Harmful alcohol use is associated with unintentional and intentional injuries, liver diseases, cardiovascular diseases, and cancer, as well as with social problems (i.e., violent behaviors, fatal accidents) [1–3]. According to a World Health Organization (WHO) report, 5.1% of the global burden of disease and injury is due to alcohol consumption, and 5.9% of all deaths worldwide were attributable to alcohol consumption in 2012 [4].

Koreans drink more alcohol than people of other nationalities and Koreans aged 15 years and older consume on average 11.9 L of pure alcohol per year compared with a global prevalence of 6.4 L in 2016 [4]. In addition, high-risk drinking is more prevalent in Korea (15.1%) compared with other countries [5]. In 2010, the alcohol-dependence rate in Korea was 4.7%, whereas the average for the Western Pacific Region was 2.3% according to the WHO [4].

There are significant gender differences in the prevalence of harmful alcohol use and the proportion of global deaths associated with alcohol. It was reported in 2012 that 7.6% of deaths in men and 4.0% in women were attributable to alcohol [4]. Alcohol consumption levels differ by gender [6,7]. It has been reported that between 2012 and 2014 men drank more than women among non-Hispanic White, Mexican, other Latino, Asian, and other non-Hispanic participants [8]. Major gender differences are present in multiple alcohol-related outcomes such as alcohol consumption amount, age of beginning drinking, and alcohol-related death rates [9]. However, a study among the Hispanic population reported no gender differences in alcohol consumption during the past month (25.2% of women and 23.8% of men), nor in heavy drinking during the past 2 weeks (19.0% of women and 19.9% of men) [10].

In a study, male adolescents’ drinking was associated with a higher level of social and enhancement motives, whereas that of female adolescents was associated with a higher level of coping motives and a lower level of conformity motives [11]. It was reported that male Korean college students drank to cope with stress, whereas females drank for enhancement motives however, both motives led to alcohol-related problems [12]. Moreover, gender differences in alcohol consumption are reportedly due to cultural differences and different biological and cultural needs [13].

According to findings reported by the WHO, the ratio of male to female heavy drinkers in South-East Asia was 10 to 9, and in the Western Pacific was 7 to 3, respectively [4]. In Korea, the prevalence of harmful alcohol use was 5-fold higher in men than women (2.5% in men and 0.5% in women) [4] and the rate of high-risk drinking in men was 23.7%, and in women was 4.2% [5]. The alcohol-dependence rate in Korea has been reported to be 7.8% in men and 1.7% in women [4]. However, few studies have investigated the factors associated with these gender differences. Therefore, in this current study gender differences in harmful alcohol use were evaluated and the factors associated with this phenomenon in Korea were examined.

In terms of age, 57.2% of men and 54.7% of women were 20–44 years, and 33.2% of men and 35.8% of women had a low or low–intermediate household income. Of the men and women, 85.5% and 78.7% respectively, had a high school education or above. Additionally, 19% and 44.4% of the men and women respectively were unemployed, 30.7% and 19.5% respectively, were unmarried, and 37.9% and 29.9% respectively, perceived their health to be good. Moreover, 43.4% and 6.2% of the men and women respectively, were currently smoking, and 41.1% and 37.7% respectively, reported walking daily for at least 30 minutes. Among the men and women, 26.6% and 28.9% respectively, reported being very stressed, and 8.3% and 15.1% respectively, stated that they felt depressed. There were significant differences between men and women in age (χ² = 8.40, p = 0.006), educational level (χ² = 110.75, p < 0.001), occupation (χ² = 1,323.81, p < 0.001), marital status (χ² = 285.26, p < 0.001), perceived health status (χ² = 113.75, p < 0.001), currently smoking (χ² = 2,587.58, p < 0.001), walking (χ² = 15.67, p = 0.001), stress (χ² = 9.06, p = 0.014), and depressive feeling (χ² = 136.10, p < 0.001; Table 1).

The prevalence of harmful alcohol use (AUDIT-K score ≥16) was 10.7% in Korean adults aged 20–64 years (18.4% in men and 3.4% in women). Among men, harmful alcohol use was more prevalent in those aged 45–64 years (21.5%) compared with 20–44 years (16.1%). However, among women harmful alcohol use was more prevalent in those aged 20–44 years (4.8%) compared with 45–64 years (1.6%). Harmful alcohol use was more prevalent in men with a high household income than in those with a low household income (20.4% and 14.8% respectively) and in those with an elementary-school education (23.5%). In contrast, in women harmful alcohol use was the most prevalent in high-school graduates (4.6%). In terms of occupation, manual laborers showed the highest rate of harmful alcohol use among men (22.2%), and for women it was sales/service workers (6.0%). A higher prevalence of harmful alcohol use was identified among men who were separated or divorced (26.0%), whereas this was true for unmarried women (6.8%). The frequency of harmful alcohol use was higher among those with a moderate health status, than among those who had a good health status among both men and women (19.1% of men and 2.9% of women). The frequency of harmful alcohol use in both men and women was higher among the currently smoking group, than among the non-smoking group (25.2% of men and 17.3% of women). Among men, the prevalence of harmful alcohol use was higher in those who did not walk for more > 30 minutes daily (19.7%) compared with those who did (16.7%). The prevalence of harmful alcohol use was higher in those who reported being stressed (25.4% in men and 6.3% in women) or experienced depressive feeling (29.8% in men and 10.2% in women; Table 2).

The prevalence of harmful alcohol use in both genders differed significantly by age, educational level, occupation, marital status, perceived health status, currently smoking, perceived stress, and the presence/absence of depressive feeling. Additionally, household income was associated with harmful alcohol use in men (Table 2).

The factors associated with harmful alcohol use differed by gender. In men, high household income, elementary school education, men who were divorced or separated, manual laborers, moderate health status, currently smoking, perceived stress, and depressive feeling were significantly associated with harmful alcohol use. In women, education to high school graduate level, sales/service workers, unmarried women, currently smoking, a moderate health status, perceived stress, and depression status were significantly associated with harmful alcohol use. The odds ratio (OR) for harmful alcohol use was highest for the currently smoking group in both men and women. The impact of currently smoking on harmful alcohol use was greater in women than in men. In men, high ORs indicated that household income, perceived stress, and depressive feeling were risk factors for harmful alcohol consumption, whereas depressive feeling, marital status, and educational level were considered as the greatest risks in women (Table 3).

A higher prevalence of harmful alcohol use was observed in men with a high household income, than in those with a low household income (OR = 1.69, p < 0.001). A higher prevalence of harmful alcohol use among men was observed with an elementary-school education compared with college graduates (OR = 1.54, p = 0.006), high-school (OR = 1.47, p < 0.001) and middle-school (OR = 1.44, p = 0.023) education. Compared with married men, a significantly lower prevalence of harmful alcohol use was observed in unmarried men (OR = 0.61, p < 0.001). There was a significantly lower prevalence of harmful alcohol use among unemployed men, than among those with professional or managerial occupations (OR = 0.65, p = 0.009). There was also a higher prevalence of harmful alcohol use among men who perceived their health to be poor, than among those who perceived their health to be good (OR = 1.46, p = 0.005). A higher prevalence of harmful alcohol use was observed in those who currently smoke, than in non-smokers (OR = 1.99, p < 0.001). Harmful alcohol use was more prevalent among men who felt stressed, than among those who did not (OR = 1.59, p < 0.001). Additionally, there was a higher prevalence of harmful alcohol use among men who had felt depressed for 2 weeks, than among men who did not feel depressed (OR = 1.53, p = 0.001; Table 3).

A significantly higher prevalence of harmful alcohol use was identified in women with a high-school education, than in those with a college education (OR = 1.78, p = 0.001). Unlike the case for men, a higher prevalence of harmful alcohol use was observed among unmarried women, than among married women (OR = 2.6, p < 0.001). There was also a significantly higher prevalence of harmful alcohol use among females currently smoking, than among female non-smokers (OR = 5.37, p < 0.001). This was also the case for women who felt stressed compared with those who did not (OR = 1.63, p < 0.001), and for women who felt depressed over 2 weeks compared with those who did not (OR = 2.9, p < 0.001; Table 3).

In this study, currently smoking was strongly associated with harmful alcohol use in both genders. Similarly, smokers had a 2-fold higher frequency of heavy drinking, than non-smokers among men, and a 3-fold higher frequency of heavy drinking among women [23]. It was also reported that among women, the prevalence of harmful alcohol use was 4.93-fold higher in smokers, than in non-smokers, whereas the prevalence among men was 3.57-fold higher in smokers, than in non-smokers [24]. The prevalence of smoking in men was 5 times higher than women due to the lower social acceptability of women’s smoking in Korea [25]. This may imply that female smokers may have relatively higher stress or depressive feelings that lead to smoking, than male smokers. Considering that smoking is statistically positively associated with drinking [23,26] female smokers may have a greater inclination toward harmful alcohol use than male smokers. There was a higher prevalence of heavy drinking and problem alcohol use among ex-smokers of both genders [23,24]. There was also a high prevalence of smoking among individuals engaged in harmful alcohol use [26]. Thus, harmful alcohol use and smoking are interrelated. Therefore, smoking cessation interventions should be part of the intervention strategies to ameliorate problematic alcohol consumption, particularly for women.

Depressive feelings were associated with harmful alcohol use by both genders in this study and this is consistent with several previous reports [27–29]. Moreover, there was a 2-fold higher prevalence of depressive feelings among Korean women (15.1%), than among Korean men (8.3%). A diagnostic assessment of depression was not included in this study but this finding indicates that a subjectively experienced depressive mood is related to harmful alcohol use. The effect of depressive feelings on harmful alcohol use was greater in women than in men. This finding is similar to a previous study reporting that women with depression had a higher rate of harmful drinking than men with depression [30]. This indicates that women with depression may be more dependent on alcohol than men with depression but, there is no clear cause of the gender difference. However, Korean women with depressive feelings have a lack of resources to relieve stress compared with Korean men in society (a robust study is necessary to investigate the possible causes of the gender differences).

Strategies to prevent or manage harmful alcohol use should consider the possibility of depression. People with alcohol dependence were reported to be at a higher risk of major depressive episodes [31]. Therefore, depressive feelings and harmful alcohol use are interrelated. However, the cross-sectional design of this study prevented causal inferences being drawn.

A high level of stress increased the risk of harmful alcohol use in both men and women (OR = 1.59 and 1.63, respectively). However, it was reported that stress levels did not significantly increase the risk of harmful alcohol use in women or men [24]. In this study harmful alcohol use was met when there was an AUDIT-K score of ≥ 16, whereas an AUDIT-K score of ≥ 12 was used in the study by Jeon and Lee in 2010 [24], suggesting that the threshold was set too high. Stress is significantly associated with the frequency of alcohol intake and heavy drinking [23]. That is, stress is not significantly related to low-level alcohol intake, but is significantly associated with heavy drinking or harmful alcohol use. Therefore, stress management should reduce the risk of harmful alcohol use in both genders.

Consistent with the findings of a previous study [24], there was a higher prevalence of harmful alcohol use among men with an elementary-school education and women with a high-school education, than among those with a college education. However, a previous study showed that a higher level of education was significantly associated with harmful alcohol use regardless of gender [32]. Discrepancies among reports may be due to the population differences, as shown in a study conducted in Korean adults [24] and another with older Swedish adults [32]. The finding in this current study are consistent with a previous study among the Korean adult population. However, further studies are needed to investigate the differences between countries and populations. It was reported that single, separated, or divorced men showed a greater risk of harmful alcohol use [33]. Another study observed a lower prevalence of harmful alcohol use among unmarried men however, a higher prevalence was reported among unmarried compared with married, women [14]. They also reported that unmarried women had a higher risk of harmful alcohol use compared with married women, but no significant relationship with marital status was reported in men [23]. A previous study, reported that married individuals of both genders had a significantly lower risk of harmful alcohol use, compared with unmarried, divorced, separated, or widowed individuals [24]. These findings may be associated with stress due to problematic spousal relationships [34].

In men, but not in women, household income, perceived health status, and occupation were associated with harmful alcohol use. Household income was significantly positively related to alcohol consumption in men, but not in women. Additionally, unemployed men had a lower risk of harmful alcohol use than men with a professional occupation. It was reported that men with a low household income had a lower risk of harmful alcohol use compared with those with a high household income [24]. The differences between men and women may be due to gender differences in economic activities, job classifications, and income levels. A significantly lower prevalence of harmful alcohol use among unemployed men was observed in the current study, but employment status was not associated with harmful alcohol use in women. Another study observed that employment increased the frequency of harmful alcohol use among men but not women [35]. Unemployed people (i.e., those who lost a job or retired) may have fewer opportunities to participate in social gatherings (i.e., drinking with coworkers after work) and less disposable income. Alcohol consumption plays an important role in nonverbal communication in Korea, as people tend to socialize while drinking [36]. Korean workers often drink with coworkers after work to relieve stress, relax, or network [37]. It was reported in 2018 that men (69.4%) are more likely than women to be the major source of household income (30.6%) [38]. Moreover, as of July 2018, 74.1% of men ≥ 15 years of age were economically active compared with 53.5% of women [38]. In this current study, 81% of men were employed, whereas 55.6% of women were employed. This may explain why household income and employment status were significantly associated with harmful alcohol use in men but not in women.

A previous study showed that household income was not significantly associated with heavy drinking [23]. In contrast, several studies have demonstrated an association between low income and harmful alcohol use [39–44]. It was also reported that people with a middle-class household income were more likely to report harmful alcohol use [45]. Previous studies of the relationship between household income and harmful alcohol use were conducted in low-to-middle income countries (e.g., India, Zimbabwe, Nepal, and Thailand), whereas this study was conducted in Korea, a high-income country. A study of high household income in Chile and Finland, both high-income countries, reported a higher prevalence of alcohol consumption and harmful alcohol use [46]. Thus, the association between harmful alcohol use and household income may differ according to the economic level of the nation in question. Further studies should explore the factors that influence the relationship between household income and harmful alcohol use.

A poor self-rated health status was associated with a higher prevalence of harmful alcohol use in men but not in women. Several studies have shown that harmful alcohol use is significantly associated with perceived health status [47–49], but these studies did not assess gender differences. Perceived health status and harmful alcohol use may also be related to gender differences, derived from more frequent drinking among men than among women, as the adverse effects of this behavior may reduce self-perceived health status. Therefore, an improved perceived health status may reduce the risk of harmful alcohol use among men.

The findings of this current study provide insight into gender differences in the prevalence of, and the factors associated with, harmful alcohol use among Korean adults. However, this study had limitations. Firstly, the cross-sectional design of this study prevented investigation into the causal associations of smoking, depression, stress, and perceived health status with harmful alcohol use. Indeed, harmful alcohol use may have influenced the observations related to these factors. Secondly, the mechanisms underlying the gender differences observed in this study were not investigated.

Information on gender differences affecting harmful alcohol consumption and associated factors, will enable identification of subgroups vulnerable to harmful alcohol use, and aide in the development of management or prevention strategies for harmful alcohol consumption. In this study gender differences were observed in the prevalence of harmful alcohol use and its associated factors The demographic factors related to harmful alcohol use (i.e., education, marital status, household income, occupation, and perceived health status) should be considered when identifying the target groups for programs that are designed to prevent and/or treat harmful alcohol use. Moreover, such programs should also address smoking, stress, and depression.