Test-retest reliability of the underlying latent factor structure of alcohol subjective response.

Abstract

Alcohol subjective experiences are multi-dimensional and demonstrate wide inter-individual variability. Recent efforts have sought to establish a clearer understanding of subjective alcohol responses by identifying core constructs derived from multiple measurement instruments. The aim of this study was to evaluate the temporal stability of this approach to conceptualizing alcohol subjective experiences across successive alcohol administrations in the same individuals. Healthy moderate alcohol drinkers (n=104) completed six experimental sessions each, three with alcohol (0.8g/kg), and three with a non-alcoholic control beverage. Participants reported subjective mood and drug effects using standardized questionnaires before and at repeated times after beverage consumption. We explored the underlying latent structure of subjective responses for all alcohol administrations using exploratory factor analysis and then tested measurement invariance over the three successive administrations using multi-group confirmatory factor analyses. Exploratory factor analyses on responses to alcohol across all administrations yielded four factors representing "Positive mood," "Sedation," "Stimulation/Euphoria," and "Drug effects and Urges." A confirmatory factor analysis on the separate administrations indicated acceptable configural and metric invariance and moderate scalar invariance. In this study, we demonstrate temporal stability of the underlying constructs of subjective alcohol responses derived from factor analysis. These findings strengthen the utility of this approach to conceptualizing subjective alcohol responses especially for use in prospective and longitudinal alcohol challenge studies relating subjective response to alcohol use disorder risk.