How much ultra-processed food is too much?

What the long-term studies found

The best evidence comes from very large studies that recorded what people ate and then followed their health for years. Two stand out.

In UK Biobank, 60,298 adults were followed for about 11 years. The quarter who got the most of their calories from ultra-processed food had about a 17% higher rate of cardiovascular disease than the quarter who got the least.

A European study of 266,666 adults in seven countries, also followed for about 11 years, linked eating roughly 260g more ultra-processed food a day with about a 9% higher rate of developing both cancer and a heart or metabolic condition.

These studies show a link, not proof: people who eat a lot of ultra-processed food differ in other ways too. And a 17% higher rate compares groups. It isn't your personal odds, and 17% more of something uncommon is still uncommon. What makes it worth taking seriously is that the same direction keeps turning up across countries. More studies are in the research notes below.

How much do people in Ireland and the UK eat?

Older Irish national survey data, analysed in a 2021 European study, put ultra-processed food and drink at about 35% of the energy women ate and 32% for men. UK survey data from 2008 to 2019 puts it at about 54% for adults.

Different years and methods, so it's not a league table, and it doesn't mean Irish diets are healthier. It does show ultra-processed food is a big, normal part of most people's week.

Is all ultra-processed food unhealthy?

No, and this is the part that gets lost.

The studies use NOVA, a system that groups foods by how they're made. FoodSussed estimates a Level 1 to 4 from the ingredients list along similar lines, but it's our own estimate, not a NOVA result (how the Levels work). Either way, the Level describes processing, not what a food gives you.

Three everyday Level 4 examples, from the packs on 30 October 2026:

• Heinz Baked Beanz (415g): flagged for modified corn flour, a starch changed for factory use. The label lists 10g of protein per half can.
• Warburtons Wholemeal Medium (400g): flagged for added wheat gluten and emulsifiers. The label lists 6.4g of fibre per 100g.
• Arla Protein Strawberry Yogurt (450g): flagged for flavouring, pectin and sweetener. The label lists 10g of protein per 100g.

Labels and recipes can change, so check the pack you have.

The European study also split ultra-processed food by type. Ultra-processed breads, cereals and plant-based alternatives weren't clearly linked with developing cancer plus a heart or metabolic condition; sweetened drinks and animal-based ultra-processed foods were. That doesn't prove any of them caused or prevented it, but it's a good reason not to treat every Level 4 food the same.

A large US study in PLOS Medicine points the same way. It followed 461,682 older adults for over 20 years and found no link between total ultra-processed food and bowel cancer. Wholegrains, fibre and dairy from ultra-processed foods were linked with lower rates; meat from ultra-processed foods with higher.

It works the other way too. A lower Level doesn't make something better for you. A short ingredients list can still be high in sugar, salt or fat.

So what should you aim for?

With no official target, the closest real-world reference is what the "low" group ate in the studies. In the UK Biobank analysis, the lowest quarter got about 21% or less of their calories from ultra-processed food; the highest got over 43%. That isn't a proven safe amount, just what lower intake looked like for ordinary adults.

What I actually do

When I started testing FoodSussed on my own shopping, I was surprised by how much of what I bought every week came back Level 4. The beans were the one that stopped me. I know they're decent food, which is exactly why I wanted to write this.

My week doesn't leave much room for perfect planning. I work full-time with long hours, I'm building FoodSussed on the side, we have kids, my wife runs her own business, and I'm in the gym four nights a week. I find the balance hard, and I'm still working on it.

What's helped is keeping a few simple meals as the default: beef burgers made from just beef on a sourdough bun, or chicken and rice. Fruit is where I struggle. I buy it with good intentions and half of it goes off, so I've moved to frozen berries, which actually get eaten.

Studying sports nutrition has pushed me the same way. Calories, protein and the overall pattern of what you eat still do most of the work. Processing is one more thing to look at on top.

Making the weekly shop work for you

Most of what I eat is decided in the shop, not the kitchen. What makes it easier for me:

• A handful of reliable meals you can make without thinking, bought every week.
• Frozen fruit and veg. They keep, and the ingredients list is often just the fruit or veg. Check the bag.
• The ingredients, not the front of the pack. "Protein" or "sourdough" on the front doesn't tell you what's in it.
• Like-for-like swaps. If two similar products do the same job at a similar price, the lower Level is an easy change.
• Keeping the Level 4 foods that pull their weight. If one gives you protein or fibre you'd otherwise miss, keep it.

You don't have to clear every Level 4 food out of the press. Knowing which of your regulars are Level 4, and why, is a good start.

Checking your own regulars

That's what I built FoodSussed for. Take a photo of the ingredients list and it estimates a Level from 1 to 4 and shows which ingredients set it, so you can compare the things you buy every week. It doesn't give food a health score; the Level sits alongside the nutrition label and what works for you.

See how FoodSussed works