Peer similarity in physical activity and weight status during physical education class

Abstract

PURPOSE:

Examined how moderate-to-vigorous physical activity (MVPA) and weight status during physical education (PE) class was related to peer MVPA and weight status.

DESIGN:

Data collected during the first and last weeks of a semester-long PE class.

SETTING:

Houston, Texas charter school.

PARTICIPANTS:

Adolescents (n=156; 13.51 ± 1.45 years old; 52.6% female; 85.4% Hispanic, 14.1% African American, 0.5% Asian) nominated up to three female and three male students they felt closest to in PE.

MEASURES:

MVPA during PE was assessed by heart rate throughout the semester. Body mass index (BMI) z-score was calculated from height and weight measured at the beginning and end of the semester.

ANALYSIS:

Exponential random graph models explored friendship existence, formation, and maintenance in these networks. Separable temporal exponential random graph models examined how MVPA impacted friendships forming or maintaining over the semester.

RESULTS:

Students with higher MVPA were more likely to receive a new nomination from a student throughout the semester (E=0.007, p<.03), but not to send a new nomination themselves. Students with higher BMI z-scores were significantly less likely to send a nomination at baseline (E=−0.11, p=.04); however, this effect was not significant at semester’s end.

CONCLUSIONS:

MVPA opportunities during PE may be a mechanism for youth to establish friendships, making PE important for adolescent social and physical health. Larger scale research is warranted.

Purpose

Physical education (PE) is an important setting for promoting moderate-to-vigorous physical activity (MVPA) children need to reach physical activity (PA) guidelines, particularly among youth from marginalized communities.¹ PE offers opportunities for socialization through MVPA with peers. However, social factors during PE can positively or negatively impact engagement. Positive social support can motivate adolescents to engage in MVPA;² whereas negative social influences, like weight-based teasing, can discourage MVPA participation.³ PE provides a naturalistic opportunity for a sociocentric social network analysis that has practical implications for how to optimize the social influences in PE to promote MVPA and prevent excess weight gain. This study examined the relationships among adolescents’ BMI, MVPA during PE, and the classmates they identified as friends. This study also examined how these factors were related to the development or maintenance of friendships across the semester.

Methods

Design

In 2018, adolescents (grades 6–12) were recruited from a public charter school in Houston, Texas.⁴ The present study used data collected from the 2018 cohort only and is focused on the first semester (12 weeks) during which all students received the same intensive lifestyle program as their PE class.⁴ Details regarding the program can be found elsewhere.^5,6 Each week consisted of one nutrition day and four active days. Active days included games and station-based activities designed to engage adolescents in MVPA. Students wore heart rate monitors each active day and were graded based on the number of minutes they were in their target heart rate zone. Behavioral modification techniques were overlaid throughout including goal setting, a token economy system (positive behavior reinforced through points students could redeem for prizes), and modeling from undergraduate students.

Sample

To be included in the present secondary analysis, participants had to be in the 2018 cohort and have complete heart rate, pre-post social network, and standardized body mass index (BMI z-score). The final analytic sample included 156 youth in four PE classes. The study was approved by the University of Houston’s Institutional Review Board.

Measures

Demographics

At baseline, participants were asked to complete a demographic questionnaire including their birthdate, gender, grade, race and ethnicity, and nativity status.

Social Networks

At the beginning and end of the semester, adolescents nominated up to three female and three male students they felt closest to in their PE class, in order of closeness, a commonly used, validated procedure for social network analysis.⁷

MVPA

Adolescents wore Polar Heart Rate monitors during class and were given a goal number of MVPA minutes (defined as ≥ 65% of age-estimated maximum heart rate). The instructor gave this goal based on the lesson plan (ranging from 10–30 minutes). From this goal, a percentage of goal minutes achieved was calculated for each day. The mean percentage of MVPA goal minutes achieved each class of the semester was calculated as the variable of interest.

BMI Z-Score

At baseline and the end of the semester, participants’ height and weight were measured in light clothing without shoes to the nearest 0.01 centimeter and kilogram respectively. Measurements were used to calculate BMI, which was translated into age and sex specific BMI z-scores, according to CDC guidelines.⁸

Analysis

Exponential random graph models (ERGM) and separable temporal exponential random graph models (STERGM) were used to explore friendship existence, formation, and maintenance in these networks. Cross-sectional ERGMs determined significant factors associated with the presence of friendship between adolescents in the networks at each independent time point. STERGMs were used to examine how MVPA during class impacted the odds of friendships forming or maintaining over the course of the semester. Models returned parameter estimates (E) and standard errors (SE) for each factor in the model. A factor was deemed significant at a p<.05 level if the parameter estimate is greater than two times the SE. Analyses were completed using statnet package in R Studio.

Results

Sample Characteristics

A greater number of participants were female (52.6%), and most were Hispanic (85.3%). Participants had a mean age of 13.51±1.45 years old and achieved a mean of 70.35% (SD=20.75) of MVPA goal minutes.

ERGM

At both timepoints, adolescents were significantly more likely to nominate adolescents as friends if they were in same class period, of the same gender, and of the same race/ethnicity (respectively for the start of the semester E= 3.28, p<0.001; E=0.39, p<0.001; E=0.18, p=0.04; and the end of semester E=3.66, p<0.001; E=0.32, p<0.001; E=0.24, p=0.002). Similarity in MVPA or BMI z-score did not increase the odds of friendship at either timepoint. MVPA was not associated with sending or receiving nominations at either timepoint. Students with higher BMI z-scores at the start of the semester were significantly less likely to send nominations within this network (E=−0.11, p=.04); this effect was not significant at the end of the semester. BMI z-scores did not significantly alter the odds of students receiving nominations at either timepoint. Supplementary Table 1 provides ERGM results.

STERGM

Similarity in MVPA did not alter the likelihood of friendships being formed or maintained over time. Students who attained a higher percent of MVPA goal minutes during class were significantly more likely to receive new nominations throughout the semester (E=0.007, p<.03). However, students who attained a higher percent of MVPA goal minutes were not more likely to send new nominations. Similarity in BMI z-score did not significantly alter the likelihood of friendships being formed or maintained over time. BMI z-scores did not significantly alter the likelihood of students forming or maintaining incoming or outgoing nominations. Supplementary Table 2 provides STERGM results.

Discussion

Summary

Consistent with previous studies and social marginalization theory, students with higher BMI z-scores were less likely to send nominations within the social network at the beginning of the semester.⁹ This effect was not significant at the end of the semester, suggesting that PE may offer an opportunity to positively alter weight-related social perceptions, possibility through improved self-perception. There was a significant positive relationship between MVPA engagement during class and the likelihood of gaining new incoming nominations over the course of the semester. This finding is aligned with previous studies demonstrating that PA can be an avenue for relationship formation.¹ Finally, inconsistent with the hypothesis, adolescents did not exhibit selection effect (preference toward similarity) during friendship formation or maintenance for MVPA or BMI z-scores. Prior studies in which a selection effect was found were longer-term and included general assessment of PA.¹⁰ Like our study, studies in which no selection effect was found have been shorter-term and occurred in a structured setting with uniform PA opportunities.¹⁰ These are important considerations for future research.

Limitations

This study was conducted in a single school, limiting the generalizability of the findings. Potentially confounding variables, such as socioeconomic status of adolescent’s family, were not collected and instructor adherence was not measured. Future research could include a qualitative component to more fully assess friendships. We assessed MVPA as PA at this intensity to help adolescents reach PA guidelines; future research should examine other PA types or intensities.

Significance

This study addresses three gaps in the MVPA and social networks literature. First, it focuses on a predominately Hispanic adolescent population who are at increased risk for chronic disease development yet underrepresented in social network activity research, adds to the small number of longitudinal studies, and was conducted in a naturalistic setting in which intervention could occur.

So What?

What is known on this topic?

Some studies have observed how MVPA and weight status influence friendship development. ¹⁰

What does this article add?

This study provides evidence for the impact of friendship on (largely) Hispanic adolescents’ MVPA engagement during PE. Findings suggest opportunities for MVPA during PE may allow for youth to establish friendships, further emphasizing the importance of PE for adolescent social and physical health.

What are the implications for health promotion practice or research?

Overall, these findings highlight the potential for PE teachers to enhance social well-being and contribute to a more inclusive and connected learning environment for adolescents through MVPA promotion in PE. Additional research is needed to understand how to optimize the social environment to promote, rather than discourage, MVPA.

Availability of data and material

The dataset supporting the results of the study is available upon request.