August 26, 2020

expert reaction to study looking at the PHE Eatwell Guide and health and environmental impacts

A study, published in BMJ Open, looked at the PHE Eatwell Guide and the health and environmental impacts.

Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“This is a very interesting piece of research, but there are some problems with interpreting it.  We can’t be sure that the improvements in survival rates, for people who stick to more of the dietary recommendations, are actually caused by them sticking to the dietary recommendations.  That might be the cause, but it might not be. In some ways, it’s not clear how big the reductions in mortality are.  And there are some questions about how typical the people who provided the data are of the national population.

“The study is observational – that is, the people, who were studied, weren’t allocated to particular diets, but just ate what they ate.  That was recorded, along with whether they died.  But there will be many differences between people who followed, or didn’t follow, dietary recommendations, apart from whether they followed the recommendations.  If some of these differences (known as confounders) are also related to death rates, it could be that the confounders are the actual cause of differences in death rates, and not the aspects of diet that were studied.  It’s possible for statistical adjustments to be made, to allow for potential confounders on which the researchers have information, and indeed these researchers generally made a good job of both making these adjustments and checking on the effects of their adjustments.  For instance, you might expect that people who stick to more of the dietary recommendations would tend to be more health-conscious in a lot of ways, and maybe it is an aspect of this tendency other than diet that is the cause of the differences in death rates.  The researchers adjusted (among other things) for how much exercise the participants took, whether they smoked, their body mass index (which might be lower in some people who think a lot about their health), and their blood pressure status.  But adjustments can’t be made for confounders on which no data are available, and one can never be sure that everything appropriate has been taken into account.  There may be other important aspects of being health-conscious that couldn’t be adjusted for.

“One potential confounder that seems not to have been adjusted for is age.  I can find no mention in the paper of the age of the participants being directly taken into account.  (The study that they were in – data from three different studies was used – was taken into account, and because the studies took place at different times and recruited people from different age groups, the average ages in the studied would have differed.  But I can’t see a mention of account being taken of the different ages of people within a particular study.)  Obviously the chance of a person dying in a given period of time depends heavily on how old they are.  If the participants’ age has no association with whether they follow dietary recommendations, then not directly accounting for age would not lead to a confounding problem – but I think it’s plausible that people in different age groups might have different patterns of following the diet recommendations, for a lot of potential reasons, so I am surprised not to see age on the list of factors that were adjusted for (or on the list of factors used for stratification, which can be though of as a different type of adjustment).

“Even though we can’t be certain that the decrease in death rates is actually caused by sticking to dietary recommendations, it’s still worth looking at how big the differences are in death rates between people who stick to different numbers of the recommendations.  People who adhered to five or more of the nine recommendations had a mortality rate 7% smaller than those who adhered to two or fewer, after adjustments, thought there’s some statistical uncertainty about the exact figure, and the difference could plausibly be somewhere else between 3% and 10%.  But 7% (or 3% or 10%) of what exactly?  In the end, we will all die, and it obviously can’t mean that 7%  of people who stick to five or more recommendations won’t die at all.  In fact, I believe, the researchers compared the risks of death in each of the three studies they looked at, over the period of follow-up for each study.  The length of the follow-up time, on average, was very different for the three different studies, ranging from 4 years for one of them to 21 years for another, and also the participants in the different studies had different age distributions, and one of them (the biggest) involved only women.  So, when they report that the death rate was 7% lower for those who stuck to five or more requirements, that’s an average across the three studies (and across men and women, in the studies that involved both).  Nothing wrong with that approach statistically, but it doesn’t allow us to say directly, for instance, what the chances of dying in the next 10 years would be for women aged 50 who stick to four or more recommendations, compared to similar women who stick to two or fewer.  This makes it somewhat difficult to judge how important a 7% reduction in death rates actually is.

“A final question is whether these results, that were based on data from three studies that participants chose to join, are really applicable to the population in general.  The overall figures in the report on how many recommendations people adhere to, and on what percentages of people adhere to each of the nine specific recommendations, come from a fourth study, the National Diet and Nutrition Survey (NDNS).  That survey takes data from a representative sample of the UK population, so the figures based on the NDNS will be reasonably representative of the UK as a whole.  The percentages that adhered to specific numbers of the nine recommendations, in the three studies that provided the data on mortality, were pretty similar to those from the NDNS.  However, that seems not to be the case for the percentages sticking to some of the nine specific recommendations.  For instance, nationally (according to NDNS), 80% of people stick to the recommendation on not eating too much total fat, but the percentages in the three studies included for the death rate data were 47%, 49%, and 62%, considerably lower.  And on the recommendation to eat at least 5 portions of fruit and vegetables daily, which nationally was met by only 26% of people, the percentages meeting it in the three studies were 49%, 56%, and 68%, considerably higher than national.  So there must be a certain amount of doubt as to whether the participants, who provided the data on the association between meeting the diet recommendations and death rates, are typical enough of the national population for the detailed results to be applicable nationally.”

Catherine Collins RD FBDA, NHS Dietitian, said:

“This ambitious review attempts to evaluate the potential impact on personal health and the environment if more of the UK population were to follow the ‘Eatwell Guide’ dietary approach – the healthy eating recommendations of Public Health England.

“Using three large UK based population studies (which increase the potential relevance of their findings), the authors conclude that if the population were to adopt the ‘Eatwell Guide’ approach en masse there would be a significant benefit to personal health, plus a reduction in the obligate (non-rainfall) ‘blue’ water demands and greenhouse gas (GHG) emissions associated with current UK dietary habits.  A win-win of personal health and environmental protection is inferred.

“To model this, population dietary records were judged against nine predefined criteria and the environmental impact of aligning each group with the Eatwell Guide recommendations was calculated.

“For foods, estimates of fruit and vegetables, oily and non-oily fish, and red or processed meat intakes were included in this modelling.

“Nutrient intakes were confined to those which typically grace front-of-pack labelling in the UK: salt, sugar, saturated and total fat intake, plus dietary fibre.

“Both food and specific nutrient intakes were used to qualify the healthiness of the ‘Eatwell Guide’ by associating a dietary score with mortality data.

“Other foods contributing to the typical UK diet, which remain part of the Eatwell Guide have not been included in the modelling.  Cereals and grains, dairy foods and even alcohol (in moderation) appear to be beneficial as part of a varied diet, yet their environmental impact has not been considered.  Their omission undermines the authors’ calculations and predictions made against the ‘Eatwell Guide’ approach.

“Of interest, the combined data from the UK Biobank, EPIC-Oxford and the Million Women Study on salt and sugar reduction, total fat, fibre and red or processed meat intake failed to demonstrate a compelling impact on population mortality with adherence to recommended guidelines.  Only an increase in fruit and vegetable intake to 5 or more portions a day was associated with a significant reduction in mortality, and from environmental data presented, a significant reduction in CO2 production, GHG emissions and water usage.

“What’s to be made of this?  Professionally, my concern is always more about the quality of the diet, rather than what is reduced or removed from it.  This analysis demonstrates the greater environmental footprint associated with meat production when compared to the impact of plant-based foods.  The most recent NDNS survey showed the averaged daily UK meat intake is within the Eatwell dietary guidelines.  However, dedicated carnivores and those following diets with a focus on increasing protein intake at the expense of carbohydrates (such as Paleo, Keto and LCHF (low-carb, high-fat) diets) must consider their dietary choices as damaging to the environment,and possibly their long term health.  We can take from this and other established work that fad diets based on a high protein intake are neither beneficial for health nor our globe.

“The potential health risks of food substitution is important – but beyond the remit of this study.  Swapping 90g of meat for an equivalent weight of potato or chickpeas reduces the environmental impact associated with food production, but doesn’t address protein, iron and vitamin B12 deficits from the swap.  Dietary adequacy relates to both the absolute amount of a nutrient within a food (e.g. the protein or fibre content of a typical portion of red meat versus a portion of red kidney beans), but also changes in nutrient bioavailability and uptake influenced by other foods.  ‘Let them eat rice/pasta/potato’ may well be a valid environmental mantra, but it isn’t a dietary one if nutritional quality of the whole diet isn’t addressed.

“Individually, we should be trying to incorporate more plant-based foods into our diet, but given our reliance on fruit and vegetable imports, this can’t be at the local environmental expense of countries supplying our demands.

“In summary, the environmental benefit predicted per capita from following the ‘Eatwell Guide’ approach is attenuated by two key aspects.  First, the incomplete modelling of the typical UK diet in this current paper, and second, the impact of a burgeoning population on the environment, due in part to the increased longevity of those citizens adopting a healthier diet and lifestyle.”

Prof Gunter Kuhnle, Professor of Nutrition and Food Science, University of Reading, said:

“This is a very important study as it investigates the impact of dietary guidelines on health.  There has been a lot of criticism of dietary guidelines, and some have blamed them for rising rates of obesity and chronic diseases.  This study compared the health of people who follow the dietary guidelines and those who do not in more than 500,000 people and found that those following these recommendations where healthier than those who did not (although the difference in mortality was small).  Following the dietary guidelines therefore is clearly associated with better health and does not increase the risk of chronic disease as is often claimed.

“The study also highlighted that only very few people follow the dietary guidelines: less than 0.1% of the general population.  While many do follow the recommendations to reduce fat, salt and meat intake, fewer than 10% consume sufficient fibre and only 25% meet the recommendations regarding fish, fruits and vegetables and free sugars.

“Dietary guidelines are therefore important to provide recommendations to the general public, and there would be a clear health benefit if more people would follow them.  It is however important to remember that dietary guidelines are for healthy people and cannot replace advice given by health professionals such as dietitians.  The study also shows that following these recommendations does not only benefit health, but has also a positive impact on the environment.

“The study itself has been very well conducted and is very robust.  The authors combined data from three large UK cohorts, and in addition used the representative National Diet and Nutrition Survey (NDNS).  The dietary assessment methods used are ideally suited to investigate dietary patterns and therefore assess adherence to diet based recommendations.”

Dr Ada Garcia, Senior Lecturer in Public Health Nutrition, University of Glasgow, said:

Does the press release accurately reflect the science?

“Adherence to guidelines was measured using all data bases, not just NDNS.

Is this good quality research?  Are the conclusions backed up by solid data?

“Yes, the large sample size allows this type of analysis although there are limitations of using secondary data that was not initially intended to fit the aims of the present study.  Using different methods for data acquisition to pool data takes into consideration lots of assumptions.

How does this work fit with the existing evidence?

“The findings of this study confirm the importance of the government Eat Well Guidance for chronic disease prevention and the need to consider environmental aspects in future population-based recommendations.

Have the authors accounted for confounders?  Are there important limitations to be aware of?

“Yes, several essential confounders were included in the analysis.  Of particular importance was the inclusion of sex as a confounder because when analysing individual cohort contributions it is clear that the Million Women Study contributed greatly to the significant findings of the overall study.

What are the implications in the real world?  Is there any overspeculation?

“The implications of this study contribute to consolidate the importance of following dietary guidelines for chronic disease prevention.  Eating more fruit and vegetables is a key recommendation, the next step is to consider practical advice in regards sustainability that take into account other factors such as cost, food choice and culture.”

Prof Andrew Salter, Professor of Nutritional Biochemistry, School of Biosciences and Future Food Beacon, University of Nottingham, and Trustee of the Academy of Nutrition Sciences, said:

“The paper by Scheelbeek et al provides an interesting analysis of the health and environmental impact of following the dietary guidelines presented in the Eatwell Guide (EWG).  It draws upon pre-existing high-quality data, to estimate the impact of adhering to the various recommendations within EWG on mortality.  It also uses existing data to model the effects of adhering to the guidelines on greenhouse gas emissions and water use.  The data sources and analysis appear appropriate and robust with due recognition of any limitations.

“Perhaps one of the most remarkable findings is the low level of adherence to the guideline currently in the UK, with only 0.1% of the population adhering to all 9 of the guidelines and 30.6% adhering to at least 5.  This is perhaps disappointing, in view of the fact that the Eatwell Plate has been the cornerstone of the UK government’s dietary policies since 2007 (albeit updated to EWG in 2016).  In view of all the uncertainties around the future of Public Health England, this calls into question what further measures the government can take to improve the diet of the country.  It also throws into question whether even more draconian measures, such as the EatLancet Diet, would ever be adopted by the majority of the population of the UK.

“The relative impact of different components of the EWG on mortality is also intriguing.  The most beneficial component appears to be adhering to the recommended fruit and vegetable intake (essentially 5-a-day).  More surprisingly, reducing red and processed meat intake to the recommended 70g, or less, day appeared to have no impact on mortality.  This appears to conflict with the widely reported assertion that reducing meat intake would have both health and environmental impacts.

“There is increasing recognition that dietary guidelines should take into consideration both health and environmental impacts of our food systems.  While a movement towards the EWG would have favorable effects on both, the authors point out that even further changes would be required to meet emission targets.  Linking the health of the individual and of the planet appears to be one strategy which may help persuade Western populations to change their dietary habits.  However, this paper illustrates how far away most of the UK population is from accepting such changes.  It is also important to recognise the impact of such a shift on our agricultural industry and reliance on food imports.  Equally, we need to ensure that any such changes do not negatively impact on the ability of the most vulnerable, and economically challenged, of our ageing population to consume a healthy and balanced diet.”

Dr Duane Mellor, Registered Dietitian and Senior Teaching Fellow, Aston Medical School, Aston University, said:

“This is a significant  analysis of over 1/2 million people’s diet looking at how ‘adherence’ to  the UK’s EatWell Guide is associated with risk of death.  This seems to suggest that there is a large reduction risk in death for those following the UK EatWell Guide, however as much of the data was collected before the guide was even published, it appears to be the case that those following the diet might not be doing as deliberately as we first might think!  Therefore, caution has to be used when using evidence from old eating habits to show that current nutritional guidelines are correct, as this risks bias, as some of these studies may have actually helped to shape the same guidelines.

“A big issue is the description that less than 0.1% of the NDNS sample adhere to all nine EatWell Guide recommendations – this data is based on only 4 days intake, and some of these recommendations, such as eating fish is not an everyday behaviour, so could potentially be missed.  Also the language itself is problematic, why do people struggle to eat healthily?  Perhaps, the use of controlling language like ‘adherence is low’ could almost  seem to place the blame with the individual.  Even though the authors conclude that governments should increase encouragement to improve diets, it does not solve the issues of social inequality and food poverty underpinning many of the health and dietary issues face in the UK.  This highlights the need for a more inclusive attitudes to supporting healthier food environments where it is easier, and also more normal to choose and enjoy healthier and minimally processed foods.  This study does very little to consider the effect of poverty on food choice, something through the current pandemic has become even more apparent in the UK as well as in other countries.

“There is also potential bias in that two of the nine EatWell Guide recommendations are based on fish, this firstly weights the score against vegetarians, and secondly with the other focus of the paper being environmental impact is ignored by the omission of a score assessing good marine stewardship within the rating of the environmental impact of diet.  Fish as a commodity is also an issue, both in terms of cost again biasing the EatWell Guide score against those with the least to spend on food, but also (excluding tinned fish) disadvantages those living in food poverty as fish require freezing or refrigeration to extend shelf life, which is not something that everyone has adequate access.

“Although green house gasses and water footprint were assessed, it is unlikely that the details in the food records would have allowed country of origin to be identified, so assumptions based on using UK data for food items risks further error.  Also when looking at the environmental impact of a person’s diet, it is vital to look at food waste not just food eaten, this study only looking at what was eaten, so only considers part of the actual impact.  Again, as fresh food with lower greenhouse gases and often water use can often need refrigeration and more frequent purchasing, this can disadvantage those without the resources and facilities.

“Although this paper is interesting, it appears to miss one of the key issues facing our society with respect to food and nutrition and ultimately which drives much of the ill health too – that is the inequality and unequal access to healthy fresh food that is affordable and people can enjoy.”

Dr Ian Johnson, Nutrition researcher and Emeritus Fellow, Quadram Institute Bioscience, said:

“In this very rigorous observational study, the authors have made an ambitious attempt to measure the impact of the government’s nutritional guidelines, summarised as The Eatwell Guide, on both total mortality as an index of public health, and on the ‘environmental footprint’ of consumers, as an index of planetary wellbeing.  It is striking that although only less than one percent of the UK population complied with all of the nutritional guidelines, those who achieved only moderate to high adherence still experienced a significant reduction in their total risk of mortality, whilst at the same time reducing their overall environmental impact.

“With regard to health, the greatest benefit was associated with the recommendation to consume higher levels of fruits and vegetables.

“The great strength of this investigation is that it used real dietary intake data to reach these conclusions, rather than a theoretical modelling approach.

“It is pleasing to see that the UK government’s dietary advice, compiled and promulgated through the work of Public Health England, has thus been validated and shown to be of very tangible benefit to public health.  This evidence is timely, coming as it does when the vital role of Public Health England in the prevention of diet-related diseases seems to have been thrown into some uncertainty by the abrupt departmental reorganisation recently announced by the government.”

Prof Dominic Moran, Professor of Agricultural and Resource Economics, Roslin Institute, University of Edinburgh, said:

“The evidence on good health and environmental outcomes related to ‘planetary health’ diets is becoming more convincing.  This study provides more epidemiological evidence, which is an advance on Lancet-EAT, which was more speculative.  But it stops short of showing (or stratifying) who eats such diets and why.  Drilling into the populations of consumer that do and don’t adhere would reveal the socioeconomic and cultural barriers that might be hard to crack without subsidy.”

Prof Judith Buttriss, Director General, British Nutrition Foundation, said:

“This study is important and supports previous analyses.  It demonstrates that better adherence to the UK’s Eatwell Guide recommendations can deliver both health (7% reduction in mortality risk) and environmental benefits in terms of dietary greenhouse gas emissions (30% reduction), compared to very low compliance.  Yet, currently, less than 0.1% of the nationally representative National Diet and Nutrition Survey (NDNS) sample adhere to all the recommendations and less than a third (30.6%) adhere to at least 5 out of the 9.  Poorest adherence was for fibre (recommended 30g/day; 7.2% achieved this) and oily fish (recommended one serving a week; 16.8% met this), while over half of the sample population met the recommendations for red and processed meat (less than 70g per day), saturated fat, total fat and salt.  [The 9 criteria concern fruit and veg, oily fish, other fish, red & processed meat, fibre, salt, free sugars, saturated fat and total fat.]

“Compared with low adherence, compliance with 5-9 recommendations (intermediate to high) was associated with a 30% lower dietary footprint in terms of greenhouse gas emissions (GHGE), but water footprint did not fall.

“Adherence to 5 or more of the 9 UK government’s Eatwell Guide recommendations was associated with a 7% lower mortality risk (compared to those with low adherence), with the biggest reduction in risk associated with adherence to consuming 5 or more servings of fruit and veg a day.  It is noteworthy that adherence to the fruit and veg recommendation was also associated with an increased blue water footprint of the diet (+28.5 L/person/day), highlighting the importance of not just considering GHGE when assessing the environmental impact of food group changes in the diet.

“Strengths of the study include use of high quality empirical data from three large prospective UK cohort studies and the NDNS, that empirical rather than modelled diets were used, and that UK-specific data were used to determine greenhouse gas emissions and water use footprints for UK produced foods.

“In the discussion, the authors caution that an increasingly large proportion of plant-based food imports originate from countries that are highly vulnerable to climate change and water shortage, implications of which need to be taken into account.  They also remind readers that dietary change is only one of a number of ways by which we can personally influence environmental impact, others include active travel and lower emission vehicles.”

‘Health impacts and environmental footprints of diets that meet the Eatwell Guide recommendations: analyses of multiple UK studies’ by Pauline Scheelbeek et al. was published in the BMJ Open at 23:30 UK time on Wednesday 26 August 2020.

DOI: 10.1136/bmjopen-2020-037554

Declared interests

Catherine Collins: “No conflict of interest.”

Prof Gunter Kuhnle: “I don’t have any conflict of interest to declare.”

Dr Ada Garcia: “None.”

Prof Andrew Salter: “I currently receive research funding from ABAgri, Herbalife and McCain.  None of these organisations have been involved in producing the comments below.”

Dr Duane Mellor: “I have no financial links, however personally I chose to follow a vegetarian diet, but will always support others to follow which ever dietary preference they have.”

Dr Ian Johnson: “No conflicts.”

Prof Dominic Moran: “I declare no interests in this or any of these studies.”

Prof Judith Buttriss: “Professor Buttriss is a trustee of the academy of Nutrition Sciences and has been a member of UK government expert/advisory committees considering topics such as nutrient profiling, food based dietary guidelines and dietary surveys.  As Director General of the British Nutrition Foundation (BNF – a UK charity that engages with healthcare professionals, academics, schools, government, the food industry and the media), she provides advice on a variety of nutrition and food related matters to stakeholders across the nutrition sector, including a range of food/beverage companies (all fees are paid directly to BNF).  BNF’s funding comes from a variety of sources including EU projects; contracts with national government departments and agencies; conferences, publications and training; membership subscriptions; donations and project grants from food producers and manufacturers, retailers and food service companies; and funding from grant providing bodies, trusts and other charities.  BNF is not a lobbying organisation nor does it endorse any products or engage in food advertising campaigns.  More details about BNF’s work, funding and governance can be found at www.nutrition.org.uk/aboutbnf.”

None others received.