Problem Dieting Behaviors Among Young Adolescents

Objective  To examine dieting, eating and exercise behaviors, use of diet pills, and vomiting or use of laxatives to lose weight among younger adolescents.

Design  Analysis of data from a modified version of the Youth Risk Behavior Survey administered to middle school students in North Carolina in 1995.

Setting  Fifty-three randomly selected middle schools in North Carolina.

Subjects  Two thousand three hundred thirty-one students in the sixth, seventh, and eighth grades.

Interventions  None.

Main Outcome Measures  Responses to questions regarding weight control practices, including vomiting or laxative use, dieting, exercise, or diet pill use.

Results  Of the students surveyed, 110 (9.7%) of the girls and 46 (4.0%) of the boys reported vomiting or using laxatives to lose weight. Among the girls, vomiting or laxative use was associated with feeling overweight, other weight loss practices, older age, being a poor student, smoking, eating more salads or vegetables, and eating more candy or other sweets (P≤.01). A logistic regression model consisting of diet pill use, dieting to lose weight, lower academic achievement, and currently trying to lose weight correctly classified 92% of female students who had or had not vomited or used laxatives. Among boys, vomiting or laxative use was associated with feeling overweight, other weight loss practices, minority racial status, smoking, frequency of eating hamburgers or other high-fat meats, and frequency of eating french fries or potato chips (P≤.01). A model consisting of diet pill use, minority race, dieting to lose weight, smoking, feeling overweight, and number of servings of hamburgers, hot dogs, or barbecue correctly classified 97% of the boys who had or had not vomited or used laxatives.

Conclusion  Younger adolescents trying to lose weight engage in a variety of problem dieting and weight loss behaviors that can compromise health and may be associated with eating disorders.

DIETING TO lose or control weight is common among adolescents in the United States.^1-5 Although efforts at weight control may be associated with healthy behavioral changes, including increasing exercise and reducing fat intake, concern exists about possible adverse outcomes associated with some weight loss behaviors. Dieting may be a contributing factor in the development of bulimic symptoms or eating disorders, such as anorexia nervosa or bulimia nervosa.^6-10 In addition, dieting may lead to inadequate nutrient intake during critical periods of growth and development.^6,11,12 Finally, cycles of weight loss and gain appear to enhance metabolic efficiency. As a result, less energy intake may be needed to maintain body weight, thus compromising future attempts at weight loss.^6,13

Several large population-based studies attest to the prevalence of weight control efforts among older adolescents. Analyzing data from the 1990 Youth Risk Behavior Survey (YRBS), a national school-based study designed to monitor the prevalence of health-risk behaviors among high school students, Serdula et al¹ found that 44% of girls and 15% of boys reported trying to lose weight. An additional 26% of girls and 15% of boys were attempting to keep from gaining weight. Thirty-four percent of girls and 15% of boys considered themselves "too fat," a perception that was most strongly associated with attempts to lose weight. Methods used to lose weight in the 7 days preceding the survey included exercise (51% of girls and 30% of boys), skipping meals (49% and 18%, respectively), using diet pills (4% and 2%, respectively), and vomiting (3% and 1%, respectively).¹

Among high school students participating in the YRBS in Massachusetts, Middleman et al² found that 61.5% of girls and 21.5% of boys were trying to lose weight. Forty-two percent of girls and 22.3% of boys considered themselves slightly or definitely overweight, and perceived overweight status was positively correlated with attempting to reduce weight. Of those trying to lose weight, 80.1% had dieted, exercised, or both in the past 7 days, and 7.8% had intentionally vomited, used diet pills, or both.²

In a longitudinal, population-based study of 1015 girls in the 9th through 12th grades in a suburban school district, French et al³ found that 63% had used no weight-loss methods during the past year, 21.5% used healthy practices (eg, increasing exercise, eliminating snacks, or decreasing fat and energy intake), and 15.2% reported the use of both healthy and unhealthy methods (eg, fasting, skipping meals, vomiting, and using diet pills). Potentially harmful weight-loss practices were infrequently reported: 4.4% vomited, 5.4% took diet pills, 1.6% used laxatives or enemas, and 1% took diuretics.³

Less is known about the weight-loss behaviors of younger adolescents, and most available information is derived from studies of small select samples. In the single population-based study addressing this issue, Story et al¹⁴ examined the prevalence of long-term dieting, described as being on a diet more than 10 times in the past year, in a sample of 36 320 public school students participating in the Minnesota Adolescent Health Survey conducted during 1987 and 1988. The prevalence of long-term dieting among girls was 7.8% for seventh and eighth graders, 13.5% for 9th and 10th graders, and 14.3% for 11th and 12th graders. Among boys, the prevalence of such dieting was lower (1.6%, 2.5%, and 2.4%, respectively).¹⁴

The purpose of the present study was to examine dieting, eating and exercise behaviors, use of diet pills, and vomiting or use of laxatives to lose weight among younger adolescents in the sixth, seventh, and eighth grades in North Carolina.

The Centers for Disease Control and Prevention developed the YRBS to assess the prevalence of health-risk behaviors among students in the 9th through 12th grades across the United States. In the spring of 1995, a modified and shortened version of the YRBS was administered to 2331 students in the sixth through eighth grades attending 53 randomly selected public middle schools in North Carolina. The sample was designed to represent the 463 public middle schools and 261 309 students attending these schools in the state. Participation was voluntary and students were assured of anonymity during administration of the survey. The response rate was 74% for schools and 86% for students within responding schools. Demographic data on age, grade, sex, race/ethnicity, and family type are presented in Table 1. For the analyses, race/ethnicity was recorded as either white, nonhispanic, or minority racial/ethnic group.

Using standard YRBS questions, the YRBS administered in North Carolina middle schools measured self-perceived body weight status, dieting behaviors, foods eaten the day prior to the survey, exercise frequency, participation in organized physical activity, academic achievement, and smoking. With respect to dieting behaviors, students were asked "Which of the following are you trying to do about your weight?" Possible responses included "lose weight," "gain weight," "stay the same weight," and "not trying to do anything about my weight." Participants were also asked the following: "Have you ever dieted to lose weight?" "Have you ever exercised to lose weight or to keep from gaining weight?" "Have you ever vomited or taken laxatives to lose weight or to keep from gaining weight?" "Have you ever taken diet pills to lose weight or to keep from gaining weight?" For this group of questions, subjects could respond yes or no.

Five questions were posed regarding dietary practices on the day preceding the survey. Students were asked, "Yesterday, how many times did you eat green salad or raw or cooked vegetables?" Analogous questions were posed for fruit or fruit juice; hamburgers, hot dogs, sausage, bacon, or barbecue; french fries or potato chips; and candy, cookies, doughnuts, pie, or cake. Possible responses were 0 times, 1 time, 2 times, or 3 or more times. Responses to these questions were analyzed to examine possible associations between eating patterns and problem dieting behaviors.

Students were asked 4 questions about physical activity. These addressed the number of days students had exercised during the past 7 days (possible responses ranged from 0-7 days), days per week of physical education classes (possible responses ranged from 0-5 days), participation in organized team athletics (possible responses were yes or no), and participation in individual sports (possible responses were yes or no).

Reliability of the YRBS has been evaluated by investigators from the Centers for Disease Control and Prevention and Westat, Inc, Rockville, Md. The 1992 YRBS questionnaire was administered twice, 14 days apart, to 1679 students in grades 7 through 12. The κ statistics for the entire test ranged from 0.145 to 0.911, with 71.7% of the 53 items considered to have good to excellent reliability (0.61-1.00).¹⁵

Weighting procedures were used to correct for the sampling scheme in the data collection. Westat Inc performed the data cleaning and the computation of the weight under contract by the Centers for Disease Control and Prevention. Weighting compensated for nonresponse and reflected the likelihood of sampling each student. The weight used for estimation is given by:W = (W₁) (W₂) (f₁) (f₂) (f₃), whereW₁ equals the inverse of the probability of school selection; W₂, the inverse of the probability of classroom selection; f₁, a school-level nonresponse adjustment factor calculated by school size;f₂, a student-level nonresponse adjustment factor calculated by school size; and f₃, a poststratification adjustment factor by sex and age.

Pairwise associations among categorical variables were examined using Pearson χ² tests followed by Cramer V statistic to assess the strength of association. Variables found to be significantly (P≤.01) associated with vomiting and laxative use were analyzed with stepwise multiple logistic regression analysis using the likelihood ratio approach.

Among students surveyed, 549 (23.8%) considered themselves overweight, 930 (40.5%) were trying to lose weight, and 521 (22.1%) were trying to keep the same weight. Measures used to lose or maintain weight included dieting (785 [34%]), exercise (1361 [59.4%]), vomiting or taking laxatives (156 [6.7%]), and taking diet pills (153 [6.8%]).

Two hundred ninety-four (25.8%) of the girls and 255 (21.8%) of the boys believed that they were slightly or very overweight (P = .18, χ²) (Table 2). Whether adolescents were trying to lose or gain weight differed significantly by sex. Of the students surveyed, 577 girls (50.6%) and 353 boys (30.5%) were trying to lose weight (P<.0001). In contrast, 264 boys (22.8%) but only 104 girls (9.1%) were trying to gain weight (P<.0001). Of the adolescents attempting to lose weight, 593 (64.1%) had dieted, 853 (92.0%) exercised, 122 (13.1%) vomited or used laxatives, and 117 (12.7%) used diet pills. Perceived weight status was correlated with attempts to change weight (r = 0.50;P<.0001). For example, 394 (83.5%) of those who considered themselves slightly overweight and 67 (88.2%) of those who believed themselves to be very overweight were attempting to lose weight. Of the students surveyed, 110 girls (9.7%) and 46 boys (4.0%) reported vomiting or using laxatives to control weight. Because girls and boys differed greatly in their attempts to change their weight, further bivariate analyses were stratified by sex.

Among girls, vomiting or laxative use was associated with feeling overweight; currently trying to lose weight; dieting, using diet pills, and exercising to control weight; school grade; older age; being a poor student; smoking; eating more salads or vegetables; and eating more candy and other sweets (Table 3). Factors not associated with vomiting or laxative use included minority racial or ethnic status, eating hamburgers or other high-fat meats, eating french fries or potato chips, and participation in physical activity, physical education, or team or other organized sports (Table 3).

When all variables significantly associated with vomiting or laxative use to control weight were entered into multiple logistic regression models, 4 variables correctly classified 92% of the girls. The use of diet pills to control weight had the largest odds ratio and partial r (Table 4). Girls who used diet pills were more than 9 times more likely to also vomit or use laxatives to control weight than those who never used these agents. Similarly, girls who had previously dieted and were currently trying to lose weight were more likely to report having vomited or used laxatives as weight control measures than nondieters. In the area of self-perceived academic status, as girls progressed on the 7-point scale from one of the best students to near the bottom of their class, the odds of vomiting or using laxatives increased. For example, with the classification of one of the best students as the reference group, girls describing themselves as far above the middle of their class were 1.24 times more likely to have vomited or used laxatives. In contrast, girls who described themselves as near the bottom of their class were 7.44 times more likely to have engaged in these high-risk practices.

Vomiting or laxative use to control weight by boys was associated with feeling overweight; currently trying to lose weight; dieting, using diet pills and exercising to control weight; minority racial or ethnic status; smoking; frequency of eating hamburgers or other high-fat meats; and eating french fries or potato chips (Table 3). In contrast to girls, age, school grade, perceived success as a student, and intake of salads or vegetables were not associated with vomiting or laxative use. However, like girls, boys' use of these high-risk weight control measures was not associated with variables reflecting physical activity (Table 3).

Six variables correctly classified 96.6% of the boys who vomited or used laxatives to control weight. In order of magnitude, the strongest associations were found with diet pill use, minority race/ethnicity, and having dieted (Table 5). Boys' risk of having vomited or used laxatives increased with the frequency of cigarette smoking. As the number of days cigarettes were smoked during the past 30 days progressed on the 7-point scale from 0 days (reference category) to all 30 days, the risk of having vomited or used laxatives increased 1.45 times for each ordinal category on the scale. Similarly, as self-described weight varied on the 5-point scale from very underweight (reference category) to very overweight, the odds ratio of having vomited or used laxatives increased 1.70 fold for each category on the scale. The odds of having vomited or used laxatives ranged from 1.44 times if hamburgers or other high-fat meats were eaten once on the day preceding the survey to 4.32 times if the student consumed these foods 3 or more times during the previous day.

In the present study, we examined the weight-control practices of students in the sixth, seventh, and eighth grades in North Carolina using data from the YRBS. To our knowledge, this is the first population-based epidemiological study to address this issue in younger adolescents. Our analysis revealed that a significant proportion of middle school students (50.6% of girls and 30.5% of boys) were trying to lose weight and that many others (22.2% of girls and 23.1% of boys) were attempting to maintain their weight. These data are comparable to those derived from studies of older adolescents enrolled in high school. Serdula et al¹ and Middleman et al² found that 44.0% and 61.5%, respectively, of girls and 15.0% and 21.5%, respectively, of boys were trying to lose weight. Also, as among high school students,^1,2,5,16 the desire of younger adolescents to lose weight was correlated with self-perceived weight status. Specifically, those students who saw themselves as overweight were attempting to reduce their weight.

Middle school students used a variety of strategies to lose weight. More than 90% of young adolescents who were attempting to lose weight exercised and nearly two thirds dieted. Although decreasing portion size and altering the composition of one's diet may be safe and effective adjuncts to weight reduction, there is evidence that adolescents often use unhealthy dieting practices.^6,11,12 In a study of 206 girls in the sixth grade, Koff and Rierdan¹¹ found that nearly one half of those who were dieting skipped meals and 20% fasted. Nutrient intake may also be compromised in adolescents who diet. Macdonald and colleagues¹² reviewed the 24-hour food records of 276 females aged 14 to 18 years. They observed that heavier girls skipped meals to lose weight, a practice that not only decreased the total nutrient intake but also diminished the mean number and amount of servings from specific food groups (eg, milk and milk products and fruits and vegetables).¹² Attesting to the potential adverse effects of weight loss on dietary composition are the results of the study by Woodward¹⁷ of 1055 Australian girls aged 12 to 15 years. He found that girls who viewed themselves as overweight, and therefore were prone to food restriction, were more likely to have low intakes of carbohydrates, iron, calcium, niacin, and fat. Reduction of calcium intake among younger female adolescents is of particular concern given the mineral's role in bone metabolism. Inadequate mineralization during adolescence may increase the risk of osteoporosis later in life.

The use of unhealthy weight-reduction practices was not uncommon among middle school students in our study. Of those trying to lose weight, 122 (13.1%) vomited or used laxatives and 117 (12.7%) used diet pills. Vomiting or laxative use was associated with other weight-control practices, including diet pill use and dieting. In fact, diet pill use and prior attempts to diet markedly increased the odds of having vomited or used laxatives to control weight. In addition, vomiting or laxative use was associated with smoking, a practice that might be viewed as an attempt to suppress the appetite. Dieting and the use of unhealthy weight-reduction measures appear to increase the risk of one's developing an eating disorder.^6,7,10,18 Neumark-Sztainer and colleagues¹⁸ investigated the association between dieting methods and binge eating in a sample of females in the 10th grade attending 3 high schools in Israel. Dieters using dangerous methods to lose weight, defined as fasting, vomiting, or taking laxatives or diet pills, were significantly more likely to binge eat than those who used more moderate strategies. Among school-aged girls in London, England, Patton et al⁷ observed that when compared with nondieters, dieters had an 8-fold increased risk of developing an eating disorder. Similarly, Marchi and Cohen¹⁰ found a significant association between efforts at weight reduction during adolescence and later development of bulimia (r = 0.667; P<.01).

Certain limitations of this study deserve mention. First, this was a secondary data analysis of a survey that was designed to monitor health-risk behaviors, not to specifically address issues of weight control. Second, the data are self-reported and the validity of the type of modified dietary recall used is not known. Finally, the survey did not permit verification by objective measures of students' perceptions of weight status. However, the purpose of our study was to examine associations between weight change intentions and behaviors and not to validate reports of body image.

Despite these limitations, the findings from this analysis indicate that most young adolescents have made attempts to lose weight and that many use potentially harmful strategies to do so. These efforts at weight reduction may lead to inadequate nutrition or serve as a risk factor for the development of an eating disorder. From a clinical perspective, these data underscore the importance of conducting routine screening for dieting and other weight-loss practices as a part of ongoing health care for young adolescents. This may be accomplished during the clinical interview or through the use of a questionnaire (eg, such as that offered as part of the Guidelines for Adolescent Preventive Services¹⁹) completed by the adolescent. If the adolescent is using unsound dieting or weight-loss practices, additional counseling or referral to a nutritionist is warranted.

Editor's Note: There's something really wrong with a world in which a large proportion of the population is at risk from starving while the others are at risk from purging.—Catherine D. DeAngelis, MD

Accepted for publication May 4, 1998.

Corresponding author: Daniel P. Krowchuk, MD, Department of Pediatrics, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157 (e-mail: krowchuk@wfubmc.edu).