Nudging consumers towards healthier choices: A systematic review of positional influences on food choice

Abstract

Nudging or ‘choice architecture’ refers to strategic changes in the environment that are anticipated to alter people’s behaviour in a predictable way, without forbidding any options or significantly changing their economic incentives. Nudging strategies may be used to promote healthy eating behaviour. However, to date, the scientific evidence has not been systematically reviewed to enable practitioners and policymakers to implement, or argue for the implementation of, specific measures to support nudging strategies. This systematic review investigated the effect of positional changes of food placement on food choice. In total, seven scientific databases were searched using relevant keywords to identify interventions that manipulated food position (proximity or order) to generate a change in food selection, sales or consumption, among normal-weight or overweight individuals across any age group. From 2576 identified articles, fifteen articles comprising eighteen studies met our inclusion criteria. This review has identified that manipulation of food product order or proximity can influence food choice. Such approaches offer promise in terms of impacting on consumer behaviour. However, there is a need for high-quality studies that quantify the magnitude of positional effects on food choice in conjunction with measuring the impact on food intake, particularly in the longer term. Future studies should use outcome measures such as change in grams of food consumed or energy intake to quantify the impact on dietary intake and potential impacts on nutrition-related health. Research is also needed to evaluate potential compensatory behaviours secondary to such interventions.

Full text

Nudging consumers towards healthier choices: a systematic review
of positional influences on food choice
Tamara Bucher
1,2
*, Clare Collins
1,2
, Megan E. Rollo
1,2
, Tracy A. McCaffrey
3
, Nienke De Vlieger
1,2
,
Daphne Van der Bend
1,2
, Helen Truby
3
and Federico J. A. Perez-Cueto
4
1
Nutrition and Dietetics, School of Health Sciences, Faculty of Health and Medicine, The University of Newcastle, University
Drive, Newcastle, Callaghan NSW 2300, Australia
2
Priority Research Centre in Physical Activity and Nutrition, The University of Newcastle, University Drive, Newcastle,
Callaghan NSW 2300, Australia
3
Department of Nutrition and Dietetics, School of Clinical Sciences, Monash University, Melbourne, VIC 3168, Australia
4
Sensory and Consumer Section, Department of Food Science, University of Copenhagen, Rolighedsvej 26,
1958 Frederiksberg C, Denmark
(Submitted 26 August 2015 –Final revision received 21 March 2016 –Accepted 22 March 2016)
Abstract
Nudging or ‘choice architecture’refers to strategic changes in the environment that are anticipated to alter people’s behaviour in a predictable
way, without forbidding any options or significantly changing their economic incentives. Nudging strategies may be used to promote healthy
eating behaviour. However, to date, the scientific evidence has not been systematically reviewed to enable practitioners and policymakers to
implement, or argue for the implementation of, specific measures to support nudging strategies. This systematic review investigated the effect
of positional changes of food placement on food choice. In total, seven scientific databases were searched using relevant keywords to identify
interventions that manipulated food position (proximity or order) to generate a change in food selection, sales or consumption, among
normal-weight or overweight individuals across any age group. From 2576 identified articles, fifteen articles comprising eighteen studies met
our inclusion criteria. This review has identified that manipulation of food product order or proximity can influence food choice. Such
approaches offer promise in terms of impacting on consumer behaviour. However, there is a need for high-quality studies that quantify the
magnitude of positional effects on food choice in conjunction with measuring the impact on food intake, particularly in the longer term. Future
studies should use outcome measures such as change in grams of food consumed or energy intake to quantify the impact on dietary intake
and potential impacts on nutrition-related health. Research is also needed to evaluate potential compensatory behaviours secondary to such
interventions.
Key words: Nudging: Choice architecture: Food position: Proximity: Order: Food choices: Environmental influences
In recent years, there has been a shift away from solely targeting
individuals to change their eating behaviours to an
approach that addresses wider, population-level factors and
involves other environmental components and stakeholders
(1)
.
Foodscapes
(2)
and food environments contribute to the
so-called ‘obesogenic environment’
(1,3)
and influence food
choices. Epidemiological data suggest that numerous small
changes towards a healthier behaviour such as improving diet
quality have the potential to have a positive impact on reducing
mortality risk
(4)
. Most healthy eating interventions in Europe
have been successful in providing consumers with information
to enable them to make better-informed food choices
(5)
.
Although they have been successful in creating awareness
among consumers, there has only been modest success in terms
of actual lifestyle changes and measurable health indicators
in the sample populations, such as weight reduction
(6)
.
Individualised behaviour change is ineffective unless it
becomes habit forming, which requires support and reinforce-
ment through structural or environmental change so that the
new behaviour is sustained. Although behavioural economics
have impacted on some policy interventions, the case for food-
related interventions remains under development, constituting a
promising area that could potentially achieve high social
benefits
(7,8)
.
Therefore, innovative intervention strategies that are able to
effectively improve food behaviours, dietary intake and impact
on health status need to be investigated and implemented. The
majority of interventions have an underlying assumption that
people make conscious and reasoned food choices, most of the
time
(9)
. This paradigm has been questioned following the
limited impact of information-based campaigns in achieving
behaviour change, and the subsequent rise in the prevalence of
obesity and other chronic diseases
(10)
. Furthermore, current
paradigms place the burden and responsibility for all food
*Corresponding author: T. Bucher, email tamara.bucher@newcastle.edu.au
British Journal of Nutrition, page 1 of 12 doi:10.1017/S0007114516001653
© The Authors 2016

choices on the individual, with the justification that everyone is
free to make healthy choices once informed
(6,11)
.
Dietary habits and food choices are the result of decisions
and actions that are based on routines that require very little
active decision making as well as reflective, elaborate decision
making where choice options are carefully considered. Choice
architecture, inspired by behavioural economics, describes the
way in which decisions are influenced based on how choices
are presented within meal environments
(12)
. The meal
environment has been defined as the room, the people, the
food, the atmosphere and the management system, particularly
when eating out of home. This suggests that the meal
environment can be modified to be more or less conducive to
support the required behaviour and as such may lead to weight
changes, either through promotion of healthier choices or
through decreased intake
(12–15)
.
Choice architecture is often used interchangeably with other
terms such as nudging, libertarian paternalism and behavioural
economics. Choice architecture is a subset of non-regulatory
behavioural interventions. Choice architecture can include one
or more of the following: provision of information (e.g. to
activate a rational choice), changes in the physical environment
(e.g. light, décor, placement, etc.), changes in the default policy
(e.g. pre-weighed salad portions v. free serving of a salad bowl)
and use of social norms and salience (e.g. comparison with
average consumers)
(16)
. Nudging has been defined as any aspect
of the choice architecture that alters people’s behaviour in a
predictable way without forbidding any options or significantly
changing their economic incentives
(15)
. Within the public health
nutrition area, this could mean altering the food environment,
such as product placement or labelling or even encouraging
consumers to sit together for their meal (social facilitation).
Furthermore, nudging interventions consist of provision of
information, changes to physical environment, changes to the
default policy and the use of social norms and salience
(16)
.
Previous studies have shown that nudging practices are
promising measures that can be used to support the promotion of
healthy eating. An example of nudging is that by changing the
size of dishware, portion sizes may be reduced leading to
unconscious changes in actual food intake
(17)
and meal compo-
sition
(18)
. Similarly, food positioning is thought to influence food
choice. Studies have shown that people eat more unhealthy food
such as chocolate if it is located more prominently
(19)
.However,it
is less clear whether minor changes in food position or item
placement, which are not accompanied by changes in effort, also
promote healthier food choices
(13,20)
.
Existing systematic reviews that have investigated the effec-
tiveness of choice architecture interventions have mainly
focused on the effectiveness of labelling and prompting
(21,22)
.
However, these types of interventions are more closely related
to the traditional behavioural interventions of information
giving
(23)
. To date, there is no systematic review that has
assessed the influence of food placement within micro-
environments on product choice and on food intake
(23)
. This
information is relevant for the support of public health inter-
ventions and relevant for operations in the food service sector.
The aims of this systematic review were to evaluate
published articles that have investigated the effect of positional
changes within microenvironments on food choice by healthy
weight and overweight individuals across all age groups and to
derive recommendations for future research in the area.
For the purpose of this review, we have defined a nudging
intervention as any intervention that involves altering the
non-economic properties or placement of objects or stimuli within
micro-environments with the intention of changing health-related
behaviour. Such interventions are implemented within the same
micro-environment in which the target behaviour is performed and
require minimal conscious engagement. In principle, these
interventions can influence the behaviour of many people
simultaneously, and they are not targeted or tailored to specific
individuals (adapted from Hollands et al.
(23)
). The present review
focuses on positional changes that affect immediate food
consumption or choice decisions of individuals (e.g. eating out of
home in a food service outlet), rather than the consumption pattern
of a family or a household over time, as it would be the case in
‘assortment structure’experiments within supermarket settings.
Methods
Details of the protocol for this systematic review were registered
on the International Prospective Register of Systematic Reviews
(PROSPERO) and can be accessed at http://www.crd.york.ac.
uk/PROSPERO/display_record.asp?ID=CRD42015016277
Criteria for study inclusion
The PICOS (Problem, Intervention, Comparison, Outcome,
Setting) approach
(24)
was used to frame the research question.
We defined ‘food choice’as all outcome measures that assessed
food selection or probability of food choice, including product
sales and food consumption (in grams or energy intake).
Positional changes were defined as all manipulations of food
order or variations in the distance of food placement relative to
consumers within microenvironments. Microenvironments
were defined as the immediate surroundings of the individuals,
such as within the home, workplace or cafeterias
(25)
.
The types of studies to be included were randomised-
controlled trials/experiments, pre–post experimental studies,
quasi experiments and naturalistic observations where at least
one research aim was to assess the influence of food posi-
tioning within a microenvironment on food choice (selection)
or sales (grams, number) and intake (grams, energy).
Studies where multiple variables were manipulated simulta-
neously along with the food position were not included. For
example, studies where foods were added or removed from the
selection or where portion sizes of healthy or unhealthy offers
were altered along with a positional change were excluded.
Study participants included only healthy, normal-weight or
overweight/obese individuals. There was no age restriction
with studies on both children and adults included. The search
included full-text articles that were published in peer-reviewed
journals in the English language.
Literature search
A systematic search was conducted using electronic databases
(Medline,Pre-Medline,Embase,CINAHL,Scopus,TheCochrane
2 T. Bucher et al.

Library and PsycINFO) until February 2015. No limit was placed
on publication date. The search term list included the following
items: choice architecture OR accessib*OR nudg* OR position* OR
(serving AND (direction OR distance)) OR proximity OR distance
AND food OR diet OR food choice OR energy intake OR caloric
restrictionORfruitsORvegetablesORhealth*ORfoodchoice.
Reference lists of included articles and key reviews in the area
were also manually searched for additional articles.
Review procedure
Two independent reviewers (T. B. and N. D. V./D. V. d. B.)
screened the titles and abstracts of all search results. Full texts of
all papers that appeared to potentially meet the inclusion cri-
teria were retrieved. The retrieved full texts were assessed by
two independent reviewers (T. B. and N. D. V.) to determine
inclusion. In case of disagreement, a third independent
reviewer made the final decision (M. E. R.).
Data extraction and synthesis
Quantitative data on study participants (age, sex, weight status),
the design (type of study, setting, manipulated variables) and the
outcomes (finding, main effect, conclusions) from the included
articles were extracted by T. B. and checked by M. E. R. To
distinguish between the magnitude of the change in effort that
was involved in the intervention, we differentiated between
minor changes (mere order change or very small distance
change within reach), medium (change of position to food that
required only a small effort, e.g. standing up, bending down) or
major positional changes (manipulations that involved a major
increase/reduction in effort, e.g. walking across a room).
Quality assessment of included studies
The quality of the included studies was assessed by two inde-
pendent reviewers (T. A. M. and H. T.) using the review evi-
dence analysis manual published by the Academy of Nutrition
and Dietetics
(26)
. The quality scores can be found in the online
Supplementary Table S1.
Results
The database search identified 2540 unique entries, which were
combined with another thirty-six articles of interest that were
identified by screening reference lists. A total of sixty-two full-
text articles were retrieved and assessed against the inclusion
criteria. In total, fifteen articles, comprising eighteen studies,
met the inclusion criteria and the data from these were
extracted and evaluated in this review (see Fig. 1).
The majority (n10) of studies were conducted in the USA;
seven were conducted in Europe, of which four were con-
ducted in the Netherlands. In one study, the country was not
reported
(27)
. There was only one study on children
(28)
.
Moreover, ten studies were conducted with university students
or staff, and for five studies the subjects were customers of
hospital cafeterias. Only one study was conducted in an Army
research centre
(29)
and one was conducted with attendees of a
health conference
(30)
.
The foods involved in the studies varied and included single
healthy or unhealthy items (water, fruit and vegetable, cereal
bars, chocolate candy or crackers) to more complex selections
within canteen buffets with between eight and eleven products
repositioned.
Among all, seven studies reported participants’weight status;
however, only two considered it in the analysis
(27,31)
. Levitz
(27)
reported that a change in dessert order affected normal-weight
and overweight people differently. In particular, the author
found that obese adults selected a greater amount of low-
energy dessert if it was made more salient. No changes were
observed if the high-energy desserts were made more
salient
(27)
.
The characteristics of the included studies are summarised in
Table 1.
Of the eighteen studies that were included, only one received
a positive quality rating
(32)
, with fourteen studies being assessed
as neutral and three as negative, because study procedures
were not described in detail and several validity questions could
not be answered clearly (online Supplementary Table S1).
Of the eighteen studies, nine investigated the effect of
distance/proximity changes on food choice, such as placing
unhealthy foods further from the consumer. The other half
assessed whether changes in product order, such as, for
example, the food sequence on a buffet, could have a beneficial
influence on food selection.
In summary, sixteen of the eighteen studies concluded that
positional changes had a positive influence on food choice. The
only two studies that did not find an effect manipulated the pro-
duct order of snacks on a computer screen (Van Kleef Study 1),
as well as within a shelf at a checkout counter in a cafeteria
(Van Kleef Study 2). However, in the field study, they found a
trend towards sales of healthy food being positively affected
(32)
.
15 full-text articles (18 studies) included
in the qualitative synthesis
62 full-text extracted and assessed for
eligibility
2576 records screened
2540 unique database
records identified
36 additional records identified
through reference lists
2514 records excluded
47 full-text ariticles
excluded
Fig. 1. Flow of information through the different phases of the review.
Nudging food position: a systematic review 3

Tab le 1. Characteristics of included studies (n18) assessing the effect of positional changes in the microenvironment on food choice*
Author, year Type of study Type of nudge Setting Country Subjects Subject age (years) Subject weight status
Engell et al., 1996 Experimental between-
subjects design
Distance/
proximity
Field study: Army
research centre dining
hall
USA (Boston) Employees of US Army
Natick Research Centre,
n36
39·5±13·2 Normal weight
(181·6±30·7) pounds,
70·6±2·4 inches)
Maas et al., 2012
(study 1)
Experimental between-
subjects design
Distance/
proximity
Laboratory The Netherlands
(Utrecht)
77 females recruited on
campus
17–38 Normal weight (BMI:
22·4±2·96 kg/m
2
)
Maas et al., 2012
(study 2)
Experimental between-
subjects design
Distance/
proximity
Laboratory The Netherlands
(Utrecht)
54 females recruited on
campus
17–38 Normal weight (BMI:
20·89 ±2·16 kg/m
2
)
Meiselman et al.,
1994 (study 1)
Experiment (repeated
measures possible)
Distance/
proximity
Field study: university
cafeteria
England
(Bournemouth)
Customers in university
cafeteria; 43 students
334 meals
18 years, 4; 19 years,
15; 20 years, 12;
21 years, 7; over
21·5 years
N/A
Meiselmann et al.,
1994 (study 2)
Experiment (repeated
measures possible)
Distance/
proximity
Field study: university
cafeteria
England
(Bournemouth)
Meals of customers in
university cafeteria; 60
students (36 male) who
consumed potato chips at
baseline
Between 18 and
62 years
N/A
Musher-Eizenman
et al., 2010
Correlation analysis Distance/
proximity
Field study: child day
care
USA (Ohio) 46 children 6·3±2·3, range:
3·4–11
8th to 98 percentile
(M =65th), 25 %
overweight (85th
percentile and higher)
Privitera et al., 2010 Between-subjects
experiment
Distance/
proximity
Laboratory USA (St.
Bonaventure,
NY)
96 (24 male) university
students
Variation 1:
19·9±1·1; variation
2: 20·1±16
BMI: 26·9±3·8and
26·4±4kg/m
2
(mean
overweight)
Privitera et al., 2014 Between-subjects
experiment
Distance/
proximity
Laboratory USA (St.
Bonaventure,
NY)
56 university students
(26 male)
19 ±0·9 BMI: 26·0±3·8kg/m
2
;
21 overweight,
15 obese
Wansink et al.,
2006
Within-subjects experiment Distance/
proximity
Field study: offices at
university
USA (Illinois) 40 female university staff
members
42·2±11·3N/A
Keller & Bucher,
2014
Experimental between-
subjects design
Order/
accessibility
Field study: university
campus
Switzerland
(Zurich)
120 students (60 male, age
24 ±3 years)
24 ±3N/A
Levitz, 1976 Naturalistic observation,
experiment (repeated
measures possible)
Order/
accessibility
Field study: hospital
cafeteria
N/A Customers in hospital
cafeteria. 3267
observations. Only
choices of normal-weight
(n2385) and obese
(n425) subjects were
analysed
N/A Normal weight and
overweight
(classification by
trained observers)
Levy et al., 2012 Longitudinal study pre–post
design
Order/
accessibility
Field study: Hospital
cafeteria
USA (Boston) 4642 employees of a hospital
cafeteria (71 % females)
41 N/A
Meyers et al., 1980 Experiment (repeated
measures possible)
Order/
accessibility
Field study: hospital
cafeteria
USA (Memphis) Customers in hospital
cafeteria. 4412
observations. Separate
analysis for normal-weight,
overweight and obese
subjects
N/A (adults) Normal weight and
overweight, assessed
by observer
4 T. Bucher et al.

It was not possible to quantify and directly compare the effect
sizes of the included studies, as the study designs were too
variable. Most of the studies were randomised-controlled
experiments, and only one study used correlation analysis to
study the relationship between distance and snack selection
(28)
.
This study found that the distance from the serving bowl sig-
nificantly predicted the number of crackers and carrot slices
consumed by children
(28)
.
Between-subject experiments were the most common study
design, whereas within-subject, repeated-measures designs
were rarely used. Only one study used a longitudinal design
(33)
,
which was a follow-up assessment of the intervention described
by Thorndike et al.
(34)
and was based on the same choice
architecture intervention. Both of these studies were retained in
the review because they had assessed different outcome
measures and were complementary.
Most studies assessed food selection or choice probability
using χ
2
tests, whereas only a few studies objectively measured
actual food intake in terms of food weight (g) or energy
(kJ/kcal) content. The intervention description and findings of
the included studies are summarised in Table 2.
Discussion
Out of eighteen studies where food position or order was
manipulated, sixteen showed a positive effect on food choice,
meaning that the participants were nudged towards a more
healthy food choice. In the two experiments
(32)
where posi-
tional changes had no impact on food choice, the degree of
manipulation was only a minor change in position, with all the
foods remaining within reach. This indicates that the strength of
the effect appears to depend on the type of positional manip-
ulation (order v. distance), as well as the magnitude of the
change, or how far away foods are placed.
Only one study assessed compensatory food choices
(35)
,
showing that changes in position resulted in compensatory
choices within same food categories. Further, movement of
potato chips to a more distant location, and hence a reduction
in chips selection, was accompanied by an increase in
starch selection choices among the foods that still remained
proximal
(35)
. For portion size changes, there is some evidence
from previous research that reducing offered portion sizes
does not result in immediate compensation
(36)
. However, in
that particular study, the intervention was conscious, and
consumers’self-control was activated by having servers ask
customers in a fast food restaurant whether they wanted to
downsize portions. Other studies, in which the overall energy of
a meal bundle for children was reduced, without the partici-
pants being aware, found that the overall energy intake was
significantly reduced
(37)
. More research on compensatory
behaviours is required to implement effective interventions in
practice.
The overall quality of the included studies was neutral. Only
a few papers described the procedures sufficiently well to allow
a clear evaluation of all validity questions. In particular, the
questions that related to subject selection, recruitment proce-
dures and comparison of study groups were unclear or not
Table 1. Continued
Author, year Type of study Type of nudge Setting Country Subjects Subject age (years) Subject weight status
Rozin et al., 2013
(study 3)
Experimental between-
subjects design
Order/
accessibility
Field study: university
cafeteria
USA
(Pennsylvania,
Philadelphia)
Customers of the university
cafeteria. Mainly
employees of the
University of Pennsylvania
N/A N/A
Thorndike et al.,
2012
Pre–post intervention Order/
accessibility
Field study: hospital
cafeteria
USA (Boston) Customers of hospital
cafeteria
N/A N/A
van Kleef et al.,
2012 (study 1)
Two-factor experimental
design, between subjects
Order/
accessibility
Laboratory The Netherlands
(Wageningen)
158 undergraduate students
(55 male)
21·8±6·7N/A
van Kleef et al.,
2012 (study 2)
Two-factor experimental
design, between subjects
Order/
accessibility
Hospital cafeteria The Netherlands
(Wageningen)
291 snack sales, customers
of hospital canteen
N/A N/A
Wansink et al.,
2013
Between-subjects
experiment
Order/
accessibility
Field study: conference
venue
USA (Illinois) 124 health conference
attendees
N/A N/A
*±indicates the standard deviation.
Nudging food position: a systematic review 5

Tab le 2. Intervention description and findings of the included studies (n18)
Author, year
Description of
intervention
Context (setting and
participants)
Magnitude of change in
effort Type of food involved
Data analysis
method
Dependent variables
(unit) Magnitude of the effect Main finding Conclusions
Engell et al.,
1996
Water pitcher on table, v.
dispenser at 20 ft or
40 ft distance
Customers in Army
research centre
dining hall
Major variation in
proximity. Large
increase in effort to
obtain water at a
dispenser across
room or in another
room
Water ANOVA Water
consumption (g)
Significant main effect of
proximity on intake:
F
2,33
=8·4, P<0·001,
post hoc tests:
significant reduction for
distant conditions
compared with
proximate condition, no
difference between the
two more distant
conditions
Major reduction of water
intake if dispenser is
further away (on table
v. 20 or 40 feet). No
difference between 20
feet and 40 feet. No
effect on other food
intake
Effort to obtain water
determined
amount
consumed
Maas et al.,
2012
(study 1)
Distance to snack bowl
was varied at 20, 70
and 140 cm
Staff/students
recruited to
laboratory on
university campus
Medium: 70 and 140 cm
proximity variation
required standing up
Candy: Chocolate
M&M’s (without
peanuts) 1 kg
Logistic regression,
ANCOVA (control
for chocolate
liking)
Amount of snack
consumed (g) and
risk of
compensatory
behaviour
Significant main effect of
proximity on intake:
F
2,73
=7·59, P=0·001
Post hoc tests:
significant reduction for
distant conditions
compared with
proximate condition, no
difference between the
two more distant
conditions
An increase in distance
had a significant effect
on the probability of
snack consumption
even for an increase
from 20 to 70 cm.
No effect for
compensatory eating
was found
Distance affected
intake, but
salience did not
Maas et al.,
2012
Distance to snack bowl
was varied at 20, 70
and 140 cm
Staff/students
recruited to
laboratory on
university campus
Medium: 20, 70 and
140 cm. 70 and
140 cm required
standing up
Candy: Chocolate
M&M’s (without
peanuts) 1 kg
ANOVA Amount of snack
consumed (g),
perception of
salience and
effort, likelihood of
consumption
Significant main effect of
proximity on intake:
F
2,51
=3·8, P=0·029,
post hoc tests:
significant reduction for
distant conditions
compared with
proximate condition, no
difference between the
two more distant
conditions
An increase in distance
had a significant effect
on the probability of
snack consumption
even for an increase
from 20 to 70 cm.
Perceived effort
increased in distant
conditions but not
salience
Significant effect of
proximity on
intake. Perceived
effort was higher
in the two distant
conditions, but not
perceived
salience
Meiselman
et al., 1994
(study 1)
Move candy from cash
point to distant snack
bar
Customers
(students) of
university
cafeteria
Major; increase of
distance (20 m) and
waiting at separate
queue.
Plus reduction in
availability, (from four
cash registers to one
snack bar)
9 food categories:
main dishes:
pizza,
alternatives,
salads, sandwich;
desserts: fruit,
accessory foods,
candy: chocolate,
chocolate
containing bars
and muesli bars
Binominal model (χ2) Candy selection with
meals (selection
rates)
Less candy selected in
nudging condition:
χ2(1) =17·78, P<0·001.
Trend towards more
total desserts:
χ2(1) =2·21, P<0·1
(ns), no effect on other
foods
Less candy was
purchased during the
intervention week.
However, participants
who chose candy in
the first week chose
more dessert fruit or
accessory foods
during the intervention
week
Major increase in
effort to obtain an
unhealthy food
can reduce the
consumption of
the food. People
may partially
compensate
unhealthy choices
Meiselman
et al., 1994
(study 2)
Move potato chips from
cash register to
distant snack bar
Customers
(students) of
university
cafeteria
Major; increase of
distance (20 m) and
waiting at separate
queue.
Plus reduction in
availability, (from four
cash registers to one
snack bar)
11 food categories:
main meal:
pizzas, starch,
vegetables,
salads, bread,
sandwiches,
dessert, fruit,
crisps, sweets/
cakes, sauces,
candy, drinks
Fleiss’s formula and
χ2tests
Potato chips
selection with
meals (selection
rates)
Less potato chips selected
in nudging condition:
χ2(1) =77·27, P<0·001.
More starch foods
during intervention:
χ2(1) =6·20, P<0·001
Increased effort reduced
potato chips selection,
reduction was
accompanied by
increased starch
selection
Varying effort can
increase or
decrease
consumption.
Foods are
substituted with
other foods (within
same food group)
Musher-
Eizenman
et al., 2010
Children were randomly
placed at varying
distances to healthy
and unhealthy snacks
Children in day care Major: children had to
stand up and come up
to the experimenter
from varying distances
and ask for more
snack
Snacks: high energy
dense animal
crackers v. carrot
slices
Hierarchical
regression
Consumption of
crackers and
carrot slices
(number of pieces
consumed)
Distance from serving bowl
predicted intake:
Distance from crackers:
β=−0·41, P<0·05
(ΔR
2
=0·17)
Distance from carrots
β=−0·38, P<0·05
(ΔR
2
=0·14)
Distance from the
serving bowl
significantly predicted
number of crackers
and carrot slices
consumed
Proximity influences
consumption of
healthy and
unhealthy snacks
in children

Table 2. Continued
Author, year
Description of
intervention
Context (setting and
participants)
Magnitude of change in
effort Type of food involved
Data analysis
method
Dependent variables
(unit) Magnitude of the effect Main finding Conclusions
Privitera et al.,
2012
Manipulation of proximity
(near v. far) and
visibility (clear v.
opaque bowl) of
healthy foods
Students recruited to
laboratory on
university campus
Medium: serving bowl
placed 2 m away on
counter or on table
within arm’s reach
Snacks (healthy):
fruits and
vegetables
ANOVA Apple and carrot
consumption
(number of pieces
consumed)
Significant effect of
distance on intake:
Apple: F
1,44
=25·46,
P<0·001;
Carrots: F
1,44
=4·52,
P<0·04
Proximity increased
intake of both, fruit
and vegetable intake
(visibility only affected
of fruit intake)
Proximate can
increase
consumption of
healthy foods.
The effect was
stronger for
apples compared
to carrots. This
might be because
fruits are sweeter
and more
appealing than
vegetables
Privitera et al.,
2014
Effect of proximity was
tested in a competitive
food environment with
healthy food and
unhealthy food at
different distances
Students recruited to
laboratory on
university campus
Medium: 2 m v. arms
reach
Snacks: apple slices
(healthy food) v.
buttered popcorn
(high-fat/
unhealthy food)
ANCOVA: BMI as
covariate
Apple and popcorn
consumption (kcal
and proportion)
Proximity influenced
intake:
Popcorn: t
17
=4·96,
P<0·001
Apple: t
16
=5·16,
P<0·001
Significant interaction of
proximity and food type
F
2,52
=16·46, P<0·001,
R
2
0·38
The food that was placed
closer to the
participants was
consumed most,
regardless of
preference
Making a low-energy
food more
proximate than a
high-energy food,
will reduce total
energy intake,
even if a high-
energy and more
preferred food is
also available but
less proximate
Wansink
et al., 2006
Manipulation of proximity
(near v. far) and
visibility (clear v.
opaque bowl) of candy
Female staff within
their offices at
university
Medium: 2 m v. arms
reach
Candy (individually
wrapped
chocolates)
ANOVA (post hoc
ttests)
Chocolate
consumption
(number of
pieces)
1·8 chocolates more
consumed if they were
proximate. Effect size
unclear
More candy consumed if
it is more proximate
Proximity increases
consumption.
People
overestimate
consumption of
less proximate
foods
Keller &
Bucher,
2014
Manipulation of snack
bar order on tray;
healthy bar at the side
v. in the middle of an
assortment
Students recruited
on campus at
university
Minor or none: only
positions of foods
within reach were
altered
Snacks: healthy
apple cereal bar
v. unhealthy
cereal bars
(chocolate cereal
bars)
χ2Cereal bar choice
(selection rates)
Significant influence of
position on selection:
χ2(2) =14·95, P<0·001
The healthy bar was
selected more often,
when it was placed in
the middle
Changing the
position of snacks
can nudge
healthier choices
Levitz, 1976 Order of desserts with
varying energy
content in within
shelves; front v. rear
position
Customers in
hospital cafeteria
Minor: change within
display
Three types of
dessert;
High energy:
cakes and pies,
1464 kJ/serving
(350 kcal/serving)
Low-energy: fruit,
gelatine, 314 kJ/
serving (75 kcal/
serving)
Moderate:
custard, pudding
χ2Dessert sales
(selection rates)
Normal-weight subjects:
low-energy dessert
more available:
χ2=4·13, P<0·05
High-energy dessert
more available:
χ2=3·96, P<0·05
Obese subjects: low-
energy dessert more
available: χ2=17·67,
P<0·05
High-energy dessert
more available: ns
Normal-weight
individuals
consistently selected
the most available
choice
Obese people chose
more low-energy
dessert if it was made
more salient. No
change for obese if
high-energy dessert
was more salient
Both, obese and
normal-weight
individuals are
responsive
changes in food
positioning
Levy et al.,
2012
2-phase intervention 1st
phase: labelling of
healthy and unhealthy
food. 2nd phase:
placement variation of
various foods
Customers in
hospital cafeteria
Minor: for sandwiches
and chips: only
positions of foods
within shelves were
altered. Eye level v.
below eye level
position.
Medium for bottled
water: bottled water
available at several
locations in nudging
condition
Beverages,
sandwiches, chips
Linear regression
(demographics as
controls)
Sales of healthy and
unhealthy foods
(percentage of
change)
Decrease of red item
purchases by 4·1%
during the Phase 2
choice architecture
intervention
Repositioning red
(unhealthy) beverages
reduced sales in
addition to the colour
coding intervention
Choice architecture
intervention
improved food
and beverage
choices among
employees from
all racial and
socio-economic
backgrounds on
top of the labelling
intervention

Table 2. Continued
Author, year
Description of
intervention
Context (setting and
participants)
Magnitude of change in
effort Type of food involved
Data analysis
method
Dependent variables
(unit) Magnitude of the effect Main finding Conclusions
Meyers et al.,
1980
Manipulation of order of
desserts with varying
energy content within
shelves; front v. rear
position
Customers in
hospital cafeteria
Minor: change within
display
Desserts: two types;
high energy:
cakes and pies
Low energy: fresh
fruit and gelatine
Multiple contingency
analysis (χ2)
Dessert sales
(selection rates)
Likelihood to choose a
dessert in front was
increased.
χ2(2) =22·3, P<0·001
(significant interaction
between dessert array
and dessert choice)
Subjects were more likely
to choose the dessert
in front. No difference
between overweight
and normal-weight
subjects
All subjects were
more likely to
select the dessert
in front
Rozin et al.,
2013
(study 3)
Manipulation of salad
order at self-service
salad bar: less
accessible middle
position v. more
accessible edge
position
Customers in
hospital cafeteria
Minor: change within
display
8 ingredients at a
salad bar:
chicken, egg,
tuna, salmon,
tomatoes, carrots,
mushrooms
cucumbers
Multiple ttests Sales (weight) from
pay-by weight
salad bar
Average sales of each
ingredient was reduced
by 8·9 % in the middle
position compared with
the edge position
(t
7
=−4·13, P<0·01,
Z-score =0·30)
Sales of each of the eight
ingredients diminished
when displayed in the
less accessible middle
row
Food positions at
self-serving pay-
by-weight salad
bar had a
significant
influence on sales
Thorndike
et al., 2012
See Levy et al.
(33)
Customers in
hospital cafeteria
Medium and minor; see
Levy et al.
(33)
Beverages,
sandwiches, chips
Logistic regression Sales of healthy and
unhealthy foods,
selection rates
Decrease of unhealthy
beverage purchase by
11·4%.
Increase of healthy
beverage purchase by
4%.
Increased sales of
bottled water by 25 %,
P>0·001
Small but significant
increases in sales by
reordering
sandwiches and chips
on shelves
Choice architecture
intervention
improved food
and beverage
choices
van Kleef
et al., 2012
(study 1)
Manipulation of snack
position (healthy foods
on top v. at bottom)
Undergraduate
students recruited
to laboratory on
university campus
Minor; only positions of
foods on screen were
altered
Snacks: an
assortment of 16
(out of 24) healthy
and unhealthy
snacks; fresh and
dried fruit and
vegetables,
savoury and salty
snacks, and
sweet biscuits and
chocolates
Logistic regression
and ANOVA
Snack choice on
screen
No significant differences
were observed in the
‘healthy snacks at the
top’conditions (30·38 %
choose healthy)
compared with the
bottom conditions
(27·85 %; (1, n158) =
1·29, P=0·34)
No significant effect of
shelf position on
snack choice
Field study showed a
trend that
consumption of
healthy foods was
affected, but that
consumption of
unhealthy foods
was not altered
van Kleef
et al., 2012
(study 2)
Manipulation of shelf
position (healthy foods
on top v. at bottom)
Customers in
hospital cafeteria
Medium; positions of
foods within a shelf
were altered, reaching
some foods required
bending down
Snacks: an
assortment of 16
healthy and
unhealthy snacks:
fresh and dried
fruit and
vegetables,
savoury and salty
snacks, and
sweet biscuits and
chocolates
ANOVA Snack sales No significant effect of
shelf arrangement on
total snack sales
(F
1,6 =
3·84, P=0·1.
Separate analysis for
healthy and unhealthy
snacks revealed that no
effect on unhealthy
items but a trend
towards higher sales of
healthy items when
healthy foods were
placed on top
(F
1,6
=5·03, P=0·07
No significant effect of
shelf position for
unhealthy foods, but a
trend to higher sales
for healthy foods, if
they were placed
more prominent
No effect of
repositioning on
choice
Wansink
et al., 2013
Food order inverted at
breakfast buffet:
healthiest to least
healthy v. least healthy
to healthiest food
Conference
participants at
conference venue
Medium; order of foods
on buffet was altered
7 item buffet: cheesy
eggs, potatoes,
bacon, cinnamon
roll, low-fat
granola, low-fat
yogurt and fruit
χ2, Maxwell tests Breakfast item
selection
(selection rates)
Significant effect of order
on choice: χ2(6) =25·1,
P<0·001, Stuart-
Maxwell test =171·2
(P<0·001, df =6)
Order significantly
influences what
people select
First foods in line
were consumed
most often

applicable. Studies were classified as unclear or not being free
from bias because of the use of cash incentives or course credit
being offered to participants. This may be an artifact of the
naturalistic setting of the studies, such as universities and
workplace canteens.
There is a lack of research investigating long-term outcomes
of positional interventions, and it is not clear whether changes
in product order or distance would have sustained effects.
Specifically, it is unclear whether a potentially positive effect of
a position change, such as placing healthy foods in obvious
positions and very close to cafeteria check-out lines, would
potentially diminish over time and that customers would return
to selecting a favoured unhealthy snack. To investigate this,
more studies have to be conducted that evaluate this. Changes
in choice need to be assessed at different time points, ideally
over several weeks and months –for example, using data from
a customer loyalty card scheme to determine sustainability of
the intervention.
Furthermore, only one study
(38)
assessed the effect of
potential covariates such as food preferences, restrained or
disinhibited eating styles or health consciousness on the out-
comes of position choice architecture interventions. It therefore
remains unclear which individuals are susceptible to nudges.
Further insight on these covariates, as well as potential influ-
ences of habit strength, is required to design effective
interventions.
A reason for these data not being reported may be that it is
important to ensure participants are not aware of the nudging
intervention, and this is likely to be the reason most field studies
did not collect this information from participants. A method that
could be used to address this limitation in future research would
be to implement interventions within settings where customer
loyalty cards are used to collect additional data on participants’
actual purchases. For this purpose, collaborations with industry
or supermarket chains could be effective. This would also have
the advantage that potential product price and positioning
interactions could be assessed.
Previous literature suggests that nudges could be inexpensive
approaches to positively impact behaviours
(15)
. In the studies
included in this review, however, there were no calculations on
potential costs and benefits. Factual data on previously
hypothesised benefits are required to make effective recom-
mendations for policymakers.
Only two studies differentiated between healthy and
overweight consumers and whether positional interventions
were different based on body weight
(27,31)
. They both
concluded that the positional nudges were effective irrespective
of weight status. Further, one study assessed socio-economic
status and reported that it had no influence on whether
positional interventions were effective
(33)
. These findings con-
cur with previous literature, which suggest that nudging effects
work via subconscious mechanisms, and therefore have equal
impact regardless of weight and socio-economic status
(39)
.
Food position can be manipulated by changing the order of
food products or by changing the distance between the food
and the consumer. Both of these nudges operate in different
ways. The mediating factor for the effect of distance on choice
is thought to be effort, whereas for change in order it is reported
to be salience
(38)
. Changes in order normally constitute only a
minor change in effort, whereas changes in distance affect the
effort required in order to obtain a food at various levels.
However, more research is needed to evaluate these two
aspects in detail. Future research should also clearly distinguish
between studies that examine nudging in terms of food order v.
food proximity or distance.
To date, very little is known about why positional nudges
could be effective, and, in particular, it remains unclear how
effects are moderated. The dual-process model
(40)
states that
human behaviour largely results as a function of two interacting
systems: the reflective system, which generates decisions based
on knowledge about facts and values, and the impulsive
system, which elicits behaviour through affective responses.
The first system requires cognitive capacity, whereas the second
system requires no cognitive effort and is driven by feelings and
immediate behaviours in response to the environment. Nudging
is thought to operate mainly through the second, automatic
system and affects all individuals equally. However, it remains
to be elucidated whether and how factors such as health
consciousness, habits or strong preferences for specific
products interact with the effects. The research of Levy et al.
(33)
suggests that once the social gradient effects are taken into
consideration, there is still an effect towards the desired
outcome in terms of food choices. This indicates that these
interventions could be powerful and that cheap nudging
interventions could potentially yield more than other elaborate
expensive campaigns do. However, further research is required
to explore this in detail.
It was not possible to conduct a meta-analysis of effect sizes
as a wide range of outcome measures were reported across
studies. Although the evidence that food position influences
food choice was consistent across studies, it was not possible to
evaluate the impact and effect size of these types of choice
architecture interventions on actual food consumption and
subsequent health outcomes. As has been advocated
previously
(6)
, harmonised indicators are required that would
allow comparability between experiments or interventions. We
therefore strongly recommend the use of energy (kJ/kcal) or
weight (g) as outcome measures of changes in food selection
and/or intake in future studies.
Strengths and limitations
This is the first systematic review that has assessed the influence
of position interventions (proximity and order) on food choice.
In addition to the terms ‘nudging’and ‘choice architecture’,we
used search terms such as ‘distance’and ‘position’. This strategy
located many articles that were beyond the topic of interest
such as access to fast food outlets, but ensured that older
literature published before the terms ‘choice architecture’and
‘nudging’became popular were included.
For the purpose of this review, we defined nudging as any
intervention that involved altering the non-economic properties
or placement of objects or stimuli within microenvironments,
with the intention of changing health-related behaviour
(adapted from Hollands et al.
(23)
). We acknowledge that vary-
ing definitions of this term exist and that a disparate definition
Nudging food position: a systematic review 9

of the term might have led to the inclusion of different studies,
and hence influence the conclusions drawn.
Literature investigating the effect of the assortment structure
on buying behaviour within supermarkets was not identified
with the present search strategy. The authors are aware that
supermarket-related shopping behaviour has been extensively
described in the marketing literature, and that it is one of the
venues where behavioural interventions may have a socially
relevant outcome
(7,41)
. This aspect was beyond the scope of the
present study, which focused mainly on out-of-home meal
service situations such as cafeterias or canteens. Factors
affecting selection at the time of consumption and the time of
purchase may differ in this situation.
In addition, it is relevant to note that there could be
differences between nudges that aim to increase or decrease
consumption, as well as between nudges that promote the choice
of healthy foods v. nudges that discourage the consumption of
unhealthy foods. As an example, it might be easier to promote the
consumption of more (healthy) food, compared with discoura-
ging the consumption of unhealthy (or preferred) food by
positional changes. Studies in which the positions of unhealthy
and healthy foods are simply switched are particularly proble-
matic, as they lack a neutral control group, which would enable
researchers to disentangle whether there was a potential bias in
effectiveness of nudging depending on the food. In the present
literature, studies that strategically investigated the efficacy of the
positional intervention depending on food type are missing.
This review did not specifically consider any grey literature.
Given the heterogeneity and the limited number of studies
retrieved via the search strategy, it is plausible that a positive
publication bias exists, although this was not assessed by the
authors. It is interesting to note that the paternalistic nature of
the concept of nudging has been discussed. In particular, it can
be argued that a positional change that results in high effort to
obtain an unhealthier food may be seen as a reduction in
freedom of choice
(42–45)
. However, owing to the ethical nature
of this discussion, it is beyond the scope of this review.
The synthesis of the study findings was undertaken in a nar-
rative format as data aggregation was limited by the hetero-
geneity of the research in this field. Nevertheless, the current
review identified gaps in the existing literature and where further
research is needed.
Recommendations for laboratory studies
Although laboratory settings are limited, well-planned experi-
ments could give insight into the strength of positional effects,
and therefore help estimate the cost-effectiveness of choice
architecture interventions in practice, particularly if repeated
measures are applied. Laboratory settings allow the follow-up
of the same individuals for data collection. Quantifiable out-
come measures such as change in energy (kJ/kcal) or weight
(g) of food selection or consumption should be used. Strong
experimental evidence including estimations of the potential
health benefits secondary to a reduction in energy intake or
consumer weight loss over time is needed to inform policy-
makers in terms of implementing choice architecture interven-
tions in public health settings.
Recommendations for field experiments
Although previous research suggested that substitution might
occur within the same product category following a choice
architecture intervention
(35)
, a trial in the Belgian city of Ghent
showed that meal choices were not compensated for later in the
day
(46)
. Hence, future research should address the issue of
compensation at the design stage and consider that compen-
satory behaviours could occur after a nudge intervention.
As for laboratory settings, we also strongly recommend the
use of energy or grams of food selected/consumed as an
objective outcome measure, to estimate effect sizes and
potential health benefits.
Furthermore, insight into factors (e.g. preferences, habit
strengths, health consciousness) that potentially influence the
effectiveness of positional interventions could be gained by
collecting more information on customers in cafeteria-style
settings –for example, via a loyalty card scheme. This would
further allow exploration of the sustainability (decay of effect over
time or potential compensatory choices) over time in these
settings.
Reporting recommendations
The eighteen studies included in this review did not consistently
describe the choice architecture intervention that was being
assessed –for example, whether ‘the nudge’was a change in
distance or in product positioning. On the other hand, the
inclusion of distance in combination with food resulted in a
large number of search results that were not relevant for the
purpose of this study.
We suggest that standardised keywords and vocabulary
could assist this field of research. Researchers should carefully
consider the wording for their reports and could adopt the
terminologies suggested by Hollands et al.
(23)
to classify choice
architecture interventions more clearly.
Advice for practice (policymakers, food retailers)
Choice architecture recommendations could support existing
dietary guidelines, and thus potentially contribute to the
adherence and compliance. Although more research is required
to quantify the magnitude of positional influences on health
outcomes, it is evident that choice architecture is important and
that food retailers influence consumption by organising and
displaying their products. Therefore, persons in charge of food
organisation or food outlet design (e.g. workplaces) need to be
aware of their responsibility to organise ‘foodscapes’in an
optimal way –for example, to stimulate consumption of healthy
foods and to reduce the consumption of unhealthy foods that
then could support healthy workplace initiatives. In practical
terms, this means that low-energy, nutrient-dense products such
as fruits and vegetables should be placed in easily accessible
and prominent positions. This is particularly applicable in large
self-serving setting such a school or work canteens or the
canteens of residences for the elderly.
Policymakers could integrate choice architecture nudging
measures to augment their existing policy documents as an
important measure to enhance the effectiveness of healthy
10 T. Bucher et al.

eating policies and procedures. In particular, this review
provides evidence for policymakers, and specifically supports
the use of positional changes as an effective manner to alter
food choice in a desirable way.
Furthermore, the results of this review could be used
for developing official recommendations regarding the imple-
mentation of choice architectural nudge interventions and to
harmonise the indicators for evaluation of the effect. A good
practice example would be to place salad at the beginning of
the buffet in school canteens in countries where meals are
provided at school.
Conclusions
Although the evidence that food position influences food choice
is consistent, it is difficult to quantify the magnitude of impact on
food choice and intake and the effect size of these choice
architecture interventions on actual food consumption and
subsequent health outcomes. Use of harmonised terminology
and indicators would allow comparability between experiments
or interventions and assist in moving this field forward.
Acknowledgements
The authors thank Debbie both for assistance with developing
the search strategy and the database searches.
T. B. received a fellowship from the Swiss National Science
Foundation (P2EZP1_159086) and the Swiss Foundation for
Nutrition Research (SFEFS) to work on this project. C. C. is
supported by an NHMRC Senior Research fellowship. F. J. A. P.
C. is supported by IAPP-Marie Curie FP7/EU grant (agreement
no. 612326 VeggiEAT). The funding sources had no influence
on the design of the study.
T. B., N. D. V. and D. V. d. B. screened the abstracts; T. B. and
M. E. R. extracted the results; and T. A. M. and H. T. performed
the quality assessment. T. B. and F. J. A. P. C. jointly wrote the
manuscript with critical input from C. C., H. T., M. E. R. and
T. A. M.
The authors declare that there are no conflicts of interest.
Supplementary material
For supplementary material/s referred to in this article, please
visit http://dx.doi.org/doi:10.1017/S0007114516001653
References
1. Hill JO, Wyatt HR, Reed GW, et al. (2003) Obesity and the
environment: where do we go from here? Science 299,853–855.
2. Mikkelsen BE (2011) Images of foodscapes: introduction to
foodscape studies and their application in the study of healthy
eating out-of-home environments. Perspect Public Health
131, 209–216.
3. Hill JO & Peters JC (1998) Environmental contributions to the
obesity epidemic. Science 280, 1371–1374.
4. Bamia C, Trichopoulos D, Ferrari P, et al. (2007) Dietary
patterns and survival of older Europeans: the EPIC-Elderly
study (European Prospective Investigation into Cancer and
Nutrition). Public Health Nutr 10, 590–598.
5. Wills J & Grunert K (2007) A review of research on consumer
response to nutrition information on food labels in Europe
from 2003–2006. Ann Nutr Metab 51,37–38.
6. Perez-Cueto FJA, Aschemann-Witzel J, Shankar B, et al. (2012)
Assessment of evaluations made to healthy eating policies in
Europe: a review within the EATWELL Project. Public Health
Nutr 15, 1489–1496.
7. Johnson EJ, Shu SB, Benedict GC, et al. (2012) Beyond nud-
ges: tools of a choice architecture. Mark Lett 23, 487–504.
8. List JA & Samek AS (2015) The behavioralist as nutritionist:
leveraging behavioral economics to improve child food
choice and consumption. J Health Econ 39, 135–146.
9. Riebl SK, Estabrooks PA, Dunsmore JC, et al. (2015) A sys-
tematic literature review and meta-analysis: the theory of
planned behavior’s application to understand and predict
nutrition-related behaviors in youth. Eat Behav 18, 160–178.
10. Brambila-Macias J, Shankar B, Capacci S, et al. (2011) Policy
interventions to promote healthy eating: a review of what
works, what does not, and what is promising. Food Nutr Bull
32, 365–375.
11. Capacci S, Mazzocchi M, Shankar B, et al. (2012) Policies to
promote healthy eating in Europe: a structured review of
policies and their effectiveness. Nutr Rev 70, 188–200.
12. Hansen PG & Jespersen A (2013) Nudge and the manipulation
of choice. A framework for the responsible use of nudge
approach to behaviour change in public policy. Eur J Risk
Regul 1,3–28.
13. Rozin P, Scott S, Dingley M, et al. (2011) Nudge to nobesity I:
minor changes in accessibility decrease food intake. Judgm
Decis Mak 6, 323–332.
14. Wansink B (2007) Mindless Eating: Why We Eat More Than
We Think. New York, NY: Bantam.
15. Thaler RH & Sunstein CR (2008) Nudge Improving Decisions
About Health, Wealth and Happiness. New Haven, CT: Yale
University Press.
16. House of Lords (2011) Behaviour Change. Science and
Technology Select Committee. 2nd Report of Session 2010–12.
London: The Authority of the House of Lords.
17. Skov LR, Lourenco S, Hansen GL, et al. (2012) Choice
architecture as a means to change eating behaviour in self-
service settings: a systematic review. Obes Rev 14, 187–196.
18. Libotte E, Siegrist M & Bucher T (2014) The influence of plate
size on meal composition. Literature review and experiment.
Appetite 82,91–96.
19. Wansink B, Painter JE & Lee YK (2006) The office candy dish:
proximity’sinfluence on estimated and actual consumption.
Int J Obesity (Lond) 30, 871–875.
20. Bar-Hillel M & Dayan E (2011) Nudge to nobesity II: menu
positions influence food orders. Judgm Decis Mak 6, 333–342.
21. Sinclair SE, Cooper M & Mansfield ED (2014) The influence of
menu labeling on calories selected or consumed: a systematic
review and meta-analysis. J Acad Nutr Diet 114, 1375–1388.
22. Campos S, Doxey J & Hammond D (2011) Nutrition labels on
pre-packaged foods: a systematic review. Public Health Nutr
14, 1496–1506.
23. Hollands GJ, Shemilt I, Marteau TM, et al. (2013) Altering
micro-environments to change population health behaviour:
towards an evidence base for choice architecture interven-
tions. BMC Public Health 13, 1218.
24. Liberati A, Altman DG, Tetzlaff J, et al. (2009) The PRISMA
statement for reporting systematic reviews and meta-analyses
of studies that evaluate health care interventions: explanation
and elaboration. PLoS Med 6, e1000100.
25. Kahn BE & Wansink B (2004) The influence of assortment
structure on perceived variety and consumption quantities.
J Consum Res 30, 519–533.
Nudging food position: a systematic review 11

26. Academy of Nutrition and Dietetics (2012) Evidence Analysis
Manual: Steps in the Academy Evidence Analysis Process.
Chicago: Research and Strategic Business Development,
Academy of Nutrition and Dietetics.
27. Levitz LS (1976) The susceptibility of human feeding behavior
to external controls. In Obesity in Perspective, NIH publication
no. 75–708, pp. 53–60 [G Bray, editor]. Washington, DC: US
Government Printing Office.
28. Musher-Eizenman DR, Young KM, Laurene K, et al. (2010)
Children’s sensitivity to external food cues: how distance to
serving bowl influences children’s consumption. Health
Educ Behav 37, 186–192.
29. Engell D, Kramer M, MalafiT, et al. (1996) Effects of effort and
social modeling on drinking in humans. Appetite 26, 129–138.
30. Wansink B & Hanks AS (2013) Slim by design: serving healthy
foods first in buffet lines improves overall meal selection.
PLOS ONE 8, e77055.
31. Meyers AW, Stunkard AJ & Coll M (1980) Food accessibility
and food choice. A test of Schachter’s externality hypothesis.
Arch Gen Psychiatry 37, 1133–1135.
32. van Kleef E, Otten K & van Trijp HC (2012) Healthy snacks at
the checkout counter: a lab and field study on the impact of
shelf arrangement and assortment structure on consumer
choices. BMC Public Health 12, 1072.
33. Levy DE, Riis J, Sonnenberg LM, et al. (2012) Food choices of
minority and low-income employees: a cafeteria intervention.
Am J Prev Med 43, 240–248.
34. Thorndike AN, Sonnenberg L, Riis J, et al. (2012) A 2-phase
labeling and choice architecture intervention to improve healthy
food and beverage choices. Am J Public Health 102,527–533.
35. Meiselman HL, Hedderley D, Staddon SL, et al. (1994) Effect of
effort on meal selection and meal acceptability in a student
cafeteria. Appetite 23,43–55.
36. Schwarz J, Riis J, Eibel B, et al. (2012) Inviting consumers to
downsize fast-food portions significantly reduces calorie
consumption. Health Aff 31, 399–407.
37. Wansink B & Hanks AS (2014) Calorie reductions and within-
meal calorie compensation in children’s meal combos. Obesity
(Silver Spring) 22, 630–632.
38. Maas J, de Ridder DT, de Vet E, et al. (2012) Do distant foods
decrease intake? The effect of food accessibility on
consumption. Psychol Health 27, Suppl. 2, 59–73.
39. Marteau TM, Ogilvie D, Roland M, et al. (2011) Judging
nudging: can nudging improve population health? BMJ 342,
d228.
40. Strack F & Deutsch R (2004) Reflective and impulsive
determinants of social behavior. Pers Soc Psychol Rev 8,
220–247.
41. Ducrot P, Julia C, Méjean C, et al. (2015) Impact of different
front-of-pack nutrition labels on consumer purchasing
intentions: a randomized controlled trial. Am J Prev Med.
42. Blumenthal-Barby JS & Burroughs H (2012) Seeking better
health care outcomes: the ethics of using the ’nudge’.Am J
Bioeth 12,1–10.
43. Gold A & Lichtenberg P (2012) Don’t call me ’nudge’: the
ethical obligation to use effective interventions to promote
public health. Am J Bioeth 12,18–20.
44. Huang CJ & Baum ML (2012) Nudge ethics: Just a game of
billiards? Am J Bioeth 12,22–24.
45. Potts M, Verheijde JL & Rady MY (2012) When a nudge
becomes a shove. Am J Bioeth 12,40–42.
46. Hoefkens C, Lachat C, Kolsteren P, et al. (2011) Posting point-
of-purchase nutrition information in university canteens does
not influence meal choice and nutrient intake. Am J Clin Nutr
94, 562–570.
12 T. Bucher et al.