Brain response to working memory over three years of adolescence: influence of initiating heavy drinking

Abstract

Objective: Many adolescents engage in heavy alcohol use. The aim of this study was to disentangle whether brain abnormalities seen in adolescent heavy drinkers are a consequence of heavy drinking, a preexisting risk factor for initiation of alcohol use, or both.

Method: Study 1 used cross-sectional functional magnetic resonance imaging (fMRI) visual working-memory (VWM) data from 15- to 19-year-olds (20 heavy drinkers, 20 controls) to identify brain regions affected by heavy adolescent alcohol use. Study 2 used longitudinal fMRI VWM data from 12- to 16-year-olds imaged before the onset of drinking and imaged again on the same scanner approximately 3 years later. Those who had transitioned into heavy drinking (n = 20) were matched to continuous nondrinkers (n = 20) on baseline alcohol risk and developmental factors (N = 40; 80 scans).

Results: Study 1 found that heavy drinkers exhibited more frontal and parietal but less occipital activation than controls, defining the regions of interest for Study 2. In Study 2, adolescents who later transitioned into heavy drinking showed less fMRI response contrast at baseline than continuous nondrinkers, which increased after the onset of heavy drinking, in frontal (1,431 μL, p = .003; η² = .19) and parietal (810 μL, p = .005; η²= .23) regions, as in Study 1. Lower baseline activation in the frontal and parietal regions predicted subsequent substance use, more so than commonly observed predictors of youth drinking (p < .05).

Conclusions: Adolescents who initiated heavy drinking showed different brain activation before the onset of drinking, then less efficient information processing after high-dose alcohol use started. This suggests neural response patterns that could be risk factors for future substance use and also supports prior neuropsychological reports indicating that initiating heavy episodic drinking in adolescence may be followed by subtle alterations in brain functioning.

Figure 1

Drinking classification for Studies 1 and 2, based on Cahalan et al. (1969) and modified based on the distribution of drinking characteristics of adolescents (Schweinsburg et al., 2005b, Squeglia et al., 2009; Tapert et al., 2004b).

Figure 2

Study 2 significant Group × Time interactions were found in two of the five hypothesized regions of interest: the right inferior parietal lobule (cluster size: 810 μL, p = .005) and left medial frontal gyrus (cluster size: 1,431 μL, p = .003). At baseline, significant differences in BOLD response contrast to high working-memory load relative to low working-memory load were observed in both regions (ps < .01). At follow-up, differences in BOLD response contrast trended toward significance (p = .10) in the right inferior parietal lobule. Although continuous nondrinkers (n = 20) did not change significantly over time, heavy drinkers (n = 20) did in both regions.