Craving scores are useful, but they don't tell the whole story. Here's how to read the measurements behind GLP-1 and addiction research.
Ask whether GLP-1 and addiction are really connected, and you'll find the answer is still being written. GLP-1 receptors sit in brain regions tied to reward, including parts of the mesolimbic dopamine system. You can check the basic neuroscience yourself by searching PubMed for 'GLP-1 receptor agonist reward' and reading the review articles that come up first. That biology is the main reason researchers ask whether these drugs could reach compulsive behaviors, not just appetite.
Tirzepatide works differently from GLP-1-only drugs because it acts on both GIP and GLP-1 receptors, as its FDA prescribing information describes. Whether that second receptor target changes anything about reward is still unknown. Researchers are now testing the question directly, and each method has strengths and blind spots. Here's how the main tools work and what they can and can't show.
The tools researchers use to measure craving
Craving is hard to measure because nobody can see it directly. Researchers often use the Food Craving Questionnaire, which asks people to rate how strongly they want specific foods, along with visual analog scales, where someone marks urge intensity on a line. The original validation papers are indexed on PubMed, and searching 'Food Craving Questionnaire validation' will bring them up. In plain words, the questionnaire captures how strongly a person reports wanting certain foods. It does not measure how much they actually eat.
Self-report has real limits. Expectations shape answers. Someone who believes a drug will help may rate urges lower, and someone who doesn't may rate them higher. Blinding and placebo groups reduce that effect, but they rarely erase it completely. If you track your own urges at home, use the same scale every day so the pattern stays readable, and bring the notes to your clinician. The TirzeBlog archive has posts on simple tracking formats you can adapt.
Some labs go further with cue-reactivity tasks. Participants look at trigger images, such as a drink or a slice of pizza, while researchers record heart rate, skin conductance, or brain activity. These tasks capture a physiological response to cues, which feels more objective than a checkbox. They still measure one moment in a lab, though, not a week of real decisions.
What animal studies show and why they do not transfer cleanly
Much of the early work on GLP-1 drugs and reward comes from rodents. Researchers have tested GLP-1 agonists on alcohol intake and on drug self-administration, often in operant chambers, where an animal presses a lever to earn a reward. When the drug reduces lever pressing, the result is often described as reduced motivation. That's an interesting signal, but it isn't proof that people want less.
Animal results don't travel cleanly to people. Doses in rodents rarely map directly onto human doses, species differ in metabolism and brain wiring, and a lab cage is nothing like a stressful week or a kitchen full of snacks. Animal models measure behavior, such as how often a rat presses a lever. They can't measure the felt experience of wanting. To see this yourself, search PubMed for 'exenatide alcohol rats' and read the methods section, not only the abstract. The methods list the dose, the schedule, and the setting, which is where most caveats live.
Human trials: what is registered and what is still open
ClinicalTrials.gov lists registered studies of GLP-1 agonists in alcohol use disorder and tobacco use. Registration matters because it creates a public record of what researchers planned to measure before results arrived. A registration is still not a result. Some trials never publish outcomes, some are withdrawn, and some are still recruiting.
Small pilot studies have real value, mainly for checking whether a protocol is feasible, whether people stay enrolled, and whether the measurements work. They rarely settle whether a drug actually reduces use. To see the current picture, search ClinicalTrials.gov by drug name and by condition, such as alcohol use disorder. Then read the status field for each entry and note whether results are posted. You'll get a more honest sense of how much evidence exists.
Practical reading guide for readers and clinicians
Headlines often blur the type of study. Before trusting a claim, work out whether it comes from a human trial, an animal study, a case report, or a survey. A few quick checks help:
- Check which species the abstract names in its first lines.
- Look for the number of participants and whether there was a placebo group.
- Identify the outcome: a craving score, a count of drinks, or a confirmed change in behavior?
If you're talking with a prescriber, a few questions are worth raising. Ask whether you have a history of substance use disorder, since that history can change how your care is monitored. Ask how they'd track mood, sleep, and alcohol use during treatment. And ask what they'd do if side effects made it hard to keep taking the medication. If you want a plain-language primer before that visit, TirzeBlog covers the basics.
Approval status matters too. According to Drugs@FDA, Mounjaro (tirzepatide) is approved for type 2 diabetes, and Zepbound (tirzepatide) is approved for chronic weight management and for moderate to severe obstructive sleep apnea in adults with obesity. Neither is approved for any addiction indication. Check the current label, since indications can change. That gap shapes insurance coverage and prescribing decisions. Prescribing a drug off label is a clinical choice, and your prescriber should explain it in plain terms. The National Institute on Drug Abuse (NIDA) emphasizes evidence-based treatment for substance use disorders, which is the standard any new option should be measured against.
What would count as real evidence
Strong evidence for an addiction treatment usually looks like a randomized, placebo-controlled trial with validated outcomes. Those outcomes often include drinks per week, days of heavy use, or abstinence confirmed by a biomarker. A drop in a craving score can be part of the story, but on its own it rarely proves a change in use.
Follow-up length matters. A few weeks of reduced use can look impressive and then fade, so trials need to run long enough to separate a short-term dip from lasting change. Adverse effects deserve equal attention. The Drugs@FDA label for tirzepatide lists gastrointestinal reactions such as nausea, diarrhea, and vomiting, and those can affect whether people keep taking a drug. Trials should report mood-related effects clearly too, since mood can shift during treatment.
Outcome choice changes what a result means. Pick two registered trials on ClinicalTrials.gov and compare their primary endpoints. One may measure a craving scale while the other measures drinks per week. A positive result on the first may not carry over to the second. Reading endpoints side by side is one of the most useful habits you can build.
Closing: what to keep in mind while the research grows
A reduction in a craving score and a clinically meaningful change in use are not the same thing. The first tells you how urges feel on a scale. The second shows whether someone's health, life, or drinking pattern has actually shifted. Keep that distinction in mind whenever a headline promises help with cravings.
If you have questions about substance use or weight treatment, bring them to your own clinician. They know your history and can weigh options a general article can't. Tirzepatide can help some people lose weight, but it does not replace addiction treatment. If substance use is part of your picture, ask about evidence-based care alongside any weight plan. Readers who want more context can find related posts on TirzeBlog.
FAQ
Does tirzepatide reduce food cravings?
Whether it does is still being tested. Self-reports are shaped by expectations and by how cravings are measured, so talk with your prescriber about what you notice.
Is tirzepatide approved to treat addiction?
No. Mounjaro and Zepbound have approved indications listed on the FDA label, and addiction is not among them. Check Drugs@FDA for the current version.
Are there human trials of GLP-1 drugs for alcohol use?
Yes, some are registered on ClinicalTrials.gov. A registration is not a result, so check each study's status and whether outcomes have been posted.
Can animal studies predict what happens in people?
Only partly. They help generate hypotheses, but differences in dosing, species, and setting limit how far the results apply to humans.
Sources
Disclaimer: This content is for informational purposes only and does not replace professional medical advice. Always consult your doctor before starting, changing or stopping any treatment.
Ozempro — Monitor GLP-1
Never miss a dose and track everything
Injection reminders, a dose history and a side-effects diary — all organized to bring to your appointment.
or download the app
