Accelerometer-Based Physical Activity and Shape and Weight Concerns Among Youth With Overweight and Obesity: A Pilot Exploratory Ecological Momentary Assessment Study

Abstract

Background:

A bidirectional association between shape and weight concerns (SWC) and physical activity (PA) has been previously documented. This relationship may be particularly salient among youth with overweight/obesity, given that social marginalization of larger bodies has been associated with elevated SWC and barriers to PA. This pilot study explores reciprocal relationships between momentary SWC and accelerometer-assessed PA behavior.

Methods:

Youth with overweight/obesity (N = 17) participated in a 14-day ecological momentary assessment protocol, during which they were prompted to respond to questions about SWC several times per day. They also continuously wore Actiwatch 2 accelerometers to capture light and moderate-to-vigorous PA behavior.

Results:

Hierarchical linear modeling revealed a unidirectional association between SWC and PA, whereby after engaging in a higher duration of PA, participants reported lower SWC. SWC did not predict subsequent PA.

Conclusion:

The findings support a negative temporal relationship between PA and SWC. While further work is needed to replicate and extend these preliminary findings, they may suggest that PA acutely benefits SWC among youth with overweight and obesity.

Introduction

Physical activity (PA) is associated with favorable physical, psychological, and social indicators of health for youth.¹ Despite these benefits, the majority of youth do not meet the PA recommendation of at least 60 minutes of moderate-to-vigorous PA (MVPA) per day.² Research has shown that PA declines during adolescence,³ and adolescents with overweight and obesity are less physically active than their nonoverweight peers.⁴ Engaging adolescents with overweight and obesity in PA requires understanding the barriers that prevent them from being active.

Shape and weight concerns (SWC; negative attitudes about one's shape and/or weight)⁵ during adolescence have been associated with lower PA, particularly for adolescents with overweight and obesity.^6–8 Adolescents with overweight and obesity have heightened SWC⁹ and experience weight-based stigmatization during PA.^10,11 When compared to children without overweight, children with overweight are more likely to report body-related barriers (e.g., body consciousness, dissatisfaction with one's body shape and weight) to PA.^12,13 Notably, evidence suggests that the link between SWC and PA may be bidirectional, whereby elevated SWC are associated with less PA, and engaging in PA is associated with lower SWC.^8,14,15

For adolescents with overweight/obesity, PA has been shown to improve SWC,^8,16 yet, the social evaluative nature of PA may also be detrimental. For instance, shame about one's physical fitness and appearance have been linked to reduced sport enjoyment¹⁷ and disengagement from PA.^13,18 Overall, the relationship between SWC and PA engagement in adolescents with overweight/obesity is complex and requires further research.

There has been limited research investigating momentary associations between SWC and PA. Such research may be informative, given that SWC and PA engagement fluctuate over the course of the day.^19,20 It may be that SWC deters engagement in PA in the short-term but engaging in PA leads to a temporary improvement in SWC. Ecological momentary assessment (EMA) techniques use mobile devices (e.g., smartphones) to gather real-time repeated self-reports of behaviors and attitudes in the natural environment. EMA is well suited to research time-varying predictors of PA, such as SWC.^19,20 To our knowledge, no studies have used EMA to research the relationship between momentary SWC and PA in youth with overweight and obesity.

A better understanding of these associations at the state-level could help researchers develop momentary interventions to promote PA and adaptive body image in this population. There has also been limited research investigating the temporal relationship between SWC and PA in youth, particularly with device-based measures of PA (e.g., accelerometry). Most research in this area relies on the use of retrospective self-report measures of PA,^6,21,22 which are subject to greater recall biases than device-based measures.²³ Intensive longitudinal methods with device-measured PA are necessary to better elucidate the temporal nature of the SWC and PA relationship among youth with overweight and obesity.

The Present Investigation

The present exploratory pilot study sought to examine how momentary SWC relates to subsequent device-measured engagement in PA (i.e., light and MVPA), and how PA engagement relates to subsequent SWC, among youth with overweight and obesity. Based on extant literature, it was hypothesized that (1) engaging in PA would be related to lower SWC in the subsequent 30 minute period, and that (2) higher SWC would be associated with lower levels of PA in the subsequent 30 minute period.

Methods

Participants

Participants were community-based youth (N = 17; M_age = 10.6 ± 1.54 years; M_BMI-z = 2.11 ± 0.47; 41.2% male; 65.8% African American) with overweight and obesity (BMI ≥85th percentile for age and sex as defined by normative data from CDC)²⁴ who were drawn from a larger pilot EMA study that investigated predictors of maladaptive eating in children and adolescents with overweight and obesity (N = 40). The procedures of the larger study have been previously documented.²⁵ The subsample included participants from the larger study who had any available raw Actiwatch data and had completed at least 1 week of EMA ratings (i.e., completed one or more EMA ratings per day for at least 7 days). Of the 40 participants in the larger study, 23 had no raw Actiwatch data and/or completed <1 week of EMA ratings and were therefore excluded. This resulted in a final sample of 17 participants. The subsample did not differ from the larger sample in terms of age, gender, BMI-z, or number of EMA ratings (p's > 0.39).

Participants were recruited from The University of Chicago and Illinois Institute of Technology through community flyers, pediatrician referrals, and phone logs from previous studies where the families consented to be recontacted.²⁵ Inclusion criteria included fluency in written and spoken English and the ability to read at or above a third-grade reading level. Exclusion criteria included medical conditions or medications influencing weight or appetite; an eating disorder (other than binge eating disorder); and concurrent participation in weight management.

Procedure

Potential participants were screened for eligibility via a telephone interview. Eligible participants attended an in-person laboratory visit with a parent or guardian. After providing written informed assent/consent, participants had their height and weight measured on a digital scale and completed baseline interviews and questionnaires. Participants and their caregivers were then trained on how to complete the EMA protocol, which was administered on a smartphone device. If needed, participants were provided with a loaner smartphone. Participants were also trained on the use of wrist-worn actigraphy monitors (Actiwatch 2, Respironics/Phillips, Bend, OR). Actiwatch and smartphone clocks were synchronized before the start of the EMA protocol.

Participants were instructed to complete three types of EMA ratings: (1) signal-contingent, (2) interval-contingent, and (3) event-contingent.²⁶ Signal-contingent ratings were completed in response to semirandom prompts delivered three times a day on weekdays (07:00–08:00 am, 03:00–04:00 pm, and 06:00–07:00 pm, so as not to conflict with participants' school schedules), and five times a day on weekends (every 2–3 hours between 08:00 am and 09:00 pm). Participants completed interval-contingent ratings in response to prompts sent once a day, before bed. Participants self-initiated event-contingent ratings following eating episodes. During all EMA ratings, participants were asked to report on SWC.

In the present study, only signal-contingent EMA ratings were used in analyses. The 1st day of the EMA rating period was a practice period, during which ≥70% compliance to EMA signals qualified participants to initiate the 14-day study period. Data from the 1-day practice period were excluded from analyses. Participants were instructed to wear the actigraphy monitors continuously (i.e., 24 hours a day) throughout the 14-day study period, except during noncompatible activities (e.g., swimming).

Upon completion of the 14-day study period, participants returned to the institution where they were assessed to return loaner smartphones if needed, complete final assessment procedures, and receive payment for participating in the study (up to $150). Study procedures were approved by each institution's Institutional Review Board.

Measures

Physical activity

Actiwatch 2 wrist-worn accelerometers were used to measure PA, consistent with previous research.²⁷ These accelerometers recorded activity counts in continuous 30-second epochs. Accumulated minutes of light and MVPA were measured from the 30-minute time period before and after each EMA prompt.

Shape and weight concerns

SWC were measured using items adapted from the Youth Eating Disorder Examination-Questionnaire (YEDE-Q)²⁸ for this pilot study. These items capture SWC in the present moment, including feelings of fatness (“How fat do you feel?”); dissatisfaction with shape and weight (“How unhappy are you… with your weight?” “…with your shape?”); and discomfort with shape and weight (“How uncomfortable or embarrassed do you feel seeing your own body?”; “How uncomfortable or embarrassed do you feel about other people seeing your body?”) via a 7-point Likert-type scale (1 = “Not at all”; 7 = “Very much”). Items were summed to create a composite score for momentary SWC (α = 0.94). The YEDE-Q has shown adequate reliability and good concurrent validity in previous samples of youth.^28,29

Statistical Analyses

Actiwatch nonwear time (defined as >60 minutes of continuous zeros) and nonvalid days (defined as days with <10 hours of wear time) were excluded from analyses.³⁰ Any 30-minute windows with <20 minutes of valid wear time were also excluded.³¹ Actiwatch-assessed periods of total PA in the 30 minutes before and after each EMA signal were time-matched to each EMA response. Every EMA survey was time-stamped upon completion, meaning total PA time in the 30 minutes before the EMA signal included the time participants spent responding to the EMA surveys. Descriptive statistics for the SWC sum score and activity time in the 30 minutes before and after EMA prompts were calculated.

Separate hierarchical linear models with time-varying covariates were used to examine temporal associations between SWC and duration of PA. This method was chosen due to the nested structure of the data, with EMA observations/signals (Level 1) nested within participants (Level 2). In addition, it is capable of handling unbalanced designs (i.e., an unequal number of observations for each participant) and is robust to missing data (e.g., skipped EMA surveys).³² Hierarchical linear models were estimated using restricted maximum likelihood estimation. This approach yields unbiased estimates when the number of Level 2 groups (e.g., participants) is small.^33,34 Using the software HLM7.0, the total duration of PA was modeled as a predictor of subsequent SWC. Then, the reverse direction was modeled using SWC as a predictor of subsequent PA. Intercepts were allowed to vary randomly across participants. Slopes were modeled to be fixed due to the small sample size.³⁵

In each model, PA was measured as the total duration of light and MVPA either 30 minutes before or after each EMA prompt. Day (coded continuously ranging from 1 to 14) and sequence of EMA prompt (coded continuously ranging from 1 to 5) were controlled for as simultaneous Level 1 predictors. This model is expressed using the following equation (using the prediction of SWC as a dependent variable and PA as a predictor):

Level 1. Model (within-person) $S W C_{t i} = π_{0 i} + π_{1 i} * (D A Y_{t i}) + π_{2 i} * (T I M E_{t i}) + π_{3 i} * (P A_{t i}) + e_{t i}$

Level 2. Model (between-person) $\begin{matrix} π_{0 i} = β_{00} + r_{0 i} \\ π_{1 i} = β_{10} \\ π_{2 i} = β_{20} \\ π_{3 i} = β_{30} \end{matrix}$

At Level 1, the outcome SWC_t (participant i's level of SWC at time t) was modeled as a function of an intercept (π_0i), and three slopes (π_1i, π_2i, π_3i). Slopes π_1i, π_2i represent the controlled effects of day and time, while slope π_3i is representing the effect of total PA_ti (participant i's total PA 30 minutes before time t). At Level 2, the intercepts (β₀₀_, β₁₀_, β₂₀_, β₃₀) reflect the mean level of SWC, and the main effects of day, time, and PA, in that order.

SWC was exchanged with PA levels, and the analyses were also run in this direction (using the accumulated light and MVPA in the 30 minutes following time t). Within-person predictors were person-mean centered and between-person predictors were group-mean centered. In the initial models, age, sex, and BMI-z were adjusted for at Level 2, but since none was significant or changed the parameter estimates, they were subsequently removed for model parsimony.

Results

The analytic sample consisted of 463 signal-contingent EMA ratings of SWC and time-matched Actiwatch-assessed minutes of PA (Level 1) across 17 participants (Level 2). Each participant completed an average of 27.2 signal-contingent EMA ratings. Overall compliance with signal-continent EMA ratings was 54%. Participants wore the Actiwatches for an average of 13.08 hours per day (range = 3.54–22.62 hours). Table 1 displays descriptive statistics for EMA variables and PA time in the 30 minutes before and following EMA signals.

Table 1.

Descriptive Statistics for Shape/Weight Concern Sum Score and Activity Time (Minutes) in the 30 Minutes Before and After Ecological Momentary Assessment Recordings (N = 17)

	Mean	SD	Minimum	Maximum
Valid wear time (before)	30.0	0.402	22.0	30.0
Valid wear time (after)	30.0	0.399	22.0	30.0
Light activity (before)	17.4	8.51	0.00	30.0
MVPA (before)	0.065	0.429	0.00	5.00
Total physical activity (before)^a	17.5	8.52	0.00	30.0
Light activity (after)	17.8	8.24	0.00	30.0
MVPA (after)	0.060	0.362	0.00	4.00
Total physical activity (after)^b	17.9	8.25	0.00	30.0
Weight/shape concerns	12.8	9.00	5.00	35.0

Note: “before” refers to the 30-minute time window before EMA signals; “after” indicates the 30-minute time window following EMA prompts.

3.2% of observations were 0.

2.5% of observations were 0.

EMA, ecological momentary assessment; MVPA, moderate-to-vigorous physical activity; SD, standard deviation; Total physical activity, sum of light activity and MVPA minutes.

Table 2 describes the results of the hierarchical linear models. There was a significant, negative association between PA and subsequent SWC, meaning that when adolescents engaged in higher levels of PA in the 30 minutes before an EMA signal than their typical average, they reported lower SWC at that signal than their typical average (β = −0.02, standard error = 0.01, p = 0.037). However, the level of SWC reported at an EMA signal did not predict the amount of PA in the 30 minutes following that signal (p = 0.744).

Table 2.

Associations Between Physical Activity and Shape/Weight Concerns (at t) with Day and Time as Level 1 Predictors (N = 17)

Model	β (SE)	p
Physical activity as a predictor of shape/weight concerns
Fixed effects
β₀₀: Intercept	12.518 (2.175)	<0.001
β₁₀: Day_(t)	0.002 (0.028)	0.937
β₂₀: Time_(t)	0.144 (0.084)	0.088
β₃₀: Total physical activity	−0.022 (0.010)	0.037
Random effects
r₀: Between-person intercept	88.649 (0.470)
e: Within-person variance	4.947 (0.111)
Shape/weight concerns as a predictor of physical activity
Fixed effects
β₀₀: Intercept	17.904 (1.602)	<0.001
β₁₀: Day_(t)	−0.086 (0.097)	0.375
β₂₀: Time_(t)	0.211 (0.334)	0.528
β₃₀: Shape/weight concerns	−0.036 (0.109)	0.744
Random effects
r₀: Between-person intercept	16.400 (0.202)
e: Within-person variance	54.630 (0.370)

SE, standard error.

Discussion

The present pilot study explored whether momentary SWC predicted subsequent engagement in PA (i.e., light and MVPA), and whether PA engagement related to subsequent SWC in adolescents with overweight and obesity. This study is the first to investigate short-term relationships between PA and SWC among youth with overweight/obesity, using EMA. Thus, it is an essential first step in understanding the momentary interplay between PA and SWC in this population. Results showed a negative temporal relationship between PA and SWC. Specifically, when adolescents with overweight and obesity engaged in higher levels of PA before an EMA signal, they reported lower scores of SWC at the signal. However, the reverse relationship was not significant; SWC did not predict subsequent PA in the 30 minutes following an EMA signal.

Higher bouts of PA (relative to participants' own averages) before an EMA signal predicted lower SWC reported at that signal in adolescents with overweight/obesity. For every additional minute of PA (relative to the individual's average level of PA), there was a decrease of 0.02 in SWC. Based on this, ∼25 minutes of additional PA would be needed to observe a 7% reduction in SWC (i.e., a 0.5 point change on the SWC scale). This effect, although small, is noteworthy, given that participants predominantly engaged in light-intensity PA during the study period. If this finding is supported by future EMA research, it may suggest that mild PA (e.g., slow walking) acutely benefits SWC among youth with excess weight status. This idea is consistent with research suggesting that marginal increases in PA can positively influence the physical self-perceptions of adolescents with obesity.³⁶

PA is also thought to improve body esteem,²² body image, and self-perceptions.³⁷ For adolescents with excess weight, PA may be a tool to improve positive body image-related constructs (e.g., body esteem), while reducing negative constructs (i.e., SWC). This is important given that adolescents with overweight and obesity are at higher risk for SWC.⁶ It may be that PA redirects adolescents' thoughts from the appearance of their body to its function, which contributes to an immediate reduction in SWC.

Although our research suggests that increased PA in adolescents with overweight and obesity is related to decreased SWC, these exploratory results may also be oversimplified. PA is a social evaluative context where adolescents report feeling appearance-related discomfort.³⁸ Weight-related teasing occurs during PA and disproportionately targets adolescents with overweight and obesity.^10,11 There may be an unmeasured third variable related to the social evaluative nature of PA that explains the relationship between PA and SWC. For instance, low exposure to weight-based teasing during PA may predict greater engagement in PA and lower SWC. Future research should utilize EMA and device-based measures to investigate how social factors could influence the relationship between PA level and SWC.

SWC were not found to predict subsequent PA. This is notable given that body-related barriers to PA (e.g., body dissatisfaction) are highly endorsed by adolescents,³⁹ particularly adolescents with overweight/obesity.¹² Body dissatisfaction (i.e., negative attitudes and feelings about one's body or body parts) and SWC are different but overlapping constructs, which may differentially impact PA.^40,41 If this finding of the study is supported by future research, then it may suggest that momentary SWC do not acutely impede PA.

Although appearance-related barriers to PA are highly endorsed by adolescents with overweight and obesity, they are not the only type of barrier.¹² For example, children (particularly girls) with overweight are more likely to endorse more resource-related (e.g., lack of convenient place to be physically active) and social barriers to PA (e.g., no one to do PA with) than youth without overweight.¹² Adolescents in our sample may face greater barriers outside of SWC which could be important in promoting PA in this population. However, previous research does suggest that adolescent body dissatisfaction may lead to disengagement from PA.³⁹ PA often occurs in a social context where adolescents report appearance-related discomforts.³⁸ SWC and body dissatisfaction could be studied separately in future research to further the understanding of the relationship between PA and body image.

It is also possible that SWC influence PA levels more longitudinally than can be captured in a 30-minute window. This is supported by research that found that adolescents with higher weight concerns were less likely to adhere to PA recommendations than adolescents with lower weight concerns, and this relationship was more pronounced in adolescents with overweight/obesity.⁴²

This study has several limitations that are important to acknowledge. The Level 2 sample size (i.e., the total number of participants) in this study was small. While this limits our ability to generalize the findings to other samples, we had adequate power to detect significant within-person associations of interest due to the large number of observations across participants (i.e., the Level 1 sample size). Due to the small Level 2 sample size, hierarchical models were estimated with fixed (as opposed to random) slopes. Investigating momentary associations between PA and SWC in a larger sample using a random-slope model may provide more robust support for an association between these variables. The sample consisted of children and younger adolescents; therefore, the results may not be generalizable to older adolescents or younger children.

Participants varied significantly in Actiwatch wear times. Research is needed to study factors that could influence wear time to increase compliance in future studies. Finally, this study did not take into account time-varying moderators (e.g., location) which could influence the observed relationships between PA and SWC.

This study also has a number of strengths. This study used EMA along with accelerometry to investigate PA behavior and SWC in the natural environment, which enhances the ecological validity of the findings. The novel use of EMA to investigate the relationship between PA and SWC provides valuable information about how these relate at the state-level among youth with excess weight status, which may inform future EMA studies. This study also included a heterogeneous community-based sample of adolescents which enhances generalizability.

Conclusion

In conclusion, this study contributes novel evidence of an inverse momentary relationship between PA behavior and SWC in youth with overweight and obesity. Thus, it provides an initial understanding of the short-term relationship between PA and SWC in this population. Future research should replicate the findings in larger samples of youth and utilize EMA to explore potential mechanisms that could explain the relationship between PA and SWC.

Impact Statement

Results from this pilot study indicated that physical activity (PA) temporally predicted lower momentary shape/weight concerns (SWC) among adolescents with overweight/obesity. Findings advance our understanding of the short-term relationship between PA and SWC in youth with overweight/obesity, which may ultimately inform interventions to promote PA and adaptive body image.

Footnotes

Acknowledgments

We thank the families who participated in this study and ProActive Kids for acting as a referral source.

Authors' Contributions

E.N.D.: Writing-original draft and writing-review and editing. I.R.: Conceptualization and writing-original draft. A.A.H.-M. and A.B.G.: Funding acquisition, resources, methodology, project administration, writing-review and editing, and supervision. E.P. and C.-H.Y.: Formal analysis and writing-review and editing. K.S. and S.W.: Writing-review and editing. S.G.E.: Funding acquisition, writing-review and editing, and supervision.

Funding Information

This work was supported by grants from the National Center for Advancing Translational Sciences (Grant No. UL1-TR000430), the National Institute of Diabetes and Digestive and Kidney Diseases (Grant Nos. K23-DK105234 and K23-DK128568), and the National Institute of Mental Health (Grant No. T32-MH082761).

Author Disclosure Statement

All authors declare that they have no conflicts of interest.

References

Poitras

, Gray

, Borghese

, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab, 2016; 41:S197–S239; doi: 10.1139/apnm-2015-0663

Singh

, Pattisapu

, Emery

. US physical activity guidelines: Current state, impact and future directions. Trends Cardiovasc Med, 2020; 30(7):407–412; doi: 10.1016/j.tcm.2019.10.002

Brodersen

, Steptoe

, Boniface

, et al. Trends in physical activity and sedentary behaviour in adolescence: Ethnic and socioeconomic differences. Br J Sports Med, 2007; 41(3):140–144; doi: 10.1136/bjsm.2006.031138

Olds

, Ferrar

, Schranz

, et al. Obese adolescents are less active than their normal-weight peers, but wherein lies the difference?. J Adolesc Heal, 2011; 48(2):189–195; doi: 10.1016/j.jadohealth.2010.06.010

Fairburn

, Belgin

. Assessment of eating disorders: Interview or self-report questionnaire?. Int J Eat Disord, 1994; 16(4):363–370; doi: 10.1002/1098-108X(199412)16:4<363::AID-EAT2260160405>3.0.CO;2-%23

Añez

, Fornieles-Deu

, Fauquet-Ars

, et al. Body image dissatisfaction, physical activity and screen-time in Spanish adolescents. J Health Psychol, 2018; 23(1):36–47; doi: 10.1177/1359105316664134

Reel

, Voelker

, Greenleaf

. Weight status and body image perceptions in adolescents: Current perspectives. Adolesc Health Med Ther, 2015; 6:149–158; doi: 10.2147/ahmt.s68344

Gualdi-Russo

, Rinaldo

, Zaccagni

. Physical activity and body image perception in adolescents: A systematic review. Int J Environ Res Public Health, 2022; 19(20):13190; doi: 10.3390/ijerph192013190

Arumuganathan

, Sager

, Mehta

, et al. Psychiatric co-morbidities and body shape dissatisfaction in adolescents with obesity—A school based case controlled study. Indian J Pediatr, 2021; 88(3):235–239; doi: 10.1007/s12098-020-03367-z

10.

Skogen

, Båtevik

, Krumsvik

, et al. Weight-Based victimization and physical activity among adolescents with overweight or obesity: A scoping review of quantitative and qualitative evidence. Front Sport Act Living, 2022; 4:1–20; doi: 10.3389/fspor.2022.732737

11.

Watanabe

, Fontana

, Silva

MP da

, et al. Association between weight-teasing and physical activity in adolescents. Rev Paul Pediatr, 2017; 35(3):309–315; doi: 10.1590/1984-0462/;2017;35;3;00005

12.

Zabinski

, Saelens

, Stein

, et al. Overweight children's barriers to and support for physical activity. Obes Res, 2003; 11(2):238–246; doi: 10.1038/oby.2003.37

13.

Stankov

, Olds

, Cargo

. Overweight and obese adolescents: What turns them off physical activity?. Int J Behav Nutr Phys Act, 2012; 9:1–15; doi: 10.1186/1479-5868-9-53

14.

Sabiston

, Pila

, Vani

, et al. Body image, physical activity, and sport: A scoping review. Psychol Sport Exerc, 2019; 42:48–57; doi: 10.1016/j.psychsport.2018.12.010

15.

Srismith

, Wider

L-M

, Wong

, et al. Influence of physical activity interventions on body representation: A systematic review. Front Psychiatry, 2020; 11:99; doi: 10.3389/fpsyt.2020.00099

16.

Gow

, Tee

MSY

, Garnett

, et al. Pediatric obesity treatment, self-esteem, and body image: A systematic review with meta-analysis. Pediatr Obes, 2020; 15(3):1–12; doi: 10.1111/ijpo.12600

17.

Pila

, Sabiston

, Mack

, et al. Fitness- and appearance-related self-conscious emotions and sport experiences: A prospective longitudinal investigation among adolescent girls. Psychol Sport Exerc, 2020; 47:101641; doi: 10.1016/j.psychsport.2019.101641

18.

Thedinga

, Zehl

, Thiel

. Weight stigma experiences and self-exclusion from sport and exercise settings among people with obesity. BMC Public Health, 2021; 21(1):1–19; doi: 10.1186/s12889-021-10565-7

19.

Dunton

. Ecological momentary assessment in physical activity research. Exerc Sport Sci Rev, 2017; 45(1):48–54; doi: 10.1249/JES.0000000000000092

20.

Fuller-Tyszkiewicz

. Body image states in everyday life: Evidence from ecological momentary assessment methodology. Body Image, 2019; 31:245–272; doi: 10.1016/j.bodyim.2019.02.010

21.

Jensen

, Steele

. Brief report: Body dissatisfaction, weight criticism, and self-reported physical activity in preadolescent children. J Pediatr Psychol, 2009; 34(8):822–826; doi: 10.1093/jpepsy/jsn131

22.

Mak

, Cerin

, McManus

, et al. Mediating effects of body composition between physical activity and body esteem in Hong Kong adolescents: A structural equation modeling approach. Eur J Pediatr, 2016; 175(1):31–37; doi: 10.1007/s00431-015-2586-5

23.

Prince

, Adamo

, Hamel

, et al. A comparison of direct versus self-report measures for assessing physical activity in adults: A systematic review. Int J Behav Nutr Phys Act, 2008; 5:1–24; doi: 10.1186/1479-5868-5-56

24.

Ogden

, Kuczmarski

, Flegal

, et al. Centers for Disease Control and Prevention 2000 growth charts for the United States: Improvements to the 1977 National Center for Health Statistics version. Pediatrics, 2002; 109(1):45–60; doi: 10.1542/peds.109.1.45

25.

Goldschmidt

, Smith

, Crosby

, et al. Ecological momentary assessment of maladaptive eating in children and adolescents with overweight or obesity. Int J Eat Disord, 2018; 51:549–557; doi: 10.1002/eat.22864

26.

Wheeler

, Reis

. Self-recording of everyday life events: Origins, types, and uses. J Pers, 1991; 59(3):339–354; doi: 10.1111/j.1467-6494.1991.tb00252.x

27.

Danielsen

, Skjåkødegård

, Mongstad

, et al. Objectively measured physical activity among treatment seeking children and adolescents with severe obesity and normal weight peers. Obes Sci Pract, 2022; 8(6):801–810; doi: 10.1002/osp4.624

28.

Goldschmidt

, Doyle

, Wilfley

. Assessment of binge eating in overweight youth using a questionnaire version of the Child Eating Disorder Examination with Instructions. Int J Eat Disord, 2007; 40(5):460–467; doi: 10.1002/eat.20387

29.

Kass

, Hurst

, Kolko

, et al. Psychometric evaluation of youth eating disorder examination questionnaire in children with overweight or obesity. Int J Eat Disord, 2018; 50(7):776–780; doi: 10.1002/eat.22693

30.

Smith

, O'Connor

, Mason

, et al. Associations between objective physical activity and emotional eating among adiposity-discordant siblings using ecological momentary assessment and accelerometers. Pediatr Obes, 2021; 16(3):e12720; doi: 10.1111/ijpo.12720

31.

, Mason

, Yi

, et al. Momentary associations between stress and physical activity among children using ecological momentary assessment. Psychol Sport Exerc, 2021; 55:101935; doi: 10.1016/j.psychsport.2021.101935

32.

Brown

. An introduction to linear mixed-effects modeling in R. Adv Methods Pract Psychol Sci, 2021; 4(1); doi: 10.1177/2515245920960351

33.

Huang

. Alternatives to multilevel modeling for the analysis of clustered data. J Exp Educ, 2016; 84(1):175–196; doi: 10.1080/00220973.2014.952397

34.

McNeish

, Stapleton

. The effect of small sample size on two-level model estimates: A review and illustration. Educ Psychol Rev, 2016; 28(2):295–314; doi: 10.1007/s10648-014-9287-x

35.

Matuschek

, Kliegl

, Vasishth

, Baayen

, Bates

. Balancing Type I error and power in linear mixed models. J Mem Lang, 2017; 94:305–315; doi: 10.1016/j.jml.2017.01.001

36.

Daley

, Copeland

, Wright

, et al. Exercise therapy as a treatment for psychopathologic conditions in obese and morbidly obese adolescents: A randomized, controlled trial. Pediatrics, 2006; 118(5):2126–2134; doi: 10.1542/peds.2006-1285

37.

Goldfield

, Kenny

, Alberga

, et al. Effects of aerobic training, resistance training, or both on psychological health in adolescents with obesity: The hearty randomized controlled trial. J Consult Clin Psychol, 2015; 83 (6):1123–1135; doi: 10.1037/ccp0000038

38.

Martins

, Marques

, Sarmento

, et al. Adolescents' perspectives on the barriers and facilitators of physical activity: A systematic review of qualitative studies. Health Educ Res, 2014; 30(5):742–755; doi: 10.1093/her/cyv042

39.

Jachyra

, Gibson

. Boys, transitions, and physical (in)activity: Exploring the socio-behavioural mediators of participation. Physiother Canada, 2016; 68(1):81–89; doi: 10.3138/ptc.2015-19LHC

40.

Allen

, Byrne

, McLean

, et al. Overconcern with weight and shape is not the same as body dissatisfaction: Evidence from a prospective study of pre-adolescent boys and girls. Body Image, 2008; 5(3):261–270; doi: 10.1016/j.bodyim.2008.03.005

41.

Taylor

BC.

Weight and shape concern and body image as risk factors for eating disorders. In: Encyclopedia of Feeding and Eating Disorders. ( Wade

, ed.) Springer: Singapore; 2016.

42.

Sampasa-Kanyinga

, Hamilton

, Willmore

, Chaput

. Perceptions and attitudes about body weight and adherence to the physical activity recommendation among adolescents: The moderating role of body mass index. Public Health, 2017; 146:75–83; doi: 10.1016/j.puhe.2017.01.002