Print

Ayer, Lynsay Allison

Psychology

2010

PhD

Alcohol use is highly prevalent among adolescents and young adults in The Netherlands (Currie et al., 2004) and is associated with negative outcomes such as motor vehicle accidents and sexually transmitted disease (NIAAA, 1997). Temperament and personality nality traits have been shown to confer risk for future alcohol abuse and personenc, particularly among individuals prone to the reinforcing effects of drinking alcohol (Levenson et al., 1987). The temperament-drinking literature is limited, however, by its reliance on variable-centered approaches to the examination of such associations which may not fully take into account the impact of trait combinations. Variable-centered approaches assume homogeneity within populations, whereas person-centered approaches can account for heterogeneity when testing relations between constructs (Von Eye & Bogat, 2006).
This study examined empirically derived adolescent temperament profiles and tested their longitudinal relation to drinking behavior in young adulthood. Socioeconomic indicators were examined as moderators of temperament's effect on later drinking. A moderating effect of SES on temperament and drinking could help to explain inconsistent findings on SES and alcohol use, and reveal whether particular temperamental and contextual factors render youth vulnerable to or protected against later drinking problems.
The current study utilized latent profile analysis (LPA) to empirically derive temperament profiles based on 6 traits (neuroticism, extraversion, boredom susceptibility, experience seeking, disinhibition, & thrill & adventure seeking) from a community sample of 827 Dutch adolescents (Mage = 14.88 years, SD =0.72). A four-profile model provided the best fit to the data, yielding the following patterns: (1) Neurotic, (2) Disinhibited Boys, (3) Dysregulated Girls, and (4) Regulated. The relation between these four profiles and young adult drinking behavior was then examined via linear mixed modeling (LMM). Interactions between temperament profiles and socioeconomic indicators (income & urbanization) were also tested in relation to later alcohol use. These longitudinal analyses were conducted on a sub-sample of 494 adolescents.
Adolescent temperament significantly predicted young adult drinking quantity and frequency above and beyond the effects of adolescent drinking, age, and sex. Specifically, Disinhibited Boys drank higher quantities (B=0.13, p<.05), and Dysregulated Girls drank more frequently (B=0.12,p<.05), relative to the Regulated group. There was weak support for direct and moderating effects of urbanization. Income evidenced neither a direct nor interactive effect on drinking. Clinical implications are described, and future research directions are discussed.