Alcohol is classified as an intoxicant and class 1 carcinogen. In 2023, the monthly drinking rate in South Korea was 55.1% [1]. Monthly drinking, high-risk drinking, and monthly binge drinking rates in South Korea in 2023 were all higher than those in 2021 [1]. The college period, which marks the transition from late adolescence to adulthood, is often characterized by independence from parents, gaining autonomy, and the development of a wide range of interpersonal relationships [2]. In early adulthood, people are likely to engage in high-risk drinking experiences, such as binge drinking, due to participation in various school events, such as back-to-school Orientation and Membership training, as well as relaxed drinking regulations and frequent drinking [3]. Their likelihood of engaging in heavy drinking increases through participation in these gatherings [3].

College drinking culture manifests itself in the process of establishing social relationships among college students, and some cultures make it difficult to refuse alcohol offered by others [4]. Therefore, cultural factors related to drinking are expected to have a significant effect on interpersonal factors in the formation of social relationships.

Interpersonal factors are related to stress and resilience and have been reported to influence problem drinking. In the case of college students, they sometimes drink for social support, but they also drink due to peer pressure. For instance, having strong interpersonal relationships is associated with higher resilience [3,5,6]. Resilience determines how individuals react to and cope with adversity, and high resilience is associated with reduced problem drinking [7].

Problem drinking is drinking that causes significant impairment in areas such as health, work, or adjustment [5]. Previous studies have shown that a wide range of health problems arise from problem drinking [5]. Especially academic stress, daily stress, interpersonal relationships, and a culture that accepts drinking affect problem drinking among college students [3,5]. In addition, gender, age, religious affiliation, smoking status, and family circumstances are significant predictors of problem drinking among college students [8]. In contrast, resilience, healthy interpersonal relationships, and a healthy drinking culture reduce problem drinking among college students [4,6,7].

However, previous studies examining the relationships only between stress and resilience, which influence drinking culture in the university context, remain scarce. In particular, few studies have comprehensively investigated the associations among general characteristics, stress, resilience, interpersonal relationships, and drinking culture as they relate to problem drinking among college students. Therefore, a multivariate approach is necessary to clarify the independent and hierarchical effects of these factors on problem drinking among college students, thereby better understanding the determinants of problem drinking in this population.

This study aimed to assess the levels of perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking; identify the levels of perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking based on demographic characteristics; examine the relationships among problem drinking, perceived stress, resilience, interpersonal relationships, and drinking culture factors; and explore the demographic and social factors affecting problem drinking among college students. This study sought to provide a foundation for developing healthy drinking habits and drinking culture, thereby preventing problem drinking behavior among college students.

An observational, monocentric, cross-sectional study was conducted among college students.

Data were collected from October 3, 2022, to October 16, 2022, using a self-report questionnaire. To prevent the spread of COVID-19, the survey was conducted online. To recruit participants, online questionnaires were distributed among the university community, local communities, and social media platforms.

The inclusion and exclusion criteria were as follows:

ㆍ Inclusion criteria: College students with drinking experience enrolled at a university in Seoul, South Korea.

ㆍ Exclusion criteria: College students who were currently being treated for a psychiatric disorder other than alcohol use disorder.

The sample size was calculated using G*Power 3.1.9.7. Based on a previous study [8], an F test was applied with a significance level of .05, a statistical power of .95, a medium effect size of .15 for multiple regression, and 19 predictors. The minimum target sample size was 217. Considering the dropout rate, a total of 226 questionnaires were collected. After excluding 6 questionnaires with insufficient responses, 220 questionnaires were finally used for the final analysis.

The variables of this study included demographic characteristics, perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking [3-8].

This study was approved by the Institutional Review Board (IRB No-2022-0301-01). Participation was voluntary, and informed consent was obtained online from all participants prior to participation. In addition, participants were informed that they could withdraw at any time during the study, their anonymity would be guaranteed, and the results of the study would not be used for any other purpose.

The questionnaire was administered anonymously to mitigate social desirability bias and ensure that participants could not be identified from their responses. In addition, participants received ethical instructions that emphasized the importance of completing the questionnaire independently, without consulting external sources such as online materials or peers. They were also informed that participation was entirely voluntary and that choosing not to participate or to withdraw at any time would have no adverse academic consequences for them.

Data were analyzed using SPSS 27.0. The frequency, percentage, mean, and standard deviation of demographic characteristics, perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking were analyzed. Cronbach's ⍺ was analyzed to verify the reliability of the perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking scales. Differences in perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking based on the participants' demographic characteristics were analyzed using t-tests and an analysis of variance. Post-hoc tests were conducted using Scheffé multiple comparison test. Pearson correlation analysis was conducted to confirm the correlations between perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking. A hierarchical regression analysis was conducted to identify the factors affecting problem drinking. The control variables were military service status (service, non-service, exemption, or not applicable), type of residence (live with family, live alone, dormitory, or other), student council activities, age at first drinking experience (middle school or elementary school, high school, or college), and smoking status.

The participants' demographic characteristics are presented in Table 1. Of the participants, 102 were men (46.4%) and 118 were women (53.6%). Regarding military service status, 20.5% had completed service, 24.1% had not served, 1.4% were exempt, and 54.1% were not applicable. Freshmen and seniors each accounted for the largest proportion (29.5%). With respect to major, engineering and information technology were the most common (38.2%), followed by health, medicine, and nursing (32.3%). For the type of residence, living alone (45.5%) and living with family (39.1%) were the most frequent. Most participants reported no religion (76.8%). The majority participated in student council activities (80.0%), and high academic performance was reported by 56.8%. For the average monthly allowance, 40.9% received less than 300,000 won, and 38.6% received 300,000~600,000 won. About half of the participants did not have a part-time job (50.5%). Most participants reported their first drinking experience during college (57.3%); 22.3% were smokers, and 21.4% had experienced family problems caused by drinking. Most reported middle (40.0%) or upper-middle (36.4%) family status.

Problem drinking differed significantly according to military service status (F=4.26, p=.006), type of residence (F=3.69, p=.013), student council activities (t=2.79, p=.006), age at first drinking experience (F=6.39, p<.001), and smoking status (t=4.66, p<.001). Higher levels of problem drinking were observed among non-military participants, those living alone or in dormitories, members of student organizations, those who first drank in middle school, and smokers, respectively.

The participants' perceived stress, resilience, interpersonal relationships, drinking culture, and problem drinking exhibited univariate normality, and skewness and kurtosis were within acceptable ranges (Table 2). Problem drinking was significantly and positively correlated with interpersonal relationships (r=.20, p<.01) and drinking culture (r=.61, p<.01), whereas perceived stress and resilience were not significantly correlated with problem drinking (Table 3).

Drinking culture (F=6.76, p=.001) and problem drinking (F=7.52, p=.001) differed significantly according to the age at first drinking experience, with higher levels observed among those who started drinking in middle school compared with those who first drank in college, whereas interpersonal relationships did not differ significantly by age at first drinking experience (Table 4).

In the hierarchical regression analysis, the variance inflation factor (VIF) ranged from 1.06 to 3.11, indicating no multicollinearity, and the Durbin-Watson value was 1.72, confirming the independence of residuals. The final model explained 43.8% of the variance in problem drinking (Adjusted R²=43.8%) (Table 5).

The results revealed that the level of problem drinking among the participants varied based on their military service status, living arrangements, student council activities, age at first drinking experience, and smoking status. Better interpersonal relationships and higher levels of drinking culture were associated with higher levels of problem drinking. Living with family, smoking status, interpersonal relationships, and drinking culture had statistically significant explanatory power with respect to problem drinking. Among these variables, drinking culture had the highest explanatory power, followed by smoking status, living with family, and interpersonal relationships. Thus, permissive drinking culture, smoking, and good interpersonal relationships were associated with problem drinking. Furthermore, living with family was associated with less problem drinking than living alone.

In this study, the frequency of problem drinking was higher among non-military students than among military students, and students who lived independently showed higher levels of problem drinking than those who lived with family. Another factor associated with problem drinking was early initiation into alcohol use, and earlier drinking onset was associated with higher levels of problem drinking in this study. Moreover, in Korean culture, smoking and drinking are often viewed as essential methods for interpersonal relationships [3,8]; therefore, engaging in both behaviors simultaneously can be observed in South Korea.

College students are likely to be exposed to alcohol and control their drinking, and the incidence of problem drinking associated with interpersonal relationships among college students can increase as festivals, school events, and other gatherings resume [2,3,8]. Our findings showed that drinking culture positively affected problem drinking among the participants. South Korea has a tolerant culture [6], and college students consider drinking to be an essential part of their social life in college [5]. Moreover, college students face challenges in finding alternatives to drinking at events, such as college festivals [6]. This permissive attitude toward drinking increases the likelihood of exposure to problem drinking, such as excessive drinking, associated with the drinking culture at events.

The results regarding differences in problem drinking based on military service status, living arrangements, student council activities, age at first drinking experience, and smoking status were similar to those of Jeong [14], who examined the factors influencing problem drinking among first- and fourth-year college students. However, these findings differed from those of Kim [8], who found significant differences in problem drinking by gender. The absence of gender differences in problematic drinking may be explained by the recent trend of increased alcohol consumption among female students, resulting in a growing number of female problematic drinkers. In addition, Jeong [14] identified differences in problem drinking based on grade, major, and family situation. By contrast, this study found no significant differences in the level of problematic drinking according to grade, major, or family economic status. This may be attributed to the fact that university students are exposed equally to drinking opportunities through student activities, regardless of their year, field of study, or financial background, thereby diminishing the influence of these variables. Conversely, this study demonstrated that younger age at first drinking experience, was associated with higher levels of problem drinking, which was in line with Jeong's findings that revealed differences in the incidence of problem drinking based on the age at first drinking [14].

The results that interpersonal relationships and drinking culture were positively correlated with problem drinking were in line with those of Kim [8], who found a significant positive correlation between drinking culture and problem drinking among college students. Furthermore, the drinking culture score in this study was 2.57 (out of 5), which was comparable to the average value of 23.52 (out of 42) reported in Kim's [12] study on college drinking culture.

This study has several limitations. First, technical challenges such as outdated computers or poor internet connections occasionally interfered with interaction in online surveys. Second, the study was conducted at a single college, limiting the generalizability of findings. Third, the Perceived Stress Scale's internal consistency was low, a generally accepted criterion, which may have affected measurement accuracy. Fourth, because few studies have comprehensively analyzed the variables of this study, there are limitations to interpreting the results.

This study was performed using sampling conveniently, generalizability of the findings is limited.

Measures should be taken to introduce education programs that inculcate healthy drinking habits in students to counteract the permissive and problematic drinking culture. As interpersonal relationships and drinking culture affect each other [4], and the first year is a time when new social groups of various sizes are formed [15], the incidence of problem drinking among first-year college students is expected to increase problem drinking in the future. Therefore, various drinking management systems, such as providing on-campus sobriety clubs that inform college students of problem drinking, developing a healthy drinking culture, and monitoring and managing students struggling with drinking habits, should be established and offered to students early in their college career. Specific recommendations derived from the results of this study include developing tailored drinking-reduction programs for students living alone or in dormitories to mitigate drinking-related problems among college students, providing integrated educational initiatives that emphasize the combined risks of smoking and problematic alcohol use, and promoting alternative social activities that foster meaningful relationships without the need for alcohol consumption.