Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
To examine rates of weight-based teasing before initiation of school-based childhood obesity prevention policies (Arkansas Act 1220 of 2003) and during the 2 years following policy implementation, as well as demographic factors related to weight-based teasing.
Analysis of consecutive random cross-sectional statewide telephone surveys conducted annually across 3 years.
Sample representative of Arkansas public school students with stratification by geographic region, school level (elementary, middle, and high school), and school size (small, medium, and large).
Parents of children enrolled in Arkansas public schools and index adolescents 14 years or older.
Statewide school-based obesity policies, including body mass index screening.
Main Outcome Measures
Survey items about weight-based teasing, other teasing, body weight and height, and sociodemographic factors, as well as school characteristics obtained from the Common Core of Data of the National Center for Education Statistics.
At baseline, 14% of children experienced weight-based teasing by parental report. The prevalence of weight-based teasing did not change significantly from baseline in the 2 years following school-based policy changes. Children and adolescents most likely to be teased because of weight were those who were overweight, obese, white, female, and 14 years or older, as well as those teased for other reasons. Adolescent report of weight-based teasing yielded similar patterns.
Although the effectiveness of school-based obesity prevention policies remains unclear, policy changes did not lead to increased weight-based teasing among children and adolescents.
Childhood obesity has reached epidemic levels. We use the terms overweight and obese herein to convey the magnitude of risk associated with excess weight in children, consistent with recent American Medical Association1 recommendations. Overweight children have a body mass index (BMI) between the 85th and 94th percentiles for age and sex according to Centers for Disease Control and Prevention growth charts.2 Children with a BMI at or above the 95th percentile for age and sex are considered obese. National data indicate that more than one-third of American children are overweight or obese, with significant increases in the prevalence of overweight and obesity during the previous 6-year period.3 Efforts to abate increasing childhood obesity rates could include methods to establish and monitor the prevalence of childhood obesity, including BMI screening.4
With Act 1220 of 2003, Arkansas became the first state to legislate statewide school policies to combat childhood obesity, with the following 5 primary provisions: (1) BMI measurement, with confidential reports to parents; (2) removal of vending machine access for all public elementary school students; (3) public disclosure of vending contracts; (4) creation of a state-level advisory committee to recommend physical activity and nutrition policy changes to the board of education; and (5) creation of school district–level advisory committees to guide local policy implementation. Of these components, BMI screening received the most criticism, including concerns that weight-based teasing would increase as a result.4,5
Research evaluating the prevalence of weight-based teasing indicates that it is common among children and adolescents, particularly among girls6,7 and overweight children.6- 9 Among adolescents spanning a range of body weights, 25.5% of girls reported weight-based teasing more than a few times per year, while 22.2% of boys reported such teasing.6 Similarly, weight-based teasing was 3 times more likely among obese children aged 10 to 14 years than among nonobese children in that age range.9 Weight-based teasing is of particular concern because it has been associated with negative psychosocial behaviors, including disordered eating6,9,10 and anxiety,7 as well as depression, suicidal ideation, and suicide attempts.10 A recent prospective study8 of middle school and high school students found that teasing about weight was a significant predictor of disordered eating behaviors 5 years later such as fasting, taking diet pills, self-inducing vomiting, and using laxatives or diuretics. Therefore, the question of whether weight-based teasing is precipitated by school-based obesity policies merits investigation.
Little, if any, research is available to inform the public and the scientific community about the potential risks of legislative approaches to address childhood obesity that are being implemented or considered by many states. The primary objective of this study was to examine whether rates of weight-based teasing increased after the implementation of Act 1220 of 2003.
Baseline cross-sectional telephone surveys among parents of Arkansas public schoolchildren were conducted in the summer of 2004, before initiation of school-based obesity policies. Follow-up surveys were repeated in each of the subsequent 2 years (2005 and 2006). Each survey included identical items focused on weight-based teasing, allowing comparison across the 3 years.
Survey sampling methods have been described in detail elsewhere.11 Briefly, a stratified multistage sampling procedure was used to ensure that the sample included parents whose children attended public schools across all geographic regions (north, northwest, southwest, central, and east), school types (elementary, middle, and high school), and school sizes (small, medium, and large). Alternative and special schools and schools with enrollments of fewer than 100 students were excluded. School size was based on tertiles of enrollment, using data from the Common Core of Data of the National Center for Education Statistics. Schools were sampled, and households with listed telephone numbers within a 15-mile radius of a selected school were sampled and contacted. Households with a child attending prekindergarten to 10th grade in a selected school were interviewed. If the household contained more than 1 eligible child, a single child was selected randomly, and that child was the focus of the interview.11
Parents verbally consented to complete the telephone interview. If the selected child was an adolescent (≥14 years) and the parent consented, the adolescent was interviewed after assent. The project was reviewed and approved by the University of Arkansas for Medical Sciences Institutional Review Board.
Surveys included questions about teasing, weight and height, demographic factors, opinions about school-based obesity policies, and attitudes, behaviors, and cognitions about dietary intake and physical activity. Survey questions assessed the parent interviewed, the family overall, and the index child. The present analyses use reports of teasing, weight and height data, and demographic characteristics.
Parents provided the index child's sex and age and the parents' race/ethnicity. Weight-based teasing was determined by asking: “Do others tease, joke, or make fun of your child because of his/her weight?” (yes or no). Non–weight-based teasing was assessed by asking: “Do others tease, joke, or make fun of your child because of other reasons?” (yes or no). The child's BMI was calculated as weight in kilograms divided by height in meters squared using parent-reported height and weight and was transformed into sex- and age-specific percentiles based on the Centers for Disease Control and Prevention growth charts.2
Adolescents were asked about their teasing experiences with the following 2 questions: “Do others tease, joke, or make fun of you because of your weight?” (yes or no) to assess weight-based teasing and “Do others tease, joke, or make fun of you because of other reasons?” (yes or no) for non–weight-related teasing. Adolescents also identified their racial/ethnic group.
Data were weighted to be representative of families with children attending public schools in Arkansas, accounting for geographic region, school level, and school size. Child weight status was categorized as not overweight (<85th BMI percentile), overweight (85th to <95th BMI percentile), or obese (≥95th BMI percentile).2 Race/ethnicity was classified as white or as nonwhite. Analyses of change in the weight-based teasing prevalence over time were conducted unadjusted for adolescent- and parent-reported data, followed by analyses adjusted for sex, school size, weight status, race/ethnicity, and non–weight-based teasing. For parent-reported data, changes in teasing were first assessed across sex, age groups, and weight status and then were adjusted for sex, school size, race/ethnicity, and non–weight-based teasing. Additional analyses examined which children were most vulnerable to weight-based teasing. Wherever appropriate, χ2 tests were used to measure unadjusted associations. Logistic regression models were used to adjust for covariates when examining change in weight-based teasing. Odds ratios (ORs) and 95% confidence intervals (CIs) were used to describe the magnitude of association between weight-based teasing and covariates. Analyses were considered significant at α = .05. All analyses were conducted using commercially available statistical software (SAS, version 9; SAS Institute Inc, Cary, North Carolina).
A total of 1551 parents were interviewed at baseline (76.1% cooperation rate). Of the households with an adolescent-aged index child, 99.0% (207 of 209) of adolescents assented to be interviewed. In the 1-year follow-up interviews, a larger sample of parents was interviewed (n = 2508), with a cooperation rate of 62.8%. Of households with an adolescent, 99.4% (478 of 481) of adolescents were interviewed. In the 2-year follow-up interviews, 2358 parents were interviewed (55.2% cooperation rate), and 100.0% (357 of 357) of adolescents were interviewed.
No significant differences in demographic characteristics were noted between baseline and follow-up samples except for age (Table 1). The proportion of children younger than 14 years ranged from 84% to 89% during the 3 years.
Fourteen percent of parents reported that their child experienced weight-based teasing at baseline, with no significant change in weight-based teasing noted in either of the following 2 years (Table 1). The unadjusted prevalence of parent-reported weight-based teasing also did not significantly change for subgroups of the sample (ie, age, sex, weight status, and race/ethnicity) over time (Table 2). After adjustment, weight-based teasing was as likely at baseline as it was at the 1-year follow-up (OR, 0.86; 95% CI, 0.62-1.18) and at the 2-year follow-up (0.99; 0.72-1.37).
Although analyses showed no significant changes in the prevalence of weight-based teasing across the period of policy implementation, there were some differences across weight-status groups. The likelihood of weight-based teasing was greater with increasing overweight. Specifically, obese youths were greater than 4 times more likely to be teased about their weight than nonoverweight youths (Table 3). Overweight youths were almost twice as likely to be teased about their weight compared with nonoverweight counterparts.
Regardless of weight status, if youths were teased for non–weight-based reasons, they were more likely to be teased about their weight. Obese, overweight, and nonoverweight youths were approximately 3 times more likely to experience weight-based teasing if they were teased for other reasons.
Aggregating across years, white youths were significantly more likely to experience weight-based teasing than nonwhite youths (Table 3). There was also a marginal trend for white girls to be more likely to experience weight-based teasing than nonwhite girls. There were no racial/ethnic differences in weight-based teasing among boys.
Although teasing rates among girls were stable during the 3 years, results from the adjusted model suggest that girls are more vulnerable to weight-based teasing than boys. However, these sex differences depended on weight status, with nonoverweight girls more likely to experience weight-based teasing than nonoverweight boys (Table 3). Overweight boys were half as likely to experience weight-based teasing compared with overweight girls. Overweight boys were no more likely to experience weight-based teasing than nonoverweight boys (OR, 1.40; 95% CI, 0.89-2.18), while overweight girls were more than twice as likely to experience weight-based teasing than nonoverweight girls (2.55; 1.63-3.99).
Obese boys and obese girls were equally likely to experience weight-based teasing, with both groups being approximately 4 times more likely to be weight-based teased than their nonoverweight counterparts (OR, 3.79; 95% CI, 2.51-5.72 for girls; and 5.50; 3.90-7.78 for boys). For boys, increased vulnerability to weight-based teasing does not occur until they are obese, but obese boys are just as likely to experience weight-based teasing as obese girls.
Sex differences in weight-based teasing were present in children and in adolescents. Weight-based teasing was 60% less likely among adolescent boys than among adolescent girls (OR, 0.37; 95% CI, 0.22-0.62). Among younger children, boys experienced significantly less weight-based teasing than girls (OR, 0.72; 95% CI, 0.55-0.95).
There were no sex differences in weight-based teasing among youths who were teased for other reasons. For boys and for girls, those who were teased for other reasons were about 3 times more likely to be teased about their weight (OR, 3.63; 95% CI, 2.42-5.43 for girls; and 2.94; 2.09-4.13 for boys), indicating that weight-based teasing may be part of a larger constellation of teasing.
Weight-based teasing was significantly less likely among younger children compared with adolescents. Age-related differences in weight-based teasing occurred only between obese youths, such that obese children were less likely to be weight-based teased than obese adolescents (Table 3). There were no significant age-related differences in weight-based teasing for overweight vs nonoverweight youths. While obese children were greater than 4 times more likely to be teased than their nonover weight peers (OR, 4.26; 95% CI, 3.16-5.76), obese adolescents were almost 9 times as likely as their nonoverweight peers to experience weight-based teasing (8.82; 95% 4.55-17.08). Therefore, although adolescents were significantly more likely to experience weight-based teasing overall, this relationship seems to be driven by high rates of weight-based teasing among obese adolescents.
There were no age-related differences in weight-based teasing among youths who were teased for other reasons. Children and adolescents who experienced teasing for other reasons were about 3 times as likely to experience weight-based teasing as their peers who did not experience non–weight-related teasing (OR, 3.34; 95% CI, 2.46-4.56 for children; and 3.09; 1.86-5.14 for adolescents).
No significant differences in relevant demographic characteristics were noted among adolescents in the baseline and follow-up samples. These results are summarized in Table 1.
Eleven percent of adolescents reported weight-based teasing at baseline, with no significant change in the prevalence noted in either of the 2 years following policy implementation (Table 1). Weight-based teasing rates according to adolescent report were lower than rates according to parental report about the adolescent, particularly in the most recent year surveyed. However, the pattern of stable weight-based teasing during the 3 years was similar for parent and adolescent informant groups.
There were no significant sex or racial/ethnic differences in weight-based teasing reported by adolescents (Table 4). Obese adolescents were 3 times more likely to report weight-based teasing than nonoverweight peers. However, overweight adolescents were not significantly more likely to report weight-based teasing than nonoverweight adolescents. Adolescents who were teased for other reasons were 4 times more likely to report weight-based teasing than adolescents who did not report other teasing.
Weight-based teasing did not increase in the 2 years after implementation of Arkansas Act 1220 in the sample overall or among obese or overweight youths in particular. Obese children and adolescents were substantially more likely to be weight-based teased, and overweight youths were at elevated risk compared with nonoverweight peers. Weight-based teasing was greater among white children than nonwhite children, with a marginal trend for increased teasing because of weight among white girls and no racial/ethnic differences among boys. Sex differences were apparent across the weight ranges except for obese children. Obese girls and obese boys were equally likely to experience weight-based teasing. Furthermore, weight-based teasing was significantly more frequent among obese adolescents than among obese children. Finally, across the full spectrum of weight status, children and adolescents teased for other reasons were more likely to be teased about weight, indicating a broader constellation of teasing.
Parents and adolescents were consistent in not reporting an increase in weight-based teasing in the 2 years following the implementation of Act 1220. Although adolescents were somewhat less likely to report weight-based teasing than their parents, similar trends emerged related to predictors of weight-based teasing. Specifically, obese adolescents were more likely to report weight-based teasing, and overweight adolescents were not more likely to report weight-based teasing than their nonoverweight peers. Adolescents who reported being teased for other reasons were also more likely to report weight-based teasing. Based on the adolescent data, there were no significant sex or race/ethnic group differences in weight-based teasing.
We know of no previous studies that evaluated weight-based teasing after the implementation of school-based obesity prevention policies. After the passage of Act 1220, some professional and lay critics argued that the increased focus on obesity would encourage weight-based teasing and other harmful behaviors among youths.12 However, our findings indicate that policy implementation activities did not result in greater weight-based teasing in schoolchildren overall or among obese and overweight children, who might be at greatest risk.
Results from the present study are largely consistent with previous literature that has reported weight-based teasing to be more prevalent among overweight children and obese children.7- 9,13 and more common among girls than among boys.7,13 Consistent with findings of no sex differences in the experience of weight-based teasing among obese children compared with nonoverweight children,9 we observed no differences in weight-based teasing among obese boys and obese girls, indicating that sex differences are apparent only at lower body weights. Weight-based teasing has previously been shown to be comparable across racial/ethnic groups in studies of adolescents6,13 and in samples limited to overweight and obese children.7 In our sample with broad age and weight ranges, weight-based teasing was more common among white youths. Future research may further examine these and other predictors of weight-based teasing to facilitate identification of youths at greatest risk for teasing and for possible negative psychosocial consequences. However, our findings indicate that weight-based teasing is part of the experience of youths regardless of an increased focus on obesity prevention.
This article about the influence of Arkansas Act 1220 of 2003 on weight-based teasing has several strengths. It is the first report (to our knowledge) to assess the influence of comprehensive school-based obesity reduction policies on weight-based teasing. Furthermore, the evaluation of Act 1220 recruited a diverse sample that allowed for broad statewide representation. This article expands the available data on adolescent weight-based teasing by surveying parents and adolescents. We are unaware of other studies that have measured weight-based teasing using multiple informants. Consideration of whether reports from parents or from adolescents are best for assessing weight-based teasing may be an important area for future investigation.
Previous research has largely not examined child age or non–weight-based teasing. Given our findings that weight-based teasing is more prevalent among adolescents compared with children, future research may focus on adolescents and examine the trajectory of weight-based teasing throughout childhood. The clustering of weight-based teasing with teasing for other reasons may help inform methods of detecting teasing and interventions for targeting teasing. Although policies to address weight-based teasing have not been implemented at a statewide level in Arkansas, a recent pilot study14 in Minnesota found that an intervention to decrease weight-based teasing was low cost and successful. The persistent rates of weight-based teasing indicate that school-based programs addressing the issue could be beneficial.
Several limitations should be considered when interpreting these findings. Because the surveys were designed to evaluate many aspects of a broad initiative, weight-based teasing was assessed with a single item. Therefore, future studies monitoring weight-based teasing in the context of school-based obesity policies may benefit from including a more comprehensive assessment that addresses teasing frequency and duration, specific perpetrators, and emotional effect. Research indicates that weight-based teasing can come from parents, siblings, and other family members, as well as from peers,6,10,15 and reported rates of teasing may differ based on the respondent. For example, parents who tease their children about weight may be less likely to report weight-based teasing. Therefore, it may be important to consider the source of the teasing and the respondent in future research. A larger sample size would allow examination of differences in weight-based teasing experiences across racial/ethnic groups, and a longitudinal sample would allow examination of relationships between policy changes and weight-based teasing for individual children. It may also be important to note that a child's weight status in the present study was based on parental report. However, such reports have been previously found to be reasonably accurate,16 and our results mirror findings reported by other weight-based teasing researchers who measured weight and height directly.6
Although further research is needed to determine whether school-based policies will reduce the childhood obesity prevalence, our findings may reassure parents, academicians, policy makers, school personnel, and health care providers that such approaches do not seem to adversely affect weight-based teasing. In fact, recent research about policies that include BMI screening in schools with parental feedback has shown initial promise for increasing parents' awareness of their child's weight status.17,18 Therefore, a comprehensive policy that includes BMI screening offers a promising approach to increase awareness of and to potentially reduce the prevalence of obesity among our nation's youths.
Correspondence: Rebecca A. Krukowski, PhD, Fay W. Boozman College of Public Health, University of Arkansas for Medical Sciences, 4301 W Markham St, Mailslot 820, Little Rock, AR 72205 (RAKrukowski@uams.edu).
Accepted for Publication: February 10, 2008.
Author Contributions: Dr Krukowski had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Krukowski, Smith West, Bursac, and Raczynski. Acquisition of data: Phillips and Raczynski. Analysis and interpretation of data: Krukowski, Smith West, Siddiqui, Bursac, Phillips, and Raczynski. Drafting of the manuscript: Krukowski, Smith West, Siddiqui, and Raczynski. Critical revision of the manuscript for important intellectual content: Smith West, Siddiqui, Bursac, Phillips, and Raczynski. Statistical analysis: Bursac. Obtained funding: Phillips and Raczynski. Administrative, technical, and material support: Phillips and Raczynski. Study supervision: Smith West and Raczynski.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grants 051737, 60284, and 30930 from the Robert Wood Johnson Foundation (Dr Raczynski).
Krukowski RA, West DS, Siddiqui NJ, Bursac Z, Phillips MM, Raczynski JM. No Change in Weight-Based Teasing When School-Based Obesity Policies Are Implemented. Arch Pediatr Adolesc Med. 2008;162(10):936-942. doi:10.1001/archpedi.162.10.936