Professional Documents
Culture Documents
1 s2.0 S0924224419302031 Main
1 s2.0 S0924224419302031 Main
A R T I C LE I N FO A B S T R A C T
Keywords: Background: There has been a growing body of literature published on ultra-processed foods (UPF). This has
Ultra-processed foods (UPFs) been driven by an ambiguous definition that can be easily applied to observational work. Nutritionally, UPF
Classification systems have been given the umbrella label of being “energy-dense, high in unhealthy types of fat, free sugars and salt, and
Nutritional profiling poor sources of protein, dietary fibre and micronutrients” (Monteiro et al., 2018). Unfortunately, such overarching
Protein
statements are imprecise and not necessarily reflective of all modern food products.
Fibre
Scope and approach: This commentary evaluates the NOVA classification system and includes an analysis of 50
Fortification
foods falling under its definition of being ‘ultra-processed’ yet classified as ‘healthy’ according to the UK Nutrient
Profiling Model. Nutritional profiles were also compared against European Regulation (EC) No 1924/2006
Nutrition Claims.
Key findings and conclusions: No statistically significant correlations were found between the number of in-
gredients and energy, saturated fat, total sugar, sodium, AOAC fibre and protein. The majority of UPFs identified
were low in saturated fat (84 percent), low in sugar (80 percent) and a source of fibre (60 percent). Thirty eight
percent were low salt and 30 percent a source of protein.
Based on the findings of this commentary it is suggested that rigorous nutritional analyses should be con-
ducted alongside ‘processing loads’. Overarching and tautological definitions could lead to the public avoiding
all UPFs and have unintended consequences such as fibre and micronutrient shortfalls.
1. Introduction fat, refined starches, free sugars and salt, and poor sources of protein,
dietary fibre and micronutrients” (Monteiro et al., 2018).
Interest in ultra-processed foods (UPF) has been mounting with di- This classification is rapidly being embedded within observational
chotomy between viewpoints becoming increasingly evident. Most re- research, partly due to its ease of application. However, concerns about
cently it has been concluded that food processing has reached a the application of such definitions are now being raised. For example,
‘crossroads’ – historically it has been important from a preservation, there are no specified cut –offs for food components e.g. energy, fat,
palatability, stability and convenience perspective (Knorr & Watzke, sugar and salt (per gram or portion) making conclusions about the
2019) yet a stronger emphasis on health and classifications systems nutritional profile of UPFs rather cumbersome (Gibney, Forde, Mullally,
now grading ‘levels’ of food processing appear to be placing processed & Gibney, 2017). It has also been pointed out that consumer confusions
foods under closer inspection. about UPF and the avoidance of foods deemed to fall within this cate-
In particular the NOVA system classifies foods into four distinct gory such as cereals and wholegrain/enriched breads could even have
categories: Group 1 – Unprocessed or minimally processed; Group 2 – inadvertent consequences and reduce folate, calcium and fibre intakes
Processed culinary ingredients; Group 3 – Processed foods and Group 4 (Jones, 2018).
– Ultra-processed foods (Monteiro, 2016). UPF (and drinks) forming Food Standards Australia and New Zealand have recently high-
Group 4 are defined as: “Industrial formulations typically with five or more lighted the need to develop nutrient profiling knowledge to enable the
and usually many ingredients.“ (Monteiro, 2016) (full definition; application of more reliable and rigorous health claims (Pivk Kupirovic
Table 1). Such a system, however, does not consider the nutritional et al., 2019). The French have recently developed the simplified nu-
profile of foods nor does it account for methods used in the culinary trition labelling system (SENS) algorithm which has been adapted from
preparation of food (Monteiro, 2016). NOVA authors also conclude the SAIN, LIM system (Tharrey, Maillot, Azais-Braesco, & Darmon,
that: “All together, they (UPF) are energy-dense, high in unhealthy types of 2017). In the UK an updated Nutritional Profiling Model has recently
https://doi.org/10.1016/j.tifs.2019.08.023
Received 14 March 2019; Received in revised form 1 August 2019; Accepted 29 August 2019
Available online 05 September 2019
0924-2244/ Crown Copyright © 2019 Published by Elsevier Ltd. All rights reserved.
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
99
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
Table 2
EFSA Nutrition Claims applied to UPF identified.
Low Energy A claim that a food is low in energy, and any claim likely to have the same meaning for the consumer, may only be made where the product does not
contain more than 40 kcal (170 kJ)/100 g for solids or more than 20 kcal (80 kJ)/100 ml for liquids. For table-top sweeteners the limit of 4 kcal (17 kJ)/
portion, with equivalent sweetening properties to 6 g of sucrose (approximately 1 teaspoon of sucrose), applies.
Low in Fat A claim that a food is low in fat, and any claim likely to have the same meaning for the consumer, may only be made where the product contains no more
than 3 g of fat per 100 g for solids or 1.5 g of fat per 100 ml for liquids (1.8 g of fat per 100 ml for semi-skimmed milk).
Low Sugars A claim that a food is low in sugars, and any claim likely to have the same meaning for the consumer, may only be made where the product contains no
more than 5 g of sugars per 100 g for solids or 2,5 g of sugars per 100 ml for liquids.
Low Sodium/Salt A claim that a food is low in sodium/salt, and any claim likely to have the same meaning for the consumer, may only be made where the product contains
no more than 0,12 g of sodium, or the equivalent value for salt, per 100 g or per 100 ml. For waters, other than natural mineral waters falling within the
scope of Directive 80/777/EEC, this value should not exceed 2 mg of sodium per 100 ml.
Very Low Sodium/Salt A claim that a food is very low in sodium/salt, and any claim likely to have the same meaning for the consumer, may only be made where the product
contains no more than 0,04 g of sodium, or the equivalent value for salt, per 100 g or per 100 ml. This claim shall not be used for natural mineral waters
and other waters.
Source of fibre A claim that a food is a source of fibre, and any claim likely to have the same meaning for the consumer, may only be made where the product contains at
least 3 g of fibre per 100 g or at least 1,5 g of fibre per 100 kcal.
High fibre A claim that a food is high in fibre, and any claim likely to have the same meaning for the consumer, may only be made where the product contains at
least 6 g of fibre per 100 g or at least 3 g of fibre per 100 kcal.
Source of Protein A claim that a food is a source of protein, and any claim likely to have the same meaning for the consumer, may only be made where at least 12 percent of
the energy value of the food is provided by protein.
High Protein A claim that a food is high in protein, and any claim likely to have the same meaning for the consumer, may only be made where at least 20 percent of the
energy value of the food is provided by protein.
5. Findings and discussion percent were low in sugar, 60 percent were a source of fibre, 38 percent
were low salt, 32 percent high in fibre, 30 percent a source of protein,
The fifty foods identified and number of ingredients are shown in 18 percent very low sodium and 12 percent low energy (Fig. 4a). Under
Fig. 1. The number of ingredients listed was, on average 13.6 but the updated 2018 model 38 of the original 50 UPFs remained healthy
ranged from 5 to 33. Forty percent of the UPFs contained ten in- (nutrient profile score < 4) and these tended to be healthier for the
gredients or less. Just under half (44 percent) had 5 to 10 ingredients EFSA categories than the original 50 (Fig. 4b). So, overall, 38 of the 50
listed. These findings demonstrate a large degree in variability in the were still healthy under the 2018 model and these tended to be heal-
number ingredients used between products. Quite often some of the thier for the EFSA categories than the original 50.
ingredients used were herbs and spices, preservative or fortificants. This When focusing on six of nutrient claims (low energy, low in satu-
yields the question about whether it is fair to judge products based on rated fat, low sugar, low salt and a source of fibre and energy) and using
‘ingredient numbers’ when some ingredients listed have a functional the 2011 model 4 percent of UPFs met EFSA nutrient claims for one
role to play by improving safety, shelf-life and nutritional profile. It is component, 20 percent met nutrition claims for two components and 50
also unclear why a cut-off of 5 ingredients or more has been used for the percent met EFSA benchmarks for three components. 1 in 5 (20 per-
NOVA definition of UPFs (Monteiro, 2016). cent) aligned with four EFSA nutrient claims and 6 percent for five of
Spearman Rank correlations between the number of ingredients and these. So all foods met EFSA nutrient claims for at least one component
individual nutritional components were also not statistically significant: and 76 percent of these UPFs had 3 or more healthy components and
Energy kJ/100g (rho = 0.15, p = 0.30), Saturated fat g/100g aligned with EFSA benchmarks (Fig. 5). This indicates that there are a
(rho = 0.10, p = 0.51), Total Sugar g/100g (rho = 0.15, p = 0.28), number of areas in which certain UPFs can meet nutritional bench-
Sodium mg/100g (rho = 0.20, p = 0.16), Dietary Fibre g/100g marks.
(rho = 0.06, p = 0.68) nor Protein g/100g (rho = −0.004, p = 0.98). Of the 50 UPFs identified 28 percent provided fortificants or other
Previously it has been reported that UPFs consumption has a ‘protein- functional components. One in five (20 percent) were fortified with
diluting effect’ – displacing protein from diets whilst driving up energy riboflavin, vitamin B6 or calcium, 18 percent with vitamin D, 14 per-
intakes (Martinez Steele, Raubenheimer, Simpson, Baraldi, & Monteiro, cent add vitamin B12 added, 12 percent were fortified with thiamine,
2018). The present commentary refutes this as 30 cent of the identified niacin or iron, 10 percent were fortified with folic acid, 6 percent
UPFs were a ‘source’ of protein according to European Nutrition Claims. provided vitamin E, 4 percent were fortified with iodine or pantothenic
Other work using mycoprotein (a component of Quorn™ - classified as acid and 2 percent had zinc or magnesium added. Omega-3 fatty acids
UPF by NOVA standards) has found its consumption to be associated were added in 6 percent of UPFs and plant stanols in 4 percent of UPF
with reduced energy intake and insulin release in overweight subjects products.
suggesting that it has satiety and appetite regulatory properties (Bottin It should be recognised that food fortification can be an important
et al., 2016). delivery vehicle for certain nutrients. For example, in one randomised
Nutritional profile scores of the 50 ‘healthy’ UPF are shown in controlled trial fortified breakfast cereal consumption over 12 weeks
Fig. 2. These ranged between −8 and 3. Lower values were generated was found to significantly improve the micronutrient intakes and status
when products were lower in energy, fat and sugar and had a higher of teenage girls, particularly in the case of vitamins B1, B2, B6, B12,
fibre and protein profile. Higher sugar and salt profiles tended to in- folate, iron (p < 0.001) and of vitamin D (p = 0.007) (Powers,
crease nutrient profile scores most frequently, particularly in the 2018 Stephens, Russell, & Hill, 2016). In another randomised controlled trial
model. There were no statistically significant correlations between the bread fortified with vitamin D eaten by healthy adults (n = 90) was
overall nutrient profile score and the number of ingredients listed effective at raising serum 25-hydroxyvitamin D levels to nearly ade-
(rho = 0.10, p = 0.49) (Fig. 3a) using the 2011 Department of Health quate levels (Nikooyeh et al., 2016). Given such findings it should be
model. The correlation between the nutrient profile score and the questioned whether fortificants should be excluded when classifying
number ingredients listed was increased using the 2018 cut-offs but the UPF using ingredients lists.
association remained non significant (rho = 0.24, p = 0.09) (Fig. 3b). Subsequently, the present commentary and analysis provides evi-
Regarding nutrient claims using the 2011 Department of Health dence contradicting the statement that: “All together, they (UPF) are
model 84 percent of the identified UPFs were low in saturated fat, 80 energy-dense, high in unhealthy types of fat, free sugars and salt, and poor
100
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
sources of protein, dietary fibre and micronutrients” (Monteiro et al., sodium, fibre or protein) nor nutrient profiling scores.
2018). In the present paper the number of ingredients did not correlate
against nutrition components (energy, saturated fat, total sugar,
101
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
6. Progressive trends diets and even pose a risk nutritionally. Elsewhere concerns have al-
ready been addressed, particularly in relation to fortified foods such as
From a food processing perspective nutrition is expected to remain wholegrain bread, fortified enriched bread and cereals that could im-
an important issue in coming years, especially as inter-relationships pact on dietary fibre, folate and calcium intakes (Jones, 2018) if these
between food and health continue to strengthen (Fardet et al., 2015). It were to be further restricted by being demonised as being ‘ultra-pro-
should be considered that food systems are changing rapidly - this in- cessed’.
cludes growing interest in plant-based approaches for economic, health As well as the nutritional profile of UPFs other defamatory remarks
and environmental motives (Kevany KM et al., 2018). have also been made. For example, it has been stated that: “Ultra-pro-
Interestingly, 42 percent of the identified ‘UPFs’ were either plant- cessed products are also troublesome from social, cultural, economic, poli-
based, meat or dairy alternatives. Twenty-eight percent were also for- tical and environmental points of view” (Monteiro et al., 2018). Yet many
tified or contained functional components. Given changing markets and progressions are also being made. For example, innovative package
movements away from traditional meat and dairy sources the exclusion schemes are evolving such as eco-friendly sugar palm starch-based films
of some of these foods deemed to be ‘ultra-processed’ could restrict as an effective biopackaging material (Ilyas, Sapuan, Ishak, & Zainudin,
102
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
103
E. Derbyshire Trends in Food Science & Technology 94 (2019) 98–104
• Processing when the right approaches are taken may be beneficial Guillard, V., Gaucel, S., Fornaciari, C., Angellier-Coussy, H., Buche, P., & Gontard, N.
(2018). The next generation of sustainable food packaging to preserve our environ-
from a food safety, preservation and nutritional stance. e.g. bread
ment in a circular economy context. Frontiers in Nutrition, 5, 121.
fortified with folic acid, fortified cereals, meat alternatives, pro- Gulland, A. (2017). Dietary guidelines should encourage a healthy planet. BMJ, 356,
biotic food products or gluten-free food products. j1254.
• Overall, food manufacturers should produce products that have the Ilyas, R. A., Sapuan, S. M., Ishak, M. R., & Zainudin, E. S. (2018). Development and
characterization of sugar palm nanocrystalline cellulose reinforced sugar palm starch
best profile possible from a food safety, nutrition and environmental bionanocomposites. Carbohydrate Polymers, 202, 186–202.
perspective. Jones, J. M. (2018). Food processing: Criteria for dietary guidance and public health?
Proceedings of the Nutrition Society, 1–15.
Kevany, K. M., Baur, G., & W, G. C. (2018). Shifting food systems: Increasing well-being
Conflicts of Interest Statement through plant-based approaches. Explore (NY). , pii: S1550-8307(17)30397-X.
Knorr, D., & Watzke, H. (2019). Food processing at a crossroad. Frontiers in Nutrition,
This review was supported by Marlow foods. The views expressed 6, 85.
Martinez Steele, E., Raubenheimer, D., Simpson, S. J., Baraldi, L. G., & Monteiro, C. A.
are those of the author alone and Marlow Foods had no role in writing
(2018). Ultra-processed foods, protein leverage and energy intake in the USA. Public
the review. The author declares no competing interests. Health Nutrition, 21, 114–124.
Monteiro, C. A., Cannon, G., Moubarac, J. C., Levy, R. B., Louzada, M. L. C., & Jaime, P. C.
Acknowledgements (2018). The UN Decade of Nutrition, the NOVA food classification and the trouble
with ultra-processing. Public Health Nutrition, 21, 5–17.
Monteiro, C. C. G., Levey, R., Moubarac, J.-C., Jaime, P., Martins, A. P., Canella, D., et al.
Many thanks to Jackie Cooper, statistical consultant for her input (2016). Food classification. Public Health Nova. The Star Shines Bright, 7, 28–38.
into the data analysis. Nikooyeh, B., Neyestani, T. R., Zahedirad, M., Mohammadi, M., Hosseini, S. H.,
Abdollahi, Z., et al. (2016). Vitamin D-fortified bread is as effective as supplement in
improving vitamin D status: A randomized clinical trial. The Journal of Cinical
References Endocrinology and Metabolism, 101, 2511–2519.
PHE (2016). The eatwell guide. London: PHE.
PHE (2018a). Annex C: Basis of the nutrient/food component and scoring for the UK NPM
Bottin, J. H., Swann, J. R., Cropp, E., Chambers, E. S., Ford, H. E., Ghatei, M. A., et al.
2004/5 and the draft 2018 NPM 1 page.
(2016). Mycoprotein reduces energy intake and postprandial insulin release without
PHE (2018b). Introduction to the consultation on the 2018 review of the UK nutrient profiling
altering glucagon-like peptide-1 and peptide tyrosine-tyrosine concentrations in
model.
healthy overweight and obese adults: A randomised-controlled trial. British Journal of
Pivk Kupirovic, U., Miklavec, K., Hribar, M., Kusar, A., Zmitek, K., & Pravst, I. (2019).
Nutrition, 116, 360–374.
Nutrient profiling is needed to improve the nutritional quality of the foods labelled
Crino, M., Barakat, T., Trevena, H., & Neal, B. (2017). Systematic review and comparison
with health-related claims. Nutrients, 11.
of classification frameworks describing the degree of food processing. Nutrition Food
Powers, H. J., Stephens, M., Russell, J., & Hill, M. H. (2016). Fortified breakfast cereal
Technology, 3https://doi.org/10.16966/2470-6086.138.
consumed daily for 12 wk leads to a significant improvement in micronutrient intake
DOH (2011). Nutrient profiling technical guidance.
and micronutrient status in adolescent girls: A randomised controlled trial. Nutrition
EFSA (2006). REGULATION (EC) No 1924/2006 OF THE EUROPEAN PARLIAMENT AND
Journal, 15, 69.
OF THE COUNCIL of 20 December 2006 on nutrition and health claims made on
SACN (2015). Carbohydrates and health. London: TSO.
foods. Official Journal of the European Union. L 404/9 https://eur-lex.europa.eu/
Sanyang, M. (2017). Sugar palm starch-based composites for packaging applications.
LexUriServ/LexUriServ.do?uri=CONSLEG:2006R1924:20080304:EN:PDF.
Bionanocomposites for Packaging Applications, 125–147.
Fardet, A., Rock, E., Bassama, J., Bohuon, P., Prabhasankar, P., Monteiro, C., et al. (2015).
Tharrey, M., Maillot, M., Azais-Braesco, V., & Darmon, N. (2017). From the SAIN,LIM
Current food classifications in epidemiological studies do not enable solid nutritional
system to the SENS algorithm: A review of a French approach of nutrient profiling.
recommendations for preventing diet-related chronic diseases: The impact of food
Proceedings of the Nutrition Society, 76, 237–246.
processing. Advances in Nutrition, 6, 629–638.
Willett, W., Rockstrom, J., Loken, B., Springmann, M., Lang, T., Vermeulen, S., et al.
Gibney, M. J., Forde, C. G., Mullally, D., & Gibney, E. R. (2017). Ultra-processed foods in
(2019). Food in the anthropocene: The EAT-lancet commission on healthy diets from
human health: A critical appraisal. American Journal of Clinical Nutrition, 106,
sustainable food systems. Lancet, 393, 447–492.
717–724.
104