Printed on: October 16th, 2025

Please note, once printed or downloaded, articles cannot be updated. By bookmarking the article and continuing to access it online, you can be sure you are reading the most up-to-date information and that you are not in breach of copyright laws.

Ultra-processed food

Ultra-processed food

There has been a significant rise in consumption of ultra-processed food (UPF) in recent decades. Over 50% of the food bought in the UK now consists of UPF. This compares with 14.2% and 13.4% in France and Italy respectively (PH Nutrition 2017;10.1017).

UPF has been increasingly associated with a wide range of negative health consequences. In the UK, it is often easy to prepare and is widely available. Higher rates of consumption are associated with social deprivation and exacerbate health inequalities through poorer health outcomes. 

This article was updated in August 2025.

In simple terms, ultra-processed foods are not ‘real’ food. They are not what we have evolved to eat. They can be defined as “industrial formulations made by deconstructing whole foods into chemical constituents, altering them and recombining them with additives into products that are alternatives to fresh and minimally processed foods and freshly made meals” (BMJ 2022;378:o1972).

There are four ‘groups’ of foods commonly referred to in research studies. They are taken from the NOVA food classification system developed by the Centre of Epidemiological studies at University São Paulo, Brazil (2018).

It is important to note that processing per se is not ‘bad’. In many ways, it has massively helped us as a population to preserve and transport food. Canned chickpeas retain fantastic nutrient qualities.

There is growing evidence that increased consumption of UPFs is associated with a range of long-term conditions, particularly cardiometabolic disease and cancers. They are commonly products which consist of highly-refined carbohydrates with added fats to help achieve a ‘bliss point’ (maximum craveability) for consumers (BMJ 2023;383:e075354).

For those who like the details, let’s explore some of the most recent studies.

A recent umbrella review of epidemiological meta-analyses in the BMJ explored this issue in just under 10 million patients (BMJ 2024;384:e077310). It found evidence of impact across a large array of body systems and conditions. It assessed the evidence based on its credibility (assigned based on predefined criteria) and quality (using the GRADE framework: grading of recommendations, assessment, development and evaluations).

• A convincing dose-dependent association with type 2 diabetes was found.
• Further convincing associations were found with adverse sleep, anxiety and combined mental health disorders.
• There were several other associations, but study quality was found to be low to very low.
• These findings mirror other studies, including a large population-based cohort study published in the BMJ. However, this study found that overall dietary quality had a greater influence on mortality rates than UPF consumption (BMJ 2024:385:e078476).

There is a substantial body of evidence from meta-analyses, systematic reviews and cohort studies, from across the world in different population settings, that consumption of ultra-processed food is associated with an increased risk of multiple health conditions (BMJ 2022;378:o1972). In 2023, the BMJ summarised evidence from several meta-analyses linking UPF to cardiac and metabolic health (BMJ 2023;383:e075294). In our table below, we have also included information from a JAMA article about the impact of UPF consumption on depression (JAMA Netw Open. 2023;6:e2334770).

There is also evidence that demonstrates a prospective association between ultra-processed food consumption and adverse lipid profiles, hyperuricaemia, NAFLD, Crohn’s disease and breast cancer (BMJ 2022;378:o1972).

An association has been demonstrated between higher consumption of UPF and an increased risk of distal colorectal cancer in men (BMJ 2022;378:e068921). A recent prospective cohort study in the Lancet also found an association between higher consumption of UPFs and an increased risk of cancer generally and cardiometabolic multimorbidity (Lancet Reg. 2023;35:100771).  

Artificial sweeteners were shown to be associated with increased cardiovascular and cerebrovascular disease in a large French observational cohort study (BMJ 2022;378:e071204). This was particularly true of aspartame, acesulfame potassium and sucralose, which are all commonly present in a wide range of processed food and drinks worldwide.

A randomised crossover feeding trial (UPDATE Nature 2025:0;0) in 55 adults compared 8 weeks of minimally-processed food (MPF) with UPF diets. Both diets were constructed to meet UK Eatwell Guide recommendations, and had no specific guidance on calorie intake (‘eat till full’).

• Primary outcome: both diets produced weight loss at 8 weeks, but MPF caused significantly greater percentage weight change (MPF −2.06% vs. UPF −1.05%), and had larger reductions in weight, BMI and fat mass.
• Secondary outcomes:
  • MPF led to greater reductions in body fat %, visceral fat rating and triglycerides, and improved craving control.
  • UPF (surprisingly!) showed greater reductions in LDL-C and some other biomarkers.
  • Energy intake was lower on MPF (self-reported ≈327kcal/day less vs. UPF), and the estimated daily energy–balance difference of ≈170kcal/day also favoured MPF.

Findings suggest that food processing matters beyond macronutrient/food-group guidance. Level of processing should be included in dietary advice and policy decisions. However, the trial was short (8 weeks per diet), mostly female and excluded some dietary groups. The results were also influenced by the order in which participants followed the diets and by varying adherence.

There is also an impact on families. A US-based prospective cohort study of nearly 20 000 predominantly white mother–child pairs demonstrated that maternal consumption of UPF in the ‘child-rearing period’ was associated with an increased risk of childhood and adolescent obesity in their children. There was a correlation between maternal consumption of UPF and offspring consumption; however, the association remained even when this was corrected for, suggesting additional mechanisms which the authors state might include peri-pregnancy consumption of UPF influencing in-utero epigenetics and offspring susceptibility to obesity in the future (BMJ 2022;379:e071767).

This remains a subject of debate! Our understanding of addiction is evolving rapidly, and, while food or specific foods are not included in the DSM-V criteria, substantial research has explored this area. The Yale Food Addiction Scale (YFAS) was developed to assess food addiction by aligning food intake behaviours with the DSM-V criteria for substance use disorder (Addiction 2023;118:589, Annu Rev Nutr 2021;41:387, Curr Addict Rep 2019;6:218).

Using this scale, two large meta-analyses revealed that 14% of adults and 12% of children meet the criteria for food addiction (Eur Eat Disord Rev 2022;30:85, Obes Rev 2021;22:e13183.). The prevalence is even higher among obese individuals awaiting bariatric surgery, at 32% (Obes Rev 2023;24:e13529), and exceeds 50% in those diagnosed with binge-eating disorder (Eur Eat Disord Rev 2022;30:85).

Some experts argue that formalising a diagnosis of food addiction may help progress our approach, management and policy (BMJ 2023;383:e075354).

“GPs to prescribe fruit and vegetables.” BBC, Friday 16 July 2021

“Education and willpower are not enough. We cannot escape this vicious circle without rebalancing the financial incentives in the food system.” National Food Strategy, 2021

Much of the solution lies in the hands of government and public health policy. We need population-level strategies to make group 1 foods more affordable and to provide support in developing skills around cooking and preparing fresh meals. Greater taxation of UPF along with additional regulation of the ‘Big Food’ industry may help us reduce our consumption of these foods; this may include limitations in advertising and clearer labelling.

These measures were suggested in the government-commissioned National Food Strategy in 2021. Most of these recommendations have subsequently been rejected, expressed politically in terms of a need for ‘cheap food’ during times of economic hardship. Time will tell as to whether this is a significant missed opportunity that further widens health inequalities. The BMJ editorial states, “nobody sensible wants food that causes illness” (BMJ 2022;378:o1972).

At an individual level as primary care clinicians, we can be honest and clear in our understanding of what these foods are and how they are associated with poor health outcomes. It does not look like (for now) we will be asked (or able) to prescribe fresh fruit and vegetables.

In some parts of the UK, social prescribers and health coaches (where available) may be able to signpost individuals and families to support with accessing fresh fruit and vegetables and with developing cookery skills. These services are far from universal.

	Ultra-processed food Highly-manufactured reconstituted food components. Cheap and widely available. Associated with a wide range of poor health outcomes. We need government and public health action to tackle this issue at population level. We may be able to signpost individual patients and families to support.
	Do you have a social prescriber or health coach in your practice? Do they have a list of resources to support people with accessing fresh fruit and vegetables, e.g. fresh food banks? Fresh food voucher schemes? Meal preparation support (provided by some local authorities and food banks)?
	Useful resources: Websites (all resources are hyperlinked for ease of use in Red Whale Knowledge) If you are interested, there are a number of charities in the UK that redistribute fresh foods with both health and environmental benefits: FareShare – what we do FoodCycle Trussell Trust Books Ultra-Processed People, Chris van Tulleken (Penguin, 2023)