A study of 2,192 Australian adults found that small daily rises in ultra-processed food track with weaker attention, even on an otherwise healthy diet.
Add one standard packet of chips to your daily routine and you can measure the cost in how well you pay attention. That is the size of the effect a research team at Monash University put a number on, and it’s harder than the usual “junk food is bad for you” warning because of how small the dose is.
The team analysed diet and cognition in 2,192 dementia-free Australians aged 40 to 70, drawn from the Healthy Brain Project (Cardoso et al., 2026). For every 10% increase in the share of daily energy coming from ultra-processed food, scores fell on standardised tests of visual attention and processing speed. That 10% is roughly one extra bag of chips or one soft drink a day.
The cost of one extra packet of chips
The participants were already eating a lot of this food. On average they pulled about 41% of their daily energy from ultra-processed products, close to the national Australian figure of 42%. So the 10% step the researchers measured sits well inside the range most people move through without noticing. Skip the snack on Monday, grab two on Friday, and you have shifted your intake by more than that across a week.
The study did not find a direct link to memory loss, but it did flag attention as the floor that learning and problem-solving are built on. Weaken the floor and everything stacked on top gets shakier, even if your memory test scores look fine for now.
Why attention is the floor
Picture the last time you read a page in a book, reached the bottom, and realised none of it stuck, so you started again from the top. That is attention faltering. Processing speed is how fast you take the page in once you are looking. Neither shows up on a memory test, because nothing reaches memory if it never registered in the first place. The Monash data found no direct hit to recall, yet a measurable one to these earlier steps. Memory is the filing cabinet. Attention decides whether anything makes it into the drawer.
What the Monash team measured
Higher ultra-processed intake came with worse attention scores and a higher dementia risk score. The risk side runs through familiar conditions like high blood pressure and obesity, which is good news of a sort, because those are things you can act on rather than fixed traits you are stuck with.
Why the Mediterranean diet doesn’t help
The attention drop held even among people whose diets otherwise looked healthy by Mediterranean standards. The lunchtime salad does not appear to cancel out the mid-afternoon packet of biscuits.
If a healthy overall pattern were doing the protective work, you would expect the salad eaters to be shielded. They were not. That suggests the cost is tied to the processing itself, not only to the nutrients the food is missing. Cardoso pointed to two suspects: ultra-processing strips out the natural cell structure of food, and it adds substances such as emulsifiers, dyes and flavour enhancers that the body never met in a whole-food diet. Other researchers have floated neuroinflammation as a route by which those additives could reach the brain.
What ultra-processed covers
Ultra-proccessed food is broader and blurrier than most people assume. The Nova system sorts all food into 3 groups by degree of processing. Group 4, ultra-processed food, covers industrial formulations made largely from substances extracted from foods plus additives, typically with five or more ingredients you would not find in a home kitchen.
Ultra-processed foods include most bread and other mass-produced baked goods, frozen pizza, instant noodles, flavored yogurt, fruit and milk drinks, diet products, baby food, and most of what is considered junk food. The Nova definition considers ingredients, processing, and how products are marketed. Nutritional content is not evaluated.
In practice that sweeps in soft drinks, packaged salty snacks, reconstituted meat products, and most heat-and-eat ready meals. A useful rule of thumb:
“If you can’t find the ingridient in your kitchen and the ingredient list reads like a chemistry set, with modified starches, protein isolates, hydrogenated oils, and colour codes, it is probably ultra-processed.”
Some food scientists argue Nova classification lumps together products of widely different nutritional value and lacks sharp criteria, so a sugar-free yoghurt and a frosted snack cake can land in the same bucket. The Monash study sidesteps part of that objection by showing an effect that tracks with processing degree even after accounting for diet quality, but the debate over the term is real and ongoing.
Not a verdict
This was a cross-sectional study which means it can show that high ultra-processed intake and weaker attention travel together but cannot prove one causes the other (PsyPost). People who eat more of this food may differ in sleep, exercise, income or stress, and any of those could feed into both the diet and the test scores.
The Monash result does not stand alone, which is why it carries weight despite that limit. It joins a run of 2025 and 2026 studies pointing the same direction, including a systematic review in which most of the included studies found a link between higher ultra-processed intake and poorer cognition (CNN). One snapshot proves little. A wall of snapshots all leaning the same way is harder to wave off.
Where to start cutting back
Pick the one ultra-processed item you eat most days, the desk-drawer biscuits or the lunchtime soft drink, and swap it for a whole-food version a few days a week. Fruit, nuts, plain yoghurt and a glass of water do the same job your snack does without the additive load.
Reading ingredient lists helps more than chasing a perfect diet. You are not aiming for zero ultra-processed food, which is unrealistic for most working people, but for a lower share of it. Cooking one extra meal at home each week nudges the ratio in your favour. So does buying the version of a staple with the shorter ingredient list. Small, repeatable changes beat a crash overhaul you abandon.
A realistic day shows how little it takes. Trade the morning pastry for oats and fruit, refill a water bottle instead of buying a second soft drink, and reach for a handful of nuts when the afternoon biscuit packet calls. You have pulled several percentage points out of your intake without counting anything or skipping a meal. The aim is a lower share across the week, not a spotless record on any single day. The diet you can keep beats the one you quit.
The bottom line
The number that should stick with you is the smallness of it. One packet of chips a day, the kind of choice you make without thinking, was enough to register on a focus test. If a small daily habit can pull your attention down, a small daily swap is exactly the size of lever you need to start pushing back, starting with your next snack.
Frequently asked questions
The study did not name a safe ceiling. It found that each 10% rise in the share of daily energy from ultra-processed food, about one packet of chips, came with lower attention scores (Cardoso et al., 2026). The takeaway is direction, not a threshold: less appears better, and the participants were already averaging 41% of their energy from these foods.
Not according to this study. The attention drop showed up even in people following a Mediterranean-style diet, which suggests the processing itself carries a cost beyond the nutrients a food lacks. A healthy lunch does not appear to buy back the harm of a processed snack later in the day.
Ultra-processed foods are industrial formulations, usually with five or more ingredients, made largely from substances extracted from foods plus additives like emulsifiers, dyes, and flavour enhancers (Nova classification). Soft drinks, packaged salty snacks, reconstituted meat products, and most ready meals qualify. If the ingredient list is full of items you would not keep in a home kitchen, it is likely ultra-processed.
The study cannot answer that directly, because it measured people at a single point in time rather than tracking them after a change. What it does show is that the dementia risk it measured runs partly through manageable conditions like high blood pressure and obesity. Reducing ultra-processed intake and weight gives you a handle on those, which is a reasonable starting point even without proof of reversal.
No. This was a cross-sectional study, so it shows that high intake and weaker attention occur together but cannot prove cause and effect (PsyPost). The result gains weight from fitting a wider run of 2025 and 2026 research pointing the same way, but it stops short of a verdict.
Leave feedback about this