Modulation of the mesolimbic reward circuitry has been postulated as an important substrate underlying the development of alcoholism and other drug addiction (Blum et al., 2000
; Robinson and Berridge, 2008
). The current study examined the hypothesis that children of alcoholics show different reward circuitry responding beyond confounding effects of substance use, which may underlie their increased AUD risk. There were two main findings. First, blunted NAcc incentive response was only observed in COAs who had low alcohol use and related problems. Second, a link between early externalizing risk, NAcc responding, and current and lifetime alcohol consumption was observed in COAs but not controls.
A modified MID task was used to probe reward circuitry response during anticipation period between cue presentation and motor responding. Though both groups showed expected VS activation during anticipation of reward and loss-avoidance, COAs demonstrated less right NAcc activation compared to controls. As there were no behavioral differences between the groups due to individually calibrated task difficulty, this activation difference reflects differing responsivity of the saliency-reward system. This blunted motivational response, however, was not supportive of the reward deficiency syndrome hypothesis of addiction vulnerability (Blum et al., 2000
). Further analyses revealed a significant family history by alcohol use interaction where reduced NAcc responding was only observed in COAs who had been low-drinkers, never binged and experienced few alcohol-related problems. High-risk-drinking COAs responded at the same level as controls, who showed no differential response across drinking risk-levels. Further, low-risk-drinking COAs who showed blunted NAcc response also had fewer precursive externalizing problems than high-risk-drinking COAs. The two COA groups did not differ in family severity index, indicating that these effects were not explained by the extent of liability within the nuclear family. Therefore, our results suggest that a lesser NAcc responsiveness to incentives may reflect a resilience mechanism, reducing risk for AUD development in COAs.
Early externalizing problems have consistently been identified as a risk factor for alcoholism (Cloninger et al., 1988
; Zucker et al., 2008
). Indeed, across our entire sample, externalizing problems in early adolescence were correlated with both current and lifetime alcohol consumption. Our results further revealed a positive association among externalizing problems, current and lifetime alcohol use, and NAcc activation that was unique to COAs. Therefore, although control and COA groups were matched on early externalizing behavior and alcohol consumption, they differed in the relationship between these variables and NAcc response to incentive anticipation. It is of interest to note that, although externalizing behavior scores at time of scanning were missing from 20% of COAs and 5% of controls, analysis of this incomplete dataset revealed consistent findings of significant positive associations between current externalizing behavior, alcohol use and NAcc activation in COAs but not controls. This suggests that a close association of behavioral risk, alcohol consumption and motivational mechanisms, may underlie addiction vulnerability in COAs. The absence of such systematic associations in controls may be a reflection of their lower overall AUD risk compared to COAs, which likely involves a more heterogeneous distribution of risk factors.
The current finding of a positive correlation between NAcc incentive activation and alcohol consumption in COAs may be interpreted in different ways. It may be driven by predisposed differences in reward circuitry reactivity. Alternatively, this may reflect a sensitization of reward circuitry following alcohol exposure, suggesting that COAs may be more susceptible to the modulating effects of alcohol on reward circuitry responding than controls. A recent study showed that healthy adults with a family history of alcoholism reported increased stimulating effects of alcohol and increased “wanting” and “liking” compared to controls with similar levels of current alcohol use (Soderpalm Gordh and Soderpalm, 2011
). Prospective studies are needed to test these hypotheses.
The finding of an opposite relationship between NAcc incentive activation and externalizing behavior in COAs versus controls was unexpected. However, Galvan et al. (2007)
demonstrated opposing correlations between NAcc activation to reward and anticipated positive and negative consequences of risk-taking. Specifically, anticipated positive consequences correlated positively with NAcc activity whereas anticipated negative consequences correlated negatively. Therefore, the opposite correlations found here between control and COA groups may relate to differences in their expectations about the outcomes of their externalizing behavior. This hypothesis requires further exploration in future studies which include measures of the anticipated outcomes of behaviors.
The present findings also raise the important issue of individual variability of AUD risk in COAs, and its relations to reward circuitry response. Transmission of familial liability factors can occur via multiple genetic effects (Goldman et al., 2005
) as well as environmental effects associated with alcoholic households (Zucker et al., 2008
). It has been shown that parental AUD history moderates the effect of dopamine transporter (DAT1
) gene 10-repeat allele on increased alcohol problems in males, with the effect present only in those with an alcoholic father (Vaske et al., 2009
). Furthermore, a recent study found a positive association between VS activation to reward anticipation and trait reward sensitivity in homozygote carriers of DAT1
10-repeat allele, but not in other allele carriers (Hahn et al., 2010
). These studies, along with the current findings, illustrate a complex interplay of genetic and environmental effects, reward-related personality traits and reward circuitry functioning in relation to differential transfer of familial risk. Although it is beyond the scope of the current study, this will be important to investigate in future studies.
The direction of the present findings differs from that observed in adult alcoholics. Detoxified alcoholics showed reduced activation in the VS during anticipation of monetary incentives, but increased activation to alcohol cues that was further associated with increased alcohol craving (Wrase et al., 2007
). Past drug exposure has been shown to modulate subsequent motivational response in animal and human models (Wyvell and Berridge, 2001
; Boileau et al., 2006
; Nocjar and Panksepp, 2007
). The findings in older alcoholics may therefore suggest a bias in reward system responsivity toward alcohol-related stimuli after continued alcohol use. More recently, an association between VS activation to incentive anticipation and self-reported impulsivity has been found in detoxified alcoholics but not in controls (Beck et al., 2009
). This supports the hypothesis that the relationship between VS responding to incentive anticipation and behavioral risk may differ for those vulnerable to addiction.
In a sample of 12 to 16 year old COAs and controls, Bjork et al. (2008b)
found no difference in monetary incentive response between the groups. The difference from current findings may be due to the developmental period at which subjects were investigated. An emerging literature supports developmental changes in reward circuitry reactivity through the adolescent years. VS response to reward anticipation was shown to increase from adolescence to adulthood (Bjork et al., 2004
; Bjork et al., 2010
). A developmental effect has also been observed in the association between NAcc response during monetary reward anticipation and expectation of positive or negative consequence of risk-taking behavior (Galvan et al., 2007
). Specifically, anticipated positive consequences were associated with more NAcc activation and greater likelihood of engaging in risky behavior in adults, whereas in children an association was found between anticipated negative consequences, lesser NAcc activation and lesser likelihood of engaging in risky behavior. Associations in both directions were found in adolescents, suggesting a developmental shift in anticipation of consequences and related modulation of NAcc responsivity during this interval. The differences in NAcc responses found in the present study may not have emerged until late adolescence/early adulthood - a critical period in development when alcohol use and AUD prevalence reach their highest levels (Grant et al., 2004
; Substance Abuse and Mental Health Services Administration, 2006
). Prospective longitudinal studies are needed to investigate this possibility.
A limitation of this study is its relatively small sample size (n=40), based on a-priori selection of subjects from a larger sample to allow for accurate matching. A larger sample would allow for more accurate characterization of associations within individual groups. In addition, the current study focused on early externalizing problems (age 12–14) as a precursive behavioral measure prior to significant drinking and drug use. Future studies will be necessary to determine how current levels of externalizing problems and incentive responding may be linked in relation to risk and resilience.
The present study demonstrates a close association between NAcc activation to incentive anticipation, precursive behavioral risk, and alcohol consumption that is unique to COAs. The differences in correlations, along with a significant interaction between family history and alcohol use, show that the relationship between reward circuitry response and AUD risks differs between COAs and controls. The consistent findings across reward and loss anticipation is of additional significance as growing evidence suggests that motivational salience rather than valence/hedonic effects is a key substrate in addiction development (Robinson and Berridge, 2008
). Taken together, our findings suggest a multilevel dynamic process whereby a blunted NAcc response to incentive anticipation is related to lower precursive behavioral risk and is further associated with lesser alcohol use and related problems in COAs, potentially reflecting a resilience mechanism. Future prospective studies with larger groups of resilient and at risk COAs and controls will help to confirm and expand these initial observations.