Reframing addiction: It’s not the destination, but the journey

In recent years, clinical addiction researchers have forged new paths. Motivated by dimensional models of addiction that reframe disordered drug use as arising from multiple processes in distinct functional domains [1], these novel approaches characterize several neurobehavioral mechanisms in the same individuals and their interactive influence on drug use. This innovative perspective integrates neuroscience findings and the vast heterogeneity in substance use phenotypes to devise more effective strategies to target treatments and interventions. Moreover, it emphasizes a need to understand the complex and varied pathways to addiction.

The report by Joyner et al. in this issue of Neuropsychopharmacology is an important contribution to the burgeoning literature in this area [2]. Notably, the findings inform a heterogeneous etiology of disordered drinking and the interplay of genetic and external factors. The authors analyzed electroencephalography (EEG) data and self-reports to determine how cognitive disinhibition (executive control domain) and alcohol cue reactivity (ACR, incentive salience domain) relate to problematic alcohol involvement (PAI) in emerging adults with past-year binge drinking but no current or previous history of alcohol use disorder (AUD). Neurophysiological data comprised event-related potential (ERP) P3 amplitudes elicited by oddball stimuli in separate visual tasks. In the rotating heads task, low amplitude Target-P3 reflected high disinhibition, whereas high amplitude ACR-P3 reflected greater alcohol cue reactivity in the picture viewing task.

Non-invasive scalp EEG is a valuable tool for understanding the relationships between alcohol use and brain function. With a high temporal resolution capable of measuring neural activity in real-time, ERPs elicited by task stimuli can be mapped to distinct cognitive processes. Together with recent advances in dual imaging techniques that can identify the neural basis of EEG signals, brain-behavior relationships are becoming clearer. Thus, ERPs measured at different stages of AUD can reveal links between dysregulated cognitive processing and phenotypic heterogeneity, as in [2]. Similarly, ERPs can be used during treatment to determine neurocognitive effects and identify sensitive indicators that can be used to monitor recovery [3].

The principal findings of the study confirmed the author’s hypotheses; after controlling for current drinking, relationships of Target-P3 and ACR-P3 with problematic alcohol involvement (PAI) were directionally opposite, differently mediated, and mutually suppressive. Target-P3 was negatively related to PAI and partially mediated by trait disinhibition, whereas ACR-P3 was positively associated with PAI regardless of disinhibition scores—thus, the smaller the Target-P3 and the greater the ACR-P3, the higher the PAI score. The relationship between Target-P3 and PAI was partly explained by trait disinhibition, possibly representing the heritable component of Target-P3. Implications include that informational processing of cognitive oddball stimuli is determined by a combination of genetic influences and external or state-related factors that impact cognitive disinhibition. It is also possible that other cognitive processes besides disinhibition contribute to Target-P3. Finally, results showed that Target-P3 and ACR-P3 are related, but after controlling for their shared variance, each P3 explained a unique variance in PAI. In other words, individuals with high PAI may exhibit low amplitude Target-P3 and low amplitude ACR-P3, high amplitude Target-P3 and high amplitude ACR-P3, or low amplitude Target-P3 and high amplitude ACR-P3.

The study findings are in line with existing evidence of the risk factors for disordered drinking; high alcohol cue reactivity and trait disinhibition positively predict alcohol use and AUD. Additionally, trait disinhibition explains a significant proportion of the genetic component of AUD. They also add to our understanding of early PAI and heterogeneity in both alcohol use phenotypes and etiology. Significant effects of alcohol cues on neural activity demonstrate that pathological processes central to the development of AUD may be established at very early stages of alcohol involvement and before a diagnosis of AUD. This highlights the sensitivity of ACR-P3 to detect the earliest alcohol-induced insults to neural function and supports its potential as a prognostic marker that can identify at-risk PAI phenotypes and deliver targeted interventions before clinical symptoms emerge. To this end, it will be critical to determine the predictive validity of ACR-P3 in longitudinal studies assessing long-term changes in drinking and associated problems. The data could also inform an understanding of heterogeneity in cue-induced craving and the selection of treatments for cue non-reactive individuals. For example, ACR-P3 may not strongly predict the later emergence of cue-induced craving, suggesting that conditioned cognitive responses are able to influence drinking via other mechanisms. In that case, treatments targeting conditioned cue responses may still be effective for persons who do not report intense cravings.

The findings also demonstrate how genetic and environmental (including alcohol-related) influences on neural function can interact to predict PAI. Finally, the results show unique influences of mechanisms in separate functional domains on alcohol use and alcohol problems, thus supporting dimensional models of addiction. As discussed by the authors, limitations include that the findings may not generalize because the sample did not include participants with AUD or more severe PAI. In addition, there is currently no evidence for longitudinal changes in P3 responses. Thus, they highlight the need for studies in clinical samples and longitudinal designs to probe how relationships between PAI, cognitive disinhibition, and cue reactivity change as drinking progresses and diagnostic symptoms appear, or as treatment is initiated and interventions implemented in clinical samples. Existing data support the test-retest reliability of ERPs, but few studies have evaluated ERPs as predictive biomarkers or examined how ERPs and behavior covary longitudinally. Thus, these areas are ripe for exploration and pivotal for advancing a neuroscientific approach to AUD treatment.

As the body of evidence documenting the dimensional nature of substance use grows, so too will an understanding of how to best intervene and target effective treatments. Undoubtedly, the theoretical models will be refined, and new paths of discovery will emerge. In this context, the findings of Joyner et al. [2] stand as a valuable contribution that highlights the complexity of early PAI from a perspective of dimensional addiction models. The study clarifies the multifaceted nature of disordered alcohol use and underscores the importance of considering both genetic predispositions and environmental influences to understand its underlying causes.

Funding

EC is supported by NIH grants R21 DA057071 and P50 AA022538.

Competing interests

The authors declare no competing interests.