GLP-1 Agonists and Compulsive Behavior
GLP-1 and Addiction
39% less food intake
Oral semaglutide reduced ad libitum daily energy intake by 39% versus placebo, demonstrating profound reward-driven appetite suppression.
Gibbons et al., Diabetes Obes Metab, 2021
Gibbons et al., Diabetes Obes Metab, 2021
View as imagePeople taking semaglutide for weight loss or diabetes report something unexpected: they lose interest not just in food, but in alcohol, cigarettes, online shopping, and gambling. These reports have accumulated across social media, clinical observations, and now emerging research, raising a question with substantial pharmacological implications. Are GLP-1 receptor agonists doing something broader to the brain's reward circuitry than simply suppressing appetite?
The answer appears to be yes. GLP-1 receptors are expressed throughout the mesolimbic dopamine system, the neural network that assigns value to rewards and drives motivated behavior.[1] When GLP-1 agonists activate these receptors, they do not just reduce the desire for food. They appear to reduce the pull of rewards more broadly. For a deeper look at where these receptors sit in the brain and what that means, see our pillar article on GLP-1 receptors in the brain's reward center.
Key Takeaways
- GLP-1 receptor activation modulates both appetite and reward-related brain areas in humans, including the insula and putamen, with effects that track with reduced desire to eat (van Bloemendaal et al., 2014)
- Oral semaglutide reduced daily ad libitum food intake by 39% and decreased cravings for sweet, salty, savory, and fatty foods in a 12-week RCT (Gibbons et al., 2021)
- Semaglutide reduced alcohol intake and relapse-like drinking in both male and female rats, demonstrating cross-reward effects beyond appetite (Aranas et al., 2023)
- Long-acting GLP-1 receptor agonists directly access and activate POMC/CART neurons in the hypothalamus, overlapping with cocaine- and amphetamine-regulated pathways (Knudsen et al., 2016)
- Postprandial GLP-1 increases after gastric bypass reduced food reward and brain reward system activation, with acute suppression of these hormones restoring reward-seeking behavior (Goldstone et al., 2016)
GLP-1 Receptors in the Reward System
The biological basis for GLP-1's effects on compulsive behavior starts with receptor distribution. GLP-1 receptors are not confined to the gut, pancreas, and hypothalamic appetite centers. They are expressed in the ventral tegmental area (VTA), nucleus accumbens, amygdala, and prefrontal cortex, brain regions that collectively form the mesolimbic reward circuit.[1]
Van Bloemendaal et al. (2014) used functional MRI to demonstrate that GLP-1 receptor activation modulates both appetite-related and reward-related brain areas in humans. They administered the GLP-1 receptor agonist exenatide to lean and obese participants and measured brain responses to food images. GLP-1 receptor activation reduced activity in the insula and putamen, regions involved in reward valuation and habit formation. The magnitude of these changes correlated with reduced desire to eat.[1]
This is a critical distinction. If GLP-1 agonists only suppressed appetite through hypothalamic satiety circuits, they would reduce hunger without affecting the rewarding quality of food or other pleasurable stimuli. The fMRI data shows they also dampen the reward signal itself. Food becomes less appealing, not just less needed. This mechanism is what opens the door to effects on non-food compulsive behaviors.
Beyond Appetite: Reducing All Categories of Food Craving
The breadth of GLP-1's reward effects is visible even within the domain of food. Kadouh et al. (2020) conducted a 16-week trial of liraglutide 3 mg versus placebo in adults with obesity. Liraglutide reduced cravings across every food category tested: sweet, salty, savory, and fatty foods. Participants also reported feeling fuller after meals, consumed less at ad libitum meals, and lost 2.7 kg of predominantly fat mass.[2]
Gibbons et al. (2021) found even more dramatic results with oral semaglutide. In a 12-week RCT in type 2 diabetes patients, oral semaglutide 14 mg reduced total daily ad libitum energy intake by 39% compared to placebo (treatment difference of 5,096 kJ). Participants ate less of everything, not just specific food types.[3]
The non-selective nature of this appetite suppression, affecting all food categories equally, is consistent with a reward-level mechanism rather than a specific taste or nutrient pathway. When a compound reduces the appeal of sweet, salty, savory, and fatty foods simultaneously, the effect is not on specific taste receptors. It is on the valuation process itself: the brain's assignment of reward value to incoming sensory signals. For more on how semaglutide produces weight loss, see our dedicated article.
Alcohol: The First Non-Food Evidence
The most developed evidence for GLP-1 effects on non-food compulsive behavior involves alcohol. Aranas et al. (2023) tested semaglutide in male and female rats using multiple alcohol-related behavioral paradigms. Semaglutide reduced voluntary alcohol intake and prevented relapse-like drinking after a period of abstinence. Both sexes showed the effect, addressing a limitation of earlier studies that had tested only male animals.[4]
The relapse prevention finding is particularly relevant. Relapse is the central challenge in alcohol use disorder treatment. A compound that reduces not just active drinking but also the resumption of drinking after a period of abstinence is targeting the core of addictive behavior: the persistent vulnerability to cue-triggered reward seeking.
Human clinical data on semaglutide and alcohol is emerging. Early-phase trials have shown that low-dose semaglutide reduced alcohol consumption during laboratory self-administration tasks, with participants reporting fewer drinks per drinking day and lower weekly craving scores. These findings are consistent with the animal data but remain preliminary. For a focused analysis of the semaglutide-alcohol relationship, see semaglutide and alcohol: why people on Ozempic lose interest in drinking.
The POMC/CART Connection: Where Addiction and Appetite Overlap
Knudsen et al. (2016) identified a specific neural target that helps explain why GLP-1 agonists affect both appetite and addiction-related behaviors. They demonstrated that long-acting GLP-1 receptor agonists like liraglutide directly access the arcuate nucleus of the hypothalamus and activate pro-opiomelanocortin (POMC) neurons. These neurons showed increased expression of cocaine- and amphetamine-regulated transcript (CART).[5]
The name "cocaine- and amphetamine-regulated transcript" is significant. CART was originally identified as a gene whose expression changed in response to cocaine and amphetamine exposure. It plays a role in both appetite regulation and reward processing. The fact that GLP-1 receptor agonists upregulate CART expression places them at a molecular intersection of appetite and addiction neurobiology. This is not a coincidence of naming. POMC/CART neurons are a genuine convergence point where metabolic signals and reward signals are integrated.
This convergence helps explain a clinical observation: the broader the reward suppression, the more appetite-related and non-appetite compulsive behaviors are affected simultaneously. For context on how this intersects with endogenous opioid reward systems, see beta-endorphin and the reward pathway. POMC neurons produce both alpha-MSH (which suppresses appetite) and beta-endorphin (which modulates reward), making them a dual-function node where GLP-1 agonists can simultaneously affect eating behavior and the broader hedonic experience.
Post-Surgical Evidence: GLP-1 and Food Reward
Goldstone et al. (2016) provided a different angle on GLP-1's reward effects by studying patients after Roux-en-Y gastric bypass (RYGB) surgery, which dramatically increases postprandial GLP-1 levels. RYGB patients showed reduced food reward behavior compared to non-surgical controls. When the researchers acutely suppressed GLP-1 (along with PYY and FGF-19) using octreotide, food reward behavior and brain reward system activation increased in RYGB patients but not in controls.[6]
This provides a causal test that is difficult to achieve with drug administration studies alone. The data shows that GLP-1 (alongside other gut hormones elevated by surgery) is actively suppressing food reward signaling. Remove it, and reward-seeking returns. This suppression-restoration design strengthens the case that elevated GLP-1 causally reduces reward system activity, rather than merely correlating with it.
Compulsive Shopping, Gambling, and Other Behaviors
The evidence for GLP-1 effects on non-substance compulsive behaviors (gambling, shopping, nail-biting) is at an earlier stage. Most current data comes from patient self-reports, social media analyses, and clinical observations rather than controlled trials. These reports consistently describe reduced interest in compulsive shopping, gambling, and other repetitive reward-seeking behaviors during GLP-1 agonist therapy.
The biological plausibility is strong. If GLP-1 agonists reduce the reward value assigned to incoming stimuli across the board, they should affect any behavior that is reinforced by the mesolimbic dopamine system. Gambling activates the same dopamine circuits as food and drugs. Compulsive shopping involves reward anticipation and dopamine-mediated reinforcement. The neural substrate is shared, so a pharmacological intervention that modulates this substrate should produce cross-domain effects.
The absence of controlled trial data means these behavioral effects remain unquantified. How large is the effect? Does it persist long-term? Does tolerance develop? Does it differ between behavioral compulsions (gambling, shopping) and substance addictions (alcohol, nicotine)? These questions are open. For a broader overview of the addiction treatment potential of this drug class, see could GLP-1 drugs treat addiction?.
Dopamine Protection: An Independent Pathway
Kim et al. (2009) identified a mechanism that may contribute to GLP-1's effects on compulsive behavior through a different pathway: neuroprotection of dopamine neurons. In a Parkinson's disease model, the GLP-1 receptor agonist exendin-4 protected dopaminergic neurons from degeneration by inhibiting microglial activation and reducing neuroinflammation.[7]
This neuroprotective effect is not directly about reward suppression. It is about maintaining the health and function of dopaminergic neurons. In the context of compulsive behavior, this finding raises the possibility that GLP-1 agonists may not only acutely dampen reward signaling but also protect the dopamine system from the neurotoxic effects of chronic overstimulation that characterize addiction. Chronic substance use damages dopaminergic neurons through oxidative stress and neuroinflammation. A compound that reduces both the compulsive drive and the neuronal damage could address addiction at two levels simultaneously.
This is speculative extrapolation from a Parkinson's model to addiction neurobiology. The dopaminergic neurons affected in Parkinson's (substantia nigra) and addiction (VTA) are different populations with different vulnerabilities. Whether GLP-1 neuroprotection extends to VTA dopamine neurons in the context of compulsive behavior has not been directly tested.
Stress, Mood, and the Compulsive Behavior Link
Guerrero-Hreins et al. (2021) reviewed GLP-1's effects on stress-related eating and mood. They found that while acute GLP-1 injection consistently stimulates the physiological stress response in rodents, chronic GLP-1 agonist exposure produces anxiolytic and antidepressant-like effects in animal models. In clinical studies, prolonged GLP-1 agonist treatment improved mood and psychological wellbeing in type 2 diabetes patients, though these improvements may be confounded by concurrent weight loss and improved glycemic control.[8]
The mood connection matters for compulsive behavior because many compulsive behaviors are triggered or maintained by negative emotional states. Compulsive eating, drinking, shopping, and gambling often function as maladaptive coping mechanisms for stress, anxiety, or depression. If GLP-1 agonists improve mood independently of weight loss (which remains unproven), they could reduce the emotional triggers that initiate compulsive behavior cycles, complementing the direct reward suppression mechanism.
Whether the mood improvement is a direct GLP-1 receptor effect or an indirect consequence of improved metabolic health is a question that current data cannot resolve. Separating the direct CNS effects of GLP-1 agonists from their downstream metabolic effects is one of the central challenges in this field.
What the Evidence Does and Does Not Support
The evidence base for GLP-1 agonists and compulsive behavior is heterogeneous. At the strongest end: food reward reduction is supported by multiple RCTs with fMRI confirmation of reward circuit modulation. Alcohol intake reduction has controlled animal data and early human trial signals. At the weakest end: effects on gambling, shopping, and other behavioral compulsions rely on observational reports without controlled comparisons.
The pharmacological mechanism connecting these domains is coherent. GLP-1 receptors in the mesolimbic system, POMC/CART neuron activation, and dopamine signaling modulation provide a plausible substrate for cross-domain reward effects. Whether this pharmacological coherence translates into clinically meaningful treatment for non-substance compulsive behaviors requires trials that have not yet been completed.
There is also a risk of over-interpretation. Not every anecdotal report of reduced compulsive behavior on GLP-1 therapy reflects a direct drug effect. Weight loss itself improves mood, self-efficacy, and engagement in health behaviors. Metabolic improvements reduce inflammation and improve brain function. Some of the reported behavioral changes may be consequences of these secondary effects rather than direct reward suppression. The question of whether GLP-1 drugs might trigger eating disorders in vulnerable individuals adds another dimension to the risk-benefit assessment.
The Bottom Line
GLP-1 receptor agonists modulate reward circuitry beyond appetite suppression, with controlled evidence for food reward reduction (fMRI-confirmed) and alcohol intake reduction (animal and early human data). Effects on gambling, shopping, and other compulsive behaviors are biologically plausible given GLP-1 receptor expression in mesolimbic dopamine circuits but lack controlled trial data. The POMC/CART neuron pathway provides a molecular intersection where appetite and addiction neurobiology converge, explaining the cross-domain effects.