There’s a moment a lot of people describe in almost the same words. A week or two into a GLP-1, they open the cupboard, stand there for a second, and realise the argument they were expecting to have with themselves just… isn’t happening. The biscuits are right there. And nothing is pulling.
People have taken to calling that background chatter “food noise”: the low, near-constant commentary about what to eat next, the snack you keep circling back to all afternoon. For some people on these medications, it goes quiet. Not “I have more willpower now” quiet. More like someone turned the volume knob down on a radio you’d stopped noticing was on.
This article is general information, not medical advice. Traqr does not treat, diagnose or manage addiction, substance use, or any medical condition. Nothing here is a reason to start, stop or change a medication, and nothing here is a claim about what will happen to you. Decisions about your treatment are for you and your prescriber.
For a long time that quiet was just a thing people reported anecdotally. Now there’s a growing pile of research trying to work out what’s going on underneath it, and some of it reaches further than the dinner plate.
A very large study, and a careful reading of it
In March 2026, researchers at WashU Medicine published a study in The BMJ that got a lot of attention. They looked at the health records of 606,434 US veterans with type 2 diabetes, comparing people who’d started a GLP-1 medication with people who’d started a different class of diabetes drug (SGLT-2 inhibitors), and followed them for up to three years.
Among people with no prior history, the GLP-1 group had lower recorded rates of developing a substance use disorder across the board: about 25% lower for opioids, 20% for cocaine, 20% for nicotine, 18% for alcohol, and 14% for cannabis. Among people who already had a substance use disorder, the differences ran to the serious end of the scale — roughly 39% fewer overdoses and around 50% fewer substance-related deaths than the comparison group.
Two things have to be said clearly about that, because they’re the whole point.
First, this is an observational study. It shows an association across a very large group of people; it does not prove the medication caused the difference, and a study like this can’t. The researchers themselves are careful about that.
Second, and more importantly for anyone reading this: this is a population finding, not a personal one. It describes averages across 600,000 people. It says nothing about what will happen to you, it is not a reason to start or stay on a medication, and Traqr is not an addiction tool or a substance-use anything. We’re a tracking app for people on GLP-1s. We’re writing about this because of why the researchers think they’re seeing it.
The reward pathway, and food noise as its everyday face
The explanation the researchers keep coming back to is that GLP-1 medications appear to act not just on the gut but on the brain’s reward and motivation circuitry — the system that decides how much you want something, and how loudly it nags at you until you get it.
Food noise is the most visible, most everyday version of that system doing its thing. The pull towards the snack and the pull towards other rewards run through overlapping machinery. So the theory goes, if a medication turns down the volume on wanting-food, it may be turning down a broader dial, and the substance-use numbers are what that looks like at the scale of a huge population.
You don’t need to buy the full theory to notice the near end of it. The quieting of food noise is the part of this you can feel. It’s also the part you can watch.
The one you can see in your own data
This is the part that connects to something concrete, not a headline about veterans.
Traqr lets you log food noise directly, on a simple 0–100 scale, alongside hunger and fullness, whenever you want to. Over a couple of weeks that turns into a trend line. You can see whether your food noise is settling, and whether it tends to be louder or quieter at different points in your dose cycle. When it drops, the app says so in plain words, “your food noise has quietened down over this period”, and when it creeps back up it says that too.
That’s your own self-reported data, not a medical measurement, and it isn’t a diagnosis of anything. But it’s useful, for one very human reason. For a lot of people the food noise going quiet is the first thing that changes, often before the scale has moved a single kilo. And the scale, frustratingly, is the thing everyone stares at.
Why this matters most on the flat days
Think about the plateau. Three, four months in, the weight loss slows or stalls for a bit, and the daily weigh-in stops rewarding you. That’s the moment a lot of people start to wonder whether any of it is still working.
The scale isn’t the only instrument on the dashboard, and on those flat weeks it may not even be the most honest one. If your food noise has dropped from a 70 to a 20 and stayed there, something is still working. Your relationship with the cupboard has changed, even on a week the number on the scale refuses to budge. Being able to see that, in your own logged data, is often the difference between “this has stopped doing anything” and “oh, right, look at that.”
That’s the reason food noise is a first-class thing you can track in Traqr, not a feeling you half-remember afterwards. A trend line doesn’t treat or manage anything. It earns its place for a plainer reason. On the days the scale goes quiet, it helps to have another honest signal that hasn’t.
The big veterans study is a look at what these medications might be doing to the reward system at a scale no individual will ever feel. The food noise trend on your own screen is the small, close-up, personal end of the same story. It’s the one you can watch settle.
Sources
- WashU Medicine / The BMJ — Cai M, Choi T, Xie Y, Al-Aly Z, “Glucagon-like peptide-1 receptor agonists and risk of substance use disorders among US veterans with type 2 diabetes: cohort study,” published 4 March 2026. Cohort of 606,434 veterans with type 2 diabetes; GLP-1 vs SGLT-2 inhibitors, followed up to three years. Reduced recorded risk of substance use disorders (opioid ~25%, cocaine ~20%, nicotine ~20%, alcohol ~18%, cannabis ~14%); among people with existing substance use disorder, ~39% fewer overdoses and ~50% fewer substance-related deaths. Observational association, not causation.
- WashU Medicine news release — “GLP-1 medications get at the heart of addiction: study” (March 2026): researchers link the effect to GLP-1 action on the brain’s reward and motivation pathways.
- “Food noise” as reported by GLP-1 users and its relationship to reward-pathway signalling — background on the term as an everyday description of appetite-related mental chatter, not a clinical diagnosis.