Impact of Alcohol-Induced Facial Flushing Phenotype on Alcohol Consumption Among Korean Adults: 2-Year Cross-Sectional Study

Abstract

Background: The alcohol-induced facial flushing phenotype (flushing) is common among East Asians. Despite a small intake of alcohol, they experience heightened levels of acetaldehyde, a group-1 carcinogen, which, in turn, causes unpleasant symptoms such as redness, acting as a robust protective mechanism against consuming alcohol. However, some individuals with this genetic trait exhibit weakened alcohol restraint, which increases the risk of developing alcohol-related cancers, such as esophageal and head or neck cancer, by several times. Although this flushing phenomenon is crucial for public health, there is a paucity of studies that have comprehensively investigated the effect of flushing or its genotype on alcohol consumption in a large group of East Asians while controlling for various sociodemographic and health-related variables at a country level.

Objective: This 2-year cross-sectional study aims to explore the effect of flushing on drinking behavior in Koreans and to examine whether the effect varies across sociodemographic and health-related factors.

Methods: We used data from the Korea National Health and Nutrition Examination Survey (KNHANES) for 2019 and 2020 conducted by the Korea Disease Control and Prevention Agency. Our sample comprised 10,660 Korean adults. The study investigated the association of 26 variables, including flushing, with drinking frequency and amount. The effect of flushing was examined with and without adjusting for the other 25 variables using multinomial logistic regression analysis. In addition, we tested the interaction effect with flushing and conducted a simple effect analysis. We used complex sample design elements, including strata, clusters, and weights, to obtain unbiased results for the Rao-Scott χ² test, 2-tailed t test, and multinomial logistic regression analysis.

Results: The suppressive effect of flushing was significant (P<.001) across all pronounced categories of alcohol consumption in 2019. The ranges of standardized regression slopes and odds ratios (ORs) were -6.70≥β≥-11.25 and 0.78≥OR≥0.50 for frequency and -5.37≥β≥-17.64 and 0.73≥OR≥0.36 for amount, respectively. The effect became somewhat stronger when adjusted for confounders. The effect also exhibited an overall stronger trend as the severity of alcohol consumption increased. The β values and ORs were consistently smaller in 2020 compared to the previous year. A simple effect analysis revealed a diminished alcohol-suppressive effect of flushing on alcohol consumption for specific groups (eg, those with low levels of education, limited family support, physical labor, or health-related issues).

Conclusions: Our findings suggest that flushing suppresses drinking in Koreans overall but has little or no effect in certain susceptible populations. Therefore, health authorities should conduct targeted epidemiological studies to assess drinking patterns and disease profiles, particularly regarding alcohol-related cancers, and establish effective preventive measures tailored to this population.

Keywords: East Asian; acetaldehyde; alcohol; alcohol consumption; aldehyde dehydrogenase 2 polymorphism; drinking behavior; facial flushing.

Figure 1

Changes in the effect of alcohol-induced facial flushing on alcohol consumption behavior with age among Korean adults in the 2019-2020 Korea National Health and Nutrition Examination Survey (KNHANES). The line graphs are presented exclusively for the results where the interaction effects of flushing and age were statistically significant at the .05 level, observed with the dependent variables of drinking frequency in 2019 and drinking amount in 2020, respectively. The expected probabilities of flushing versus nonflushing groups as a function of age at each level of drinking frequency (A-F) and drinking amount (G-L). (A and G) The expected probability of the flushing group is slightly higher than that of its counterpart, and the trend increases with age for both groups. (B and H) The flushing group shows a noticeably higher expected probability at younger ages compared to its counterpart, but the gap gradually decreases as age increases. (C) The difference between the 2 groups is minimal, and both exhibit decreasing trends with increasing age. (I) The flushing group exhibits a linear decrease in the expected probability, while its counterpart shows a negative quadratic trend with increasing age. (D), (E), (J), (K), and (L): in all these categories representing relatively high degrees of drinking frequency and amount, the expected probability of the flushing group is lower than that of its counterpart. This group difference appears larger at younger ages but, by and large, decreases as age increases. (F) The expected probabilities of both groups increase steadily as age increases at the highest level of drinking frequency. This suggests that the alcohol consumption behavior of those drinking as frequently as ≥4 times per week would be quite different from that of the rest, regardless of the presence of flushing.