Large analysis links ultra-processed foods to higher disease and death risk

What happened

People who ate the most ultra-processed food had higher risks of several major diseases and of dying from any cause than people who ate the least, according to a new pooled analysis, as first reported by Healthline News.

Healthline says the work was published Sept 22 in Family Medicine and Community Health, a BMJ journal, and covered its story on Sept 28, 2026.

What the analysis found

Per Healthline, this is a meta-analysis of prospective cohort studies: 51 studies, nearly 9 million adults, follow-up of 2 to 32 years. Participants were 18 or older and free of the disease being studied at the start.

Researchers compared the highest with the lowest UPF intake and also looked at dose-response. UPFs were identified with the NOVA classification, and diet came from food-frequency questionnaires, 24-hour recalls or diet histories.

The dose-response numbers are the ones most useful for a grocery run. For each extra 100 grams a day of UPFs, Healthline reports:

• Cardiovascular events: 14% higher risk
• Hypertension: 11% higher risk (the highest-vs-lowest comparison for hypertension was not statistically significant)
• Cancer: 4% higher
• All-cause death: 3% higher
• Metabolic syndrome or diabetes: 2% higher

Separately, a 10-percentage-point rise in the share of daily intake from UPFs was tied to 16% higher risk of metabolic syndrome or diabetes.

Healthline's piece gives no confidence intervals, heterogeneity statistics, quality ratings or funding details. It also doesn't say whether the 16% figure for metabolic syndrome or diabetes is one pooled estimate across the two conditions.

What it does not show

Cohort data can show that two things travel together. It can't show that one drives the other. Healthline's headline says UPFs "worsen health," which is stronger than this design supports, and "linked to" is the honest verb.

Healthline doesn't discuss confounding or measurement error. Both are standard problems when people self-report what they eat, and people who eat more UPFs may differ from those who eat fewer in other ways.

The category itself is also contested. Healthline notes there is no commonly accepted UPF definition, and NOVA draws criticism. A 2026 Nature Food perspective argues back, saying policymakers dismiss UPF risks by alleging "conceptual imprecision and limited evidence."

Funding matters here too. The Lancet's UPF series was funded by Bloomberg Philanthropies, and lead author Carlos Monteiro was part of the team that developed NOVA. That doesn't invalidate the work, but it is context.

The "UPF" label also lumps very different foods together. One dietitian quoted by Healthline pointed out that a sugary soda and a fiber-rich packaged whole-grain bread can both count as ultra-processed while differing nutritionally. She also said no known cutoff exists where UPFs become harmful.

What it means for your cart

Don't panic-purge your pantry. The more defensible takeaway is about proportion: the dose-response framing suggests the more of your diet comes from UPFs, the higher the associated risk.

The dietitian's advice, as relayed by Healthline, is to keep UPFs from being the foundation of your diet rather than eliminating them. That fits the evidence we can see.

• Look at what your meals are built on. If most of your calories come from packaged, ready-made items, that is the pattern the data flags.
• Don't treat all UPFs as equal. Nutrition still differs between a soda and a whole-grain bread.
• Swap a few staples you eat daily rather than overhauling everything, since the risk figures scale with how much you eat.
• Treat the percentages as rough signals, not personal forecasts.

Context: where this fits

This isn't the first pooled look. A 2024 BMJ umbrella review by Lane and colleagues, available via PMC, rated evidence for several outcomes as convincing or highly suggestive. But certainty by GRADE was mostly low or very low.

Examples from that review: cardiovascular mortality at a relative risk of 1.50 (very low certainty), all-cause mortality at 1.21 (low), and type 2 diabetes dose-response at 1.12 (moderate).

An updated narrative umbrella review covered 12 lead meta-analyses and 13 outcomes. Higher UPF intake was linked to higher risk in all but one, with no significant dose-response for NAFLD. Certainty was high for one outcome, moderate for six and low for six.

The authors flag the observational basis and non-rigorous dose-response analyses as gaps, yet still say the evidence supports population-wide strategies to cut UPF intake.

In November 2025, a Lancet series concluded that UPFs are displacing traditional diets and worsening diet quality, and it called for global policy action.

There is also a randomized crossover trial in Nature Medicine comparing ultra-processed and minimally processed diets that both followed UK healthy-eating guidelines. We only reviewed its title and introduction, so we aren't reporting its results.