Personality and alcohol consumption: Pooled analysis of 72,949 adults from eight cohort studies

Abstract

Background: The role of personality as a determinant of alcohol consumption has long been debated, but prospective evidence is scarce.

Methods: We performed individual participant meta-analysis to examine the association between the Five-Factor Model personality traits (extraversion, neuroticism, agreeableness, conscientiousness, and openness to experience) and alcohol consumption using data from eight cohort studies sampled from the USA, UK, Germany, and Australia (total n=72,949; mean age=50 years, 54% female). Alcohol consumption was categorized into abstinence, moderate consumption, and heavy consumption.

Results: After adjustment for age, sex, and race, higher extraversion (odds ratio for 1 standard deviation's increase in the score; 95% confidence interval: 1.14; 1.01-1.29) and lower conscientiousness (0.89; 0.79-1.00) were associated with increased risk of transitioning from moderate to heavy alcohol consumption over time, and also with heavy alcohol consumption. Lower extraversion (0.91; 0.85-0.98), higher agreeableness (1.09; 1.02-1.15), and lower openness (0.90; 0.86-0.95) were associated with increased odds of transitioning from moderate consumption to abstinence as well as with alcohol abstinence.

Conclusion: Findings from this individual-participant meta-analysis suggest that high and increasing alcohol consumption is more common among extraverts and those low on conscientiousness whereas high agreeableness and low openness to experience may increase odds of reducing alcohol consumption and preferring abstinence.

Keywords: Addiction epidemiology; Drinking; Meta-analysis; Psychology; The Five-Factor Model.

Figure 1

Associations between the Five-Factor Model personality traits and (a) heavy alcohol consumption in baseline, (b) transition from moderate to heavy alcohol consumption, (c) abstinence in baseline, and (d) transition from moderate alcohol consumption to abstinence. Values are odds ratios per 1 standard deviation increment in personality trait. I² indicates the degree of heterogeneity in effect size across studies, and p(I²) gives the statistical significance for the heterogeneity. See Supplement Figures 1, 2, 3, and 4 for study-specific results.