The Effect of Alcohol Consumption Behavior Changes on Gastric Cancer Risks Stratified by Sex in South Korea

Yonghoon Choi; Jieun Jang; Hyeong Ho Jo; Nayoung Kim

doi:10.4143/crt.2024.591

Articles

Page Path: HOME > Cancer Res Treat > Volume 58(1); 2026 > Article

Original Article Gastrointestinal cancer The Effect of Alcohol Consumption Behavior Changes on Gastric Cancer Risks Stratified by Sex in South Korea: Yonghoon Choi¹, Jieun Jang², Hyeong Ho Jo³, Nayoung Kim^1,4; Cancer Research and Treatment : Official Journal of Korean Cancer Association 2026;58(1):232-241.
DOI: https://doi.org/10.4143/crt.2024.591
Published online: April 1, 2025

¹Department of Internal Medicine and Research Center for Sex- and Gender-Specific Medicine, Seoul National University Bundang Hospital, Seongnam, Korea

²Division of Clinical Research, Research Institute, National Cancer Center, Goyang, Korea

³Department of Internal Medicine, Daegu Catholic University School of Medicine, Daegu, Korea

⁴Department of Internal Medicine and Liver Research Institute, Seoul National University College of Medicine, Seoul, Korea

Correspondence: Nayoung Kim, Department of Internal Medicine and Research Center for Sex- and Gender-Specific Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro 173 Beon-gil, Bundang-gu, Seongnam 13620, Korea
Tel: 82-31-787-7008 E-mail: nakim49@snu.ac.kr

^*Yonghoon Choi and Jieun Jang contributed equally to this work.

• Received: June 24, 2024 • Accepted: March 31, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

2,267 Views
106 Download

prev next

Full Article

Download PDF

Abstract

Purpose
The effect of behavior changes in alcohol drinking on gastric cancer (GC) development, and the sex differences in those effects have not yet been fully elucidated. This study investigated the effect of behavior changes in alcohol drinking on the GC risk by sex.
Materials and Methods
The cohort consisted of 310,192 Koreans (≥ 40 years) from the National Health Insurance Service–Health Screening Cohort with a median follow-up period of 12 years. Subjects were classified according to alcohol consumption behavior changes (non-drinker, quitter, reducer, sustainer, and increaser). The independent effect of changes in alcohol drinking patterns or concurrent effect of alcohol on GC risk were evaluated using the Cox proportional hazard regression.
Results
In males, non-drinkers showed a lower risk of developing GC (hazard ratio [HR], 0.91; 95% confidence interval [CI], 0.84 to 0.98), whereas increasers showed a higher risk of GC than sustainers (HR, 1.11; 95% CI, 1.02 to 1.20). Starting to drink alcohol, even at a mild level, was associated with an increased GC risk, while a decreased GC risk was induced when alcohol consumption dose decreases to a mild from a moderate level among males. However, in females, only substantial change of alcohol consumption dose from non- to heavy-drinking was associated with increased GC risk (HR, 1.97; 95% CI, 0.98 to 3.96).
Conclusion
These results suggest that alcohol abstinence can reduce the risk of developing GC, particularly among males.
Key words: Alcohol abstinence, Stomach neoplasms, Sex characteristics, Primary prevention

Introduction

Gastric cancer (GC) is the fifth most common malignancy worldwide; in Eastern Asia, including South Korea which has the highest incidence rate [1,2]. Helicobacter pylori infection and cigarette smoking reportedly explain more than 80% of all GC cases [3,4], while alcohol consumption has also been suggested to contribute [5]. Although previous studies of the association between alcohol consumption and GC risk have suggested that moderate and/or heavy alcohol consumption increases the risk of GC development [6,7], few studies have focused on how changes in alcohol consumption behavior affect it. In addition to this, females are generally considered to be more vulnerable to exposure to alcohol, that females are less capable of metabolizing ethanol and more likely to exhibit the harmful effects of accumulated acetaldehyde than males because of lower body weight and higher body fat ratio [8,9]. However, research focusing sex differences on the correlation between changes in alcohol consumption behavior and GC risk is lacking.

South Korea is building a vast amount of national health data through the National Health Insurance Service (NHIS), which is operated as a single system and includes 97% of the total population. In addition, the National Health Screening (NHS) includes 72% of the total population, thereby ensuring representativeness of the national population [10]. Data collected by the NHS include information on demographic (age, sex) and socioeconomic (income) factors, health behaviors such as alcohol consumption and smoking status, medical history, family history, lab test results, and healthcare usage [10,11]. Furthermore, these data are collected repeatedly, thus they can be used to assess the impact of health behavior changes on health outcomes.

From this background, we aimed to evaluate the effects of changes in alcohol consumption behavior on GC risk based on a large-scale retrospective cohort, the National Health Insurance Service–Health Screening Cohort (NHIS-HEALS), focusing on the sex- and gender-specific stratification.

Materials and Methods

1. Study design and data source

This retrospective cohort study was based on the NHIS-HEALS [12], which consisted of 514,866 subjects randomly sampled from participants who underwent the NHS programs provided by the NHIS during 2002-2003 in Korea (Fig. 1A). This cohort was followed up annually through 2019 for healthcare usage and death information. The information provided by the NHIS-HEALS includes socioeconomic and demographic factors, health behaviors, medical history, family history, laboratory test results, and healthcare usage.

2. Study population

To select appropriate participants with information on alcohol consumption behavior changes, those for whom baseline information was unavailable for alcohol consumption frequency and drinking amount per occasion were excluded (2002-2003) (Fig. 1A). We set the interval for measuring alcohol consumption behavior changes to at least 2 years and not more than 5 years (Fig. 1B). Therefore, subjects for whom no alcohol consumption behavior information was available for 2005-2006 (baseline year, 2002) as well as those for whom no information was available for 2006-2007 (baseline year, 2003) were excluded. In addition, subjects with a past history of cancer defined as those with claims data with a major diagnosis code starting with “C” were also excluded before the date of the first alcohol consumption behavior measurement. In the analysis, we considered the sex data collected by the NHS as biological sex.

3. Operative definition

1) Alcohol consumption behavior changes

We calculated daily alcohol consumption (g/day) in two periods (2002-2003 and 2005-2007) using the information about alcohol drinking frequency per week, drink amount per occasion, and amount of ethanol per drink (g). In cases where multiple measurements of alcohol consumption were available within a period, we used the first collected measurement. Subjects were classified according to daily alcohol consumption amount as follows: (1) male: non-drinker, mild drinker (< 15 g/day), moderate drinker (15-29.9 g/day), heavy drinker (≥ 30 g/day); and (2) female: non-drinker, mild drinker (< 7.5 g/day), moderate drinker (7.5-14.9 g/day), and heavy drinker (≥ 15 g/day). Based on the alcohol drinking amount in 2002-2003 and 2005-2007, participants were further classified into five groups: (1) non-drinkers who did not drink continuously; (2) quitters who had been drinkers in 2002-2003 but stopped consuming alcohol in 2005-2007; (3) reducers who reduced but did not quit their daily alcohol consumption amount in 2002-2003 and 2005-2007; (4) sustainers who maintained their level of alcohol consumption; and (5) increasers who increased their level of alcohol consumption. We used the alcohol consumption behavior change between 2002-2003 and 2005-2007 as the exposure under the assumption that alcohol consumption pattern in 2005-2007 would be maintained during the follow-up period. To check if the above assumptions are appropriate, we selected subjects with information of alcohol consumption pattern in both periods of 2005-2007 and 2014-2017. Then, we calculated the proportion of subjects who had sustained, quit, reduced, and increased alcohol consumption between these two periods. We conducted this study under the assumption that alcohol consumption behavior would not change after the second measurement of alcohol consumption (2005-2007). Therefore, we verified the number of participants whose alcohol consumption changed between the second measurement point (2005-2007) and the subsequent time point (T3, 2014-2017), and this was represented using a Sankey diagram. The GC patients were excluded from the above analysis since it is possible that they had perceived their health as poor and changed health behaviors such as alcohol drinking, as they approached the time of diagnosis.

2) GC incidence and follow-up

GC cases were defined as patients with medical claims using the International Classification of Diseases, 10th revision (ICD-10) code of C16 together with an admission history in the NHIS-HEALS. The index date was defined as the date of the last collection of alcohol consumption data. The study subjects were followed up from the index date to the date of GC diagnosis, death, or the last date of follow-up (December 31, 2019), whichever occurred first.

4. Statistical analyses

We tested differences in baseline characteristics according to changes in alcohol drinking behavior using the chi-squared test for categorical variables and analysis of variance for continuous variables. The baseline characteristics included sociodemographic factors such as age, sex, and income; health behaviors such as smoking status and physical activity; body mass index (BMI); fasting glucose level; total cholesterol; blood pressure; and comorbidities including diabetes mellitus, hypertension, and dyslipidemia before the last alcohol drinking information collection.

GC risk was evaluated according to changes in alcohol consumption behavior by setting sustainers as the reference group based on Cox proportional hazard regression. To consider multiple testing correction in the analysis, we applied an adjusted significance level to test for significance by dividing 0.05 by the number of predictors in the Cox proportional hazard regression model.

After performing the sensitivity analyses considering various latency periods (2-year, 3-year, and 4-year), we decided to consider the 3-year of latency period to assess the association alcohol consumption behavior change, smoking, and GC development. Since we assumed that at least a 3-year latency period is needed to assess the impact of changes in alcohol drinking behavior on GC incidence, GC cases diagnosed within 3 years from the start of follow-up were excluded. We defined potential confounders as baseline characteristics that differed according to changes in alcohol consumption behavior and were associated with GC incidence based on a univariate analysis. Potential confounders including age; sex; smoking status; physical activity; BMI; and history of diabetes mellitus, hypertension, and dyslipidemia were included as exploratory variables in the Cox proportional hazard regression adjustment. Although some values for potential confounders were missing—including smoking status, income, BMI, and physical activity—the proportion of missing values was < 1% except for smoking status (4.4%). To handle these missing values, we performed data imputation using the PROC MI procedure in SAS software. All data management and statistical analyses were performed using the SAS software ver. 9.4 (SAS Institute Inc.).

Results

1. Study population

The analyses were performed using data from the 310,192 subjects, including 5,083 GC cases considering 3-year latency period. The participants’ baseline characteristics are summarized in Table 1. The non-drinker group accounted for approximately half of the overall study participants (n=147,022), while the proportion of participants with reduced alcohol consumption was the smallest (n=14,480). The distributions of age, sex, and smoking status in the non-drinker group differed from those in the other groups, particularly in the reducer and sustainer groups. Specifically, the mean age at baseline was the highest (57.5 years), and > 70% of the non-drinker group was female, whereas the mean age in the reducer (54.5 years) and sustainer (52.6 years) groups was lower than that in the non-drinker group, and < 12% of each group was female. Additionally, former and current smokers comprised < 20% of the non-drinker group, whereas > 50% of the reducer and sustainer groups consisted of former and current smokers. Additionally, we selected subjects whose information of alcohol consumption pattern in both 2005-2007 and 2014-2017 exists and who had not been diagnosed as GC during follow-up period. Then we checked the changes in alcohol consumption pattern between two time points (2005-2007 and 2014-2017). Comparison of NHIS-HEALS data and Korean National Health and Nutrition Examination Survey (KNHANES) data showed some discrepancy (S1 Table). Approximately, 70% of them sustained their alcohol consumption pattern in 2005-2007 until 2014-2017 (S2 Fig.).

2. GC risk according to alcohol consumption dose change

To consider alcohol consumption dose in evaluating the association between alcohol consumption behavior change and GC risk, subjects were stratified according to alcohol consumption dose (non-drinker, mild, moderate, and heavy drinkers) in 2002-2003 (Table 2). GC risk was assessed according to alcohol consumption dose in 2005-2007, using as the reference group those who had maintained the same alcohol consumption dose between 2002-2003 and 2005-2007. For non-drinkers in 2002-2003, increasing alcohol consumption dose to mild level was associated with increased GC risk (hazard ratio [HR], 1.14; 95% confidence interval [CI], 1.01 to 1.29). We also found an elevated GC risk in subjects who changed their alcohol consumption from non-drinking in 2002-2003 to heavy drinking in 2005-2007, though this association was marginally significant (HR, 1.37; 95% CI, 0.99 to 1.90). In other groups, including mild, moderate, and heavy drinkers in 2002-2003, the alcohol consumption behavior change was not related to differences in GC risk.

For male non-drinkers in 2002-2003, increasing alcohol consumption dose into mild level was associated with elevated GC risk (HR, 1.17; 95% CI, 1.02 to 1.34) (Table 2, Fig. 2A). Conversely, decreased GC risk was observed in males who had decreased alcohol consumption dose from moderate level into mild level (HR, 0.78; 95% CI, 0.61 to 0.99). For male heavy drinkers in 2002-2003, alcohol consumption dose reduction and abstinence did not result in differences in the risk of GC.

In females, increasing alcohol consumption (from non-drinking to heavy drinking) was associated with a substantially increased GC risk (HR, 1.97; 95% CI, 0.98 to 3.96) (Table 2, Fig. 2B). Except for the alcohol consumption behavior change from non-drinking to heavy drinking, none of the other alcohol consumption behavior changes resulted in differences in GC risks in females.

3. GC risk according to alcohol consumption behavior change

At first, GC risk was lower in the non-drinker versus sustainer group (HR, 0.93; 95% CI, 0.86 to 0.99). In contrast, increased GC risk was observed in the increaser group compared to the sustainer group (HR, 1.09; 95% CI, 1.01 to 1.18) (S3 Table). In addition, a sensitivity analysis was performed by excluding GC cases diagnosed within 2 years, 3 years, and 4 years from follow-up to consider the latency period.

When GC cases diagnosed within 3 years were excluded from baseline, we observed elevated GC risk in the increaser compared to sustainer (HR, 1.11; 95% CI, 1.02 to 1.20) (Table 3). In specific, decreased and increased GC risks compared to the sustainer were observed in non-drinker and increaser groups (HR, 0.89; 95% CI, 0.81 to 0.97 for the non-drinker; HR, 1.12; 95% CI, 1.01 to 1.22 for the increaser) in males. However, no differences in GC risk according to changes in alcohol consumption behavior were observed in females. Based on this result, we applied a 3-year latency period to subsequent analyses.

Discussion

In this study, we aimed to assess the association between changes in alcohol consumption behaviors and GC risk according to sex. As a result, compared to sustaining the alcohol consumption dose, increasing the dose was associated with an elevated GC risk, while not consuming alcohol continuously was related to a decreased GC risk. When alcohol dose at baseline was considered, starting to drink alcohol, even at a mild level, was associated with an increase in GC risk. Furthermore, lowering the amount of alcohol consumed to a mild level reduced the risk of GC, even in individuals who had consumed alcohol at a moderate level. These changes in GC risk according to alcohol consumption behavior changes were more pronounced in males than in females.

Recently, studies have been widely conducted to clarify the link between high-risk behaviors such as alcohol [13-16], smoking [14-16], obesity [16], and metabolic syndrome [17,18], and gastrointestinal tract cancers using big data. Some of these studies above were conducted using the database generated by the NHS in South Korea, and it has been validated to have sufficient accuracy in operationally defining gastrointestinal cancers such as colon cancer [19], pancreatic cancer [20], and bile duct cancer [21] using ICD-10 codes within the data from NHS.

In studies of GC, the association between alcohol consumption and the risk of either gastric cardia or non-cardia cancer in several Western studies [14-16]. In opposite, a meta-analysis showed a positive association between excessive alcohol consumption (> 50 g/day) and a higher relative risk for gastric non-cardia cancer [5]. In addition, recent two large-scale studies in South Korea reported that even small amount of alcohol drinking was associated with increased risks of GC [22,23]. The difference in frequency of polymorphism on the aldehyde dehydrogenase (ALDH)-2 may be one of possible mechanisms for the conflicting results in the association between alcohol and the GC risk [5]. It is well known that Asian populations show relatively higher prevalence of the variant genotype (poor metabolizer) than other populations, and this may account for the stronger association between alcohol and GC risk in Asian studies [24-26].

The rate of ethanol absorption is influenced by various factors including the amount of ethanol consumed, body composition, gastric emptying, and enzymatic activity [27]. Since females have less body water volume, blood concentration of ethanol becomes higher than males when the same amount of alcohol is consumed. Additionally, the absorption rate of ethanol is lower due to the high body fat ratio, and the gastric emptying rate is known to be slower in females than in males [9,28]. Above all, males are known to have 70%-80% higher gastric ALDH activity than females. In contrast, females have a higher hepatic ALDH activity [27] and are presumed to exhibit higher concentrations of toxic metabolites such as acetaldehyde in the liver [29].

However, in the present study, the effect of changes in alcohol consumption on the risk of developing GC was more pronounced in males than in females. Recently, similar results were found in large-scale cohort studies in South Korea, where current alcohol drinkers showed a greater risk of GC than non-drinkers in males, but not in females [23,30]. Similarly, Chinese [31] and Japanese [32] studies and a meta-analysis [33] reported a significant association between drinking and GC only in males. Those studies estimated that the incidence of GC in males is about twice that of females, and that light drinking may have a protective effect against GC in females. Another study also suggested the possibility that high testosterone levels in males were associated with heavy drinking and alcohol abuse [34]. In addition to this, these results might be derived from females’ tendency to hide their drinking or smoking behaviors. There have been reports of “stigmatization of female alcohol drinking” [35-37]. In addition, South Korean study also pointed out the possibility of misclassification, underreporting, and omission of information regarding alcohol consumption habits [23]. Another reason could be that the diffuse-type GC is more frequent than intestinal-type in females [38]. The diffuse-type GC in females is more affected from estrogen receptor–α [39] but intestinal-type, which is highly frequent in males, is dependent on environmental factors such as smoking and alcohol. The last possible reason is that the prevalence of GC is two times frequent in males than in females similar to alcohol drinking which easily cause the statistical significance for alcohol drinking.

Nevertheless, our study based on NHIS-HEALS differs somewhat from the results of the KNHANES in Korea (S1 Table), but it is not thought to be a significant difference in unreliable levels. In spite of these situations, our study found an increase in the risk of GC in females when alcohol consumption rapidly increased from non-drinking to heavy drinking. However, the proportion of current smokers was approximately two times higher in females who had dramatically increased their alcohol consumption than in the other groups. This suggests the possibility of the synergistic effects of alcohol consumption and smoking on GC development particularly in females. Furthermore, increasing the amount of alcohol consumption and smoking at the same time may pose an additional GC risk. Thus, strong warning is needed for female smokers who rapidly increased their alcohol consumption.

Health behaviors, including alcohol consumption and smoking, are the leading causes of cancer, thus modified by individuals’ beliefs, sentiments, and health conditions [40,41]. However, most previous studies evaluated the effects of alcohol consumption under the assumption that the initially measured drinking behavior is continuously maintained. Our study has suggested clear message based on several strong points. First, it used data for a large-scale cohort constructed by sampling from a representative database and repeated measures. Thus, it was possible to estimate the risk of GC according to changes in alcohol consumption behavior. Second, sex-based differences were considered in all analyses to evaluate the association between changes in alcohol consumption behaviors and GC risk. As mentioned above, the harmful effects of alcohol drinking differ between males and females both biologically [9,28] and in typical health behavior [35-37]. Moreover, considering previous reports that self-reported alcohol drinking information is less accurate among females than males, considering sex-based differences is essential for assessing the impact of alcohol. Third, a 3-year latency period was considered to minimize reverse causation. Changes in health behavior can be motivated when people are diagnosed with a chronic illness or their perceived susceptibility to disease is high [42]. In other words, changes to one’s health behavior may be a result of one’s health condition.

Our study has some limitations, that the results were not completely consistent results in all groups in our study and some groups showed an increased risk of GC despite a decrease in alcohol consumption although the difference was not significant. These are thought to be due to the following reasons. First, there is a possibility that the response was not accurate especially in females as described above [35-37], as this was a questionnaire-based study. Second, alcohol consumption pattern of some subjects had changed during the follow-up period. Though we found evidence supporting that approximately 70% of subjects had maintained their alcohol consumption pattern for a long-term period (more than 7 years), there were subjects who change their alcohol consumption dose during follow-up and these changes were not considered in the analysis. Third, it was not possible to include the clinicopathological characteristics of GC, such as histologic type or stage and H. pylori infection status, a major carcinogen of GC in this study. Lastly, the NHIS-HEALS database consists of health screening participants aged > 40 years old, so we could not assess the impact of alcohol consumption behavior changes on early-onset GC.

In spite of these limitations, our study showed that abstinence from alcohol consumption continuously and increasing alcohol consumption were associated with a decreased and elevated GC risk, respectively, compared with sustainers particularly among males. In addition, even if people drank at a moderate level in the past, the risk of GC decreases when they reduce alcohol consumption to a mild level. These results are consistent with the latest research using Korean NHIS data that reported a reduced risk of developing GC in decreasing-heavy drinkers compared to steady-heavy drinkers [43]. Furthermore, we attempted to analyze the effect of alcohol consumption behavior changes according to sex, unlike previous study [39] that has only been conducted on males. We found that the changes in GC risk were more pronounced in males and relatively not significant in females, and we raised the need for follow-up studies to overcome the limitations of questionnaire-based research in females.

In conclusion, our study suggests that alcohol abstinence and quitting smoking together can reduce the risk of developing GC, particularly among males.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).

crt-2024-591_S1_Table.pdf

crt-2024-591_S2_Fig.pdf

crt-2024-591_S3_Table.pdf

NOTES

Ethical Statement

The study protocol was reviewed and approved by the institutional review board (IRB) of the Seoul National University Bundang Hospital (IRB No. X-2209-780-901). This study was carried out in accordance with the recommendations of the Declaration of Helsinki for biomedical research involving human subjects and the Guidelines for Good Clinical Practice. Informed consent was waived by the IRB.

Author Contributions

Conceived and designed the analysis: Choi Y, Jang J, Jo HH, Kim N.

Collected the data: Choi Y, Jang J.

Contributed data or analysis tools: Jang J.

Performed the analysis: Jang J.

Wrote the paper: Choi Y, Jang J.

Manuscript review and final approval: Choi Y, Jang J, Jo HH, Kim N.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant (RS 2024-00337453) funded by the Korea government (MSIT). In addition, it was supported by Seoul National University Bundang Hospital Research fund (02-2023-0012). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Fig. 1.

Study scheme. (A) Flowchart of study subject selection. (B) Time points to measure alcohol consumption behavior change and to start follow-up. The orange box indicates the period to measure alcohol consumption frequency and dose in 2002. For subjects whose alcohol consumption behavior was measured in 2002, their alcohol consumption behaviors after change were measured during 2005-2006 (red box). For subjects whose alcohol consumption behavior before change was measured in 2003, the measurement periods before and after the change were defined as the green and blue boxes. The follow-up was designated to begin at the time of the last alcohol consumption behavior (red arrow line). NHIS-HEALS, National Health Insurance Service-National Health Screening Cohort.

Fig. 2.

Gastric cancer risk depending on alcohol consumption behavior change and sex. (A) Gastric cancer risk according to alcohol consumption behavior change between 2002-2003 and 2005-2007 among males. (B) Gastric cancer risk according to alcohol consumption behavior change between 2002-2003 and 2005-2007 among females. CI, confidence interval; HR, hazard ratio. An asterisk (*) indicates statistical significance.

Table 1.

Baseline characteristics of study subjects according to alcohol consumption behavior change^a)

Characteristic	Non-drinker (n=147,022)	Quitter (n=34,928)	Reducer (n=14,480)	Sustainer (n=75,209)	Increaser (n=38,553)	p-value
Baseline alcohol consumption level
Non-drinker	147,022 (100)	0	0	0	23,830 (61.8)	< 0.01
Mild	0	28,739 (82.3)	0	64,484 (85.7)	11,496 (29.8)
Moderate	0	3,419 (9.8)	7,674 (53.0)	4,941 (6.6)	3,227 (8.4)
Heavy	0	2,770 (7.9)	6,806 (47.0)	5,784 (7.7)	0
Age (yr)	57.5±9.5	55.5±9.2	54.5±8.6	52.6±7.8	54.2±8.8	< 0.01
Sex
Male	43,685 (29.7)	23,428 (67.1)	13,054 (90.1)	66,466 (88.4)	27,928 (72.4)	< 0.01
Female	103,337 (70.3)	11,500 (32.9)	1,426 (9.9)	8,743 (11.6)	10,625 (27.6)
Smoking status
Non-smoker	128,245 (87.2)	28,185 (80.7)	5,456 (37.7)	31,848 (42.4)	19,629 (50.9)	< 0.01
Former smoker	4,752 (3.2)	1,970 (5.6)	2,115 (14.6)	12,671 (16.8)	4,811 (12.5)
Current smoker	10,271 (7.0)	3,549 (10.2)	5,800 (40.0)	25,609 (34.0)	11,521 (30.0)
Missing	3,754 (2.6)	1,224 (3.5)	1,109 (7.7)	5,081 (6.8)	2,592 (6.7)
Income
Q1	22,810 (15.5)	4,868 (13.9)	1,824 (12.6)	7,668 (10.2)	5,055 (13.1)	< 0.01
Q2	22,464 (15.3)	4,835 (13.8)	1,899 (13.1)	7,636 (10.2)	5,382 (14.0)
Q3	24,123 (16.4)	5,628 (16.1)	2,549 (17.6)	10,438 (13.9)	6,413 (16.6)
Q4	29,402 (20.0)	7,090 (20.3)	3,217 (22.2)	15,858 (21.1)	8,050 (20.9)
Q5	47,802 (32.5)	12,425 (35.6)	4,966 (34.3)	33,544 (44.6)	13,582 (35.2)
Missing	421 (0.3)	82 (0.2)	25 (0.2)	65 (0.1)	71 (0.2)
BMI (kg/m²)	23.9±3.0	24.0±2.9	24.2±2.8	24.1±2.7	24.0±2.8	< 0.01
Physical activity/wk
0	81,186 (55.2)	20,939 (60.0)	6,071 (41.9)	25,708 (34.2)	16,874 (43.8)	< 0.01
1-2	32,021 (21.8)	7,166 (20.5)	4,717 (32.6)	29,057 (38.6)	11,539 (29.9)
≥ 3	31,648 (21.5)	6,145 (17.6)	3,463 (23.9)	19,607 (26.1)	9,622 (25.0)
Missing	2,167 (1.5)	678 (1.9)	229 (1.6)	837 (1.1)	518 (1.3)
Fasting glucose, mean±SD	96.9±25.4	98.8±27.9	102.7±31.7	99.4±26.3	99.7±27.4	< 0.01
Total cholesterol, mean±SD	200.3±37.3	197.7±37.1	197.6±37.0	197.4±35.6	197.6±36.6	< 0.01
SBP, mean±SD	125.1±16.7	125.9±16.4	129.0±16.4	126.8±15.7	127.0±16.4	< 0.01
DBP, mean±SD	77.3±10.5	78.6±10.6	81.0±10.7	79.9±10.5	79.5±10.7	< 0.01
DM past history
No	127,477 (86.7)	30,567 (87.5)	12,677 (87.5)	67,379 (89.6)	34,172 (88.6)	< 0.01
Yes	19,545 (13.3)	4,361 (12.5)	1,803 (12.5)	7,830 (10.4)	4,381 (11.4)
HTN past history
No	110,158 (74.9)	26,802 (76.7)	11,106 (76.7)	59,516 (79.1)	30,355 (78.7)	< 0.01
Yes	36,864 (25.1)	8,126 (23.3)	3,374 (23.3)	15,693 (20.9)	8,198 (21.3)
Dyslipidemia past history
No	124,632 (84.8)	30,219 (86.5)	12,653 (87.4)	66,248 (88.1)	33,851 (87.8)	< 0.01
Yes	22,390 (15.2)	4,709 (13.5)	1,827 (12.6)	8,961 (11.9)	4,702 (12.2)

Values are presented as number (%) or mean±SD. BMI, body mass index; DBP, diastolic blood pressure; DM, diabetes mellitus; HTN, hypertension; Q, quintile; SBP, systolic blood pressure; SD, standard deviation.

^a) With consideration of 3-year latency period.

Table 2.

Risk of gastric cancer incidence according to alcohol consumption dose change^a)

Alcohol dose in 2002-2003^b)	Alcohol dose in 2005-2007^b)	Total			Male			Female
Alcohol dose in 2002-2003^b)	Alcohol dose in 2005-2007^b)	No. of cases	Incidence rate^c)	HR (95% CI)^d)	No. of cases	Incidence rate^c)	HR (95% CI)^e)	No. of cases	Incidence rate^c)	HR (95% CI)^e)
Non-drinker	Non-drinker	1,917	102.3	1.00	936	170.0	1.00	981	74.1	1.00
	Mild	347	133.1	1.14 (1.01-1.29)	279	177.3	1.17 (1.02-1.34)	68	65.9	1.05 (0.82-1.35)
	Moderate	42	154.4	1.08 (0.80-1.48)	39	210.2	1.16 (0.84-1.60)	3	34.7	0.58 (0.19-1.81)
	Heavy	38	197.4	1.37 (0.99-1.90)	30	220.3	1.27 (0.88-1.83)	8	142.2	1.97 (0.98-3.96)
Mild	Non-drinker	511	138.5	1.07 (0.96-1.20)	418	175.4	1.05 (0.93-1.18)	93	71.3	1.28 (0.89-1.83)
	Mild	1,171	137.6	1.00	1,121	149.9	1.00	50	48.5	1.00
	Moderate	175	178.6	1.15 (0.98-1.35)	170	202.8	1.17 (0.99-1.37)	5	35.4	0.76 (0.30-1.91)
	Heavy	92	185.9	1.17 (0.95-1.45)	91	199.7	1.19 (0.96-1.47)	1	25.5	0.47 (0.06-3.43)
Moderate	Non-drinker	76	180.0	0.93 (0.68-1.26)	67	208.2	0.90 (0.65-1.24)	9	89.7	1.51 (0.43-5.21)
	Mild	153	155.7	0.80 (0.63-1.02)	143	166.0	0.78 (0.61-0.99)	10	82.4	1.44 (0.45-4.64)
	Moderate	125	200.5	1.00	121	220.1	1.00	4	54.3	1.00
	Heavy	94	234.1	1.13 (0.86-1.47)	91	242.2	1.10 (0.84-1.44)	3	116.4	2.00 (0.44-9.09)
Heavy	Non-drinker	58	170.8	1.04 (0.74-1.46)	53	204.2	1.08 (0.76-1.52)	5	62.5	0.42 (0.10-1.86)
	Mild	79	160.1	0.87 (0.66-1.14)	77	167.7	0.87 (0.65-1.15)	2	58.8	0.58 (0.09-3.60)
	Moderate	70	191.0	1.00 (0.75-1.34)	69	205.0	1.02 (0.76-1.36)	1	33.5	0.43 (0.04-4.17)
	Heavy	135	186.7	1.00	132	193.4	1.00	3	74.0	1.00

CI, confidence interval; HR, hazard ratio.

^a) With consideration of 3-year latency period,

^b) Male: mild, < 15 g/day; moderate, 15-29.9 g/day; heavy, ≥ 30 g/day; female: mild, < 7.5 g/day; moderate, 7.5-14.9 g/day; heavy, ≥ 15 g/day,

^c) Cases per 100,000 person-year,

^d) Adjuster for age, sex, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia,

^e) Adjuster for age, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia.

Table 3.

Association between alcohol consumption behavior change and gastric cancer risk^a)

Alcohol consumption behavior change	Total			Male			Female
Alcohol consumption behavior change	No. of cases	Incidence rate^b)	HR (95% CI)^c)	No. of cases	Incidence rate^b)	HR (95% CI)^d)	No. of cases	Incidence rate^b)	HR (95% CI)^d)
Non-drinker	1,917	102.3	0.92 (0.85-1.00)	936	170.0	0.89 (0.81-0.97)	981	74.1	1.11 (0.84-1.45)
Quitter	645	144.9	1.05 (0.96-1.16)	538	181.5	1.05 (0.95-1.17)	107	72.0	1.19 (0.86-1.65)
Reducer	302	163.9	0.99 (0.87-1.12)	289	174.4	0.97 (0.86-1.11)	13	70.2	1.37 (0.75-2.51)
Sustainer	1,431	145.2	1.00	1,374	157.8	1.00	57	49.8	1.00
Increaser	788	159.3	1.11 (1.02-1.21)	700	196.3	1.12 (1.01-1.22)	88	63.7	1.14 (0.81-1.59)

CI, confidence interval; HR, hazard ratio.

^a) With consideration of 3-year of latency period,

^b) Cases per 100,000 person-year,

^c) Adjuster for age, sex, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia,

^d) Adjuster for age, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia.

REFERENCES

1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394–424. Article PubMed PDF
2. Lopez MJ, Carbajal J, Alfaro AL, Saravia LG, Zanabria D, Araujo JM, et al. Characteristics of gastric cancer around the world. Crit Rev Oncol Hematol. 2023;181:103841.Article PubMed
3. Kumar S, Metz DC, Ellenberg S, Kaplan DE, Goldberg DS. Risk factors and incidence of gastric cancer after detection of Helicobacter pylori infection: a large cohort study. Gastroenterology. 2020;158:527–36. Article PubMed PMC
4. Shin A, Park S, Shin HR, Park EH, Park SK, Oh JK, et al. Population attributable fraction of infection-related cancers in Korea. Ann Oncol. 2011;22:1435–42. Article PubMed
5. Tramacere I, Negri E, Pelucchi C, Bagnardi V, Rota M, Scotti L, et al. A meta-analysis on alcohol drinking and gastric cancer risk. Ann Oncol. 2012;23:28–36. Article PubMed
6. Wu H, Chen HL. The association between heavy alcohol use and gastric cancer. Am J Gastroenterol. 2021;116:2470–1. Article
7. Deng W, Jin L, Zhuo H, Vasiliou V, Zhang Y. Alcohol consumption and risk of stomach cancer: a meta-analysis. Chem Biol Interact. 2021;336:109365.Article PubMed
8. Mezey E, Sharma S, Rennie L, Potter JJ. Sex differences in gastric alcohol dehydrogenase activity in Sprague-Dawley rats. Gastroenterology. 1992;103:1804–10. Article PubMed
9. Kershenobich D, Haddad L, Marte LJ, Vargas F, de la Fuente JR, Zapata L. Alcohol metabolism in healthy subjects. Gastroenterology. 1993;105:308–9. Article PubMed
10. Kyoung DS, Kim HS. Understanding and Utilizing Claim Data from the Korean National Health Insurance Service (NHIS) and Health Insurance Review & Assessment (HIRA) Database for Research. J Lipid Atheroscler. 2022;11:103–10. Article PubMed PMC PDF
11. Kim MK, Han K, Lee SH. Current trends of big data research using the Korean National Health Information Database. Diabetes Metab J. 2022;46:552–63. Article PubMed PMC PDF
12. Seong SC, Kim YY, Park SK, Khang YH, Kim HC, Park JH, et al. Cohort profile: the National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) in Korea. BMJ Open. 2017;7:e016640Article PubMed PMC
13. Jin EH, Han K, Shin CM, Lee DH, Kang SJ, Lim JH, et al. Sex and tumor-site differences in the association of alcohol intake with the risk of early-onset colorectal cancer. J Clin Oncol. 2023;41:3816–25. Article PubMed PMC
14. Steevens J, Schouten LJ, Goldbohm RA, van den Brandt PA. Alcohol consumption, cigarette smoking and risk of subtypes of oesophageal and gastric cancer: a prospective cohort study. Gut. 2010;59:39–48. Article PubMed
15. Freedman ND, Abnet CC, Leitzmann MF, Mouw T, Subar AF, Hollenbeck AR, et al. A prospective study of tobacco, alcohol, and the risk of esophageal and gastric cancer subtypes. Am J Epidemiol. 2007;165:1424–33. Article PubMed
16. Lindblad M, Rodriguez LA, Lagergren J. Body mass, tobacco and alcohol and risk of esophageal, gastric cardia, and gastric non-cardia adenocarcinoma among men and women in a nested case-control study. Cancer Causes Control. 2005;16:285–94. Article PubMed PDF
17. Jin EH, Han K, Lee DH, Shin CM, Lim JH, Choi YJ, et al. Association between metabolic syndrome and the risk of colorectal cancer diagnosed before age 50 years according to tumor location. Gastroenterology. 2022;163:637–48. Article PubMed
18. Huang D, Shin WK, De la Torre K, Lee HW, Min S, Shin A, et al. Association between metabolic syndrome and gastric cancer risk: results from the Health Examinees Study. Gastric Cancer. 2023;26:481–92. Article PubMed PDF
19. Hwang YJ, Kim N, Yun CY, Yoon H, Shin CM, Park YS, et al. Validation of administrative big database for colorectal cancer searched by International Classification of Disease 10th Codes in Korean: a retrospective big-cohort study. J Cancer Prev. 2018;23:183–90. Article PubMed PMC
20. Hwang YJ, Park SM, Ahn S, Lee JC, Park YS, Kim N. Accuracy of an administrative database for pancreatic cancer by international classification of disease 10(th) codes: a retrospective large-cohort study. World J Gastroenterol. 2019;25:5619–29. Article PubMed PMC
21. Hwang YJ, Park SM, Ahn S, Lee J, Park YS, Kim N. Diagnostic accuracy of administrative database for bile duct cancer by ICD-10 code in a tertiary institute in Korea. Hepatobiliary Pancreat Dis Int. 2020;19:575–80. Article PubMed
22. Choi YJ, Lee DH, Han KD, Kim HS, Yoon H, Shin CM, et al. The relationship between drinking alcohol and esophageal, gastric or colorectal cancer: a nationwide population-based cohort study of South Korea. PLoS One. 2017;12:e0185778Article PubMed PMC
23. Lee HW, Huang D, Shin WK, de la Torre K, Song M, Shin A, et al. Frequent low dose alcohol intake increases gastric cancer risk: the Health Examinees-Gem (HEXA-G) study. Cancer Biol Med. 2022;19:1224–34. Article PubMed PMC
24. Shin CM, Kim N, Lee HS, Lee DH, Kim JS, Jung HC, et al. Intrafamilial aggregation of gastric cancer: a comprehensive approach including environmental factors, Helicobacter pylori virulence, and genetic susceptibility. Eur J Gastroenterol Hepatol. 2011;23:411–7. PubMed
25. Lewis SJ, Smith GD. Alcohol, ALDH2, and esophageal cancer: a meta-analysis which illustrates the potentials and limitations of a Mendelian randomization approach. Cancer Epidemiol Biomarkers Prev. 2005;14:1967–71. Article PubMed PDF
26. Yokoyama A, Muramatsu T, Ohmori T, Yokoyama T, Okuyama K, Takahashi H, et al. Alcohol-related cancers and aldehyde dehydrogenase-2 in Japanese alcoholics. Carcinogenesis. 1998;19:1383–7. Article PubMed
27. Bizzaro D, Becchetti C, Trapani S, Lavezzo B, Zanetto A, D’Arcangelo F, et al. Influence of sex in alcohol-related liver disease: pre-clinical and clinical settings. United European Gastroenterol J. 2023;11:218–27. Article PubMed PMC PDF
28. Baraona E, Abittan CS, Dohmen K, Moretti M, Pozzato G, Chayes ZW, et al. Gender differences in pharmacokinetics of alcohol. Alcohol Clin Exp Res. 2001;25:502–7. Article PubMed
29. Kwo PY, Ramchandani VA, O’Connor S, Amann D, Carr LG, Sandrasegaran K, et al. Gender differences in alcohol metabolism: relationship to liver volume and effect of adjusting for body mass. Gastroenterology. 1998;115:1552–7. Article PubMed
30. Yoo JE, Han K, Shin DW, Kim D, Kim BS, Chun S, et al. Association between changes in alcohol consumption and cancer risk. JAMA Netw Open. 2022;5:e2228544Article PubMed PMC
31. He Z, Zhao TT, Xu HM, Wang ZN, Xu YY, Song YX, et al. Association between alcohol consumption and the risk of gastric cancer: a meta-analysis of prospective cohort studies. Oncotarget. 2017;8:84459–72. Article PubMed PMC
32. Li Y, Eshak ES, Shirai K, Liu K, Dong JY, Iso H, et al. Alcohol consumption and risk of gastric cancer: the Japan Collaborative Cohort Study. J Epidemiol. 2021;31:30–6. Article PubMed PMC
33. Bae JM. Sex as an effect modifier in the association between alcohol intake and gastric cancer risk. World J Gastrointest Oncol. 2021;13:453–61. Article PubMed PMC
34. Erol A, Ho AM, Winham SJ, Karpyak VM. Sex hormones in alcohol consumption: a systematic review of evidence. Addict Biol. 2019;24:157–69. Article PubMed PMC PDF
35. Wilsnack SC, Wilsnack RW. International gender and alcohol research: recent findings and future directions. Alcohol Res Health. 2002;26:245–50. PubMed PMC
36. Walker S, Higgs S, Terry P. Estimates of the absolute and relative strengths of diverse alcoholic drinks by young people. Subst Use Misuse. 2016;51:1781–9. Article PubMed
37. Livingston M, Callinan S. Underreporting in alcohol surveys: whose drinking is underestimated? J Stud Alcohol Drugs. 2015;76:158–64. Article PubMed
38. Jo HH, Kim N, Jang J, Choi Y, Park J, Park YM, et al. Impact of body mass index on survival depending on sex in 14,688 patients with gastric cancer in a tertiary hospital in South Korea. Gut Liver. 2023;17:243–58. Article PubMed PMC
39. Kang S, Park M, Cho JY, Ahn SJ, Yoon C, Kim SG, et al. Tumorigenic mechanisms of estrogen and Helicobacter pylori cytotoxin-associated gene A in estrogen receptor alpha-positive diffuse-type gastric adenocarcinoma. Gastric Cancer. 2022;25:678–96. Article PubMed PDF
40. Salathe M, Vu DQ, Khandelwal S, Hunter DR. The dynamics of health behavior sentiments on a large online social network. EPJ Data Sci. 2013;2:4.
41. Sheeran P, Wright CE, Avishai A, Villegas ME, Rothman AJ, Klein WMP. Does increasing autonomous motivation or perceived competence lead to health behavior change? A meta-analysis. Health Psychol. 2021;40:706–16. Article PubMed
42. Newsom JT, Huguet N, McCarthy MJ, Ramage-Morin P, Kaplan MS, Bernier J, et al. Health behavior change following chronic illness in middle and later life. J Gerontol B Psychol Sci Soc Sci. 2012;67:279–88. Article PubMed PMC
43. Bui TT, Han M, Luu NM, Tran TPT, Lim MK, Oh JK. Cancer Risk According to alcohol consumption trajectories: a population-based cohort study of 2.8 million Korean men. J Epidemiol. 2023;33:624–32. Article PubMed PMC

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

The Effect of Alcohol Consumption Behavior Changes on Gastric Cancer Risks Stratified by Sex in South Korea

Cancer Res Treat. 2026;58(1):232-241. Published online April 1, 2025

DOI: https://doi.org/10.4143/crt.2024.591

XML Download

Abstract
Introduction
Materials and Methods
Results
Discussion
Electronic Supplementary Material
NOTES
REFERENCES

The Effect of Alcohol Consumption Behavior Changes on Gastric Cancer Risks Stratified by Sex in South Korea

Fig. 1. Study scheme. (A) Flowchart of study subject selection. (B) Time points to measure alcohol consumption behavior change and to start follow-up. The orange box indicates the period to measure alcohol consumption frequency and dose in 2002. For subjects whose alcohol consumption behavior was measured in 2002, their alcohol consumption behaviors after change were measured during 2005-2006 (red box). For subjects whose alcohol consumption behavior before change was measured in 2003, the measurement periods before and after the change were defined as the green and blue boxes. The follow-up was designated to begin at the time of the last alcohol consumption behavior (red arrow line). NHIS-HEALS, National Health Insurance Service-National Health Screening Cohort.

Fig. 2. Gastric cancer risk depending on alcohol consumption behavior change and sex. (A) Gastric cancer risk according to alcohol consumption behavior change between 2002-2003 and 2005-2007 among males. (B) Gastric cancer risk according to alcohol consumption behavior change between 2002-2003 and 2005-2007 among females. CI, confidence interval; HR, hazard ratio. An asterisk (*) indicates statistical significance.

Fig. 1.

Fig. 2.

The Effect of Alcohol Consumption Behavior Changes on Gastric Cancer Risks Stratified by Sex in South Korea

Characteristic	Non-drinker (n=147,022)	Quitter (n=34,928)	Reducer (n=14,480)	Sustainer (n=75,209)	Increaser (n=38,553)	p-value
Baseline alcohol consumption level
Non-drinker	147,022 (100)	0	0	0	23,830 (61.8)	< 0.01
Mild	0	28,739 (82.3)	0	64,484 (85.7)	11,496 (29.8)
Moderate	0	3,419 (9.8)	7,674 (53.0)	4,941 (6.6)	3,227 (8.4)
Heavy	0	2,770 (7.9)	6,806 (47.0)	5,784 (7.7)	0
Age (yr)	57.5±9.5	55.5±9.2	54.5±8.6	52.6±7.8	54.2±8.8	< 0.01
Sex
Male	43,685 (29.7)	23,428 (67.1)	13,054 (90.1)	66,466 (88.4)	27,928 (72.4)	< 0.01
Female	103,337 (70.3)	11,500 (32.9)	1,426 (9.9)	8,743 (11.6)	10,625 (27.6)
Smoking status
Non-smoker	128,245 (87.2)	28,185 (80.7)	5,456 (37.7)	31,848 (42.4)	19,629 (50.9)	< 0.01
Former smoker	4,752 (3.2)	1,970 (5.6)	2,115 (14.6)	12,671 (16.8)	4,811 (12.5)
Current smoker	10,271 (7.0)	3,549 (10.2)	5,800 (40.0)	25,609 (34.0)	11,521 (30.0)
Missing	3,754 (2.6)	1,224 (3.5)	1,109 (7.7)	5,081 (6.8)	2,592 (6.7)
Income
Q1	22,810 (15.5)	4,868 (13.9)	1,824 (12.6)	7,668 (10.2)	5,055 (13.1)	< 0.01
Q2	22,464 (15.3)	4,835 (13.8)	1,899 (13.1)	7,636 (10.2)	5,382 (14.0)
Q3	24,123 (16.4)	5,628 (16.1)	2,549 (17.6)	10,438 (13.9)	6,413 (16.6)
Q4	29,402 (20.0)	7,090 (20.3)	3,217 (22.2)	15,858 (21.1)	8,050 (20.9)
Q5	47,802 (32.5)	12,425 (35.6)	4,966 (34.3)	33,544 (44.6)	13,582 (35.2)
Missing	421 (0.3)	82 (0.2)	25 (0.2)	65 (0.1)	71 (0.2)
BMI (kg/m²)	23.9±3.0	24.0±2.9	24.2±2.8	24.1±2.7	24.0±2.8	< 0.01
Physical activity/wk
0	81,186 (55.2)	20,939 (60.0)	6,071 (41.9)	25,708 (34.2)	16,874 (43.8)	< 0.01
1-2	32,021 (21.8)	7,166 (20.5)	4,717 (32.6)	29,057 (38.6)	11,539 (29.9)
≥ 3	31,648 (21.5)	6,145 (17.6)	3,463 (23.9)	19,607 (26.1)	9,622 (25.0)
Missing	2,167 (1.5)	678 (1.9)	229 (1.6)	837 (1.1)	518 (1.3)
Fasting glucose, mean±SD	96.9±25.4	98.8±27.9	102.7±31.7	99.4±26.3	99.7±27.4	< 0.01
Total cholesterol, mean±SD	200.3±37.3	197.7±37.1	197.6±37.0	197.4±35.6	197.6±36.6	< 0.01
SBP, mean±SD	125.1±16.7	125.9±16.4	129.0±16.4	126.8±15.7	127.0±16.4	< 0.01
DBP, mean±SD	77.3±10.5	78.6±10.6	81.0±10.7	79.9±10.5	79.5±10.7	< 0.01
DM past history
No	127,477 (86.7)	30,567 (87.5)	12,677 (87.5)	67,379 (89.6)	34,172 (88.6)	< 0.01
Yes	19,545 (13.3)	4,361 (12.5)	1,803 (12.5)	7,830 (10.4)	4,381 (11.4)
HTN past history
No	110,158 (74.9)	26,802 (76.7)	11,106 (76.7)	59,516 (79.1)	30,355 (78.7)	< 0.01
Yes	36,864 (25.1)	8,126 (23.3)	3,374 (23.3)	15,693 (20.9)	8,198 (21.3)
Dyslipidemia past history
No	124,632 (84.8)	30,219 (86.5)	12,653 (87.4)	66,248 (88.1)	33,851 (87.8)	< 0.01
Yes	22,390 (15.2)	4,709 (13.5)	1,827 (12.6)	8,961 (11.9)	4,702 (12.2)

Alcohol dose in 2002-2003^b)	Alcohol dose in 2005-2007^b)	Total			Male			Female
Alcohol dose in 2002-2003^b)	Alcohol dose in 2005-2007^b)	No. of cases	Incidence rate^c)	HR (95% CI)^d)	No. of cases	Incidence rate^c)	HR (95% CI)^e)	No. of cases	Incidence rate^c)	HR (95% CI)^e)
Non-drinker	Non-drinker	1,917	102.3	1.00	936	170.0	1.00	981	74.1	1.00
	Mild	347	133.1	1.14 (1.01-1.29)	279	177.3	1.17 (1.02-1.34)	68	65.9	1.05 (0.82-1.35)
	Moderate	42	154.4	1.08 (0.80-1.48)	39	210.2	1.16 (0.84-1.60)	3	34.7	0.58 (0.19-1.81)
	Heavy	38	197.4	1.37 (0.99-1.90)	30	220.3	1.27 (0.88-1.83)	8	142.2	1.97 (0.98-3.96)
Mild	Non-drinker	511	138.5	1.07 (0.96-1.20)	418	175.4	1.05 (0.93-1.18)	93	71.3	1.28 (0.89-1.83)
	Mild	1,171	137.6	1.00	1,121	149.9	1.00	50	48.5	1.00
	Moderate	175	178.6	1.15 (0.98-1.35)	170	202.8	1.17 (0.99-1.37)	5	35.4	0.76 (0.30-1.91)
	Heavy	92	185.9	1.17 (0.95-1.45)	91	199.7	1.19 (0.96-1.47)	1	25.5	0.47 (0.06-3.43)
Moderate	Non-drinker	76	180.0	0.93 (0.68-1.26)	67	208.2	0.90 (0.65-1.24)	9	89.7	1.51 (0.43-5.21)
	Mild	153	155.7	0.80 (0.63-1.02)	143	166.0	0.78 (0.61-0.99)	10	82.4	1.44 (0.45-4.64)
	Moderate	125	200.5	1.00	121	220.1	1.00	4	54.3	1.00
	Heavy	94	234.1	1.13 (0.86-1.47)	91	242.2	1.10 (0.84-1.44)	3	116.4	2.00 (0.44-9.09)
Heavy	Non-drinker	58	170.8	1.04 (0.74-1.46)	53	204.2	1.08 (0.76-1.52)	5	62.5	0.42 (0.10-1.86)
	Mild	79	160.1	0.87 (0.66-1.14)	77	167.7	0.87 (0.65-1.15)	2	58.8	0.58 (0.09-3.60)
	Moderate	70	191.0	1.00 (0.75-1.34)	69	205.0	1.02 (0.76-1.36)	1	33.5	0.43 (0.04-4.17)
	Heavy	135	186.7	1.00	132	193.4	1.00	3	74.0	1.00

Alcohol consumption behavior change	Total			Male			Female
Alcohol consumption behavior change	No. of cases	Incidence rate^b)	HR (95% CI)^c)	No. of cases	Incidence rate^b)	HR (95% CI)^d)	No. of cases	Incidence rate^b)	HR (95% CI)^d)
Non-drinker	1,917	102.3	0.92 (0.85-1.00)	936	170.0	0.89 (0.81-0.97)	981	74.1	1.11 (0.84-1.45)
Quitter	645	144.9	1.05 (0.96-1.16)	538	181.5	1.05 (0.95-1.17)	107	72.0	1.19 (0.86-1.65)
Reducer	302	163.9	0.99 (0.87-1.12)	289	174.4	0.97 (0.86-1.11)	13	70.2	1.37 (0.75-2.51)
Sustainer	1,431	145.2	1.00	1,374	157.8	1.00	57	49.8	1.00
Increaser	788	159.3	1.11 (1.02-1.21)	700	196.3	1.12 (1.01-1.22)	88	63.7	1.14 (0.81-1.59)

Table 1. Baseline characteristics of study subjects according to alcohol consumption behavior changea)

With consideration of 3-year latency period.

Table 2. Risk of gastric cancer incidence according to alcohol consumption dose changea)

CI, confidence interval; HR, hazard ratio.

With consideration of 3-year latency period,

Male: mild, < 15 g/day; moderate, 15-29.9 g/day; heavy, ≥ 30 g/day; female: mild, < 7.5 g/day; moderate, 7.5-14.9 g/day; heavy, ≥ 15 g/day,

Cases per 100,000 person-year,

Adjuster for age, sex, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia,

Adjuster for age, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia.

Table 3. Association between alcohol consumption behavior change and gastric cancer riska)

CI, confidence interval; HR, hazard ratio.

With consideration of 3-year of latency period,

Cases per 100,000 person-year,

Adjuster for age, sex, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia,

Adjuster for age, smoking status, physical activity, body mass index, past history of diabetes mellitus, hypertension, and dyslipidemia.