Relationships on the rocks: A meta-analysis of romantic partner effects on alcohol use

The partner influence hypothesis postulates one partner's alcohol use influences the other partner's alcohol use over time. Although several studies have examined the partner influence hypothesis, the magnitude and gender-specific nature of partner influences on alcohol use are unclear and have yet to be examined meta-analytically. We addressed this by conducting a traditional bivariate meta-analysis and two-stage meta-analytic structural equation modeling across 17 studies (N = 10,553 couples). Studies that assessed both romantic partners' alcohol use at a minimum of two time-points were selected. Results suggest romantic partners do influence one another's drinking, to a small but meaningful degree, with women (β =.19) exerting a statistically stronger (p <.05) influence than men (β =.12). Results also suggest time lag between assessment, alcohol indicator, married, and year of publication may moderate partner influence. Thus, social influences on individual alcohol use include important partner influences. These influences can serve either risk or protective functions. Given the economic, social, and health consequences associated with alcohol misuse, advancing knowledge of social risk factors for alcohol misuse is essential. Therefore, assessment and treatment of alcohol misuse should extend beyond the person to the social context. We encourage clinicians to consider involving romantic partners when assessing and treating alcohol misuse.