Ultra-Processed Foods and Binge Eating
Ultra-Processed Foods and Binge Eating
Ultra-Processed Foods and Binge Eating
1
NOVA-4 foods and binge eating
ST1 Neurosurgery
St George’s Hospital,
London
2
NOVA-4 foods and binge eating
Abstract
OBJECTIVE
There is increasing evidence of the impact of ultra-processed foods on multiple
metabolic and neurobiological pathways, including those involved in eating behaviours,
both in animals and in humans. In this pilot study, we aimed to explore ultra-processed
foods and their link with disordered eating in a clinical sample.
METHODS
This was a single site, retrospective observational study in a specialist eating disorder
service using self report on the electronic health records. Patients with a DSM-5
diagnosis of anorexia nervosa (AN), bulimia nervosa (BN) or binge eating disorder (BED)
were randomly selected from the service database in Oxford from 2017 to 2019. The
recently introduced NOVA classification was used to determine the degree of industrial
food processing in each patient’s diet. Frequencies of ultra-processed foods were
analysed for each diagnosis, at each mealtime and during episodes of bingeing.
RESULTS
71 female and 3 male patients were included in the study. 22 had AN, 25 BN and 26 had
BED. Patients with AN reported consuming 55% NOVA-4 foods, as opposed to
approximately 70% in BN and BED. Binge foods were 100% ultra-processed.
DISCUSSION
Further research into the metabolic and neurobiological effects of reducing ultra-
processed food intake on bingeing behaviour is needed.
Key words: Anorexia nervosa, bulimia nervosa, binge eating disorder, ultra-processed
food, metabolic and neurobiological effect.
3
NOVA-4 foods and binge eating
INTRODUCTION
There has been little research into the dietary intake of people with eating disorders, despite
the clear relevance to the condition (Ayton & Ibrahim, 2019). The mainstream psychological
therapies encourage the re-introduction of regular eating (NICE, 2017), and emphasize the
importance of widening food choices according to individual preferences, including cultural
beliefs; but they pay no attention to personal metabolic and neurobiological responses, and
usually include ultra-processed foods (Hart, Marnane, McMaster, & Thomas, 2018).
In general, ultra-processed foods are high in sugars and fats and low in natural protein
(Mendonca et al., 2016; Monteiro et al., 2018). The recently proposed NOVA classification
system provides a framework for categorising the degree of modern industrial food
processing(Monteiro et al., 2019). It comprises four groups:
- Group 1: unprocessed or minimally processed foods (e.g.: fruit, vegetables, meat,
fish);
- Group 2: processed culinary ingredients used in home cooking, such as flour, olive oil,
butter, cream, dried fruit, herbs;
- Group 3: processed foods, such as tinned food, cheeses, freshly made breads;
- Group 4: ultra-processed foods (e.g. soft drinks, sweet or savoury packaged snacks,
reconstituted meat products, pre-prepared frozen dishes, diet products).
NOVA-4 definition includes: ‘Ultra-processed foods, …, are not modified foods but formulations
made mostly or entirely from substances derived from foods and additives and include other
sources of energy and nutrients not normally used in culinary preparations. A multitude of
sequences of processes is used to combine the usually many ingredients and to create the final
product (hence ‘ultra-processed’). (Monteiro et al., 2019). Many of these ingredients, including
high level of sugars, are metabolically active (Lustig, 2017), and may have an addictive
potential (Schulte, Potenza, & Gearhardt, 2017).
Studies based on the NOVA classification show exponential growth in the consumption of
ultra-processed products and confirm that they have gradually been displacing unprocessed
or minimally processed foods and freshly prepared meals (Freire et al., 2018). The recent U.K.
National Diet and Nutrition Survey (NDNS) found that 56.8% of foods (ranging from 30 to
almost 80%) consumed in the UK are ultra-processed, and in parallel, the dietary intake of
carbohydrates, free sugars, total fats increased significantly while the consumption of protein
decreased (Rauber et al., 2018). Ecological and epidemiological studies have shown a link
between the consumption of NOVA-4 foods and increased rates of obesity and metabolic
disorders (Juul, Martinez-Steele, Parekh, Monteiro, & Chang, 2018; Nardocci et al., 2019;
Wang et al., 2017), indicating that they may reduce satiety and stimulate overeating (Fardet,
Mejean, Laboure, Andreeva, & Feron, 2017; Juul et al., 2018; Monteiro et al., 2019; Nardocci
et al., 2019). A recent, carefully controlled randomized trial comparing ad libitum ultra-
processed with unprocessed diets found that people on ultra-processed foods consumed
500 kcal/day more calories, with increased consumption of carbohydrate and fat but not
protein (K. D. Hall et al., 2019), associated with widespread endocrine changes. This finding
was also consistent with recent evidence, from both animal (Bortolin et al., 2018) and human
experiments (Small & DiFeliceantonio, 2019), which showed that ultra-processed foods
interact with various hormonal and neurobiological systems that affect food intake.
Recent metabolic research has also highlighted that responses to common foods vary from
person to person, this being influenced by a number of factors, including insulin resistance,
4
NOVA-4 foods and binge eating
sleep, stress, exercise and the microbiome (Zeevi et al., 2015). These differences are
associated with impaired glucose tolerance and insulin sensitivity (H. Hall et al., 2018). The
implications of this novel research have been largely unexplored in the eating disorder field,
even though several studies have highlighted the metabolic underpinning of eating
disorders, including insulin dysregulation (Duncan et al., 2017; Ilyas et al., 2018). In addition,
several ingredients in ultra-processed foods impair the gut microbiome and intestinal
permeability (Lerner & Matthias, 2015). This is relevant to patients with restricting eating
disorders, who often use low fat products in an attempt to control their calorie intake, and
experience abdominal complaints, which may drive further restriction (Halmos, Mack, &
Gibson, 2019).
To our knowledge there have been no studies examining the links between NOVA-4 foods
and eating disorders. Clinical experience suggests that patients do consume a high
proportion of ultra-processed foods: these can be low calorie products (Schebendach et al.,
2017), or as described in Fairburn’s book ‘Overcoming binge eating’, binge foods, such as ice
cream, doughnuts, cookies, etc. (Fairburn, 2013).
Objective: following our review of the relevant literature (Ayton & Ibrahim, 2019), this pilot
study aimed to explore the degree of industrial processing in the food choices of patients
with eating disorders. We hypothesised that ultra-processed food consumption may be
associated with abnormal eating behaviours, such as bingeing.
METHOD:
This was a cross-sectional retrospective observational study using routinely collected data
from the electronic health records in Cotswold House, Oxford, a specialist NHS eating
disorder service for adults (over 18 years old). Patients with an ICD-10 diagnosis of anorexia
nervosa, bulimia nervosa and binge eating disorders were randomly selected from the
service database between 2017 and 2019. The service uses both ICD-10 and DSM-5
diagnostic criteria. A senior author (AA) checked the accuracy of the clinical diagnoses based
on the assessment documentation, and for the purposes of this paper, we used DSM-5
diagnostic criteria. As part of the assessment, clinicians documented dietary intake by asking
the patient to describe ‘a typical food intake per day over the past two weeks’, including
specific mealtimes, bingeing and purging. To reduce recall bias we did not attempt to
estimate total daily energy intake. Only the types and frequency of food choices, but not the
amounts were included in the analysis. Subjective and objective binges were analysed
together. The NOVA classification (Monteiro et al., 2018) was used to categorize the degree
of industrial food processing in each patient’s diet by two of the authors (JD & AA)
independently. Foods were categorised into Nova 4 groups based on the patients’ reports at
each mealtimes. Differences in coding between the authors were resolved by joint discussion
referring to the NOVA definitions (Monteiro et al., 2019). Patients with inadequate dietary
information on their assessment were excluded from the study. We also reviewed the health
records for evidence of glucose intolerance, as this is an important factor of individual
metabolic response to foods, and is associated with binge eating disorder (Ilyas et al., 2018).
The study was approved by Oxford Health Foundation Trust Audit Department. SPSS-22 was
used for descriptive statistics and Pearson’s s chi-square test was used to compare
categorical variables.
5
NOVA-4 foods and binge eating
RESULTS
Seventy-three patients were included from the original pool of 79 patient records. Three
were male and 71 were female. Six were excluded owing to insufficient recording of dietary
intake details during assessment by the clinician involved. Twenty-two patients had a DSM-5
diagnosis of anorexia nervosa (AN) (21 restrictive and 1 binge-purging subtype), 25 had
bulimia nervosa (BN) and 26 binge eating disorder (BED). Length of illness was on average 7-
8 years for people with anorexia and bulimia nervosa, and more than 20 years for binge
eating disorders (Table 1). Seven patients with BED and one patient with BN had impaired
fasting glucose (6.1-6.9 mmol/dl), and one patient with BED had type 2 diabetes (in total,
30.7% with BED had impaired glucose tolerance or type 2 diabetes).
Patients binged 100% on NOVA-4 foods. Examples included: chocolate, ice-cream, crisps,
sandwiches, biscuits, cakes, pizza, smoothies and donuts.
DISCUSSION
To our knowledge this is the first observational study exploring the frequency of NOVA-4
ultra-processed food choices of people with eating disorders. This was a retrospective
casenote study using routine data. Both patients and clinicians were blind to the research
hypothesis. Our main finding is patients (regardless of the type of eating disorder) binged on
100% of NOVA-4 foods. Binge foods were high in carbohydrate and fats. Meals were often
missed during the day by all diagnostic groups, indicating that dietary restriction is shared
between people with eating disorders. This is consistent with Fairburn’s transdiagnostic
model of eating disorders (Fairburn et al., 2015).
The link between obesity and NOVA-4 consumption has been shown in a number of studies,
but to our knowledge, no study has examined the association with binge eating, or eating
disorders (Juul et al., 2018; Mendonca et al., 2016).
6
NOVA-4 foods and binge eating
A recent carefully controlled randomized crossover trial on a metabolic ward (K. D. Hall et al.,
2019) found that NOVA-4 foods resulted in overconsumption. Participants consumed an
excess of 500kcal/day when they were placed on an ad libitum, ultra-processed diet, as
compared with a minimally processed one by using a wide range of foods with similar
appearance and similar macronutrient content. Participants did not report differences in
palatability, suggesting that the increased food intake was driven by metabolic mechanisms.
The hunger hormone ghrelin, fasting glucose and insulin levels were increased during the
ultra-processed diet, whilst appetite-suppressing hormone peptide tyrosine tyrosine (PYY)
reduced. The research team carefully tried to control macronutrient intake, but found that
participants on ultra-processed diet increased both carbohydrate and fat intake, whilst
protein intake remained the same. This is consistent with the protein leverage hypothesis,
which suggests excess intake is driven by the dilution of dietary protein (Gosby, Conigrave,
Raubenheimer, & Simpson, 2014; Martinez Steele, Raubenheimer, Simpson, Baraldi, &
Monteiro, 2018).
A recent meta-analysis found that anorexia nervosa is associated with increased insulin
sensitivity whilst bulimia nervosa and binge-eating disorders are associated with insulin
resistance (Ilyas et al., 2018). Our findings are consistent with this, as 30% of patients with
binge eating disorder had impaired glucose tolerance. This suggests that metabolic factors
may contribute to binge eating. Furthermore, there is increasing recognition that there are
significant individual differences in responses to common foods (Valdes, Walter, Segal, &
Spector, 2018; Zeevi et al., 2015). This is partially driven by the microbiome, which is
influenced by the diet (Rothschild et al., 2018). These findings may explain why some people
develop certain metabolic or eating disorders in the modern food environment, while others
do not.
Animal (DiFeliceantonio & Small, 2019) and human experiments (Small & DiFeliceantonio,
2019) have demonstrated that ultra-processed foods also alter the neurobiological reward
pathways involved in eating behaviours. Neuroimaging studies have shown that food cues,
which are predictive of calories, activate the striatum in humans. The magnitude of these
responses is regulated by metabolic signals, and this process influences food intake
according to energy requirement (DiFeliceantonio et al., 2018). However, non nutritive
sweeteners, which are not found in nature, and combinations of sugars and fats (common in
NOVA-4 products), disrupt the brain’s ability accurately to estimate the energy value of
foods and result in supra-additive effects, which may drive overconsumption (Small &
DiFeliceantonio, 2019). Although these studies were carried out using only a small range of
foods, these discoveries have important implications for our understanding of eating
disorders. Small’s group has developed a novel two stage model bringing together the
biological and psycho-social factors influencing eating behaviours (Small & DiFeliceantonio,
2019). One system directly reflects the nutritional value of foods and relies on metabolic
signals reaching the brain. This nutrient-sensing system plays a critical role in regulating
striatal dopamine and reward. The second, conscious system influences food choices based
on beliefs on healthfulness, cost, etc, which are heavily targeted by advertising and by the
food environment. Patients with eating disorders choose diet products (which are commonly
NOVA-4), believing them to be healthier options, unaware of the metabolic and
neurobiological effects that impair accurate sensing of nutrient content by the brain and
result in uncontrollable eating during a binge episode.
7
NOVA-4 foods and binge eating
As this was a pilot observational study, there are limitations. We used routinely collected
data, rather than the standard food frequency questionnaires. However, food frequency
questionnaires have been heavily criticized, due to recall and reporting bias (Archer, Marlow,
& Lavie, 2018). For example, the UK NDNS survey (Rauber et al., 2018) reported daily intake
of only 1764.7 Kcal, which was inconsistent with the obesity epidemic, and therefore
indicates significant underreporting, To improve the reliability and accuracy of self reports, in
our study, we only attempted to code the type of foods and not the amounts consumed, and
we did not attempt to estimate the overall calorie intake.
8
NOVA-4 foods and binge eating
References
Archer, E., Marlow, M. L., & Lavie, C. J. (2018). Controversy and debate: Memory-Based Methods
Paper 1: the fatal flaws of food frequency questionnaires and other memory-based dietary
assessment methods. J Clin Epidemiol, 104, 113-124. doi:10.1016/j.jclinepi.2018.08.003
Ayton, A., & Ibrahim, A. (2019). The Western diet: a blind spot of eating disorder research?-a
narrative review and recommendations for treatment and research. Nutr Rev.
doi:10.1093/nutrit/nuz089
Bortolin, R. C., Vargas, A. R., Gasparotto, J., Chaves, P. R., Schnorr, C. E., Martinello, K. B., . . .
Moreira, J. C. F. (2018). A new animal diet based on human Western diet is a robust diet-
induced obesity model: comparison to high-fat and cafeteria diets in term of metabolic and
gut microbiota disruption. Int J Obes (Lond), 42(3), 525-534. doi:10.1038/ijo.2017.225
DiFeliceantonio, A. G., Coppin, G., Rigoux, L., Edwin Thanarajah, S., Dagher, A., Tittgemeyer, M., &
Small, D. M. (2018). Supra-Additive Effects of Combining Fat and Carbohydrate on Food
Reward. Cell Metab, 28(1), 33-44 e33. doi:10.1016/j.cmet.2018.05.018
DiFeliceantonio, A. G., & Small, D. M. (2019). Dopamine and diet-induced obesity. Nat Neurosci,
22(1), 1-2. doi:10.1038/s41593-018-0304-0
Duncan, L., Yilmaz, Z., Gaspar, H., Walters, R., Goldstein, J., Anttila, V., . . . Bulik, C. M. (2017).
Significant Locus and Metabolic Genetic Correlations Revealed in Genome-Wide Association
Study of Anorexia Nervosa. Am J Psychiatry, 174(9), 850-858.
doi:10.1176/appi.ajp.2017.16121402
Eldridge, A. L., Piernas, C., Illner, A. K., Gibney, M. J., Gurinovic, M. A., de Vries, J. H. M., & Cade, J. E.
(2018). Evaluation of New Technology-Based Tools for Dietary Intake Assessment-An ILSI
Europe Dietary Intake and Exposure Task Force Evaluation. Nutrients, 11(1).
doi:10.3390/nu11010055
Fairburn, C. G. (2013). Overcoming binge eating (Second edition ed.). New York, London: The
Guilford Press.
Fairburn, C. G., Bailey-Straebler, S., Basden, S., Doll, H. A., Jones, R., Murphy, R., . . . Cooper, Z.
(2015). A transdiagnostic comparison of enhanced cognitive behaviour therapy (CBT-E) and
interpersonal psychotherapy in the treatment of eating disorders. Behav Res Ther, 70, 64-71.
doi:10.1016/j.brat.2015.04.010
Fardet, A., Mejean, C., Laboure, H., Andreeva, V. A., & Feron, G. (2017). The degree of processing of
foods which are most widely consumed by the French elderly population is associated with
satiety and glycemic potentials and nutrient profiles. Food Funct, 8(2), 651-658.
doi:10.1039/c6fo01495j
Freire, W. B., Waters, W. F., Roman, D., Jimenez, E., Burgos, E., & Belmont, P. (2018). Overweight,
obesity, and food consumption in Galapagos, Ecuador: a window on the world. Global
Health, 14(1), 93. doi:10.1186/s12992-018-0409-y
Gosby, A. K., Conigrave, A. D., Raubenheimer, D., & Simpson, S. J. (2014). Protein leverage and
energy intake. Obes Rev, 15(3), 183-191. doi:10.1111/obr.12131
Hall, H., Perelman, D., Breschi, A., Limcaoco, P., Kellogg, R., McLaughlin, T., & Snyder, M. (2018).
Glucotypes reveal new patterns of glucose dysregulation. PLoS Biol, 16(7), e2005143.
doi:10.1371/journal.pbio.2005143
Hall, K. D., Ayuketah, A., Brychta, R., Cai, H., Cassimatis, T., Chen, K. Y., . . . Zhou, M. (2019). Ultra-
Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized
9
NOVA-4 foods and binge eating
Controlled Trial of Ad Libitum Food Intake. Cell Metab, 30(1), 67-77 e63.
doi:10.1016/j.cmet.2019.05.008
Halmos, E. P., Mack, A., & Gibson, P. R. (2019). Review article: emulsifiers in the food supply and
implications for gastrointestinal disease. Aliment Pharmacol Ther, 49(1), 41-50.
doi:10.1111/apt.15045
Hart, S., Marnane, C., McMaster, C., & Thomas, A. (2018). Development of the "Recovery from
Eating Disorders for Life" Food Guide (REAL Food Guide) - a food pyramid for adults with an
eating disorder. J Eat Disord, 6, 6. doi:10.1186/s40337-018-0192-4
Ilyas, A., Hubel, C., Stahl, D., Stadler, M., Ismail, K., Breen, G., . . . Kan, C. (2018). The metabolic
underpinning of eating disorders: A systematic review and meta-analysis of insulin
sensitivity. Mol Cell Endocrinol. doi:10.1016/j.mce.2018.10.005
Juul, F., Martinez-Steele, E., Parekh, N., Monteiro, C. A., & Chang, V. W. (2018). Ultra-processed food
consumption and excess weight among US adults. Br J Nutr, 120(1), 90-100.
doi:10.1017/S0007114518001046
Lerner, A., & Matthias, T. (2015). Changes in intestinal tight junction permeability associated with
industrial food additives explain the rising incidence of autoimmune disease. Autoimmun
Rev, 14(6), 479-489. doi:10.1016/j.autrev.2015.01.009
Lustig, R. H. (2017). Processed Food-An Experiment That Failed. JAMA Pediatr, 171(3), 212-214.
doi:10.1001/jamapediatrics.2016.4136
Martinez Steele, E., Raubenheimer, D., Simpson, S. J., Baraldi, L. G., & Monteiro, C. A. (2018). Ultra-
processed foods, protein leverage and energy intake in the USA. Public Health Nutr, 21(1),
114-124. doi:10.1017/S1368980017001574
Mendonca, R. D., Pimenta, A. M., Gea, A., de la Fuente-Arrillaga, C., Martinez-Gonzalez, M. A., Lopes,
A. C., & Bes-Rastrollo, M. (2016). Ultraprocessed food consumption and risk of overweight
and obesity: the University of Navarra Follow-Up (SUN) cohort study. Am J Clin Nutr, 104(5),
1433-1440. doi:10.3945/ajcn.116.135004
Monteiro, C. A., Cannon, G., Levy, R. B., Moubarac, J. C., Louzada, M. L., Rauber, F., . . . Jaime, P. C.
(2019). Ultra-processed foods: what they are and how to identify them. Public Health Nutr,
1-6. doi:10.1017/S1368980018003762
Monteiro, C. A., Cannon, G., Moubarac, J. C., Levy, R. B., Louzada, M. L. C., & Jaime, P. C. (2018). The
UN Decade of Nutrition, the NOVA food classification and the trouble with ultra-processing.
Public Health Nutr, 21(1), 5-17. doi:10.1017/S1368980017000234
Nardocci, M., Leclerc, B. S., Louzada, M. L., Monteiro, C. A., Batal, M., & Moubarac, J. C. (2019).
Consumption of ultra-processed foods and obesity in Canada. Can J Public Health, 110(1), 4-
14. doi:10.17269/s41997-018-0130-x
NICE. (2017). NICE Guidance 69. Eating Disorders: Recognition and treatment.
Rauber, F., da Costa Louzada, M. L., Steele, E. M., Millett, C., Monteiro, C. A., & Levy, R. B. (2018).
Ultra-Processed Food Consumption and Chronic Non-Communicable Diseases-Related
Dietary Nutrient Profile in the UK (2008(-)2014). Nutrients, 10(5). doi:10.3390/nu10050587
Rothschild, D., Weissbrod, O., Barkan, E., Kurilshikov, A., Korem, T., Zeevi, D., . . . Segal, E. (2018).
Environment dominates over host genetics in shaping human gut microbiota. Nature,
555(7695), 210-215. doi:10.1038/nature25973
Schebendach, J., Klein, D. A., Mayer, L. E. S., Attia, E., Devlin, M. J., Foltin, R. W., & Walsh, B. T.
(2017). Assessment of the motivation to use artificial sweetener among individuals with an
eating disorder. Appetite, 109, 131-136. doi:10.1016/j.appet.2016.11.026
10
NOVA-4 foods and binge eating
Schulte, E. M., Potenza, M. N., & Gearhardt, A. N. (2017). A commentary on the "eating addiction"
versus "food addiction" perspectives on addictive-like food consumption. Appetite, 115, 9-
15. doi:10.1016/j.appet.2016.10.033
Small, D. M., & DiFeliceantonio, A. G. (2019). Processed foods and food reward. Science, 363(6425),
346-347. doi:10.1126/science.aav0556
Toribio-Mateas, M. A., & Spector, T. D. (2017). Could food act as personalized medicine for chronic
disease? Per Med, 14(3), 193-196. doi:10.2217/pme-2016-0017
Valdes, A. M., Walter, J., Segal, E., & Spector, T. D. (2018). Role of the gut microbiota in nutrition and
health. BMJ, 361, k2179. doi:10.1136/bmj.k2179
Wang, D., Hawley, N. L., Thompson, A. A., Lameko, V., Reupena, M. S., McGarvey, S. T., & Baylin, A.
(2017). Dietary Patterns Are Associated with Metabolic Outcomes among Adult Samoans in a
Cross-Sectional Study. J Nutr, 147(4), 628-635. doi:10.3945/jn.116.243733
Wark, P. A., Hardie, L. J., Frost, G. S., Alwan, N. A., Carter, M., Elliott, P., . . . Cade, J. E. (2018). Validity
of an online 24-h recall tool (myfood24) for dietary assessment in population studies:
comparison with biomarkers and standard interviews. BMC Med, 16(1), 136.
doi:10.1186/s12916-018-1113-8
Zeevi, D., Korem, T., Zmora, N., Israeli, D., Rothschild, D., Weinberger, A., . . . Segal, E. (2015).
Personalized Nutrition by Prediction of Glycemic Responses. Cell, 163(5), 1079-1094.
doi:10.1016/j.cell.2015.11.001
11
NOVA-4 foods and binge eating
12
NOVA-4 foods and binge eating
13