Can Ozempic Cure Addiction?
Mary started drinking regularly in the early nineties, when she was thirteen. Her father had recently married a Danish woman and moved the family to Denmark, which has one of the highest teen-drinking rates in Europe. (Sixteen-year-olds are allowed to buy beer.) As Mary got older, her alcohol consumption accelerated—she could drink eighteen beers in a sitting—and a cycle of inebriation and hangovers blurred her days. She tried rehab, Alcoholics Anonymous, and a medication called Antabuse, which provokes nausea in combination with alcohol. None of them worked for her. Mary (a pseudonym) got married, had a child, and became a social worker; her alcohol tolerance increased so much that, when she binged, her friends barely registered that she was buzzed. She remembered a close friend who also drank heavily telling her, “I don’t get it. Why do you say you have an alcohol addiction?”
“The difference between you and me,” Mary replied, “is that I think about alcohol all the time—when to drink, how to drink, where to drink, how much to drink.”
Last year, at a bar, Mary noticed that the same friend had hardly sipped her drink. She told Mary that she’d started taking Ozempic for weight loss. “If I have more than two beers now, I go outside and barf,” the friend said. Mary was perplexed. Ozempic, or semaglutide, originated as a diabetes medicine; more recently, as an obesity treatment, it has made its manufacturer, Novo Nordisk, one of Europe’s most valuable companies. What did it have to do with drinking? The next day, Mary saw an advertisement on Facebook: a nearby clinical trial was studying semaglutide’s effects on alcohol addiction. She enrolled.
Once a week, researchers blindfolded Mary and injected her with a solution. They didn’t tell her whether she’d received semaglutide or a placebo. But, a few weeks into the study, she lost her taste for beer. She switched to white wine, then stopped drinking altogether. “People talk about Ozempic getting rid of food noise,” she told me. “For me, it took out alcohol noise.”
One way to think about addiction is as a battle between first- and second-order desires. On some level, you want a drink; on another, you don’t want to want a drink. Strikingly, Mary felt that semaglutide had not only curbed her first-order desire to drink but also helped her act on second-order desires. “Because alcohol was no longer an issue, all of this energy was released,” she said. “I finally had the mental space to think, What type of life do I want?” She’d had marital problems for years; within months of starting the trial, she left her wife. “I’d never dared make that decision,” she told me.
As the researchers increased the dose of the medication, however, Mary’s appetite vanished. Some days, she ate as few as two hundred calories. She had been overweight when the study began; she lost fifty-five pounds in five months. One afternoon, she cried in a grocery store because she was surrounded by food and didn’t want to eat. “I knew what I had to do, but there was no hunger,” she said. “I felt desperate.” Mary was supposed to take the drug for six months, but, after the weight loss continued even on a lower dose, she had to stop the drug two weeks early. Still, she maintained her lower weight with a strict diet-and-exercise regimen. Her cravings for alcohol returned, but she felt better able to control them. “These drugs can give you the break you need to jump-start changes in your life,” she told me. “People know how much they affect your body. I don’t think they realize how much they affect your mind.”
When I asked addiction researchers about the promise of medications like Ozempic, they sounded as enthusiastic as Mary. GLP-1 medications—colloquially known as GLP-1s—mimic a naturally occurring hormone called glucagon-like peptide 1, which scientists have historically associated with digestion. In the traditional telling, after a meal, GLP-1 attaches to receptors in the pancreas, the gastrointestinal tract, and the brain; it stimulates the release of insulin, slows the passage of food through the stomach, and signals to our brains that we’re full. It breaks down within minutes. But it’s increasingly clear that GLP-1 affects much more than eating. GLP-1 drugs, which bind to the body’s receptors for hours or days, are now being studied for all sorts of addictions.
The results from the trial that Mary participated in should come out later this year, but a small study published in 2021 showed that smokers given exenatide, the first GLP-1 on the market, were nearly twice as likely to stop smoking as those given a placebo. An analysis of hundreds of thousands of electronic health records found that people with an addiction to opioids who happened to be prescribed GLP-1s were forty per cent less likely to overdose. Some scientists think that the medications may even help with behavioral addictions, such as gambling and compulsive shopping. Research into GLP-1s, scientists hope, could deepen our understanding of what addiction is. “Essentially all addiction medications to date have been specific to a particular type of addiction,” Heath D. Schmidt, a neuroscientist at the University of Pennsylvania, told me. “GLP-1s might be telling us that there’s some kind of universal pathology when it comes to addiction. And that they’re part of how we fix it.”
America’s first randomized trial of semaglutide for heavy alcohol use began in 2023 at the medical campus of the University of Colorado Anschutz, not far from Denver. On a frigid morning in December, I met its principal investigator, Joseph Schacht, in a sleek glass building within view of the snow-capped Rocky Mountains. Schacht, a lean clinical psychologist with boyish features, has spent much of his career searching for medications to treat alcohol addiction. Since the turn of the twenty-first century, drinking-related deaths have roughly doubled in the U.S., yet no new medication has been approved for alcohol addiction in decades. Schacht’s field has faced a basic challenge: risky drug use is usually hidden and stigmatized, which makes it difficult to study and treat. “I fundamentally believe that a medication you can take in the privacy of your home—that doesn’t require disclosing to others that you have this problem—is essential to be able to offer,” he told me.
Schacht escorted me to a storage room unlike any I’ve visited at a research laboratory. Stacked on a countertop were cocktail mixers: cranberry juice, pineapple juice, Fever-Tree pink-grapefruit soda. Inside a wooden cupboard were shelves of alcohol: Chardonnay, Zinfandel, Tito’s vodka, Jose Cuervo tequila, Mount Gay rum. Where another lab might have pipettes, this one had shot glasses. Schacht had agreed to put me through a cue-reactivity test—which exposes participants to a drug to see how strongly they want it—and to image parts of my brain involved in cravings for alcohol. Before the test, his staff will ask participants what they like to drink. (When participants have specified an ultra-top-shelf liquor, Schacht has been known to ask, “O.K., how often do you actually drink that?”)
My drink of choice is a Negroni. Across the hall from the storage room, I sat down at a table with an iPad on it. Melina Kilen, a disarmingly confident research assistant in a blue sweater and long earrings, arrived with a tray of bartending supplies. She unscrewed the cap from a bottle of gin, poured a shot into a glass over ice, and added sweet vermouth and Campari. I watched the liquid change from clear to golden to amber. Finally, she ran an orange peel around the rim, dropped it into the drink, and left the room.
A man’s voice emanated from a nearby laptop: “When you hear a high tone”—ding!—“pick up the glass, bring it to your nose, and smell the beverage. When you hear a low tone”—dong!—“stop smelling the beverage and move the glass away from your nose. Do not drink!” For five minutes, I followed instructions, inhaling the drink’s bitter and bright aromas and then putting the glass down again. I was hungry and tired, having flown in late the previous night, and I was surprised that even in a laboratory I wanted to take a sip. Then the test was over; the untouched Negroni was rather cruelly taken away.
On the iPad, I answered some questions. The drink was alluring, I wrote, but not in an all-consuming way. In contrast, one of Schacht’s trial participants, a former I.C.U. nurse whom I’ll call Susan, told me that her cue-reactivity test was “torture.” She remembered sniffing a glass of wine for what seemed like forever. “I wanted that drink so bad,” she said. Before the trial, she’d often have a bottle of wine and a margarita or two per night.
Addiction is more than a subjective sensation; its patterns can be observed in the brain. Schacht told me that when specific brain regions respond more to alcohol than to, say, food, that’s “very predictive of alcohol-use disorder.” He led me to a nearby building that housed MRI scanners. After changing into scrubs, I lay on a padded table that glided into a cylindrical machine. I was holding a remote control that allowed me to rate the strength of my urge to consume alcohol. “Thumb for extreme,” a technician told me. “Pinkie for none.”
For roughly an hour, I was shown images on a screen. I gazed at beautiful landscapes and abstract paintings to establish a baseline, but most of the images were of foods or drinks: glasses of frothy beer nestled in snow; a bottle of red wine next to a bowl of grapes; celery, apples, doughnuts, burgers. Occasionally, a prompt asked me to rate my desire to drink.
Later, on a computer, Schacht pulled up scans that showed what addiction looks like in the brain. Harmful alcohol use is strongly correlated with bright-red splotches in two specific areas, he said. The first, a circle where the hemispheres meet, is the ventral striatum. “That’s where all these dopamine-producing neurons release their dopamine in response to alcohol and other drugs,” Schacht said. Notably, the area is also rich in receptors for naturally occurring GLP-1. The other area, an elongated oval near the front of the brain, is in the medial prefrontal cortex, which is involved in higher-level evaluations of rewards. “It says, ‘Hey, this is something worth chasing!’ ” Schacht told me. Together, the splotches represent a kind of neural signature for craving. “It’s what we hope GLP-1 medications can dampen,” he said.
I was struck by how different the images of my brain looked: largely black and white, with only a faint dusting of color. “Good news,” Schacht said. “You don’t have alcohol-use disorder.” On some level, I’d already understood addiction to be a treatable disease, not a personal failing. Still, these scans helped me appreciate how deeply addiction is rooted in neurobiology. A mere photograph of alcohol—to say nothing of a sip—was enough to send a person’s reward centers into a frenzy. Our decisions still matter; the red circles are known to grow brighter when they’re conditioned by repeated use of a drug. But it was counterproductive, even biologically incoherent, to shame a person for having overwhelming cravings. If GLP-1 drugs prove successful, they might reset not only people’s addictions but also society’s perceptions.
Schacht is still analyzing data from the fifty people in his trial, but he shared some preliminary findings. Before the study, participants consumed nearly seven drinks a day; two months later, people on semaglutide drank half as much. The number of days on which they drank heavily—four or more drinks for women, five or more for men—fell from roughly two-thirds to a quarter. “My emotional response to alcohol was totally cut off,” Susan told me. Before she joined the study, if she found herself in the wine aisle of a grocery store, she’d buy five bottles. Afterward, she told me, “my brain recognized it—like, Oh, that’s what you used to want—but my body had no desire for it.” She said that when she repeated the cue-reactivity test at the end of the trial, the glass of wine “had absolutely no pull.” Strikingly, the drug didn’t make people any more likely to abstain from alcohol. It only led them to consume fewer drinks. “This, to me, is the most meaningful result,” Schacht said. “Most patients don’t want to be completely abstinent. They want to drink like a ‘normal person.’ ”
Aristotle argued that the path to a life well lived runs through moderation. Courage lies somewhere between cowardice and recklessness, generosity between stinginess and extravagance. “It is best to rise from life as from a banquet, neither thirsty nor drunken,” he is often quoted as saying. GLP-1s have earned a reputation as “moderation molecules.” For many people, these drugs might restore a sense of control, offering a path away from excess. “GLP-1s are the most exciting drugs for alcohol-use disorder I’ve seen in my career,” Schacht said. “This could be our Prozac moment.”
This doesn’t mean that GLP-1s will be a panacea. When Prozac débuted, in the late eighties, it represented a fundamental shift in the treatment of depression. Millions of people who lacked effective options for a distressing condition suddenly had one; by 2008, antidepressants were the third most common prescription drugs in the U.S., behind cholesterol medications and painkillers. Yet Prozac’s legacy is complicated. Its reputation as a miracle drug outpaced evidence of its medical benefit, and it propagated a narrative of depression as a chemical imbalance that later proved incomplete, if not untrue. Many people took it and said that they no longer felt depressed, but many also said that they no longer felt like themselves.
Recently, I spoke with a fortysomething woman I’ll call Jessica, who struggled with untreated A.D.H.D. for decades. She told me that her “treatment” was food and alcohol. For much of her life, she binged on both; when she entered menopause, she started to put on weight. She saw an advertisement for a GLP-1 medication from a telehealth company and decided to try it out. A few days later, she received a box of syringes in the mail.
Jessica lost twenty pounds in the four months after she started injecting herself, but her drinking was largely unaffected. Meanwhile, she grew more anxious. Her husband noticed her mood worsening; she lost interest in activities she’d previously enjoyed. An avid gardener, she had planned to refresh her yard with six Japanese maple trees, but now she couldn’t bring herself to plant them. “That’s when I was, like, Whoa, this isn’t my normal,” she said. She covered the windows looking into her garden so that she wouldn’t have to see her withering plants. For her, the GLP-1 medication didn’t moderate so much as agitate and numb. Eventually, she decided that the psychological side effects weren’t worth the weight loss. She stopped taking the medication and regained the weight, but her mood improved. “It was pretty immediate—the return to feeling like myself,” she said.
Jessica seemed to be describing anhedonia, a struggle to experience pleasure. Last year, the New York Times interviewed two dozen people whose relationships were affected by GLP-1 drugs, including a woman who lost interest in sex. In “Magic Pill,” the journalist Johann Hari attributes a muted mood to Ozempic. “I didn’t feel as excited for the day ahead,” he writes. “I was often emotionally dulled.” On Reddit, users discuss numerous psychological changes. “I wasn’t enjoying my family at all, or myself, or my life,” one person wrote. “I don’t want to be a zombie.” Jens Juul Holst, a Danish scientist credited with helping discover GLP-1, worried, in 2023, that GLP-1 medications would make life “so miserably boring that you can’t stand it any longer and you have to go back to your old life.”
These anecdotes raise the possibility that the so-called moderation molecule has an alter ego—as a desire dampener that can go too far. Maybe some people who take GLP-1s lose interest in drugs and alcohol because they lose interest in everything. In 2023, after scores of reports of suicidal thoughts and self-harm among people taking GLP-1s, the European Medicines Agency initiated a safety review. The agency ultimately concluded that the available evidence did not establish a causal link between the medications and suicidal thoughts or behaviors. (The U.S. Food and Drug Administration noted in 2024 that it could not “definitively rule out that a small risk may exist,” but requested last month that suicide warnings be removed from GLP-1 drugs.) The data remain frustratingly ambiguous. Observational studies, which tend to be less precise than randomized controlled trials, have found that GLP-1s might harm or help a person’s mental health. The Nature network of journals recently published an article suggesting that GLP-1s reduce the risk of suicidal thoughts by more than half—and another article saying that they more than double the risk of suicidal behavior. Nearly all the experts I spoke with said that we needed more research.
When we take a medicine in the hope that it will curb a particular desire, we’re meddling with a complex system, Sarah Kawasaki, the chief of addiction services at Penn State Health, told me. Kawasaki portrayed addiction as a legitimate evolutionary adaptation—the pursuit of pleasure—gone haywire. Enjoyment of food encouraged our ancestors to nourish themselves; enjoyment of sex encouraged them to procreate. “When it’s not in its extreme form, pleasure-seeking is what led to the success of this species,” Kawasaki said. Tampering with it can have side effects. “You need joy in life,” she told me. “What happens when you blanket a society in GLP-1s? Do we all turn into boring, listless people?” A small minority of the population seems to report anhedonia after taking GLP-1s, and other mental-health consequences—irritability, insomnia, apathy, brain fog—are probably only slightly more common. But the rapid spread of these medications is, essentially, a vast social experiment. Drugs are generally approved based on controlled studies of several hundred individuals. Now tens of millions of people are taking GLP-1s, and we’re discovering, in real time, the full range of their effects.
It’s surprisingly difficult to say what turns a given behavior into an addiction. The answer probably has to do with pleasure, and with strong cravings, but a person can of course experience both without being addicted to anything. In “The House at Pooh Corner,” Winnie-the-Pooh remarks that “although Eating Honey was a very good thing to do, there was a moment just before you began to eat it which was better than when you were.” Meanwhile, Tantalus, the Greek mythological figure who lent his name to the word “tantalize,” was condemned to hunger and thirst in a pool beneath a fruit tree; the water ebbed whenever he tried to drink, and the fruit receded whenever he tried to eat. We don’t usually think of Pooh or Tantalus as having addictions. Yet a person in the throes of one might identify with either character, depending on whether the drug is within reach.
Opponent-process theory, refined in the seventies by the psychologists Richard Solomon and John Corbit, holds that the body balances out the high of a drug with the opposing force of a comedown or a withdrawal. In their view, addiction tends to dull the high but not the low, causing a disruption in a person’s “hedonic homeostasis.” In other words, people start taking a drug because it makes them feel good but go on taking it to avoid feeling bad; the dread of deprivation eventually dominates.
In the eighties, Kent Berridge, a neuroscientist at the University of Michigan, tried to study these dynamics on a biochemical level. The neurotransmitter dopamine was understood as a driver of pleasure and therefore a mediator of addiction; in one experiment, Berridge and his colleagues gave sugar water to dopamine-deprived rats. If pleasure depended wholly on dopamine, then the rats presumably wouldn’t have enjoyed the sweetness, but curiously they appeared to experience just as much pleasure as normal rats did. What changed was the rats’ willingness to work for the sugar water. A dearth of dopamine seemed to destroy their motivation.
In another study, Berridge linked the effects of dopamine with pleasant stimuli (sugar water and cocaine) and an unpleasant stimulus (a shock from an electrified rod). Each stimulus was paired with laser stimulation of a brain region called the amygdala, mimicking a dopamine hit. Oddly, the rats became just as entranced by the rod as they were by the sugar water or the cocaine. They seemed to want something that they clearly didn’t like. “I thought, My God, it’s the perfect prototype for addiction,” Berridge told me. “The wanting gets totally divorced from the liking.” People with addictions may seek a drug even when it no longer makes them feel good. For an addiction treatment to succeed, then, it might need to do more than render a drug unpleasant.
Many researchers believe that Ozempic and its chemical cousins act through the brain’s mesolimbic pathway, which is sometimes called the reward system. Alcohol, nicotine, cocaine, and opioids increase dopamine release in the pathway; so do gambling and social media. “Sex, drugs, sports cars—those things just light up the mesolimbic system,” Berridge said. “You want world peace, a good career, a nice family? Those are great wants, but they’re not mesolimbic dopamine wants.” This may explain why we don’t become addicted to our ideals or our families, and why these commitments aren’t usually affected by GLP-1 drugs. The specific ways that GLP-1s affect the mesolimbic system may also help explain how, at least in some people, they induce moderation. In animal studies, GLP-1s have been shown to limit spikes of dopamine, but not baseline levels, in the brain. When mice on the medications are given cocaine, they experience smaller surges of dopamine than normal mice, but otherwise they maintain adequate amounts of the neurotransmitter. The drugs may calm the water without draining the pool.
If this explanation is correct, then it’s just as Mary said: these drugs may have their most surprising effects not in the gut but in our brains. Indeed, when researchers knock out GLP-1 receptors in rodent brains, the drugs no longer combat obesity. (They still work as an anti-diabetes medicine.) Yet these clues about how GLP-1s work have given rise to another mystery. As scientists have tinkered with GLP-1s, transforming a short-lived peptide into long-lasting injections, the molecules have grown large enough that they shouldn’t be able to cross the blood-brain barrier. So how do they impact the brain? Scientists have theories. Maybe the drugs transmit signals through the vagus nerve, which connects the brain to other organs; maybe they boost production in the brain’s own small GLP-1 factory; maybe they seep in through small windows that are less protected, such as the teardrop-shaped area postrema, sometimes called the vomit center. (The latter might have something to do with a side effect of GLP-1s: nausea.) “The truth is, no one really knows,” Lorenzo Leggio, a physician scientist and a clinical director at the National Institutes of Health, told me.
Berridge’s discovery—that it’s possible to want something you no longer like—has become widely influential in addiction medicine. But Patricia (Sue) Grigson, the chair of the department of neuroscience and experimental therapeutics at Penn State, believes that there’s a third dimension. Addiction, Grigson told me, is more than psychological; she sees it as a physiological drive similar to that for food or water. There’s wanting and liking—and then there’s needing. A few months ago, Grigson showed me around a basement at Penn State’s medical school which smelled of fur, feed, and sawdust. On one side stood a dozen or so cages occupied by snow-white rats. On the opposite wall was an empty chamber next to a small black platform and a tangle of tubing. “That’s where the fentanyl goes,” Grigson said. The tubing could be connected, via a catheter, directly to a rat’s bloodstream. Using this contraption, Grigson has performed sophisticated experiments on GLP-1s and addiction. By repeatedly licking an empty spout, rats could “earn” an infusion of fentanyl. (“We’ve never had an animal overdose in the chamber,” Grigson said, solemnly.)
In one experiment, rats addicted to fentanyl were treated with either liraglutide, an older GLP-1 medication, or salt water. At first, the fentanyl syringe was kept empty—no matter what the rats did, they weren’t getting any more of the drug. Rats that had been injected with salt water engaged in drug-seeking behavior: they lapped at the spout more than a hundred times. “It didn’t do them a lick of good!” Grigson said. Rats on liraglutide, however, stopped after a few dozen attempts. They barely seemed to want the drug.
When Grigson’s team examined the brains of the rats on liraglutide, fentanyl-seeking wasn’t causing its usual spikes of activity in the mesolimbic pathway; GLP-1s appeared to lower the perceived rewards of the drug. “Now, here’s the really interesting thing,” Grigson said. The medication also dramatically suppressed activity in the locus coeruleus, an area involved in withdrawal and aversion. GLP-1s seemed to be having a moderating effect not only because they made fentanyl less pleasurable—but also because they made abstinence less painful.
Kawasaki, at Penn State Health, considered Grigson’s rat studies so astonishing that they began working together. “Humans are not rats—for the most part, anyway,” Kawasaki told me. “It was clear that we needed to test these medications in humans.” Decision-making is far more complex in humans than in other animals. The regions of the human brain that are affected by GLP-1s are embedded in a large cerebral cortex, which is involved in higher-order functions such as learning, memory, and consciousness. “The cortex is the cause and the cure,” Kawasaki said. “People have agency in a way that rats don’t. Ultimately, we can still make horrible decisions for ourselves.”
Kawasaki collaborated on a human trial for opioid addiction led by Grigson and Scott Bunce, a clinical psychologist at Penn State. Because liraglutide was being tested on a vulnerable population, they recruited twenty people at a residential treatment center where they could closely monitor participants. (“Someone might say, ‘These people are already thin,’ ” Kawasaki told me. “ ‘Now you’re going to give them a weight-loss drug?’ ”) Ten people received a placebo, and ten received liraglutide. The results were promising: people on liraglutide reported forty per cent fewer cravings. The differences were especially pronounced in the evenings, when cravings tend to peak.
Kawasaki is now helping lead a trial of hundreds of people with opioid addiction. This time, participants live in the community, not in a treatment center, and they are on multiple medications: semaglutide, as well as either methadone or buprenorphine, two opioid treatments that curb cravings of drugs such as heroin. “People with hypertension walk around on three or four blood-pressure medicines all the time,” Kawasaki said. “If one medication doesn’t work, we add another and then another until the condition is under control. We haven’t had that option with addiction, and it’s been profoundly limiting.”
In the lobby of her clinic, a squat brick building in Harrisburg, Pennsylvania, Kawasaki introduced me to Adrienne Pierce, a fifty-year-old woman in the trial. Pierce, who wore a bandanna and camouflage leggings, had tattoos of butterflies and flowers on her arms. In a counselling space that resembled a living room, she told me that she’d cycled through foster homes as a kid. She was raped at sixteen and afterward began using marijuana, alcohol, crack, and eventually heroin. She had a son who was taken into state custody shortly after birth, and she spent most of her twenties in prison. Once she was released, she got married and swore off drugs, but, a few months after the birth of another son, she discovered that her wife had an addiction to Percocet. Pierce’s own drug use returned; a familiar cycle repeated, and the authorities took her second son. Afterward, she achieved another period of sobriety. But it was upended after her wife died, of cardiac issues related to past drug use.
“I was willing to give Ozempic a shot because I’d never been able to stay completely clean,” Pierce told me. During prior attempts to quit, she had compulsively snacked on candy and gained weight, but this time her weight had stayed steady. The last time she started methadone treatment, it took her six months to stop using heroin; on methadone and semaglutide, it took six weeks. Pierce still had access to drugs and needles, but she had “no desire to touch them,” she said. “It was, like, I don’t even want to get high!” I was struck by how much her description echoed what Mary and Susan had told me. In each case, GLP-1s seemed to ease the pull of an addictive substance in the mind. It was as though by quieting the brain’s response to certain rewards—perhaps dimming the red splotches that I’d seen on the MRI scans—the medications transformed an overpowering emotional response into something that could be seen with detachment, from a distance.
Kawasaki was encouraged by Pierce’s experience. Still, she cautioned against prescribing GLP-1s for opioid addiction before larger and longer-term studies validate their use. There’s a danger in thinking that GLP-1s are a cure, she argued—and not only because the drugs have different effects in different people. “They do nothing to address the why of addiction,” she said: trauma, loneliness, pain, stress, poverty. “Those whys are tangible at first, but over time, as using drugs becomes a coping strategy, they disappear. The whys become an invisible part of what makes it hard for people to make it through the day.”
Pierce is now attending a community college. She has a relationship with her younger son; she’s pursuing a degree in psychology and hopes to open a mental-health center for at-risk teen-agers. “Almost all of us addicts have mental-health issues,” she told me. “People do drugs to numb the pain. If we were talked to when we were younger, maybe we wouldn’t have gotten into drugs in the first place.” GLP-1s are at best a partial solution. They’re not a substitute for mental-health care. And yet, for Pierce, they quieted not only cravings but also ruminations and traumatic memories that triggered her drug use. “The reason I couldn’t get clean in the past is because of the thoughts in my head,” she said. “Ozempic took all that away.”
It’s a truism in medicine that a drug works only if you take it. As many as half of patients on GLP-1 medications discontinue them within a year, whether because of side effects, access, or cost. Efforts to treat addiction with GLP-1s may face a deeper challenge: even when people stay on them, it’s not clear how long cravings will remain at bay. GLP-1s induce weight loss by altering your sense of fullness; eating enough to maintain your body weight starts to feel like eating too much. The drugs don’t require you to ignore hunger, as a diet would. “The magic sauce here is that people lose weight and they’re less hungry while it’s happening,” Randy Seeley, the director of the Michigan Nutrition Obesity Research Center, said. “If I taped your mouth shut, you’d lose weight, but you’d be more hungry.” Still, when people on GLP-1s reach a new weight plateau—usually within months—they often start to feel hungry again. When it comes to weight loss, this is O.K., since you were never supposed to stop eating altogether. But Seeley fears that a person’s appetite for drugs will return, too, which could prompt a recurrence of drug use.
Clinical trials have generally observed patients for weeks or months; addiction is a chronic disease that can last a lifetime. Even some of Grigson’s rats developed a tolerance to GLP-1s: after they’d been on a dose for a few weeks, their drug-seeking returned. “We’re in this phase right now where some people think these drugs cure everything,” Seeley said. “The hypothesis is that they target food, and then they have all these great spillover effects all over the place. My view is that the spillover is temporary. When the weight loss stops, the spillover stops.” If so, the medications could loosen the grip of an addiction temporarily, without inducing any kind of sustained sobriety. The GLP-1 bubble could still pop.
If trials like Joseph Schacht’s continue to produce promising results, the F.D.A. could formally approve GLP-1s for the treatment of addiction. But the drugs don’t necessarily need the F.D.A.’s blessing to gain new and widespread uses. (Aspirin was used to prevent repeat heart attacks for decades before the F.D.A. approved it for that use; propranolol, which is officially a blood-pressure drug, is often given for anxiety.) Soon, doctors may have to determine for themselves how to prescribe GLP-1s. How can we predict which patients will benefit, and at what cost? Are GLP-1s a bridge to sobriety or something that patients will need to take indefinitely? If researchers can’t address these questions quickly, patients will supply their own answers.
Susan, the former I.C.U. nurse, told me that, after she’d finished Schacht’s trial, she’d gone looking for a way to stay on a GLP-1 drug. A friend eventually directed her to a nurse practitioner who ran a beauty spa. The N.P. wouldn’t formally prescribe the medication for alcohol addiction, Susan told me, but agreed to do so under the cover of her “wanting to lose a few pounds.” Susan now keeps the drug at home and injects herself when she feels that she’s at risk of drinking. Last year, after her husband was given a diagnosis of Stage IV cancer, she briefly restarted the medication. “I knew I needed the GLP to get through that,” she told me. A few months ago, they took a long trip to celebrate a milestone wedding anniversary; Susan injected herself the week before they left. “I didn’t want to be caught at a happy hour and not be content with a Diet Coke,” she said. As a physician, I felt uneasy hearing about the ad-hoc use of a powerful medicine. Susan wasn’t following a doctor’s orders; she was essentially self-medicating. And yet I couldn’t dismiss the results. GLP-1s had helped her more than anything else had. When I last spoke with her, she was still sober. ♦
This article has been updated to clarify that Mary is a pseudonym.