Understanding eating topography: The key to reducing energy intake in humans?

Lead Research Organisation: University of Bristol
Department Name: Experimental Psychology

Abstract

According to the World Heath Organisation, obesity has more than doubled since 1980. Estimates suggest that obesity now affects around 1.5 billion people. This is worrying because the health and economic consequences are very clear. Obesity is a major risk factor for cardiovascular diseases, diabetes, and some cancers.
One of the key observations about obesity is that not everyone becomes obese, even when they live in the same community, or even the same family. This means that some people appear to be 'protected.' Obesity researchers are interested in understanding why this is the case because this protection may hold the key to an effective treatment, or even a way to prevent obesity in the first place.
For a long time, researchers and health professionals have suspected that obesity is associated with a particular eating style, eating quickly in particular. Indeed, it is sometimes said that we should chew our food several times in order to feel satisfied and to 'aid digestion.' Recently, researchers have begun to explore this idea systematically. The results are striking. For example, under controlled conditions, it would seem that eating at a slower rate produces both an increase in self-reported fullness and a reduction in meal size. Moreover, when we look at people across an entire country, we find that eating rate is a good predictor of bodyweight, even in large-scale studies. In 2010 researchers started to look at ways to reduce eating rate to see if this might be used to lower bodyweight. Their results were impressive. They used device called a mandometer to encourage children to eat at a slower rate. This training produced a clinically significant reduction in bodyweight, which was sustained 12 months post treatment.
The prospect that we can manipulate eating behaviour to reduce energy intake is tantalizing because this approach has potential as an effective treatment for obesity. Moreover, an opportunity exists to change our eating behaviour by manipulating the physical characteristics of food. If this can be achieved then we may be able to design foods to encourage behaviours (e.g., slow eating) that reduce our calorie intake from meal to meal. Importantly, for these benefits to be realised, we need to discover the underlying mechanism. This is an important objective of this project.
In the first instance we will develop a method to quantify and characterise 'eating topography' - collectively, the pattern of behaviours associated with eating; swallow rate, bite size, eating rate, and so on. With this tool, we will run a series of experiments to identify specific aspects of eating topography that influence our food intake.
In a second set of experiments we will focus on the mechanism. Two hypotheses will be tested. Firstly, we will explore the prospect that a causal relationship exists between specific aspects of eating topography and the hunger and fullness that we experience at the end of a meal and during the period between meals. There are two reasons why this relationship might exist. Eating topography may change levels of hormones that control our appetite. Alternatively, it may influence the formation of memory for a meal - a process that is known to influence the amount of food that we eat at a subsequent meal.
Our second hypothesis relates to the eating topography that is associated with particular foods. If a food is eaten with a topography that makes us feel full then we may remember this relationship. In future, when we encounter that food again, we may expect the food to be more filling and select a smaller portion. By this account, eating topography influences our energy intake by changing the way we make decisions about portion size, before a meal begins.

Technical Summary

There is mounting evidence that eating rate has a marked effect on energy intake in humans. For example, under controlled conditions, eating at a slower rate is found to promote self-reported fullness and to reduce energy intake. Differences in eating rate are meaningful because they predict variability in bodyweight in large-scale population studies. Moreover, significant and sustained (12 months post-treatment) reductions in BMI and body fat have been reported when obese children are trained to moderate their rate of eating using a 'mandometer.'
Given the role that eating rate plays in energy intake it is surprising that little is known about the underlying process. In Phase I we will use EMG and an eating pattern monitor to quantify and characterise 'eating topography.' This will be achieved by taking simultaneous measures of bite size, eating rate, swallow rate, inter-bite interval, and so on. Using this measure, we will identify specific aspects of eating topography that influence energy intake from meal to meal.
In Phase II we will identify the process by which eating topography comes to influence energy intake. Specifically, we will explore two hypotheses. First, we will determine whether a causal relationship exists between eating topography and satiety in the inter-meal interval. In particular, we will measure direct effects of eating topography on appetite-regulating hormones (ghrelin, CCK, PYY, GLP-1, and insulin) and indirect effects on attention and 'memory for recent eating.' Second, we will explore whether the effects of eating topography are learned and expressed in the 'expected satiety' of foods and in decisions about portion size, before a meal begins.

Planned Impact

We see three areas where our project could have considerable social and economic impact:

1. FOODS FOR WEIGHT MANAGEMENT
Several sources indicate that eating topography (eating rate in particular) influences energy intake from meal to meal. Indeed, this relationship appears to have an accumulative effect on energy balance because it is a good predictor of BMI in large population studies. For the first time, we will expose the underlying process. In addition, we will identify the key aspects of eating topography (chew rate, inter-bite interval, etc) that influence energy intake.
In collaboration with our industry partner, we will use this understanding to explore ways in which foods might be modified to encourage these specific patterns of eating. If this is successful, then this approach might be applied to enhance the efficacy of commercial products that are designed to confer benefits for weight loss. In relation to this form of impact, Study 8 is important, because it illustrates how the fundamental research in this project might be applied. We also note that this project is submitted under the LINK scheme. During triannual meetings we will reserve half a day to discuss ways in which our results might be exploited and applied to promote weight loss. More generally, both the PI and Co-I have regular contact with researchers in several major food companies. We will use this network to ensure that our findings can be exploited at the earliest opportunity.

2. TREATMENT OF OBESITY
In part, our interest in eating topography stems from the recent discovery that by manipulating eating rate it is possible to achieve a significant and sustained (12 months post treatment) reduction in BMI and body fat in otherwise obese children. As a direct result of this project, we will have a better understanding of how this intervention works. Moreover, with this understanding we will be well-placed to identify ways in which it might be improved, simplified, and implemented more widely and cost effectively.
To achieve this impact we will collaborate with Professor Julian Hamilton-Shield, who was directly involved in the clinical trial establishing the efficacy of the mandometer. We have already agreed to explore ways in which our research findings might be implemented in a clinical setting. As part of this process we will also involve Dr Jeremy Burn, who has considerable expertise in the field of electronic instrumentation. Again, Dr Burn already collaborates with our group. He has agreed to work alongside this project to explore ways in which modifications to eating behaviour might be achieved in a home environment.
Finally, we will explore opportunities to collaborate with Dr Andrew Johnson who treats obese patients at the Avon Obesity Service (Southmead NHS Hospital, Bristol). This collaboration will provide an ideal opportunity to evaluate the clinical significance of our research findings in an adult obese population.

3. PROMOTING BEST PRACTICE IN DIETARY BEHAVIOUR
Our research has considerable potential to inform policy makers and heath professionals. By the end of this project we will be well-positioned to comment on ways in which distraction and specific eating behaviours contribute to overeating. To assist with evidence-based practice, results from our research will be submitted to target journals that are likely to reach a broader audience of health professionals and policy makers (e.g., the American Journal of Clinical Nutrition).
Finally, to realise the broader impact of our results we recognise the benefits of demonstrating that eating behaviour predicts future weight gain. Researchers in the Nutrition and Behaviour Unit are currently conducting prospective studies of this kind using the 'Freshman 15' model. In this context, we intend to include measures of eating behaviour as predictors of weight gain.

Publications

10 25 50

 
Description This BBSRC-DRINC funded project demonstrated an important role for expectations about satiety in guiding portion-size decisions. This conceptual approach and the psychophysical techniques developed to quantify expected satiety have been employed by industry in large scale product testing and in product development studies.

n the first phase of this project, we developed a method to quantify and characterise "eating topography" - collectively, the pattern of behaviours associated with eating; eating rate, oral sensory exposure time, chews per mouthful and so on. This was achieved using a novel combination of video recordings of mouth movements and real-time measures of plate weight via a hidden balance scale set into a table (see photos below).

?

In our first two experiments, we explored whether it is possible to increase post-meal fullness and reduce meal size, simply by capitalizing on natural variation in the oral processing of commercially available pre-packaged meals (Ferriday et al., 2016). Across 20 commercially available pre-packaged meals, we found that foods that were eaten slowly delivered more satiation and satiety, and had a higher expected fullness. We also found that natural variations in the oral processing of pre-packaged meals reduced meal size. These differences in food intake during the ad libitum meal were not compensated at a subsequent snacking opportunity one hour later. Based on all of these findings, we concluded that altering food form to encourage increased oral processing might represent a viable target for food manufacturers to help nudge consumers to manage their weight.

We have also completed other experiments exploring the prospect that beliefs about the satiating quality of a food and visual cues of portion size can have a direct effect on how we eat a food. In a recent study we showed that eating rate is modified by visual feedback associated with the perceived rate of change in food volume during a meal. We are the first group to expose this implicit process. This work suggests that eating rate is not determined solely by the structural properties of food being eaten (Wilkinson, Ferriday, et al., 2016).

In the second phase of our project we completed a set of experiments where we focussed on the mechanism underlying effects of eating topography on energy intake. Two hypotheses were tested. First, we explored the prospect that a causal relationship exists between specific aspects of eating topography and the hunger and fullness that we experience at the end of a meal and during the period between meals. There are two reasons why this relationship might exist. Eating topography may change levels of hormones that control our appetite. Alternatively, it may influence the formation of memory for a meal - a process that is known to influence the amount of food that we eat at a subsequent meal (Ferriday et al., 2015). Our second hypothesis relates to the eating topography that is associated with particular foods. If a food is eaten with a topography that makes us feel full then we may remember this relationship. In future, when we encounter that food again, we may expect the food to be more filling and select a smaller portion. By this account, eating topography influences our energy intake by changing the way we make decisions about portion size, before a meal begins.
Exploitation Route This research might be used to inform the design of foods that confer benefits for weight loss.
Sectors Agriculture, Food and Drink,Communities and Social Services/Policy,Healthcare,Other

URL http://www.bristol.ac.uk/psychology/research/brain/nbu/currentprojects/bbsrclink.html
 
Description This research was a BBSRC-LINK funded project. Outputs and knowledge gained from this work has informed the design and testing of food products in a commercial context by our LINK partner. Academic and industry researchers have since used our approach to show how the texture of a food can be manipulated to effect changes in the meal size (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0093370)
First Year Of Impact 2015
Sector Agriculture, Food and Drink
Impact Types Economic

 
Description FP7 EU consortium (Nudge-it)
Amount £567,528 (GBP)
Funding ID Grant Agreement 607310 (Nudge-it) 
Organisation European Commission 
Department Seventh Framework Programme (FP7)
Sector Public
Country European Union (EU)
Start 01/2014 
End 12/2018
 
Title Eating topography 
Description An initial objective of this project was to develop a technique for measuring a range of parameters associated with eating (e.g., bite rate, bite size, number of chews, oral sensory exposure time) - a measure of eating topography. In the first phase of this project we generated a novel technique using a combination of real-time measures of plate weight and video-recordings of the mouth. By combining these approaches, it is possible to obtain a comprehensive assessment of variables that are thought to influence food intake and appetite (e.g., bite size, eating rate, oral-sensory exposure time and chews per mouthful). To extract data from these recordings, we have also developed; i) a standardised video coding scheme that can be used to quantify a range of oral processing behaviours, and ii) a MATLAB script that filters dynamic plate-weight data to plot cumulative food intake curves. This research tool has already been implemented in a number of studies in our group. We have already shared this methodological tool with our industry partners (Nestle Research Center) and we have a paper in preparation that provides a full description of our methodology and associated MATLAB code. 
Type Of Material Improvements to research infrastructure 
Year Produced 2013 
Provided To Others? Yes  
Impact To date, researchers using these types of methodologies have either had to possess considerable technical knowledge and/or they have paid companies/ research organisations for the equipment to collect this data. Our software is free and easy to implement. Since we are publishing our MATLAB script in full and in open access, our methodology will be available for use by any research group at no cost. 
 
Title Keeping pace with your eating: Visual feedback affects eating rate in humans 
Description Raw data for the paper entitled "Keeping pace with your eating: Visual feedback affects eating rate in humans". Abstract: Deliberately eating at a slower pace promotes satiation and produces a sustained reduction in body weight. Therefore, understanding factors that affect eating rate should be given high priority. Eating rate is affected by the physical/textural properties of a food, by motivational state, and by portion size and palatability. This study explored the prospect that eating rate is also influenced by a hitherto unexplored cognitive process that uses ongoing perceptual estimates of the volume of food remaining in a container to adjust intake during a meal. A 2 (amount seen; 300ml or 500ml) x 2 (amount eaten; 300ml or 500ml) between-subjects design was employed (10 participants in each condition). In two 'congruent' conditions, the same amount was seen at the outset and then subsequently consumed (300ml or 500ml). To dissociate visual feedback of portion size and actual amount consumed, food was covertly added or removed from a bowl using a peristaltic pump. This created two additional 'incongruent' conditions, in which 300ml was seen but 500ml was eaten or vice versa. We repeated these conditions using a savoury soup and a sweet dessert. Eating rate (ml per second) was assessed during lunch. After lunch we assessed fullness over a 60-minute period. In the congruent conditions, eating rate was unaffected by the actual volume of food that was consumed (300ml or 500ml). By contrast, we observed a marked difference across the incongruent conditions. Specifically, participants who saw 300ml but actually consumed 500ml ate at a faster rate than participants who saw 500ml but actually consumed 300ml. Participants were unaware that their portion size had been manipulated. Nevertheless, when it disappeared faster or slower than anticipated they adjusted their rate of eating accordingly. This suggests that the control of eating rate involves visual feedback and is not a simple reflexive response to orosensory stimulation. 
Type Of Material Database/Collection of data 
Year Produced 2015 
Provided To Others? Yes  
 
Title Variation in the oral processing of everyday meals is associated with fullness and meal size; a potential nudge to reduce energy intake 
Description Raw data for the paper entitled "Variation in the oral processing of everyday meals is associated with fullness and meal size; a potential nudge to reduce energy intake". Abstract: Laboratory studies have demonstrated that experimental manipulations of oral processing can have a marked effect on energy intake. Here, we explored whether variations in oral processing across a range of unmodified everyday meals could affect post-meal fullness and meal size. In Study 1, female participants (N=12) attended the laboratory over 20 lunchtime sessions to consume a 400-kcal portion of a different commercially available pre-packaged meal. Prior to consumption, expected satiation was assessed. During each meal, oral processing was characterised using; i) video-recordings of the mouth and ii) real-time measures of plate weight. Hunger and fullness ratings were elicited pre- and post-consumption, and for a further three hours. Foods that were eaten slowly had higher expected satiation and delivered more satiation and satiety. Building on these findings, in Study 2 we selected two meals (identical energy density) from Study 1 that were equally liked but maximised differences in oral processing. On separate days, male and female participants (N=24) consumed a 400-kcal portion of either the 'fast' or 'slow' meal followed by an ad libitum meal (either the same food or a dessert). When continuing with the same food, participants consumed less of the slow meal. Further, differences in food intake during the ad libitum meal were not compensated at a subsequent snacking opportunity an hour later. Together, these findings suggest that variations in oral processing across a range of unmodified everyday meals can affect fullness after consuming a fixed portion and can also impact meal size. Modifying food form to encourage increased oral processing (albeit to a lesser extent than in experimental manipulations) might represent a viable target for food manufacturers to help to nudge consumers to manage their weight. 
Type Of Material Database/Collection of data 
Year Produced 2015 
Provided To Others? Yes  
 
Description Collaboration with the Community Mandolean randomised trial (ComMando) 
Organisation University of Bristol
Department School of Social and Community Medicine
Country United Kingdom 
Sector Academic/University 
PI Contribution The Nutrition and Behaviour Unit (NBU) has been involved in a collaboration with the Community Mandolean randomised trial (ComMando), which was funded through the National Institute for Health Research (NIHR) health technology assessment (HTA) programme (Ref: 09/127/04). The original aim of this trial was to use quantitative and qualitative methods to investigate the efficacy and cost-effectiveness of using a computer device, Mandolean (previously Mandometer) to retrain pro-obesogenic eating behaviours (speed of eating and portion size determination) in obese children and their families as an adjunct to standard lifestyle education in primary care clinics. The NBU developed computer software for use by practice nurses to obtain precise estimates of 'ideal portion size' and 'expected satiety' across a range of commonly consumed foods. The plan was to compare these estimates across groups (all 600 participants) at 0, 12, and 24 months. Together, these data were going to be used to determine whether Mandolean training modifies beliefs about satiety, and the extent to which this change is reflected in the selection of smaller portion sizes. However, the study was terminated after the feasibility phase in July 2013. Nevertheless, during the course of the ComMando feasibility trial, 62 overweight/obese children between the ages of 5 - 11 and their parents completed the NBU computer tasks. This is an unusually large sample of young children that completed measures of theoretical importance, including assessments of ideal portion size, the maximum portion size they could imagine consuming, measures of ranked preference and liking (these measures are outlined in further detail in the protocol section below). To our knowledge, to date, only one published study has explored expectations about fullness in children (Hardman, McCrickerd, Brunstrom, 2011, American Journal of Clinical Nutrition). In addition, since each child's parent also completed the measures about their child, we can also explore the relationship between parental beliefs and their child's beliefs. Therefore, we conducted a further study at @Bristol Science centre to recruit a large control sample of healthy weight children and their parents for comparison purposes. Our results suggested that overweight children choose to consume more within a meal but do not demonstrate a greater tolerance of larger portions. In addition, parents demonstrated greater accuracy in predicting their child's preference where there was concordance between parent and child BMI. The results of these studies are currently being prepared for publication by the research team (Potter et al, In preparation).
Collaborator Contribution Professor Julian Hamilton Shield and his team were instrumental in securing the funding and executing/managing the ComMando trial. They are also assisting in interpreting the findings from our follow up studies and providing feedback on paper drafts.
Impact The results from these studies have already been presented at a national conference (the British Feeding and Drinking Group conference). An abstract from this presentation has been published in the journal Appetite (see Potter et al , 2014). A paper based on the results of this study are currently being prepared for submission to a journal (Potter et al, In preparation).
Start Year 2012
 
Description 'The Incredible Edible Show' - An interactive event based at a local science museum - 'Explore at-Bristol' 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact Helped to develop and deliver demonstrations of psychological controls of food intake. Audience was primarily school-age children and their parents.

Feedback from the public-science exhibition staff was extremely positive. This led to a longer-term collaboration to enable further public dissemination of activities in this museum.
Year(s) Of Engagement Activity 2012
 
Description Advisory panel for @Bristol Science and Discovery Center 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact Professor Jeffrey Brunstrom and Dr Danielle Ferriday were members of a scientific advisory panel for a local science and discovery centre (@Bristol). Their specific role was to assist in scientific fact checking and idea generation for interactive displays in the development of their Food! exhibit. This exhibit aims to be an educational yet fun and interactive exhibit for children and their guardians to learn about where their food comes from and the science behind food development/food choice. Brunstrom and Ferriday attended meetings to assist in the development of interactive displays and shared current scientific findings to assist the team @Bristol. They also assisted in the development of online materials for interested parties. The exhibit launched on 23rd July 2014.

@Bristol receives hordes of visitors each year and the Food! exhibit has already received lots of positive feedback/interest. Through the explore more website, any parent/child who is interested can find out further information.
Year(s) Of Engagement Activity 2012,2013,2014
URL http://www.at-bristol.org.uk/2337
 
Description BBC Radio 4 Food Programme 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Participation in radio broadcast which sparked discussion about food cravings and preferences.

Increased public awareness regarding the factors which contribute to food choice.
Year(s) Of Engagement Activity 2012
URL http://www.bbc.co.uk/programmes/p00yyr41
 
Description Café Scientifique meeting, Explore at-Bristol 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact Talk generated a Q&A afterwards which was followed by informal discussions with several members of the public.

One member of the audience asked me about job opportunities and access to a career in science.
Year(s) Of Engagement Activity 2012
 
Description Interactive demonstration @Bristol science and discovery centre 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Public/other audiences
Results and Impact Christina Potter (Ph.D. student) designed an interactive demonstration with project staff at At-Bristol for visitors at the Science centre. In the demo, we presented various portion sizes of a snack food (wine gums). The wine gums were presented in different portion sizes on 9 plates, ranging from 2g (one wine gum) up to 386g (a full plate of wine gums). Children had the option to try a wine gum to eat. As a visual feedback, we displayed bar charts where children could place a sticker underneath a portion size which corresponds to their ideal portion as well as their maximum portion.

As many of our computerised tasks involve portion size selection, we felt that this educational activity paired well with the research task and helped to illustrate the difference between "ideal" portions and "maximum" portions. The researchers then used conversation cues for child/family groups, while being careful to avoid any assertive message about eating. These cues were implemented to generate a discussion with the family. Further, any questions that parents or children had about the research tasks, such as explaining the more scientific concepts of our measures (i.e. expected satiety), were answered by the research staff on site.

Example cues:
'How do you know when you're full?' This cue lead to conversations about physical and mental sensations of fullness and hunger.
'What do you think happens to food once it's inside your body?' This cue lead to conversations about your body using food to give you energy to do move/play, to allow you to grow, to help your body to recover if you're ill or hurt yourself.
"What is the most you could possibly eat?"
"What is your perfect amount to eat"

The overall feedback from both parents and children who took part in the activity was extremely positive. Many parents commented that they hadn't ever considered several of the topics that we discussed with their child (such as portion tolerance) and they were interested to learn more. The activity often lead to conversations about how children, particularly those younger in age, have "eyes bigger than their stomach" when selecting both their ideal and maximum potion sizes.
Year(s) Of Engagement Activity 2013,2014
 
Description Media interest (Portion size, expected satiety, and the role of dietary variability) 
Form Of Engagement Activity A press release, press conference or response to a media enquiry/interview
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Current research projects were highlighted on BBSRC website.

Web coverage of our projects highlighted our current lines of research, attracting new interest in our research team.
Year(s) Of Engagement Activity 2013
URL http://www.bbsrc.ac.uk/news/health/2013/130204-f-would-you-like-to-supersize-that.aspx
 
Description Stall at Food Matters Live 2016 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact More than 13,500 experts from across the food, health and nutrition sectors attended Food Matters Live 2016. The Nutrition and Behaviour Unit hosted a stall at this event and featured a Powerpoint projection which detailed current research projects/ideas, methodologies, facilities. Recent scientific publications were also available for distribution. This stall sparked interest, conversations and requests for further information.
Year(s) Of Engagement Activity 2016
URL https://www.foodmatterslive.com/