Sarah Boseley Health editor 

Ultra-processed foods may be linked to cancer, says study

Findings suggest increased consumption of ultra-processed foods tied to rise in cancers, but scientists say more research is needed
  
  

Ultra-processed foods can include noodles, ready meals, cakes and confectionery which contain additives, preservatives, flavourings and colourings – and often high levels of sugar, fat and salt.
Ultra-processed foods can include noodles, ready meals, cakes and confectionery which contain additives, preservatives, flavourings and colourings – and often high levels of sugar, fat and salt. Photograph: Premier Foods/PA

“Ultra-processed” foods, made in factories with ingredients unknown to the domestic kitchen, may be linked to cancer, according to a large and groundbreaking study.

Ultra-processed foods include pot noodles, shelf-stable ready meals, cakes and confectionery which contain long lists of additives, preservatives, flavourings and colourings – as well as often high levels of sugar, fat and salt. They now account for half of all the food bought by families eating at home in the UK, as the Guardian recently revealed.

A team, led by researchers based at the Sorbonne in Paris, looked at the medical records and eating habits of nearly 105,000 adults who are part of the French NutriNet-Santé cohort study, registering their usual intake of 3,300 different food items.

They found that a 10% increase in the amount of ultra-processed foods in the diet was linked to a 12% increase in cancers of some kind. The researchers also looked to see whether there were increases in specific types of cancer and found a rise of 11% in breast cancer, although no significant upturn in colorectal or prostate cancer.

“If confirmed in other populations and settings, these results suggest that the rapidly increasing consumption of ultra-processed foods may drive an increasing burden of cancer in the next decades,” says the paper in the British Medical Journal.

France is one of the few countries that already specifically warns its people against high levels of ultra-processed foods in the diet on “the precautionary principle”, said Mathilde Touvier, lead author of the study. The foods have already been linked to obesity, but the association with cancer is new.

“The results are very strong – very consistent and quite compelling,” she said. “But we have to be cautious. It is the first study. We should not be alarmist. These results need to be confirmed in other prospective studies.”

Ultra-processed food is a definition created by a group of scientists led by Prof Carlos Monteiro in Brazil, a country which also has national dietary guidelines urging they be eaten as little as possible. The classification system, called Nova, puts foods into four groups – raw or minimally processed foods including seeds, fruit, eggs and milk; processed culinary ingredients such as oils and butter; processed foods including bottled vegetables and canned fish and cheeses; and ultra-processed, which are “formulations made mostly or entirely from substances derived from foods and additives”.

Critics of ultra-processed foods say the processing strips out most of the nutrients, but Touvier says they do not believe the low nutritional value of biscuits and sweets and cakes are the cause of the raised cancer risk they saw. “We did a statistical analysis to try to see if the whole association was only due to poor nutritional quality,” she said. “But the results do not depend on that.”

She said they needed to do more research to figure out whether any rise in cancer is down to the high load of sugar, fat and salt or possibly the additives. “We need to understand the mechanism,” she said. “Maybe in the future we will have an idea whether one or two molecules are the problem and not all the ultra-processed foods.”

Her team now have a massive database of all the additives in specific foods, by commercial names and brands. Over the years to come, she said, “we will be able to quantify the chronic exposure [of people] to the food additives”. They will be able to study their effects alone and in combination with other additives, she believes, to find out whether there is “a cocktail effect”.

Other scientists questioned whether it was practical to group foods as ultra-processed. “The term ultra-processed food is difficult to define in terms of food quality, and is not widely used by nutritional scientists,” said Tom Sanders, professor emeritus of nutrition and dietetics at King’s College London. “From a nutritional standpoint, this classification seems arbitrary and based on the premise that food produced industrially has a different nutritional and chemical composition from that produced in the home or by artisans. This is not the case.

“The approach of categorising dietary patterns that depend on industrially processed food in relation to disease risk is novel but probably needs refining before it can be translated into practical dietary advice.”

A spokesman for the Food and Drink Federation said they were proud of the industry’s track record on reformulation “but recognise that obesity and diet-related diseases are a complex and serious issue and more needs to be done. We believe a whole diet and lifestyle approach, which includes consideration of net calorie intake, and not just the role of individual nutrients or ingredients, is the correct way to tackle such issues.

“Processed food should not be demonised – by working closely with our partners throughout the food supply chain, we can use processing positively to ensure all sectors of society have access to safe, affordable food.”

Tam Fry, spokesman for the National Obesity Forum, said: “A lot of research has limitations and the scientists here are honest enough to acknowledge that theirs needs more work to be conclusive. But there is no smoke without fire: we should heed their fears – and read food labels more carefully. Huge quantities of everyday processed food have excessive levels of sugar, fat and salt stuffed in them and it’s all listed on the packaging.”

 

Leave a Comment

Required fields are marked *

*

*