Weight-Related Concerns and Behaviors Among Overweight and Nonoverweight Adolescents:  Implications for Preventing Weight-Related Disorders FREE

OBESITY AMONG children and adolescents has reached epidemic proportions in the United States. One of the Healthy People 2010 objectives is to "reduce the proportion of children and adolescents who are overweight or obese."¹(chap 19, p13) The aim is to decrease the percentage of youth (aged 6-19 years) with body mass index (BMI) values above the 95th percentile from 11% to 5%.¹ However, there is no evidence that the prevalence of obesity is decreasing; rather, data^2- 3 suggest a continuing upward trend. The high prevalence of obesity among youth and its potentially serious psychosocial and physical consequences⁴ have made obesity one of the greatest contemporary public health issues. Numerous questions exist regarding effective strategies for preventing and treating obesity among youth.^5- 11

Another public health concern related to eating and weight is the high prevalence of body dissatisfaction, unhealthy weight control behaviors, and other disordered eating patterns, particularly among adolescent girls.^12- 18 Excessive weight-related concerns and behaviors have potentially serious consequences for youth in their impact on psychosocial development, dietary intake, physical growth, and the development of eating disorders.^19- 25 It could be argued that we need not be concerned about weight-related concerns and behaviors in light of more pressing concerns regarding obesity. However, there is reason for concern if nonoverweight youth are attempting weight loss and if unhealthy weight control behaviors are being used by youth, regardless of their weight status. As with obesity prevention, questions exist regarding the most effective strategies for preventing excessive weight concerns and unhealthy weight control/disordered eating behaviors.^26- 29

To date, the fields of obesity prevention and dieting/eating disorder prevention have been quite separate. However, it could be argued that it is possible, and practical, to view these issues as overlapping and to work toward developing integrative interventions that address the broad range of weight-related disorders. To develop effective interventions aimed at preventing obesity and unhealthy dieting/disordered eating, a greater understanding of the types of weight-related concerns and behaviors among youth is needed. In developing interventions for obesity prevention, it is particularly important to be cognizant of the specific weight-related concerns and behaviors among overweight youth.

The present study was designed to (1) examine specific weight-related concerns and behaviors in a large population-based sample of adolescent girls and boys; (2) compare weight-related concerns and behaviors among underweight, nonoverweight, moderately overweight, and very overweight adolescents; and (3) explore the overlap between overweight status and unhealthy weight-related behaviors and consider implications for interventions.

The overall study population consists of 4746 adolescents from 31 public middle schools and high schools from urban and suburban school districts in the St Paul and Minneapolis, Minn, area. The mean age of the study population was 14.9 years (SD, 1.7 years); 34% were in junior high school and 66% were in high school. The racial/ethnic background of the participants was as follows: 48% white, 19% African American, 19% Asian American, 6% Hispanic, 4% Native American, and 4% mixed or other. Participants were equally divided by sex.

Data for the present study were drawn from Project EAT (Eating Among Teens), a comprehensive study of adolescent nutrition and obesity. Trained research staff administered surveys within school classes and assessed height and weight within a private area. Study procedures were approved by the University of Minnesota Human Subjects' Committee and by the research boards of the participating school districts. Consent procedures were done in accordance with the requests of the participating school districts; in some schools, passive consent procedures were used, while in others, active consent procedures were required. The response rate for student participation was 81.5%; the main reasons for lack of participation were absenteeism and failure to return consent forms within schools requiring active consent.

The Project EAT survey is a 221-item self-report instrument assessing a range of factors of potential relevance to nutritional health and obesity among adolescents. The development of the survey was guided by focus group discussions with youth,³⁰ a theoretical framework (Social Cognitive Theory) for understanding factors influencing eating behavior,^31- 32 a review of the literature for existing instruments,^33- 38 numerous reviews by professionals from different disciplines and adolescents with different backgrounds, and several pilot tests of the survey.

Weight status was based on height and weight measurements taken by trained research staff using standardized equipment and procedures. Body mass index values were calculated according to the following formula: weight in kilograms divided by the square of height in meters. Sex- and age-specific cutoff points were based on reference data from the Centers for Disease Control and Prevention growth charts.^39- 40 For this study, respondents were classified as underweight (BMI, <15th percentile), average weight (BMI, 15th-<85th percentile), moderately overweight (BMI, 85th-<95th percentile), and very overweight (BMI, ≥95th percentile). Among the adolescent girls, the mean BMI values were 16.6, 20.9, 26.0, and 32.9 for the underweight, average-weight, moderately overweight, and very overweight categories, respectively. Among the adolescent boys, the mean corresponding BMI values were 17.0, 20.9, 25.1, and 31.6.

Weight-related concerns assessed included the following: perceived weight status, weight disparity, body satisfaction, and care about controlling weight. Perceived weight status was assessed with the following question: "At this time do you feel that you are . . . (1) very underweight, (2) somewhat underweight, (3) about the right weight, (4) somewhat overweight, or (5) very overweight?" Weight disparity was based on 2 questions ("How much do you weigh?" and "At what weight do you think you would look best?") and was calculated as desired weight as a percentage of reported weight. Responses of less than 100% indicate that one's desired weight was lower than one's reported weight, while responses greater than 100% indicate that one's desired weight was higher than one's reported weight. Body satisfaction was assessed with a modified version of the Body Shape Satisfaction Scale,³⁵ which included 10 items assessing satisfaction with different body parts (eg, height, weight, stomach, and hips), with 5 Likert response categories ranging from "very dissatisfied" to "very satisfied" (Cronbach α = .92). Responses were categorized as low, moderate, and high based on distributions within the study population, with one third of the population in each category. Care about controlling weight was assessed with the following question: "How much do you care about controlling your weight . . . (1) not at all, (2) a little bit, (3) somewhat, or (4) very much?"

Several weight-related behaviors were assessed in this study. Currently trying to lose weight was assessed with the following question: "Are you currently trying to . . . (1) lose weight, (2) stay the same weight, (3) gain weight, or (4) I am not trying to do anything about my weight?" Trying to lose or maintain weight during the past year was assessed with the following question: "During the past year, have you done anything to try to lose weight or keep from gaining weight?" Long-term dieting (≥5 times per year) was assessed with the following question: "How often have you gone on a diet during the last year? By ‘diet,' we mean changing the way you eat so you can lose weight . . . (1) never, (2) 1 to 4 times, (3) 5 to 10 times, (4) more than 10 times, or (5) I am always dieting." Extreme weight control practices during the past week were assessed with the following question: "During the past week, did you do any of the following to lose weight or keep from gaining weight . . . (1) made myself vomit (throw up), (2) took diet pills, or (3) used laxatives?" Respondents using any of these methods were categorized as using extreme methods in the past week. Respondents were also asked if they had ever been told they had an eating disorder by a health professional: "Has a doctor ever told you that you have an eating disorder such as anorexia nervosa, bulimia nervosa, or binge-eating disorder? (yes or no)." Binge eating was assessed with the following question: "In the past year, have you ever eaten so much food in a short period of time that you would be embarrassed if others saw you (binge eating)? (yes or no)." Healthy, unhealthy, and extreme weight control behaviors during the past year were assessed with the following question: "Have you done any of the following things in order to lose weight or keep from gaining weight during the past year? (yes or no for each method)." Responses classified as healthy weight control behaviors included (1) exercised, (2) ate more fruits and vegetables, (3) ate less high-fat foods, and (4) ate less sweets. Responses classified as unhealthy weight control behaviors included (1) fasted, (2) ate very little food, (3) used a food substitute (powder or a special drink), (4) skipped meals, and (5) smoked more cigarettes. Responses classified as extreme weight control behaviors included (1) took diet pills, (2) made myself vomit, (3) used laxatives, and (4) used diuretics.

Sex, school level, ethnicity/race, and socioeconomic status were based on self-report. The prime determinant of socioeconomic status was parental educational level, defined by the higher level of educational attainment of either parent. Other variables taken into account in assessing family socioeconomic status included family eligibility for public assistance, eligibility for free or reduced-cost school meals, and employment status of the mother and father.

Percentages of youth reporting weight-related concerns and behaviors were examined across sex and weight status among adolescent girls and boys separately. χ² Values were calculated by analysis-of-variance models that included the school as a random component of variance. Selected outcome variables were dichotomized to investigate, separately by sex, the associations with weight status, adjusted for sociodemographic characteristics. Logistic regression models, which included the school as a random effect and the average-weight group as a reference group, were used to generate adjusted odds ratios and 95% confidence intervals for weight status. Odds ratios are statistically significant (P<.05) when 1.0 is not included in the 95% confidence interval. All analyses were carried out using SAS statistical software, release 8.0.⁴¹

There were modest, albeit statistically significant (P<.001), differences in weight status across sex. Among adolescent girls, 4.6% were underweight, 62.8% were average weight, 20.0% were moderately overweight, and 12.6% were very overweight. Among adolescent boys, 5.7% were underweight, 63.1% were average weight, 14.6% were moderately overweight, and 16.6% were very overweight. In contrast, as seen in Table 1, large differences were noted for nearly all of the weight-related concerns and behaviors. Adolescent girls expressed greater concerns and were more likely to report engaging in weight-related behaviors than adolescent boys; all differences were statistically significant (P<.001), except for ever told had an eating disorder (P = .28), laxative use (P = .02), and diuretic use (P = .79). Among both sexes, healthy weight control practices were commonly reported; 85.4% of the adolescent girls and 69.9% of the adolescent boys reported healthy practices during the past year. However, 56.9% of the adolescent girls and 32.7% of the adolescent boys reported unhealthy practices and 12.4% of the adolescent girls and 4.6% of the adolescent boys reported extreme practices (Table 1). While adolescent girls were more likely than adolescent boys to express weight-related concerns and to engage in weight-related behaviors, significant numbers of adolescent boys also reported these concerns and behaviors.

For most adolescent girls, perceived weight status was in accordance with actual weight status (Table 1). For example, 81.2% of the very overweight adolescent girls and 68.3% of the moderately overweight adolescent girls perceived themselves as somewhat overweight or very overweight. However, there were some notable exceptions in which perceived weight status differed from actual weight status. Weight disparity was strongly correlated with weight status: high percentages of very overweight adolescent girls (86.5%) desired to weigh less than 90% of their self-reported weight, and high percentages of underweight adolescent girls (68.1%) desired to weigh more than their self-reported weight. Still, a high percentage of the average-weight adolescent girls (63.8%) desired to weigh less than their self-reported weight. Body satisfaction was inversely associated with weight status. Caring about controlling weight (somewhat or very much) was reported by high percentages of average, moderately overweight, and very overweight adolescent girls; percentages of underweight adolescent girls who reported caring were considerably lower. Strong correlations were found between weight status and nearly all of the weight-related behaviors among the adolescent girls; the highest use was reported by the very overweight adolescent girls, and the lowest use was reported by the underweight adolescent girls. Among very overweight adolescent girls, 95.4% reported healthy weight control practices, 76.0% reported unhealthy practices, and 17.9% reported extreme practices during the past year. Among adolescent girls, the only differences that were not statistically significant across weight status were for ever told had an eating disorder (P = .77), vomited (P = .19), and took diuretics (P = .90). Differences were statistically significant for binge eating (P = .03), fasted (P = .002), smoked more cigarettes (P = .04), and took laxatives (P = .02). All other differences were highly statistically significant (P<.001).

Among the adolescent boys, there were also strong correlations between weight-related concerns and behaviors and weight status (Table 1), although some of the trends differed from those found among the adolescent girls. Most very overweight adolescent boys (75.2%) perceived themselves as somewhat or very overweight, but only 38.3% of the moderately overweight adolescent boys perceived themselves in this way. Furthermore, in contrast to the adolescent girls, only a small percentage of the average-weight adolescent boys perceived themselves as somewhat or very overweight (6.6%), while 23.9% perceived themselves as underweight. Weight disparity was strongly associated with weight status among the adolescent boys, as was found among the adolescent girls. However, in contrast to the adolescent girls, nearly half (49.2%) of the average-weight adolescent boys indicated that they wanted to weigh more than their self-reported weight. Body satisfaction was lowest among the very overweight adolescent boys, followed by the underweight adolescent boys. This was a different pattern from that found among the adolescent girls (in that underweight adolescent girls expressed the highest level of body satisfaction). Caring about controlling weight seemed to be highest among the moderately and very overweight adolescent boys. For most of the weight-related behaviors, there were direct associations with weight status; very overweight adolescent boys were most likely to report their use, followed by moderately overweight adolescent boys. However, for some of the more extreme weight control practices, there seemed to be a U-shaped association with weight status, and their use tended to be highest among very overweight and underweight adolescent boys. These associations were further explored in multivariate analyses (described later). Among adolescent boys, all differences across weight status were statistically significant (P<.01 or P<.001), except for ever told had an eating disorder (P = .12), smoked more cigarettes (P = .80), took laxatives (P = .80), took diuretics (P = .25), and took diet pills (P = .03).

Associations between weight-related concerns and behaviors and weight status were further examined, controlling for school level, race, and socioeconomic status. Using average-weight youth as the reference group, odds ratios and 95% confidence intervals were calculated for adolescent girls (Table 2) and boys (Table 3). In general, patterns were similar to those found in the bivariate analyses previously described. Among adolescent girls and boys, compared with average-weight youth, overweight youth were more likely to perceive themselves as overweight, desire to weigh less, and express body dissatisfaction. Similarly, overweight youth were more likely to engage in weight control practices and binge eating than average-weight youth. This pattern was apparent between both sexes.

Underweight adolescent girls were less likely to report weight-related concerns and behaviors than average-weight adolescent girls (Table 2). Similarly, underweight adolescent boys were less likely than average-weight adolescent boys to express several weight-related concerns (eg, care about controlling weight) and to engage in general weight control practices (eg, tried to lose or maintain weight in the past year) and healthy weight control practices (Table 3). However, compared with average-weight adolescent boys, underweight adolescent boys were at greater risk for low body satisfaction and extreme weight control behaviors in the past week.

This study aimed to assess weight-related concerns and behaviors in a large sample of adolescents and to compare these concerns and behaviors across sex and weight status. In light of the high and rapidly increasing prevalence of obesity among youth,^1- 3 we were particularly interested in examining weight-related concerns and behaviors among overweight youth to guide the development of appropriate prevention and treatment interventions.

Adolescents in the study population, particularly girls, reported a high prevalence of weight-related concerns and behaviors. It was encouraging that healthy weight control practices (eg, decreasing fat intake) were more commonly reported than practices considered to be unhealthy (eg, skipping meals) or extreme (eg, self-induced vomiting). Many of the healthy weight control behaviors (eg, increasing physical activity), used judiciously, are recommended for all adolescents, regardless of their weight status.¹ There is considerable debate about whether we need to be concerned about the high prevalence of dieting behaviors among adolescents, particularly adolescent girls.²⁷ Previous studies^{12,33,42- 43} have included general questions about dieting and/or weight loss attempts, and youth may define dieting or trying to lose weight differently,^44- 45 making interpretations from such studies somewhat difficult. The finding that many youth in the present study reported using specific healthy weight control behaviors might suggest that we need not be overly concerned with the high prevalence of self-reported weight control attempts found across studies. However, in addition to healthy weight control behaviors, many nonoverweight and overweight youth reported engaging in unhealthy and extreme weight control behaviors. Furthermore, a high percentage of average-weight girls perceived themselves as overweight, desired to weigh less, and expressed body dissatisfaction. These findings suggest that we do need to be concerned about the high prevalence of dieting among youth, particularly among adolescent girls.

Many of the overweight youth in the present study accurately perceived themselves as overweight, reported that they cared about controlling their weight, and had engaged in healthy weight control behaviors during the past year. These findings suggest that most overweight teenagers are motivated to achieve a healthier weight and have made some positive steps toward that goal. This suggests that, for most youth, little time within interventions needs to be devoted toward identifying who is overweight, discussing the importance of weight control, and superficially reviewing desirable weight control behaviors. However, time may need to be devoted to skill development in learning how to successfully engage in healthy weight control behaviors, such as increasing physical activity and increasing fruit and vegetable intake, decreasing fat intake, and decreasing sweets in one's diet. Previous studies^1,46- 49 among youth have found high levels of fat intake, low levels of physical activity, and low levels of fruit and vegetable intake. Therefore, many adolescents who report the use of these behaviors for weight control purposes may not be implementing them adequately.

The high prevalence of binge eating and unhealthy or extreme weight control behaviors exhibited by overweight youth (especially by overweight girls) demonstrates a need to also address these behaviors within interventions for overweight youth. Individual factors associated with binge eating should be identified (eg, excessive dietary restraint), and alternative behaviors for dealing with triggers to binge eating should be developed. The potential dangers of unhealthy or extreme weight control behaviors and their ineffectiveness for long-term weight control should be discussed. Previous studies^50- 51 have found that body dissatisfaction is a strong predictor of unhealthy weight control practices, suggesting that to decrease the use of unhealthy weight control behaviors, it may be important to develop interventions that simultaneously aim for improved body satisfaction and the development of realistic weight goals.

While most overweight youth seemed motivated to achieve a healthier weight through healthy means, this was not always the case. For example, about one fifth of the very overweight girls and one fourth of the very overweight boys did not perceive themselves as somewhat or very overweight, suggesting that they may not be interested in losing weight. This finding supports the value of assessing readiness for change and developing an appropriate intervention plan.⁵² For the few overweight youth who do not perceive a need for weight control, the focus will need to be on encouraging an accurate awareness of the individual's weight status and eating and physical activity habits and on getting the individual to understand the potential harms of obesity and the benefits of adopting healthy lifestyle behaviors to prevent unhealthy weight gain. As the adolescent perceives a need to change, the focus can be shifted toward increasing self-efficacy and skill development for behavioral change. Family and peer support networks may be necessary for overweight youth, regardless of the stage of readiness for change in behaviors.

This study had several strengths that enhance our ability to draw conclusions from the findings. The population-based nature of the sample allows for more generalization than would be possible from a clinic-based sample of youth. Another strength was the collection of actual height and weight measurements; most large population-based studies of youth rely on self-reported height and weight data^33,53 or do not even include questions assessing height and weight.⁵⁴ A major strength of the study was the assessment of various weight-related concerns and behaviors. Typically, population-based surveys of youth address a broad range of health-related issues; therefore, questions regarding weight-related concerns and behaviors have been quite general.^{33,53,55- 56} However, questions still remain unanswered regarding the frequency and the intensity of specific weight control behaviors reported. Furthermore, as in any study assessing self-reported behaviors, questions of validity arise. Finally, because of the cross-sectional nature of the study, we can only state that overweight status was associated with increased unhealthy weight control behaviors. We cannot determine whether these behaviors led to a higher prevalence of obesity or whether overweight youth chose to engage in these behaviors in an attempt to lose weight.

Traditionally, interventions aimed at preventing and treating obesity focus on increasing physical activity and decreasing calorie (energy) intake. Interventions aimed at preventing disordered eating/eating disorders tend to focus more on promoting a positive body image and on dieting prevention. Recently, there have been discussions about decreasing existing disparities between the obesity and eating disorder fields. These discussions have arisen because of an increased recognition of some of the similarities between these conditions and a need for taking both into account in the development of interventions, particularly primary prevention interventions. Our findings suggest a need for identifying and addressing shared predictive factors for obesity and disordered eating behaviors (ie, unhealthy weight control and binge-eating behaviors), because for many youth they co-occur. Models examining shared predictive factors for disordered eating/eating disorders and obesity need to be developed and tested. Different intervention strategies that address the broad spectrum of weight-related concerns and behaviors need to be developed, implemented, and evaluated. For example, within interventions for overweight adolescents, there is a need to address body image issues and unhealthy dieting behaviors in addition to focusing on changes in eating and physical activity behaviors. In interventions aimed at preventing disordered eating/eating disorders, there may be a need to discuss and provide skills for healthful weight management. Through the development, evaluation, and refinement of such programs, it may be possible to decrease the high prevalence of disordered eating and obesity.

Accepted for publication September 19, 2001.

This study was supported by grant MCJ-270834 from the Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Service Administration, US Department of Health and Human Services, Rockville, Md (Dr Neumark-Sztainer).

We thank the students and staff from the St Paul and Osseo, Minn, school districts for participating in this study.

Studies suggest that weight-related concerns and behaviors, including obesity, body dissatisfaction, and healthy and unhealthy weight control behaviors, are prevalent among adolescents. However, large population-based studies of youth are limited in their ability to explore interrelations between weight status and weight-related behaviors because of inadequate assessments of these variables. In the present study, measured height and weight and a range of weight-related concerns were assessed in a large and diverse population of adolescents. The high prevalence of overweight youth reporting weight control behaviors indicates that most overweight youth are aware that they are overweight, but may need intensive interventions to provide them with adequate skills to engage in healthy eating and increased physical activity. The high prevalence and co-occurrence of obesity and unhealthy weight control behaviors found in this study suggest that there is a need for interventions that address the broad spectrum of weight-related disorders.

Corresponding author and reprints: Dianne Neumark-Sztainer, PhD, MPH, RD, Division of Epidemiology, School of Public Health, University of Minnesota, 1300 S Second St, Suite 300, Minneapolis, MN 55454.