Research Review: Ultra-Processed Food and Mortality

In this article, our expert Dr Kieron Rooney – Associates Professor at The University of Sydney & nationally recognised expert in nutrition research – reviews Ultra-processed food intake and all-cause mortality: DRECE cohort study by Ferreiro et al. 2021, and breaks it down to give us the vital parts we need to know. 


1) What was the high-level summary of the research?

The risk of dying increases with the amount of ultra-processed food products in one’s diet.


2) What did the study try to measure?

This study monitored the risk of what is referred to as “all-cause mortality” (death from any cause) during the 26 years that spanned 1991 – 2017 in Spain. The authors wanted to see if this risk was influenced by what people ate.


3) How was the study undertaken?

The study design described in this paper is called a longitudinal cohort survey. This simply means following the same group of people over a long period of time. Practically speaking, the authors gained access to a diet survey that was completed by 4679 individuals living in Spain in 1991. The survey is what is called a food frequency questionnaire and asked individuals to estimate over the previous 12 months how often they may or may not have eaten over 100 different foods.

Once the authors had this survey data, they classified the food items according to degree of “processing” as described by the NOVA classification.

The NOVA classification allocates foods into 1 of 4 groups depending on the level of processing the food has undergone. Group 1 for example are fruits and vegetables; Group 2 are culinary ingredients such as salts and oils; Group 3 are processed products with 3-4 ingredients while Group 4 are ultra-processed foods with multiple, highly refined ingredients and additives.

The authors then reviewed the National Institute of Statistics in Spain to see how many, if any, of these 4679 participants had died during the 26 years between completing the survey in 1991 and December 31, 2017.

Once they had both the diet and the incidence of death data, they used statistics to see if there was any increased risk of likelihood of dying in the 26-year period with the foods reported being eaten in 1991.


4) What did the study find?

The average age of individuals in the study was 30 years, and on average, this group of participants were eating about a quarter of their total daily energy from Group 4 ultra-processed foods products such as sugar sweetened beverages, milkshakes, meat and meat products like sausages and deli meats, dairy products, cakes pastries, sweets and cookies.

Of the 4679 individuals that completed the food frequency questionnaire in 1991, 450 had died by December 31, 2017.

And…. the higher the amount of Group 4 ultra-processed food products in the diet, the higher the risk of being one of those 450 individuals.

The actual risk reported by the study was a 15% increased likelihood of all-cause mortality for every 10% increase in the proportion of energy in the diet from group 4 products.

The authors also made some changes to the diet survey responses to see if the risk of all-cause mortality could be reduced and found that replacing Group 4 ultra-processed foods with Group 1 foods (fruits and veges) resulted in a statistically significant improvement and a reduced risk of all-cause mortality.


5) Is there any other research out there that supports these findings or contradicts it?

Absolutely, very similar studies have been completed on population data in France, the USA and Italy. Some of these studies have been completed in larger cohorts with the French study completed in over 100 000 individuals looking at cancer incidence in the first 8 years following diet survey. Each of these cohort studies are singing the same song – the higher the proportion of energy in your diet that comes from Group 4 ultra-processed food products. The higher the risk of health problems and disease.

To date, there are no studies that contradict this outcome.


6) How much weight should we give this research?

This study is interesting, and the statistical analysis is strong. Importantly it confirms data coming out of Europe and America that are generally telling us the same thing – the rise of ultra-processed foods is dangerous for our health. However, the sample size is relatively small, and there was only 1 diet survey completed in 1991 and how representative that was of each participant’s diet for the next 26 years is obviously a stretch.


7) What does this mean for your work/research/industry?

What this does for me is add strength to the need to continue to unpack the role of ultra-processed foods in the development of disease and empower individuals to make changes towards reducing the consumption of these foods.


8) What’s the key takeaway for us to take from this research?

If you can find the motivation and capacity to switch out an ultra-processed food product for its minimally processed equivalent – do it!


9) Will you be doing anything differently because of this research?

Not so much differently, but perhaps with more vigour, as it confirms the need for more current prospective studies in which we actively remove these group 4 foods from people’s diets and see if it improves their health as predicted.


Dr Kieron Rooney is a biochemist, nutritionist and exercise physiologist from the University of Sydney whose research focuses on how what we eat, influences our metabolism.