September 19, 2018

expert reaction to study looking at nutritional quality of food and risk of cancer

Researchers, publishing in PLOS, report that the consumption of food products with lower nutritional quality is associated with  a higher risk of cancer.

Dr Elizabeth Lund, Independent Consultant in Nutrition and Gastrointestinal Health, said:

“The authors of this paper have looked at the overall quality of people’s diets and their chance of getting cancer over roughly 15 years after assessing their diet.  Assessing the impact of overall diet quality has the benefit of highlighting the importance of total diet rather than one specific food that could then be labelled as either “healthy” or “unhealthy”.  It is the diet that is healthy or unhealthy, generally not specific foods.

“The aim of the study was to investigate the use of a number called the Nutri-Score based on simple food labelling similar to the traffic light system we see on some foods sold in the UK.  The authors are in effect checking this scoring system gives sensible results in terms of what we already know about diet and chronic disease.

“After correcting for a large number of factors known to affect cancer risk, such as how fat people are, the authors show a small benefit of eating a diet low in energy, sugar, saturated fat and salt and high in fibre, protein, fruits and vegetables and nuts.  The risk of getting cancer in the 15 year follow up was about 7 people per 1000 people in the group eating the healthiest diet as assessed by the “Nutri-Score” system compared to 8 people per 1000 in those assessed to be eating the poorest quality diet.

“As expected bowel cancer was most affected by quality of diet followed by prostate and breast, but these are common cancers so it’s important to do all we can to reduce the risk of getting them both at an individual and a population level, and this method of food labelling could be useful in helping to achieve this.”

Prof David Spiegelhalter, Winton Professor for the Public Understanding of Risk, University of Cambridge, said:

“The association reported in this study between the nutritional quality of food that people ate and their risk of getting cancer is very weak and certainly does not justify the authors’ bold conclusions.

“To put their claim of an ‘increased risk’ in perspective, it means that 1,000 people eating the diet they class as ‘lowest quality’ would get just 1 extra cancer between them in 20 years, compared to 1,000 people on the ‘highest quality’ diet: that is 15 instead of 14.  This is a tiny effect and might well be explained by inadequate adjustment for other related factors such as social class.  Even the authors admit in the paper that these are ‘rather weak’ associations.  This is essentially a null study, in spite of the claims about an increased risk.”

Prof Tom Sanders, Professor emeritus of Nutrition and Dietetics, King’s College London, said:

“Nutrient profiling (NP) is claimed to be a measure of the healthiness of different foods.  The nutrient profiling model used in the UK gives scores to foods based on the levels of energy (kcal), saturated fat, salt and sugar per 100 grams of food.  If the food contains mainly fruit, vegetables, nuts and beans, and/or is rich in protein or dietary fibre, there is a downward adjustment of the score; so low scores are healthier and high scores are less healthy.  Foods that get higher scores are cakes, biscuits, puddings, some high sugar-containing breakfast cereals, confectionery, crisps, sugary drinks, fats including olive oil, red and processed meat including sausages and burgers, cheese, lasagne, pizza, some oily fish such as smoked mackerel, whole milk and condiments such as tomato sauce.

“What this study has done is to calculate dietary score by averaging the nutrient profiling scores for every food reported as being consumed in a food frequency questionnaire, which asks questions about 124 items of food and drink.  It has then made adjustments for the contribution made by each food to the total energy intake.  This is an improvement on the profiling system used in the UK which does not make an allowance for contributions made to total intake.  The study also uses a French adaptation of profiling which treats dairy products more favourably.

“As expected, there was a trend for lower intakes of fruit and vegetables and less fish, and higher intakes of meat especially red and processed meat, with increasing dietary score – higher scores are given for more unhealthy foods.  But, the energy intake was about 500 kcal lower in the group reporting the lowest score compared to those with the highest score – meaning those who ate more healthy foods also reported to eat fewer calories in total.  However, body mass index (a measure of obesity linked to increased cancer risk) was actually highest in the group with the lowest score!  This would suggest under-reporting of food intake in that group, so the method of measuring diet here may not be very robust.

“The paper refers to the diets with high scores as being of low nutritional quality.  However, it is important to stress that diets with higher scores cannot accurately be described as nutritionally inadequate or being ‘junk food’, as high score foods include things like beefburgers and cheddar cheese.  Even nutritionally adequate diets (i.e. those that meet all nutrient requirements) may increase risk of cancer especially if they contain carcinogens or are eaten in excess.

“The study divided the participants into five equal groups according to the calculated dietary score and compared the risk of cancer in those in the top 20 percent with those in the bottom 20 percent, and has made statistical adjustments for known contributions to risk of cancer.  The main finding is that the risk of all cancers was 7 percent higher in those with the highest score (the least healthy diets).  In general, diet is believed to be responsible for about a third of all cancers, but other environmental factors such as occupational exposures (e.g. asbestos), ionising irradiation, exposure to some chemicals and infection (e.g. hepatitis B, papilloma virus, leukaemia virus, Helicobacter pylori) are also important.  Dietary habits are often intertwined with other adverse exposures especially among manual and unskilled workers in urban environments.  There were differences in smoking habit and alcohol intake between groups, with a trend for these habits to be greater in those with higher scores (eating the least healthy foods).  When adjustments were made for smoking and alcohol, the risk of lung cancer and cancer of the aerodigestive tract such as mouth and throat cancer (both known to be linked to smoking and the latter to alcohol), and stomach cancer, were no longer significantly linked to diet.  Consequently, the abstract and press release are not justified in claiming the dietary score is linked to cancer at these sites.

“Previous studies on diet and cancer in the same group of people have shown red and processed meat intake to be associated with a 35 percent greater risk of colorectal cancer.  In this report, the increased risk of colorectal cancer was weaker (11%) among those with a high dietary score (eating the least healthy diets) and there was no significant association at all for men.

“The other differences reported of cancer incidence between groups are of marginal statistical significance and, given that so many comparisons were made, some statistically significant differences may have arisen by play of chance.  The results of this study are disappointing particularly as the score did not predict which people would become overweight or obese, which was one main reasons for introducing nutrient profiling in the UK.  So, what I take from this study isn’t that unhealthy diets don’t cause cancer, but that the method used in this study of scoring the healthiness of foods may not be particularly reliable.”

Declared interests

Dr Elizabeth Lund: “I have no specific interests to declare. I am an independent consultant who used to be a Research Leader at IFR now called Quadram Institute of Biological Sciences. I am an honorary senior lecturer at The University of East Anglia (School of Biological Sciences).”

Prof David Spiegelhalter: “No conflicts of interest.”

Prof Tom Sanders: “Honorary Nutritional Director of HEART UK.  Scientific Governor of the British Nutrition Foundation.  He is now emeritus but when he was doing research at King’s College London, the following applied: Tom does not hold any grants or have any consultancies with companies involved in the production or marketing of sugar-sweetened drinks.  In reference to previous funding to Tom’s institution: £4.5 million was donated to King’s College London by Tate & Lyle in 2006; this funding finished in 2011.  This money was given to the College and was in recognition of the discovery of the artificial sweetener sucralose by Prof Hough at the Queen Elizabeth College (QEC), which merged with King’s College London.  The Tate & Lyle grant paid for the Clinical Research Centre at St Thomas’ that is run by the Guy’s & St Thomas’ Trust, it was not used to fund research on sugar.  Tate & Lyle sold their sugar interests to American Sugar so the brand Tate & Lyle still exists but it is no longer linked to the company Tate & Lyle PLC, which gave the money to King’s College London in 2006.  Tom also used to work for Ajinomoto on aspartame about 8 years ago.  Tom was a member of the FAO/WHO Joint Expert Committee that recommended that trans fatty acids be removed from the human food chain.  Tom has previously acted as a member of the Global Dairy Platform Scientific Advisory Panel and Tom is a member of the Programme Advisory Committee of the Malaysian Palm Oil Board.  In the past Tom has acted as a consultant to Archer Daniel Midland Company and received honoraria for meetings sponsored by Unilever PLC.  Tom’s research on fats was funded by Public Health England/Food Standards Agency.”