Peer, Parent, and Media Influences on the Development of Weight Concerns and Frequent Dieting Among Preadolescent and Adolescent Girls and Boys
Objective. To assess prospectively the influence of peers, parents, and the media on the development of weight concerns and frequent dieting.
Design. Prospective cohort study.
Setting. Questionnaires mailed annually to participants throughout the United States.
Participants. One-year follow-up of 6770 girls and 5287 boys who completed questionnaires in 1996 and 1997 and were between 9 and 14 years of age in 1996.
Main Outcome Measure. Onset of high levels of concern with weight and dieting frequently to control weight.
Results. During 1 year of follow-up, 6% of girls and 2% of boys became highly concerned with weight and 2% of girls and 1% of boys became constant dieters. Peer influence was negligible. Independent of age and body mass index, both girls (odds ratio [OR]): 1.9; 95% confidence interval [CI]: 1.1–3.1) and boys (OR: 2.7; 95% CI: 1.1–6.4) who were making a lot of effort to look like same-sex figures in the media were more likely than their peers to become very concerned with their weight. Moreover, both girls (OR: 2.3; 95% CI: 1.1–5.0) and boys (OR: 2.6; 95% CI: 1.1–6.0) who reported that their thinness/lack of fat was important to their father were more likely than their peers to become constant dieters.
Conclusions. Our results suggest that parents and the media influence the development of weight concerns and weight control practices among preadolescents and adolescents. However, there are gender differences in the relative importance of these influences.
The association between body weight of children and their parents is known to be influenced by both genetic and cultural factors. However, little is known about the relationship between the weight concerns and weight control practices of parents and their children. Concerns with weight and body shape and dieting to lose weight are common among preadolescent and adolescent girls1–6 and have been suggested as risk factors for eating disorders of at least subsyndromal severity.7 Although weight concerns are less common among males, recent data suggest that these concerns are becoming more prevalent8; however, little is known about how weight-related concerns arise in either gender.
The development of eating disorders is believed to be related to personal factors, family and peer influences and pressures, and sociocultural pressures. Stice et al9 have postulated that “internalization of the thin-ideal body image” for women in Western culture results in dissatisfaction with weight and shape, because the ideal body shape and weight are unattainable for most women. Moreover, family, peer, and media pressures to be thin also increase or sustain body dissatisfaction by repeatedly reminding women how far away they are from the ideal shape and weight.
The mass media are believed to encourage girls to form unrealistically thin body ideals,10–15 which are unattainable for most females, but their impact on boys has not been investigated. Several cross-sectional studies have observed a positive association between exposure to beauty and fashion magazines and higher levels of weight concerns or eating disorder symptoms in girls.16 The most likely mechanisms through which frequency of exposure to the mass media increases the likelihood that a girl becomes concerned with her weight or develops eating disorder symptoms are via encouraging girls to compare their bodies with those depicted in the media19and inspiring them to try to look like the unrealistic and unhealthy models and actresses in the media.
According to the theory of Stice et al,9 one should expect overweight females to be more dissatisfied with their weight because they are far away from the ideal weight and shape glorified by popular western culture. Thus, it is not surprising that several population-based studies have reported a positive association between weight and engaging in bulimic behaviors.20–24 Obesity is a large public health problem among both children and adults in the United States,25 thus, a relatively large proportion of youth may be at risk for developing weight concerns, dieting, and bulimic behaviors.
The media is not the only source of pressure to be thin. Levine et al26 suggested that weight control behaviors among young girls are modeled partially on their mothers' behavior. Although there is a paucity of data, several studies have observed that girls whose mothers diet and are concerned with their weight and shape are more likely than their peers to developing unhealthy weight control practices.27–28 Data are lacking on boys, but Smolak et al29 observed the comments by mothers had a larger impact than those by fathers and that daughters were more effected than were sons by the comments.
In addition to being affected by the comments and behaviors of their parents, preadolescents and adolescents are influenced by their peers. To gain acceptance by their peers, adolescents may resort to adopting the perceived beliefs and behaviors practiced by members of their peer group.30–33 Levine et al26 assessed the relationship of weight control practices of middle school girls and their peers. They observed that the weight control practices of the girls were significantly related to those of their peers. Moreover, peer pressure10 has been found to be a risk factor for developing bulimic behaviors and disordered eating but has not been assessed prospectively as a risk factor for becoming highly concerned with weight among girls and boys.
To prevent the development of unhealthy weights and weight control practices, it is important to know the factors associated with the development of high levels of concern with weight and the initiation of frequent dieting, a known risk factor for disordered eating.10,34 To determine the relative contribution of peer, parental, and media influences on the development of weight concerns and dieting practices among girls and boys, we analyzed data from ∼12 000 preadolescents and adolescents in a prospective cohort study.
The Growing Up Today Study (GUTS) was established in 1996 by recruiting children, who were 9 to 14 years of age, of women participating in the Nurses' Health Study II (NHS II).35Using the NHS II data, we identified mothers who had children 9 to 14 years of age. In total, we identified over 53 000 children in the eligible age range. We wrote a detailed letter to the mothers, explaining that the purpose of GUTS was to study the predictors of weight change during adolescence and sought parental consent to enroll their children. The mothers who responded to the invitational letter provided us with the names, age, and gender of 25 000 children in the eligible age range. In October 1996, we mailed letters and baseline questionnaires to the 13 261 girls and 13 504 boys whose mothers had granted us consent to invite them to participate in GUTS. The invitation letter to the child explained the study and asked them to complete the questionnaire if they wished to participate. The letter assured the children that no information they provided would be made available to anyone, including their parents. Approximately 68% of the girls (n = 9039) and 58% of the boys (n = 7843) returned completed questionnaires, thereby assenting to participate in the cohort. Because ongoing participation is crucial to the validity of cohort studies, we did not make efforts to increase baseline participation from the unwilling children (ie, the nonrespondents).
Mothers who gave us permission to contact their children were slightly less likely to smoke (8% vs 10%) and were slightly leaner (25.3 vs 25.7 kg/m2) and younger (37.7 vs 37.8 years) than were the women who did not grant us permission; however, the differences were quite minimal. In addition, among those given permission to participate, children who decided to participate had slightly leaner (25.2 vs 25.5 kg/m2) mothers than did children who did not return the baseline questionnaire. However, there was no difference in terms of age between ages of mothers of children who decided to join the study and those who declined to participate.
After 2 mailings of follow-up questionnaires and a reminder e-mail, ∼81% of girls (n = 7299) and 72% of boys (n = 5653) returned the 1997 questionnaire. Girls and boys who returned only the baseline questionnaire were similar to children who completed both questionnaires in terms of the prevalence of binge eating (girls: 2.2% vs 1.9%; boys: 1.1% vs .7%), constant dieting (girls: 3% vs 2.4%; boys: 1.3% vs. 1.1%), and overweight (girls: 17.9% vs 18.7%; boys: 25.0% vs 25.7%). However, girls who returned only the baseline questionnaire were more likely than were girls who completed both questionnaires to have been trying to lose weight in 1996 (36% vs 30%).
Information From Mothers (NHS II Participants)
Women participating in NHS II have been sent questionnaires biennially since 1989. Body mass index (BMI [kg/m2]) was calculated from self-reported weight and height information collected in 1995. Women were classified as weight cyclers if they had intentionally lost 10 or more pounds 3 or more times between 1989 and 1993. In 1995 women were asked whether they were likely to buy low-calorie foods and whether the following statements were true: “I eat anything I want, whenever I want,” “I pay a great deal of attention to changes in my figure.” From these 3 questions, a summary score from 0 (inattentive to diet, weight, and shape) to 3 (attentive to diet, weight, and shape) was created. The distribution was as follows 0 (3.3% girls, 3.5% boys), 1 (55.5% girls, 56.8% boys), 2 (31.4% girls, 31.3% boys), and 3 (9.8% girls, 8.4% boys).
Information From Girls and Boys
Questions adapted from the junior high school version of the McKnight Risk Factor Survey (MRFS)36 were used to measure weight concerns, attitudes, and behaviors. The MRFS for junior high school and high school students uses 5 response categories: never/not at all, a little, sometimes/pretty much, a lot, and always/totally. The 7 MRFS-based domains included on the questionnaire are concern with weight (4 questions), importance of thinness to peers (3 questions), importance of thinness to adults (3 questions), teasing and comments about weight by peers (8 questions), teasing and comments about weight by adults (3 questions), social eating (2 questions), and influence of the media (1 question).
Among females, thinness and low body fat are associated with one another, and both are considered to be positive attributes. Among males, however, the 2 are not necessarily related to one another, and they are not both considered to be positive attributes. As with females, low body fat is desirable, but thinness may be associated with weakness or femininity and, therefore, is not considered a desirable characteristic by young males. Therefore, to make the MRFS appropriate for boys, the questions on thinness were replaced with questions inquiring about the importance of not being fat or desiring not to be fat and the question assessing concern about gaining 2 lbs was deleted. One question on how often the mother had tried to lose weight during the past year and 2 others assessing the importance to each of the parents that the child be thin (or not fat for boys) were adapted from the MRFS.
The MRFS-based domain of concern with weight was modeled as the mean response on a scale of 1 to 5; other MRFS-based questions and domains were dichotomized. The concern with weight subscale was comprised of the following questions: “In the past year, how often have you felt fat?”, “In the past year, how often have you thought about wanting to be thinner”, “In the past year, how often have you worried about having fat on your body?”, and (for the girls), “In the past year, how much have you worried about gaining 2 pounds?” Thinness/lack of fatness was considered very important (to parent or peer) if the child indicated the degree of importance as “a lot” or “totally.” Because so few boys reported that lack of fat was very important, the cutoff was lowered to include boys who reported it was “pretty much” important. Girls and boys were considered to be making a lot of effort to look like females/males in the media if they responded “a lot” or “totally” to the question, “In the past year, how much effort have you made to look like the girls or women (boys or men) you see on television, in movies, or in magazines?” Mothers were classified as frequent dieters if their child reported that during the last year the mother was trying to lose weight “a lot” or “always.”
The parent influence questions included on the girls questionnaire were: “During the past year how important has it been to your father that you be thin?” (“Father means the adult man in your life who acts most like a father to you”), and “During the past year how important has it been to your mother that you be thin?” Boys were asked, “During the past year how important has it been to your father that you not be fat?” (“Father means the adult man in your life who acts most like a father to you”), and “During the past year how important has it been to your mother that you not be fat?” The questions were assessed individually as well as combined into a measure of importance to both parents.
Dieting was assessed with a question adapted from the Youth Risk Behavior Surveillance System questionnaire.37 The questions asks, “During the past year, how often did you diet to lose weight or to keep from gaining weight?” The response categories for the frequency of dieting to control weight during the past year were: “never,” “less than once a month,” “1 to 3 times a month,” “once a week,” “2 to 6 times per week,” or “every day.”
We calculated BMI using self-reported weight and height information collected in 1996 and computed quintiles of gender- and age-specificz score of BMI for use in analyses.
Reliability and Validity of Self-Report Measures: Weight Concerns and Influences, Height, and Weight
The test–retest correlations and Cronbach's αs of the 7 MRFS-based domains included on the questionnaire have been evaluated among girls in junior high school. Both were found to be moderate to high for the subscales relevant to our study: concern with weight (test–retest = .84; Cronbach's α = .86), influence of the media (test–retest = .66; only 1 question so no Cronbach's α), teasing or comments about weight by peers (test–retest = .76; Cronbach's α = .85), teasing or comments about weight by adults (test–retest = .63; Cronbach's α = .67), importance of thinness to peers (test–retest = .62; Cronbach's α = .54), importance of thinness to parents (test–retest = .59; Cronbach's α = .75), and change in eating patterns around peers (test–retest = .77; Cronbach's α = .63).36
The validity of self-reported weight and height among preadolescents and adolescents has been investigated by several groups of researchers. Shannon et al,38 in a sample of 806 sixth graders, report moderately high correlations for weight and height for both boys (r = .90; r = .74) and girls (r = .84; r = .62). More children tended to underreport than to overreport their weight, with the taller and heavier children and girls showing the greatest tendency to do so. We did not find a systematic bias in reported heights. Among 1657 adolescents 12 to 16 years of age in the National Health and Nutrition Examination Survey III study, Strauss39 observed high correlations between self-reported weight and actual weight (r = .87 and .94, depending on gender or race) and self-reported height and actual height (ranged from r = .82–.91). Moreover, the use of self-reported weight and height resulted in the correct classification of weight status in 94% of children. In addition, in a nationally representative sample of youth from the National Longitudinal Study of Adolescent Health, Goodman et al40 observed that the correlation between BMI calculated form self-report versus measured height and weight was .92 and only 3.8% of the youth were misclassified as obese using self-reported BMI. Thus, the results suggest that preadolescents and adolescents provide information on weight and height that is as valid as the information provided by adults.
The 2 outcomes were the incidence of high level of weight concerns and the incidence of constant dieting. Both outcomes were defined with data provided by the girls and boys.
Constant dieting to lose weight was defined as reporting to always (ie, every day) be on a diet to lose weight. High level of concern with weight was defined as a mean score on the MRFS weight concern subscale of at least 4, corresponding to “a lot” or “totally.”
Participants included 7299 girls and 5653 boys who completed questionnaires in 1996 and 1997, excluding those whose mothers did not complete NHS II questionnaires in 1989, 1993, or 1995 (245 girls and 179 boys), and participants who did not complete the questions on weight concerns or dieting in 1996 (26 girls and 24 boys) or 1997 (14 girls and 19 boys), were younger than 9 or older than 14 years old in 1996 (138 girls and 70 boys) or provided no or implausible information on height or weight for height (106 girls and 83 boys), thus leaving 6770 girls and 5287 boys for analysis. Children who did not report the importance of thinness to their friends (26 girls), how much effort they make to look like same sex figures in the media (33 girls and 73 boys), how important their thinness (or lack of fatness) is to their father (16 girls and 27 boys) or mother (41 girls and 18 boys), or how often their mother diets (65 girls and 38 boys) were also excluded from multivariate analyses. Children whose mother did not provide information on weight cycling (1534 girls and 1258 boys) or did not complete the 3 questions on concern with weight and shape (823 girls and 593 boys) were excluded from specific analyses.
All analyses were stratified by gender and conducted with SAS software.41 To account for the correlation between siblings, generalized estimating equations, using the logit link, were used for all multivariate analyses (SAS Proc Genmod, Cary, NC). Children who were cases at baseline (ie, who were highly concerned with weight or constant dieters in 1996) were excluded from the relevant analyses. All statistical models adjusted for age and age-specificz score of BMI. In addition, all models predicting weight concerns in 1997 adjusted for level of concern with weight in 1996 (1, 2, or 3—children with values of 4 and 5 were excluded as prevalent cases). Likewise, all models predicting becoming a constant dieter adjusted for frequency of dieting in 1996 (“never,” “less than once a month,” “1–3 times a month,” “once a week,” or “2–6 times per week,” children reporting “every day” were excluded as prevalent cases). All P values are 2-sided, withP < .05 considered statistically significant.
In 1996, ∼9% of girls and 4% of boys were highly concerned with their weight and 2% of girls and 1% of boys reported always being on a diet to lose or maintain weight. Over 1 year of follow-up, 6% of girls and 2% of boys became highly concerned with weight and 2% of girls and 1% of boys became constant dieters.
Cumulative incidence of becoming highly concerned with weight (Fig 1) and starting to diet constantly (Fig 2) increased across the quintiles of BMI in 1996. Associations with BMI were stronger among boys (odds ratio [OR]: 2.2–2.6; P < .001) than among girls (OR: 1.2–1.4; P < .01) because outcomes were very rare among boys in lower quintiles of BMI in 1996.
The children's BMI was associated with the mother's BMI (Spearmanr = .29 for girls; r = .30 for boys). After control for the child's BMI, the mother's BMI was not significantly associated with either outcome (data not shown). Among girls, but not boys, age was modestly associated with the development of constant dieting (OR: 1.2 per year of age; 95% confidence interval [CI]: 1.0–1.4).
Girls who reported that it was important to either parent that they were thin were twice as likely as their peers to become highly concerned with weight (OR: 2.1; 95% CI: 1.1–4.0; Table 1), whereas among the boys the development of concern with weight was unrelated to the perceived importance to the mother or the father that the boy not be fat. Peer influences were not predictive in either gender of becoming highly concerned with weight. Girls (OR: 1.9; 95% CI: 1.1–3.1; Table 2) and boys (OR: 2.7; 95% CI: 1.1–6.4;Table 3) who reported making a lot of effort to look like same-sex figures in the media were more likely to become weight concerned.
Girls (OR: 2.3; 95% CI: 1.1–5.0; Table 2) and boys (OR: 2.6; 95% CI: 1.1–6.0; Table 4) who reported that it was important to their father that they be thin (girls)/not fat (boys) were more likely than their peers to become constant dieters. The child's perception that it was important to the mother that the child be thin/not fat was unrelated to development of dieting (data not shown). There was a suggestion that girls whose mothers reported being attentive to their diet, weight, and shape were less likely to become constant dieters (OR: .8; 95% CI: .6–1.0); however, the difference was not significant (P = .07).
Weight concerns, which developed among girls across the weight spectrum, were less common, but more strongly related to BMI, among boys. BMI was similarly associated with the development of constant dieting. The associations were stronger among the boys because both outcomes were very rare among boys, but not girls, in the lower quintiles of BMI in 1996. Thus, boys, but not girls, who became highly concerned with their weight or began to diet constantly may have been partially justified in their concerns and actions because of their excessive weight. Among both girls and boys, constant dieting was relatively rare, but approximately twice as many girls became constant dieters over the 1-year follow-up.
To our knowledge, this is the first study to assess the relative importance of peer, parental, and media influences on the development of weight concerns and frequent dieting among both girls and boys. Independent of age and BMI, parental influences were predictive of becoming highly concerned with weight and becoming a constant dieter. Among both girls and boys, making considerable effort to look like same-sex figures in the media was predictive of developing weight concerns and becoming a constant dieter.
The results suggest that children who perceive that their mother is frequently trying to lose weight were more likely to become highly concerned with weight (boys) or constant dieters (girls), thus offering support for the role mothers play in the transmission of cultural values about desirable body weight and shape.27–29 An unexpected finding was that the child's perception of the importance of thinness/lack of fatness to the father was more important than their perception of its importance to the mother. It is unclear whether the child's perceptions are accurate reflections of the parents' beliefs; however, regardless of their accuracy, further study is needed to understand the role of fathers in the development of weight concerns and weight control behaviors.
Our results are not in complete agreement with those of Stice42 who assessed the influence of peers, family, and the media on the development of bulimic behaviors among 218 adolescent girls. Similar to our findings, he did observe a significant effect of family influences; however, in contrast to our findings, he observed that peers but not the media were predictive of developing bulimic behaviors. There are several possible explanations for the discrepancies between his findings and those that we observed. First, his sample was older and it is possible that peer influences are more important among adolescents than preadolescents. Second, our sample was comprised of children throughout the United States, whereas his sample was comprised of girls who were seniors in 2 high schools in an urban southwestern location. It is possible that the prevalence of weight concerns and bulimic behaviors vary by region and that the determinants of these outcomes show geographic variation. Future research is needed to address these issues. Another possible explanation for the discrepancy in results is that our sample was much larger; thus, we had more statistical power to pick up differences in media influence.
The media have been criticized for promoting excessively thin body images and playing a role in the development of weight concerns and disordered eating among females.10 The impact of the media on weight and shape concerns and weight control behaviors of boys has not been studied in detail. If the barrage of images of females with perfect bodies has a deleterious effect on young girls, it seems plausible that the recent increase in the number of media images of males baring their carefully sculpted chests or chests devoid of any body fat may have a similar negative impact on boys. More research is needed to better understand media influences.
A strength of the study is that both self-reported weight38–40 and the questions on weight concerns and influences36 have been validated among preadolescents and adolescents. Moreover, a prospective design and relatively high follow-up rate guard against the effect of biased sampling on the outcome, which can result in incorrect inference. However, there are several limitations to the present study. The information on the importance to parents that their daughter is thin or that their son is not fat was collected from the participants. Thus, it is possible that the child's perceptions are inaccurate reflections of the parents' beliefs. Another limitation is that we did not measure hours per week exposed to various media sources. Thus, our measure should be interpreted as influence of the media, not exposure to the media.
Overall, our data suggest that weight-related issues of parents are transmitted to their children. Given the substantial health risks associated with overweight, the weight concerns of some of the parents may be justifiable. However, to maintain a healthy weight and to ensure that their children do not become either overweight or overly concerned with weight, it is advisable that parents remind themselves that they serve as role models and, therefore, should attempt to adopt the diet and activity patterns that they would like their children to emulate. Because, like obesity, excessive weight concerns are hard to treat, efforts should be made to prevent these conditions. Our results suggest that to prevent children and adolescents from developing extreme concern with weight and unhealthy weight control behaviors, peers, parents, and the media must be targeted for intervention.
Drs Field and Colditz were partially supported by Grant DK 46200 from the Boston Obesity Nutrition Research Center. Additional funding was provided by Research Grant DK-46834 from the National Institutes of Health, and First Independent Research Support and Training Award (R29) Grant HL57871-01 from the National Institutes of Health, and the Kellogg Company.
We thank Susan Bryson for her helpful comments and suggestions on the statistical analysis and interpretation.
- Received November 4, 1999.
- Accepted May 23, 2000.
Reprint requests to (A.E.F.) Channing Laboratory, 181 Longwood Ave, Boston, MA 02115. E-mail:
- GUTS =
- Growing Up Today Study •
- NHS II =
- Nurses' Health Study II •
- BMI =
- body mass index •
- MRFS =
- McKnight Risk Factor Survey •
- OR =
- odds ratio •
- CI =
- confidence interval
- Casper RC,
- Offer D
- Moses N,
- Banilivy MM,
- Lifshitz F
- Field AE, Cheung L, Wolf AM, Herzog DB, Gortmaker SL, Colditz GA. Exposure to the mass media and weight concerns among girls.Pediatrics 1999;103(3). URL:http://www.pediatrics.org/cgi/content/full/103/3/e36
- Harrison K,
- Cantor J
- Patton GC,
- Selzer R,
- Coffey C,
- Carlin JB,
- Wolfe R
- Goodman E,
- Hinden B,
- Khandelwal S
- SAS Institute Inc. SAS User's Guide: Statistics. 4th ed. Cary, NC: SAS Institute; 1990
- Copyright © 2001 American Academy of Pediatrics