Although their individual significance is well-documented, the interaction effects of age, sex, and alcohol use disorder (AUD) have undergone little systematic investigation. Here, we extend prior work interrogating sex and group (AUD vs. community comparison [CC]) by probing the main and interaction effects of age on emotion processes as well as two conventional neuropsychological tests. Main effects for age and group were anticipated; however, interaction effects comprise our primary focus. While sex differences in AUD prevalence are commonly reported, sex differences within AUD samples are inconsistently found. Therefore, we pose our inquiry regarding sex as exploratory. In this study, 54 CC (n=30 women) and 55 AUD (n=14 women) subjects completed a battery that included an emotional face discrimination task, the Trail-Making Test-B (TMT-B), and the Digit Symbol Substitution Test (DSST). The initial/full models included the main and interaction effects of age (as a continuous variable; 25-59 years of age), sex, and group (AUD, CC). In analysis of the emotional face discrimination task, performance on a non-affective face discrimination task was entered as a covariate. Analysis of emotion identification revealed group and age main (p=.02; d=.53 & .003; d=.50, respectively) and interaction effects (p=.05; d=.41). The latter suggested that age and emotion processing performance were positively correlated in the AUD group, but unrelated in the CC group. Notably, neither sex, main, nor interaction effects achieved significance. Using the full model, analysis of the TMT-B and DSST failed to show sex effects or reveal expected performance decrement in the AUD group. To clarify the latter, simple models including only group as well as correlations between age and performance by group for each task were conducted. These analyses demonstrated the expected AUD-related deficits and suggested differential relationships between age and neurocognitive performance as a function of both group and task. Outcomes across tasks emphasize the need to reframe aging effects, particularly in the context of AUD.
Read full abstract