PAPER Weight-related attitudes and behaviors of adolescent boys

International Journal of Obesity (2002) 26, 1579–1587 ß 2002 Nature Publishing Group All rights reserved 0307–0565/02 $25.00 www.nature.com/ijo

PAPER Weight-related attitudes and behaviors of adolescent boys and girls who are encouraged to diet by their mothers JA Fulkerson*, MT McGuire, D Neumark-Sztainer, M Story, SA French and CL Perry Division of Epidemiology, University of Minnesota, Minneapolis, Minnesota, USA OBJECTIVE: To assess the relationships between mothers’ reports of dieting and encouraging adolescents to diet and adolescents’ reports of their own dieting practices and weight-related concerns. DESIGN: Cross-sectional study of parent interviews and adolescent surveys in an ethnically-diverse sample. SUBJECTS: A total of 810 adolescents (n ¼ 381 boys and n ¼ 429 girls) and their mothers. RESULTS: Mothers’ dieting was associated with their adolescent girls’ weight-related concerns and behaviors, but these associations were not significant after adjusting for girls’ body mass index (BMI). In contrast, mothers’ encouragement for sons to diet was associated with sons’ binge eating, dieting and other weight-control behaviors, even after controlling for sons’ BMI. Compared with mothers who did not encourage their child to diet, mothers who encouraged their child to diet were significantly heavier women and were more likely to view their child as overweight. Forty-three percent of boys and 46% of girls who were encouraged by their mothers to diet were classified as nonoverweight by federal guidelines. CONCLUSIONS: Boys who are encouraged by their mothers to diet may be at risk for health-compromising eating and dieting behaviors, particularly binge-eating, fasting, eating a little bit of food and skipping meals. Parents who are concerned about their children’s weight should be educated to encourage healthy eating habits and physical activity to promote their children’s health, including healthy weight control. International Journal of Obesity (2002) 26, 1579 – 1587. doi:10.1038=sj.ijo.0802157 Keywords: encouragement; dieting; weight control; parental influence

Introduction Childhood overweight is a major public health problem and its prevalence continues to rise in the USA.1,2 Childhood obesity prevention and treatment approaches focus on promoting physical activity and healthy eating so that weight loss can be achieved without adverse effects on growth or lean body mass.3,4 Dieting behaviors are commonly reported by adolescents, particularly among girls.5 – 7 Adolescent reports of dieting may refer to a variety of weight-control behaviors, some of which may be healthy (eg exercise and lowering excessive fat intake) and others which may be unhealthy (eg very restrictive dieting). Healthy weight-control behaviors may be desirable for overweight adolescents; *Correspondence: JA Fulkerson, Division of Epidemiology, University of Minnesota, 1300 South Second Street, Minneapolis, MN 55454-1015, USA. E-mail: [email protected] Received 13 November 2001; revised 6 June 2002; accepted 19 June 2002

however, unhealthy weight-control behaviors are not appropriate for any adolescent, regardless of their weight status. Moreover, even moderate levels of dieting have been found to increase the likelihood of the development of an eating disorder.8 Therefore, it is important to consider the positive and negative effects of dieting as well as the type of dieting to prevent childhood overweight. Social influences (eg media, peers, family) regarding dieting and other weight-control practices are particularly important during adolescence.9,10 Sociocultural pressures to be thin via the media are strong,11,12 and peer dieting attitudes and behaviors have also been shown to be very important to adolescents.10 However, studies have shown that parents, particularly mothers, can have a substantial influence on their children’s eating and dieting patterns.13 Mothers are most frequently the source of food preparation and dieting information for their children.14 Even girls as young as 5 y of age have been found to be influenced by their mothers’ dieting behaviors.15

Encouragement to diet by mothers JA Fulkerson et al

1580 There are two primary ways in which parents can influence dieting behaviors in their children: indirect influence and direct influence. Indirect influence occurs by modeling dieting behaviors. In contrast, direct influence occurs by making verbal statements regarding dieting (eg encouraging child to diet, commenting on child’s weight). There is some evidence of a modest relationship for indirect parental influence of dieting and dieting behaviors in their pre-adolescent daughters;16 however, this relationship does not appear among adolescent daughters17,18 or pre-adolescent sons.19 Smolak and colleagues19 found a stronger relationship for direct vs indirect parental influence (as reported by parents) and their child’s dieting behaviors (as reported by child). Studies have suggested that this relationship may be stronger for girls than boys.19,20 In particular, when parents directly encourage their children to diet, the children are more likely to diet. For instance, studies of adolescent girls9,18,21 found that, compared with girls who did not report parental encouragement to diet, girls who reported parental encouragement to diet were significantly more likely to diet. Similar findings were seen in a study of pre-adolescent girls; girls who reported that their mother commented on their weight were significantly more likely to diet.22 These studies did not include the parents’ own report of encouraging their child to diet or making comments about their child’s weight. One study of adolescent girls that did assess both child and parent reports of parental comments about daughter’s weight23 found that dieting was associated with girls’ reports, but not parents’ reports, of mother’s comments about weight. None of these studies included boys as participants. In a study with younger children, Thelen and Cormier20 surveyed 9 – 10.5 y-old boys and girls and their parents to assess the relationships between children’s dieting behaviors and direct parental reports of encouragement to diet. Girls’ and boys’ body mass indexes (BMIs) were significantly related to parents’ own report of their encouragement of their child to diet; however, after controlling for child’s body weight, parental encouragement to diet was not significantly related to dieting frequency among girls or boys. Given the age of their sample and the low participation rate (35%), their results may not reflect the relationship between parental encouragement to diet and dieting behaviors and attitudes in adolescent-aged girls and boys. To date, three studies have investigated indirect and direct parental influences on dieting behaviors in boys and girls from the perspectives of both parents and children;19,20,24 however the ‘child’ participants were either young children19,20 or college-age adults.24 Thus, two important sets of relationships have not been investigated: (1) simultaneous examination of parental modeling of dieting behaviors and parents’ direct encouragement to diet on dieting practices of adolescent boys and girls; and (2) the relationships between parents’ reported encouragement to diet and dieting practices for both adolescent boys and girls. One of the most interesting pieces of information missing from this body of literature is data regarding weight-related concerns and International Journal of Obesity

behaviors of parents who encourage their children to diet. Perhaps these parents are overweight themselves and perceive their children to be overweight, or are interested in helping their children prevent excessive weight gain. Because parents are often involved in changing children’s behaviors to prevent childhood overweight=obesity and may be important change agents for modifying adolescents’ behaviors, it is important to understand the characteristics of parents who encourage their children to diet, and the effects of that encouragement. The present study assessed the relationships among mothers’ own reported dieting practices and encouragement for their adolescents to diet with adolescent-reported dieting practices and weight-related concerns. We hypothesized that both mothers’ dieting behavior (indirect influence) and mothers’ encouragement of adolescents to diet (direct influence) would be associated with the adolescents’ weight-related concerns and dieting behaviors, with stronger associations for direct influence. In addition, weight-related concerns, dieting practices, and perceptions of adolescent’s weight of mothers who report they ‘encourage their adolescent to diet’ were assessed. We hypothesized that two types of factors may influence a mother’s encouragement behavior regarding her adolescent’s weight: (1) her own weight-related concerns; and (2) her perceptions of her adolescent’s weight.

Method Subjects Participants in the present study were 810 adolescents (n ¼ 381 boys and n ¼ 429 girls) and their female guardians. The adolescents were a subset of those who participated in the survey portion of Project EAT (Eating Among Teens), a study of 4746 adolescents from 31 middle- and high-schools in the Minneapolis=St Paul seven-county metropolitan area. The 810 adolescents who constituted the subset were randomly selected from the larger study in a stratified design in order to increase racial=ethnic heterogeneity of the sample. The parents of the ethnically- diverse subset (n ¼ 1045) of students who completed the school survey were contacted for a telephone interview. The parent telephone survey response rate was 86.6%, with a total of 902 completed parent interviews. The parent participants of the present study included only the 810 female guardians who participated in the telephone interviews (90% of the completed parent interviews), because generally mothers have more influence on the eating behavior of their children than male guardians. Ninety-two percent of the female guardians were mothers (n ¼ 742), 2% stepmothers (n ¼ 20), and 6% other female guardians (n ¼ 48). All female guardians are referred to as ‘mothers’ hereafter. Among the 810 adolescents, 47% were boys, 60% were in senior high school, and the mean age was 14.6 years (s.d. ¼ 1.7). Thirty-one percent of the adolescents were white, 25% black, 17% Asian, 14% Hispanic, 10% Native American, and 3% other=mixed.


1581 Of the 810 mothers in this study, 47% were white, 23% black, 13% Asian, 10% Hispanic, and 7% other=mixed. Sixtytwo percent of the mothers were married. Mothers were equally distributed between high-, middle- and low-SES groups. Eighty-three percent of mothers were working outside of the home.

Procedure For the adolescent sample, trained research staff administered a 221-item survey during class time and directly measured their height and weight in a private area in the school. Study procedures were approved by the University of Minnesota Human Subjects’ committee and by the participating schools districts’ research boards. Passive and active consent procedures were fulfilled in accordance with the participating school board procedures. Mothers of these adolescents were interviewed by telephone by trained interviewers.

Measures Adolescents’ weight and BMI. Research staff directly measured each adolescent’s height using a portable stadiometer and measured weight using a portable digital scale. Body mass index was calculated according to the formula: weight (kg)=height (in meters squared). Using gender- and agespecific cut-off points based on reference data from the Centers for Disease Control (2000)1 growth tables, adolescents were classified as underweight (BMI < 15th percentile), average weight (15th percentile BMI < 85th percentile), moderately overweight (85th percentile BMI < 95th percentile), and very overweight (BMI > 95th percentile). Adolescents’ weight-related attitudes, concerns and weight-control practices. The weight-related attitudes and concerns that were assessed included: weight dissatisfaction, caring about controlling weight, thinking about being thinner, worrying about weight gain, weighing self often, weight=shape importance and perception of current weight. The following adolescent weight-control practices were also assessed at the item level: past year dieting, current dieting and binge eating (see Appendix for survey items and response categories). Summary score of adolescents’ healthy weight-control behaviors. Adolescents were asked to report if they had practiced any of the following behaviors as methods of weight-control during the past year: exercise, ate more fruit and vegetables, ate less high-fat foods, ate fewer sweets. These items were summed to form a total healthy weightcontrol behavior scale (range ¼ 0 – 4, Cronbach’s a ¼ 0.82), and the variable was analyzed by comparing adolescents who engaged in at least one healthy weight-control behavior with adolescents who did not engage in any healthy weightcontrol behaviors.

Summary score of adolescents’ unhealthy weight-control behaviors — less extreme. The prevalence of practicing unhealthy weight-control behaviors during the past year was assessed. Respondents were asked individual items about fasting, eating little food, food substitutes (liquid meals), skipping meals and smoking cigarettes, and responses were summed to form a total ‘less extreme’ unhealthy weight-control behavior scale (range ¼ 0 – 5, Cronbach’s a ¼ 0.66). The variable was analyzed by comparing adolescents who engaged in at least one less extreme unhealthy weight-control behavior with adolescents who did not engage in any less extreme unhealthy weight-control behaviors. Summary score of adolescents’ unhealthy weight-control behaviors — more extreme. Adolescents were asked to report if during the past year they had engaged in taking diet pills, laxatives, diuretics or vomiting. These items were summed to form a total ‘more extreme’ unhealthy weightcontrol scale (range ¼ 0 – 4, Cronbach’s a ¼ 0.64). The variable was analyzed by comparing adolescents who had engaged in at least one extreme unhealthy weight-control behavior with adolescents who had not engaged in any extreme unhealthy weight-control behaviors. Mothers’ weight and weight-related behaviors. Mothers reported their current weight and height, and BMI was calculated from the self-report data according to the formula: weight (kg)=height (meters squared). Respondents were classified into one of four groups: underweight (BMI < 18.5), normal weight (BMI ¼ 18.5 – 24.9), overweight (BMI ¼ 25.0 – 29.9), and obese (BMI > 30.0) according to federal clinical guidelines.25 Mothers were asked how frequently they dieted to lose weight or keep from gaining weight (the four response options were dichotomized into ‘not at all’ vs ‘a little bit’, ‘somewhat’ and ‘very much’. Mothers reported how satisfied

Table 1 Body mass index and weight status of mothers and adolescent boys and girls a

BMI (mean, standard deviation) Weight status Underweight (%) Average weight (%) Overweight (%) Obese (%)

b

b

Mothers (n ¼ 804)

Boys (n ¼ 380)

Girls (n ¼ 429)

26.6 (5.95)

23.3 (4.90)

23.5 (5.13)

1.4 44.6 29.1 24.9

5.3 58.4 17.9 18.4

2.8 60.7 21.7 14.9

a

NHLBI (1998). Underweight, BMI < 18.5; average weight, BMI ¼ 18.5 – 24.9; overweight, BMI ¼ 25.0 – 29.9; and obese, BMI > 30.0. b Weight status categories classified by using percentiles according to age and gender. Underweight, BMI < 15th percentile; average weight, 15th percentile BMI < 85th percentile; overweight, 85th percentile BMI < 95th percentile; obese (very overweight), BMI > 95th percentile.

International Journal of Obesity


1582 Table 2 Unadjusted prevalence estimates and adjusted odds ratios of adolescent boys’ and girls’ weight-related attitudes and behaviors by mothers’ dieting behavior Boys

Girls

Mother’s dieting a

Mother’s dieting a

No (n ¼ 137) %

Yes (n ¼ 242) %

Weight-related concerns Dissatisfied with weight Care about controlling weight Think about being thinner Worried about weight gain Weigh myself often Weight=shape importance Feel somewhat overweight

15.0 35.6 32.8 33.3 32.9 62.4 25.6

21.5 33.3 31.5 34.5 28.6 59.0 25.0

1.65 0.89 0.78 0.92 0.78 0.85 0.74

Dieting Dieted in the past year Currently dieting to lose weight

23.7 19.1

Adolescent attitudes and behaviors

Binge eating (past year) Embarrassed by binge-eating Weight control behaviors (past year) Healthy weight controlc Less extreme unhealthy weight controlc c Extreme unhealthy weight control

No (n ¼ 158) %

Yes (n ¼ 269) %

3.72) 1.64) 1.54) 1.78) 1.46) 1.56) 1.62)

34.8 39.7 58.1 60.1 27.7 75.0 38.5

44.9 51.3 64.5 73.8 34.8 83.4 49.8

1.36 1.57 1.12 1.69 1.39 1.65 1.26

30.1 25.2

1.18 (0.58, 2.41) 1.23 (0.56, 2.71)

52.3 39.2

58.1 52.8

1.11 (0.64, 1.94) 1.58 (0.87, 2.86)

4.4

6.9

1.48 (0.37, 5.97)

14.8

15.2

0.94 (0.43, 2.02)

67.2 30.7 2.9

70.9 32.2 4.6

1.07 (0.56, 2.07) 0.89 (0.47, 1.72) 1.69 (0.30, 9.65)

82.2 52.2 10.2

89.9 62.9 11.2

1.95 (0.90, 4.23) 1.48 (0.85, 2.57) 0.77 (0.31, 1.91)

OR (99% CI)

(0.73, (0.48, (0.40, (0.47, (0.41, (0.47, (0.34,

b

b

OR (99% CI)

(0.75, (0.91, (0.63, (0.94, (0.77, (0.84, (0.67,

2.46) 2.72) 1.99) 3.05) 2.51) 3.22) 2.36)

a

Referent group for logistic analyses. Logistic analyses adjusted for the effects of adolescents’ observed BMI and school. Summary scores in which adolescents who did not endorse any behaviors were compared to adolescents who reported at least one behavior.

b c

they were with their current body weight (the five response options were dichotomized into ‘completely satisfied’, ‘somewhat satisfied’, ‘neutral’ vs ‘somewhat dissatisfied’, ‘completely dissatisfied’.) Mothers’ perceptions of adolescents and encouragement to diet. Mothers were asked how they would describe their adolescent’s current weight (response options included: very underweight, somewhat underweight, or about right vs somewhat overweight or very overweight). They were also asked the extent to which they ‘encouraged their adolescent to diet to control his=her weight’ (the four response options were dichotomized into ‘not at all’ vs ‘at least a little bit’, ‘somewhat’ and ‘very much’).

Analysis overview In all the analyses, the two independent variables of interest were mothers’ dieting status and mothers’ encouragement of the adolescent to diet as reported by the mother. Group differences between dieting and nondieting mothers were analyzed using t-test and chi-square analyses. Group differences between mothers who encouraged their adolescent to diet and mothers who did not encourage dieting were analyzed using t-test and chi-square analyses. Logistic regression analyses to assess the relationships between mothers’ dieting status (as reported by the mothers) International Journal of Obesity

and adolescents’ weight concerns and behaviors (as reported by the adolescents) were conducted with SAS 8.2.26* Unadjusted percentages of weight concerns and behaviors for each group are presented in Table 2 to allow comparison across studies. However, all of the odds ratios and 99% confidence intervals shown in the table are from the logistic regression analyses statistically controlling for the effects of BMI and school. Conservative 99% confidence intervals were used to decrease the likelihood of significant effects due to chance while conducting multiple comparisons. All analyses were conducted separately for boys and girls. As presented on Table 3, the same series of analyses were conducted to assess

*Most of the variables in the study were assessed with items whose response options were polytomous or quasicontinuous rather than strictly dichotomous or strictly continuous; therefore, two approaches were taken for data analysis. First, analyses were conducted with logistic regression with the assumption that the variables were categorical. Second, analyses were conducted with t-tests with the assumption that the variables were continuous. The results of the two approaches were compared and the findings were quite consistent. Thus, for ease of interpretation, logistic regression analyses with dichotomous dependent variables are presented.


1583 Table 3 Unadjusted prevalence estimates and adjusted odds ratios of adolescent boys’ and girls’ weight-related attitudes and behaviors by mother’ encouragement to diet Boys

Girls

Mother’s encouragement to diet a

Mother’s encouragement to diet a

No (n ¼ 251) %

Yes (n ¼ 128) %

Weight-related concerns Dissatisfied with weight Care about controlling weight Think about being thinner Worried about weight gain Weigh myself often Weight=shape importance Feel somewhat overweight

15.3 30.9 25.1 24.4 27.8 55.4 15.4

26.0 40.8 46.4 54.2 34.1 69.8 43.9

1.46 1.38 1.40 2.14 1.06 1.48 1.90

Dieting Dieted in the past year Currently dieting to lose weight

17.8 12.2

Binge eating Embarrassed, seen binge-eating

Adolescent attitudes and behaviors

Weight control behaviors (past year) Healthy weight controlc c Less extreme unhealthy weight control c Extreme unhealthy weight control

No (n ¼ 254) %

Yes (n ¼ 172) %

3.27) 2.67) 2.83) 4.29) 2.13) 2.90) 4.21)

36.8 41.9 57.1 64.8 31.6 76.8 36.4

47.6 54.7 69.9 74.9 33.1 85.4 58.9

0.78 1.88 1.00 1.17 1.13 1.30 0.96

46.5 43.3

2.53 (1.24, 5.18) 3.22 (1.47, 7.05)

50.8 42.0

63.9 56.8

0.97 (0.53, 1.79) 0.85 (0.45, 1.60)

2.0

14.6

7.32 (1.54, 34.82)

13.9

16.4

1.08 (0.46, 2.52)

59.3 21.5 1.2

89.7 51.2 9.5

3.44 (1.44, 8.22) 2.48 (1.26, 4.87) 5.48 (0.88, 34.07)

86.2 52.6 9.1

88.3 68.4 13.5

0.81 (0.34, 1.92) 1.31 (0.72, 2.41) 1.13 (0.42, 3.05)

b

OR (99% CI)

(0.66, (0.71, (0.69, (1.07, (0.53, (0.75, (0.86,

OR (99% CI)b

(0.41, (1.03, (0.53, (0.61, (0.60, (0.60, (0.49,

1.48) 3.43) 1.90) 2.27) 2.12) 2.81) 1.86)

a

Referent group for logistic analyses. Logistic analyses adjusted for the effects of adolescents’ observed BMI and school. Summary scores in which adolescents who did not endorse any behaviors were compared to adolescents who reported at least one behavior.

b c

the relationships between mothers’ encouragement to diet and adolescents’ weight concerns and behaviors. To assess racial=ethnic differences in the relationships between mothers’ encouragement to diet and adolescents’ weight concerns and behaviors, logistic regression analyses were conducted for each of the following racial=ethnic groups (separately for boys and girls): white, black, Asian, Hispanic, Native American. All analyses controlled for the effects of BMI and school. Again, 99% confidence intervals were used to detect statistical significance. Logistic regression analyses were not conducted for the mixed=other racial= ethnic group due to very small sample sizes. These analyses are important and feasible given the racial=ethnic diversity of the sample; however, the findings should be interpreted with caution due to the small sample sizes once the groups were split by gender (eg some cells less than n ¼ 10). The same series of logistic regression analyses as above were conducted to examine the relationships between mothers’ encouragement to diet and adolescents’ weight concerns and behaviors for the three SES groups.

Results Mothers’ weight-related concerns and behaviors As shown in Table 1, on average, mothers were heavier than recommended by federal guidelines.25 Almost one-quarter of

mothers were classified as obese and a similar percentage were classified as overweight, resulting in more than half of the mothers classified into the overweight or obese categories. Dieting was a commonly reported behavior by mothers as 63.4% reported that they dieted to lose weight or keep from gaining weight. Mothers who reported dieting were significantly heavier (BMI: mean ¼ 27.6, s.d. ¼ 5.9) than mothers who reported no dieting (BMI: mean ¼ 25.1, s.d. ¼ 5.7), t(769) ¼ 75.68, P < 0.001). Similarly, mothers who reported dieting also reported being more dissatisfied with their body weight (mean ¼ 3.3, s.d. ¼ 1.2) than mothers who reported no dieting (mean ¼ 2.6, s.d. ¼ 1.3), t(730) ¼ 77.76, P < 0.001). Mothers’ perception of adolescents’ current weight did not differ significantly by mothers’ dieting status. More than one-third of mothers (37.3%) reported encouraging their adolescent to diet. About one-third (33.8%) of mothers of sons reported encouraging their sons to diet and 40.4% of mothers of daughters reported encouraging their daughters to diet. Mothers who encouraged their adolescents to diet were significantly heavier (BMI: mean ¼ 28.2, s.d. ¼ 6.7) than mothers who did not encourage their adolescents to diet (BMI: mean ¼ 25.8, s.d. ¼ 5.3), t(470) ¼ 7 5.05, P < 0.001). Mothers’ own body weight satisfaction did not differ significantly by encouragement status. Compared with mothers who did not encourage International Journal of Obesity


1584 their adolescents to diet, mothers who encouraged their adolescents to diet were significantly more likely to report that their adolescents were currently overweight (8.7 vs 38.3%, w2 (1) ¼ 102.3, P < 0.001). Mothers’ encouragement to diet was not significantly related to the adolescents’ age or grade level.

Adolescents’ BMI and overweight classification More than one-third of boys and one-third of girls were classified as overweight or obese using direct measurements (see Table 1). Boys who were encouraged to diet by their mothers were significantly heavier (mean ¼ 25.8, s.d. ¼ 5.9) than boys who were not encouraged to diet (mean ¼ 22.0, s.d. ¼ 3.7), t(171) ¼ 76.52, P < 0.001. Of the boys who were encouraged to diet, 57.0% were classified as overweight or obese, 41.3% were classified as average weight, and 1.7% were classified as underweight. Girls who were encouraged to diet by their mothers were significantly heavier (mean ¼ 26.2, s.d. ¼ 6.1) than girls who were not encouraged to diet (mean ¼ 21.9, s.d. ¼ 3.6), t(215) ¼ 77.80, P < 0.001. Of the girls who were encouraged to diet, 54.0% were classified as overweight or obese, 45.3% were classified as average weight, and less than 1% were classified as underweight.

Adolescents’ weight-related concerns, attitudes and behaviors Mothers’ dieting status was not significantly related to boys’ reports of weight-related concerns, weight-control behaviors, or binge eating regardless of whether analyses controlled for boys’ BMI (see Table 2). Mothers’ dieting behavior was significantly associated with girls’ reports of worrying about weight gain and current dieting; however, these relationships were not significant after controlling for girls’ BMI. That is, when adolescent body mass index was taken into account, adolescents’ weight concerns and dieting behaviors were not more prevalent for girls or boys who had mothers who reported dieting (compared with non-dieting mothers). When boys’ BMI was not taken into account, mothers’ encouragement to diet was significantly associated with boys’ reports of thinking about being thinner, worrying about weight gain, weight and shape importance, feeling somewhat overweight, past year and current dieting, binge eating, and healthy and unhealthy weight-control behaviors. As shown in Table 3, even after controlling for boys’ BMI, boys who were encouraged to diet were significantly more likely to be worried about gaining weight, currently dieting, and have dieted and binged in the past year than boys who were not encouraged to diet. They were also more likely to report both healthy and unhealthy weight-control behaviors. Prior to controlling for girls’ BMI, analyses revealed that mothers’ encouragement to diet was significantly related to girls’ reports of caring about controlling weight, thinking about being thinner, feeling somewhat overweight, past year International Journal of Obesity

and current dieting, and less extreme unhealthy weightcontrol behaviors. However, the only significant difference between the two groups of girls after controlling for girls’ BMI was that girls who were encouraged by their mothers to diet were significantly more likely to care about controlling their weight than girls who were not encouraged to diet. Therefore, far fewer associations were found between mothers’ encouragement to diet and adolescent weightcontrol behaviors for the girls in the study than for the boys.*

Racial=ethnic and socioeconomic status differences After statistically controlling for the effects of boys’ BMI and school, white boys were the only boys to report significantly higher rates of dieting in the past year (OR ¼ 6.4, 99% CI ¼ 2.0, 20.3; Wald’s w2 (1) ¼ 9.85, P < 0.01) and current dieting (OR ¼ 5.3, 99% CI ¼ 1.1, 26.6; Wald’s w2 (1) ¼ 7.15, P < 0.01) when their mothers encouraged them to diet. The significant relationships between mothers’ encouragement to diet and less extreme unhealthy weight-control methods found for the entire sample of boys in the analyses reported on Table 3 held only at the trend level (P < 0.05) for white boys only in the subgroup analyses. None of the associations between mothers’ encouragement to diet and girls’ weight concerns and dieting behaviors were statistically significant for white, black, Asian, Hispanic or Native American girls after controlling for girls’ BMI and school. After statistically controlling for the effects of boys’ BMI and school, boys at the lowest socioeconomic status (SES) level were significantly more likely to report practicing healthy weight-control behaviors when their mothers encouraged them to diet than boys who were not encouraged to diet by their mothers (OR ¼ 5.2, 99% CI ¼ 1.2, 21.5; Wald’s w2 (1) ¼ 8.8, P < 0.01). This relationship was not significant for boys in the higher two SES groups. None of the associations between mothers’ encouragement to diet and girls’ weight concerns or dieting behaviors were statistically significant within SES level after controlling for girls’ BMI and school.

Discussion The present study examined associations between mothers’ direct and indirect dieting behaviors and their adolescents’ self-reported weight concerns and dieting behaviors.

*The influence of mothers’ weight on adolescent weight concerns and dieting was examined by including mothers’ BMI as an independent variable in the logistic regression models along with the independent variable assessing mothers’ encouragement to diet. Including mothers’ BMI in the model did not change the significant findings for mothers’ encouragement on adolescent weight concerns and dieting practices.


1585 Mothers’ self-reported dieting behaviors were not associated with weight concerns or dieting behaviors in their sons and daughters after statistically controlling for adolescents’ BMI. In contrast, mothers’ encouragement of adolescents to diet was associated with several weight-related concerns and weight-control behaviors in their sons even after controlling for the effects of BMI; only one significant relationship was found for daughters. The most striking finding of the present study is the extent of weight loss behaviors reported by boys who were encouraged to diet by their mothers, even after controlling for boys’ BMI. The boys in the present study who were encouraged to diet by their mothers were twice as likely as boys who were not encouraged to diet to worry about weight gain, two to three times more likely to diet, seven times more likely to binge eat, and two to three times more likely to engage in both healthy and unhealthy weight-control practices. The fact that these relationships did not hold for girls suggests that the influence of mothers’ encouragement may not enhance the already existing high levels of weight concern and dieting behaviors in girls of this age. Girls continue to be at high risk for health compromising eating and dieting behaviors regardless of parental encouragement to diet. Interestingly, in the present study, for boys who are encouraged to diet, prevalence rates of weight-related concerns such as weighing frequency, and feeling overweight as well as binge eating and dieting practices were midway between the prevalence rates for the two groups of girls. That is, boys who were encouraged to diet by their mothers were showing similar levels of weight-related concerns and behaviors as girls. This suggests that perhaps boys who are encouraged by their mothers to diet should be considered a group at-risk for health-compromising eating and dieting behaviors, particularly, binge-eating, fasting, eating small amounts of food and skipping meals. Boys who are encouraged to diet by their mothers may be more at risk of dieting and engaging in some unhealthy weight-control practices than boys who do not receive maternal encouragement to diet; however, they may not report more weight concerns. Perhaps boys respond to the direct parental messages about dieting by practicing weight-control behaviors without much weight-related concerns or fears about their weight. Alternatively, the nonsignificant differences in weight concerns between boys who were encouraged to diet and boys who were not encouraged to diet may be completely related to BMI since the analyses controlled for the effects of BMI; overweight boys are more weight-concerned than nonoverweight boys and these differences were controlled for in the analyses. The gender differences found in the present study conflict with earlier research.19 Smolak and colleagues found a greater number of significant associations between parental comments about weight and weight loss attempts in girls than boys while the present study found the opposite. This could be explained by the fact that the previous study assessed relationships between children’s behavior and parent’s comments about children’s ‘weight’ rather than parental comments

about encouraging their children to ‘diet’. In addition, the previous study did not control for BMI and the children in the study were younger (pre-adolescent rather than adolescent). The present study finding for girls corroborate those from one previous study.23 However, present study findings conflict with much of the research to date. The present study found only one difference (caring about controlling weight) between the weight-control practices of girls who were or were not encouraged to diet. In contrast, Dixon and colleagues18 found that girls who were encouraged by their parents to diet were more likely to smoke cigarettes, use laxatives and diuretics and practice fasting than their counterparts, and Schreiber and colleagues22 found an association between mothers’ comments on daughters’ weight and chronic dieting. The differences between the present study findings and those of previous research are probably due to the fact that the present study controlled for girls’ BMI. Prior to controlling for BMI, the analyses of the present study found a relationship between mothers’ encouragement to diet and girls’ reports of past and current dieting and less extreme unhealthy weight-control behaviors (which included fasting and smoking cigarettes). In addition, the source of information differed in the studies. In both previous studies, assessment of parental encouragement was from adolescent self-report rather than from a parent source. The findings of the present study corroborated previous research findings17,18 that do not support the relationship between mothers’ dieting status (indirect influence) and increased weight-related concerns or behaviors in their adolescent children. Hence, modeling dieting behaviors may not necessarily increase the likelihood that adolescent children will diet, particularly if the child’s body mass index is taken into account. These findings give credence to the theory that parents’ own dieting behaviors may be less influential on their adolescent’s behavior than their direct verbal statements regarding their adolescent’s attributes.19 Interestingly, while mothers’ indirect influence (via dieting) was not related to adolescent dieting practices, mothers’ BMI was related to encouraging children to diet. Mothers who encouraged their children to diet were significantly heavier than mothers who did not encourage their children to diet. However, mothers’ encouragement to diet remained significantly related to boys’ weight concerns and dieting even after controlling for mothers’ BMI. Almost half of the girls and almost half of the boys who were encouraged to diet by their mothers were not classified as overweight.1 The findings for girls are similar to those found by Dixon and colleagues.18 However, many of the mothers were overweight. It is possible that overweight mothers were worried about their adolescent’s weight because of their own experiences with weight-control and it may have led them to be vigilant regarding their adolescent’s weight. This scenario leads to two perspectives when trying to understand the role of parents in adolescent weight-control. On the one hand, parents may impart their own weight-related concerns on their adolescents regardless International Journal of Obesity


1586 of the adolescents’ weight status, thereby, imposing unhealthy, unrealistic norms on adolescents who are in the normal weight range.19 This is particularly important given that about half of the mothers in the present study are considered overweight or obese, a finding consistent with rates in the US adult population.4 Parents who are concerned about their children’s weight should be educated to encourage healthy eating habits18 and physical activity to promote their children’s health, including healthy weight-control.4 Although the adolescents who were encouraged to diet in the present study reported engaging in more healthy weightcontrol behaviors than adolescents who were not encouraged to diet, they also reported engaging in more restrictive dieting behaviors. A focus on dieting for weight-control may contribute to an overemphasis on the thinness ideal18 and over time, these behaviors may lead to an increased risk for the development of obesity.27 The present study had many strengths, particularly the ability to simultaneously assess indirect and direct parental encouragement of adolescents to diet and adolescent weightrelated concerns and behaviors for both males and females. To date, the present study is the first to assess the relationships between parental encouragement to diet and weight-related concerns for adolescent boys. In addition, data regarding encouragement to diet was reported by parents rather than children, thus reflecting intentional encouragement of adolescent dieting by mothers. The racial=ethnic diversity of the sample also provided the opportunity to examine the relationships between mothers’ and adolescents’ weight concerns and behaviors in a relatively heterogeneous sample; some subgroup analyses were also possible. Also, adolescent BMI was collected by objective, trained researchers rather than self-report giving credence to the accuracy of these values. The present study had limitations that must be taken into consideration when interpreting the results. For instance, the general nature of the meaning of ‘dieting’ when mothers were asked if they encourage their son or daughter to ‘diet’ prohibited the assessment of different types of dieting. Mothers in the study may have defined dieting in a variety of ways, including restricting intake and promoting eating healthier foods to name a few. Another limitation was the self-report nature of mothers’ BMI. Future research focused on understanding differences in the type of dieting that is encouraged and the associated adolescent weight-related concerns and behaviors would help direct healthy weight management initiatives. Furthermore, future research with large samples of ethnic minorities should further examine the relationships between parental encouragement to diet and adolescent weight-control practices by race=ethnicity in order to identify potential high-risk groups.

Acknowledgements This study was supported by grant R40 MC 00125 from the Maternal and Child Health Bureau (Title V, Social Security International Journal of Obesity

Act), Health Resources and Services Administration, Department of Health and Human Services. We thank Jillian Moe and Scott Mulert for project coordination, and Rose Hilk and Peter Hannan for data management. References 1 Centers for Disease Control and Prevention National Center for Health Statistics. CDC Growth Charts: United States 2000; Advance Data; 314 (8 June revised). 2 Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents. Arch Pediatr Adolesc Med 1995; 149: 1085 – 1091. 3 Epstein LH, Wing RR. Behavioral treatment of childhood obesity. Psychol Bull 1987; 101: 331 – 342. 4 US Department of Health and Human Services. Healthy People 2010, Conference edn, Vol II. US Department of Health and Human Services: Washington, DC; 2000. 5 French SA, Perry CL, Leon GR, Fulkerson JA. Dieting behaviors and weight change history in female adolescents. Health Psychol 1995; 14: 548 – 555. 6 Keel PK, Fulkerson JA, Leon GR. Disordered eating precursors in pre- and early adolescent girls and boys. J Youth Adolesc 1997; 26: 203 – 216. 7 Neumark-Sztainer D, Story M, Resnick MD, Blum RW. Lessons learned about adolescent nutrition from the Minnesota Adolescent Health Survey. J Am Diet Assoc 1998; 98: 1449 – 1456. 8 Patton GC, Selzer R, Coffey C, Carlin JB, Wolfe R. Onset of adolescent eating disorders: Population based cohort study over three years. Br Med J 1999; 318: 765 – 768. 9 Huon G, Lim J, Gunewardene A. Social influences and female adolescent dieting. J Adolesc 2000; 23: 229 – 232. 10 Huon GF, Walton CJ. Initiation of dieting among adolescent females. Int J Eat Disord 2000; 28: 226 – 230. 11 Berel, S, Irving, LM. Media and disturbed eating: an analysis of media influence and implications for prevention. J Primary Prev 1998; 18: 415 – 430. 12 Wertheim, EH, Paxton, SJ, Schutz, HK, Muir, SL. Why do adolescent girls watch their weight? An interview study examining sociocultural pressures to be thin. J Psychosom Res 1997; 42: 345 – 355. 13 Edmunds H, Hill AJ. Dieting and the family context of eating in young adolescent children. Int J Eat Disord 1999; 25: 435 – 440. 14 Shur EA, Sanders M, Steiner H. Body dissatisfaction and dieting in young children. Int J Eat Disord 2000; 27: 74 – 82. 15 Abramovitz BA, Birch LL. Five-year-old girls’ ideas about dieting are predicted by their mothers’ dieting. J Am Diet Assoc 2000; 100: 1157 – 1163. 16 Hill AJ, Weaver C, Blundell JE. Dieting concerns of 10-year-old girls and their mothers. Br J Clin Psychol 1990; 29: 346 – 348. 17 Byely L, Archibald AB, Graber J, Brooks-Gunn J. A prospective study of familial and social influences on girls’ body image and dieting. Int J Eat Disord 2000; 28: 155 – 164. 18 Dixon R, Adair V, O’Connor S. Parental influences on the dieting beliefs and behaviors of adolescent females in New Zealand. J Adolesc Health 1996; 19: 303 – 307. 19 Smolak L, Levine M, Schermer F. Parental input and weight concerns among elementary school children. Int J Eat Disord 1999; 25: 263 – 271. 20 Thelen MH, Cormier JF. Desire to be thinner and weight control among children and their parents. Behav Ther 1995; 26: 85 – 99. 21 Strong KG, Huon GF. An evaluation of a structural model for studies of the initiation of dieting among adolescent girls. J Psychosom Res 1998; 44: 315 – 326. 22 Schreiber GB, Robins M, Striegel-Moore R, Obarzanek E, Morrison JA, Wright DJ. Weight modification efforts reported by black and white preadolescent girls: National Heart, Lung, and Blood Institute Growth and Health Study. Pediatrics 1996; 98: 63 – 70.

Encouragement to diet by mothers JA Fulkerson et al 23 Keel PK, Heatherton TF, Harnden JL, Hornig CD. Mothers, fathers and daughters: Dieting and disordered eating. Eat Disord: Treat Prev 1997; 5: 216 – 228. 24 Baker CW, Whisman MA, Brownell KD. Studying intergenerational transmission of eating attitudes and behaviors: Methodological and conceptual questions. Health Psychol 2000; 19: 376 – 381. 25 National Heart, Lung and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. National Institutes of Health publication; 1998.

26 SAS Institute Inc. SAS=STAT Software 8.2. SAS Institute: Cary, NC; 1999 – 2001. 27 Stice E, Cameron RP, Killen JD, Hayward C, Taylor CB. Naturalistic weight-reduction efforts prospectively predict growth in relative weight and onset of obesity among female adolescents. J Consult Clin Psychol 1999; 67: 967 – 974.

1587

Appendix Table A1 Concepts Weight dissatisfaction

Care about controlling weight Think about being thinner Worry about gaining weight Weigh self often Weight=shape importance

Perception of current weight

Past year dieting

Current dieting

Binge eating

Adolescent survey items and response categories ‘How satisfied are you with your . . . weight?’ Students indicated their satisfaction on a five-point scale from ‘very dissatisfied’ to ‘very satisfied’, and response options were dichotomized into ‘very dissatisfied=dissatisfied’ and ‘neutral=satisfied=very satisfied’. ‘How much do you care about . . . controlling your weight?’ The four response options were dichotomized into ‘not at all’, ‘a little bit’, ‘somewhat’ vs ‘very much’. How much do you agree or disagree with the statement ‘I think a lot about being thinner’. The question had four response options that were dichotomized into ‘strongly disagree’, ‘disagree’ vs ‘agree’, ‘strongly agree’. How much do you agree or disagree with the statement ‘I am worried about gaining weight’. The question had four response options that were dichotomized into ‘strongly disagree’, ‘disagree’ vs ‘agree’, ‘strongly agree’. How much do you agree or disagree with the statement ‘I weight myself often’. The question had four response options that were dichotomized into ‘strongly disagree’, ‘disagree’ vs ‘agree’, ‘strongly agree’. ‘In the past 6 months, how important has weight or shape been in how you feel about yourself?’ (the four response options were dichotomized into ‘not very important’ vs ‘played a part in how I felt about myself’, ‘among the main things that affected how I felt about myself’, ‘the most important things that affected how I felt about myself’). ‘At this time, do you feel that you are: very underweight, somewhat underweight, about the right weight, somewhat overweight, very overweight?’ The somewhat overweight and very overweight categories were compared to the remaining categories. ‘How often have you gone on a diet during the last year? By dieting we mean changing the way you eat so you can lose weight’. The five response options were dichotomized into ‘never’ vs ‘at least once’ by collapsing the following categories: ‘1 – 4 times’, ‘5 – 10 times’, ‘more than 10 times’, and ‘I am always dieting’. ‘Are you currently trying to: lose weight, stay the same weight, gain weight, I am not trying to do anything about my weight?’ The four response options were dichotomized into ‘lose weight’ vs ‘stay the same weight’, ‘gain weight’, and ‘I am not trying to do anything about my weight’. ‘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)?’ Response options were ‘yes’ or ‘no’.

International Journal of Obesity