Regular versus episodic drinking in Swedish women: Reporting of regular drinking may be less biased by social desirability

Aim: To describe the personality and social characteristics associated with regular and episodic alcohol consumption in a Swedish cohort of women.

Methods: 406 women aged 38 and 50 who participated in the Population Study of Women in Gothenburg in 2004-2005 with complete data on the key variables are included. Regular alcohol use was based on frequencies of alcoholic beverage consumption, reported to examining physicians. Regular drinking was defined as those consuming wine, beer, or spirits at least twice weekly. Episodic drinking was defined as consumption of six drinks or more on a single occasion at least once during the last year. Personality traits were studied using the self-administered 57-item Eysenck Personality Inventory, which includes a 'lie scale' measuring the tendency toward social desirability, a 'neuroticism scale', and an 'extraversion scale'. Logistic regression was used to estimate odds ratios (OR) for regular and episodic drinking, respectively, in relation to standardized (SD) personality scores and selected social characteristics.

Results: 49% of the women reported episodic drinking, while 58% reported regular drinking, and 34% reported both. Women with a higher tendency toward socially desirable responses were less likely to report episodic drinking (odds ratio [OR] = 0.67, 95% confidence interval [CI] = 0.53-0.84) per standard deviation (SD), a trait that was not associated with regular drinking.

Conclusions: The strong inverse association between the propensity to lie scale with episodic but not regular drinking suggests that episodic drinking is subject to greater social desirability and under-reporting biases than regular drinking. Our findings indicate that this type of problem drinking may be missed in medical examinations, which limits the ability of health professionals to intervene.