Content uploaded by Ralph Hertwig
Author content
All content in this area was uploaded by Ralph Hertwig on Feb 12, 2018
Content may be subject to copyright.
REG ART INCL REV
Public Beliefs About Obesity Relative to Other Major Health
Risks: Representative Cross-Sectional Surveys in the USA, the UK,
and Germany
JuttaMata, PhD1,2 • RalphHertwig, PhD1
Published online: XX XXXX 2018
© The Society of Behavioral Medicine 2018
Abstract
Background Overweight and obesity are among the lead-
ing risk factors for death worldwide. Scientists believe
that the increase in obesity is primarily due to environ-
mental changes and thus favor obesity prevention meas-
ures targeting the environment. However, it is less clear
what lay people perceive as causes of obesity, and which
measures they deem acceptable and promising in fight-
ing it.
Purpose This article compares lay beliefs about obesity
with beliefs about other major health risks sharing cer-
tain similarities with obesity (alcohol and tobacco de-
pendence, depression) in three countries with high
obesity rates.
Methods Computer-assisted face-to-face interviews
with representative samples in the UK (N=1,216) and
Germany (N= 973) and an online survey in the USA
(N=982) tapping beliefs about locus of responsibility,
liability for treatment costs, and effectiveness of policy
measures.
Results In each country, respondents attributed respon-
sibility for obesity primarily to the individual; the same
pattern emerged for alcohol and tobacco dependence,
but not for depression (ps < .01). The higher the attri-
bution of personal responsibility, the more strongly
respondents endorsed individual liability for treatment
costs (ps < .01). Respondents judged information and
fiscal policies as most and least effective, respectively, in
obesity prevention.
Conclusions Respondents’ views about obesity are simi-
lar to those about addictions; however, they regard fiscal
and regulatory policies as less effective for obesity than
for addictions. Raising awareness about environmental
drivers of obesity and framing policy measures by ref-
erence to the fight against tobacco and alcohol could
increase public support of obesity-targeted policies.
Keywords Representative survey • Personal responsibil-
ity • Obesity • Alcohol dependence • Tobacco depend-
ence • Depression
Introduction
Overweight and obesity are among the leading risk fac-
tors for death worldwide [1]. Policymakers, scientists,
and many citizens agree that the global obesity epi-
demic requires a forceful response. There is less agree-
ment, however, about the form this response should take.
Public health specialists generally attribute the rise in
obesity over recent decades to dramatic environmental
changes [2–4]. Accordingly, many proposed policy meas-
ures target the environment—for example, by imposing
surcharges on products that directly harm health, con-
tain no beneficial nutrients, and for which healthier alter-
natives are available (e.g., taxing obesogenic drinks [5])
or by restricting food marketing and sale (e.g., banning
advertisements for high-sugar children’s products [6]).
JuttaMata
mata@uni-mannheim.de
1 Center for Adaptive Rationality (ARC), Max Planck Institute
for Human Development, 14195 Berlin, Germany
2 Department of Social Sciences, University of Mannheim,
68161 Mannheim, Germany
ann. behav. med. (2018) XX:1–14
DOI: 10.1093/abm/kax003
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
It is less clear, however, what lay people think about the
causes of the obesity epidemic and which measures they
deem acceptable and promising in fighting it. Do they
agree with the diagnosis of a primarily environmental
disease or do they side with the food industry, regarding
diet to be principally a matter of personal responsibil-
ity rather than a justified target of regulatory and fiscal
measures [7]?
The goal of this study is to elicit and analyze lay beliefs
about obesity as compared with other global health risks,
with a focus on locus of responsibility, liability for treat-
ment costs, and effectiveness of policy measures. To this
end, we compare three countries with very high obesity
rates [8]: the USA, the UK, and Germany. Any differ-
ences observed between the three countries are likely
attributable to cultural, economic, or other differences,
rather than to differences in obesity prevalence.
To provide a frame of reference for lay beliefs about
obesity, we also obtained respondents’ beliefs about
three other major health risks: alcohol dependence, to-
bacco dependence, and depression. These risks were
chosen, first, because they are hypothesized to share
certain similarities with obesity and, second, because ef-
fective prevention and intervention policies have already
been successfully implemented for some of them. In
terms of similarities, it has been suggested that obesity
should be categorized as a substance dependence, akin
to alcohol or tobacco dependence [9]. Some individuals
with obesity would indeed fulfill the criteria for sub-
stance dependence (e.g., continued use despite physical
problems [10]). Other research has emphasized the links
between obesity and stress, thus raising the possibility of
obesity being a stress-related disorder, similar to depres-
sion: most prominent models of the etiology of depres-
sion assume that susceptible individuals are more likely
to become depressed when faced with chronic stress or a
stressful life event [11]. Chronic stress can also cause ex-
cessive consumption of high-calorie foods and, in turn,
weight gain (see [12] for a review).
In terms of intervention and prevention policies,
researchers and policymakers in all three countries have
endorsed and implemented hard paternalistic interven-
tions, such as fiscal and regulatory measures, as well as
softer measures, such as public information campaigns
and health warning labels, to combat alcohol and to-
bacco dependence. Although controversial when intro-
duced, such measures now commonly meet with broad
public approval. For example, surveys in the USA and
Germany have shown that most people now support
smoking bans in restaurants and other public areas [13,
14].
Public health researchers have suggested that the
obesity epidemic should likewise be addressed by fiscal
and regulatory measures [5, 10]. However, public sup-
port for such measures (e.g., taxes on high-calorie food
or supersized soft drinks) is presently low in Germany
[15], the UK, and especially the USA [16].
Research Aims and Hypotheses
Our representative study of the US, UK, and German
public compared lay beliefs about obesity with respect
to the locus of responsibility, liability for treatment
costs, and effectiveness of prevention policies with corre-
sponding beliefs about alcohol dependence, tobacco de-
pendence, and depression. In this article, we analyze the
following questions.
Locus of responsibility
Does the public attribute obesity to personal responsi-
bility, thus endorsing the causal model advocated by the
food industry, or do they attribute it to changes in the
environment, thus subscribing to the causal model advo-
cated by many public health experts? Furthermore, how
does obesity compare with addictions and depression in
terms of lay attributions of responsibility?
Liability for treatmentcosts
If respondents attribute a health risk to personal respon-
sibility, are they also more inclined to consider those
afflicted as being individually liable for treatment costs?
How does assignment of liability for treatment costs
compare across obesity, addictions, and depression?
Effectiveness of policy measures
What kind of policy measures do respondents consider
most effective in preventing obesity—and how does this
compare with policies implemented to fight tobacco and
alcohol dependence?
Methods
Respondents and Procedure
A total of 3,171 respondents from the USA (508 male,
474 female; aged 18–93years), the UK (607 male, 609
female; 18–93 years), and Germany (429 male, 544 fe-
male; 14–99years) were surveyed. All samples were rep-
resentative of the country’s population with respect to
age, gender, region, and other participant characteristics
described in Table 1. To account for cultural specifici-
ties, we assessed socioeconomic status differently in each
country: in the USA, respondents gave their annual
household income and level of education; in the UK,
they indicated their social class (“upper middle class” to
“lowest level of subsistence”) and whether they worked
full-time; in Germany, respondents reported their type of
work. In addition, respondents were representative with
2 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
Table1 Respondent Characteristics (Weighted)
% USA % UK % Germany
Gender Male 48.4 49.0 48.8
Female 51.6 51.0 51.2
Age 14‒19a4.0 2.5 7.1
20‒29 17.2 18.3 13.6
30‒39 16.2 15.5 13.4
40‒49 15.9 18.7 19.3
50‒59 21.6 16.3 16.3
60+ 25.2 28.7 30.2
Socioeconomic status Lowest level of subsistence 14.3
Working class 13.9
Lower middle class/skilled working class 49.5
Middle class 18.2
Upper middle class 4.0
Employment status In full-time work 55.0
Not in full-time work 45.0
Type of work Blue-collar worker 24.5
White-collar worker 31.6
Self-employed 7.5
Retired/not in work 32.1
Other 4.3
Annual household income Under $15,000 11.6
$15,000–less than $20,000 2.4
$20,000–less than $25,000 3.5
$25,000–less than $30,000 6.2
$30,000–less than $40,000 11.3
$40,000–less than $50,000 6.5
$50,000–less than $75,000 18.7
$75,000–less than $100,000 14.0
$100,000–less than $125,000 12.7
$125,000–less than $150,000 5.2
$150,000 and over 7.9
Education Less than high school 11.0
High school graduate 30.1
Some college/2-year degree 29.1
College graduate 17.4
Postgraduate school 12.5
Size of household 1 person 18.6 22.2
2 persons 33.5 38.2
3 persons 21.4 17.9
≥4 persons 26.5 21.7
Race Caucasian (White) 67.3
African-American (Black) 11.4
Asian or Pacic Islander 4.3
Hispanic 14.4
American Indian, Alaskan Native 2.7
(Table1 Continued)
ann. behav. med. (2018) XX:1–14 3
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
respect to race/ethnicity in the USA, with respect to size
of household in Germany and the UK, and with respect
to size of place of residence in Germany. The three sam-
ples were obtained using quota sampling, a systematic
sampling method that determines the proportion of indi-
viduals to be sampled from each subcategory [17]. The
resulting samples were stratified, and sampling weights
were applied to reflect the population structure with re-
spect to the subcategories described for each country (see
below for details).
Respondents were recruited by an international market
research company (Gesellschaft fuer Konsumforschung,
GfK). In Germany and the UK, respondents partici-
pated in a computer-assisted personal interview in their
homes. In the USA, respondents were recruited using
address-based sampling (part of the KnowledgePanel®)
and answered online questionnaires. Respondents
without Internet access were provided with a laptop
and free Internet access to complete the online sur-
veys. In all three countries and independent of survey
mode (face-to-face vs. online), participants sat in front
of a computer screen and inserted their responses into
the computer. The ethics committee of the Max Planck
Institute for Human Development approved the study.
Interview Questions
The questions were developed in German and then
translated into English by a certified translator for
English and German. A block of questions was pre-
sented for each health risk; the order of presentation
of the four blocks was randomized. With the exception
of the name of the risk, the wording of the questions
was identical across the four health risks: obesity, al-
cohol dependence, tobacco dependence, and depres-
sion. By way of illustration, we present the questions
concerning obesity.
Locus of responsibility
“To what extent are obese individuals responsible
for their weight themselves?” Responses were given
ona scale from 0 to 100 (or “don’t know”; modified
from [18]).
Liability for treatmentcosts
“Suppose obese individuals have to undergo treatment
because they are not able to get their weight under con-
trol alone. Should these individuals bear the costs of
treatment themselves?” Response options were “yes”
and “no.” Respondents who answered “yes” were then
asked what proportion (0%‒100% or “don’t know”) of
the treatment costs individuals should cover (modified
from [19]).
Effectiveness of policy measures
“How effective is measure X in preventing obesity?”
For obesity, alcohol dependence, and tobacco depend-
ence, respondents rated the effectiveness of the follow-
ing four policy measures on a scale from 0 to 100 (or
“don’t know”): (i) high taxes, (ii) nutritional or warning
labels, (iii) limiting availability or consumption in public
spaces, and (iv) banning or limiting advertising. These
measures were derived from the following references:
high taxes (on junk food [5]; alcohol [20]; tobacco [21]);
nutrition or warning labels (improved nutrition labels
[22]; warning labels on alcohol [23] and tobacco [24]);
limits on availability or consumption in public spaces
(banning soda vending machines in schools and at the
workplace [25]; policies to reduce general availability of
alcohol [20]; policies making more places smoke free
[21]); bans or limits on advertisements (for obesogenic
foods and drinks [6, 26]; for alcoholic drinks [27]; for
tobacco products [28]).
% USA % UK % Germany
Size of place of residence <2,000 inhabitants 5.8
2,000–19,999 inhabitants 36.6
20,000–99,999 inhabitants 27.5
100,000–499,999 inhabitants 14.2
≥500,000 inhabitants 15.9
Household net income (categories) Low/below average 23.8 30.8 30.1
Medium/about average 36.5 16.4 34.8
High/above average 39.8 14.2 11.4
No response 0.0 38.6 23.7
aIn the USA and UK, respondents in this age group were 18–19years old. High household net income=USA: $75,000 and over; UK:
£35,000 and over; Germany: €43,200 and over; medium income=USA: $30,000–$74,999; UK: £17,500–£34,999; Germany: €24,000–
€43,199; low income=USA: less than $30,000; UK: less than £17,500; Germany: less than €24,000. Samples were not representative with
respect to the household net income category (last table row).
Table1 (Continued)
4 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
Statistical Analyses
To achieve representativeness of the data for the US,
UK, and German populations, we applied sampling
weights in the descriptive analyses. The sampling weights
were different for each country and were based on the
participant characteristics reported in Table1 (i.e., the
sampling weights for the USA were based on gender,
age, annual household income, education, and race; the
procedure for the UK and Germany was analogous).
To control for the different sampling probabilities, we
included the variables used to calculate the sampling
weights in the parametric inference statistics (repeat-
ed-measures analyses of variance [ANOVAs], logistic
regression analyses, and regression analyses). Effect
sizes are given as η2. As a rule of thumb, an η2 of about
0.01 or below is regarded as small, an η2 of about 0.06
as medium, and an η2 of about 0.14 or above as large
[29]. Analyses were carried out using SPSS Version 24,
including the Complex Surveys Package [30].
Only the German sample included participants
younger than 18years of age (n=38 participants were
between 14 and 17years; 3.9% of the sample). To allow a
more equivalent comparison of results across countries,
we also recalculated all analyses, limiting the German
sample to participants aged 18years and older. All coef-
ficients from these analyses were equivalent in size and
direction to those from the full sample.
Results
To What Extent Is the Individual Held Responsible?
In all three countries, respondents attributed high levels
of responsibility for becoming obese to the individual
(Fig. 1). Responsibility for alcohol dependence and, in
particular, tobacco dependence was also primarily attrib-
uted to the individual. In contrast, across all countries,
depressed individuals were held to be less responsible for
their condition.
Three repeated-measures ANOVAs indicated that
attributions of personal responsibility differed sig-
nificantly across the four health risks, but were similar
across the three countries. When comparing obesity with
the other three health risks, within-subject constrasts
indicated that by far the largest difference was between
obesity and depression, followed by obesity and to-
bacco dependence in all three countries. The effect size
of the difference between obesity and alcohol depend-
ence was very small and was significant only in the UK
Fig.1. Attributions of personal responsibility: “To what extent are obese individuals/alcohol-dependent individuals/individuals who smoke
tobacco/individuals suffering from depression responsible for their weight/alcohol dependence/tobacco dependence/depression themselves?”
(0: not responsible at all; 100: fully responsible). The plot widths represent the density of the raw data distributions; the bandwidth of each
bean is determined by the difference between the smallest and largest density of the raw data per country. The lines represent the weighted
mean. For exact p values, see Table2.
ann. behav. med. (2018) XX:1–14 5
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
and Germany (see Table 2 for results of statistical signif-
icance tests).
To What Extent Should the Individual Be Liable for
TreatmentCosts?
About a third of respondents in the UK and Germany
believed that obese people should bear the costs of their
obesity treatment. This proportion was larger in the
USA, at nearly 45% (Fig.2). Across all three countries,
individual liability for treatment costs was most strongly
endorsed for tobacco dependence. As with attributions
of personal responsibility, the pattern of findings for
depression was distinct from that emerging for the other
health risks: only a small proportion of respondents—
and this proportion was again largest in the USA—
believed that people with depression should pay for the
costs of their treatment. Averaged across all four health
risks, the proportion of respondents who considered the
individual to be liable for treatment costs was consider-
ably higher in the USA (43.2%) than in the UK (32.3%)
or Germany (29.6%).
In each country, a Cochran’s Q test for dependent
binary variables showed that beliefs about individ-
ual liability for treatment costs differed across the four
health risks (after Bonferroni corrections, only p values
smaller or equal to .001 are considered statistically sig-
nificant): USA: Q(3)=276.34, p < .001; UK: Cochran’s
Q(3)=487.85, p < .001; and Germany: Q(3)=552.45,
p < .001. To test for differences between beliefs about
obesity and the other three health risks, we conducted
McNemar tests using Bonferroni correction to adjust
p values for multiple tests. Across the USA, UK, and
Germany, there was no significant difference between
beliefs about treatment liability for obesity versus
alcohol dependence, USA: Χ2 = 0.37, p = .562; UK:
Χ2=11.27, p=.003; Germany: Χ2=5.06, p=.025. In all
three countries, endorsement of individual liability for
treatment costs was significantly lower for obesity than
for tobacco dependence, UK: Χ2=22.78, p < .001; USA:
Χ2=31.00, p < .001; Germany: Χ2= 118.87, p < .001,
and significantly higher for obesity than for depression,
UK: Χ2=247.74, p < .001; USA: Χ2=122.78, p < .001;
Germany: Χ2=188.82, p <.001.
Was attribution of personal responsibility positively
associated with the belief that individuals should be
liable for treatment costs? We used logistic regression
analyses to test for this association (see Table3). Across
all countries and health risks, for every additional point
(up to a maximum of 100) that respondents attributed
individual responsibility for a health risk, the odds of
endorsing individual liability for its treatment costs
increased significantly—by between 3% (UK, Germany)
and 4% (USA) for obesity, and by between 2% and 4%
for the other health risks. Consistent with the previous
results, the odds of an increase were higher in the USA
than in the UK or Germany.
Which Policy Measures Are Judged to Be Effective in
Targeting Obesity?
We considered four policies designed to reduce the con-
sumption of potentially harmful substances, such as
sweet/fatty foods, alcohol, and tobacco: (i) high taxes, (ii)
limiting availability or consumption in public spaces, (iii)
regulating marketing (i.e., banning or limiting advertis-
ing), and (iv) labeling and warnings, see Fig.3.
How did respondents judge the effectiveness of
these policies? For each country, we ran three repeat-
ed-measures ANOVAs, each comparing judgments of
effectiveness of one policy across the three health risks
(depression was not included in these analyses; see
Table4 for statistical parameters). In all three countries,
taxation was judged as less effective in preventing obe-
sity than in preventing alcohol or tobacco dependence.
The effect sizes of the differences were consistently large,
Table2 Statistical Difference Values for Answers to the Question “To What Extent Are Obese Individuals/Alcohol-Dependent Individuals/
Individuals Who Smoke Tobacco/Individuals Suffering From Depression Responsible for Their Weight/Alcohol Dependence/Tobacco
Dependence/Depression Themselves?”
Main effect across the four health risks Within-subject contrasts
USA F(3, 1713)=561.75, p < .001, η2=0.50 O–A F(1, 571)=1.48, p = .225, η2=0.003
O–T F(1, 571)=71.17, p < .001, η2=0.11
O–D F(1, 571)=678.95, p < .001, η2=0.54
UK F(3, 3108)=1096.44, p < .001, η2=0.51 O–A F(1, 1036)=53.25, p < .001, η2=0.05
O–T F(1, 1036)=230.94, p < .001, η2=0.18
O–D F(1, 1036)=1255.12, p < .001, η2=0.55
Germany F(3, 2580)=1598.88, p < .001, η2=0.65 O–A F(1, 860)=27.85, p < .001, η2=0.03
O–T F(1, 860)=294.39, p < .001, η2=0.26
O–D F(1, 860)=1814.81, p < .001, η2=0.68
After Bonferroni corrections, only p values smaller or equal to .001 are considered statistically signicant. O obesity; A alcohol depend-
ence; T tobacco dependence; D depression.
6 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
with the exception of a medium-sized difference for obe-
sity versus alcohol in the UK. Furthermore, in all three
countries, high taxes were considered to be less effect-
ive than any of the other policies in preventing obesity.
Conversely, across all countries, understandable nutri-
tion labeling was regarded as the most effective policy
for preventing obesity. It was also considered to be sub-
stantially more effective than labels warning about the
dangers of alcohol, and moderately more effective than
labels warning about the dangers of tobacco products.
We also conducted three repeated-measures ANOVAs
comparing participants’ judgments of the effectiveness
of the four policies in the context of obesity. Across all
countries, the perceived effectiveness differed signifi-
cantly between the four policy areas (all ps < .001, η2
between 0.20 and 0.33; see Supplementary Table S1 and
Supplementary Fig. S1). We therefore conducted paired
comparisons to contrast the perceived effectiveness of the
four policy measures. In all three countries, the perceived
effectiveness of taxation was lowest and that of labeling
was highest. In the UK, banning or limiting advertising
was perceived as the second most effective policy meas-
ure and limiting availability or consumption in public
spaces as the third most effective; in Germany, this order
was reversed; and in the USA, these two policies were
perceived as similarly effective (see Supplementary Table
S1 for all statistical coefficients).
Does Level of Household Income Influence Beliefs About
Locus of Responsibility, Liability for Treatment Costs,
and Effectiveness of Policy Measures?
Not only does the magnitude of the four health risks
differ across socioeconomic groups, the four policies
Fig.2. Should aficted individuals have to pay for treatment themselves? Proportions of responses, separately for the four health risks
(alcohol: alcohol dependence, tobacco: tobacco dependence).
ann. behav. med. (2018) XX:1–14 7
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
discussed may affect these groups differently (e.g., higher
taxation). Therefore, we tested how net household
income related to locus of responsibility, liability for
treatment costs, and effectiveness of policy measures. To
this end, we reran all analyses reported above, examining
the influence of three levels of household income in each
country (low, medium, high). The following patterns
emerged (see Supplementary Tables S2–S5 for the results
of all statistical tests): it was only in the USA that attri-
butions of personal responsibility for the four health
risks differed by income level (interaction effect between
the main effect across the four health risks and house-
hold income, F(3, 1713)=5.30, p=.001, η2=0.01). This
effect was driven by differences in attributions of respon-
sibility for obesity versus depression: people with a high
or medium income attributed almost twice as much
responsibility for obesity than for depression to the indi-
vidual (78.8 for obesity vs. 42.6 for depression in the high
income group; 78.8 for obesity vs. 47.0 for depression in
the medium income group); in the low income group,
the difference between the two health risks was much
smaller (73.7 for obesity vs. 50.7 for depression). In con-
trast, we did not find any influence of income level on
attributions of responsibility for any of the four health
risks in the UK or Germany (Supplementary TableS2).
Next, we examined whether income level influenced
participants’ beliefs about individual liability for treat-
ment costs across the health risks, running Cochran’s Q
tests separated by income level. Across all three coun-
tries and income levels, the proportion of participants
who believed that afflicted individuals should pay for
treatment themselves differed across the four health
risks (Supplementary Table S3). Paired comparisons
of obesity with each of the three other health risks
revealed that income level did not drive differences in
the proportion of participants endorsing individual li-
ability for treatment costs for obesity versus alcohol
dependence or obesity versus depression. Across the
three countries and income levels, a larger proportion
of participants endorsed individual liability for treat-
ment costs for tobacco dependence than for obesity,
but the difference in proportions was significant in
only five of the nine comparisons (three income levels
× three countries).
Across the three countries, income level did not affect
the relation between beliefs about individual responsi-
bility for a health risk and endorsements of individual
liability for its treatment costs (Supplementary Table
S4), with one exception: in the USA, for each decrease
in income level (i.e., from high to medium or medium
to low), the odds of endorsing individual liability for
the treatment costs for depression were roughly halved.
Further, income level did not affect the perceived effect-
iveness of the policy measures across the three countries
(Supplementary Table S5).
Table3 Results of Logistic Regression Analyses Predicting How Beliefs About Individual Responsibility for Obesity/Alcohol Dependence/Tobacco Dependence/Depression Relate to
Endorsements of Individual Liability for Treatment Costs
Factors included
Obesity Alcohol dependence Tobacco dependence Depression
B (SE) OR
95% CI
of OR B (SE) OR
95% CI
of OR B (SE) OR
95% CI
of OR B (SE) OR
95% CI
of OR
USA Constant −0.61 (0.64) 0.54 −0.16 (0.64) 0.85 −0.16 (0.68) 0.85 −0.80 (0.65) 0.45
Proportion individual responsibility 0.04 (0.01) 1.04 1.03–1.05 0.04 (0.01) 1.04 1.03–1.05 0.04 (0.01) 1.04 1.03–1.04 0.03 (0.004) 1.03 1.02–1.04
UK Constant −2.13 (0.47) 0.12 −1.20 (0.42) 0.30 −1.00 (0.43) 0.37 −3.27 (0.66) 0.04
Proportion individual responsibility 0.03 (0.003) 1.03 1.03–1.04 0.02 (0.003) 1.02 1.01–1.02 0.02 (0.003) 1.02 1.01–1.02 0.03 (0.004) 1.03 1.02– 1.04
Germany Constant −2.59 (0.64) 0.08 −4.31 (0.64) 0.01 −3.36 (0.66) 0.04 −5.14 (0.99) 0.01
Proportion individual responsibility 0.03 (0.01) 1.03 1.02–1.04 0.04 (0.01) 1.04 1.03–1.05 0.03 (0.01) 1.03 1.02–1.04 0.02 (0.01) 1.02 1.01–1.03
The ORs represent the increase in the odds of endorsing individual liability for treatment costs, per additional point increase (up to a maximum of 100)in the attribution of personal
responsibility. 95% CI 95% condence interval; OR odds ratio; SE standard error.
8 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
Fig.3. Judgments of effectiveness of policies targeting obesity, alcohol dependence, and tobacco dependence (0: no effect; 100: very
strong effect; alcohol: alcohol dependence, tobacco: tobacco dependence). The plot widths represent the density of the raw data distribu-
tions, the bandwidth of each bean is determined by the difference between the smallest and largest density of the raw data per country.
The lines represent the weighted mean. For exact p values, see Table4.
ann. behav. med. (2018) XX:1–14 9
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
Discussion
Statement of Principal Findings
Respondents in the USA, UK, and Germany attributed
responsibility for obesity primarily to the individual. This
pattern of attribution also held for alcohol dependence
and, to an even greater extent, for tobacco dependence.
Thus, in terms of personal responsibility, people placed
obesity closer to alcohol and tobacco dependence than
to a stress-related mental disorder, depression. Likewise,
they placed obesity closer to substance dependencies
in terms of perceived liability for treatment costs, with
similar patterns of findings emerging across the four
health risks in all three countries: respondents’ levels of
endorsement of individual liability for treatment costs
for obesity were similar to those for alcohol dependence,
and also much closer to those for tobacco dependence
than for depression. Furthermore, respondents who
tended to attribute personal responsibility for health
risks also considered the individuals affected to be more
accountable for the costs incurred. Respondents in all
three countries believed intelligible nutrition labeling—
the least intrusive and restrictive measure—to be the
most effective policy (among those considered) for pre-
venting obesity, and taxes to be the least effective policy.
Last but not least, across all three countries, the level of
household income had limited influence on respondents’
beliefs about locus of responsibility, liability for treat-
ment costs, and effectiveness of policy measures.
Strengths, Weaknesses, and Future Research
To our knowledge, this is the first investigation to com-
pare lay theories of obesity and of other major health
risks thought to share certain similarities with obesity.
Further, it is the first study to use the same items to elicit
lay beliefs about major health risks across representa-
tive samples in three countries. The findings identify a
gap between lay and expert beliefs about the causes of
obesity: although there is growing agreement among
experts that the rapid weight gain of the last four decades
Table4 Statistical Difference Values for Answers to the Question “How Effective Is Measure X in Preventing Obesity/Alcohol Dependence/
Tobacco Dependence/Depression?”
Main effect across four health risks Within-subject contrasts
High taxes USA F(2, 1368)=294.14, p < .001, η2=0.30 O–A F(1, 684)=227.95, p < .001, η2=0.25
O–TD F(1, 684)=486.30, p < .001, η2=0.42
UK F(2, 2140)=167.80, p < .001, η2=0.14 O–A F(1, 1070)=69.87, p < .001, η2=0.06
O–T F(1, 1070)=313.31, p < .001, η2=0.23
Germany F(2, 1792)=171.86, p < .001, η2=0.16 O–A F(1, 896)=165.03, p < .001, η2=0.16
O–T F(1, 896)=283.47, p < .001, η2=0.24
Limiting availability or
consumption in public
spaces
USA F(2, 1390)=190.03, p < .001, η2=0.22 O–A F(1, 695)=242.30, p < .001, η2=0.26
O–T F(1, 695)=270.55, p < .001, η2=0.28
UK F(2, 2170)=177.0, p < .001, η2=0.14 O–A F(1, 1085)=219.85, p < .001, η2=0.17
O–T F(1, 1085)=274.25, p < .001, η2=0.20
Germany F(2, 1846)=15.26, p < .001, η2=0.02 O–A F(1, 923)=7.00, p = .008, η2=0.01
O–T F(1, 923)=7.25, p = .007, η2=0.01
Banning or limiting
advertising
USA F(2, 1330)=73.79, p < .001, η2=0.10 O–A F(1, 665)=48.28, p < .001, η2=0.07
O–T F(1, 665)=131.84, p < .001, η2=0.17
UK F(2, 2174)=35.0, p < .001, η2=0.03 O–A F(1, 1087)=0.57, p=.450, η2=0.00
O–T F(1, 1087)=54.46, p < .001, η2=0.05
Germany F(2, 1788)=9.48, p < .001, η2=0.01 O–A F(1, 894)=1.62, p = .203, η2=0.00
O–T F(1, 894)=16.74, p < .001, η2=0.02
Labeling and warnings USA F(2, 1352)=94.85, p < .001, η2=0.12 O–A F(1, 676)=164.93, p < .001, η2=0.20
O–T F(1, 676)=33.88, p < .001, η2=0.05
UK F(2, 2174)=106.0, p < .001, η2=0.09 O–A F(1, 1087)=198.16, p < .001, η2=0.15
O–T F(1, 1087)=59.81, p < .001, η2=0.05
Germany F(2, 1834)=384.96, p < .001, η2=0.30 O–A F(1, 917)=423.12, p < .001, η2=0.32
O–T F(1, 917)=613.71, p < .001, η2=0.40
After Bonferroni corrections, only p values smaller or equal to .001 are considered statistically signicant. O obesity; A alcohol depend-
ence; T tobacco dependence; D depression.
10 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
has been largely driven by the obesogenic modern food
environment [31], lay people in the three countries under
investigation tend to hold the individual responsible. We
also analyzed the impact of one important indicator of
socioeconomic status, namely, household income, across
the three countries. Future research needs to examine
additional indicators of socioeconomic status. Although
participants entered their survey responses into a per-
sonal computer in all three countries, the different survey
modes (computer-assisted face-to-face interviews in the
UK and Germany vs. online surveys in the USA) may
have affected responses (e.g., [32]). However, given the
similarity of responses and response patterns across the
two survey modes (e.g., concerning perceptions about in-
dividual responsibility for the four health risks), we be-
lieve that any impact of the difference in survey modes
is limited. Other potential limitations are that, like any
self-report measure, our surveys are subject to response
bias, and that individual knowledge or attitudes may also
have influenced respondents’ answers. Despite random-
ization of question blocks, moreover, order effects are
possible. Admittedly, our focus on countries with high
obesity rates is also a limitation, but our concern was to
exclude the obesity rate itself as the cause of potentially
divergent public beliefs.
Conclusion
In 2014, more than 1.9 billion adults worldwide were
overweight or obese [1]. The fundamental cause of
obesity is an energy imbalance between calories con-
sumed and calories expended. One of the two key levers
to fight the obesity epidemic is therefore the number of
daily calories consumed. How this can be achieved will
depend substantially on the framing of this health crisis.
If framed as a matter of personal (ir)responsibility, it will
be addressed differently than if framed as a crisis driven
in no small part by other factors (e.g., an obesogenic en-
vironment, corporate misbehavior, lack of government
regulations).
It is important to acknowledge that obesity is brought
about by myriad factors and is likely the result of
an interaction between environment and individual.
Therefore, there is unlikely to be a silver bullet—that is,
a single lever that can be used to contain or even reverse
the obesity epidemic. Helping individuals with obesity
to take responsibility for factors they can control (e.g.,
weight-related behaviors) and not unduly attributing
responsibility to those they cannot control (e.g., envir-
onmental characteristics) could attenuate some of the
guilt, poor self-acceptance, and stigma that people with
obesity experience [33, 34]. That being said, behavioral
interventions on obesity are rarely successful in the long
term [35–37]. Thus, focusing on prevention, particularly
by designing our modern environment to make it less
obesogenic, will likely be a key force in combating the
obesity epidemic.
There were some notable similarities and differences
in views across countries. For instance, US respond-
ents were more likely to endorse individual liability for
treatment costs than were German or UK respondents.
This finding is consistent with a pattern observed by
Branson and colleagues [16], showing the USA to stand
out among wealthier nations as the country least in
favor of government interventions. It is also consistent
with the degree of public funding of the healthcare
system: in 2013, 48% of healthcare costs in the USA
were publicly funded, relative to 83% in the UK and
77% in Germany [38].
Our results show that the US, UK, and German
public strongly believe individuals to be personally re-
sponsible for obesity and, similarly, for tobacco and al-
cohol dependence. Although it is unclear to what extent
the public has adopted the food and soda industries’
framing of the problem [7, 39], this belief has policy
implications. For instance, attribution to individuals is,
as our results show, positively associated with the belief
that individuals should be personally liable for treatment
costs. Furthermore, the public’s emphasis on personal
responsibility may also explain why information (intelli-
gent labeling) is rated to be most effective in preventing
obesity, and taxation to be least effective. The former can
be interpreted as boosting the individual’s competence
to exercise personal responsibility, whereas taxes on un-
healthy food can be understood as a one-size-fits-all pen-
alty that is unfair to those who consume fast food only
as a raretreat.
Yet public opinions change and evolve. In all three
countries, respondents rated high taxes as effective in
reducing tobacco consumption. Over a period of dec-
ades, the US public has transformed from a smok-
ing-tolerant culture to one accepting and supporting
bans on the marketing and consumption of tobacco
(e.g., creating smoke-free public places), as well as high
taxation of tobacco products [31]. Lessons learned in
overcoming opposition to fiscal and regulatory inven-
tions in the context of smoking might help policymak-
ers to raise public support for corresponding measures
addressing obesity [40].
Our results highlight one obesity prevention meas-
ure that already enjoys public support, namely, intelli-
gible food labeling. In Germany and the UK, nutrition
labels have been mandated by EU regulations since
December 2016 [41]. The UK has additionally imple-
mented an improved front-of-pack labeling system [42].
In May 2016, the US Food and Drug Agency (FDA)
launched a new, more comprehensive food label includ-
ing a declaration of added sugars and realistic portion
sizes [43]. Despite this important progress, neither the
EU nor the FDA legislation mandates understandable
ann. behav. med. (2018) XX:1–14 11
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
and user-friendly front-of-package labeling (such as the
traffic light system), the type of labeling that consumers
consult most often [44].
Taxing of unhealthy foods and drinks, such as sug-
ar-sweetened beverages, is still at an early stage. The
World Health Organization (WHO) recently called for
a 20% tax on sugar-sweetened beverages. Berkeley was
the first US city to impose such a tax [45]. The UK gov-
ernment has published draft legislation for a tax on sug-
ar-sweetened drinks to begin in 2018 [46]. Germany is
currently not expected to impose such a tax (e.g., [47]).
First attempts to limit access to sugar-sweetened
beverages and foods high in sugar, salt, or fat have
been made in schools: in 2005, both the UK govern-
ment [48] and California [45] banned vending machines
selling such products. We are not aware of any plans
in Germany to institute a similar ban in public spaces.
Regarding limits or bans on advertising, the UK has
again implemented the strongest and most far-reaching
policies, with advertising of products high in fat, salt,
or sugar being banned from programs aimed at children
aged between 4 and 15years since 2008. In the USA, the
Children’s Food and Beverage Advertising Initiative,
launched in 2007, has issued a list of products that
may be advertised to children. However, in 2014, more
than half of the products on the list exceeded the rec-
ommended limit for saturated fat, trans fat, sugar, and
sodium [49]. To our knowledge, Germany does not re-
strict the content or timing of television advertisements
aimed at children (e.g., [50]).
To summarize, the available public record suggests
that, of the three countries surveyed, the UK has most
forcefully implemented policies to target obesity. In the
USA, a number of policies apply only at the city or
state level; thus, there is considerable variation across
the country. In Germany, comparably little effort seems
to have been made to implement obesity prevention
policies. This pattern mirrors the regulations and pol-
icies implemented to control tobacco consumption:
on the Tobacco Control Scale, the UK ranks as the
country most forcefully implementing tobacco control
policies; Germany ranks 26th (among 31 ranked coun-
tries) [51]. The USA was not ranked on the Tobacco
Control Scale but has implemented a number of regula-
tory measures [21]. Worldwide, countries are only now
beginning to implement policies to curtail and prevent
obesity. The efficacy of many of these policies, as well
as their effects on different population groups, is yet to
be evaluated. Yet effective policies also require public
support. Understanding lay people’s beliefs about what
is possibly the most significant global risk to public
health, and how those beliefs relate to public support
of policy measures, promises to be an important step in
orchestrating individual and collective responses to the
obesity crisis.
Supplementary Material
Supplementary material is available at Annals of
Behavioral Medicine online.
Acknowledgements We are grateful to Susannah Goss, Kate
Pleskac, and Valerie Chase for editing the manuscript. We also
thank Mattea Dallacker, Marianne Hauser, Emelie Letzsch, Rui
Mata, Andrea H. Meyer, Michael Schulte-Mecklenbeck, Petra
Kühner-Knaup, Sarah Otterstetter, and Françoise Weber for
their help with earlier versions of this manuscript, and Nicole
Engelhardt and the library of the Max Planck Institute for Human
Development. No financial disclosures were reported by the
authors of this paper.
Authors’ Statement of Conflict of Interest and Adherence to Ethical
Standards Authors Jutta Mata and Ralph Hertwig declare that
they have no conflict of interest. All procedures, including the
informed consent process, were conducted in accordance with the
ethical standards of the responsible committee on human experi-
mentation (institutional and national) and with the Helsinki
Declaration of 1975, as revised in 2000.
Compliance with Ethical Standards
Authors’ Contributions Jutta Mata (JM) and Ralph Hertwig (RH)
conceived the paper and designed the study; JM coordinated the
data collection and analyzed the data; JM and RH wrote the paper.
Ethical Approval All procedures performed in studies involving
human participants were in accordance with the ethical standards
of the institutional and/or national research committee and with
the 1964 Helsinki declaration and its later amendments or compar-
able ethical standards.
Informed Consent Informed consent was obtained from all indi-
vidual participants included in the study.
References
1. World Health Organization. Fact sheet N°311: obesity and
overweight. World Health Organization Media Centre.
Available at http://www.who.int/mediacentre/factsheets/
fs311/en/. Accessibility verified December 13, 2017.
2. Hill JO, Peters JC. Environmental contributions to the obe-
sity epidemic. Science. 1998;280(5368):1371–1374.
3. Estabrooks PA, Fisher EB, Hayman LL. What is needed to
reverse the trends in childhood obesity? Acall to action. Ann
Behav Med. 2008;36(3):209–216.
4. Sallis JF, Carlson JA, Mignano AM, etal. Trends in presenta-
tions of environmental and policy studies related to physi-
cal activity, nutrition, and obesity at Society of Behavioral
Medicine, 1995–2010: a commentary to accompany the
Active Living Research Supplement to Annals of Behavioral
Medicine. Ann Behav Med. 2013;45(suppl 1):S14–S17.
5. Franck C, Grandi SM, Eisenberg MJ. Taxing junk food to
counter obesity. Am J Public Health. 2013;103(11):1949–1953.
6. Harris JL, Pomeranz JL, Lobstein T, Brownell KD. A crisis
in the marketplace: how food marketing contributes to child-
hood obesity and what can be done. Annu Rev Public Health.
2009;30:211–225.
12 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
7. Brownell KD, Warner KE. The perils of ignoring history: Big
Tobacco played dirty and millions died. How similar is Big
Food? Milbank Q. 2009;87(1):259–294.
8. World Health Organization. Data for saving lives.
WHO Global InfoBase. Available at https://apps.who.int/
infobase/Comparisons.aspx. Accessibility verified December
13, 2017.
9. Marcus MD, Wildes JE. Obesity: is it a mental disorder? Int J
Eat Disord. 2009;42(8):739–753.
10. Gearhardt AN, Corbin WR, Brownell KD. Food addiction:
an examination of the diagnostic criteria for dependence. J
Addict Med. 2009;3(1):1–7.
11. Hammen C. Stress and depression. Annu Rev Clin Psychol.
2005;1:293–319.
12. Adam TC, Epel ES. Stress, eating and the reward system.
Physiol Behav. 2007;91(4):449–458.
13. Gilpin EA, Lee L, Pierce JP. Changes in population attitudes
about where smoking should not be allowed: California ver-
sus the rest of the USA. Tob Control. 2004;13(1):38–44.
14. Mons U, Jazbinsek D, Kahnert S. Smoke-free restaurants
in Germany 2012: majority of smokers in favor of smoking
ban for the first time [in German]. Available at http://www.
dkfz.de/de/tabakkontrolle/download/Publikationen/AdWfP/
AdWfP_Rauchfreie_Gaststaetten_2012.pdf. Accessibility
verified December 13, 2017.
15. Morgan A. Germans (75%) reject extra taxes on unhealthy
foods [in German]. mingle-Trend. Available at http://min-
gle-trend.respondi.com/de/deutsche-75-lehnen-extras-
teuer-auf-ungesunde-lebensmittel-ab/. Accessibility verified
December 13, 2017.
16. Branson C, Duffy B, Perry C, et al. Acceptable behaviour?
Public opinion on behaviour change policy. London: Ipsos
MORI: Social Research Institute. Available at https://
www.ipsos.com/sites/default/files/publication/1970-01/
sri-ipsos-mori-acceptable-behaviour-january-2012.pdf.
Accessibility verified December 13, 2017.
17. Särndal C-E, Swensson B, Wretman J. Model Assisted Survey
Sampling. New York, NY: Springer; 1992.
18. McFerran B, Mukhopadhyay A. Lay theories of obesity
predict actual body mass. Psychol Sci. 2013;24(8):1428–1436.
19. Sikorski C, Luppa M, Schomerus G, Werner P, König HH,
Riedel-Heller SG. Public attitudes towards prevention of
obesity. PLoS One. 2012;7(6):e39325.
20. Room R, Babor T, Rehm J. Alcohol and public health.
Lancet. 2005;365(9458):519–530.
21. World Health Organization. WHO Report on the Global
Tobacco Epidemic, 2011 warning about the dangers of
tobacco. Geneva: World Health Organization. Available at
http://www.who.int/tobacco/global_report/2011/en/index.
html. Accessibility verified December 13, 2017.
22. Hawkes C. Nutrition Labels and Health Claims: The Global
Regulatory Environment. Geneva: World Health Organization;
2004.
23. Thomson LM, Vandenberg B, Fitzgerald JL. An exploratory
study of drinkers views of health infor mation and warning labels
on alcohol containers. Drug Alcohol Rev. 2012;31(2):240–247.
24. Hammond D. Health warning messages on tobacco prod-
ucts: a review. Tob Control. 2011;20(5):327–337.
25. Taber DR, Chriqui JF, Powell LM, Chaloupka FJ. Banning
all sugar-sweetened beverages in middle schools: reduction of
in-school access and purchasing but not overall consumption.
Arch Pediatr Adolesc Med. 2012;166(3):256–262.
26. McClure AC, Tanski SE, Gilbert-Diamond D, et al.
Receptivity to television fast-food restaurant market-
ing and obesity among U.S. youth. Am J Prev Med.
2013;45(5):560–568.
27. Smith LA, Foxcroft DR. The effect of alcohol advertising,
marketing and portrayal on drinking behaviour in young
people: systematic review of prospective cohort studies. BMC
Public Health. 2009;9:51.
28. Harris F. Effects of the 2003 advertising/promotion ban in
the United Kingdom on awareness of tobacco marketing:
findings from the International Tobacco Control (ITC) Four
Country Survey. Tob Control. 2006;15:iii26–iii33.
29. Cohen J. Statistical Power Analysis for the Behavioral
Science. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates;
1988.
30. IBM Corp. IBM SPSS Statistics for Macintosh, Version 24.0.
Armonk, NY: IBM Corp.; 2016.
31. Rozin P. The process of moralization. Psychol Sci.
1999;10(3):218–221.
32. Duffy B, Smith K, Terhanian G, et al. Comparing data
from online and face-to-face surveys. Int J Mark Res.
2005;47(6):615–639.
33. Carr D, Friedman MA. Is obesity stigmatizing? Body
weight, perceived discrimination, and psychological well-
being in the United States. J Health Soc Behav. 2005;
46(3):244–259.
34. Puhl RM, Heuer CA. The stigma of obesity: a review and
update. Obesity (Silver Spring). 2009;17(5):941–964.
35. Fildes A, Charlton J, Rudisill C, Littlejohns P, Prevost AT,
Gulliford MC. Probability of an obese person attaining
normal body weight: cohort study using electronic health
records. Am J Public Health. 2015;105(9):e54–e59.
36. Jeffery RW, Drewnowski A, Epstein LH, et al. Long-term
maintenance of weight loss: current status. Health Psychol.
2000;19(1S):5–16.
37. Wing RR, Phelan S. Long-term weight loss maintenance. Am
J Clin Nutr. 2005;82(1 suppl):222S–225S.
38. OECD. Health at a Glance 2015. OECD Publishing.
Available at http://www.oecd-ilibrary.org/social-issues-migra-
tion-health/health-at-a-glance-2015_health_glance-2015-en.
Accessibility verified December 13, 2017.
39. Nestle M. Soda Politics: Taking on Big Soda (and Winning).
Oxford, UK: Oxford University Press; 2015.
40. Marteau TM, Hollands GJ, Fletcher PC. Changing human
behavior to prevent disease: the importance of targeting
automatic processes. Science. 2012;337(6101):1492–1495.
41. The European Parliament and the Council of the
European Union. Regulation (EU) No 1169/2011 of the
European Parliament and of the Council. Off J Eur Union.
2011;304:18–63.
42. Department of Health, Government of the United
Kingdom. 2010 to 2015 government policy: obe-
sity and healthy eating. Available at https://www.gov.
uk/government/publications/2010-to-2015-govern-
ment-policy-obesity-and-healthy-eating/2010-to-2015-gov-
ernment-policy-obesity-and-healthy-eating. Accessibility
verified December 13, 2017.
43. U.S. Food & Drug Administration. Changes to the Nutrition
Facts Label. U.S. Department of Health and Human Services.
Available at http://www.fda.gov/Food/GuidanceRegulation/
GuidanceDocumentsRegulatoryInformation/
LabelingNutrition/ucm385663.htm. Accessibility verified
December 13, 2017.
44. Grunert KG, Wills JM, Fernández-Celemín L. Nutrition
knowledge, and use and understanding of nutrition informa-
tion on food labels among consumers in the UK. Appetite.
2010;55(2):177–189.
45. World Health Organization. Fiscal policies for diet and
prevention of noncommunicable diseases: technical meet-
ing report, May 5–6, 2015, Geneva, Switzerland. Available
ann. behav. med. (2018) XX:1–14 13
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
at http://apps.who.int/iris/bitstream/10665/250131/1/978
9241511247-eng.pdf. Accessibility verified December 13,
2017.
46. BBC News. UK pushes ahead with sugar tax. Available at
http://www.bbc.com/news/health-38212608. Accessibility
verified December 13, 2017.
47. Süddeutsche Zeitung. What would a sugar tax bring?
[in German]. Available at http://www.sueddeutsche.de/
gesundheit/ernaehrung-was-wuerde-eine-zuckersteuer-
bewirken-1.3206627. Accessibility verified December 13,
2017.
48. Dimbleby H, Vincent J. The School Food Plan. Available at
http://www.schoolfoodplan.com/wp-content/uploads/2013/
07/School_Food_Plan_2013.pdf. Accessibility verified
December 13, 2017.
49. Schermbeck RM, Powell LM. Nutrition recommendations and
the Children’s Food and Beverage Advertising Initiative’s 2014
approved food and beverage product list. Prev Chronic Dis.
2015;12:1–6.
50. Capacci S, Mazzocchi M, Shankar B, etal. Policies to pro-
mote healthy eating in Europe: a structured review of poli-
cies and their effectiveness. Nutr Rev. 2012;70(3):188–200.
51. Joossens L, Raw M. The tobacco control scale 2010 in Europe.
Available at https://www.krebshilfe.de/fileadmin/Downloads/
PDFs/Kampagnen/TCS_2010_Europe.pdf. Accessibility ver-
ified December 13, 2017.
14 ann. behav. med. (2018) XX:1–14
Downloaded from https://academic.oup.com/abm/advance-article-abstract/doi/10.1093/abm/kax003/4823848
by guest
on 12 February 2018
