Abstract
Background: Marriage is one of the most frequently examined sources of social support and has been shown to protect against alcohol use and abuse. This study examines the relationship between perceived marital strain and support, and alcohol use controlling for additive genetic influence. Methods: Data from monozygotic (MZ) (n = 320) and dizygotic (DZ) (n = 464) twin pairs from the second wave of the National Survey of Midlife Development in the United States (MIDUS II) were used to test whether past year marital strain and support were associated with recent alcohol use. Generalized linear mixed models (GLMM) were estimated, allowing us to control for additive genetic and shared environmental influences as variance components. Results: Marital strain and support had positive, statistically significant associations with alcohol use. However, only the relationship between marital strain and alcohol use remained after controlling for variance in alcohol use attributed to genetics. Conclusions: After accounting for genetics, midlife adults still appear to cope with marital strain via alcohol use. However, this coping is unlikely to result in heavy episodic drinking or alcohol use disorder without other compounding factors.
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.
Notes
1 A multiple imputation (MI) algorithm in STATA 13 was used to retain missing observations. Since missingness approached 20% in some measures, including marital, familial, and peer strain/support, we imputed 50 data sets, and missing values were replaced with mean scores from across the 50 imputed multivariate normal (MVN) data sets, which was necessary as STATA 13’s MI function cannot be combined with the GLLAMM package. (See for post-imputation descriptive statistics by zygotic category).
2 Even though it is possible that these measures of marital strain/support precede alcohol use, it is more likely that respondents appraised all survey items using the most recent, relevant memories (Greene, Citation1986). It is therefore unlikely that responses to these survey items reflect the time ordering inherent to the survey items.