FramingtheConsequencesofChildhoodObesitytoIncreasePublicSupportforObesityPreventionPolicy1.pdf
Framing the Consequences of Childhood Obesity to Increase Public Support for Obesity Prevention Policy Sarah E. Gollust, PhD, Jeff Niederdeppe, PhD, and Colleen L. Barry, PhD, MPP
Policy reports and news articles often frame the problem of childhood obesity by emphasizing its consequences on health, medical costs, and other nonmedical outcomes. One objective in emphasizing these consequences might be to heighten public and policymaker attention to obesity and engage sectors of the public who would otherwise be uninterested. The 2012 Institute of Medicine report Accelerating Progress in Obesity Prevention: Solving the Weight of the Nation made this objective explicit, stating that
[t]he committee sees the primary catalyst for activation of its system of recommendations as a heightened awareness of the potential cata- strophic consequences of the high rates of obe- sity in the United States for quality of life and the national budget, and a common understanding of the critical role that must be played by individuals in every sector.1(p385)
Descriptions of the consequences of obesity abound in the mass media, policy sphere, and scholarly research. The 2012 HBO documen- tary Weight of the Nation included an episode titled “Consequences,” which described the health, economic, and social consequences of increasing rates of obesity for the United States. The aforementioned Institute of Medicine report included a chapter describing the physical health, psychosocial, and functional conse- quences of obesity over the life course.1 Aca- demic studies provide evidence of childhood obesity’s consequences on longevity,2 medical care costs,3,4 childhood bullying and teasing,5
and American military readiness.6 In fact, a Sep- tember 2012 report by retired military leaders argued for tougher nutrition standards for food sold in schools, citing as motivation data that 25% of17- to 24-year-old Americans are above the weight thresholds for military service.7
Strategically emphasizing this diversity of consequences has potential to expand the perceived seriousness of childhood obesity to segments of the population who may otherwise be disinclined to see the issue as a social problem. Traditional American individualistic values lead to the common view that obesity is
an individual issue, for which personal re- sponsibility—not the government—is the solu- tion.8,9 As a result, Americans assign primary responsibility for reducing childhood obesity to parents rather than government, schools, or the food and beverage industry.10,11 Moreover, re- search shows that people who believe that re- sponsibility to address obesity lies with the food industry are significantly more likely to support obesity-related policies, with no such link for those believing responsibility falls on individ- uals.12 This research suggests that messages that contextualize obesity in broader societal terms— the responsibility of schools, communities, the food and beverage industry, and the govern- ment, not just children and parents—may be effective in increasing public support for, and consequently the political feasibility of, obesity policies such as soda taxes and stronger food marketing regulations that experts consider crit- ical for public health improvement.1,13
Research in social psychology, communica- tion, and political science offers support for the idea that strategically framing a societal
problem can influence judgments of responsi- bility and opinion about policy solutions.14---17
The term framing is used in many different ways,15 but framing in communication is de- fined as selecting some aspect of an issue to make it more salient.17 When issues are framed to activate core values—such as equality or security—these messages are particularly per- suasive because individuals for whom the message resonates are more likely to pay attention to and accept the message, and less likely to resist or counterargue it.18---20 It is likely that the various social, economic, and health consequences of obesity available in the public discourse resonate differently on the basis of individuals’ political predispositions, which serve as a proxy for values.21 Previous research has shown large differences in opinion about obesity policy on the basis of political ideology22 as well as differences in response to messages about obesity by political ideology.12
Although previous research has demonstrated that framing the causes of obesity can influence support for policies,12,22---24 no earlier work has
Objectives. We examined the effects of messages describing consequences of
childhood obesity on public attitudes about obesity prevention policy.
Methods. We collected data from 2 nationally representative Internet-based
surveys. First, respondents (n = 444) evaluated the strength of 11 messages
about obesity’s consequences as reasons for government action. Second, we
randomly assigned respondents (n = 2494) to a control group or to treatment
groups shown messages about obesity consequences. We compared groups’
attitudes toward obesity prevention, stratified by political ideology.
Results. Respondents perceived a message about the health consequences of
childhood obesity as the strongest rationale for government action; messages
about military readiness, bullying, and health care costs were rated particularly
strong by conservatives, moderates, and liberals, respectively. A message
identifying the consequences of obesity on military readiness increased conser-
vatives’ perceptions of seriousness, endorsement of responsibility beyond the
individual, and policy support, compared with a control condition.
Conclusions. The public considers several consequences of childhood obesity
as strong justification for obesity prevention policy. Activating new or un-
expected values in framing a health message could raise the health issue’s
salience for particular subgroups of the public. (Am J Public Health. 2013;103:
e96–e102. doi:10.2105/AJPH.2013.301271)
RESEARCH AND PRACTICE
e96 | Research and Practice | Peer Reviewed | Gollust et al. American Journal of Public Health | November 2013, Vol 103, No. 11
examined whether framing the consequences of obesity affects public opinion.
To fill this gap, we examined the effects of messages describing consequences of childhood obesity on Americans’ attitudes about obesity prevention. We expected that liberals and con- servatives would differ in their judgments about the role of government in obesity prevention. Moreover, we expected that different obesity consequence messages would activate different values, such as protecting children from bully- ing, the fairness of health inequalities, or the importance of strong national defense, and, in turn, that the degree to which these values resonate among the public would vary by po- litical ideology. Thus, we expected that conse- quence frames would have different effects on political subgroups. Our research consisted of 2 sequential studies. First, we examined the public’s evaluations of the strength of 11 mes- sages about obesity consequences as a rationale for government action. Second, using an ex- perimental design, we tested whether 4 highly rated consequence messages affected public perceptions of (1) the seriousness of the prob- lem, (2) responsibility for addressing it, and (3) support for obesity prevention policies.
STUDY 1
The first study was a Web-based public opinion survey fielded in February 2011 by GfK Knowledge Networks (KN), a survey firm that maintains a nationally representative panel of respondents. The KN panel has served as the sampling frame for many studies on public health and medical topics.22,25---27 The firm recruits panel participants through address- based probability sampling, providing Internet access and hardware to those who lack it. It calculates a series of statistical weights to make the resulting sample nationally representative (i.e., by gender, age, race/ethnicity, education, US Census region, metropolitan area, and Internet access), including selection weights that adjust for differential probability of selec- tion into the sample and demographic post- stratification weights that adjust for survey nonresponse and panel attrition.
The survey completion rate (defined as the percentage of KN panel participants who were randomly selected and invited to complete the survey who did so) was 67.9%, resulting in
a final sample of 444 respondents. As mentioned previously, GfK KN uses probability sampling at the first stage of recruitment when individuals are initially approached to participate in the online panel, and the panel recruitment rate was 16.6%. All respondents were aged between 18 and 64 years and diverse across sociodemo- graphic characteristics. (See Appendix A, avail- able as a supplement to the online version of this article at http://www.ajph.org, for descriptive statistics of this sample.)
Survey Design
The goal of the survey was to assess public evaluations of common messages about the consequences of childhood obesity, to under- stand which messages might be effective in persuading the public to consider reducing childhood obesity as a responsibility of the government. We constructed 11 messages culled from an analysis of news media cover- age28 and national reports.29---31 We standard- ized messages in length, structure, and use of a single illustrative statistic (Table 1). Respon- dents were asked to read each message and answer “Is the statement a strong or weak reason for the government doing something about childhood obesity?” where 1 is a “very weak reason” and 7 is a “very strong reason.”
We measured respondents’ political ideology with a standard item asking participants to self- identify as “extremely liberal” (1), “liberal” (2), “slightly liberal” (3), “moderate, middle of the road” (4), “slightly conservative” (5), “conserva- tive” (6), or “extremely conservative” (7). We collapsed this to form 3 groups: liberals (n = 124), moderates (n = 168), and conservatives (n = 145); 7 respondents did not complete this item.
We calculated the proportion of respondents (overall and stratified by political ideology) who indicated that the message was a “strong reason for government doing something” about childhood obesity (answering a 5, 6, or 7 on the scale; results were substantively equivalent when we used the mean). All analyses applied the KN survey weights described previously to produce nationally representative estimates.
Results
Table 1 illustrates the proportion of re- spondents (n = 444) who evaluated each of 11 messages as a strong reason for the gov- ernment to address childhood obesity. The
long-term health consequences message was rated as strong by 60.0% of the full sample, the highest across all messages. Messages about racial/ethnic or socioeconomic disparities and discrimination were rated as the weakest rea- sons for government action, across the political spectrum. Although none of the messages were endorsed by more than 50% of conservatives as being strong reasons for the government to act, the long-term health consequences mes- sage and the message describing the conse- quences of obesity on military readiness were rated highest for this group (42.0% and 41.6% of conservatives, respectively). Moderates judged a message about bullying as the second-strongest (62.1%) after long-term health. More than three fourths (76.4%) of liberals rated the message about health care costs as a strong reason for the government to act—the highest-rated message among liberals.
Comparing conservatives’ and liberals’ strength judgments across messages reveals an expected gulf in beliefs: for all but 2 messages, conservatives were significantly less likely than liberals to indicate that the messages were strong. Liberals and conservatives rated the messages about military readiness and weight- based bullying as equally strong rationales for government action, suggesting that these mes- sages may be the least polarizing. On the basis of these results, we selected the health message (strongest overall), the health care costs mes- sage (strongest for liberals), the bullying mes- sage (strongest for moderates, after health), and military readiness (tied with health as strongest for conservatives) for use in study 2, a ran- domized experiment to examine the causal impact of these messages on public perceptions.
STUDY 2
We fielded a Web-based randomized ex- periment in January---February 2012, also with KN. The survey completion rate was 68.6%, resulting in a final sample of 2494 respon- dents. (See Appendix A, available as a supple- ment to the online version of this article at http://www.ajph.org for descriptive statistics of the sample.)
Experimental Design
To maximize external validity, we embed- ded the health consequences message in all of
RESEARCH AND PRACTICE
November 2013, Vol 103, No. 11 | American Journal of Public Health Gollust et al. | Peer Reviewed | Research and Practice | e97
TABLE 1—Respondents’ Evaluations of Childhood Obesity Consequence Messages as “Strong Reasons for the Government
to Do Something About Childhood Obesity” Stratified by Political Ideology: Web-Based Public Opinion Survey, February 2011
Message
Full Sample (n = 444),
% (95% CI)
Conservatives (n = 145),
% (95% CI)
Moderates (n = 168),
% (95% CI)
Liberals (n = 124),
% (95% CI)
Long-term health: As obese children are more likely to become obese
adults, they are at increased risk for type 2 diabetes, heart disease,
arthritis, and several cancers. Research indicates that about 70%
of obese children are at high risk for heart disease in adulthood.
60.0 (54.4, 65.5) 42.0 (32.4, 51.6) 63.9 (55.4, 72.4) 71.1 (61.4, 80.7)
Mortality: Obesity is expected to cause 112 000 deaths per year in the
United States. Research indicates that the current generation of
children is on track to have shorter lives than their parents because
of increasing rates of obesity.
57.9 (52.3, 63.5) 38.5 (28.9, 48.1) 61.7 (60.1, 80.3) 70.2 (60.0, 80.3)
Health care costs: Childhood obesity substantially increases health care
costs. Research indicates that childhood obesity costs the health care
system $14.3 billion per year, much of which comes from public funds.
57.4 (51.8, 63.0) 36.2 (27.0, 45.4) 58.7 (49.8, 67.6) 76.4 (67.5, 85.2)
Psychological: Obese children are more likely than nonobese children
to feel sad, lonely, and anxious. Research indicates that obese
teenagers are almost 2 times more likely to report symptoms of
depression compared with their nonobese peers.
55.6 (49.9, 61.2) 35.7 (26.2, 45.1) 58.6 (49.7, 67.4) 69.1 (59.1, 79.1)
Bullying: Being obese affects children’s social development because
obese children are more likely to be bullied and teased. Research
indicates that 34% of obese teenage girls have low self-esteem,
compared with only 8% of nonobese teenage girls.
53.9 (48.3, 59.6) 38.3 (28.7, 47.9) 62.1 (53.4, 70.8) 57.7 (46.9, 68.5)
Short-term health: Children who are obese have problems sleeping
and are more prone to certain illnesses. For instance, research
indicates that more than 100 000 American children suffer from
asthma each year because of their excess body weight.
53.7 (48.0, 59.3) 36.8 (27.5, 46.2) 56.6 (47.6, 65.6) 66.7 (56.8, 76.5)
Workforce costs: Because obese children are more likely to become
obese adults, the US workforce may become less productive in the
future. Research indicates that obesity results in about $117 billion
in lost wages and other costs to employers annually.
52.1 (46.5, 58.0) 36.4 (27.1, 45.8) 56.0 (47.1, 64.8) 61.9 (51.2, 72.6)
Military readiness: Young people are increasingly rejected from the
military for being obese, threatening the military’s ability to protect
the nation. Research indicates that 12% of men and 35% of women
of military age are ineligible for duty because of weight problems.
49.8 (44.1, 55.5) 41.6 (31.8, 51.3) 48.5 (39.5, 57.6) 57.4 (46.7, 68.1)
Socioeconomic status disparity: Obesity is more common among children
living in families with lower incomes. Research indicates that rates of
obesity are more than 40% higher among children who are poor
compared with children who are not poor.
49.8 (44.1, 55.5) 30.1 (21.0, 39.2) 54.2 (45.3, 63.2) 63.5 (52.7, 74.3)
Racial/ethnic disparity: Obesity is more common among children from
certain racial and ethnic groups. Research indicates that rates of
obesity are more than 45% higher among Mexican American and
African American children compared with non-Hispanic White children.
48.3 (42.5, 54.0) 31.9 (22.6, 41.2) 51.5 (42.4, 60.6) 58.5 (47.8, 69.1)
Discrimination: Obese children face disadvantages when they grow up.
Research indicates that obese adults are much more likely to be
discriminated against on the job, and they earn up to 6% less
than nonobese employees.
47.7 (42.0, 43.5) 32.1 (22.8, 41.5) 52.8 (43.8, 61.7) 56.5 (45.6, 67.3)
Note. CI = confidence interval. Table entries display proportions of each group who evaluated the messages as “strong” reasons for the government to act as well as the 95% CIs.
RESEARCH AND PRACTICE
e98 | Research and Practice | Peer Reviewed | Gollust et al. American Journal of Public Health | November 2013, Vol 103, No. 11
the consequence messages that respondents read. This design is based on unpublished findings from a news media content analysis of childhood obesity articles.28 Although half (54.0%) of all news stories examined men- tioned any health, psychological, or economic consequences of obesity, 92% of the stories that mentioned any consequences included a mention of the health consequences. We randomly assigned respondents (n = 2494) to 1 of 6 groups reading 6 different messages (Table 2): no information about childhood obesity (group 1), a statement about the mag- nitude of the problem of childhood obesity with no mention of consequences (group 2), a statement about the magnitude of the
problem and its health consequences (group 3), and 3 messages that included the statement about the magnitude of the problem and its health consequences along with 1 of 3 addi- tional nonhealth consequences: health care costs, weight-based bullying, and military readiness (groups 4, 5, and 6, respectively).
Measures
We assessed 3 sets of dependent variables. We measured seriousness of childhood obesity with a single item: “In your opinion, how serious a problem is obesity among children?” (where 1 = “not at all serious” and 7 = “very serious”).
We measured responsibility for addressing childhood obesity by asking, “Please tell me
how much responsibility you believe each of the following should have for addressing the problem of childhood obesity in the United States”: government, food and beverage com- panies, schools, parents, and children (where 1 = “hardly any” and 7 = “a great deal”).
We asked respondents about their support for policies aimed at curbing childhood obesity being considered at the national, local, and state levels, based on a review of legislative databases.32,33 We measured support for 14 policies (where 1 = ”strongly oppose” and 7 = “strongly support”). We conducted an ex- ploratory factor analysis of all policies, yielding a single factor (eigenvalue = 3.26) containing 7 items related to population-level diet and physical activity promotion (the other items related to, for instance, legal protections for the overweight and health insurance incentives, are not examined here). These policies in- cluded levying a tax on sugar-sweetened bev- erages, requiring minimum physical activity requirements in schools, prohibiting junk food marketing in children’s television program- ming, implementing school-based surveillance of students’ body mass indices, prohibiting junk food and soda in school cafeterias, providing incentives for grocery stores to open in areas with low access to healthy food, and prohibiting toys in fast food kids’ meals. (See Appendix B, available as a supplement to the online version of this article at http://www.ajph.org, for item wording.) We measured policy support as a scale (Cronbach a = 0.788) by averaging re- sponses to these 7 policies (models for indi- vidual policies are available from the authors).
The main explanatory variables were the randomly assigned indicator variables repre- senting each of the 5 experimental groups, with the excluded category representing the no- exposure control group. As in the survey, we measured political ideology on a 7-point scale, collapsed to form 3 groups: liberals (n = 682), moderates (n = 876), and conservatives (n = 880); 56 respondents did not complete this item.
Analysis
We confirmed that assignment to groups was random by testing for differences in par- ticipant characteristics across groups with the v2 test and analysis of variance. (See Appendix A, available as a supplement to the online version of this article at http://www.ajph.org,
TABLE 2—Messages Randomly Assigned to Experimental Groups in a Web-Based
Experiment, January–February 2012
Experimental Group (Group No.) Message Viewed
Control (1) No exposure
Magnitude of problem (2) We’ve heard a lot lately about how more children are obese today than in
previous generations. Rates of obesity have more than tripled among children
and adolescents over the past 30 years. a
Long-term health consequences (3) As obese children are more likely to become obese adults, they are at increased
risk for type 2 diabetes, heart disease, arthritis, and several cancers. Research
indicates that about 70% of obese children are at high risk for heart disease
in adulthood.
Long-term health + health care costs
consequences (4)
As obese children are more likely to become obese adults, they are at increased
risk for type 2 diabetes, heart disease, arthritis, and several cancers. Research
indicates that about 70% of obese children are at high risk for heart disease
in adulthood. In addition, childhood obesity substantially increases health care
costs. Childhood obesity costs the health care system $14 billion per year,
much of which comes from public funds.
Long-term health + bullying
consequences (5)
As obese children are more likely to become obese adults, they are at increased
risk for type 2 diabetes, heart disease, arthritis, and several cancers. Research
indicates that about 70% of obese children are at high risk for heart disease
in adulthood. In addition, being obese affects children socially because obese
children are more likely to be bullied. The most common reason high-school
students are bullied is because of their weight, and 84% of students have
observed weight-related bullying or teasing in their school.
Long-term health + military readiness
consequences (6)
As obese children are more likely to become obese adults, they are at increased
risk for type 2 diabetes, heart disease, arthritis, and several cancers. Research
indicates that about 70% of obese children are at high risk for heart disease
in adulthood. In addition, young people are increasingly rejected from the
military for being obese, threatening the military’s ability to protect the nation.
Currently, 12% of men and 35% of women of military age are ineligible for
duty because of weight problems.
aParticipants in groups 3 through 6 read the “magnitude of the problem” message (2) as an introduction before their assigned consequence message.
RESEARCH AND PRACTICE
November 2013, Vol 103, No. 11 | American Journal of Public Health Gollust et al. | Peer Reviewed | Research and Practice | e99
for statistical tests of randomization.) Next we estimated ordinary least squares regressions to examine the effects of each consequence mes- sage on each outcome (perceptions of serious- ness, responsibility for solving the problem, and policy support), stratified by participants’ political ideology; results were substantively equivalent when estimated by using ordered logit models for the seriousness and responsi- bility outcomes.
Because of the multiple tests performed in the stratified regression analyses with multiple outcomes, we focused our attention on the results on statistically significant (P < .05) find- ings that evidenced consistency beyond that which would be expected on the basis of chance. Following established conventions for
survey-based experiments,34 we did not in- clude any covariates as control variables. Be- cause of random assignment to treatment group, covariates are not necessary to produce valid estimates of the experimental effect; in- clusion of covariates did not change message effects (results examining relationships be- tween demographic covariates and key out- comes are available from authors). All analyses applied the KN weights to produce nationally representative estimates.
Results
Table 3 displays the effects of the conse- quence messages on public perceptions of the seriousness of obesity. Compared with a no-exposure control condition, each
consequence message (long-term health, health care costs, childhood bullying, and military readiness) significantly increased participants’ perceptions of the seriousness of childhood obesity, as did a message that only described the problem’s magnitude.
Stratification by political ideology revealed that all of the messages significantly increased conservatives’ judgments of the problem’s se- riousness compared with the control condition, whereas none of the messages significantly increased moderates’ or liberals’ perceptions of the problem’s seriousness compared with the control. Whereas conservatives in the control group rated childhood obesity a 5.19 on the 7-point seriousness scale (see intercept coeffi- cient, conservative column) versus liberals’ 5.87 rating in the control group, the messages boosted conservatives’ perceptions of serious- ness by between 0.49 and 0.80 points depending on the message, making their per- ceptions nearly indistinguishable from those of moderates and liberals.
Table 4 presents a summary of the results from 6 regression models reporting the effects of consequence messages on attitudes about responsibility for solving the problem of child- hood obesity (models 1---5) and on support for obesity prevention policy (model 6). Full re- gression model results are available in Appen- dix C (available as a supplement to the online version of this article at http://www.ajph.org). The table displays only the statistically signifi- cant (P < .05) effects of the consequence mes- sages (compared with the control group) on the outcomes, stratified by political ideology. Results reveal a consistent pattern related to the effects of the military readiness message:
TABLE 3—Effects of Consequence Messages on Perceptions of Seriousness of Childhood Obesity Stratified by Political Ideology:
Web-Based Randomized Experiment, January–February 2012
Message Full Sample, Coeff (SE) Liberals, Coeff (SE) Moderates, Coeff (SE) Conservatives, Coeff (SE)
Magnitude only (Ref = no exposure) 0.28* (0.12) 0.09 (0.19) 0.23 (0.22) 0.49** (0.18)
Health consequences (Ref = no exposure) 0.34* (0.13) 0.32 (0.18) 0.24 (0.26) 0.43* (0.19)
Health + health care costs (Ref = no exposure) 0.35** (0.12) 0.13 (0.17) 0.35 (0.20) 0.49* (0.19)
Health + bullying (Ref = no exposure) 0.41** (0.12) 0.18 (0.19) 0.29 (0.22) 0.80*** (0.18)
Health + military readiness (Ref = no exposure) 0.34** (0.12) 0.27 (0.19) 0.18 (0.22) 0.49** (0.19)
Intercept 5.50*** (0.09) 5.87*** (0.14) 5.54*** (0.17) 5.19*** (0.14)
No. 2481 678 872 877
*P < .05; **P < .01; ***P < .001.
TABLE 4—Significant Effects of Consequence Messages on Attitudes About
Responsibility for Solving the Problem of Childhood Obesity and Support for Obesity
Prevention Policy Stratified by Political Ideology: Web-Based Randomized Experiment,
January–February 2012
Outcome Liberals Moderates Conservatives
Model 1: responsibility of government Health care costs**(–) None Military readiness*(+)
Model 2: responsibility of food and
beverage companies
None None Military readiness**(+)
Model 3: responsibility of schools None Military*(+) Military readiness*(+)
Model 4: responsibility of parents None None None
Model 5: responsibility of children Health*(+) None Magnitude**(+)
Health*(+)
Military readiness*(+)
Model 6: obesity prevention policy support None None Military readiness*(+)
Note. Table entries display the consequence message treatment group that is significantly different from the control group for each outcome, the direction of effect (in parentheses), and the P value of the coefficient. Full regression model results are available in Appendix C (available as a supplement to the online version of this article at http://www.ajph.org). *P < .05; **P < .01.
RESEARCH AND PRACTICE
e100 | Research and Practice | Peer Reviewed | Gollust et al. American Journal of Public Health | November 2013, Vol 103, No. 11
this message significantly increased conserva- tives’ attributions of government, food and beverage company, and school responsibility for addressing childhood obesity, and in- creased conservatives’ support for obesity pre- vention policies, compared with the control condition. However, this message also signifi- cantly increased conservatives’ beliefs that children themselves are responsible for solving the problem, compared with the control con- dition, as did messages about the magnitude of the problem and health consequences.
DISCUSSION
Our findings support conventional wisdom that framing obesity in terms of its multiple consequences can broaden attention to obesity as a serious societal problem warranting gov- ernment action. Survey findings demonstrated that a majority of respondents considered the long-term health consequences of childhood obesity a strong justification for government action. As anticipated, conservatives were less likely than liberals to support rationales for government involvement, consistent with a worldview that endorses limited government intervention into individual behavioral mat- ters.8 The survey also revealed relatively high evaluations of the strength of messages about obesity’s non---health-related consequences, including health care costs, bullying, and mili- tary readiness. By contrast, respondents eval- uated messages about disparities in obesity rates as weak rationales for government action.
Although the survey results showed that liberals and moderates were generally more supportive of a government role in obesity, the randomized experiment revealed that mes- sages about consequences did not usually shift their opinions. This is likely a result of ceiling effects; given liberals’ and moderates’ already strong perceptions of the problem’s seriousness and of the desirability of multisector responsi- bility, messages had no effects. By contrast, several messages significantly increased conser- vatives’ perceptions of the issue’s seriousness.
The message describing the consequences of childhood obesity upon military readiness was particularly promising for garnering the atten- tion of conservatives. This message increased conservatives’ perceptions that nonindividual actors (the government, food and beverage
companies, and schools) bear responsibility for addressing obesity, and increased their support for policy action. This linkage between obesity and military readiness has attracted high-level policy attention. For instance, the preface of the 2010 White House Task Force on Childhood Obesity Report included a statement by a de- fense department official: “We have an obesity crisis in the country. . . . These are the same young people we depend on to serve in times of need and ultimately protect this nation.”31(p3)
Our data support the wisdom of such commu- nication efforts, suggesting that a message link- ing a problem traditionally considered under the domain of public health to national defense has the potential to shift public opinion among conservatives. This message was likely effective because of its novelty, and also because it tapped into values beyond those—such as equality and social responsibility—that are typically associated with public health.35 However, re- framing public health issues in terms of other domains and activating alternative values could also raise the possibility of backlash among subgroups who do not find the link with the new domain credible or appropriate (note that lib- erals ranked the message about military readi- ness as among the weakest).36
Our findings underscore some challenges and potential negative outcomes of framing the consequences of obesity. First, although the military readiness frame did increase conser- vatives’ assignment of responsibility to external actors, there was no corresponding decrease in their attribution of responsibility to parents. High levels of assignment of responsibility to parents—across the political spectrum10—ap- pear difficult to shift.
Second, several messages increased respon- dents’ attribution of responsibility for address- ing obesity to obese children themselves. Obese children are already the targets of stigmatization, which leads to negative health and psychosocial outcomes such as poor self- esteem, unhealthy weight control behaviors, and poor mental health.37 Health promotion campaigns should be cautious that any mes- sages about excessive weight in children do not increase the levels of stigma or blame Ameri- cans hold toward this stigmatized group.38-40
Heightened blame might cause negative out- comes for obese children and could also be counterproductive to expanding public support
for policy actions if such messages increase an individualized focus on the problem.41
Pretesting health messages for such unintended consequences should become standard for public health campaigns because, as the 2012 Institute of Medicine report articulated, “the case for addressing the obesity epidemic cannot be made at the expense of obese people.”1(p104)
Limitations
These findings must be considered in light of some study limitations. First, although use of a population-based sample and messages con- structed from actual media and policy docu- ments extends the study’s external validity, the setting of the Internet-based survey is dissimilar from the real-life context in which people are exposed to or select information about obesity within a competitive media environment.34,42
Second, Internet surveys have been criticized for issues of incomplete coverage and selec- tion.43 Knowledge Networks endeavors to min- imize these issues by recruiting probability- based samples and providing Internet access to those who lack it. In addition, it is unlikely, although not impossible, that the final sample was biased toward people with a particular in- terest in obesity, as the survey invitation did not indicate the study topic.
Third, our study was limited to people younger than 65 years, but senior citizens are an important political constituency whose views on obesity policy should be explored in future work. Finally, other outcomes not ex- amined here might have elucidated mecha- nisms of message effects. For instance, arousal of anger is an important indicator of resistance to communication,44 but we are unable to determine here whether any of the messages we examined had the unintended effects of inducing anger among respondents, which may compromise message effectiveness.
Conclusions
We examined the effects of 4 messages about the consequences of childhood obesity, revealing that a message describing the conse- quences of obesity on military recruits at- tracted the interest of political conservatives. Health researchers continue to advance evi- dence on the social, economic, and psychological consequences of childhood obesity, so it is likely that new messages will emerge in the research
RESEARCH AND PRACTICE
November 2013, Vol 103, No. 11 | American Journal of Public Health Gollust et al. | Peer Reviewed | Research and Practice | e101
literature and be transmitted in the media and policy arenas. In parallel, researchers in political and social psychology are amassing new data on the values that underlie different worldviews in an increasingly polarized citizenry.45 Future research should continue to explore the effects— both intended and unintended—of public health messages, especially those linking public health problems to novel domains and activating diverse values. Such information can support the design of evidence-based communication campaigns, one critical step toward attaining heightened public and policymaker attention to, and ulti- mately actions for, improving public health.10 j
About the Authors Sarah E. Gollust is with the Division of Health Policy and Management, the University of Minnesota School of Public Health, Minneapolis. Jeff Niederdeppe is with the Depart- ment of Communication, Cornell University, Ithaca, NY. Colleen L. Barry is with the Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD. Correspondence should be sent to Sarah Gollust, Division
of Health Policy and Management, University of Minnesota School of Public Health, 420 Delaware St SE, MC 729, Minneapolis, MN 55455 (e-mail: [email protected]). Re- prints can be ordered at http://www.ajph.org by clicking the “Reprints” link. This article was accepted January 29, 2013.
Contributors All authors conceptualized and designed the study. S. E. Gollust conducted the data analysis and drafted the article. C. L. Barry and J. Niederdeppe provided critical revisions to the article, contributed to the data interpre- tation, and approved the final article.
Acknowledgments The authors gratefully acknowledge funding from the Robert Wood Johnson Foundation Healthy Eating Re- search Program (grants 69173 and 68051).
Human Participant Protection The research was determined to be exempt under 45 CFR 46.101 (b) Category 2 by the Johns Hopkins Bloom- berg School of Public Health institutional review board.
References 1. Institute of Medicine. Accelerating Progress in Obesity Prevention: Solving the Weight of the Nation. Washington, DC: National Academies Press; 2012.
2. Olshansky SJ, Passaro DJ, Hershow RC, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med. 2005;352(11):1138---1145.
3. Cawley J. The economics of childhood obesity. Health Aff (Millwood). 2010;29(3):364---371.
4. Cawley J, Meyerhoefer C. The medical care costs of obesity: an instrumental variables approach. J Health Econ. 2012;31(1):219---230.
5. Eisenberg ME, Neumark-Sztainer D, Story M. Associations of weight-based teasing and emotional
well-being among adolescents. Arch Pediatr Adolesc Med. 2003;157(8):733---738.
6. Cawley J, Maclean JC. Unfit for service: the impli- cations of rising obesity for US military recruitment. Health Econ. 2012;21(11):1348---1366.
7. Mission: Readiness. Still too fat to fight. Available at: http://www.rwjf.org/content/dam/farm/reports/reports/ 2012/rwjf401381. Accessed November 7, 2012.
8. Kersh R. The politics of obesity: a current assess- ment and look ahead. Milbank Q. 2009;87(1):295---316.
9. Brownell KD, Kersh R, Ludwig DS, et al. Personal responsibility and obesity: a constructive approach to a controversial issue. Health Aff (Millwood). 2010;29(3): 379---387.
10. Barry CL, Gollust SE, Niederdeppe J. Are Americans ready to solve the weight of the nation? N Engl J Med. 2012;367(5):389---391.
11. Evans WD, Finkelstein E, Kamerow D, Renaud J. Public perceptions of childhood obesity. Am J Prev Med. 2005;28(1):26---32.
12. Niederdeppe J, Shapiro MA, Porticella N. Attribu- tions of responsibility for obesity: narrative communica- tion reduces reactive counterarguing among liberals. Hum Commun Res. 2011;37(3):295---323.
13. Dorfman L, Wallack L. Moving nutrition upstream: the case for reframing obesity. J Nutr Educ Behav. 2007;39(2 Suppl):S45---S50.
14. Iyengar S. Is Anyone Responsible? Chicago, IL: University of Chicago Press; 1991.
15. Chong D, Druckman JN. Framing theory. Annu Rev Polit Sci. 2007;10:103---126.
16. Scheufele D. Framing as a theory of media effects. J Commun. 1999;49(1):103---122.
17. Entman R. Framing: toward clarification of a frac- tured paradigm. J Commun. 1993;43(4):51---58.
18. Zaller JR. The Nature and Origins of Mass Opinion. Cambridge, UK: Cambridge University Press; 1992.
19. Nelson TE, Garst J. Values-based political messages and persuasion: relationships among speaker, recipient, and evoked values. Polit Psychol. 2005;26(4):489---515.
20. Taber CS, Lodge M. Motivated skepticism in the evaluation of political beliefs. Am J Pol Sci. 2006;50 (3):755---769.
21. Gollust SE, Lantz PM, Ubel PA. The polarizing effect of news media messages about the social determinants of health. Am J Public Health. 2009;99(12):2160---2167.
22. Barry CL, Brescoll VL, Brownell KD, Schlesinger M. Obesity metaphors: how beliefs about the causes of obesity affect support for public policy. Milbank Q. 2009;87(1):7---47.
23. Major LH. Break it to me harshly: the effects of intersecting news frames in lung cancer and obesity coverage. J Health Commun. 2009;14(2):174---188.
24. Oliver JE, Lee T. Public opinion and the politics of obesity in America. J Health Polit Policy Law. 2005;30 (5):923---954.
25. Davis MM, Fant K. Coverage of vaccines in private health plans: what does the public prefer? Health Aff (Millwood). 2005;24(3):770---779.
26. Tarini BA, Singer D, Clark SJ, Davis MM. Parents’ concern about their own and their children’s genetic disease risk: potential effects of family history vs genetic test results. Arch Pediatr Adolesc Med. 2008;162(11): 1079---1083.
27. Rothman EF, Edwards EM, Heeren T, Hingson RW. Adverse childhood experiences predict earlier age of drinking onset: results from a representative US sample of current or former drinkers. Pediatrics. 2008;122(2): e298---e304.
28. Barry CL, Jarlenski M, Grob R, Schlesinger M, Gollust SE. News media framing of childhood obesity in the United States from 2000 to 2009. Pediatrics. 2011; 128(1):132---145.
29. Institute of Medicine. Preventing Childhood Obesity: Health in the Balance. Washington, DC: National Acade- mies Press; 2004.
30. Institute of Medicine. Food Marketing to Children and Youth: Threat or Opportunity? Washington, DC: National Academies Press; 2006.
31. White House Task Force on Childhood Obesity. Report to the President. Solving the Problem of Childhood Obesity Within a Generation. Washington, DC: Office of the President of the United States; 2010.
32. Winterfield A, Shinkle D, Morandi L. State Actions to Promote Healthy Communities and Prevent Childhood Obesity: Summary and Analysis of Trends in Legislation. Denver, CO: National Conference of State Legislatures; 2012.
33. Rudd Center for Food Policy & Obesity. Legislation database. 2012. Available at: http://www.yaleruddcenter. org/legislation/search.aspx. Accessed November 7, 2012.
34. Mutz D. Population-Based Survey Experiments. Princeton, NJ: Princeton University Press; 2011.
35. Wallack L, Lawrence R. Talking about public health: developing America’s “second language.” Am J Public Health. 2005;95(4):567---570.
36. Myers TA, Nisbet MC, Maibach EW, Leiserowitz AA. A public health frame arouses hopeful emotions about climate change. Clim Change. 2012;113(3-4): 1105---1112.
37. Puhl RM, Latner JD. Stigma, obesity, and the health of the nation’s children. Psychol Bull. 2007;133(4):557---580.
38. Byrne S, Niederdeppe J. Unintended consequences of obesity prevention messages. In: Cawley J, ed. The Oxford Handbook of the Social Science of Obesity. New York, NY: Oxford University Press; 2011.
39. Barry CL, Gollust SE, McGinty E, Niederdeppe J. Effects of messages from a media campaign to increase public awareness of childhood obesity. Obesity. Epub ahead of print July 9, 2013.
40. Puhl R, Luedicke J, Peterson JL. Public reactions to obesity-related health campaigns: a randomized con- trolled trial. Am J Prev Med. 2013;45(1):36---48.
41. Barry CL, Brescoll V, Gollust SE. Framing childhood obesity: how individualizing the problem affects public support for prevention. Political Psychol. 2013; 34 (3):327---349.
42. Chong J, Druckman JN. A theory of framing and opinion formation in competitive elite environments. J Commun. 2007;57(1):99---118.
43. Couper M. Web surveys: a review of issues and approaches. Public Opin Q. 2000;64(4):464---494.
44. Dillard JP, Shen LJ. On the nature of reactance and its role in persuasive health communication. Commun Monogr. 2005;72(2):144---168.
45. Haidt J. The Righteous Mind: Why Good People Are Divided by Politics and Religion. New York, NY: Pantheon; 2012.
RESEARCH AND PRACTICE
e102 | Research and Practice | Peer Reviewed | Gollust et al. American Journal of Public Health | November 2013, Vol 103, No. 11
Copyright of American Journal of Public Health is the property of American Public Health Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
