The hunger signal: GLP-1 drugs and autistic interoception

Weight-loss drugs like semaglutide - the GLP-1s - work by changing how the body senses hunger and fullness. Many autistic people sense hunger and fullness differently. So does the drug work differently for us? Better, worse, or just strangely?

1 The sense most people have never heard of

In Chapter 3 of Loving Variant Minds, I called interoception the most important sense you’ve never heard of. It’s the sense that reports on the inside of your body: hunger, thirst, temperature, needing the toilet, the racing heart that tells you you’re anxious. Many autistic people describe it as unreliable, and it can fail in both directions.

Some of us don’t notice hunger until we’re shaking and irritable. We forget to drink. We push through illness without registering how unwell we are. Partners end up asking “have you eaten today?” like a parent, and both people find it frustrating. Others have the opposite experience: hunger that never quite switches off, a meal that ends without the body noticing it has ended, and food taking up far more headspace than it seems to for anyone else. The only autistic case report in this whole area is a young man of that second kind.

Now look at what a GLP-1 drug actually does.

GLP-1 is a hormone your gut releases when food arrives. It’s one of the messengers that tells your brain the meal is over. Semaglutide - sold as Ozempic and Wegovy - is a manufactured copy of that messenger, built to last far longer in the body than the natural one. Liraglutide is an earlier drug from the same family. The advertised effect is weight loss. What it really does is change the hunger signal itself.

So there’s a drug that works directly on a body signal, and a group of people known for reading that signal differently. That seemed worth asking about.

And this isn’t a small group. Autistic people are more likely to be obese than non-autistic people: a meta-analysis of fifteen studies, covering nearly fifty million people, put the odds at roughly one and a half to two and a half times higher. ¹ A larger review of ninety-five studies found obesity in about a fifth of autistic people overall, and nearly a third of autistic adults. ² So the gap widens as autistic children grow into autistic adults. The same review found that being underweight was also more common, and also rose with age. As a group, we’re more likely to be at either end - heavier or lighter - and that matters for what comes later.

2 One man, and a lot of mice

There’s very little to go on.

Several teams have given semaglutide to BTBR mice, a strain used as a model of autism. A 2024 study reported that repeated doses reduced autism-like behaviours without affecting movement. ³ Another, from 2026, put it down to changes in gut bacteria and metabolism. ⁴ Both are real papers in real journals. But neither one looks at interoception. The first is about DNA repair and oxidative stress; the second is about the gut-brain connection. Neither is testing my question, and mouse behaviour is a long way from human experience.

Then there’s one human being. In 2019 a Finnish team published a case report on a twenty-year-old autistic man with moderate intellectual disability. ⁵ He had severe food obsessions, binge eating and aggressive behaviour. On liraglutide, the food obsessions dropped away quickly, his weight fell by twelve to thirteen per cent, and - the interesting part - the aggression and his other, non-food compulsions eased too. No side effects were recorded. The authors were careful to say that one patient is one patient, and called for proper trials.

One detail in that case stays with me. His weight gain had been caused by the antipsychotic he’d been given for his behaviour in the first place. That’s a common autistic story, and it’s one reason autistic people end up as candidates for weight-loss drugs at all.

As for trials, I couldn’t find a single registered trial of a GLP-1 drug in autistic people. ⁶ There are active semaglutide trials for alcohol dependence, opioid dependence, Alzheimer’s, asthma, arthritis and teenage bone health.

So the entire human evidence for my question is one case report, from 2019.

3 The comparison that makes me nervous

If you want to know whether a drug that works on fullness helps people whose sense of fullness is unusual, there’s an obvious place to look. Prader-Willi syndrome is a genetic condition in which the hunger switch doesn’t turn off. People with it live with relentless hunger, and obesity is the main cause of early death.

Liraglutide has been properly tested there. A trial across several centres ran for fifty-two weeks in children and teenagers with Prader-Willi syndrome, including sixteen weeks against a placebo. ⁷ It missed both of its main targets: the change in body mass index at sixteen weeks, and at fifty-two weeks. Hunger scores shifted a little in the teenagers, enough that the authors thought it worth studying further, but the trial didn’t show what it set out to show.

That’s sobering, and anyone excited about my question should keep it in mind. In the one group where the fullness signal is definitely atypical, the drug did worse than it does in the general population, not better. That’s the opposite of the hopeful reading.

4 Maybe it isn’t autism at all

There’s a bigger problem, and it’s with my starting point.

The idea that autistic people have atypical interoception isn’t settled. People are still arguing about it.

The strong version came from Quattrocki and Friston in 2014, who argued that a failure of interoception is what produces autism in the first place. ⁸ Within a year, Brewer, Happé, Cook and Bird published a direct reply. ⁹ They argued that the interoception problems belong not to autism but to alexithymia - difficulty identifying and describing your own emotions - which turns up alongside autism in about half of cases. In 2016, Shah and colleagues published a study whose title says it plainly: alexithymia, not autism, is linked to impaired interoception. ¹⁰ A 2018 study went further, arguing that impaired interoception isn’t at the heart of either one. ¹¹

The evidence runs the other way too. A 2023 study found reduced interoceptive accuracy and awareness in autistic participants, with alexithymia playing a part in that rather than explaining it away. ¹² A 2016 review found enough to justify taking interoception in autism seriously. ¹³ So it’s contested, and nobody has won the argument yet.

Two more things make it messier.

The first is how it’s measured. Much of this research rests on one task, unchanged since 1981: sit still, don’t take your pulse, and count how many heartbeats you can feel. The closer your count is to your actual heart rate, the better your interoception is judged to be. ¹⁴ The task has been heavily criticised, including by researchers who’d used it themselves, because people may be guessing from what they believe about their heart rate rather than actually feeling it. ¹⁵ A field built on a shaky measure produces shaky findings, in both directions.

The second is more specific to my question, and I think it’s the sharpest objection. Interoception isn’t one thing. It varies from one part of the body to another, and the evidence that being good at sensing one predicts being good at sensing another is mixed. Almost all the autism research has looked at the heart. GLP-1 drugs act on the stomach. Even if the link between autism and interoception were rock solid for heartbeats, it wouldn’t automatically carry over to hunger. Anyone who looks at this properly would need to measure the stomach, and measure alexithymia separately.

5 The risk nobody is measuring

If I had to make the case for this on public health grounds rather than curiosity, I wouldn’t start with weight loss. I’d start with harm.

Avoidant/restrictive food intake disorder - ARFID - isn’t about body image. It’s restricted eating driven by sensory aversion, by fear of choking or vomiting, or just by a lack of interest in food. It overlaps heavily with autism. A 2025 meta-analysis found autism in about sixteen per cent of people with ARFID, and ARFID in about eleven per cent of autistic people. ¹⁶ A genetically informed study of a large autism cohort put the figure higher, at about a fifth, and suggested ARFID is routinely missed. ¹⁷

Remember too that the reviews finding more obesity in autistic adults also found more underweight, rising with age alongside it. ² Eating too little sits right beside eating too much, in the same group of people.

Now put a drug on top of that. A drug whose main job is to take away interest in eating, which commonly causes nausea, and which changes how appealing food feels.

I’m not saying this causes harm. I’m saying nobody has looked, that the way it could cause harm is obvious once you say it out loud, and that autistic people are already being prescribed these drugs. Not having a trial doesn’t make the risk go away. It just means nobody has measured it.

6 And then you stop

There’s a part of this I didn’t think about until someone asked me, and it may matter most.

These drugs usually aren’t a course of treatment with an end. When people stop taking them, the weight tends to come back. In the follow-up to one of the big semaglutide trials, people who came off the drug regained two-thirds of what they’d lost within a year, and the improvements in blood pressure and blood sugar went with it. ¹⁸ The authors’ conclusion was that obesity behaves like a chronic condition, and stopping the treatment tends to undo the result.

But people do stop. They stop because of cost, side effects, supply, moving, or because they think they’re done. And here’s where I was wrong. I expected to find established psychological support for people coming off these drugs, so I could ask why autistic people were being left out of it. There isn’t any. A 2025 review says plainly that behavioural support isn’t standard even when these drugs are first prescribed, and that what happens psychologically after people stop is poorly understood. ¹⁹

Think about what stopping means for the sense this piece is about.

For a year or two, the drug has been managing your hunger signal for you. Then it stops, and the signal is handed back. That’s strange for anyone: hunger arriving with a force it hasn’t had in a long time, and the mental noise around food coming back after a stretch of quiet. If your interoception was unreliable before the drug, it’s stranger still. You’re being handed back a signal you never read well, after a long time of not having to read it at all, and nobody has told you what to expect.

I don’t know whether that’s harder for autistic people than for anyone else. I suspect it might be, but I can argue it either way, which tells me it needs measuring rather than assuming. What I’m confident about is that nobody is studying it, and that someone coming off one of these drugs with no warning, and a returning hunger they can’t make sense of, is at risk of deciding they’ve failed.

What I think

The drugs are already in autistic bodies. It’s happening at scale, in every country where they’re prescribed, and it’ll keep happening. Whether it should be allowed isn’t really the question anymore; the market has answered that.

My guess is that the weight loss would look much the same for autistic and non-autistic people, because the drug probably works below the level of conscious body awareness. But I think how it feels would differ a lot, and that some autistic people would do badly in ways a trial that only weighs people would never pick up. A patient who quietly stops eating can get recorded as a treatment success, if the only number anyone’s watching is the weight.

Knowing more would change two things. The first is in the doctor’s office. If the way someone’s interoception differs predicts how they respond, prescribers have a reason to ask about autism before they write the prescription, and to watch for effects they’d never expect in other patients. Nobody looks for a side effect they’ve never been told about. Asking takes one conversation, and not asking could cost someone their relationship with food.

The second is in therapy. A psychologist can’t prepare someone for an effect nobody has written down. If we knew how these drugs tend to land on a nervous system like ours, a therapist could say in advance: you may find food stops interesting you altogether, and given how you already relate to eating, that might be a big deal - so let’s agree now what we’ll do if it happens. That’s a very different conversation from having it six weeks later, with someone who has decided something is wrong with them.

Neither needs new law, or anyone’s permission. And there’s a reason to get to it soon. Because there’s no standard support for coming off these drugs yet, there’s nothing for autistic people to be left out of. That makes the gap bigger, but easier to fill, because there’s nothing in place to argue with. If the first real guidance on coming off these drugs were written with autistic interoception in mind from the start, rather than added later as an afterthought, that would be worth more than anything I could ask of a single prescriber. And the first big wave of people coming off these drugs is arriving now.

⁂

---