A series of articles warns about the threat of ultra-processed foods to public health

What do you think of the articles overall? Are they of good quality?

"These are three excellent articles that highlight the role that ultra-processed foods have played in worsening the diet of the global population by displacing fresh foods. This leads to diets that are of poorer nutritional quality, high in sugars and unhealthy fats and low in fiber and protein, and also exposes the population to harmful chemicals and additives.

In addition, public health policies to reduce the consumption of ultra-processed foods have been evaluated. Many of these policies are similar to those already implemented in Spain by the Spanish Food Safety and Nutrition Agency, such as the elimination of these products from school cafeterias. Finally, it is indicated that the food industry is primarily responsible for the massive consumption of ultra-processed foods and that their production must be regulated."

What implications could this have?

"It could be a boost to public health policies aimed at reducing the consumption of these foods. The quality of the diet needs to be improved because many of the health problems we see in our society are due to an unhealthy diet.

In Spain, the Nutrition and Obesity Study Observatory met on November 18 to share many of the strategies that have been implemented in our country over the last year to reduce the prevalence of obesity in adults and children, which is very high. Some of these strategies involve reducing the population's exposure to the consumption of ultra-processed foods."

Are there any important limitations to consider?

"Most of the scientific evidence linking the consumption of ultra-processed foods to a multitude of chronic diseases has been obtained through observational studies, which have methodological problems in proving that this consumption actually causes the diseases studied, because other types of studies, such as clinical trials, cannot be done to demonstrate the cause-and-effect association. However, the studies conducted to date are sufficient to implement measures to protect the population from consuming this type of food."

The paper by Monteiro and colleagues raises further concerns about ultra-processed food (UPF), providing evidence from a comprehensive review of the literature, including studies across the globe.  The meta-analysis, where results from previous epidemiology studies are averaged to examine whether they agree, shows that a wide range of chronic diseases are associated with increased consumption of ultra-processed foods.  Despite the very different study designs used to study UPF, and the widespread geographic distribution of these studies, the results show that some major chronic diseases ailing modern life are associated with increased consumption of ultra-processed food.

The complication with the analysis though is a lot of other things have happened to our food systems across this time, so association may not be causation, as the authors freely admit.  The authors also consider randomised controlled trials, which are specifically set up to test causation – this approach provides much better evidence to investigate whether UPF contributes to ill health.  The problem here is that there are a lot less of these types of trials (only two), and the main conclusion I take from the evidence is we need more randomised controlled trials to be sure of the results.  There is also a lack of evidence for mechanism in terms of what things are causing and contributing to disease.  The authors discuss whether the adverse effects of ultra-processed foods are  caused by increased free sugar, increased saturated fats, high salt content, low fibre, trans fats, acrylamide, endocrine disruptors, hyper-palatability, high non-beverage energy density, disrupted food structures, soft-texture, low-content of health-protective phytochemicals, and toxic contaminants – this is a fairly comprehensive list of all the things we are currently worried about in nutrition!  Not all ultra processed foods will have these issues, and indeed some processing we know is beneficial – for example folate fortification of bread through flour and current efforts in the UK by the food industry to reduce salt, sugar and saturated fat in certain products are the positive side of food processing.

Despite these reservations with the understanding of mechanisms that UPF may cause ill health, the overall message is that we urgently need to understand how food processing on this scale influences our health and how ultra-processed food has a major impact on our risk of developing chronic diseases.

The authors conflate a large number of disparate concepts under the umbrella of 'ultra-processed foods'.  The concept of 'ultra-processed' foods, their impact on our health and the adoption of the NOVA classification tool is highly contentious.  There is huge pushback on these concepts from many nutrition experts as the idea that the processing of foods is to blame for dietary impacts on human health is far too basic an approach.  Simply correlating intake of these foods with the incidence of disease does not prove causality.  This also completely ignores the nuances of food in the context of our wider lives.

The authors surmise that 'ultra-processed' foods are inherently unhealthy due to the level and nature of their processing.  There is little convincing, high-quality evidence that 'ultra-processed' foods are inherently unhealthy.  Indeed, the authors of this paper base their claims on relatively weak evidence such as observational studies and narrative reviews.  To understand the true impacts 'ultra-processed' foods have on health we need numerous large-scale and robust randomised controlled trials.  We also need to move beyond these simplistic views of 'good' and 'bad' foods and instead truly address the fundamental issues in our food system that limit access and affordability of healthy foods.

The claims about the health effects of ultra-processed foods are largely based on observational data.  Randomised clinical trials - the gold standard in nutrition research - have not given cause for concern: while they have shown that some ultra-processed foods can result in overconsumption, they have also shown that a diet with ultra-processed foods that follows current dietary guidelines is not detrimental to health.

Estimating ultra-processed food intake in the general public is difficult, as most methods used in nutrition research have never been designed to estimate UPF intake.  Claims about consumption at a population scale therefore need to be interpreted very carefully as some of these figures are based on very crude assumptions, e.g. that all bread consumed in the UK is ultra-processed.

Many observational studies rely on dietary assessment methods that are unable to estimate actual UPF intake because they do not distinguish e.g. between UPF and non-UPF versions of the same foods such as bread, yoghurt or cereals.  The authors claim that such measurement errors would be inconsequential without explaining why this should be the case.

Without reliable data on ultra-processed food intake at the level of the individual - i.e. not purchase data - it is impossible to make any claims about the health effects of these foods.  The difficulty of estimating intake is not a new criticism and has been raised by many scientists, including the UK’s Scientific Advisory Committee on Nutrition (SACN).  It is therefore very disappointing that neither the authors nor the reviewers have engaged with this criticism.