Effects of parent-only childhood obesity prevention programs on BMIz and body image in rural preteens

Body Image 16 (2016) 143–153

Contents lists available at ScienceDirect

Body Image journal homepage: www.elsevier.com/locate/bodyimage

Effects of parent-only childhood obesity prevention programs on BMIz and body image in rural preteens Galen Eldridge a , Lynn Paul b , Sandra J. Bailey b , Carrie Benke Ashe a , Jill Martz a,1 , Wesley Lynch c,∗ a

4-Health Project, Montana State University Extension, Bozeman, MT 59717, USA Department of Health and Human Development, Montana State University, Bozeman, MT 59717, USA c Department of Psychology, Montana State University, Bozeman, MT 59717, USA b

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 17 October 2014 Received in revised form 13 December 2015 Accepted 29 December 2015 Available online 4 February 2016 Keywords: Body image Body esteem Body size perception Overweight dissatisfaction Obesity prevention Rural preteens

This experiment compared body image (BI) and BMI changes resulting from two parent-only obesity prevention interventions aimed at 8–12 year olds. Parents in the experimental intervention attended ten face-to-face educational sessions, while parents in the minimal (control) intervention received similar mailed information. Parent-child dyads (N = 150) were semi-randomly assigned to intervention groups. Children were assessed before, after, and 6 months following the interventions; children did not attend experimental intervention sessions. Child BI assessments included weight and size perception, weight management goals, body esteem, and appearance attitudes. Significant effects included small decreases in BMIz scores and overweight dissatisfaction, as well as improvements in aspects of body esteem and appearance attitudes. Some BI effects were gender-specific. Decreases in overweight dissatisfaction were greater following the experimental treatment. Neither treatment reduced body size misperception. Thus, parent-only obesity prevention interventions can reduce body weight and body image concerns among rural preteens. © 2016 Elsevier Ltd. All rights reserved.

Introduction The prevalence of poor body image among both girls and boys is well established. Studies of 7–12 year olds found that 50–80% of girls and 35–55% of boys are dissatisfied with their bodies (Goncalves, Silva, Gomes, & Machado, 2012; Parkinson, Drewett, Le Couteur, Adamson, & Gateshead Millennium, 2012; Rolland, Farnill, & Griffiths, 1996; Wood, Becker, & Thompson, 1996). Recognizing that child body image is considered a multidimensional construct having behavioral, cognitive, emotional, and cultural aspects, we operationally defined “body image” in terms of variables derived from four previously validated assessment instruments (see “Body image measures” below for details). Among these, we defined the emotional aspect referred to as “body dissatisfaction” in terms

∗ Corresponding author. Tel.: +1 406 994 3801/406 585 8622; fax: +1 406 994 3804. E-mail addresses: [email protected] (G. Eldridge), [email protected] (L. Paul), [email protected] (S.J. Bailey), [email protected] (C.B. Ashe), [email protected], [email protected] (J. Martz), [email protected] (W. Lynch). 1 Current address: Department of Recreation, Park and Tourism Sciences, Texas A&M University, College Station, TX, USA. http://dx.doi.org/10.1016/j.bodyim.2015.12.003 1740-1445/© 2016 Elsevier Ltd. All rights reserved.

of the numerical difference in figure ratings between perceived and ideal body sizes based on the Truby and Paxton (2002) Child Body Image Scale. Previous studies have found that body dissatisfaction is a risk factor not only for disordered eating behaviors and clinically diagnosable eating disorders but also for depression, lower health-related quality of life, lower psychosocial functioning, and suicidal ideation and behaviors (Chung & Joung, 2012; Stice, Hayward, Cameron, Killen, & Taylor, 2000; Wilson, Latner, & Hayashi, 2013). Many programs have been created to try to prevent or reverse the development of body dissatisfaction in children and adolescents. The optimal age to participate in programs designed to promote positive body esteem is unclear and is likely to vary between countries, cultures, and possibly even specific communities (Fenton, Brooks, Spencer, & Morgan, 2010). Some researchers suggest that media literacy should begin in childhood before cultural ideals about appearance are internalized (Tiggemann, 2002). Others posit that young adolescents may possess inadequate insight, given that their abstract reasoning skills are still developing, which may limit their ability to benefit from these sorts of programs; these researchers suggest that prevention programs may be most effective around ages 15–16 (Stice, Shaw, & Marti, 2007).

144

G. Eldridge et al. / Body Image 16 (2016) 143–153

By the time children reach their teens, they may already have body image issues, which may be hard to reverse (Skemp-Arlt, 2006). Therefore, targeting preteens by giving them tools to maintain a positive body image before body dissatisfaction and changes in eating habits have begun may be the most successful approach (Skemp-Arlt, 2006). Parents can help children and adolescents maintain a healthy body image by modeling healthy and active lifestyles, rejecting the cultural messages around body shape perfection, emulating a positive body image, maintaining awareness toward body shape discrimination, acknowledging that a person’s physical appearance says very little about their character and value, and talking about these issues with their children (Skemp-Arlt, 2006). Prevention programs for obesity and eating disorders/body image issues have tended to adopt very different approaches to eating behavior and body weight (Neumark-Sztainer, 2003). For example, obesity prevention programs might emphasize dieting and weight loss, while eating disorder and body dissatisfaction prevention programs emphasize cessation of dieting and acceptance of one’s current weight (Neumark-Sztainer, 2003). Obesity prevention programs may inadvertently serve to heighten awareness of body weight and increase body dissatisfaction among children and adolescents (O’Dea, 2004). Therefore, it seems prudent to include promotion of positive body image in obesity prevention programs, by employing an integrated approach to prevention, balancing the importance of healthy lifestyle habits, including a healthy, balanced diet with lifelong physical activity, and acceptance of the human body, including its height, weight, and shape (O’Dea, 2004). Since parents play a large part in shaping the environment of preteens and body image is a critical issue in adolescent development, some healthy lifestyle programs for non-clinical child populations in this age group have included parents in familybased, home-based, or mother and daughter-based interventions (Niemeier, Hektner, & Enger, 2012). Body image is sometimes included as part of the curricula, which usually concentrate on nutrition and physical activity (Debar et al., 2012; Harter, 2012; Hystad, Steinsbekk, Odegard, Wichstrom, & Gudbrandsen, 2013; Olvera, Leung, Kellam, Smith, & Liu, 2013; Pinard et al., 2012; Ransdell, Detling, Taylor, Reel, & Shultz, 2004). Most studies including body image have not reported on preteen body image changes after the intervention (Debar et al., 2012; Harter, 2012; Hystad et al., 2013; Olvera et al., 2013). One family-based program, which did not include a control group, found that preteen body dissatisfaction significantly improved from baseline to 1 month, but not baseline to post-program at 3 months (Pinard et al., 2012). Another study of mothers and daughters, which also did not include a control group, found an increase in girls’ perceived body attractiveness after the 12-week program (Ransdell et al., 2004). Another less common type of program consists of parent-only classes. Parent-only programs can be effective in terms of decreased cost and healthy weight management for overweight youth (Golan & Crow, 2004; Golan, Kaufman, & Shahar, 2006; Golan, Weizman, Apter, & Fainaru, 1998; Golley, Magarey, Baur, Steinbeck, & Daniels, 2007; Janicke et al., 2008; Janicke et al., 2009; Munsch et al., 2008; Shelton et al., 2007). To our knowledge, only three child obesity prevention or management studies that included parent-only groups have assessed child body image variables and none of these addressed body image as part of the intervention (Beech et al., 2003; Estabrooks et al., 2009; Jansen, Mulkens, & Jansen, 2011). Jansen and colleagues found a main effect of time (pre-, post-, and 3-month follow-up of the 10-week program) for changes in preteen shape and weight concerns; the 7–13 year olds in both the parent-only intervention program and the wait-list control group significantly decreased their shape and weight concerns (Jansen et al., 2011). Another weight management study that included a parent-only intervention group found that after the 12-week program, girls in

the global self-esteem control group had significantly more concern about weight and shape than either the girl-only or parent-only intervention groups, but there was no difference between groups in perceived or ideal body size post-intervention (Beech et al., 2003). In a third study, with three parent-only interventions, baseline scores were in the healthy range on a survey about weight dissatisfaction, body dissatisfaction, and unhealthy eating behaviors and there was no increase in scores after the program (Estabrooks et al., 2009). Although body image issues were assessed, none of these parent-only healthy lifestyle programs specifically addressed body image issues as part of the curricula. One study with parent-only and parent-child groups addressed body concept in the curriculum but did not measure changes in the 8–12 year old children’s body image (Munsch et al., 2008). One mother-only intervention program aimed specifically to discourage the development of body-related problems, rather than obesity prevention or management, among middle school girls (Corning, Gondoli, Bucchianeri, & Salafia, 2010). Results were mixed, with significantly lower body dissatisfaction for the intervention group compared to the wait-list control group after the 4-week program according to one body dissatisfaction measure, but no difference in scores for another body dissatisfaction measure (Corning et al., 2010; Garner, Olmstead, & Polivy, 1983). As this brief overview suggests, there has been little research on whether integration of body image into a parent-centered childhood obesity prevention curriculum can positively influence child body image. To our knowledge the present study is the first to integrate body image topics into training and to assess changes in body image in a parent-only child obesity prevention or weight management program. Thus, the present study examined preteen body image-related results of the 4-Health Educational Program, a research and outreach project that aimed to promote healthy lifestyles and prevent childhood obesity development among rural Montana families. The specific goal of this experimental study was to evaluate the efficacy of a parent-only face-to-face (experimental) healthy lifestyles intervention to improve child body image, by assessing changes in BMI and several body image variables, as a function of gender and time of assessment, by comparing children in the Experimental group with those in a mailed information minimal intervention (Control) group. Three main research questions were evaluated: (1) Do measures of body image improve significantly among 8–12 year old children whose parents participate in an 8-month, 10-session face-to-face child obesity prevention program addressing nutrition, physical activity, parenting, and body image? (2) Do measures of body image improve significantly among 8–12 year old children whose parents receive only mailed information (10 packets over 8 months) about nutrition physical activity, parenting, and body image? (3) Is there a significant difference in body image improvements of children whose parents participate in the above two treatment conditions? Method The background, rationale, and study design have been described in detail previously (Lynch et al., 2012). The following is a brief summary of methods relevant to the present report. Research Design USDA County Extension Agents (Agents), with a background in Family and Consumer Sciences and the 4-H Youth Development Program, delivered the face-to-face intervention during 10 ninety-minute face-to-face meetings, attended by parents only, over an 8-month period between late September and late April. In the following pages, we refer to this group as the Experimental

G. Eldridge et al. / Body Image 16 (2016) 143–153

group. The minimal intervention Control group consisted of parents and children who received printed materials mailed to their homes approximately every 3 weeks over the same 8 months. An Extension Food and Nutrition Specialist and the 4-Health Program Project Director trained Agents to deliver the face-to-face intervention during a day-long training session held in August. Only Agents delivering the Experimental intervention attended the training. Recruitment materials describing either the Experimental or Control intervention were provided to each Agent depending on their treatment group assignment. Although Agents were aware that the study involved two treatment conditions, they only received information relevant to one. The research design was a 2 (Experimental vs. Control) × 3 (preintervention, post-intervention, 6-month follow-up) mixed design. Each of the Experimental intervention training modules aimed to promote healthy lifestyles among the rural Montana participants and their families by addressing specific topics and objectives (see below). Based on the advice of expert consultants, the minimal intervention (Control) was devised as a “current best-practices” control that offered participants written information derived from USDA and Academy of Nutrition and Dietetics sources (see below). Thus, both treatment groups received an intervention. A semi-random cluster design was dictated by several constraints resulting from the widely dispersed participant sample and applied nature of the study. Because Extension Agents have other responsibilities, their agreement to participate in the study had to be obtained before assignment to treatment groups. Thus, Agents (representing groups of parents of children currently participating in 4-H) were first recruited from all regions of Montana and their preferences for facilitating the Experimental or Control intervention were determined. Again, because their participation was voluntary, Extension Agents could not be assigned randomly to treatments but instead were assigned to treatment conditions based both on their treatment group preference, if any, and their geographic location, such that pairs of adjacent counties with similar participant demographics were assigned to each treatment condition. Once assigned to treatment groups, Agents recruited parent participants from within the 4-H clubs in their county or an adjacent county, with the aim of recruiting 7–10 parent-child dyads who met participation criteria. Agents were asked not to disclose the existence of treatment conditions other than the one to which they and their participants were assigned. Inclusion Criteria To be eligible for participation, parents and children in both groups agreed to participate in assessment sessions immediately before (Pre) and after the 8-month intervention (Post) as well as 6 months following the intervention (6-month follow-up). The children needed to be between 8 and 12 years old at the preintervention session and taking part in the 4-H Youth Development Program. Written informed consent for participation in the study was obtained from each parent, who also provided consent for her or his child. Children indicated their assent on the same form. Neither parent nor child body weight was considered in determining eligibility. All components of this project were reviewed and approved by the Montana State University Institutional Review Board for the Protection of Human Subjects. Participants Participants were recruited from 21 Montana rural counties. The participant sample was drawn exclusively from families who had at least one child enrolled in the in Montana’s 4-H Youth Development Program. Adult participants recruited into both the Experimental and Control groups were required to be the parent or legal

145

Table 1 Demographic characteristics of participants who completed the follow-up assessment (N = 104). Demographic variable Parent gender Female Child gender Female Parent ethnicity White or Caucasian American Indian or Alaskan Native Other Child ethnicity White or Caucasian American Indian or Alaskan Native Other Annual household (family) income Less than $14,999 $15,000 to $24,999 $25,000 to $34,999 $35,000 to $49,999 $50,000 to $74,999 $75,000 to $99,999 $100,000 or more

% 96.2 70.5 98.1 1.0 1.0 97.1 1.0 1.9 1.9 3.8 4.8 26.0 39.4 14.4 9.6

guardian of a child between the ages of 8 and 12 years at the start of the study. There were 150 parent-child dyads that began the program; 76 dyads were recruited into the Experimental group and 74 to the Control group. Of these, 72 and 64 dyads, respectively, completed the post-program assessment, for a Pre to Post completion rate of 91%. Fifty-eight and 47 dyads, respectively, completed the follow-up assessment, for a completion rate of 70% from pre-intervention to follow-up. Mean age of parent participants at pre-intervention was 40.7 ± 5.4 years. Mean age of child participants was 10.6 ± 1.35 years. At pre-intervention 18.9% of children and 30.4% of parents were categorized as “overweight,” while 16.9% of children and 39.2% of parents were categorized as “obese.” No parents or children were “underweight.” Table 1 shows additional participant demographics of dyads that completed the follow-up assessment. Experimental Curriculum Development and Delivery Montana State University Extension Specialists on the research team utilized existing research, focus group results, and selected components from existing research-based curricula to develop a pilot version of the Experimental curriculum. The curriculum was revised after an 8-month pilot test in three counties. For additional details on the pilot study and background to intervention development see Lynch et al. (2012). During the pilot phase, Agents had weekly conference calls with the Extension Food and Nutrition Specialist and Project Director to give feedback, clarify areas where there were questions, and check for consistency in how the Experimental intervention was delivered. After piloting the initial curriculum, adjustments were made based on feedback from parents, Agents offering the intervention, and research on the topics. During the full intervention, post-session facilitator evaluations were completed and study staff visited sites during sessions to check intervention fidelity for the Experimental group. Experimental intervention curriculum topics and objectives. Food and Nutrition topics and objectives included recommendations such as understanding general nutrition recommendations; increasing times when families eat together; practicing the principles of normal, healthy eating; and avoiding unhealthy weight-control practices. Examples of Physical Activity objectives include understanding general physical activity recommendations

146

G. Eldridge et al. / Body Image 16 (2016) 143–153

and concepts; creating an accessible physical environment that promotes an active lifestyle; and promoting physical activity via whole-family communication. Body Image objectives aimed to encourage size and body acceptance of self and others; understand media and environmental influences on the development of body image; and teach and model healthy self-esteem, respect, and confidence. Parenting and Family Communication objectives included recommendations to practice good communication skills; provide high levels of love, warmth, and boundaries; advocate for the preteen child; and provide opportunities for the preteen child to grow and develop his or her own identity. Parenting and family communication examples often centered on food and nutrition, physical activity, or body image, but were skills that could be applied to any topic. Body image topics were integrated throughout the 8-month intervention and were addressed via facilitator presentation, information in the parent guide, informational slides, take-home body image environment activities, discussions of parenting styles and stages of child development, a media impact activity using short videos about body image in boys and girls, goal-setting around body image, and scenario-based conversations about communicating with preteens about body image issues. Mailed information minimal (Control) intervention materials. The minimal (Control) intervention consisted of written information based on the USDA’s ChooseMyPlate website (United States Department of Agriculture, n.d.) and the Academy of Nutrition and Dietetics’ – formerly the American Dietetic Association – internet resources (Hayes, American Dietetic Association, & American Dietetic Association Foundation, n.d.). The general topic areas covered by these materials were the same as those included in the Experimental intervention; however, in this case, participants were given no explicit instructions about the use of provided information. Despite the lack of explicit instructions, however, all parents were aware they were participating in a “healthy living” research study. Body image was specifically addressed in one page titled ‘Tips for talking to your child about body image,’ which was included in the final mailing of the intervention (Hayes et al., n.d.). Procedures At each assessment session parents and children each provided anthropometric data and also completed unique (parent or child) survey packets. Participants completed the same set of assessments at each session. All assessments were carried out with groups of participants at county Extension offices (or nearby locations). Assessments were scheduled during the 3 weeks preceding and following program delivery and at 6 months after the postintervention data collection session. Teams of data collectors (3–4 members per team), hired and trained for each data collection period, traveled to each county data collection site at the scheduled times. Data collectors received two days of intense training on the research protocol, use of equipment, and how to interact with the program participants so that data collection was highly consistent and sensitive to participants. The children’s extensive battery of questionnaires assessed their attitudes and behaviors related to all aspects of the interventions including physical activity, food and nutrition, body image, appearance, body esteem, self-esteem, and quality of life. Parents completed questionnaires about themselves and their child in terms of physical activity and food and nutrition, their own body image and parenting style, and their child’s quality of life. In the following pages only the data related to child body mass index (BMIz), body size perception, and body image broadly defined are presented.

Parent-child dyads in both the Experimental and Control groups received $50 for completing each assessment session. All parentchild dyads that completed the study also received a $200 stipend at the post-program assessment session. Completing the study, for parents in the Experimental group, meant attending at least eight out of the 10 face-to-face sessions and the post-program assessment session. Experimental group parent participants also received a small travel stipend depending on the distance traveled to and from sessions and the number of sessions attended. Measures Anthropometric measures. Height, weight, resting heart rate, and resting systolic and diastolic blood pressure of parents and children were measured by the trained data collectors. Participants could not see the measurement results and were not told their height or weight, unless they specifically requested this information. Height was measured in centimeters using a Seca 217 stadiometer and weight was measured in pounds, using a Tanita WB-110A scale, both procedures following the National Health and Nutrition Examination Survey (NHANES) protocol (Centers for Disease Control and Prevention, 2007). Height and weight were both measured two times, unless the first two readings were >0.5 cm different, for height, or >0.5 pounds different, for weight. In that case, a third measurement was taken. Using median height and weight measures, a body mass index (BMI) score was calculated using the standard procedure [BMI = weight (kg)/height (m)2 ]. For children, this calculated BMI score was then converted to an age- and gender-specific BMI z-score (BMIz) using the CDC’s SAS program for CDC growth charts (Centers for Disease Control and Prevention, n.d.). Body image measures. Body image is presumed to be a multidimensional construct, which has been assessed in children by a variety of methods (Smolak, 2004). Recognizing this, as well as the limitations of time and the cognitive capabilities of 8–12 year old children, we selected the following four instruments in order to gain a broad assessment of the “body image” construct. Where appropriate Cronbach’s alphas (˛) for each subscale score are indicated in Table 2. Weight self-perception and weight management goal. Two questions derived from the Youth Risk Behavior Survey (YRBS) were included as part of the body image assessment (Centers for Disease Control and Prevention, 2009). One question (Weight Self-Perception.) simply asked children “How do you describe your weight?” Response choices were on a 5-point scale (1 = very underweight and 5 = very overweight). The other question (Weight Management Goal) asked, “Which of the following are you trying to do about your weight?” Choices were 1 = lose weight, 2 = stay about the same weight or do nothing, and 3 = gain weight. Although complete psychometric research is lacking, previous studies provide some evidence for the validity of these questions. For example, overestimation of weight, derived from the weight self-perception question and BMI percentile derived from self-reported height and weight, are both positively correlated with higher rates of disordered eating behaviors (Eichen, Conner, Daly, & Fauber, 2012). Similarly, responses to the weight management goal question are positively correlated with regular participation in vigorous physical activity, strengthening exercises, and smoking (Brener, Eaton, Lowry, & McManus, 2004). Perceived body size, ideal body size, body size misperception, and overweight dissatisfaction. The Children’s Body Image Scale (CBIS) was used to assess children’s own body size perception

G. Eldridge et al. / Body Image 16 (2016) 143–153

147

Table 2 Means (SD) for all dependent variables by group, gender, and time (pre, post, 6-month follow-up). Dependent variable

Female

Male

Experimental

Control

Experimental

Control

n = 41 3.12 (0.71) 3.05 (0.77) 3.20 (0.87)

n = 31 3.06 (0.77) 2.97 (0.80) 3.23 (0.62)

n = 18 3.00 (0.97) 2.78 (0.80) 3.00 (0.77)

n = 13 3.38 (0.51) 3.31 (0.75) 3.38 (0.65)

n = 41 1.54 (0.55) 1.41 (0.50) 1.54 (0.55)

n = 31 1.71 (0.64) 1.61 (0.56) 1.52 (0.57)

n = 18 1.67 (0.59) 1.89 (0.76) 1.94 (0.73)

n = 13 1.46 (0.52) 1.85 (0.69) 1.85 (0.69)

n = 36 1.69 (1.14) 1.44 (1.13) 1.86 (1.22)

n = 27 1.78 (1.36) 1.41 (1.12) 1.78 (1.19)

n = 14 1.79 (1.19) 1.79 (0.89) 1.79 (1.37)

n=7 0.86 (1.21) 1.42 (1.40) 1.14 (1.35)

n = 36 0.91 (0.81) 0.61 (0.73) 0.61 (0.64)

n = 27 0.63 (0.84) 0.59 (0.89) 0.56 (0.70)

n = 14 0.79 (0.89) 0.43 (0.51) 0.71 (0.61)

n=7 1.29 (0.95) 0.71 (0.95) 0.71 (0.76)

n = 32 2.06 (1.04) 2.25 (1.05) 2.06 (1.22)

n = 27 2.13 (1.11) 2.22 (1.19) 2.02 (1.29)

n = 15 1.93 (0.88) 2.13 (1.25) 2.33 (0.90)

n=9 2.44 (0.53) 2.22 (1.09) 2.56 (1.01)

n = 32 11.05 (2.08) 12.09 (1.33) 11.87 (1.79)

n = 27 10.74 (2.71) 11.37 (1.88) 11.56 (1.97)

n = 15 10.27 (2.31) 10.73 (2.55) 12.13 (1.64)

n=9 11.78 (2.44) 11.77 (1.92) 11.89 (1.90)

n = 27 1.81 (0.40) 1.78 (0.51) 1.96 (0.19)

n = 15 1.73 (0.46) 1.67 (0.62) 1.67 (0.49)

n=9 1.67 (0.50) 1.89 (0.33) 1.44 (0.53)

n = 31 16.59 (4.91) 16.58 (5.47) 16.68 (4.97)

n = 15 11.07 (2.91) 9.60 (3.89) 9.27 (3.73)

n = 13 11.49 (3.43) 9.69 (2.50) 9.31 (3.30)

n = 31 18.78 (6.92) 18.68 (7.08) 19.64 (6.89)

n = 15 14.20 (4.87) 12.73 (4.96) 13.40 (5.40)

n = 13 17.31 (4.87) 15.92 (4.99) 15.08 (4.75)

n = 15 5.73 (2.05) 4.40 (2.13) 6.73 (2.02)

n = 13 6.62 (1.76) 5.31 (2.36) 5.08 (2.10)

YRBS-Weight Self-Perception Pre Post 6-month follow-up YRBS-Weight Management Goal Pre Post 6-month follow-up CBIS-Body Size Perception Pre Post 6-month follow-up CBIS-Overweight Dissatisfaction Pre Post 6-month follow-up BES-Weight (˛ = .77; 3 items) Pre Post 6-month follow-up BES-Appearance (˛ = .87; 13 items) Pre Post 6-month follow-up BES-Attribution (˛ = .61; 5 items)

n = 32 Pre 1.78 (0.42) Post 1.88 (0.34) 6-month follow-up 1.94 (0.25) SATAQ-Awareness – Girls (˛ = .78; 6 items); Boys (˛ = .53; 4 items) n = 39 Pre 16.17 (4.66) Post 14.49 (6.08) 6-month follow-up 14.14 (5.38) SATAQ-Internalization – Girls (˛ = .86; 8 items); Boys (˛ = .72; 7 items) n = 39 Pre 19.46 (6.46) Post 17.33 (7.72) 6-month follow-up 16.13 (7.43) SATAQ-Muscular Look Boys (˛ = .31; 2 items) Pre Post 6-month follow-up

Note: Data shown are for participants who completed all three assessments and provided complete data for each dependent variable associated with a given assessment instrument. For BES and SATAQ subscales, Cronbach’s alphas (˛) and the number of items included for each dependent variables are shown to the right of the subscale name.

and body size satisfaction (Truby & Paxton, 2002). Seven genderspecific body shape figures, ranging in size from very underweight (1) to very overweight (7), were presented to the children. The figures were developed to represent standardized BMI percentiles for healthy children. The seven gender-specific body shape figures were based on photographs of children at the 3rd, 10th, 25th, 50th, 75th, 90th, and 97th BMI percentile and each figure represents the range of BMI from the mid-point of the figure below and above (Truby & Paxton, 2002). Thus, a specific range of actual BMI scores corresponds to each of the gender-specific figures. For example, the actual body size figure for boys with measured BMI of 14.0–14.6 was the most underweight figure (1), while actual body size figure of boys with a measured BMI of 28.5–29.0 was the most overweight figure (7). The figure corresponding to the child’s measured BMI is considered his/her actual body size. Each child was asked to respond to two statements by circling one of the seven

figures for each statement. (A) “Identify the body figure most like your own.” This selection is considered the child’s perceived body size. (B) “Identify the body figure you would most like to have.” This selection is considered the child’s ideal body size. Using these three figure selections, two dependent variables were calculated. The first representing body size misperception, was calculated as the difference between the actual and perceived body sizes (actual minus perceived), so that positive values represented the most common misperception; namely that perceived body size was smaller than actual body size. The second dependent variable, representing weight dissatisfaction, was calculated as the difference between the perceived and ideal body sizes (perceived minus ideal). In this case, positive values represent dissatisfaction due to a larger perceived than ideal body size. Truby and Paxton reported significant correlations between the perceived-ideal discrepancy and verbal questions about body size

148

G. Eldridge et al. / Body Image 16 (2016) 143–153

perception and preference (Truby & Paxton, 2002) Perceived-ideal discrepancy was also significantly correlated to Body Esteem Scale scores (Truby & Paxton, 2002). Because most children were dissatisfied with their body size due to a perception that their bodies were larger than ideal, both at pre- and post-intervention (i.e., 90.2% of valid cases), we limited our analysis to this group. In what follows, we refer to this variable as “Overweight Dissatisfaction.” Body esteem. The Body Esteem Scale (BES) is a measure of children’s self-evaluation of their own body and appearance (Mendelson, White, & Mendelson, 1996), which has three subscales: BES-Appearance (general feelings about appearance), BES-Weight Satisfaction, and BES-Attribution (evaluations attributed to others about one’s body and appearance). For each subscale, higher scores represent greater body esteem. Based on previous psychometric analyses (Mendelson et al., 1996) the above three factor scores were calculated. Mendelson et al., reported acceptable internal reliability, Cronbach’s ˛ = .87 for BESAppearance and ˛ = .77 for BES-Weight Satisfaction (Mendelson et al., 1996). They did not determine internal reliability for BESAttribution since it consists of only two items (Mendelson et al., 1996). Appearance attitudes. The Sociocultural Attitudes Towards Appearance Questionnaire (SATAQ) is designed to assess the child’s level of endorsement of societal appearance ideals (Heinberg, Thompson, & Stormer, 1995). SATAQ total scores and specific subscale scores were also calculated separately for girls and boys. For girls, subscale scores were calculated for awareness of appearance ideals (SATAQ Awareness) and internalization of ideals (SATAQ Internalization). For boys, subscale scores were calculated for awareness of ideals, internalization of ideals, and concern about muscular looks (SATAQ Muscular Look). In general, higher scores on each of these scales represent greater appearance-related concerns. For both genders SATAQ total scores and subscale scores have shown adequate internal reliability, Cronbach’s ˛ = .75 to .91 (Smolak, Levine, & Thompson, 2001). Results All statistical analyses were carried out using the GLM repeated measures ANOVA procedure of SPSS for Mac (ver. 20). Because random sampling was not possible due to the applied nature of the study (see above), females were over represented in both treatment groups. Therefore, Treatment (Group) and Gender were entered as between-subjects variables and Time of Assessment (Time) was entered as within-subjects variable. Four separate multivariate analyses (2 Group × 2 Gender × 3 Time) were carried out, one analysis for each of the four measurement scales (YRBS, CBIS, BES, and SATAQ), with subscale scores serving as dependent variables in each case. SATAQ data for males and females were analyzed separately with three dependent variables included for males and two for females. Table 2 shows the means and SDs for each univariate dependent variable by Group, Gender, and Time. Univariate and multivariate outliers were removed prior to any analyses. BMI z-scores In designing this study, we had assumed that each of the subscale scores derived from the above measurement scales would contribute to an overall measure of “body image.” Furthermore, based on previous research, we assumed that this overall measure of “body image” would be associated with BMI. In order to evaluate these assumptions, we regressed pre-intervention BMIz scores on the subscale scores shown in Table 2. Results based on

Table 3 Mean BMIz (SD) by group and gender across three assessment times. n Experimental Male Female Control Male Female Total

Pre

Post

6-month follow-up

17 41

0.71 (0.92) 0.54 (0.91)

0.60 (0.90) 0.48 (0.90)

0.58 (0.89) 0.54 (0.90)

13 32

1.11 (0.98) 0.34 (0.84)

1.02 (1.13) 0.33 (0.81)

1.03 (1.10) 0.50 (0.97)

103

0.57 (0.92)

0.52 (0.91)*

0.60 (0.94)

Note: Data shown represent only those participants who completed all three assessments. For both groups, the overall (Total) BMIz decrease from pre- to post-intervention was small but statistically significant (* p < .05); however the preintervention to 6-month follow-up difference was not significant (p > .05).

pre-intervention data confirmed that overall, the subscale scores included in this analysis accounted for more than half the variance in BMIz scores both for boys, R2 = .68, F(10, 42) = 13.45, p < .001, and girls, R2 = .55, F(9, 90) = 13.10, p < .001. Although the 4-Health Program was intended to reduce obesity risk rather than to promote weight loss, the well-established association between body weight and body image suggested that changes in BMIz might be associated with corresponding changes in body image and perhaps future obesity risk. Thus, a preliminary (2 Group × 2 Gender × 3 Time) ANOVA was carried out with genderand age-adjusted BMIz scores as the dependent variable. Results showed a significant multivariate effect of Time, Wilks’ lambda F(2, 98) = 3.64, p < .05, 2 = .07. Within-subjects contrasts revealed this effect was primarily due to a significant decrease in BMIz from preto post-intervention, F(1, 99) = 5.60, p < .05, 2 = .05. At follow-up, mean BMIz for both groups recovered to near pre-intervention levels, such that the pre-intervention to follow-up contrast was not significant (p > .05). Table 3 shows the mean BMIz (+SD) by group and gender at each assessment time. Of the 137 children who completed the post-intervention assessment, only one moved from a “normal weight” to “underweight” based on her BMI percentilefor-age (9.57 to 3.80). Weight Self-Perception and Weight Management Goals Two items derived from the Youth Risk Behavior Survey were included in a multivariate ANOVA. Results showed a significant multivariate effect of Time, Wilks’ lambda F(4, 396) = 2.41, p < .05, 2 = .02, as well as a significant multivariate Gender × Time interaction, Wilks’ lambda F(4, 396) = 3.70, p < .05, 2 = .04. Both of these effects were very small. Univariate tests failed to show a significant Time effect on either weight self-perception (p > .05) or weight management goals (p > .05). However, a significant univariate Gender × Time interaction revealed that girls maintained or slightly increased their weight loss goals over the three assessment sessions, whereas boys reported a progressive decrease in their weight loss goals. Notice that this could also be considered a shift among boys toward a weight gain goal, although their mean score remained below 2 (i.e., “do nothing” or “stay the same”). Over the course of the three assessments, the percentage of boys reporting an actual weight-gain goal (score >2) went from 8.3% at pre-intervention to 22.7% at post-intervention, and 18.8% at follow-up. T-tests for gender differences in weight management goals at each assessment time were statistically significant both at post-intervention, Mdiff = .38, t(102) = 2.98, p < .01, r = .28 and at follow-up, Mdiff = .38, t(102) = 2.97, p < .01, r = .28. It should be noted that 72% of children in this study were above their age- and gender-specific BMIz scores and nearly half of each gender group (male = 47.8%; females = 48.5%) described themselves as having the goal of losing weight.

G. Eldridge et al. / Body Image 16 (2016) 143–153

Body Size Misperception and Overweight Dissatisfaction Analyses of data derived from the Children’s Body Image Scale showed a significant multivariate effect of Time, Wilks’ lambda F(4, 77) = 3.15, p < .05, 2 = .14. Univariate tests revealed this effect was primarily due to a significant decrease in Overweight Dissatisfaction, F(2, 160) = 5.68, p < .01, 2 = .07, but not Body Size Misperception (p > .05), and was statistically significant from preto post-intervention, F(1, 80) = 8.65, p < .01, 2 = .10, as well as from pre-intervention to follow-up, F(1, 80) = 6.22, p < .05, 2 = .07. Again, the effect sizes were small. Although nearly all participants underestimated their actual size, and generally reported overweight dissatisfaction, none of the between-group or between-gender effects, or their interactions, was significant. Body Esteem All three BES subscale scores were included in an initial MANOVA for effects on body esteem. Results of this analysis showed a significant multivariate effect of Gender, Wilks’ lambda F(3, 77) = 2.77, p < .05, 2 = .10, and a significant multivariate (Gender × Time) interaction, Wilks’ lambda F(6, 74) = 2.43, p < .05, 2 = .16. Univariate tests showed that the Time effect was due primarily to a significant increase in the BES-Appearance subscale score, which began to increase during the intervention and continued beyond the post-intervention assessment, to become statistically significant by follow-up, F(1, 79) = 7.49, p < .01, 2 = .09. Univariate tests also showed that the Gender × Time interaction was due to an increase in the BES-Attribution score among females but a decrease among males across the three assessment periods, F(1, 79) = 6.24, p < .05, 2 = .07. T-tests for gender differences at each time-point showed a significant difference only at follow-up, Mdiff = 3.52, t(99) = 3.52, p < .001. Appearance Attitudes: Boys SATAQ scores were analyzed separately for males and females. For boys, there was a relatively large and significant multivariate effect of Time, Wilks’ lambda F(6, 21) = 2.71, p < .05, 2 = .44, as well as a relatively large and significant multivariate (Group × Time) interaction, Wilks’ lambda F(6, 21) = 2.95, p < .05, 2 = .46. Univariate tests showed the Time effect was due mainly to a significant decrease in SATAQ-Awareness, which was statistically significant from pre- to post-intervention, F(1, 26) = 6.59, p < .05, 2 = .20, as well as from pre-intervention to follow-up, F(1, 26) = 6.77, p < .05, 2 = .21. The significant multivariate (Group × Time) interaction reflected a significant change in the SATAQ-Muscular Look score for boys. In this case, the interaction was only significant over the period from pre-intervention to follow-up, F(1, 26) = 6.30, p < .05, 2 = .20, reflecting the fact that boys in the Experimental group reported an increase in their agreement that there is a positive association between being muscular and “looking good” from preintervention to follow-up, F(1, 26) = 6.59, p < .05, 2 = .44, whereas boys in the Control group continued to report less agreement that there is a positive association between being muscular and “looking good” over this same period. T-tests at each time-point indicated a significant group difference only at follow-up, t(28) = 2.50, p < .05. Appearance Attitudes: Girls Among girls, although the multivariate Group × Time interaction was not significant (p > .05), there was a small but significant univariate Group × Time interaction effect on Internalization of Appearance ideals, F(2, 136) = 3.88, p < .05, 2 = .05. In this case, girls in both groups reported a decrease in Internalization from pre- to post-intervention. For girls in the Experimental group this decrease

149

continued through the follow-up assessment. For girls in the Control group, however, there was a mean increase in Internalization scores from post-intervention to follow-up. T-tests at each time point showed that only the group difference at follow-up was statistically significant, Mdiff = 3.52, t(68) = 2.04, p < .05.

Discussion The face-to-face (Experimental) intervention and the minimal (Control) intervention were both modestly effective at improving some body image measures, thus potentially reducing future obesity risk. Both interventions also resulted in decreases in BMIz scores from pre- to post-intervention, although these decreases were relatively small and were effectively reversed by the 6month follow-up. Unfortunately, such a reversal of BMI decreases at follow-up is not uncommon for child and adolescent obesity prevention programs (Ajie & Chapman-Novakofski, 2014; Hamel & Robbins, 2013; Shirley et al., 2015; Stice, Shaw, & Marti, 2006). Although children in this study were not selected on the basis of body weight, most were above their age- and gender-specific BMIz scores and higher BMIz scores were strongly associated with greater body image concerns, regardless of gender. As noted earlier, weight loss was not a goal of the current study. However, when considered in the context of other changes, such as enhanced body esteem, decreased weight dissatisfaction, and lessened internalization of cultural ideals about weight, the small decreases in BMIz found in this study support the notion that parent-only interventions implemented during the preteen years may be effective at decreasing obesity risk. Other recent studies have also found parent-only programs produce significant decreases in child BMIz scores (Janicke et al., 2008; Magarey et al., 2011; Munsch et al., 2008; Resnick et al., 2009). Earlier studies, reported that parentonly programs were more successful for child weight loss than child-only or parent-child programs (Golan & Crow, 2004; Golan, Fainaru, & Weizman, 1998; Golan et al., 2006). How changes such as those seen here can be made more persistent remains to be determined. The effectiveness of our minimal intervention may suggest that continued engagement with parents, by providing them with updated and relevant healthy living information, should be explored as a way to extend these BMI improvements. Both interventions had additional effects on body esteem. First, there was a small but significant decrease in weight dissatisfaction (BES-Overweight Dissatisfaction) from pre- to post-intervention, which remained significant at 6-month follow-up. Second, there was a small but significant increase in body esteem based on appearance (BES-Appearance) for both genders and a signficant gender difference in body esteem based on attributions to others (BES-Attribution), but only at the 6-month follow-up. Previous cross-sectional studies have reported that BES-Appearance scores decreased as age increased (Frisen, Lunde, & Kleiberg, 2013; Mendelson, White, & Mendelson, 1995; Mendelson et al., 1996); but longitudinal studies have suggested that boys’ general satisfaction with appearance begins to increase around age 13 (Holsen, Jones, & Birkeland, 2012; Rosenblum & Lewis, 1999). Thus, it is uncertain whether the increase seen in the present study was due to our interventions or to age. Previous studies have not reported gender (Frisen, Lunde, & Berg, 2015; Frisen et al., 2013) or age (Frisen et al., 2013; Mendelson et al., 2001) differences in BES-Attribution scores. Similarly, we found no significant gender differences in BESAttribution scores during the interventions but a significant gender difference, mainly due to a decrease in the attribution score among males, at follow-up. At present we have no explanation for this apparent gender difference six months following the intervention. As noted earlier, Truby and Paxton (2002) found that BES scores are significantly correlated with body dissatisfaction assessed by

150

G. Eldridge et al. / Body Image 16 (2016) 143–153

the CBIS. Thus, given abundant evidence suggesting that body dissatisfaction is a risk factor for eating disorders and numerous other mental and physical disorders, these small improvements in body image are notable. Furthermore, the fact that these effects were found in both intervention groups suggests that engaging parents of preteens in a long-term program to promote healthy living or simply providing them with written information about best practices may have significant beneficial effects on important dimensions of body image. Although children in this study were not selected on the basis of body weight, nearly half of both gender groups reported wanting to lose weight at pre-intervention. Interestingly, however, while girls persisted in this weight-loss goal throughout the study, boys moved significantly away from their weight loss goal and a substantial percentage of them reported wanting to gain weight at the completion of the interventions and at follow-up. This may reflect the fact that as boys enter their teenage years they are uniquely subject to media and peer messages consistent with the idea that having a larger body is a more desirable goal. Self-perception of body size and body weight changed little over the course of the present study. At pre-intervention children in both treatment groups, regardless of gender, when asked to describe their weight, reported it was “about right.” By contrast, they generally perceived themselves as substantially larger than their “ideal” body size based on the CBIS measure. Rand and Resnick (2000) reported similar results, showing that most people consider their own body size within the range of acceptable sizes, even though their body size might not match their “ideal.” In the present study, despite the apparent discrepancy in self-perception based on different methods of assessment, neither self-perception changed significantly over the course of the study. Virtually all participants misperceived their body size as smaller than their “actual” size and reported an “ideal” size even smaller than their “perceived” size. Misperception of body weight or size has been significantly associated with all 32 health risk behaviors measured by the Youth Risk Behavior Survey; these are grouped into the domains of safety and violence, mental health, substance abuse, and dieting and physical activity (Jiang, Kempner, & Loucks, 2014). Similarly, a number of studies have found that inaccurate perceptions of overweight among normal weight adolescents are associated with higher rates of eating disorders and unhealthy weight control behaviors, as well as mental health problems (Al Mamun et al., 2007; Ali, Fang, & Rizzo, 2010; Eichen et al., 2012; Talamayan, Springer, Kelder, Gorospe, & Joye, 2006). Indeed, one study found that normal weight 14-year-olds who perceived themselves as overweight had more mental health problems seven years later than normal weight adolescents who perceived themselves to be ‘about the right weight’ (Al Mamun et al., 2007). The consistency and persistence of body weight and size misperception seen in the present study may suggest a potentially serious problem needing to be addressed by future intervention programs. Despite major differences in the methods of delivery of the two interventions in the current study, the experimental intervention proved superior in only two cases, both involving moderate sized differences in appearance attitudes. In one case, girls in both groups decreased their scores on the SATAQ-Internalization subscale from pre- to post-intervention. However, this decrease only continued at follow-up for girls in the Experimental group; the initial decrease was reversed at follow-up among girls in the Control group. Thus, only girls in the Experimental group continued to decrease their adoption (internalization) of the “thinness-ideal” 6 months after the intervention ended. Similarly, McVey and colleagues found that girls in an 8-month school-based eating disorder prevention program and a comparison group of girls decreased their SATAQ-Internalization scores pre- to post-intervention, but

only the girls in the intervention group maintained the decrease at the 6-month follow-up (McVey, Tweed, & Blackmore, 2007). In the second case, we found a similar difference between the two interventions among boys. In this case, both groups decreased their scores on the SATAQ-Muscular Look subscale from pre- to postintervention but this decrease was reversed at follow-up for boys in the Experimental group, resulting in a moderately large and significant difference between groups by follow-up. Why Muscular Look scores increased at follow-up only among boys in the Experimental group is not clear. Further examination of our data indicated no significant correlation between Muscular Look scores and either child age or BMIz scores. However, the low reliability of this two-item scale (˛ = .31) may be notable. It is also possible that variables not measured in the current study differed between the two groups. For example, Verstuyf, Petegem, Vansteenkiste, Soenens, and Boone (2013) suggested that for adolescent boys different styles of identity exploration are associated differentially with changes in adoption of the muscular ideal. Given the few differences in outcomes for the two very different intervention methods, one must ask whether such time and resource intensive programs as the 4-Health face-to-face educational program are justified, when a simple mail-delivery of comparable information seems to have nearly equal benefit. Although we can only speculate, the answer to this question may depend on the types of participants and their commitment to the healthy-living goals of the program. Most of our parent participants had a long-standing involvement in the 4-H Youth Development Program and were committed to their children’s health. Health is one of the four clover leafs of 4-H emblem. Moreover, they had established relationships with their County Extension Agents, and were familiar with health research programs at Montana State University. Such involvement and commitment to their family’s health may explain why participants in both treatment groups took full advantage of the information provided, regardless of the method of delivery. Limitations related to the rural and applied nature of the present study are not unique to this study. Rural health researchers commonly encounter methodological challenges including unique cultural and social issues, limited availability of appropriate materials and instrumentation, special privacy concerns, and difficulties recruiting an adequate sample, the last of which can lead to statistical problems (Bigbee & Lind, 2007). As noted earlier, it was not practical in the present study to randomly assign participants to groups or to balance the numbers of males and females in each group. As a result, girls made up the largest proportion of participants in both treatment groups and overweight boys were over-represented in the Control group. Such non-random variables pose a potentially serious limitation for experimental studies of obesity risk and prevention carried out in applied settings. Although the gold standard for evaluation of intervention effectiveness is the randomized controlled trial, practical considerations often preclude its use. Despite this, it is possible to conduct methodologically sound evaluation research (Kirkwood, Cousens, Victora, & deZoysa, 1997). It is increasingly recognized that successful evaluations of public health interventions will often entail the use of research designs other than randomized controlled trials (Victora, Habicht, & Bryce, 2004). Randomized controlled trials often produce inconsistent results and may have limited generalizability (Curtis, Kennedy-Martin, Faries, Robinson, & Johnston, 2014; Horwitz, 1987; Rabeneck, Viscoli, & Horwitz, 1992). Furthermore, results of randomized controlled trials and non-randomized studies may not yield substantially different results (Ioannidis et al., 2001; McKee et al., 1999). It has been suggested that variations are often no greater than those between different randomized controlled trials or among non-randomized studies (McKee et al., 1999).

G. Eldridge et al. / Body Image 16 (2016) 143–153

Another limitation related to the significant Time effects noted in the present study, is the fact that the assessments occurred at different times of the year and several months apart. This is potentially important for two reasons. First, the length of time between assessments represents a substantial part of each child’s life, so that age-related changes in body image may have occurred between the two assessment times, regardless of treatments. Of course, this could be assessed by the addition of a no-intervention control group. Second, although we are not aware of any objective research on this issue, seasonal differences in body weight or physical activities may have influenced body image differences at the three assessment times. Had we included a no-intervention control group in addition to the two intervention groups compared in this study we might have observed more substantial treatment group differences. Unfortunately, financial and other considerations (noted earlier) made it impossible to expand this experiment to three groups. Given the option of either including a no-intervention control group or a minimal intervention control group, we opted for the minimal intervention group on the grounds that finding differences between an experimental and no-intervention control would only tell us that our intervention was superior to doing nothing. With the inclusion of a minimal intervention control group, we were able to show that mailed information was in many cases as effective as the face-toface intervention. This study has implications for practitioners and clinicians. Although in many cases neither intervention was superior, the data suggest that parent-only programs may help preteens develop healthier body image assessments, particularly improved weight satisfaction. For practitioners the findings point to the importance of helping parents understand how their children view their own bodies during preadolescence. Educating parents during a time in their child’s development when they still have influence on their child’s day-to-day decisions may be particularly beneficial. Programs specifically targeting parents may raise their awareness of body image changes in their children and encourage them to play an active role in promoting a healthy body image. This awareness can set the stage for parents to initiate constructive conversations with their children about this sensitive and important topic. Clinicians, such as counselors and dietitians who work with preadolescents are encouraged to involve parents in treatment plans and goals, emphasizing healthy lifestyle habits and acceptance of differences in the human body including weight and shape. Finally, this study demonstrates the feasibility of using Cooperative Extension Service offices to deliver child obesity prevention programs to rural families. Overall participation from pre- to postintervention was over 90%. This high attendance and completion rate is not surprising given participants’ ongoing involvement in the 4-H Youth Development Program. Parents were interested in their children’s health (one emphasis of 4-H), they had developed relationships with and respect for their County Extension Agents, and they were familiar with the quality of Montana State University research programs. In addition, Agents took care to schedule sessions at times suitable for the Experimental group parents and parents enjoyed the camaraderie during classes. The $250 stipend for Experimental and Control group parent-child dyads attending the pre- and post-intervention assessment sessions ($50 for preintervention, $200 for post-intervention) certainly also contributed to completion rates for both groups. We conclude that parentonly obesity prevention programs such as those used here can improve the body image of rural preadolescent youth. Our assessments revealed numerous changes among average rural children that resulted from these interventions, although many of these changes were small to moderate in size and some were specific to only one gender.

151

Acknowledgments The USDA’s National Institute of Food and Agriculture (NIFA) and the Agriculture and Food Research Initiative (AFRI) program on Childhood Obesity Prevention provided funding for the current research. The CSREES award number is 2009-55215-05334. Neither the funding agency nor its representatives played any part in the research design, data collection, analysis, interpretation of the results, or the decision to submit the present manuscript for publication in Body Image. The content of this paper is solely the responsibility of the authors and does not represent the views of Montana State University, the Montana State Extension Service, the USDA, or other agencies. We wish to specifically acknowledge the contributions and the dedication of the Extension Agents and 4H parents and children who participated in this research. We also acknowledge the support of Montana State University departments of Health and Human Development and Psychology and their staff, particularly Jan Feist and Brenda Lewis, respectively. References Ajie, W. N., & Chapman-Novakofski, K. M. (2014). Impact of computer-mediated, obesity-related nutrition education interventions for adolescents: A systematic review. Journal of Adolescent Health, 54, 631–645. http://dx.doi.org/10.1016/j. jadohealth.2013.12.019 Al Mamun, A., Cramb, S., McDermott, B. M., O’Callaghan, M., Najman, J. M., & Williams, G. M. (2007). Adolescents’ perceived weight associated with depression in young adulthood: A longitudinal study. Obesity, 15, 3097–3105. http://dx.doi.org/10. 1038/oby.2007.369 Ali, M. M., Fang, H., & Rizzo, J. A. (2010). Body weight, self-perception and mental health outcomes among adolescents. Journal of Mental Health Policy and Economics, 13, 53–63. Beech, B. M., Klesges, R. C., Kumanyika, S. K., Murray, D. M., Klesges, L., McClanahan, B.,. . . & Pree-Cary, J. (2003). Child- and parent-targeted interventions: The Memphis GEMS pilot study. Ethnicity & Disease, 13, S1–S40. Bigbee, J. L., & Lind, B. (2007). Methodological challenges in rural and frontier nursing research. Applied Nursing Research, 20, 104–106. http://dx.doi.org/10.1016/ j.apnr.2007.01.001 Brener, N. D., Eaton, D. K., Lowry, R., & McManus, T. (2004). The association between weight perception and BMI among high school students. Obesity Research, 12, 1866–1874. http://dx.doi.org/10.1038/oby.2004.232 Centers for Disease Control and Prevention. (n.d.). A SAS Program for the CDC Growth Charts. Retrieved from http://www.cdc.gov/nccdphp/dnpao/ growthcharts/resources/sas.htm Centers for Disease Control and Prevention. (2007). National Health and Nutrition Examination Survey: Anthropometry Procedures Manual.. Retrieved from www. cdc.gov/nchs/data/nhanes/nhanes 07 08/manual an.pdf Centers for Disease Control and Prevention. (2009). 2009 Middle School Youth Risk Behavior Survey.. Retrieved from ftp://ftp.cdc.gov/pub/data/yrbs/2009/2009 ms questionnaire.pdf Chung, S., & Joung, K. (2012). Risk factors related to suicidal ideation and attempted suicide: Comparative study of Korean and American youth. Journal of School Nursing, 28, 448–458. http://dx.doi.org/10.1177/1059840512446704 Corning, A. F., Gondoli, D. M., Bucchianeri, M. M., & Salafia, E. H. B. (2010). Preventing the development of body issues in adolescent girls through intervention with their mothers. Body Image, 7, 289–295. http://dx.doi.org/10.1016/j.bodyim. 2010.08.001 Curtis, S. E., Kennedy-Martin, T., Faries, D. E., Robinson, S., & Johnston, J. A. (2014). A scoping literature review on the external validity of randomized controlled trial populations. Value in Health, 17, A182. http://dx.doi.org/10.1016/j.jval.2014.03. 1060 Debar, L. L., Stevens, V. J., Perrin, N., Wu, P., Pearson, J., Yarborough, B. J.,. . .& Lynch, F. (2012). A primary care-based, multicomponent lifestyle intervention for overweight adolescent females. Pediatrics, 129, E611–E620. http://dx.doi.org/ 10.1542/peds.2011-0863 Eichen, D. M., Conner, B. T., Daly, B. P., & Fauber, R. L. (2012). Weight perception, substance use, and disordered eating behaviors: Comparing normal weight and overweight high-school students. Journal of Youth and Adolescence, 41, 1–13. http://dx.doi.org/10.1007/s10964-010-9612-8 Estabrooks, P. A., Shoup, J. A., Gattshall, M., Dandamudi, P., Shetterly, S., & Xu, S. (2009). Automated telephone counseling for parents of overweight children: A randomized controlled trial. American Journal of Preventive Medicine, 36, 35–42. http://dx.doi.org/10.1016/j.amepre.2008.09.024 Fenton, C., Brooks, F., Spencer, N. H., & Morgan, A. (2010). Sustaining a positive body image in adolescence: An assets-based analysis. Health & Social Care in the Community, 18, 189–198. http://dx.doi.org/10.1111/j.1365-2524.2009.00888.x Frisen, A., Lunde, C., & Berg, A. I. (2015). Developmental patterns in body esteem from late childhood to young adulthood: A growth curve analysis. European Journal of Developmental Psychology, 12, 99–115. http://dx.doi.org/10.1080/17405629. 2014.951033

152

G. Eldridge et al. / Body Image 16 (2016) 143–153

Frisen, A., Lunde, C., & Kleiberg, A. N. (2013). Body esteem in Swedish children and adolescents: Relationships with gender, age, and weight status. Nordic Psychology, 65, 65–80. http://dx.doi.org/10.1080/19012276.2013.796080 Garner, D. M., Olmstead, M. P., & Polivy, J. (1983). Development and validation of a multidimensional Eating Disorder Inventory for anorexia-nervosa and bulimia. International Journal of Eating Disorders, 2, 15–34. http://dx.doi.org/10.1002/ 1098-108x(198321)2:2<15::aid-eat2260020203>3.0.co;2-6 Golan, M., & Crow, S. (2004). Targeting parents exclusively in the treatment of childhood obesity: Long-term results. Obesity Research, 12, 357–361. http://dx.doi. org/10.1038/oby.2004.45 Golan, M., Fainaru, M., & Weizman, A. (1998). Role of behaviour modification in the treatment of childhood obesity with the parents as the exclusive agents of change. International Journal of Obesity & Related Metabolic Disorders, 22, 1217. http://dx.doi.org/10.1038/sj.ijo.0800749 Golan, M., Kaufman, V., & Shahar, D. (2006). Childhood obesity treatment: Targeting parents exclusively v. parents and children. British Journal of Nutrition, 95, 1008–1015. http://dx.doi.org/10.1079/BJN20061757 Golan, M., Weizman, A., Apter, A., & Fainaru, M. (1998). Parents as the exclusive agents of change in the treatment of childhood obesity. American Journal of Clinical Nutrition, 67, 1130–1135. Golley, R. K., Magarey, A. M., Baur, L. A., Steinbeck, K. S., & Daniels, L. A. (2007). Twelve-month effectiveness of a parent-led, family-focused weightmanagement program for prepubertal children: A randomized, controlled trial. Pediatrics, 119, 517–525. http://dx.doi.org/10.1542/peds.2006-1746 Goncalves, S., Silva, M., Gomes, A. R., & Machado, P. P. P. (2012). Disordered eating among preadolescent boys and girls: The relationship with child and maternal variables. Nutrients, 4, 273–285. http://dx.doi.org/10.3390/nu4040273 Hamel, L. M., & Robbins, L. B. (2013). Computer- and web-based interventions to promote healthy eating among children and adolescents: A systematic review. Journal of Advanced Nursing, 69, 16–30. http://dx.doi.org/10.1111/j.1365-2648. 2012.06086.x Harter, M. (2012). Evaluation of the SHAPEDOWN weight loss program conducted through a hospital in north central Indiana: A family approach. Ball State University. Hayes, D., American Dietetic Association, & American Dietetic Association Foundation. (n.d.). Promoting positive body image. Retrieved from http://www.eatright. org/kids/article.aspx?id=6442459384&terms=body+image Heinberg, L. J., Thompson, J. K., & Stormer, S. (1995). Development and validation of the Sociocultural Attitudes Towards Appearance Questionnaire. International Journal of Eating Disorders, 17, 81–89. http://dx.doi.org/10.1002/1098108X(199501)17:1<81::AID-EAT2260170111>3.0.CO;2-Y Holsen, I., Jones, D. C., & Birkeland, M. S. (2012). Body image satisfaction among Norwegian adolescents and young adults: A longitudinal study of the influence of interpersonal relationships and BMI. Body Image, 9, 201–208. http://dx.doi. org/10.1016/j.bodyim.2012.01.006 Horwitz, R. I. (1987). Complexity and contradiction in clinical trial research. The American Journal of Medicine, 82, 498–510. http://dx.doi.org/10.1016/00029343(87)90450-5 Hystad, H. T., Steinsbekk, S., Odegard, R., Wichstrom, L., & Gudbrandsen, O. A. (2013). A randomised study on the effectiveness of therapist-led v. self-help parental intervention for treating childhood obesity. British Journal of Nutrition, 110, 1143–1150. http://dx.doi.org/10.1017/s0007114513000056 Ioannidis, J. P. A., Haidich, A. B., Pappa, M., Pantazis, N., Kokori, S. I., Tektonidou, M. G.,. . .& Lau, J. (2001). Comparison of evidence of treatment effects in randomized and nonrandomized studies. Journal of the American Medical Association, 286, 821–830. http://dx.doi.org/10.1001/jama.286.7.821 Janicke, D. M., Sallinen, B. J., Perri, M. G., Lutes, L. D., Huerta, M., Silverstein, J. H., & Brumback, B. (2008). Comparison of parent-only vs family-based interventions for overweight children in underserved rural settings: Outcomes from Project STORY. Archives of Pediatrics & Adolescent Medicine, 162, 1119–1125. http://dx. doi.org/10.1001/archpedi.162.12.1119 Janicke, D. M., Sallinen, B. J., Perri, M. G., Lutes, L. D., Silverstein, J. H., & Brumback, B. (2009). Comparison of program costs for parent-only and family-based interventions for pediatric obesity in medically underserved rural settings. Journal of Rural Health, 25, 326–330. http://dx.doi.org/10.1111/j.1748-0361.2009.00238.x Jansen, E., Mulkens, S., & Jansen, A. (2011). Tackling childhood overweight: Treating parents exclusively is effective. International Journal of Obesity (London), 35, 501–509. http://dx.doi.org/10.1038/ijo.2011.16 Jiang, Y., Kempner, M., & Loucks, E. B. (2014). Weight misperception and health risk behaviors in youth: The 2011 US YRBS. American Journal of Health Behavior, 38, 765–780. http://dx.doi.org/10.5993/ajhb.38.5.14 Kirkwood, B. R., Cousens, S. N., Victora, C. G., & deZoysa, I. (1997). Issues in the design and interpretation of studies to evaluate the impact of community-based interventions. Tropical Medicine & International Health, 2, 1022–1029. http://dx. doi.org/10.1046/j.1365-3156.1997.d01-188.x Lynch, W. C., Martz, J., Eldridge, G., Bailey, S. J., Benke, C., & Paul, L. (2012). Childhood obesity prevention in rural settings: Background, rationale, and study design of ‘4-Health,’ a parent-only intervention. BMC Public Health, 12, 255. http://dx.doi. org/10.1186/1471-2458-12-255 Magarey, A., Perry, R., Baur, L., Steinbeck, K., Sawyer, M., Hills, A.,. . . & Daniels, L. (2011). A parent-led family-focused treatment program for overweight children aged 5 to 9 years: The PEACH RCT. Pediatrics, 127, 214–222. http://dx.doi.org/ 10.1542/peds.2009-1432 McKee, M., Britton, A., Black, N., McPherson, K., Sanderson, C., & Bain, C. (1999). Interpreting the evidence: Choosing between randomised and non-randomised studies. British Medical Journal, 319, 312–315. http://dx.doi.org/10.1136/bmj. 319.7205.312

McVey, G., Tweed, S., & Blackmore, E. (2007). Healthy Schools-Healthy Kids: A controlled evaluation of a comprehensive universal eating disorder prevention program. Body Image, 4, 289–295. http://dx.doi.org/10.1016/j.bodyim.2007.01. 004 Mendelson, B. K., Mendelson, M. J., & White, D. R. (2001). Body-esteem scale for adolescents and adults. Journal of Personality Assessment, 76, 90–106. http://dx. doi.org/10.1207/S15327752JPA7601 6 Mendelson, B. K., White, D. R., & Mendelson, M. J. (1995). Children’s global selfesteem predicted by body-esteem but not by weight. Perceptual and Motor Skills, 80, 97–98. http://dx.doi.org/10.2466/pms.1995.80.1.97 Mendelson, B. K., White, D. R., & Mendelson, M. J. (1996). Self-esteem and body esteem: Effects of gender, age, and weight. Journal of Applied Developmental Psychology, 17, 321–346. http://dx.doi.org/10.1016/S0193-3973(96)90030-1 Munsch, S., Roth, B., Michael, T., Meyer, A. H., Biedert, E., Roth, S.,. . . & Margraf, J. (2008). Randomized controlled comparison of two cognitive behavioral therapies for obese children: Mother versus mother-child cognitive behavioral therapy. Psychotherapy and Psychosomatics, 77, 235–246. http://dx.doi.org/10. 1159/000129659 Neumark-Sztainer, D. (2003). Obesity and eating disorder prevention: An integrated approach? Adolescent Medicine Clinics, 14, 159–173. Niemeier, B. S., Hektner, J. M., & Enger, K. B. (2012). Parent participation in weightrelated health interventions for children and adolescents: A systematic review and meta-analysis. Preventive Medicine, 55, 3–13. http://dx.doi.org/10.1016/j. ypmed.2012.04.021 O’Dea, J. A. (2004). Evidence for a self-esteem approach in the prevention of body image and eating problems among children and adolescents. Eating Disorders, 12, 225–239. http://dx.doi.org/10.1080/10640260490481438 Olvera, N., Leung, P., Kellam, S. F., Smith, D. W., & Liu, J. (2013). Summer and follow-up interventions to affect adiposity with mothers and daughters. American Journal of Preventive Medicine, 44, S258–S266. http://dx.doi.org/10.1016/j.amepre.2012. 11.018 Parkinson, K. N., Drewett, R. F., Le Couteur, A. S., Adamson, A. J., & Gateshead Millennium, S. (2012). Earlier predictors of eating disorder symptoms in 9-year-old children: A longitudinal study. Appetite, 59, 161–167. http://dx.doi.org/10.1016/ j.appet.2012.03.022 Pinard, C. A., Hart, M. H., Hodgkins, Y., Serrano, E. L., McFerren, M. M., & Estabrooks, P. A. (2012). Smart Choices for Healthy Families: A pilot study for the treatment of childhood obesity in low-income families. Health Education & Behavior, 39, 433–445. http://dx.doi.org/10.1177/1090198111425686 Rabeneck, L., Viscoli, C. M., & Horwitz, R. I. (1992). Problems in the conduct and analysis of randomized clinical trials – Are we getting the right answers to the wrong questions? Archives of Internal Medicine, 152, 507–512. http://dx.doi.org/ 10.1001/archinte.152.3.507 Rand, C. S. W., & Resnick, J. L. (2000). The “good enough” body size as judged by people of varying age and weight. Obesity Research, 8, 309–316. http://dx.doi. org/10.1038/oby.2000.37 Ransdell, L. B., Detling, N. J., Taylor, A., Reel, J., & Shultz, B. (2004). Effects of home- and university-based programs on physical self-perception in mothers and daughters. Women & Health, 39, 63–81. http://dx.doi.org/ 10.1300/J013v39n02 05 Resnick, E. A., Bishop, M., O’Connell, A., Hugo, B., Isern, G., Timm, A.,. . .& Geller, A. C. (2009). The CHEER study to reduce BMI in elementary school students: A schoolbased, parent-directed study in Framingham, Massachusetts. Journal of School Nursing, 25, 361–372. http://dx.doi.org/10.1177/1059840509339194 Rolland, K., Farnill, D., & Griffiths, R. (1996). Children’s perceptions of their current and ideal body sizes and body mass index. Perceptual and Motor Skills, 82, 651–656. http://dx.doi.org/10.2466/pms.1996.82.2.651 Rosenblum, G. D., & Lewis, M. (1999). The relations among body image, physical attractiveness, and body mass in adolescence. Child Development, 70, 50–64. http://dx.doi.org/10.1111/1467-8624.00005 Shelton, D., LeGros, K., Norton, L., Stanton-Cook, S., Morgan, J., & Masterman, P. (2007). Randomised controlled trial: A parent-based group education programme for overweight children. Journal of Paediatrics and Child Health, 43, 799–805. http://dx.doi.org/10.1111/j.1440-1754.2007.01150.x Shirley, K., Rutfield, R., Hall, N., Fedor, N., McCaughey, V. K., & Zajac, K. (2015). Combinations of obesity prevention strategies in US elementary schools: A critical review. Journal of Primary Prevention, 36, 1–20. http://dx.doi.org/10.1007/ s10935-014-0370-3 Skemp-Arlt, K. M. (2006). Body image dissatisfaction and eating disturbances among children and adolescents: Prevalence, risk factors, and prevention strategies. Journal of Physical Education, Recreation & Dance, 77, 45–51. http://dx.doi.org/ 10.1080/07303084.2006.10597813 Smolak, L. (2004). Body image in children and adolescents: Where do we go from here? Body Image, 1, 15–28. http://dx.doi.org/10.1016/s1740-1445(03)00008-1 Smolak, L., Levine, M. P., & Thompson, J. K. (2001). The use of the Sociocultural Attitudes Towards Appearance Questionnaire with middle school boys and girls. International Journal of Eating Disorders, 29, 216–223. http://dx.doi.org/10.1002/ 1098-108x(200103)29:2<216::aid-eat1011>3.0.co;2-v Stice, E., Hayward, C., Cameron, R., Killen, J., & Taylor, C. (2000). Body-image and eating disturbances predict onset of depression among female adolescents: A longitudinal study. Journal of Abnormal Psychology, 109, 438–444. http://dx.doi. org/10.1037//0021-843X.109.3.438 Stice, E., Shaw, H., & Marti, C. (2006). A meta-analytic review of obesity prevention programs for children and adolescents: The skinny on interventions that work. Psychological Bulletin, 132, 667–691. http://dx.doi.org/10.1037/0033-2909.132. 5.667

G. Eldridge et al. / Body Image 16 (2016) 143–153 Stice, E., Shaw, H., & Marti, C. (2007). A meta-analytic review of eating disorder prevention programs: Encouraging findings. Annual Review of Clinical Psychology, 3, 207–231. http://dx.doi.org/10.1146/annurev.clinpsy.3.022806.091447 Talamayan, K. S., Springer, A. E., Kelder, S. H., Gorospe, E. C., & Joye, K. A. (2006). Prevalence of overweight misperception and weight control behaviors among normal weight adolescents in the United States. The Scientific World Journal, 6, 365–373. http://dx.doi.org/10.1100/tsw.2006.70 Tiggemann, M. (2002). Media influences on body image development. In T. F. Cash & T. Pruzinsky (Eds.), Body image: A handbook of theory, research, and clinical practice (pp. 91–98). New York: Guilford. Truby, H., & Paxton, S. J. (2002). Development of the Children’s Body Image Scale. British Journal of Clinical Psychology, 41, 185–203. http://dx.doi.org/10.1348/ 014466502163967 United States Department of Agriculture. (n.d.). Retrieved from http://www. choosemyplate.gov

153

Verstuyf, J., Petegem, S. V., Vansteenkiste, M., Soenens, B., & Boone, L. (2013). The body perfect ideal and eating regulation goals: Investigating the role of adolescents’ identity styles. Journal of Youth and Adolescence, 43, 284–297. http://dx. doi.org/10.1007/s10964-013-9949-x Victora, C. G., Habicht, J. P., & Bryce, J. (2004). Evidence-based public health: Moving beyond randomized trials. American Journal of Public Health, 94, 400–405. http:// dx.doi.org/10.2105/ajph.94.3.400 Wilson, R., Latner, J., & Hayashi, K. (2013). More than just body weight: The role of body image in psychological and physical functioning. Body Image, 10, 644–647. http://dx.doi.org/10.1016/j.bodyim.2013.04.007 Wood, K., Becker, J., & Thompson, J. (1996). Body image dissatisfaction in preadolescent children. Journal of Applied Developmental Psychology, 17, 85–100. http:// dx.doi.org/10.1016/S0193-3973(96)90007-6