select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
A study published in JAMA Neurology looks at association between consumption of ultra-processed foods and cognitive decline.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“This is another in the steady stream of studies on ultra-processed foods and various possible health consequences. In my view, it shares with other similar studies some quite serious difficulties of interpretation, and there is an additional interpretation issue that is more specific to this particular piece of research.
“First, it’s important to be clear on what the researchers are claiming. The abstract (summary) of the research paper says, in its Conclusions section, that “A higher percentage of daily energy consumption of ultraprocessed foods was associated with cognitive decline among adults from an ethnically diverse sample.” (Something very similar is said in the brief summary provided by the journal.) It’s important that this refers to an association, a correlation, and that doesn’t necessarily mean that the cognitive decline was caused by eating more ultra-processed foods.
“And further than that, the researchers in fact found that some types of cognitive decline with age occurred in all four of their groups of participants, defined by the percentage of their daily energy that came from consuming ultra-processed foods. That’s hardly surprising – it’s a sad fact of life that pretty well all of us gradually lose some of our cognitive functions as we go through middle and older age. What the researchers found was rather more subtle – effectively they compared how fast their measures of cognition were declining with age in groups that ate different amounts of ultra-processed foods, and found that the rate of decline was rather slower in the group that got less than a fifth of their food energy from ultra-processed foods than in the other three groups that got more of their energy from these foods. But, overall, the difference in the speed of cognitive decline between the groups was considerably less than the overall speed of cognitive decline in age.
“What this actually means is completed, I think. The researchers did not find a difference between the groups who ate different amounts of ultra-processed food in the rate of decline in their performance on memory tasks, which is perhaps strange given that they introduce their research paper by writing about dementia. (Dementia does involve more than memory loss, of course, but memory loss is an important aspect.) The researchers did find a reasonably clear difference between the group that consumed the smallest proportion of ultra-processed foods (less than a fifth of their energy intake), and the other groups, but not between any of the three other groups that consumed increasing amounts of ultra-processed food. I’m not a dietician, but is it really probable that there’s some sort of threshold consumption of ultra-processed food, beyond which faster cognitive decline occurs, but that it doesn’t occur any faster in increasing consumption levels above that threshold?
“The researchers recorded a ‘healthy diet’ or ‘heathy eating’ score for the participants, which was not solely based on consumption of ultra-processed food. In one data analysis, they looked separately at the effect of ultra-processed food consumption on the rate of decline of cognitive function in participants who had a low healthy diet score and those who had a high healthy diet score. (There were people from all four of the ultra-processed food groups in both of these healthy diet groups, it appears.) In those with a high healthy diet score, the researchers found no difference in the speed of cognitive decline between the four ultra-processed food groups, but they did find evidence of a difference in those with a low healthy diet score. That would be strange if the differences in cognitive decline were caused entirely or mostly by the consumption of ultra-processed food, and the researchers do discuss possible reasons for this difference.
“But the study doesn’t establish that differences in speed of cognitive decline are caused by ultra-processed food consumption anyway. That’s because its an observational study. The researchers didn’t tell the participants what to eat – they simply recorded it. So there are many differences between the groups that ate different amounts of ultra-processed food, apart from their consumption of ultra-processed food. Those differences could be, in part or in whole, the actual cause of the association that was found between ultra-processed food consumption and the rate of decline of cognitive function. Of course the researchers realised this, and they made statistical adjustments to allow for some of the other differences between the groups. However, they do, rightly, acknowledge that they may not have adjusted for everything relevant, so we still can’t be sure about cause and effect.
“Maybe the difference in the rate of cognitive decline between the group who got less than 20% of their food energy from ultra-processed foods and the others was, to some extent anyway, caused by the differences in consumption of ultra-processed foods – at least in those whose diets were not considered to be generally good. Or maybe it wasn’t. We can’t be sure. That’s why the researchers conclude that “limiting consumption of ultraprocessed food could be associated with reduced cognitive decline in middle-aged and older adults.” Note the ‘could’. If the consumption of ultra-processed food causes the differences in rate of cognitive decline, then eating less of it might slow cognitive decline, but if the cause is something else, then that won’t happen. And the findings using the healthy diet score are taken into account, shouldn’t general recommendations to improve diet be very important too anyway?
“In any case, quite how findings from any study of ultra-processed food relate to advice on diet is complicated by the fact that the ultra-processed category is defined very broadly. Some previous studies on associations between health and ultra-processed food consumption have found that associations depend on which particular type of ultra-processed food are being eaten. And the basis for the classification and its clarity have been criticised, for instance by a study in France earlier this year* that found that food and nutrition experts were not consistent with one another in allocating foods to the various categories involved.
* Braesco et al. (March 2022) “Ultra-processed foods: how functional is the NOVA system?”, European Journal of Clinical Nutrition, https://www.nature.com/articles/s41430-022-01099-1
Prof Gunter Kuhnle, Professor of Nutrition and Food Science, University of Reading, said:
“”Ultra-processed foods” have become a fashionable term to explain associations between diet and ill health, and many studies have attempted to show associations. Most studies have been observational and had a key limitation: it is very difficult to determine ultra-processed food intake using methods that are not designed to do so and so authors need to make a lot of assumptions. Bread and meat products are often classed as “ultra-processed”, even though this is often wrong.
“The same applies to this study: the method used to measure ultra-processed food intake was not designed for that task and relied on assumptions. This makes it virtually impossible to draw any conclusions.
“There are some interesting findings in this study: in contrast to many other studies, ultra-processed food intake was associated with higher income and better education which probably explains that those with the high intake appear to have a much lower prevalence of heart disease.
“The authors compared cognition between different categories of ultra-processed food intake. The results show only a very weak – if any – relationship. Only when comparing those with comparatively low ultra-processed food intake with everyone else – a rather unusual approach – did they find a statistically significant difference, but it is not clear whether this is clinically meaningful.
“Diet is one of several factors that can affect cognitive function and it is important to understand the relationship in order to make recommendations. However, such recommendations need to be based on robust data, taking into consideration risks and benefits and further implications, for example the cost of different foods and the respective impact on health. The current study does not provide strong evidence for a relationship between estimated ultra-processed food intake and cognitive function.”
Dr Duane Mellor, Registered Dietitian and Senior Teaching Fellow, Aston Medical School, Aston University, said:
“This is a large study from Brazil exploring the effects of what people said they ate at the start of the study might be associated with aspects of memory and cognition 8 years later.
“Although the idea of ultra processed food is interesting, it is noticeable that those who consumed more of these foods, also consumed more calories, and it is a shame the researchers did not include information in the paper about intakes of added sugar, salt and fat as it is not possible if using the label ultra processed foods is just another way of saying a less healthy dietary pattern. The paper suggests it controls for healthy diet score, and explored the association between low healthy diet score, intake of ultra processed foods and cognition. It appears that having a less healthy diet is associated with a lower cognition score.
“Although this study is interesting, it does not clearly show an association between ultra processed food and cognition as the effect seen was small, and as for the younger participants. It is perhaps, not the case that eating more ultra processed foods is a problem, it may be more the case they were eating less minimalist processed foods such as vegetables, fruit, nuts, seeds and pulses. So, it does not really change how we should try to eat to maintain good brain function and cognition, as we should try to eat less foods which are high in added sugar, salt and fat (which would includes many of the foods classified as being ultra processed) while eating more in terms of both quantity and variety of vegetables, fruit, nuts, seeds and pulses which are known to be beneficial for both our cognitive and overall health.”
‘Association Between Consumption of Ultraprocessed Foods and Cognitive Decline’ by Natalia Gomes Gonçalves et al. was published in JAMA Neurology at 16:00 UK TIME on Monday 5 December 2022.
DOI: 10.1001/jamaneurol.2022.4397
Declared interests
Dr Duane Mellor: “No conflicts of interest to declare.”
Prof Gunter Kuhnle: “No conflicts.”
Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee. My quote above is in my capacity as an independent professional statistician.”