GLP-1 Medications and Ultra-Processed Foods: Why Preferences Shift

Research is starting to show GLP-1 medications tilt shopping carts away from ultra-processed foods, but the shift has real limits worth understanding.

You go to the grocery store a couple of months after starting a GLP-1, and the cart looks different. Fewer chips. The cookie aisle no longer calls your name. That bag of candy at checkout, the one you used to toss in without thinking, stays on the shelf. This is not your imagination, and it is not willpower suddenly arriving out of nowhere. Researchers have started measuring this shift, and the most interesting part is where it shows up: mostly in the ultra-processed, hyper-palatable corner of the diet. Let's walk through what the evidence actually shows, where it is solid, and where it gets oversold.

What "ultra-processed" means, and why it matters here

Ultra-processed foods (often shortened to UPF) are the industrially formulated items engineered for maximum appeal: chips, packaged sweets, cookies, many baked goods, sugary drinks, and similar products built around refined starches, added sugars, fats, salt, and additives. They are calorie-dense, easy to overeat, and designed to be hard to stop eating. If a medication changes your relationship with food, this is exactly the category where you would expect to see it first, because these are the foods that most strongly hijack the brain's reward response.

That is the throughline of this post. GLP-1 medications appear to reduce the pull specifically toward these hyper-palatable items, and a large purchasing study now shows that shift landing in real shopping carts. But the story has honest limits, and a good clinic will tell you both halves.

The grocery-cart signal: what the purchasing study found

The clearest real-world evidence comes from a peer-reviewed study in the Journal of Marketing Research, built on verified household transaction records from Numerator's panel of roughly 150,000 U.S. households, with about 2,623 GLP-1-user households identified and adoption surveyed in four waves between late 2023 and mid 2024. Researchers compared each household's basket in the roughly six months before starting the medication to the six months after.

Within six months of starting, households with at least one GLP-1 user cut grocery spending by about 5.3%. The reductions were not spread evenly across the store. They concentrated in the calorie-dense, ultra-processed categories: savory snacks such as chips fell about 10.1%, with similarly large declines in sweets, cookies, and baked goods. One category actually went up in a statistically meaningful way: yogurt. So the pattern was not simply "buy less of everything." It was a measurable tilt away from the junk-food end of the aisle.

Two honest caveats belong right next to that number. First, the effect was income-dependent. Higher-income households (above $125,000 a year) cut roughly 8.2%, about $690 a year when annualized, close to double the overall figure. So the magnitude depends on the population you are looking at. Second, the authors themselves described the shift as "modest, relative" rather than a wholesale dietary overhaul. This is a nudge in the cart, not a transformation into a whole-foods household overnight.

Why the pull toward ultra-processed food weakens

The purchasing data tells you what happens; the mechanism tells you why. GLP-1 receptor agonists act on the brain's appetite and reward circuits, reducing how reactive you are to food cues and how motivated you feel to pursue palatable food. Neuroimaging work associates this class of medication with decreased activation in reward-related regions, including the insula, putamen, and amygdala, when people see palatable-food cues. Lower cue reactivity lines up neatly with a lower preference for energy-dense foods, which is exactly the pattern the grocery study picked up.

This is also the machinery behind what many patients call food noise, the repetitive, intrusive preoccupation with food that keeps running even when you are full. It is worth being precise here: food noise is a patient-reported, lay description, not an established clinical diagnosis, and the tools to measure it are still early. Many people on these medications report that it quiets down, and the mechanism makes that plausible, but the rigorous evidence is still developing.

What the controlled trials add

Two small but careful randomized trials give the preference story a firmer footing. In a 12-week randomized, double-blind, placebo-controlled crossover trial, once-weekly semaglutide in 30 adults with obesity reduced how much people ate when allowed to eat freely by about 24%, and produced a relatively lower preference for high-fat, energy-dense foods, along with fewer cravings and better reported control of eating. Mean weight loss was 5.0 kg on the medication versus a slight gain on placebo.

For tirzepatide, a secondary analysis of a phase 1 trial in 55 adults with obesity reported significant reductions in reported food cravings and food preference, particularly for high-fat and high-sugar or sweet foods, alongside greater weight loss. These are meaningful findings, but keep the frame honest: they are small studies measuring preference and intake in controlled settings, not proof of what your whole diet looks like month after month in real life. They tell you the pull weakens. They do not tell you the plate automatically becomes perfect.

The honest limits: this is a tool, not a cure

Here is where a trustworthy clinic separates itself from the hype. Several findings keep the claims grounded.

• The effect tracks with being on the medication. In the purchasing study, about one-third of users discontinued during the window. When they stopped, food spending reverted to pre-adoption levels, and baskets became slightly less healthy than baseline, partly from more candy and chocolate. That argues strongly against any claim that these drugs permanently change how people eat.
• The reward-suppression effect may be temporary. A Penn Medicine brain-imaging study suggested tirzepatide's dampening of reward and food-cue brain activity may not be a permanent rewiring. So "these medications fix the brain's response to hyper-palatable food for good" overshoots what we know.
• Diet quality does not automatically become good. In a cross-sectional analysis of adults using a GLP-1 for weight loss, the mean Healthy Eating Index was only 54 out of 100, well below the goal, with suboptimal fruit, vegetable, whole-grain, seafood and plant-protein, and dairy intake. Diet quality did not differ by how long people had been on the medication. Eating less junk is not the same as eating well.
• The pivotal trials mostly did not measure diet quality at all. A scoping review found that most major GLP-1 trials tracked weight and intake, not dietary quality. So the cart-level UPF signal, while real, is observational and category-level, not a validated diet-quality endpoint. Claiming these drugs "improve diet quality" outruns the trial evidence.

Put simply: GLP-1 medications can make ultra-processed foods less compelling and measurably shrink the junk-food share of your cart, but they do not build a nutrition plan for you, and the changes lean on staying on the medication and building durable habits.

How to lean into higher-quality eating while the pull is lower

The reduced craving for hyper-palatable food is best thought of as a window. It is easier to change what you eat when the constant tug toward chips and sweets has softened. Use it deliberately rather than assuming the medication does the whole job.

• Anchor every meal on protein. A joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association, and The Obesity Society sets nutritional priorities for people on GLP-1 therapy, stressing adequate protein to help protect lean mass, plus micronutrient sufficiency and higher-quality whole-food patterns. Our guide to what to eat on a GLP-1 walks through practical meals.
• Replace, don't just remove. If chips leave the cart, something should fill the space: fruit, yogurt (the one category that rose in the study), vegetables, whole grains, and lean or plant proteins. An emptier cart is not automatically a healthier one.
• Shop your environment. When the reward pull is quieter, it is a good moment to reshape what is within reach at home, since the food environment keeps steering choices long after the appointment ends.
• Think in patterns, not perfection. A Healthy Eating Index of 54 has a lot of room to grow, and small, repeatable upgrades beat a short-lived overhaul. Our piece on nutrition science on a GLP-1 goes deeper on the why.

Where medication fits, and where it does not

At New Hope Weight Loss and Wellness, Dr. Anjmun Sharma, MD frames these medications the way the evidence supports: a real tool that reduces the pull toward hyper-palatable, ultra-processed food, paired with a nutrition plan that does the work the drug cannot. A note on options you may see marketed: compounded semaglutide and tirzepatide are not FDA-approved and not brand-identical, and results vary by individual. Ozempic and Wegovy are trademarks of Novo Nordisk; Mounjaro and Zepbound are trademarks of Eli Lilly; New Hope is not affiliated with either company. And a GLP-1 is not a treatment for any non-obesity condition. Never start, stop, change, or skip any prescription on your own; that decision belongs with your prescriber.

The takeaway is not that a medication reprograms your taste forever. It is more useful and more honest than that. For a while, the foods that used to run the show get quieter, and the cart tilts away from the ultra-processed aisle. What you build during that window, meal by meal, is what tends to last.

Frequently asked questions

This article is informational only and not medical advice. Speak with a licensed physician before starting or changing any GLP-1 therapy. Individual results vary. New Hope Weight Loss is a physician-supervised medical weight loss clinic in Costa Mesa, CA. Eligibility for treatment is determined during the medical consultation. Compounded semaglutide and compounded tirzepatide are not the same products as Wegovy®, Ozempic®, Mounjaro®, or Zepbound®.