Processed foods and Type 2 diabetes – new EPIC data

Aim

Over 540 million people worldwide currently have type 2 diabetes mellitus (T2DM), a fourfold increase in cases over the past few decades, and this number is predicted to rise.

Ultra-processed food (UPF) consumption is increasing as well and makes up a sizable portion of dietary intake.

Prior meta-analyses have shown a 35% increase in T2DM incidence in those with the highest UPF consumption relative to those with the lowest, and a 12% increase in T2DM risk for every 10% increase in UPF content in diets.

This prospective cohort analysis of European Prospective Investigation into Cancer and Nutrition (EPIC) data examined the association between degree of food processing and incident T2DM.

Method

The study analysed EPIC data from eight countries (The Netherlands, Denmark, Sweden, France, Germany, Italy, Spain, and the United Kingdom (UK)) where T2DM incidences were reported.

Participants were excluded if follow up time was zero, had T2DM at baseline, implausible anthropometric measurements, or reported extreme energy intake.

Using the Nova classification, foods were coded into four categories: (1) Unprocessed/minimally processed foods (MPFs): whole or processing using basic methods such as drying, boiling, or freezing. (2) Processed culinary ingredients (PCIs): foods directly obtained from MPFs or nature (i.e. table sugar, olive oil). (3) Processed foods (PFs): simple industrial combinations of MPFs and PCIs, potentially using basic preservation methods such as canning or bottling (i.e . tinned fish, fruits in syrup). (4) UPF: food made from formulations of industrially modified ingredients that contain additives such as colours, flavours, emulsifiers, or gelling agents.

Key findings

311,892 participants were included in the analysis.

At baseline, average age was 52.5 years (standard deviation (SD): 9.4), BMI was 25.7 kg/m2 (SD: 4.1), and 63.5% of the participants were female.

Average follow-up time was 10.9 years (SD: 2.4), during which 14,236 (4.6%) participants were diagnosed with T2DM.

Across all countries, the average percentage contribution of UPF to total daily dietary intake in grams was 13.0% (SD: 7.8), the UK has the highest contribution with a daily intake at 17.4% (SD: 8.0).

One of the article's four objectives was to divide UPF into sub-groups and examine the relationship between an increase in consumption of 10% g/day and incident T2DM (Figure 1).

Each 10% g/day higher intake of the savoury snacks (HR: 2.77 (95% CI: 1.09–7.05)), animal-based products (HR: 2.25 (95% CI: 1.96–2.57)), ready-to-eat/heat mixed dishes (HR: 1.16 (95% CI: 1.01–1.35)), and artificially and sugar-sweetened beverage (HR: 1.25 (95% CI: 1.22–1.28)) subgroups were associated with higher incident T2DM.

Subgroups of breads, biscuits and breakfast cereals (HR: 0.65 (95% CI: 0.57–0.73)), sweets and desserts (HR: 0.89 (95% CI: 0.84–0.95)), and plant-based alternatives (HR: 0.46 (95% CI: 0.26–0.82)) were associated with lower incident T2DM.

Sauces, spreads, and condiments (p = 0.074); alcoholic drinks (p = 0.744) and other UPF (p = 0.870) were not associated with incident T2DM.

Figure 1:

Conclusion

This large-scale analysis of eight European countries showed that higher intake of UPF was linked to higher incident T2DM and replacing UPF with lower degrees of food processing is associated with lower incident of T2DM.

In this analysis, some UPFs were linked to increased occurrence T2DM such as savoury snacks, animal-based products, ready meals and sugary beverages, where other UPFs such as ultra-processed breads, biscuits, breakfast cereals, sweets, sauces, spreads, desserts and plant-based alternatives were linked to a neutral or lower risk.

It should be noted that information on diet and covariates was only collected from participants at baseline.

Over time, dietary changes, and other lifestyle factors, such as alcohol consumption, smoking, and physical activity, might affect the risk of T2DM and may introduce residual confounding.

Therefore, to validate their findings, the authors call for more research with cohorts that provide repeated dietary information.

Save article as PDF

Save article as PDF