The future of reward and relief drinking profiles: Considerations for social motives, stability of profiles, and tailored interventions

Understanding pathways and motives underlying drinking has become useful for characterizing individuals diagnosed with an alcohol use disorder (AUD, as defined by the DSM-5 (American Psychiatric Association and Association, 2013)), and to a lesser extent individuals who are drinking heavily or at hazardous levels (Bujarski and Ray, 2014). Specifically, reward and relief profiles represent different motivations for alcohol use. Reward drinking has been characterized as drinking for positive reinforcement (e.g., stimulating, mood enhancing, and rewarding effects of alcohol); whereas relief drinking is characterized as drinking for negative reinforcement (e.g. drinking to relieve negative emotional states, withdrawal symptoms, stress; (Grodin et al., 2019). These profiles have been hypothesized to correspond with stages of the addiction cycle and with associated neurobiology (Koob and Volkow, 2016), and correspond with the incentive motivation model of alcohol use (Verheul et al., 1999). Importantly, reward and relief phenotypes have promise for developing pharmacotherapy precision medicine approaches (Witkiewitz et al., 2019) and predicting differentiable responses to alcohol administration (Grodin et al., 2019).

In the current paper in Alcohol: Clinical and Experimental Research, Hebden et al. (2023) examined reward and relief drinking subgroups in binge drinking college students. Importantly, this study examined the benefit of including a social drinking motive for the first time. Using latent profile analysis, they found evidence for three drinking profiles: low-reward/low-relief, high-reward/low-relief, and high-reward/high-relief.

The benefit of including “social reward” in the reward profile

The authors used items from the Drinking Motives Questionnaire-Revised-Short Form (DMQ-R-SF) (Kuntsche et al., 2005) to assess reward enhancement (e.g. “Because you like the feeling”), reward social (e.g. “Because it makes social gatherings more fun”), and relief (e.g. “Because it helps when you feel depressed or nervous”). They found a benefit of including the reward-social items in their models, such that indices of model fit for reward-relief profiles were better when the reward-social items were included. Interestingly, the frequency of endorsing reward-social items was highest for high-reward/high-relief drinkers and high-reward/low-relief drinkers, indicating that, for reward drinkers, the reward-social motives were more motivating than some of the reward-enhancement motives. The inclusion of social enhancing items is an important step in the refinement of reward and relief phenotypes, especially in college-aged drinkers. However, the DMQ-R-SF does not include drinking corresponding to social relief items.

Social motives may correspond to relieving negative states, for example, drinking for loneliness or drinking due to feelings of social anxiety or tension, which may also be critical to the relief phenotype (King et al., 2023). The concept of social relief is consistent with the social-attributional framework, in which alcohol’s effect in social settings is based on perceptions of social rejection (Fairbairn and Sayette, 2014). Specifically, in this framework, alcohol is proposed to enhance mood when individuals are anticipating social rejection (Fairbairn and Sayette, 2014). Ostracism, social rejection, and drinking in an unfamiliar social setting all increase alcohol consumption (Bacon and Engerman, 2018, Laws et al., 2017, Fairbairn et al., 2018), further supporting the construct of a relief-social drinking motive. Of note, this study did not identify a drinking profile associated with low-reward/high-relief, which has been previously identified in adult samples with AUD (Witkiewitz et al., 2019). This is consistent with a previous study using the DMQ-R-SF in young adults (Roos et al., 2021) and with evidence that young adults report greater enhancement motives than coping motives (Kuntsche et al., 2005). However, it is possible that a low-reward/high-relief profile could be identified with the inclusion of relief-social items, for example from the Young Adult Alcohol Motives Scale (King et al., 2023), which includes a social tension reduction factor, and with college-aged samples with more AUD symptomology, which should be investigated in future studies with young adult samples.

Stability of Drinking Profiles Over the Lifespan

Hebden et al. (2023) and others have postulated that reward and relief drinking profiles reflect phases of the addiction cycle (Koob and Volkow, 2016), with reward drinking profiles reflecting the binge-intoxication stage and relief drinking profiles representing the withdrawal-negative affect stage. However, this assertion requires further consideration as there are important clinical implications if research reveals that reward/relief profiles are fixed profiles or that drinking profiles change over time. Clarifying whether reward and relief drinkers are fixed profiles or vary across the lifespan would help determine how precision medicine approaches are applied. If reward and relief drinkers are fixed profiles, this would allow for a singular assessment of drinking motives to determine appropriate treatment planning across the life course. By contrast, if research revealed that drinking profiles change over time, this would require clinicians to repeatedly assess for drinking motives and adapt treatment over time. The current literature on the stability of drinking motives is limited with some studies indicating that High, Low, and Positive Reinforcing motive classifications are stable over a 12-month period in college students (Arterberry, 2015) and other studies highlighting that coping and enhancement motives change across early to mid-adulthood (18–35) (Littlefield et al., 2010). As Hebden et al. (2023) note, future longitudinal studies should assess the stability of reward and relief drinking profiles across the lifespan.

It is also possible that drinking profiles are more likely to change in particular periods of the lifespan. Given that the mean age of onset for AUD is 20.3 (Seeley et al., 2019) and given that the binge-intoxication stage of the addiction cycle reflects an earlier period in the course of AUD, the 20s may be a period in which reward versus relief drinking profiles are more likely to change and after this period, drinking profiles may stabilize. If data support this hypothesis, clinicians could conceptualize reward and relief profiles as fixed after a particular age (around age 30) and apply fixed precision medicine recommendations based on these factors for the duration of their patient’s life. Longitudinal work measuring drinking motives from 18–35 provide initial support for this hypothesis, such that enhancement motives appear to decrease until age 29 and then stabilize, whereas coping motives remain more variable throughout the measurement period (Littlefield et al., 2010). This hypothesis would also influence long-term intervention applications for the present study, given that it assessed reward/relief drinking profiles at the age of mean AUD onset.

Regardless of the fixed versus changing nature of reward and relief profiles, it may be preferable to consider the addiction cycle and reward and relief drinking profiles as parallel rather than convergent theories. Specifically, the idea that relief drinkers are in the withdrawal-negative affect stage may be limiting. Given that the withdrawal-negative affect stage is defined by using a substance primarily to relieve withdrawal and negative affect caused by discontinuation of substance use, conflating the withdrawal-negative affect stage to relief drinking profiles may fail to capture the subset of individuals with AUD who initiate alcohol use to cope with stress and negative affect. Notably, drinking to relieve withdrawal versus stress may be difficult to differentiate given that over time distress due to withdrawal and distress due to external stressors become less distinguishable for individuals with AUD (Baker et al., 2004). However, examining motivations for initiating alcohol use and motivations for continued use later in the course of AUD may both have relevance to the field. Relatedly, the binge-intoxication phase of the addiction cycle is believed to be an earlier phase in the addiction cycle. However, there may be reward drinkers who have had AUD for multiple years who continue to engage in alcohol use for its rewarding effects (see (King et al., 2021)). As such, it may be limiting to develop research questions which view the addiction cycle and reward and relief drinking profiles as convergent theories.

Reward Relief Profiles as Targets for Brief Interventions prior to AUD

Profiling young adult, college drinkers for reward and relief drinking subgroups may be particularly helpful for targeted brief interventions. Extant work has already shown promise for brief interventions which include education on drinking to cope as beneficial in reducing drinking to cope with anxiety and depression (Blevins and Stephens, 2016), which may be beneficial for individuals in the high-reward/high-relief profile, and individuals in a low-reward/high-relief profile, if one were to be identified in college students. Individuals in the high-reward/low-relief profile, which was more commonly endorsed (with the highest amount of binge-drinking at follow-up), may benefit from an intervention targeting social enhancement and reward motives. Motivational enhancement has been shown to reduce drinking and alcohol-related consequences, particularly in women with high social and enhancement motives for drinking (LaBrie et al., 2008). Together this suggests that identifying individuals at risk for heavy drinking (high-reward/low-relief) and/or alcohol-related consequences (high-reward/high-relief) and providing targeted interventions may be a promising approach for improving outcomes in college-aged drinkers. One concern related to the profiling approach used by Hebden et al. (2023) is that it requires advanced statistical modeling, i.e., latent profile analyses, to identify reward and relief drinking profiles. While this approach is appropriate for research and is useful in identifying subgroups with multiple motives (e.g., high reward and high relief), it would be difficult to implement in university or clinical settings. Simplified approaches, such as cut-off scores or single-item questions, are needed to facilitate timely implementation to large-scale platforms and ultimately provide much-needed resources for at-risk individuals. In order to accomplish this goal, research will need to identify the most important items for each profile and develop easy scoring cut-offs, or binary yes/no options. Work in this area has already begun, such that a short, 10-item reward relief scale has been developed and cutoff scores for profiles have been proposed (Votaw et al., 2022). Future work will need to validate these cutoffs in additional samples and incorporate relief social items as reviewed above.

Reconsidering the Term “Phenotype”

It may also be worth reconsidering the use of the term phenotype in categorizing reward and relief drinking profiles. Phenotype, defined as the observable characteristics that are produced by the interaction of the genotype and the environment (Merriam-Webster, 2023). There is a growing discourse in the field of psychology highlighting the risks of adopting biological explanations for psychological phenomena before an empirical basis has been established (Miller, 2010, Thomas and Sharp, 2019) and similar concerns may arise with the usage of the genetically driven term phenotype. While the terminology arose out of a desire to parallel genetic profiles discussed in precision medicine, introducing biological terminology to a topic largely driven by psychological factors (i.e., motivations for drinking) may be misleading. For example, it may be unhelpful to compare the phenotypic tendency to develop keloid scars (which is determined based on genetic and environmental factors) to a process in which cognitions and emotions drive a behavioral outcome (drinking for reward or relief purposes). Specifically, there is no known genetic basis for reward versus relief drinking profiles. While some research has identified a genetic basis for social, coping, and conformity motives for alcohol use in women, there is no evidence for heritable influences on enhancement motives (Agrawal et al., 2008). Moreover, there is no known genetic basis for reward versus relief drinking profiles themselves. While a genetic basis may be identified in the future and some of the measures used to define the profiles have a genetic basis (Agrawal et al., 2008), the extant literature does not provide the level of confidence necessary to adopt a biological term for these profiles and thus, would be premature. As Miller (2010) asserts, rushing to reduce psychological phenomena to biological events may stymie scientific progress, preventing the field from empirically evaluating the relationships between biological and psychological phenomena. Similarly, adopting the term phenotype to describe drinking profiles may lead to erroneous assumptions about a genetic basis for these profiles. Moreover, if profiles are not stable across the lifespan or if they relate directly to stages of the addiction cycle, phenotype may be an imprecise term for describing reward and relief drinking profiles and instead, we may need terminology that acknowledges that these profiles are not fixed. Of note, we have previously used phenotype to describe reward and relief drinking profiles in our own work (Grodin et al., 2019), but have increasingly recognized that this terminology may be imprecise to describe these subgroups at present.

Are non-AUD samples appropriate for studying neurobiological theories of AUD?

The study of college samples that are not entirely comprised of individuals with AUD may hinder the translation of the neurobiological models. Preclinical animal models capture significant levels of alcohol intake and an abstinence syndrome marked by withdrawal. In the absence of these phenotypes being represented in the human condition, the forward translation of the neurobiological models is at best, incomplete. This implies that if hypotheses from these neurobiological models were not supported, it may be the results of a sample that does not reflect sufficient AUD symptomatology to adequately translate the full range of neurobiological adaptions captured in the preclinical studies.

Conclusion

In conclusion, Hebden et al. (2023) have furthered the emerging research into reward and relief drinking profiles by identifying subgroups in non-treatment-seeking young adults. The inclusion of reward-social drinking motives is an important contribution and reveals new questions surrounding the inclusion of similar relief-related social drinking motives. As discussed by Hebden et al. (2023), longitudinal and lifespan studies of reward relief profiles are sorely needed in order to move forward with applying precision medicine approaches. Targeted interventions may be implemented based on current reward and relief profiles, but it is unknown how stable these profiles are and if individuals will need to be re-profiled in the future to receive the appropriate intervention or treatment. An important consideration surrounds the use of non-AUD samples to identify reward and relief profiles, which are designed to translate neurobiological models of addiction. Understanding the impact of sample severity on these models will be critical for translation in this area. Ultimately, the goal of these and other profiling efforts is to refine and improve precision prevention and treatment efforts. As such, reward and relief profiles hold promise in both prevention and treatment areas in samples across the drinking and severity spectrum.