GLP-1 Agonists in 2026: Beyond Diabetes and Weight Loss
The evidence is outrunning the prescription pad
Here's the thing about GLP-1 receptor agonists in 2026: the data keeps landing faster than the guidelines can catch up. Five years ago these were a diabetes footnote. Then they became the weight-loss story of the decade. Now they're showing up in heart failure trials, kidney trials, liver trials, and addiction trials, and the prescribing habits haven't caught up to any of it.
So if you're standing in front of a patient and wondering "am I treating their sugar, their weight, their heart, or something else?" you're asking the right question. That's the whole point of this piece. Let's walk through what the data actually shows, organ by organ, and then figure out who should be on these drugs and who really shouldn't.
First, the part everyone already believes
GLP-1s lower glucose and they cause weight loss. Semaglutide at the top dose takes off roughly 15 percent of body weight; tirzepatide, the GIP/GLP-1 combo, gets to 20 to 22 percent. That's the part that made the headlines, and it's real. But it's also the least interesting part now, because the glucose and the weight were never the whole story.
The headline shift happened with the SELECT trial. Semaglutide in people with obesity and cardiovascular disease who did NOT have diabetes cut the composite of cardiovascular death, heart attack, or stroke by about 20 percent. That single result reframed the entire class. A GLP-1 stopped being "a diabetes drug that also makes you lose weight" and became "a cardiovascular risk-reduction drug that happens to work in obesity." If a patient has obesity and established CVD, the GLP-1 is now a legitimate part of the heart conversation, not just the diabetes conversation.
Heart failure: the preserved ejection fraction story
This is where 2024 and 2025 got exciting. The SUMMIT trial put tirzepatide into patients with heart failure with preserved ejection fraction (HFpEF) who also had obesity and cardiovascular disease, and it reduced the combined risk of worsening heart failure events and cardiovascular death.
That matters because HFpEF is the heart failure that's been maddeningly hard to treat. For years the answer to "what drug improves outcomes in HFpEF?" was basically "lose weight and manage your blood pressure." Now there's a drug with an outcomes signal in exactly that population. A couple of Lancet studies around the same time showed GLP-1s improving functional capacity and heart failure stability in obesity-related HFpEF, and there's even emerging sex-specific data on tirzepatide in this group.
But be precise about the gap. The benefit is clearest in HFpEF, the preserved ejection fraction side. For HFrEF, the reduced ejection fraction patients, especially those with a very low EF, we still don't have solid outcomes data. So if you're thinking about a GLP-1 for a heart failure patient, the EF number matters a lot, and the honest answer for the low-EF patient is "we don't know yet."
The kidneys
The FLOW trial is the one to remember here. Semaglutide in patients with diabetes and advanced kidney disease, stage 3b and 4, significantly slowed the rate of GFR decline and reduced cardiovascular death. That's a hard renal outcome, not just a surrogate.
Where it sits in the algorithm: SGLT2 inhibitors are still first-line for diabetic kidney disease. But when you can't use an SGLT2 inhibitor, the GLP-1 is the recommended move. So "renal protection" has quietly become a real reason to reach for these drugs, not just a bonus.
The liver
The ESSENCE study is the one that made hepatologists sit up. In metabolic-associated steatohepatitis (MASH) with fibrosis, semaglutide reversed steatohepatitis in about 63 percent of patients versus 34 percent on placebo at 72 weeks, and more than a third of the treated patients had reduced fibrosis. The five-year clinical endpoints, like progression to cirrhosis, are still on the way, but the histologic signal at 72 weeks is strong.
For the clinician, the takeaway is that a patient with MASH and fibrosis is now a reasonable candidate for a GLP-1 on liver grounds, not just weight or diabetes grounds. It's one more organ on the list.
Addiction: the frontier that actually surprised people
This is the one that caught most of us off guard, and it's the reason "beyond diabetes and weight loss" is really the headline.
The first randomized trial came out of JAMA Psychiatry in early 2025. It was small, 48 adults with alcohol use disorder, nine weeks, and it used the lowest clinical doses of semaglutide, 0.25 mg up to 0.5 mg, which is about a fifth of the weight-loss dose. They ran a "laboratory bar" design where people could drink their preferred alcohol in a controlled setting before and after treatment. The semaglutide group drank significantly less, had lower breath alcohol levels, fewer cravings, fewer heavy drinking days, and, in the subgroup that smoked, fewer cigarettes per day. The effect sizes were potentially bigger than what you see with the existing FDA-approved addiction medications.
And it wasn't an isolated result. A Nature Communications analysis of over 682,000 records found semaglutide associated with about a 50 percent lower risk of new alcohol use disorder diagnoses and lower recurrence. A separate analysis of 1.3 million electronic health records found roughly a 50 percent lower incidence of alcohol intoxication events. A Finnish cohort of 228,000 people found about a 36 percent lower risk of alcohol-related hospitalizations.
The proposed mechanism is in the brain's reward circuitry. GLP-1 receptors sit in the ventral tegmental area and the nucleus accumbens, and semaglutide appears to dampen the dopamine "want" signal. That's still preclinical and being mapped, but it fits the observations.
Now the honest caveats, because they matter. The RCT was small and short. The participants weren't actively seeking treatment. And the FDA has not approved any GLP-1 for alcohol use disorder. Phase 3 trials are underway, with both Lilly and Novo Nordisk investing in this space. So this is cautious optimism, not a prescription. If a patient is struggling with alcohol, the answer today is still the approved options, naltrexone, acamprosate, and disulfiram, plus the behavioral work. But the trajectory is hard to ignore, especially given that only three drugs are approved for an addiction that affects roughly 29 million Americans and kills about 178,000 a year.
So who SHOULD be on them?
By 2026 the indication list has genuinely broadened. Reasonable candidates include:
- Type 2 diabetes with cardiovascular disease or high cardiovascular risk.
- Obesity with cardiovascular disease, where the SELECT data applies.
- Heart failure with preserved ejection fraction plus obesity, where SUMMIT applies.
- Diabetic kidney disease when an SGLT2 inhibitor isn't an option.
- MASH with fibrosis, where the liver data is emerging.
- Alcohol use disorder, where the data is promising but not yet approved.
The unifying thread is that the drug is increasingly chosen for the organ you're trying to protect, not just the number you're trying to move.
And who should NOT be on them?
This is the part that separates a good prescriber from a careless one. The hard stops and cautions:
- Medullary thyroid carcinoma or MEN2: there's a boxed warning. Personal or family history of MTC or multiple endocrine neoplasia type 2 is a contraindication, based on rodent thyroid C-cell tumor data. The human relevance is uncertain, but the warning is real and you should honor it.
- Pregnancy: not recommended, based on animal reproductive toxicity. Hold it before a planned conception.
- Type 1 diabetes: these are not a substitute for insulin. In an insulin-deficient patient you risk DKA and you're not treating the underlying problem.
- Diabetic retinopathy: rapid glucose lowering has been associated with worsening retinopathy in one trial. If the eyes are already compromised, slow down and monitor.
- History of pancreatitis: use caution. The absolute risk is low, but it's the safety question everyone asks, so it deserves a deliberate answer rather than a shrug.
- Severe GI intolerance: if nausea, vomiting, or early satiety are derailing the patient, that's a reason to re-titrate or stop, not to push through.
And there's a non-medical one that's arguably the biggest in practice: cost and coverage. Off-label use, like obesity without diabetes, runs into prior authorization walls constantly. The right drug at the wrong price is a prescription that never gets filled.
What's coming
A few things are on the horizon that will keep reshaping this. Oral formulations are maturing, which matters a lot for access and cost. Multi-agonists are the next wave: tirzepatide already combines GIP and GLP-1, and triple agonists targeting glucagon, GLP-1, and GIP are in late-stage development with even bigger weight loss. And the indication list keeps growing, which is exactly the problem and the opportunity. The question is no longer whether a GLP-1 helps. It's whether it's the right tool for this specific patient's specific organ.
The bottom line for the bedside
When you're deciding whether a patient should be on a GLP-1 in 2026, run through this:
- Name the organ you're actually treating. Glucose, weight, heart, kidney, liver, or something else. The indication drives the decision.
- Check the EF if it's a heart failure patient. The benefit is clearest in preserved EF; the reduced-EF data is still thin.
- Screen the contraindications before you celebrate the benefits: MTC/MEN2, pregnancy, type 1 diabetes, active retinopathy, pancreatitis history.
- Titrate slowly for the GI effects and don't be a hero about pushing a dose a patient can't tolerate.
- Be honest about what's approved and what's emerging. Heart failure, kidney, and liver have real data. Addiction is promising but not approved. Say that plainly.
- Factor in cost and coverage, because the best drug in the world is useless if it can't be filled.
The class has outgrown its original label. That's a good thing, but it means the easy answer, "it's for diabetes and weight," is no longer enough. The better answer is the specific one: here's the organ, here's the evidence, here's why this patient, and here's why not that one.