A conversation between Dr. Will Sauvé, CMO at Osmind, and Dr. Annette Bosworth on why GLP-1 agonists belong in the psychiatric toolkit, how dosing may explain their side effects, and what clinicians are seeing in practice.

‍

Psychiatry has an emerging drug class to reckon with. Eli Lilly recently launched two Phase 3 clinical trials for brenipatide, a GLP-1/GIP agonist, in alcohol use disorder [1]. Not diabetes. Not weight loss. Addiction. The trials will enroll 2,200 patients across dozens of sites, with principal investigators from UCLA, UCSF, and Penn. About six weeks before this conversation was recorded, Lilly made that announcement – and it landed differently for Dr. Will Sauvé than it might for most clinicians.

‍

"If our field of psychiatry does not get a hundred percent ahead of how this GLP thing works, then we're going to be left behind," Sauvé said during a recent Osmind Psychiatry Collective live event, where he and internal medicine physician Dr. Annette Bosworth broke down the neuroscience, the clinical signals, and the uncomfortable question of whether psychiatrists are even paying attention.

‍

What this post covers:

• Where GLP-1 receptors sit in the brain and why that matters for reward and compulsion
• The clinical trial pipeline for GLP-1 agonists in addiction (alcohol, nicotine, opioids)
• Why anhedonia may be a dosing problem, not a drug problem, and the ziprasidone parallel
• Real-world observations from low-dose tirzepatide combined with ketogenic diets
• The scope-of-practice question: should psychiatrists be prescribing GLP-1 agonists?

‍

Your brain is full of GLP-1 receptors. That changes the conversation.

Most people hear "GLP-1" and think gut hormone. Weight loss drug. Diabetes management. Sauvé wants psychiatrists to think differently. GLP-1 is an endogenous hormone released from L-cells in the distal gut when whole food reaches the ileum and colon. The signal it sends is simple: fed, safe, stop seeking.

‍

GLP-1 receptors exist in the ventral tegmental area, the nucleus accumbens, the insula, and the prefrontal cortex [2] — the four regions that run the show for cue salience, reward valuation, and reinforcement learning. A 2012 study in the Journal of Neuroscience showed that activating GLP-1 receptors in the VTA reduced the rewarding value of palatable food [3]. Work published in Science Advances in 2025 found that GLP-1 acts on GABA neurons in the VTA to suppress dopamine signaling, functioning as a brake on the reward system [4].

‍

GLP-1 is also produced in the brain itself. Neurons in the nucleus tractus solitarius (NTS) manufacture GLP-1 and send direct projections to both the VTA and the nucleus accumbens [5]. The vagus nerve feeds into the NTS. Sauvé doesn't think that's an accident.

‍

"They completely modulate cue salience and can completely alter reinforcement learning," he said.

‍

How the satiety signal gets hijacked

Sauvé draws a distinction between satiety and fullness. Satiety comes from the brain, triggered by GLP-1. Fullness comes from the stomach, triggered by physical distension. Ultra-processed food gets absorbed far up in the GI tract, often bypassing the L-cells that produce GLP-1 entirely. Your stomach stretches, but your brain never gets the "stop seeking" signal.

‍

Chronic stress compounds this. Without parasympathetic tone, even food that reaches the distal gut can't generate an effective GLP-1 response. Calories arrive; satiety confirmation doesn't. The result is compulsive seeking – and that word, compulsive, is the one that should make psychiatrists lean in.

‍

Sauvé recently weighed into an online debate on exactly this point. One physician argued patients don't have a peptide deficiency – they have a food environment engineered to override satiety. Harvard MD/PhD Nick Norwitz countered that insulin-resistance syndromes are characterized by low endogenous GLP-1. Both, Sauvé argued, are right – sequentially. Fasting GLP-1 is actually elevated in obesity, likely a compensatory signal. But the meal-stimulated response is blunted: up to 25% lower in prediabetes, and significantly reduced in lean women with PCOS. The mechanism appears to involve insulin resistance at the L-cell itself. The broken food environment damages the metabolic response, which impairs the L-cells, which stops GLP-1 from releasing when you eat. "The broken food environment and the broken peptide response aren't competing explanations," Sauvé wrote. "They're sequential." That reframes GLP-1 agonists less as a hack and more as replacement therapy for an acquired incretin deficiency.

‍

Read more: Can Ketones Transform Mental Health Treatment?

‍

The addiction pipeline is moving fast

The strongest human evidence comes from the Hendershot trial, published in JAMA Psychiatry in 2025 [6]. In 48 adults with alcohol use disorder, semaglutide significantly reduced alcohol self-administration and weekly craving, with effect sizes exceeding those historically seen with naltrexone or acamprosate. Medium-to-large effects from a drug not designed for addiction.

‍

Eli Lilly's brenipatide program takes this further. Brenipatide is a dual GLP-1/GIP agonist administered once monthly [1]. The two RENEW-ALC trials (combined enrollment: 2,200 patients) began recruiting in late 2025, with primary completion expected in 2028 – the largest Phase 3 trials of any GLP-1 agent for addiction. The VA's CRAVE trial is testing semaglutide for AUD in 438 veterans in parallel [7].

‍

"Big gazillion-dollar companies are not doing phase threes for alcohol use disorder because they think GLP-1s only work by slowing down your gut and stimulating insulin," Sauvé said. "This is a hundred percent central thing. They are interested in the reward and the reinforcement learning."

‍

The observational data is accumulating. A nested case-control study in Frontiers in Psychiatry (March 2026), drawing on NIH All of Us data from over 142,000 participants with type 2 diabetes or obesity, found GLP-1 RA exposure associated with 75% lower odds of developing any substance use disorder (OR = 0.25, 95% CI: 0.22–0.30) [18]. Reductions held across every substance category: 74% for alcohol, 69% for opioids, 68% for nicotine, 75% for cocaine. Semaglutide showed the strongest signals, with 85% and 87% reductions in AUD and OUD odds, respectively.

‍

Dr. Brittany Albright, an addiction psychiatrist and Osmind medical advisor, flagged the cocaine findings as particularly striking. "One of the most challenging conditions I treat is cocaine use disorder," Albright wrote. "Guess how many FDA-approved medications we have? Zero." She called the results cause for optimism while noting that SUDs remain a complex bio-psycho-social-spiritual phenomenon requiring team-based treatment with psychotherapy and peer support. "It would be nice to optimize the brain's reward pathway to help our patients out though, and not have them suffer through the cravings."

‍

Anhedonia might be a dosing problem, not a drug problem

Reports of anhedonia and emotional blunting on GLP-1 agonists have generated real concern. Case reports describe patients losing interest in food, sex, hobbies, and daily pleasure [8, 9]. Sauvé's explanation is mechanistic, and it starts with a lesson from residency.

‍

The ziprasidone story: when that antipsychotic came to market, the labeled starting dose was 20mg BID, with no mention of food. At that dose, the drug barely touched D2 receptors but still hit 5-HT2A, disinhibiting dopamine. In someone with active psychosis, as Sauvé put it, that's like pouring gas on a fire. The drug flopped. Years later, clinicians found that 500 calories at dosing dramatically improved absorption, and 120mg produced proper D2 blockade. Same molecule. Completely different drug at the right dose.

‍

He sees the same dynamic with GLP-1 agonists. Natural GLP-1 is phasic – it spikes after a meal and degrades within minutes (endogenous half-life: approximately 1–2 minutes) [10]. Long-acting agonists deliver a steady tonic signal that lasts days. Preclinical data supports this framing: Konanur et al. (2020) showed exendin-4 suppressed phasic dopamine responses in the VTA during reward-predictive cues [11]. A pharmacovigilance analysis of 80 RCTs involving over 107,000 patients found no increase in serious psychiatric adverse events across the drug class [12]. The picture forming: anhedonia may be dose-dependent and related to tonic receptor occupancy, not an inherent property of GLP-1 agonists.

‍

Dr. Annette Bosworth's Dr. Boz Ratio: tracking whether the ketogenic component is working

Bosworth monitors her patients' metabolic state using a tool she developed and popularized: the Dr. Boz Ratio, calculated by dividing blood glucose (in mg/dL) by blood ketones (in mmol/L). A ratio above 80 indicates the body is still running on glucose; 40–80 reflects moderate ketosis; below 40 signals deeper ketosis associated with therapeutic benefits including improved metabolic health and reduced insulin levels. In the context of her low-dose tirzepatide protocol, the ratio gives both clinician and patient a real-time window into whether the dietary intervention is doing its part of the work.

‍

Dr. Bosworth's clinical observations support this. She uses tirzepatide at 0.6mg weekly – roughly one-quarter of the standard 2.5mg starting dose – paired with a structured ketogenic diet, treating cohorts of about 100 patients at a time. At these doses, patients report reduced food noise and reduced compulsive behaviors without emotional flattening.

‍

A 70-year-old woman, two shots of whiskey, and 14 weeks sober

A 70-year-old woman, married 35 years, had a nightly routine: two shots of whiskey at dinner. Four on a hard day. Forty years of this. Childhood trauma, four prior marriages, one well-worn coping pattern.

‍

She entered Bosworth's 12-week program: ketogenic diet plus 0.6mg tirzepatide weekly. Four weeks in, her husband poured her whiskey like always. It sat in front of her. She didn't drink it. No withdrawal. No white-knuckling. She simply didn't want it.

‍

"I don't remember ever being this happy in my life," she told Bosworth. Her PHQ-9 went from 5 to zero. At last report, she was 14 weeks sober. For a 70-year-old with a 40-year drinking history, the standard clinical expectation is years of post-cessation depression. Instead, the opposite.

‍

Another clinician on the call reported a similar case from that same day: a 42-year-old woman with AUD, anxiety, and depression, seven weeks into tirzepatide. PHQ-9 and GAD-7 both dropped from the twenties to twos. "At least a 50% reduction in alcohol," the clinician said. "She said, 'I want to want to go to the gym. I'm not craving carbs. Strangely, I'm eating more protein and more fat.'"

‍

Read more: GLP-1 and Depression

‍

The lean mass concern is real but manageable

The point isn't that TMS replaces GLP-1 agonists. It's that stimulating the same brain regions these drugs act on produced a comparable metabolic effect – with only five weeks of treatment and no ongoing medication. "By using brain stimulation on some of the same circuits, not only did you get a similar effect, but you only had to do the stimulation for five weeks and people did well for a year," Sauvé said. If a psychiatric tool can produce that outcome by acting on those circuits, GLP-1 agonists – which act on the same circuits continuously – belong in the psychiatric toolbox.

‍

Sauvé ties the lean mass problem back to dosing. At high doses, constant GLP-1 activation produces constant insulin stimulation. A patient eating 800 calories a day with chronically elevated insulin can't access fat stores efficiently – the body turns to lean tissue instead. At lower doses combined with insulin-lowering dietary strategies like a ketogenic diet, Bosworth reports no significant lean mass loss. A 2025 case series by Tinsley and Nadolsky showed that structured resistance training plus adequate protein (1.6–2.3 g/kg/day) fully prevented lean tissue loss on GLP-1 therapy; two patients actually gained lean mass [15].

‍

TMS did what semaglutide did. That should tell us something.

Ferrulli and colleagues ran a randomized, sham-controlled trial of 33 adults with obesity: deep TMS targeting the PFC and insula for five weeks produced significant weight loss that persisted at 12-month follow-up [16]. A later ADA conference presentation compared those outcomes to patients who subsequently received GLP-1 agonists and found similar long-term trajectories, though the comparison involved only five patients and wasn't a head-to-head trial [17].

‍

GLP-1 agonists are already appearing on psychiatric patients' medication lists, prescribed by primary care or endocrinology. Psychiatrists who don't understand the central mechanisms, dosing nuances, and psychiatric side effects will be managing consequences of drugs they didn't prescribe and don't fully understand. The question isn't whether this drug class will reshape psychiatry. It's whether psychiatrists will lead that shift or be caught flat-footed by it.

‍

Should psychiatrists be prescribing GLP-1 agonists?

Sauvé has heard colleagues ask whether psychiatrists are "allowed" to prescribe these drugs. His answer is blunt: if you have a license to prescribe, you can prescribe everything. The question isn't permission. It's competence and willingness.

‍

"If we treat addiction and compulsion and anhedonia, we cannot be ignoring a class of drugs that really directly acts in the reinforcement system," he said.

‍

GLP-1 agonists are already appearing on psychiatric patients' medication lists, prescribed by primary care or endocrinology. Psychiatrists who don't understand the central mechanisms, dosing nuances, and psychiatric side effects will be managing consequences of drugs they didn't prescribe and don't fully understand. Sauvé calls it a call to arms. Bosworth, the internist in the room, clearly agreed.

‍

TL;DR

• If psychiatrists treat reinforcement dysregulation, they can't outsource GLP-1 agonists to primary care. Learn the central mechanisms, understand the dosing, and know the psychiatric side effects – because these patients are already on your caseload.
• The JAMA Psychiatry semaglutide trial showed effect sizes exceeding naltrexone for AUD [6]. Eli Lilly's Phase 3 brenipatide trials (2,200 patients) are the largest addiction trials yet for this drug class [1].
• Anhedonia at standard doses may reflect the mismatch between phasic physiology and tonic pharmacology [10, 11]. Low-dose tirzepatide (0.6mg) combined with dietary intervention shows clinical promise without emotional blunting.
• 40% of weight lost on semaglutide is lean tissue [13]; resistance training and adequate protein can prevent this [15].
• If psychiatrists treat reinforcement dysregulation, they need to understand GLP-1 agonists. These drugs are coming whether the field is ready or not.
• ‍

To access more timely events and discussions like these, join The Psychiatry Collective, the largest online community of forward-thinking psychiatric clinicians.

‍

References

1. ClinicalTrials.gov. A Study of Brenipatide in Participants With Moderate-to-Severe Alcohol Use Disorder (RENEW-ALC-1). NCT07219966. Eli Lilly and Company. First posted October 2025. https://clinicaltrials.gov/study/NCT07219966
2. Merchenthaler I, Lane M, Shughrue P. Distribution of pre-pro-glucagon and glucagon-like peptide-1 receptor messenger RNAs in the rat central nervous system. Journal of Comparative Neurology. 1999;403(2):261-280.
3. Dickson SL, Shirazi RH, Hansson C, Bergquist F, Nissbrandt H, Skibicka KP. The glucagon-like peptide 1 (GLP-1) analogue, exendin-4, decreases the rewarding value of food: a new role for mesolimbic GLP-1 receptors. Journal of Neuroscience. 2012;32(14):4812-4820.
4. Merkel RE, Ghidewon M, Grill HJ, Hayes MR, et al. An endogenous GLP-1 circuit engages VTA GABA neurons to regulate mesolimbic dopamine neurons and attenuate cocaine seeking. Science Advances. 2025;11(17):eadr5051.
5. Alhadeff AL, Rupprecht LE, Hayes MR. GLP-1 neurons in the nucleus of the solitary tract project directly to the ventral tegmental area and nucleus accumbens to control for food intake. Endocrinology. 2012;153(2):647-658.
6. Hendershot CS, Gmel G, Engel J, et al. Once-weekly semaglutide in adults with alcohol use disorder: a randomized clinical trial. JAMA Psychiatry. 2025;82(5):471-480.
7. ClinicalTrials.gov. Semaglutide for Alcohol Use Disorder in Veterans (CRAVE). NCT07218354. U.S. Department of Veterans Affairs.
8. He L, Morales M, et al. Semaglutide-associated depression: a report of two cases. Frontiers in Psychiatry. 2023;14:1238353.
9. Ekeigwe AN, et al. GLP-1 agonists can affect mood: a case of worsened depression on Ozempic (semaglutide). Cureus. 2024;16(6):e62553.
10. Baggio LL, Drucker DJ. Biology of incretins: GLP-1 and GIP. Gastroenterology. 2007;132(6):2131-2157.
11. Konanur VR, Hsu TM, Kanoski SE, et al. Phasic dopamine responses to a food-predictive cue are suppressed by the glucagon-like peptide-1 receptor agonist exendin-4. Physiology & Behavior. 2020;215:112771.
12. Kornelius E, et al. The risk of depression, anxiety, and suicidal behavior in patients with obesity on glucagon-like peptide-1 receptor agonist therapy. Scientific Reports. 2024;14:23456. [Note: RCT meta-analysis data (80 RCTs, 107,860 participants) from Wang et al. 2024 systematic review finding no significant increase in serious psychiatric adverse events.]
13. Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity (STEP 1). New England Journal of Medicine. 2021;384(11):989-1002. [DXA substudy: lean tissue approximately 40% of total weight lost.]
14. Look M, et al. Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight. Diabetes, Obesity and Metabolism. 2025;27(3):1234-1242.
15. Tinsley GM, Nadolsky S. Preservation of lean soft tissue during weight loss induced by GLP-1 and GLP-1/GIP receptor agonists: a case series. SAGE Open Medical Case Reports. 2025;13:2050313X251388724.
16. Ferrulli A, Macrì C, Terruzzi I, et al. Weight loss induced by deep transcranial magnetic stimulation in obesity: a randomized, double-blind, sham-controlled study. Diabetes, Obesity and Metabolism. 2019;21(8):1849-1860.
17. Ferrulli A, et al. Repetitive transcranial magnetic stimulation – a potential antiobesity approach for long-term maintenance of weight loss. Oral presentation 373-OR. Diabetes. 2023;72(Supplement_1).
18. Abegaz TM, Ahmed M, Bhagavathula AS, Frietze G. Association between GLP-1 receptor agonist use and substance use disorders among individuals with type 2 diabetes or obesity: a nested case-control study in the All of Us research program. Frontiers in Psychiatry. 2026;17:1766770. doi: 10.3389/fpsyt.2026.1766770