Latent class analysis of drinking behaviors and predictors of latent class membership among college students in the Republic of Korea

Objectives: We aimed to investigate the drinking patterns of college students and successful predictors of latent class membership (LCM). Methods: We analyzed cross-sectional data obtained from 1,338 Korean college students who had consumed at least one standard drink in their lifetime. Primary data were collected during December 2016. We conducted latent class analysis to investigate drinking patterns and successful predictors of latent class membership. Results: A three-latent class model best fit our data: low-risk drinking (34%), intermittent binge drinking (37%), and habitual binge drinking with negative experiences (29%) classes. In addition, successful predictors of LCM were gender, college year, monthly allowance, sensation seeking, age of drinking onset, current smoking, peer pressure, and friends' drinking/smoking. Conclusions: Given that heavy drinkers were at greater risk of experiencing negative drinking consequences, health professionals should provide heavy-drinking college students with interventions that aim to equip students with protective behavioral strategies when drinking. In addition, in developing and providing interventions aiming to reduce at-risk drinking, health professionals should take into account the successful predictors of LCM.