Behavioral, craving, and anxiety responses among light and heavy drinking college students in alcohol-related virtual environments

Abstract

Drinking-related behavior in college students represents a public concern with consequences for health and academic performance. The aim of the present study was to determine which measures (behavioral and self-reported measures of craving and anxiety) differentiate best between light- and heavy-drinking college students when exposed to a virtual reality (VR) alcohol-cue environment. 25 college students participated in this study, of whom 13 were light drinkers (standard drink units (SDU)/month ≤ 10) and 12 heavy drinkers (SDU)/month ≥ 11). Participants completed the Alcohol Use Disorder Identification Test (AUDIT) before exposure to the VR environment. Heavy drinkers scored higher than light drinkers on AUDIT. The virtual environment consisted of four situations: restaurant, bar, chill-out area, and bedroom, where participants could choose alcoholic or non-alcoholic beverages. An Oculus Rift DK2 headset was used as the HMD. In each situation, craving and anxiety were self-reported on a visual analog scale (VAS, from 0 to 10). The results showed differences between groups in the type of beverage chosen in the VR situations, whereby heavy drinkers chose alcoholic drinks more frequently. However, no statistically significant differences were found between groups in craving or anxiety levels reported on the VAS during VR exposure. Heavy-drinking students show a preference for alcoholic beverages in all VR situations compared with light drinkers, but do not experience different levels of craving or anxiety as assessed with VAS. If virtual environments are used to detect heavy drinking cases, behavioral parameters such as choosing between alcoholic or non-alcoholic cues seem more suitable than self-reports of craving or anxiety. Nevertheless, future studies are necessary to determine whether more objective measures of craving and anxiety (eye tracking or psychophysiological responses) perform better than selfreports in differentiating between heavy and light drinking.