Etiological (Dis)Similarity Between Individual Alcohol Use Disorder Symptoms

Previous literature indicates that overall heritability of alcohol use disorder (AUD) is approximately 50%. However, there is little research examining etiological similarity between individual AUD symptoms or hazardous... [ view full abstract ]

Previous literature indicates that overall heritability of alcohol use disorder (AUD) is approximately 50%. However, there is little research examining etiological similarity between individual AUD symptoms or hazardous drinking behavior. Waves 1 and 3 of the National Longitudinal Study for Adolescent to Adult Health (Add Health) sibling subsample (319 MZ pairs (145 male; 144 female) and 476 DZ pairs (117 male; 131 female; 204 opposite) were analyzed. Past-year endorsement of outcomes related to three AUD symptoms were assessed: hazardous behavior ([a] fighting or [b] having regretted sex while drinking), interference (having trouble with [a] school or work, [b] romantic partners, or [c] friends), and sickness/aftereffects ([a] vomiting or [b] hangovers).  Past-year risky drinking based on the NIAAA guidelines was also assessed: drinking more than 7/14 drinks per week (for women and men respectively) and no binge drinking. Waves 1 and 3 were assessed to examine both adolescent (W1 age M = 15.65, SD = 1.74) and young adult (W3 age M = 21.89, SD = 1.75) etiology, as well as stability across time. Univariate biometric models indicated that heritability varied across outcomes. Percentage attributed to additive genetic variance ranged from 0% (trouble with school/work; fighting) to 60% (having regrettable sex) in adolescence, and from 0 % (fighting) to 36% (vomiting) in young adulthood. Variance was typically non-significant across all outcomes at both age periods. Variance attributed to unique environment in young adulthood was the most consistent factor, ranging from 43% (trouble at school/work) to 73% (having regrettable sex). Unique environment factors were typically significant in adolescence and young adulthood. Bivariate biometric analyses were then estimated in which all outcomes were paired with each other individually and with risky drinking, at each age period (i.e., 28 models per wave), as well as biometric models that analyzed similarity over waves. Results indicated a range of rG, rC, and rE scores, highlighting the variability in etiological similarity across symptoms, and in relation to risky drinking behavior itself.  There was also little stability in relations between additive genetic and environmental variance over adolescence and young adulthood for symptoms over adolescence and young adulthood.