GLP-1 Agonists In Addiction: Rationale And Drug Development

By Tariq M. Ghafoor, MD, medical director, Aurora Behavioral Healthcare, Tempe, Arizona; founder, Transitions Center and AddictionRehab.com

The GLP-1 receptor agonist class — exemplified by semaglutide and liraglutide — has become one of the most consequential developments in metabolic medicine over the past decade. Their efficacy in type 2 diabetes and obesity is well established. What was not anticipated was the accumulating evidence suggesting that these agents may also influence reward-driven behaviors, including substance use disorders (SUDs).

This convergence is not coincidental. The neural circuits regulating food intake and those governing substance reinforcement share substantial anatomical and functional overlap. Both involve dysregulation of reward valuation, inhibitory control, and cue-driven motivation, raising the possibility that agents targeting metabolic signaling may exert broader effects on compulsive behavior.

The clinical implications extend beyond substance use. Behavioral addictions, including binge eating disorder, compulsive gambling, and hypersexual behavior, share neurobiological substrates with SUDs, and the mechanistic rationale for GLP-1 receptor activity in these domains is at least as compelling. This remains among the most underexplored frontiers in the field.

The addiction pharmacotherapy landscape continues to face significant unmet need. For drug discovery professionals, the key question is not simply whether GLP-1 receptor agonists will play a role in addiction, but how to translate a credible mechanistic signal into clinically validated, optimized therapeutic strategies.

Mechanistic Rationale

GLP-1 is an incretin hormone produced in intestinal L-cells and brainstem neurons. Its receptors are expressed not only in pancreatic and peripheral tissues but throughout the central nervous system, including the nucleus accumbens (NAc), ventral tegmental area (VTA), prefrontal cortex, and hypothalamus — regions central to reward processing, impulse control, and stress reactivity.

GLP-1 receptor activation in these areas appears to modulate mesolimbic dopaminergic activity, the primary substrate of addiction-related reinforcement. Preclinical evidence suggests several plausible mechanisms:

• attenuation of substance-induced dopamine release in the NAc
• reduction of cue-induced reinstatement of drug-seeking behavior
• modulation of stress-responsive circuits implicated in relapse vulnerability
• decreased impulsivity, a transdiagnostic risk factor across substance and behavioral addictions.

It is important to note that the precise neural substrates mediating these effects remain under active investigation. Evidence points toward both direct central GLP-1 receptor activity and indirect modulation via hindbrain-VTA projections, but these pathways have not been fully characterized in humans. The mechanistic picture, while biologically plausible, should not be overstated in the absence of definitive human neuroimaging or circuit-level data.

The overlap between food reward and substance reward circuits is particularly relevant to behavioral addictions. Binge eating disorder, for instance, involves many of the same dopaminergic and inhibitory control disruptions seen in alcohol and stimulant use disorders — and the GLP-1 system’s well-documented role in appetite and satiety signaling provides a direct mechanistic bridge that warrants more systematic investigation.

Current Evidence Landscape

Preclinical Data

The preclinical literature is substantial and relatively consistent across substance classes. In rodent models, GLP-1 receptor agonists have been shown to:

• reduce voluntary alcohol consumption and relapse-like drinking^1,2
• decrease nicotine self-administration and withdrawal-associated anxiety
• attenuate opioid conditioned place preference and reinstatement behavior
• reduce cocaine and amphetamine seeking in multiple behavioral paradigms.

These effects appear to be mediated, at least in part, by central GLP-1 receptor activity rather than peripheral metabolic changes — an important mechanistic distinction with implications for drug design.

Clinical And Observational Evidence

Clinical data remain limited but are generating meaningful signals. Key findings include:

• A randomized, placebo-controlled trial of exenatide in alcohol use disorder³ demonstrated reductions in cue-induced craving, particularly among individuals with higher body weight, suggesting a potential interaction with metabolic phenotype.
• An analysis using pharmacovigilance data from VigiBase identified associations between GLP-1 receptor agonist use and reduced reports of substance use-related adverse events.⁴
• Retrospective and claims-based analyses have reported lower rates of alcohol-related hospitalizations and tobacco use among patients prescribed semaglutide or liraglutide for metabolic indications, though these findings are subject to significant confounding and selection bias.
• Preliminary data from ongoing trials of semaglutide in alcohol use disorder are anticipated to provide more rigorous evidence within the next one to two years.

The current clinical evidence base should be regarded as hypothesis generating. No adequately powered, preregistered randomized controlled trial has yet established efficacy for any substance use disorder or behavioral addiction. Observational signals, while consistent in direction, cannot be interpreted as causal.

Addiction Treatment Gaps: Context For Development

The potential relevance of GLP-1 receptor agonists must be understood against the backdrop of long-standing limitations in addiction pharmacotherapy:

• Opioid use disorder (OUD): Buprenorphine and methadone are effective but require adherence to strict regulatory frameworks; naltrexone is underutilized due to tolerability and patient preference issues.
• Alcohol use disorder (AUD): Naltrexone and acamprosate demonstrate modest effect sizes; disulfiram is limited by adherence. Notably, naltrexone does demonstrate cross-SUD activity across AUD and OUD, providing a partial precedent for non-substance-specific pharmacotherapy, though not a true mechanism-agnostic approach.
• Nicotine use disorder: Varenicline and bupropion are moderately effective; cessation remains challenging at the population level.
• Stimulant use disorders: No FDA-approved pharmacotherapy exists.
• Behavioral addictions (gambling, binge eating, hypersexual behavior): Treatment is largely behavioral, with pharmacotherapy playing a limited off-label role.

GLP-1 receptor agonists, if validated, could represent a meaningful shift: targeting higher-order regulatory mechanisms shared across addictive behaviors rather than substance-specific receptor pathways. Whether this translates into a broad-spectrum effect or a more limited indication will depend critically on the specificity of neurobiological mechanisms and the rigor of forthcoming trials.

Drug Discovery And Development Implications

Repurposing Vs. Novel Development

Existing approved agents offer a near-term pathway via indication expansion, with established safety data and known pharmacokinetic profiles. However, current GLP-1 receptor agonists were not designed for CNS indications. Their central nervous system penetration is limited, and their dosing regimens and side effect profiles may not be optimal for addiction populations, particularly those with psychiatric comorbidities or without metabolic disease.

This gap creates a clear opportunity for next-generation compounds with enhanced CNS bioavailability, modified receptor selectivity, or dual-agonist profiles (e.g., GLP-1/GIP or GLP-1/glucagon co-agonists) that may provide differentiated activity on reward circuits.

Target Validation And Biomarkers

Addiction lacks the validated surrogate biomarkers available in metabolic disease (HbA1c, weight). Clinically meaningful endpoints — craving, relapse, functional outcomes — are variable and context dependent. Developing reliable biomarkers will be essential.

• Neuroimaging (fMRI-based reward circuit activation, dopamine receptor availability via PET) may serve as mechanistic proof-of-concept tools in early trials.
• Digital phenotyping — using passively collected behavioral data from smartphones or wearables — offers scalable, ecological measures of craving, sleep disruption, and behavioral patterns that may be more sensitive to treatment effects than traditional instruments.
• Pharmacogenomic stratification, given preclinical data suggesting differential effects based on metabolic phenotype, warrants early investigation.

Trial Design Considerations

Addiction trials face unique methodological challenges: high dropout rates, strong placebo responses, population heterogeneity, and difficulty standardizing behavioral co-interventions. Key design decisions will include:

• primary endpoint selection: reduction in use vs. abstinence vs. functional outcomes
• duration of treatment and post-discontinuation follow-up to assess durability
• stratification by metabolic phenotype, addiction severity, and psychiatric comorbidity
• integration with behavioral therapies, which may interact synergistically with pharmacological effects on craving and impulsivity.

Regulatory Considerations

Existing approvals provide a potential regulatory foundation for label expansion. However, addiction indications will require their own robust efficacy and safety demonstrations. The FDA’s Breakthrough Therapy designation pathway may be available if early trial data are compelling, potentially accelerating development timelines.

Safety Considerations

A rigorous assessment of GLP-1 receptor agonists in addiction cannot omit the current safety landscape, including neuropsychiatric signals that are directly relevant to addiction populations.

• Gastrointestinal tolerability (nausea, vomiting) is the most common adverse effect and may disproportionately impact adherence in populations already prone to treatment dropout.
• Suicidality signal: Both the FDA and EMA have reviewed reports of suicidal ideation and self-injurious behavior in patients receiving GLP-1 receptor agonists, including semaglutide. As of mid-2024, regulatory agencies concluded that available data did not establish a causal relationship, but active post-marketing surveillance continues. Given that depression, suicidality, and impulsivity are prevalent across addiction populations, this signal warrants specific attention in addiction trial designs, including prospective monitoring using validated instruments.
• Thyroid C-cell effects observed in rodent models remain a theoretical concern, though relevance to humans has not been established.
• Risk-benefit profile in individuals without metabolic disease — the likely target population for addiction indications — has not been defined and should be a primary focus of early Phase 2 trials.

Limitations And Unanswered Questions

Several fundamental questions remain unresolved and should temper current enthusiasm:

• Durability: Whether reductions in substance use persist beyond the treatment period, or whether continuous pharmacotherapy is required, is unknown.
• Mechanism attribution: The relative contribution of direct neurobiological effects vs. weight loss, improved metabolic health, or general behavioral changes has not been disentangled — a critical gap for development programs targeting non-obese populations.
• Phenotypic specificity: Addiction is not a monolithic condition. Individual differences in genetic vulnerability, comorbid psychiatric illness, substance class, and use pattern make it unlikely that a single mechanism will confer broad efficacy. Stratification and precision-based approaches will be essential.
• Behavioral addiction data: Evidence specifically addressing gambling disorder, compulsive sexual behavior, or other behavioral addictions remains largely absent from the clinical literature, despite strong mechanistic rationale.

Future Directions

The path forward requires coordinated effort across preclinical science, clinical research, industry, and regulatory strategy:

• Large-scale randomized controlled trials in clearly defined addiction populations, with adequate power, preregistered endpoints, and meaningful follow-up durations
• Mechanistic studies combining human neuroimaging with pharmacokinetic data to establish brain target engagement and clarify the neural substrates of behavioral effects
• Investment in CNS-optimized analogs or combination therapies, including co-agonist molecules designed with central reward circuit activity in mind
• Integration of digital phenotyping and passive behavioral monitoring as trial outcome measures, which may enable more sensitive, real-world detection of treatment response
• Dedicated investigation of behavioral addictions — an area of substantial clinical burden and commercial opportunity that remains insufficiently addressed by the current evidence base

More broadly, the GLP-1 receptor agonist story in addiction reflects a growing convergence between metabolic and neuropsychiatric drug discovery. The shared biology of appetite dysregulation and addictive behavior is not a coincidence — it is an underexplored design principle. Whether the current class of agents proves directly efficacious or primarily serves as a proof of concept for next-generation CNS-targeted compounds, the translational signal is scientifically meaningful and deserves continued rigorous investigation.

Conclusion

GLP-1 receptor agonists represent a scientifically credible and potentially consequential avenue in addiction medicine. The convergence of metabolic and reward neurobiology provides a robust theoretical foundation, and early data, while preliminary, are directionally consistent across substance classes and behavioral domains.

The field is at an inflection point. Translating this signal into clinically validated therapeutics will require rigorous trial designs, thoughtful patient stratification, frank attention to safety, and an honest accounting of what the current evidence does and does not establish.

For drug discovery professionals, the imperative is not to determine whether GLP-1 receptor biology is relevant to addiction — the preclinical and mechanistic data make a compelling case that it is. The imperative is to design the next generation of studies and compounds that can convert a promising signal into a meaningful treatment advance for one of medicine’s most persistent and undertreated conditions.

References

1. Egecioglu E, Engel JA, Jerlhag E. The glucagon-like peptide 1 analogue, exendin-4, attenuates the rewarding properties of psychostimulant drugs in mice. Addict Biol. 2013;18(4):597–609.
2. Farokhnia M, Grodin EN, Lee MR, et al. Exenatide, a glucagon-like peptide-1 receptor agonist, attenuates alcohol drinking in heavy-drinking individuals: a randomized, double-blind, placebo-controlled trial. Mol Psychiatry. 2021;26(9):4553–4570.
3. Klausen MK, Thomsen M, Wortwein G, et al. The role of glucagon-like peptide 1 (GLP-1) in addictive disorders. Br J Pharmacol. 2022;179(4):625–641.
4. Quddos F, Hassamal S, Haglund M, et al. Semaglutide and liraglutide reduce alcohol intake and relapse in animal models and produce divergent in vitro effects at the GLP-1 receptor. PLOS ONE. 2023;18(11):e0293730.

About The Author

Tariq M. Ghafoor, MD, is board certified in general adult and addiction psychiatry by the American Board of Psychiatry and Neurology. He serves as medical director at Aurora Behavioral Healthcare in Tempe, Arizona, where he oversees clinical programs in addiction medicine and co-occurring disorders. He is the founder of Transitions Center, a behavioral health treatment organization, and the digital platform AddictionRehab.com — a resource dedicated to evidence-based education and treatment access. His clinical interests include pharmacotherapy innovation, behavioral addictions, and the intersection of metabolic and reward neurobiology.

Disclosures: The author declares no conflicts of interest. No funding was received for preparation of this article.