinker 52 136,461 Ref Ref 
Occasional drinker 15 55,876 0.85 (0.47–1.52) 0.88 (0.49–1.58) 
1–149 g/week 41 110,046 1.08 (0.71–1.63) 1.13 (0.75–1.71) 
150–299 g/week 21 78,811 0.74 (0.44–1.23) 0.76 (0.45–1.27) 
≥300 g/week 32 90,619 1.05 (0.67–1.66) 1.04 (0.66–1.65) 
Ptrend     0.897 0.910 
Negative 
Nondrinker 30,124 Ref Ref 
Occasional drinker 10 27,141 2.22 (0.84–5.90) 2.35 (0.88–6.26) 
1–149 g/week 24 86,013 1.44 (0.62–3.37) 1.49 (0.64–3.48) 
150–299 g/week 38 93,370 2.06 (0.91–4.65) 1.98 (0.87–4.48) 
≥300 g/week 40 124,596 1.64 (0.72–3.72) 1.50 (0.65–3.42) 
Ptrend     0.384 0.670 
Pinteraction     0.159 0.168 
 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Positive 
Nondrinker 52 136,461 Ref Ref 
Occasional drinker 15 55,876 0.85 (0.47–1.52) 0.88 (0.49–1.58) 
1–149 g/week 41 110,046 1.08 (0.71–1.63) 1.13 (0.75–1.71) 
150–299 g/week 21 78,811 0.74 (0.44–1.23) 0.76 (0.45–1.27) 
≥300 g/week 32 90,619 1.05 (0.67–1.66) 1.04 (0.66–1.65) 
Ptrend     0.897 0.910 
Negative 
Nondrinker 30,124 Ref Ref 
Occasional drinker 10 27,141 2.22 (0.84–5.90) 2.35 (0.88–6.26) 
1–149 g/week 24 86,013 1.44 (0.62–3.37) 1.49 (0.64–3.48) 
150–299 g/week 38 93,370 2.06 (0.91–4.65) 1.98 (0.87–4.48) 
≥300 g/week 40 124,596 1.64 (0.72–3.72) 1.50 (0.65–3.42) 
Ptrend     0.384 0.670 
Pinteraction     0.159 0.168 

Abbreviation: Ref, reference.

aAdjusted for age (continuous) and area.

bAdjusted for age (continuous), area, history of diabetes, BMI, family history of cancer, coffee intake, smoking (pack-year), and status of sports.

Table 4.

Association between alcohol drinking and pancreatic cancer risk among men by smoking status.

 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Never smoker 
Nondrinker 16 59,827 Ref Ref 
Occasional drinker 28,874 1.00 (0.39–2.60) 1.04 (0.40–2.71) 
1–149 g/week 18 61,456 1.27 (0.64–2.51) 1.31 (0.66–2.60) 
150–299 g/week 12 37,429 1.36 (0.64–2.92) 1.40 (0.65–3.00) 
≥300 g/week 37,661 1.00 (0.43–2.32) 1.02 (0.44–2.36) 
Ptrend   0.686 0.657 
Former smoker 
Nondrinker 18 50,033 Ref Ref 
Occasional drinker 17,771 1.21 (0.48–3.10) 1.14 (0.44–2.91) 
1–149 g/week 17 52,161 1.03 (0.53–2.02) 1.00 (0.51–1.96) 
150–299 g/week 14 44,731 0.96 (0.47–1.96) 0.93 (0.46–1.89) 
≥300 g/week 16 50,499 1.03 (0.51–2.07) 0.99 (0.49–2.00) 
Ptrend     0.941 0.876 
Current smoker 
Nondrinker 37 97,499 Ref  Ref 
Occasional drinker 14 40,925 1.10 (0.59–2.05) 1.12 (0.60–2.09) 
1–149 g/week 41 93,951 1.30 (0.83–2.03) 1.34 (0.86–2.11) 
150–299 g/week 39 99,917 1.14 (0.72–1.79) 1.13 (0.72–1.78) 
≥300 g/week 48 139,720 1.03 (0.66–1.60) 0.98 (0.63–1.53) 
Ptrend     0.914 0.904 
Pinteraction     0.997 0.996 
 Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Never smoker 
Nondrinker 16 59,827 Ref Ref 
Occasional drinker 28,874 1.00 (0.39–2.60) 1.04 (0.40–2.71) 
1–149 g/week 18 61,456 1.27 (0.64–2.51) 1.31 (0.66–2.60) 
150–299 g/week 12 37,429 1.36 (0.64–2.92) 1.40 (0.65–3.00) 
≥300 g/week 37,661 1.00 (0.43–2.32) 1.02 (0.44–2.36) 
Ptrend   0.686 0.657 
Former smoker 
Nondrinker 18 50,033 Ref Ref 
Occasional drinker 17,771 1.21 (0.48–3.10) 1.14 (0.44–2.91) 
1–149 g/week 17 52,161 1.03 (0.53–2.02) 1.00 (0.51–1.96) 
150–299 g/week 14 44,731 0.96 (0.47–1.96) 0.93 (0.46–1.89) 
≥300 g/week 16 50,499 1.03 (0.51–2.07) 0.99 (0.49–2.00) 
Ptrend     0.941 0.876 
Current smoker 
Nondrinker 37 97,499 Ref  Ref 
Occasional drinker 14 40,925 1.10 (0.59–2.05) 1.12 (0.60–2.09) 
1–149 g/week 41 93,951 1.30 (0.83–2.03) 1.34 (0.86–2.11) 
150–299 g/week 39 99,917 1.14 (0.72–1.79) 1.13 (0.72–1.78) 
≥300 g/week 48 139,720 1.03 (0.66–1.60) 0.98 (0.63–1.53) 
Ptrend     0.914 0.904 
Pinteraction     0.997 0.996 

Abbreviation: Ref, reference.

aAdjusted for age (continuous) and area.

bAdjusted for age (continuous), area, history of diabetes, BMI, family history of cancer, coffee intake, smoking (pack-year), facial flushing, and status of sports.

The HRs and 95% CIs of the incidence of pancreatic cancer associated with the frequency of alcohol consumption in men and women and with binge drinking among men are shown in Table 5. No significant associations were observed.

Table 5.

Association between alcohol consumption patterns and pancreatic cancer risk.

 Number of subjects, n (%)Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Association with frequency of alcohol drinking 
Men 
 Non/occasional drinkers 10,518 (23.1) 72 208,320 Ref Ref 
 1–2/week 8,346 (18.3) 54 173,883 1.15 (0.80–1.65) 1.24 (0.86–1.79) 
 3–4/week 5,317 (11.7) 31 108,266 1.01 (0.66–1.54) 1.06 (0.69–1.64) 
 ≥5/week 21,396 (46.9) 158 426,155 1.17 (0.88–1.56) 1.18 (0.87–1.60) 
Ptrend       0.341 0.405 
Women 
 Non/occasional drinkers 39,605 (78.8) 234 827,323 Ref Ref 
 1–2/week 7,100 (14.1) 37 151,141 1.15 (0.81–1.64) 1.07 (0.74–1.56) 
 3–4/week 1,586 (3.2) 33,272 0.70 (0.29–1.70) 0.61 (0.25–1.51) 
 ≥5/week 1,944 (3.9) 40,741 0.76 (0.36–1.61) 0.62 (0.29–1.35) 
Ptrend       0.563 0.217 
Pinteraction       0.478 0.503 
Association with binge drinking among men 
Amount of binge drinking (g/time) 
 <70 9,164 (68.3) 71 169,984 Ref Ref 
 70–<140 2,136 (15.9) 13 41,089 0.97 (0.53–1.79) 1.06 (0.58–1.95) 
 140–<210 668 (5.0) 13,112 0.56 (0.13–2.33) 0.64 (0.15–2.66) 
 ≥210 1,452 (10.8) 27,933 1.01 (0.47–2.17) 1.09 (0.50–2.37) 
Ptrend       0.808 0.953 
Frequency of binge drinking (per month) 
 No binge drinking 2,199 (8.9) 12 22,824 Ref Ref 
 1 4,492 (20.0) 29 73,400 0.88 (0.45–1.75) 0.92 (0.46–1.82) 
 2–3 6,510 (26.3) 44 107,304 0.96 (0.50–1.85) 1.03 (0.53–1.98) 
 ≥4 11,068 (44.8) 45 154,745 0.70 (0.36–1.34) 0.71 (0.37–1.38) 
Ptrend       0.200 0.204 
 Number of subjects, n (%)Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Association with frequency of alcohol drinking 
Men 
 Non/occasional drinkers 10,518 (23.1) 72 208,320 Ref Ref 
 1–2/week 8,346 (18.3) 54 173,883 1.15 (0.80–1.65) 1.24 (0.86–1.79) 
 3–4/week 5,317 (11.7) 31 108,266 1.01 (0.66–1.54) 1.06 (0.69–1.64) 
 ≥5/week 21,396 (46.9) 158 426,155 1.17 (0.88–1.56) 1.18 (0.87–1.60) 
Ptrend       0.341 0.405 
Women 
 Non/occasional drinkers 39,605 (78.8) 234 827,323 Ref Ref 
 1–2/week 7,100 (14.1) 37 151,141 1.15 (0.81–1.64) 1.07 (0.74–1.56) 
 3–4/week 1,586 (3.2) 33,272 0.70 (0.29–1.70) 0.61 (0.25–1.51) 
 ≥5/week 1,944 (3.9) 40,741 0.76 (0.36–1.61) 0.62 (0.29–1.35) 
Ptrend       0.563 0.217 
Pinteraction       0.478 0.503 
Association with binge drinking among men 
Amount of binge drinking (g/time) 
 <70 9,164 (68.3) 71 169,984 Ref Ref 
 70–<140 2,136 (15.9) 13 41,089 0.97 (0.53–1.79) 1.06 (0.58–1.95) 
 140–<210 668 (5.0) 13,112 0.56 (0.13–2.33) 0.64 (0.15–2.66) 
 ≥210 1,452 (10.8) 27,933 1.01 (0.47–2.17) 1.09 (0.50–2.37) 
Ptrend       0.808 0.953 
Frequency of binge drinking (per month) 
 No binge drinking 2,199 (8.9) 12 22,824 Ref Ref 
 1 4,492 (20.0) 29 73,400 0.88 (0.45–1.75) 0.92 (0.46–1.82) 
 2–3 6,510 (26.3) 44 107,304 0.96 (0.50–1.85) 1.03 (0.53–1.98) 
 ≥4 11,068 (44.8) 45 154,745 0.70 (0.36–1.34) 0.71 (0.37–1.38) 
Ptrend       0.200 0.204 

Abbreviation: Ref, reference.

aAdjusted for age (continuous) and area.

bAdjusted for age (continuous), area, history of diabetes, BMI, family history of cancer, coffee intake, smoking (pack-year), facial flushing, and status of sports.

This study also analyzed the association between alcohol intake and pancreatic cancer risk according to the alcohol subtypes of sake, shochu/awamori, and beer among men. Results showed that each alcohol type had no significant association with the risk of pancreatic cancer (Supplementary Table S1).

Table 6 shows the association between alcohol consumption and pancreatic cancer risk when the population was limited to men whose alcohol intake habits remained unchanged from the baseline survey to the 5-year follow-up survey. Regular drinkers who consumed an alcohol amount of 1 to 299 g/week showed a significant association with pancreatic cancer risk compared with non/occasional drinkers (multivariable-adjusted HR, 1.73; 95% CI, 1.15–2.60). When analyzed stratified by smoking status, alcohol consumption was significantly associated with pancreatic cancer risk in never smokers, with multivariable-adjusted HRs of 4.81 (95% CI, 1.51–15.31) and 5.48 (95% CI, 1.57–19.16) for 1 to 299 and ≥300 g of ethanol consumption per week (Ptrend = 0.007), respectively; meanwhile, this trend was not observed in former and current smokers.

Table 6.

Association between the habit of alcohol drinking unchanged from baseline survey to 5-year follow-up and pancreatic cancer risk in men and by smoking status.

 Number of subjects, n (%)Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Men 
 Non/occasional drinker 9,327 (37.7) 43 144,090 Ref Ref 
 1–299 g/week 8,395 (33.9) 66 129,748 1.77 (1.20–2.62) 1.73 (1.15–2.60) 
 ≥300 g/week 7,022 (28.4) 48 108,204 1.59 (1.04–2.43) 1.44 (0.91–2.28) 
Ptrend       0.027 0.121 
Never smoker 
 Non/occasional drinker 2,693 (44.4) 42,874 Ref Ref 
 1–299 g/week 2,240 (36.9) 15 35,521 4.60 (1.50–14.10) 4.81 (1.51–15.31) 
 ≥300 g/week 1,134 (18.7) 17,788 5.13 (1.53–17.21) 5.48 (1.57 -19.16) 
Ptrend       0.006 0.007 
Former smoker 
 Non/occasional drinker 2,136 (35.5) 12 32,161 Ref Ref 
 1–299 g/week 2,322 (38.6) 17 35,615 1.39 (0.65–2.96) 1.19 (0.54–2.61) 
 ≥300 g/week 1,556 (25.9) 11 23,929 1.31 (0.56–3.06) 1.04 (0.43–2.56) 
Ptrend       0.499 0.918 
Current smoker 
 Non/occasional drinker 4,458 (35.5) 27 68,466 Ref Ref 
 1–299 g/week 3,801 (30.3) 34 58,182 1.56 (0.94–2.61) 1.56 (0.91–2.68) 
 ≥300 g/week 4,296 (34.2) 28 65,952 1.18 (0.69–2.05) 1.18 (0.64–2.15) 
Ptrend       0.520 0.611 
Pinteraction       0.201 0.193 
 Number of subjects, n (%)Number of casesPerson-yearsHRa (95% CI)HRb (95% CI)
Men 
 Non/occasional drinker 9,327 (37.7) 43 144,090 Ref Ref 
 1–299 g/week 8,395 (33.9) 66 129,748 1.77 (1.20–2.62) 1.73 (1.15–2.60) 
 ≥300 g/week 7,022 (28.4) 48 108,204 1.59 (1.04–2.43) 1.44 (0.91–2.28) 
Ptrend       0.027 0.121 
Never smoker 
 Non/occasional drinker 2,693 (44.4) 42,874 Ref Ref 
 1–299 g/week 2,240 (36.9) 15 35,521 4.60 (1.50–14.10) 4.81 (1.51–15.31) 
 ≥300 g/week 1,134 (18.7) 17,788 5.13 (1.53–17.21) 5.48 (1.57 -19.16) 
Ptrend       0.006 0.007 
Former smoker 
 Non/occasional drinker 2,136 (35.5) 12 32,161 Ref Ref 
 1–299 g/week 2,322 (38.6) 17 35,615 1.39 (0.65–2.96) 1.19 (0.54–2.61) 
 ≥300 g/week 1,556 (25.9) 11 23,929 1.31 (0.56–3.06) 1.04 (0.43–2.56) 
Ptrend       0.499 0.918 
Current smoker 
 Non/occasional drinker 4,458 (35.5) 27 68,466 Ref Ref 
 1–299 g/week 3,801 (30.3) 34 58,182 1.56 (0.94–2.61) 1.56 (0.91–2.68) 
 ≥300 g/week 4,296 (34.2) 28 65,952 1.18 (0.69–2.05) 1.18 (0.64–2.15) 
Ptrend       0.520 0.611 
Pinteraction       0.201 0.193 

Abbreviation: Ref, reference.

aAdjusted for age (continuous) and area.

bAdjusted for age (continuous), area, history of diabetes, BMI, family history of cancer, coffee intake, smoking (pack-year), facial flushing, and status of sports.

The same analysis using data from the baseline survey and 10-year follow-up survey was conducted, but no significant association was observed (Supplementary Table S2). We also conducted another analysis in which individuals limited to men with increased or decreased alcohol consumption from the baseline survey to 5-year follow-up referring to those whose alcohol drinking habits unchanged from the baseline, and no significant association was observed either in those with increased or decreased alcohol consumption (Supplementary Materials and Methods; Supplementary Table S3).

This population-based prospective study investigated the association between alcohol consumption and risk of pancreatic cancer in Japan. Although the results showed no significant association between alcohol consumption and the risk of pancreatic cancer as overall results, an association was observed when the population was limited to men whose alcohol consumption habit remained unchanged from the baseline survey to the 5-year follow-up survey; a particularly strong association was observed in never smokers.

The absence of an association with alcohol consumption at baseline did not change when the analyses were performed with stratification by facial flushing responses or smoking status. A meta-analysis reported that the association between high alcohol consumption and pancreatic cancer risk is suggested mostly in case–control studies because the data on previous exposure collected after the diagnosis of cancer may be affected by the diagnosis; moreover, recall bias and publication bias may occur, which cannot be ignored (19). This study was a large-scale prospective cohort study; therefore, bias in the baseline information would not affect the results. In addition, compared with a previous report that used the same cohort study (13), our study had a longer follow-up period and a more finely divided categories of the amount of alcohol consumption. A longer follow-up period can allow the detection of small associations, and a finer division of categories can support the investigation of the trend in the amount of alcohol consumption.

In this study, an association was limited only to men whose drinking habits remained unchanged for 5 years, although several previous studies have shown an association between high alcohol intake and pancreatic cancer risk (13). The average amount of alcohol consumed in Japan is less than that in most Western countries. The alcohol consumption per year among adults from 2007 to 2017 is 7.1 L per capita (15.6 g/day) in Japan, whereas that in many European countries is higher than 8.9 L per capita (19.5 g/day) of the average across the Organisation for Economic Co-operation and Development countries (20). In addition, the proportion of dependent drinkers is much higher in many western countries than in Japan. Therefore, the prevalence of the highest alcohol consumption may be higher in Western countries than in Japan. In fact, although the highest alcohol consumption in this study was ≥450 g/week, which is similar to the definition of heavy drinkers (i.e., alcohol consumption of >420 g/week) in a previous study (21), the average amount of alcohol consumed in the highest consumption category in previous studies may be higher than that in this study, which was the reason for the positive association reported in previous studies.

We also investigated the association between alcohol consumption patterns and the risk of pancreatic cancer. A previous study demonstrated that binge drinking (alcohol consumption of >70 g/day) is associated with an increased risk of pancreatic cancer in men (22). This study also analyzed the association of pancreatic cancer risk with binge alcohol consumption; however, no significant association was demonstrated in the amount and frequency of binge drinking.

A previous study showed an increased risk of all-cause mortality among those who had fewer liver holidays, which refer to those who abstain from drinking (23). Liver holidays are considered important because daily heavy drinkers are consistently exposed to acetaldehyde compared with liver holiday takers. The association with the frequency of alcohol consumption was investigated in this study, and no significant association was observed between men and women.

According to a previous study that analyzed the European Prospective Investigation into Cancer and Nutrition (EPIC) study, the risk of pancreatic cancer was strongly associated with beer and spirit/liquor intake than with wine intake (21), and the ingredients in the alcohol can be confounders. In the case of beer, the intake may dose-dependently increase the amylase secretion and potentially disturb the exocrine activity of the pancreas, which increases the risk of pancreatic cancer (24). Results of the analysis of each alcohol subtype were not significant. The data of the EPIC study were based on the data from Europe, whereas those of this study were based on the data from Japan. The absence of the association between pancreatic cancer risk and beer intake may due to the fact that the amount of beer consumed in Japan is less than that consumed in European countries; therefore, the association might not be observed in Japan. When calculated as pure ethanol, the 90th percentile of daily beer intake is 20 g/day in Japan, whereas it is 40 g/day in Germany. If there are ingredients other than alcohol in beer that are associated with pancreatic cancer risk and the amount varies among beers produced in each country, it may cause the different results of the association.

Ethanol is metabolized by ADH1B to acetaldehyde and sequentially converted to acetate by ALDH2. ADH1B and ALDH2 are key enzymes for alcohol metabolism. The ADH1B His/His polymorphism, which is more prevalent in Japan than in Western countries, metabolizes faster than the ADH1B Arg allele (25). The ALDH2 Glu504Lys polymorphism encodes an inactive subunit (26) and is associated with high accumulation of acetaldehyde. More than 40% of Japanese individuals have an inactive form of ALDH2 (27). A previous study suggested that the risk of pancreatic cancer increases with alcohol consumption in individuals with the ADH1B His/His and ALDH2 504 Lys alleles (12). Acetaldehyde accumulation causes a facial flushing response to alcohol consumption (28). A previous study showed that the increased risk of alcohol-related cancers in men with a facial flushing response is associated with alcohol consumption, suggesting that the facial flushing response can be used as a reference for active ADH1B and inactive ALDH2 (29), although no significant association was found between facial flushing response and pancreatic cancer risk in this study. Therefore, we analyzed the association between alcohol consumption and pancreatic cancer risk after stratification by facial flushing response to evaluate the effect modification. Although our study showed no significant association, the HRs in individuals without facial flushing responses were higher than the HRs in those with the response. If acetaldehyde accumulation due to active ADH1B and inactive ALDH2 is associated with pancreatic cancer risk, the HRs should be higher in those with a facial flushing response. A sensitivity analysis was conducted by excluding individuals whose facial flushing response was unknown because the proportion of unknown response was much higher in nondrinkers compared with that in the other categories of alcohol consumption, and potential bias may have caused the difference in the association in the analysis. The results of the sensitivity analysis remained unchanged. Therefore, inactive ALDH2 is possibly associated with pancreatic cancer risk separately from alcohol intake (27). The facial flushing response as a phenotype is the combined effect of ADH1B and ALDH2 polymorphisms and is not particularly correlated with each polymorphism (28); however, a previous study observed a significant interaction between ALDH2 polymorphism and alcohol consumption for head and neck/esophageal cancers (30). In addition, the proportion of flushers was less among heavy drinkers than among never drinkers with the ALDH2 Glu/Lys genotype, suggesting that the flushing status might change with high and long-term alcohol exposure (31).

Smoking is an independent risk factor for pancreatic cancer (32–34) and is another potential confounder (35). A previous study showed an association between alcohol consumption and the risk of pancreatic cancer when limited to never smokers (36). A subgroup analysis stratified by smoking status was conducted among men, and no significant association was found for each smoking status (current, former, and never smokers).

In this study, a significant association between alcohol consumption and pancreatic cancer risk was observed when the population was limited to men with 5-year unchanged alcohol intake habit. Results suggested that the association may be stronger in this population because the influence of changes in alcohol consumption was eliminated. A sensitivity analysis using the amount of alcohol intake at baseline and carried out after excluding the first 2 and 5 years of follow-up showed no significant association. The result of this sensitivity analysis demonstrated that the increased risk in men with unchanged alcohol intake habit was not be due to the misclassification of alcohol consumption based on the changes by medical symptoms prior to diagnosis, and exposures were classified independently from the symptoms. When data from the baseline survey and 10-year follow-up survey were analyzed, no association was observed. This may be because the sample size in each category of exposure was limited due to the change in drinking habits of each individual from the baseline survey to the 10-year follow-up survey. In addition, the follow-up period from the 10-year follow-up survey is shorter than from 5-year follow-up survey. Although a significant association was observed in the limited population of men whose alcohol intake habits remained unchanged for 5 years, no previous studies have focused on investigating the unchanged habits. Hence, further investigations using other cohort studies are required.

This study has several limitations. First, the lifestyle characteristics of individuals were available only at baseline, and there could have been misclassifications to some extent in the exposure category. After the baseline data were obtained, the lifestyle factors used for adjustment, such as smoking, changed during the follow-up period and were not correctly reflected in the classification. Second, data of some confounding factors, such as the facial flushing response, were not fully available. This categorization may have impacted the results. Third, unmeasured variables and confounding factors could have affected the results, although the statistical model was adjusted for several possible confounding factors. For instance, one of the important risk factors for pancreatic cancer is a family history of pancreatic cancer; however, because the population with data on the family history of pancreatic cancer was relatively small, this study used family history of cancer instead. If this information can be obtained in future studies, the multivariable-adjusted model should be further adjusted. Fourth, non-proportional hazards were observed in never and current male smokers for the association with the amount of alcohol consumed at baseline and in women for the association with the drinking frequency. No significant association was observed in the analyses of these populations, and the effect of non-proportional hazards on the overall findings of this study would be limited. Hence, other statistical analysis methods should be used in future investigations. Finally, despite this large-scale cohort study, the sample size of patients with pancreatic cancer was limited. The results of the subgroup analysis may be unclear owing to the small sample size in each category.

In conclusion, although no significant association was observed between alcohol consumption and the risk of pancreatic cancer in the total population, a significant association was observed only when the population was limited to men with 5-year unchanged alcohol drinking habit.

T. Yamaji reports grants from Ministry of Health, Labour and Welfare of Japan during the conduct of the study. No disclosures were reported by the other authors.

Y. Okita: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. T. Sobue: Conceptualization, methodology, writing–review and editing. L. Zha: Conceptualization, methodology, writing–review and editing. T. Kitamura: Conceptualization, methodology, writing–review and editing. M. Iwasaki: Writing–review and editing. M. Inoue: Writing–review and editing. T. Yamaji: Writing-review and editing. S. Tsugane: Supervision, funding acquisition, writing–review and editing. N. Sawada: Funding acquisition, project administration, writing–review and editing.

The members of the JPHC-based prospective study are listed in the following website (as of April 2020): https://epi.ncc.go.jp/en/jphc/781/8510.html. We would like to thank the Akita, Iwate, Nagano, and Okinawa Cancer Registries for providing their incidence data. This study was supported by the National Cancer Center Research and Development Fund (since 2011), and Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare of Japan (from 1989 to 2010) to N. Sawada and S. Tsugane.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Epidemiology, Biomarkers & Prevention Online (http://cebp.aacrjournals.org/).

1.
Katanoda
K
,
Hori
M
,
Matsuda
T
,
Shibata
A
,
Nishino
Y
,
Hattori
M
, 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
.
2.
Ministry of Health Law
.
Vital Statistics Japan
;
2019
. Available from: https://www.e-stat.go.jp/.
3.
Siegel
RL
,
Miller
KD
,
Jemal
A
.
Cancer statistics, 2020
.
CA Cancer J Clin
2020
;
70
:
7
30
.
4.
Monitoring of Cancer Incidence in Japan - Survival 2009–2011 Report. Center for Cancer Control and Information Services, National Cancer Center,
2020
.
5.
Matsuda
T
,
Ajiki
W
,
Marugame
T
,
Ioka
A
,
Tsukuma
H
,
Sobue
T
.
Population-based survival of cancer patients diagnosed between 1993 and 1999 in Japan: a chronological and international comparative study
.
Jpn J Clin Oncol
2011
;
41
:
40
51
.
6.
Yadav
D
,
Lowenfels
AB
.
The epidemiology of pancreatitis and pancreatic cancer
.
Gastroenterology
2013
;
144
:
1252
61
.
7.
Maisonneuve
P
,
Lowenfels
AB
.
Risk factors for pancreatic cancer: a summary review of meta-analytical studies
.
Int J Epidemiol
2015
;
44
:
186
98
.
8.
Barone
E
,
Corrado
A
,
Gemignani
F
,
Landi
S
.
Environmental risk factors for pancreatic cancer: an update
.
Arch Toxicol
2016
;
90
:
2617
42
.
9.
Norat
T
,
Aune
D
,
Vieira
AR
,
Chan
D
,
Rosenblatt
DN
,
Vieira
R
.
WCRF/AICR systematic literature review continuous update project report: diet, nutrition, physical activity and pancreatic cancer
.
2018
.
Available from
: https://www.wcrf.org/wp-content/uploads/2021/02/pancreatic-cancer-slr.pdf.
10.
Higuchi
S
,
Matsushita
S
,
Murayama
M
,
Takagi
S
,
Hayashida
M
.
Alcohol and aldehyde dehydrogenase polymorphisms and the risk for alcoholism
.
Am J Psychiatry
1995
;
152
:
1219
21
.
11.
Asakage
T
,
Yokoyama
A
,
Haneda
T
,
Yamazaki
M
,
Muto
M
,
Yokoyama
T
, et al
.
Genetic polymorphisms of alcohol and aldehyde dehydrogenases, and drinking, smoking and diet in Japanese men with oral and pharyngeal squamous cell carcinoma
.
Carcinogenesis
2007
;
28
:
865
74
.
12.
Kanda
J
,
Matsuo
K
,
Suzuki
T
,
Kawase
T
,
Hiraki
A
,
Watanabe
M
, et al
.
Impact of alcohol consumption with polymorphisms in alcohol-metabolizing enzymes on pancreatic cancer risk in Japanese
.
Cancer Sci
2009
;
100
:
296
302
.
13.
Luo
J
,
Iwasaki
M
,
Inoue
M
,
Sasazuki
S
,
Otani
T
,
Ye
W
, et al
.
Body mass index, physical activity and the risk of pancreatic cancer in relation to smoking status and history of diabetes: a large-scale population-based cohort study in Japan—the JPHC study
.
Cancer Causes Control
2007
;
18
:
603
12
.
14.
Tsugane
S
,
Sawada
N
.
The JPHC study: design and some findings on the typical Japanese diet
.
Jpn J Clin Oncol
2014
;
44
:
777
82
.
15.
Tsugane
S
,
Fahey
MT
,
Sasaki
S
,
Baba
S
.
Alcohol consumption and all-cause and cancer mortality among middle-aged Japanese men: seven-year follow-up of the JPHC study cohort I. Japan public health center
.
Am J Epidemiol
1999
;
150
:
1201
7
.
16.
Shimazu
T
,
Inoue
M
,
Sasazuki
S
,
Iwasaki
M
,
Kurahashi
N
,
Yamaji
T
, et al
.
Alcohol and risk of lung cancer among Japanese men: data from a large-scale population-based cohort study, the JPHC study
.
Cancer Causes Control
2008
;
19
:
1095
102
.
17.
Ishiguro
S
,
Sasazuki
S
,
Inoue
M
,
Kurahashi
N
,
Iwasaki
M
,
Tsugane
S
.
Effect of alcohol consumption, cigarette smoking and flushing response on esophageal cancer risk: a population-based cohort study (JPHC study)
.
Cancer Lett
2009
;
275
:
240
6
.
18.
Fritz
A
,
Percy
C
,
Jack
A
,
Shanmugaratnam
K
,
Sobin
LH
, et al
.
International classification of diseases for oncology, 3rd ed.
World Health Organization
;
2000
.
Available from
: https://apps.who.int/iris/handle/10665/42344.
19.
Wang
YT
,
Gou
YW
,
Jin
WW
,
Xiao
M
,
Fang
HY
.
Association between alcohol intake and the risk of pancreatic cancer: a dose-response meta-analysis of cohort studies
.
BMC Cancer
2016
;
16
:
212
.
20.
Health at a glance 2019
.
OECD indicator
;
2019
.
Available from
: https://www.oecd-ilibrary.org/social-issues-migration-health/health-at-a-glance-2019_4dd50c09-en.
21.
Naudin
S
,
Li
K
,
Jaouen
T
,
Assi
N
,
Kyrø
C
,
Tjønneland
A
, et al
.
Lifetime and baseline alcohol intakes and risk of pancreatic cancer in the European prospective investigation into cancer and nutrition study
.
Int J Cancer
2018
;
143
:
801
12
.
22.
Gupta
S
,
Wang
F
,
Holly
EA
,
Bracci
PM
.
Risk of pancreatic cancer by alcohol dose, duration, and pattern of consumption, including binge drinking: a population-based study
.
Cancer Causes Control
2010
;
21
:
1047
59
.
23.
Marugame
T
,
Yamamoto
S
,
Yoshimi
I
,
Sobue
T
,
Inoue
M
,
Tsugane
S
.
Patterns of alcohol drinking and all-cause mortality: results from a large-scale population-based cohort study in Japan
.
Am J Epidemiol
2007
;
165
:
1039
46
.
24.
Gerloff
A
,
Singer
MV
,
Feick
P
.
Beer and its non-alcoholic compounds: role in pancreatic exocrine secretion, alcoholic pancreatitis and pancreatic carcinoma
.
Int J Environ Res Public Health
2010
;
7
:
1093
104
.
25.
Bosron
WF
,
Li
TK
.
Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism
.
Hepatology
1986
;
6
:
502
10
.
26.
Li
H
,
Borinskaya
S
,
Yoshimura
K
,
Kalina
N
,
Marusin
A
,
Stepanov
VA
, et al
.
Refined geographic distribution of the oriental ALDH2*504Lys (nee 487Lys) variant
.
Ann Hum Genet
2009
;
73
:
335
45
.
27.
Miyasaka
K
,
Hosoya
H
,
Tanaka
Y
,
Uegaki
S
,
Kino
K
,
Shimokata
H
, et al
.
Association of aldehyde dehydrogenase 2 gene polymorphism with pancreatic cancer but not colon cancer
.
Geriatr Gerontol Int
2010
;
10
Suppl 1
:
S120
6
.
28.
Matsuo
K
,
Wakai
K
,
Hirose
K
,
Ito
H
,
Saito
T
,
Tajima
K
.
Alcohol dehydrogenase 2 His47Arg polymorphism influences drinking habit independently of aldehyde dehydrogenase 2 Glu487Lys polymorphism: analysis of 2,299 Japanese subjects
.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
1009
13
.
29.
Ono
A
,
Inoue
M
,
Sawada
N
,
Saito
E
,
Yamaji
T
,
Shimazu
T
, et al
.
Impact of alcohol drinking on cancer risk with consideration of flushing response: the Japan public health center-based prospective study cohort (JPHC study)
.
Prev Med
2020
;
133
:
106026
.
30.
Chang
JS
,
Hsiao
JR
,
Chen
CH
.
ALDH2 polymorphism and alcohol-related cancers in Asians: a public health perspective
.
J Biomed Sci
2017
;
24
:
19
.
31.
Oze
I
,
Matsuo
K
,
Hosono
S
,
Ito
H
,
Kawase
T
,
Watanabe
M
, et al
.
Comparison between self-reported facial flushing after alcohol consumption and ALDH2 Glu504Lys polymorphism for risk of upper aerodigestive tract cancer in a Japanese population
.
Cancer Sci
2010
;
101
:
1875
80
.
32.
Bosetti
C
,
Lucenteforte
E
,
Silverman
DT
,
Petersen
G
,
Bracci
PM
,
Ji
BT
, et al
.
Cigarette smoking and pancreatic cancer: an analysis from the international pancreatic cancer case-control consortium (Panc4)
.
Ann Oncol
2012
;
23
:
1880
8
.
33.
Raimondi
S
,
Maisonneuve
P
,
Löhr
JM
,
Lowenfels
AB
.
Early onset pancreatic cancer: evidence of a major role for smoking and genetic factors
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
1894
7
.
34.
Pandol
SJ
,
Apte
MV
,
Wilson
JS
,
Gukovskaya
AS
,
Edderkaoui
M
.
The burning question: why is smoking a risk factor for pancreatic cancer?
Pancreatology
2012
;
12
:
344
9
.
35.
International AIfWCRF
.
2012
Pancreatic cancer | Continuous Update Project Report
.
Food, nutrition, physical activity, and the prevention of pancreatic cancer
. Available from: http://www.dietandcancerreport.org.
36.
Gapstur
SM
,
Jacobs
EJ
,
Deka
A
,
McCullough
ML
,
Patel
AV
,
Thun
MJ
.
Association of alcohol intake with pancreatic cancer mortality in never smokers
.
Arch Intern Med
2011
;
171
:
444
51
.

Supplementary data