Overweight Children's Barriers to and Support for Physical Activity


Washington University School of Medicine, Department of Psychiatry, 660 South Euclid, Campus Box 8134, St. Louis, MO 63110. E-mail: wilfleyd@msnotes.wustl.edu


Objective: As the epidemic of overweight increases among youth, research needs to examine factors that may influence children's participation in weight-related health behaviors. This study examined overweight children's perceived barriers to and support for physical activity compared with nonoverweight children.

Research Methods and Procedures: Barriers to and support for physical activity were examined among 84 overweight children attending a summer fitness camp or a university-based weight loss clinic. Barriers and support levels were then compared with those of 80 nonoverweight children of a similar age range.

Results: Body-related barriers were the most predominant barrier type among overweight youth, especially among overweight girls. Overweight children, particularly girls, reported significantly higher body-related, resource, and social barriers to physical activity compared with nonoverweight children and lower levels of adult support for physical activity.

Discussion: Overweight children may be particularly vulnerable to body-related barriers to physical activity, and reducing such barriers may serve as physical activity intervention points most relevant for overweight youth. Future interventions may also benefit from enhancing support for physical activity from adults and peers.


Pediatric obesity is associated with negative physical and psychological health consequences (1)(2). Obesity in childhood is a significant public health problem that requires comprehensive prevention and intervention efforts (3). Critical elements in efficacious child weight control and prevention programs have included strategies to increase physical activity and/or decrease sedentary behavior (4)(5)(6). Interventions for overweight children incorporating a physical activity component may be improved by a better understanding of factors that promote or interfere with physical activity. A potential determinant of physical activity may be perceived barriers to engaging in these behaviors. Perceiving greater barriers to physical activity is among the most consistent negative correlate of children's actual physical activity (7)(8)(9). Conversely, children's perceived ability to overcome barriers is positively related to higher levels of both moderate and vigorous intensity physical activity (10).

Some barriers may impede physical activity more than others (11), and some may be specific to overweight children. The most common physical activity barriers identified by a cross-section of children across the weight spectrum include lack of time because of other obligations (e.g., homework), personal/individual barriers (e.g., lack of interest), and environmental barriers (e.g., weather, no equipment). Body consciousness has been shown to be a barrier for female adolescents in the general population (7) and may be an especially important obstacle to physical activity for overweight children, as these children report lower levels of body-esteem compared with normal weight peers (12). However, it remains unknown whether barriers to physical activity identified among children in general are similar to those perceived by overweight children.

Whereas perceptions of barriers may hinder physical activity, perceived support may facilitate children's physical activity. Parental support, through active involvement during treatment, is integral to overweight children's weight loss success (13). Among community samples of children across the weight spectrum, parental support seems to be consistently positively related to greater amounts of physical activity among older children (9)(14). Parental verbal support (e.g., encouragement) and instrumental support (e.g., transportation to physical activity) are positively related to children's activity levels (15)(16). Peer influences also become increasingly important as children age, and peers have been shown to influence the amount of physical activity in which children engage (17)(18). Overweight children experience higher rates of stigmatization and social isolation compared with leaner children (19)(20), but it is unclear whether overweight children perceive less support for physical activity from parents and peers.

In the current study, overweight children are compared with nonoverweight children to explore how perceived barriers to and support for physical activity differ across child weight status. It is hypothesized that overweight children will report more barriers to physical activity, specifically more body-related barriers, compared with nonoverweight children. It is further hypothesized that overweight children will report greater social-related barriers to physical activity, but no differences are expected between overweight and nonoverweight children on resource barriers. Finally, it is hypothesized that overweight children will report lower levels of adult and peer-related support for physical activity compared with nonoverweight children. For all sets of analyses, gender and age are also examined as factors associated with perceived physical activity barriers and support.

Research Methods and Procedures


The present sample consists of a subset of a group of children who participated in a comprehensive study of eating and physical activity behaviors. This sample was drawn from the following sources: the overweight sample from either a summer fitness camp in Southern California (84%) or a university-based weight loss clinic in San Diego, California (16%), and a convenience sample of nonoverweight children from schools in Westchester County, part of suburban New York City (42.5%), and Southern California (57.5%).

Of the recruited overweight sample of 139, children were excluded from the present analyses if they were outside the selected age range of 8 to 16 years old (n = 11), they did not meet overweight criteria of being at or above the 95th body mass index (BMI)1 percentile for age and gender (26; n = 15), or they had incomplete data (i.e., were missing two or more items from any barrier or support composite; n = 29). The final overweight (OV) sample consisted of 84 children (70% girls) with a mean age of 12.6 ± 2.2 years, a mean BMI of 32.2 ± 5.7 kg/m2, a mean percent overweight of 71.8 ± 24.9%, and a mean BMI z-score of 2.2 ± 0.3. The overweight sample self-reported as 78.8% white, 6.3% Hispanic, 2.5% African American, 3.8% Asian, 1.3% Native American, and 7.5% multiracial or other ethnicity. Maternal education, self-reported by mothers and used as a proxy for familial socioeconomic status, indicated that 54.3% were college graduates.

Nonoverweight children included 5th and 6th grade boys and girls (between the ages of 10 and 14 years) attending an elementary or middle school in California or New York (n = 95). There were no significant differences between the nonoverweight children from New York and California on examined demographic variables or any of the dependent variables, so these samples were combined for analyses. Children were excluded if they were above the 85th percentile of BMI for age and gender, which is considered at-risk for overweight (21) (n = 9), or had incomplete data (n = 6). This left a final nonoverweight (non-OV) sample of 80 children (49% girls) with a mean age of 12.1 ± 0.9 years, a mean BMI of 18.0 ± 2.2 kg/m2, a mean percent overweight of −2.1 ± 11.2%, and a mean BMI z-score of −0.3 ± 0.9. Nonoverweight children were not individually queried about ethnicity or maternal education, and information was not obtained from parents. However, this nonoverweight sample was drawn from school districts with estimated average ethnicity rates of 70% white, 14% Hispanic, 3% African American, 1% Asian, 5% Native American, and 7% multiracial or other ethnicity; 63.3% of children in the school districts had mothers who were college graduates (22).

Weights were measured to the nearest 0.25 lb using a balance beam scale, and heights were measured to the nearest 0.25 in. Missing heights and weights were obtained from medical records. Percent overweight was calculated as the percentage a child's actual BMI was above the 50th percentile of BMI for the child's age and gender provided in the Centers for Disease Control and Prevention National Center for Health Statistics 2000 growth curves (21). BMI percentiles and z-scores were also determined using these growth curves and the related procedures accompanying the growth curves.

Parental consent and child assent were obtained before data collection. This study was approved by the Committee for the Protection of Human Subjects at San Diego State University.


Questionnaires were administered to children by research assistants, who were available throughout assessments to address children's questions about questionnaire items.

Barriers to Physical Activity

A scale was developed to assess children's perception of 21 barriers to their physical activity. Items were generated from prior research on barriers to physical activity in the population (7)(23). Barrier items were prefaced by “How often do the following things prevent you from getting physical activity?” with children choosing a response on a 5-point scale [e.g., never (1) to very often (5); see Table 1 for individual items]. Items were categorized and grouped into body-related, resources, social, fitness, and inconvenience barriers. Mean scores were used as composite scores for each barrier type category. Internal consistencies determined by Cronbach's α were as follows: body-related (0.92), resources (0.77), social (0.73), fitness (0.82), and convenience (0.58); α was comparable across older and younger children.

Table 1.  Means ± SD of barriers to PA by gender and age group for overweight children
 Total (n = 84)Boys (n = 25)Girls (n = 59)8 to 12 years (n = 40)13 to 16 years (n = 44)
  1. Response range: 1 = never to 5 = often.

  2. PA, physical activity.

Body-related2.55 ± 1.31.96 ± 0.82.80 ± 1.52.15 ± 1.22.92 ± 1.4
 Self-conscious about my looks when I do PA2.60 ± 1.52.28 ± 1.12.73 ± 1.62.10 ± 1.33.05 ± 1.5
 Self-conscious of my body when I do PA2.55 ± 1.51.84 ± 0.92.85 ± 1.62.23 ± 1.52.84 ± 1.5
 Don't want people to see my body when I do PA2.51 ± 1.41.76 ± 0.82.83 ± 1.52.13 ± 1.32.86 ± 1.5
Convenience2.45 ± 0.92.31 ± 0.72.51 ± 0.92.24 ± 0.92.64 ± 0.8
 I have too much homework2.74 ± 1.32.64 ± 1.22.78 ± 1.42.55 ± 1.42.91 ± 1.3
 Lack of time2.50 ± 1.22.32 ± 1.12.58 ± 1.22.07 ± 1.12.89 ± 1.2
 Weather is too bad2.11 ± 1.01.96 ± 0.82.17 ± 1.12.10 ± 1.12.12 ± 1.0
Resource2.20 ± 0.82.00 ± 0.82.29 ± 0.82.00 ± 0.82.39 ± 0.8
 Lack of convenient place to do PA2.35 ± 1.12.28 ± 1.12.38 ± 1.22.25 ± 1.32.44 ± 1.0
 Lack of interest in PA2.29 ± 1.22.16 ± 1.12.34 ± 1.22.03 ± 1.12.52 ± 1.2
 Lack of skills2.20 ± 1.21.76 ± 0.92.39 ± 1.31.90 ± 1.22.48 ± 1.2
 Lack of equipment2.19 ± 1.12.00 ± 0.82.28 ± 1.22.00 ± 1.22.36 ± 1.0
 Lack of knowledge on how to do PA1.96 ± 1.01.80 ± 1.12.03 ± 1.01.78 ± 1.12.14 ± 1.0
Social2.09 ± 0.81.92 ± 0.82.17 ± 0.81.94 ± 0.92.24 ± 0.7
 Do not have anyone to do PA with me2.46 ± 1.32.24 ± 1.22.56 ± 1.32.25 ± 1.32.66 ± 1.2
 I'm chosen last for teams2.35 ± 1.42.12 ± 1.22.44 ± 1.52.07 ± 1.42.59 ± 1.4
 No one at my skill level to do PA with me2.11 ± 1.21.72 ± 0.82.28 ± 1.32.05 ± 1.32.16 ± 1.1
 Friends don't like to do PA1.94 ± 1.01.76 ± 0.82.02 ± 1.11.80 ± 0.92.07 ± 1.1
 Friends tease me during PA/sports1.61 ± 1.01.76 ± 1.01.54 ± 1.01.53 ± 1.01.68 ± 1.0
Fitness2.07 ± 0.91.77 ± 0.62.19 ± 0.91.90 ± 0.82.22 ± 0.9
 Too overweight to do PA2.37 ± 1.42.04 ± 1.22.51 ± 1.52.07 ± 1.32.64 ± 1.5
 Physical activity is too much work2.08 ± 1.01.96 ± 0.92.14 ± 1.11.97 ± 1.22.18 ± 0.9
 I do not like how my body feels when I do PA2.12 ± 1.21.68 ± 0.82.31 ± 1.31.98 ± 1.22.25 ± 1.2
 Being active is physically uncomfortable1.88 ± 1.11.56 ± 0.62.02 ± 1.21.70 ± 1.12.05 ± 1.0
 Physical activity is too hard1.87 ± 0.91.60 ± 0.61.98 ± 1.01.73 ± 1.02.00 ± 0.8

The non-OV sample was administered questionnaires during a health class, and because time was limited, they received a subset of the questionnaire items to complete the assessment in this shorter time. Therefore, only the body-related, resources, and social composite scores were calculated for this group.

Support for Physical Activity

This 10-item measure was created to assess perceived level of different types of social support (e.g., verbal encouragement, transportation) from different sources (e.g., adult or peer) for physical activity. The types of support assessed were generated based on research on physical activity for children and adolescents (16)(24). Support items were prefaced by “During a typical week, how often did [source] provide [type of support]?,” and items were rated on a 5-point scale [e.g., never (1) to daily (5); see Table 2 for individual items]. Composite scores were calculated separately for adult and peer support across types of support. The adult support score was calculated as the average between male and female adult support items, and to prevent a bias toward lower support values for children from single parent households, support items missing from one parent were not calculated into the average. Internal consistency was evaluated by Cronbach's α for female (0.74; n = 74) and male adult support (0.73; n = 67) and by a correlation (r = 0.54, p < 0.001) for peer support because it was made up of only two items. The non-OV sample was administered all support items.

Table 2.  Means ± SD of support for physical activity by gender and age grouping for overweight children
 Total (n = 84)Boys (n = 25)Girls (n = 59)8 to 12 years (n = 40)13 to 16 years (n = 44)
  1. Response range: 1 = never to 5 = daily.

  2. PA, physical activity.

Adult2.93 ± 0.93.03 ± 0.82.89 ± 0.82.95 ± 0.82.90 ± 0.8
 Female adult told you PA was good for your health?3.44 ± 1.33.39 ± 1.13.46 ± 1.43.49 ± 1.53.40 ± 1.1
 Male adult told you PA was good for your health?3.37 ± 1.43.67 ± 1.13.26 ± 1.53.39 ± 1.43.36 ± 1.5
 Female adult encouraged you to do PA?3.25 ± 1.03.17 ± 0.83.28 ± 1.13.15 ± 1.13.33 ± 0.9
 Male adults encouraged you to do PA?3.16 ± 1.23.26 ± 1.43.12 ± 1.23.17 ± 1.23.15 ± 1.3
 Female adult provided transportation to PA?2.85 ± 1.23.00 ± 1.42.79 ± 1.22.87 ± 1.32.84 ± 1.2
 Male adult provided transportation to PA?2.53 ± 1.32.67 ± 1.42.47 ± 1.22.63 ± 1.42.44 ± 1.2
 Male adult done PA with you?2.49 ± 1.42.71 ± 1.62.39 ± 1.22.59 ± 1.22.39 ± 1.5
 Female adult done PA with you?2.42 ± 1.22.40 ± 1.32.43 ± 1.22.43 ± 1.12.41 ± 1.3
Peer3.03 ± 1.13.34 ± 0.82.89 ± 1.23.26 ± 1.12.81 ± 1.0
 How often do friends do PA/sports with you?3.25 ± 1.23.79 ± 0.83.03 ± 1.33.50 ± 1.23.02 ± 1.1
 How often do friends encourage you to do PA?2.81 ± 1.32.84 ± 1.22.79 ± 1.33.14 ± 1.32.52 ± 1.2

Data analysis

Among the OV sample, two-way ANOVAs were used to examine main and interaction effects for gender and age (8 to 12 vs. 13 to 16 years) on barrier type composite scores; given the five barrier type composite scores, the criterion for significance was adjusted to p < 0.01 (0.05/5). Paired Student's t tests were conducted separately by gender to compare the frequency of barrier types among each other; given the 10 comparisons, the criterion for significance was adjusted to p < 0.005 (0.05/10). Two-way ANOVAs were also used to examine gender and age effects on perceived support from adults and peers, and paired Student's t tests were used to examine differences in adult and peer support (adjusted significance level, p < 0.025; 0.05/2).

To determine the effects of weight status on perceived barriers to and support for physical activity, a subset of the OV sample was compared with the non-OV sample. The subset OV sample was restricted to the age range (10 to 14 years) of the non-OV sample. Fifty-seven participants of the original OV sample were included in the weight status comparison analyses. This subset OV sample was similar to the full OV sample; they were 68.4% girls, had a mean age of 12.4 ± 1.4 years, had a mean BMI of 31.9 ± 4.9 kg/m2, had a mean percent overweight of 72.0 ± 26.6%, and had a mean BMI z-score of 2.2 ± 0.3. This sample self-reported as 79.6% white, 7.4% Hispanic, 3.7% African American, 3.7% Asian, and 5.6% multiracial or other ethnicity. Barrier composite scores were compared by three separate two-way ANOVAs with weight status and gender as the between-subjects factors and an adjusted statistical significance value of p < 0.017 (0.05/3). Similarly, adult and peer support composite scores were evaluated with two separate ANOVAS, with an adjusted statistical significance value of p < 0.025 (0.05/2). To explore specific barrier and support differences by weight status, independent sample Student's t tests on the 23 individual barriers and support items were conducted, using an adjusted statistical significance value of p < 0.002 (0.05/23).

SPSS version 10.0 (SPSS Inc., Chicago, IL) was used for analyses, and all statistical tests were two-tailed.


Overweight Sample

Barrier composite (body, convenience, resource, social, fitness) scores are presented by gender and age group for the OV sample in Table 1. There was a significant main effect of gender for body-related barriers, with OV girls reporting greater body-related barriers to physical activity than OV boys [F (1, 80) = 7.08, p < 0.009]. OV girls and boys did not differ significantly on any other barriers composite score. There were no significant two-way interactions of gender × age group or main effects of age group for any of the other barriers composite scores. For OV girls, body-related barriers were of higher frequency than resource (p < 0.001), social (p < 0.001), and fitness (p < 0.001) barriers to physical activity. These girls also reported greater convenience than social barriers to physical activity (p < 0.002). The only difference in types of barriers to physical activity that differed among OV boys was that convenience barriers were greater than fitness barriers (p < 0.001; see Table 1).

Mean levels of perceived adult (range, 1.5 to 5) and peer (range, 1 to 5) support for physical activity within the overweight sample are presented in Table 2 by gender and age group. There were no significant gender × age group interactions for either adult or peer support and there were no significant main effects of age group or gender on adult or peer support for physical activity. Perceived level of support for physical activity did not differ by source (adult or peer) among either OV girls or boys.

Comparison of Overweight and Nonoverweight Children

There were no significant weight status × gender interactions for barrier or support composite scores. There were significant weight status main effects, with overweight participants reporting higher body-related [F (1, 133) = 29.1, p < 0.001], resource [F (1, 133) = 13.6, p < 0.001], and social [F (1, 140) = 14.3, p < 0.001] barriers compared with non-OV children (see Table 3). Because of the differences in gender distribution among the subset OV and non-OV samples and evidence of gender effects among the complete OV sample, subset OV to non-OV comparisons were also done separately for girls and boys. Similar to above, OV girls reported significantly more body-related (p < 0.001), resource (p < 0.003), and social (p < 0.001) barriers to physical activity in comparison to non-OV girls. In contrast, for boys, the only significant difference was that OV boys had higher body-related barriers than non-OV boys (p < 0.002), with no significant differences for resource and social barriers by boys’ weight status. As seen in Table 3, the OV children reported significantly higher frequencies for 6 of the 13 barriers compared with the non-OV children, with the most frequent barriers to physical activity all concerning body-related issues.

Table 3.  Means ± SD of barriers to and support for physical activity for OV and non-OV children of similar age range (10 to 14 years old)
 OV Total (n = 57)OV Boys (n = 18)OV Girls (n = 39)Non-OV Total (n = 80)Non-OV Boys (n = 41)Non-OV Girls (n = 39)
  • Response range: 1 = never to 5 = often for barriers and 1 = never to 5 = daily for support.

  • *

    Significant differences between OV and non-OV samples (p < 0.001).

  • Significant differences between OV and non-OV samples (p < 0.013).

  • PA, physical activity.

 Body-related*2.60 ± 1.31.94 ± 0.72.90 ± 1.41.44 ± 0.61.36 ± 0.51.52 ± 0.7
  Self-conscious of my body when I do PA*2.67 ± 1.51.89 ± 0.83.03 ± 1.61.53 ± 0.81.41 ± 0.71.64 ± 1.0
  Self-conscious about my looks when I do PA*2.58 ± 1.42.22 ± 0.92.74 ± 1.61.45 ± 0.81.39 ± 0.71.51 ± 0.8
  Don't want people to see my body when I do PA*2.54 ± 1.41.72 ± 0.82.92 ± 1.41.34 ± 0.71.25 ± 0.51.44 ± 0.8
 Resource*2.21 ± 0.81.97 ± 0.72.32 ± 0.81.69 ± 0.61.60 ± .051.79 ± 0.6
  Lack of interest in PA*2.40 ± 1.22.22 ± 1.12.49 ± 1.21.71 ± 0.81.59 ± 0.71.85 ± 1.0
  Lack of convenient place to do PA2.34 ± 1.22.28 ± 1.12.37 ± 1.21.89 ± 1.01.80 ± 0.91.97 ± 1.1
  Lack of skills2.18 ± 1.21.67 ± 0.82.41 ± 1.31.63 ± 0.81.46 ± 0.71.81 ± 0.9
  Lack of equipment2.16 ± 1.12.00 ± 0.82.24 ± 1.21.75 ± 0.91.71 ± 0.81.79 ± 1.0
  Lack of knowledge on how to do PA1.93 ± 1.01.67 ± 0.92.05 ± 1.01.46 ± 0.71.40 ± 0.61.51 ± 0.7
 Social*2.12 ± 0.91.89 ± 0.72.23 ± 0.91.61 ± 0.51.59 ± 0.51.62 ± 0.5
  Do not have anyone to do PA with me2.51 ± 1.32.17 ± 1.12.67 ± 1.42.00 ± 1.02.15 ± 1.01.85 ± 1.0
  I'm chosen last for teams*2.35 ± 1.42.22 ± 1.32.41 ± 1.51.51 ± 0.81.49 ± 0.91.54 ± 0.8
  No one at my skill level to do PA with me2.14 ± 1.21.61 ± 0.92.38 ± 1.21.80 ± 1.01.68 ± 0.81.92 ± 1.1
  Friends don't like to do PA1.93 ± 1.11.61 ± 0.72.08 ± 1.21.53 ± 0.61.46 ± 0.61.61 ± 0.7
  Friends tease me during PA/sports*1.68 ± 1.11.83 ± 1.01.62 ± 1.11.19 ± 0.51.17 ± 0.41.21 ± 0.5
 Adult2.96 ± 0.82.97 ± 0.72.96 ± 0.83.33 ± 0.93.05 ± 0.93.64 ± 0.8
  Female adult told you PA was good for your health?3.49 ± 1.23.41 ± 0.93.53 ± 1.33.57 ± 1.43.05 ± 1.54.08 ± 1.1
  Male adult told you PA was good for your health?3.35 ± 1.53.56 ± 1.23.26 ± 1.63.40 ± 1.53.03 ± 1.63.78 ± 1.2
  Female adult encouraged you to do PA?3.20 ± 1.03.24 ± 0.83.18 ± 1.13.32 ± 1.33.06 ± 1.23.58 ± 1.3
  Male adults encouraged you to do PA?3.15 ± 1.23.12 ± 1.33.17 ± 1.13.47 ± 1.33.42 ± 1.43.53 ± 13
  Female adult provided transportation to PA?*2.98 ± 1.13.00 ± 1.22.97 ± 1.13.91 ± 1.23.51 ± 1.34.30 ± 0.9
  Male adult provided transportation to PA?2.66 ± 1.32.67 ± 1.42.66 ± 1.33.25 ± 1.23.07 ± 1.23.46 ± 1.2
  Female adult done PA with you?2.46 ± 1.22.33 ± 1.22.53 ± 1.32.52 ± 1.32.06 ± 1.02.97 ± 1.3
  Male adult done PA with you?2.43 ± 1.32.56 ± 1.62.36 ± 1.23.15 ± 1.12.98 ± 1.13.34 ± 1.2
 Peer3.03 ± 1.03.17 ± 0.72.96 ± 1.13.41 ± 1.13.33 ± 1.13.50 ± 1.1
  How often do friends do PA/sports with you?*3.26 ± 1.13.72 ± 0.83.05 ± 1.24.23 ± 1.14.31 ± 0.94.15 ± 1.2
  How often do friends encourage you to do PA?2.79 ± 1.32.61 ± 1.22.88 ± 1.32.65 ± 1.52.48 ± 1.52.84 ± 1.4

There were no significant weight status × gender interactions for either peer or adult support, and there were no main effects of weight status on peer support for physical activity. Overweight children reported receiving lower levels of adult support for physical activity than non-OV children [F (1, 133) = 6.5, p < 0.013; see Table 3]. In analyses within sex, OV girls indicated having lower levels of adult support than non-OV girls (p < 0.001), but there were no support composite score differences between OV and non-OV boys. In analyses of specific support items, OV children reported less support than non-OV children on 2 of 10 support items (see Table 3).


In this study, overweight children, particularly overweight girls, reported body-related concerns as frequent barriers to physical activity. Overweight girls reported higher body-related barriers to physical activity than overweight boys and indicated body consciousness and concern about others seeing their bodies while being active as the most common type of barrier to physical activity. Indeed, the perception of body awareness or consciousness as a potent barrier to physical activity seemed unique to overweight youth, as body-related barriers were the only type of barrier that differed between overweight and nonoverweight boys. Such body consciousness has not previously been cited among the most significant barriers to physical activity for general youth or adult samples (7)(25)(26).

Weight control or other health-promoting interventions for overweight youth may need to encourage physical activity that minimizes body awareness or more proactively fosters better body-esteem among overweight children. Helping children learn to appreciate their bodies for what they can do, rather than how they look, may help to reduce feelings of self-consciousness or embarrassment that may prevent participation in activity. For example, interventions may include components that tie specific body parts (e.g., legs) to activities that children could not do without them (e.g., walk or run). Similarly, encouraging children to focus on how to make their bodies work better and become more fit may provide a concrete, motivating goal for them. Programs can be designed to help children create realistic goals (e.g., run for 15 minutes) that they can build on over time (e.g., increase to 20 minutes of running).

In addition, helping children reduce the value of shape, weight, and overall appearance in self-worth will be central for improving body image. For example, interventions may have children identify positive role models and the characteristics that make that role model important. A child may indicate that his/her parent is a positive role model because he/she is intelligent and caring. Making a clear distinction between the most valued attributes in friends and family members (e.g., intelligent and caring) and the lesser importance of physical appearance (e.g., weight or body shape) is important for reducing social pressures to look a certain way, which may in turn, improve body-esteem. Programs can also provide education on the problems with making judgments based on appearances alone and what might happen if you choose friends completely based on how a person looks. Interventions such as these may lead to better adherence with physical activity or other weight control behaviors among overweight children, who exhibit lower body-esteem than their lean counterparts (12)(27).

It is noteworthy that higher ratings of barriers to physical activity for overweight vs. nonoverweight children were not limited to barrier types associated with possible weight stigmatization, such as body-related or social barriers. Overweight children reported more frequent barriers to physical activity than nonoverweight children, regardless of barrier type, although these differences in nonbody type barriers seemed exclusive to girls. Contrary to hypotheses, overweight children endorsed higher ratings of resource barriers to physical activity (e.g., lack of convenient place to be active, lack of equipment) than nonoverweight youth. Overweight children could be reporting higher levels of barriers across this category because this is a socially acceptable reason for low levels of exercise; however, overweight children did not report higher levels of all barriers, and there is no empirical basis for knowing which barriers would or would not be considered socially acceptable. Whereas the average frequency of reported barriers occurring was similar to that seen in prior research among youth across the weight spectrum (7), greater barriers among overweight youth highlight a potentially important impediment to promoting physical activity within this population.

The general gender difference observed in overall perceived barriers to physical activity, with girls reporting higher barriers, seems to be a reliable finding (7)(11). This may be part of gender-based differences in the socialization around physical activity (7). Gender differences among barriers to physical activity may help explain the divergent levels of actual physical activity in late adolescence between boys and girls, but the factors promoting greater perceived barriers to physical activity and actual level of physical activity by girls remain unclear (7). Attempts to reduce barriers to physical activity in children need to target girls. Findings in this study regarding support for physical activity also were gender specific. That is, overweight girls reported receiving less adult support for physical activity than their lean counterparts, whereas overweight boys reported similar levels of adult and peer support for physical activity as nonoverweight boys.

The specific type of lower support (i.e., peer support) may be especially troublesome for children as they move through adolescence and their physical activity becomes more influenced by peers (18). Given the significant social stigma associated with being overweight (28)(29), the combination of perceived higher barriers to and lower support for physical activity may be significant impediments to overweight youth obtaining more physical activity, which may perpetuate their overweight status. It is clear that parental involvement is essential in weight control attempts among younger children (30). However, as with the unknown impact of improving body-esteem on physical activity, effects of interventions that work to augment peer support for overweight children's physical activity remain generally untested (for an exception, see Ref. (31)).

Support from parents or caregivers has been shown to be related to children's physical activity (9), especially if such support is action-oriented (e.g., being active with children), rather than simply providing prompts to be more active (15)(16). In this study, older overweight girls reported receiving the least adult support, possibly reducing the likelihood of their participation in more active behaviors (16). Unlike their male counterparts, family support may be especially pertinent to adolescent girls to be vigorously active (32).

Present findings highlight the importance of considering body-related barriers to physical activity when evaluating and attempting to intervene with children's physical activity, especially for overweight children. In addition, this study replicated the low-to-moderate frequencies of reported barriers to physical activity among youth across the weight spectrum (7). For instance, the level of convenience barriers identified as pressing by both the overweight and nonoverweight youth in the present study are similar to those identified by general adolescent (7)(11) and adult samples (25)(26). This is especially true for time constraints as a commonly reported barrier to physical activity.

Conclusions in this study are limited to treatment-seeking overweight children, and it is unknown whether these results can be generalized to the majority of overweight children. Comparisons are lacking between treatment-seeking and nontreatment-seeking overweight children on barriers to and support for physical activity. Some studies have examined weight status differences in constructs, such as athletic competence, and social, physical, and athletic self-worth, which may be related to the barrier categories in this study. Physical self-worth and athletic competence seem lower in overweight than nonoverweight children (12)(33), with even lower physical self-worth in treatment-seeking than nontreatment-seeking overweight children (33). The relationship between weight status and social acceptance is less consistent (12)(34)(35). Future studies need to examine whether various perceived competencies and aspects of self-esteem are related to perceived barriers to and support for physical activity and whether such factors differ within overweight children based on treatment-seeking status.

Additional limitations to this study include the cross-sectional nature of the data and the comparison of children with different weight status sampled from different populations (clinic and camp vs. school samples). In addition, the sample was primarily white, making generalizations to other ethnic groups limited. Future research into barriers to children's physical activity would benefit from obtaining longitudinal objective physical activity data from a larger and more ethnically diverse sample of children. This would help contextualize findings in terms of clinical implications of barriers as related to activity levels. In addition, it would be helpful to obtain objective evaluation of the perceived barriers where possible (e.g., objective evaluation of physical activity equipment availability). Particularly with barrier and support types that are more objective (resource barriers rather than body-related barriers), it could be that overweight children have inaccurate perceptions of barriers and support, and in fact have similar resources for physical activity as nonoverweight children; pediatric interventions might target both inaccurate perception of and actual levels of barriers and support. Research on physical activity barriers and support may also benefit from randomized trials that implement interventions targeting specific high-ranking barriers, such as body-related barriers among overweight children, or targeting the increase of peer support, and then evaluating physical activity outcomes.

As the epidemic of child overweight continues to rise in the United States (36)(37)(38), research needs to focus on the most effective means of reducing and preventing obesity. Interventions that focus on increasing support while reducing barriers to physical activity, particularly barriers that may be unique to overweight children, may be most promising. This may be an especially important treatment focus for overweight girls, who report the highest relative levels of physical activity barriers and the lowest levels of adult support for physical activity. This study helps identify several potential intervention targets among children that may guide future program development for improving the health status of our youth.


This study was supported by the National Institute of Child Health and Human Development (NIH Grant HD 36904). We thank the campers, students, and patients for their participation in this study, as well as Nancy Lenhart, the director of Camp La Jolla, the camp at which part of the study was performed. The data collection efforts of Amy Kusel Epner in the school settings were also greatly appreciated.


  • 1

    Nonstandard abbreviations: BMI, body mass index; OV, overweight; non-OV, nonoverweight.