Ultra-Processed Foods and Long-Term Health

1. Higher ultra-processed food consumption was associated with adverse outcomes across multiple areas of health.

Greater exposure to ultra-processed foods was significantly associated with higher risk in 32 of 45 pooled analyses — 71%.

The strongest evidence involved mortality, cardiometabolic outcomes and common mental health outcomes.

However, the strength and quality of evidence differed substantially between individual outcomes.

2. Higher consumption was associated with greater mortality risk.

Greater exposure to ultra-processed foods was associated with a 21% higher risk of all-cause mortality when higher and lower exposure groups were compared.

This association was classified as highly suggestive evidence, with low certainty according to GRADE.

For cardiovascular mortality, greater exposure was associated with a 50% higher risk.

This association met the statistical criteria for convincing evidence, although its GRADE certainty was very low.

This distinction is important.

A strong statistical association does not necessarily mean that the underlying evidence is of equally high quality.

3. Type 2 diabetes showed one of the strongest and more consistent associations.

In the dose-response analysis, greater ultra-processed food consumption was associated with a 12% higher risk of type 2 diabetes.

This association met the criteria for convincing evidence and received a moderate GRADE rating.

This was among the strongest combinations of statistical credibility and evidence quality identified in the review.

4. The overall direction of the evidence was consistent, but certainty remained limited.

Across all outcomes, 93% of the pooled analyses produced estimates in the direction of greater risk with higher ultra-processed food exposure.

Not all of these associations were statistically significant.

The GRADE assessment also showed important limitations:

4 pooled analyses were rated moderate quality,

22 were rated low quality,

19 were rated very low quality.

None were rated high quality.

Healthy aging depends on reducing the long-term burden of diseases that shorten lifespan and healthspan.

Cardiovascular disease and type 2 diabetes are particularly important because they can influence health over many years before major complications develop.

The associations identified in this review therefore have direct relevance to healthy aging.

The findings also suggest that evaluating a diet only by individual nutrients may not provide the complete picture.

Ultra-processed foods often differ from minimally processed foods in several ways. They may have higher energy density, contain more added sugars, sodium or saturated fat, and provide less fiber and fewer micronutrients.

Processing can also alter food structure, eating rate and satiety.

However, the mechanisms responsible for the observed long-term associations are not yet fully established.

This is one of the broadest evaluations of the epidemiological evidence on ultra-processed foods and health.

Its strength lies not only in its scale, but also in the rigorous assessment of the underlying evidence. The researchers examined heterogeneity, prediction intervals, small-study effects, excess significance bias and the contribution of the largest studies. They also evaluated evidence using both credibility criteria and GRADE.

However, the evidence was predominantly observational.

People who consume more ultra-processed foods may differ in many other ways from those who consume less. Differences in overall diet, physical activity, smoking, body weight, socioeconomic circumstances and other lifestyle factors can influence long-term health.

Residual confounding therefore remains possible.

Ultra-processed foods are also a heterogeneous category.

Different products classified as ultra-processed can have very different nutritional profiles. Some individual subcategories have not shown the same associations as overall ultra-processed food consumption.

The evidence should therefore not be interpreted as showing that every food classified as ultra-processed has the same effect on health.

Experimental evidence nevertheless provides additional context.

A randomized crossover trial published in Nature Medicine in 2025 compared minimally processed and ultra-processed diets that both followed UK healthy eating guidelines. Fifty-five adults with overweight or obesity were randomized, and 50 contributed to the intention-to-treat analysis. Both eight-week diets resulted in weight loss, but weight loss was significantly greater during the minimally processed diet. Physical activity did not significantly differ between the dietary periods.

This newer trial strengthens evidence that food processing may influence some short-term outcomes beyond conventional healthy-eating recommendations. It does not, however, establish the long-term effects on cardiovascular disease, mortality or longevity observed in epidemiological studies.