The Influence of Friends and Psychosocial Factors on Physical Activity and Screen Time in Normal and Overweight Adolescents: A Mixed-Methods Analysis

Abstract

Background:

Understanding factors that influence physical activity (PA) and sedentary behavior is crucial to develop interventions to improve adolescents’ health-related behaviors.

Purpose:

To compare the influence of friends and psychosocial factors on moderate-to-vigorous physical activity (MVPA) and screen time (ST) between normal weight (NW) and overweight (OW) adolescents.

Methods:

In all, 21 OW and 21 NW adolescents wore accelerometers and completed questionnaires assessing MVPA, ST, and psychosocial variables. The MVPA and ST were assessed in nominated friends. Adolescents participated in focus groups assessing influence on activity behaviors.

Results:

There were no differences in MVPA; however, NW adolescents reported less ST than OW adolescents (8.9 vs 13.1 h/wk, P = .04). For OW adolescents, friends’ ST (P = .002) and psychosocial factors (P = .05) were associated with ST, while only PA self-efficacy was associated with MVPA. For NW adolescents, only friends’ MVPA (P = .04) was associated with self-reported PA. Exploratory analyses revealed differences among weight status and gender. Focus group discussions revealed that friends influenced both OW and NW adolescents’ MVPA; however, this appeared to be more apparent for NW males, while psychosocial factors played a role in both OW and NW females. The OW adolescents reported that friends were more of an influence on their ST levels, while NW adolescents indicated that their ST was not affected by their friends’ behaviors.

Conclusions:

Interventions to increase MVPA and/or decrease ST may need to be tailored for NW and OW adolescents.

Keywords

weight status qualitative social influence sedentary behavior

Introduction

Regular moderate-to-vigorous physical activity (MVPA) during adolescence is associated with numerous physiological and psychological benefits, such as decreased risk of obesity, type 2 diabetes, and depression.^1,2 Despite all of these benefits, only 8% of US adolescents accumulate the recommended 60 minutes per day of MVPA.³ In contrast, screen time (ST) behavior (watching television, using the computer, and playing video games) is associated with many adverse health outcomes, including increased risk of obesity, cardiovascular disease, type 2 diabetes, and depression.⁴ A recent report indicates that US adolescents reported approximately 8 to 8.5 hours per day of total media use.⁵ Although it is possible to have high levels of both MVPA and ST, engaging in MVPA does not completely protect youth from the negative effects of ST.¹ In addition, ST may have greater health risks than other forms of sedentary behavior.⁴ For example, previous evidence has shown that ST, compared to other forms of ST, is associated with unhealthy dietary behaviors and lower amounts of energy expenditure.⁴ Therefore, it is critical to understand factors that influence adolescent MVPA and ST to develop effective interventions to improve these health-related behaviors.⁶

One factor that has been shown to affect activity behavior in adolescents is weight status. Research suggests that overweight (OW) adolescents are less likely to engage in MVPA than normal weight (NW) adolescents and are more likely to have higher ST behaviors.^7,8 Unfortunately, OW adolescents are more likely to experience the negative physiological and psychological health consequences of an inactive lifestyle, which tend to increase in severity as they enter adulthood.⁹ Therefore, there may be an even greater urgency to better understand the factors that affect activity behavior in OW adolescents, especially as they may differ from NW adolescents.

Social cognitive theories assume that personal factors, environmental factors, and behavior interact in a reciprocal fashion.¹⁰ In this study, we focus on self-efficacy (SE) or the belief in one’s ability to attain a behavioral goal. Central to SE levels are sources of efficacy information, including mastery experiences, verbal persuasion, vicarious experiences, and physiological/psychological states.¹⁰ In this study, we examined the influence of friends and enjoyment as sources of efficacy. Evidence has demonstrated that SE in one’s ability to engage in PA is one of the strongest predictors of PA behavior.¹¹ Unfortunately, OW adolescents have reported lower levels of SE, which may help to explain lower levels of MVPA.¹² A number of studies have examined correlates of youth MVPA with demographic (younger age, boys, and Caucasian) and psychosocial (PA SE, PA enjoyment, and PA benefits) variables associated with greater PA,¹³ although these results were for a general sample of adolescents and did not examine weight status. Less is known about correlates of ST in adolescents, but evidence suggests that ST enjoyment and low SE for PA may be associated with higher levels of ST.¹⁴

Social influences may also affect activity behavior; however, MVPA and ST are less well understood with most data based on the individual’s reports of the perceived social support they receive from family and/or friends.^15,16 Previous research indicates that other health behaviors, such as smoking, drug use, and snacking, are influenced by friends’ behavior.^17,18 Recent studies that have measured the behaviors of adolescents’ nominated friends indicate positive associations between individual and friends’ MVPA and ST.^19
-21 Additionally, a study by Jago et al found that children reported friends having an influence on their activity behavior through 3 main mechanisms: (1) modeling of PA, (2) verbal encouragement of PA, and (3) coparticipation in PA with participants.²² In addition, the influence of friends’ behaviors on MVPA and ST may be modified by an individual’s weight status.²³ In a study by Salvy et al, while both NW and OW adolescents demonstrated increased motivation to be physically active when a peer was present, OW adolescents’ motivation to be physically active was greater than the NW adolescents.²⁴ In addition, the OW adolescents cycled a greater distance when a peer was present, as compared to the NW adolescents.²⁴ Experimental evidence suggests that, in addition to increasing PA when a peer is present, OW adolescents may also decrease ST when a peer is present.²⁵ Therefore, analyses attempting to identify friend influences on adolescent MVPA and ST need to also consider the weight status of the individual. Although a few of these studies have found that friends’ PA predicts change in adolescent PA, the mechanisms by which friends’ PA influences adolescent PA are unclear. Questions such as these may be better answered with a qualitative approach. To better understand the influence of friends on PA and ST, Jago et al suggested that qualitative research methods, in addition to a quantitative analysis, would be beneficial. However, few studies have employed such methods.^22,16 The current study utilizes a mixed-methods approach to examine the associations between NW and OW adolescents’ activity and their friends’ activities, as well as examine potential reasons for these associations.

The purpose of this study was to determine, through a mixed-methods approach, whether there were differences in the relationship among nominated friends’ ST and MVPA behaviors, psychological correlates (SE, enjoyment, and barriers to MVPA), and individual MVPA and ST levels according to weight status. We hypothesized that the results from our quantitative analysis would show differences in these relationships between NW and OW adolescents, and that our qualitative data would provide a better understanding for why these differences exist.

Methods

Participants

Participants were 21 NW and 21 OW adolescents matched on sex, race (white vs other), and school level (middle school grades 6-8 vs high school grades 9-12; Table 1). The current study was a part of a larger parent study that examined 152 middle and high school males and females. Schools, located in central Virginia, were randomly chosen to participate in the parent study, and a total of 4 middle schools and 4 high schools opted to participate in the study. Potential participants were identified and randomly chosen from class rosters of participating schools. Participants were classified as “OW/obese” if they had a body mass index (BMI) greater than the 85% for adolescents their height and age (n = 21). A comparison group of 21 NW adolescents, matched by gender, race, and school level, was also included. A parent/guardian signed an informed consent form and the participant signed an assent form. All methods were approved by the institutional review board.

Table 1.

Participant Characteristics for NW and OW Adolescents.^a

	NW Participants, n = 21	OW Participants, n = 21
Demographics
Age	14.7 (1.5)	14.7 (1.6)
BMI percentile	0.49	0.92^b
Females	8 (38)	8 (38)
High school	11 (52)	11 (52)
Caucasian	14 (67)	14 (67)
Behaviors
MVPA (min/d)	52.8 (22.5)	45.9 (18.6)
ST (h/wk)	8.9 (5.2)	13.1 (7.6)^b
Psychosocial factors mean scores
PA self-efficacy	11.6 (2.6)	11.1 (2.8)
No PA enjoyment	4.6 (2.3)	6.5 (3.2)^b
PA barriers	9.6 (2.5)	9.3 (2.7)
ST self-efficacy	30.4 (5.1)	26.8 (5.3)^b
ST enjoyment	35.5 (6.2)	44.8 (5.2)^c
Nominated friends’ variables
Number of nominated friends	1.7 (1.0)	1.4 (0.7)
Friends’ MVPA (h/wk)	8.9 (1.6)	9.0 (2.2)
Friends’ ST (h/wk)	9.6 (5.0)	12.4 (5.5)^c

Abbreviations: MVPA, moderate-to-vigorous physical activity; NW, normal weight; OW, overweight; PA, physical Activity; SD, standard deviation; ST, screen time.

^aMean (SD) or %.

^bSignificantly different P ≤ .05.

^cMarginal P ≤ .10.

Focus Groups

Students from each school level were selected to participate in focus groups, consisting of 4 to 12 participants of the same sex.²⁶ All participants were invited to participate in the focus groups; however, 2 high schools and 1 middle school elected to opt out of focus groups due to time constraints. Therefore, a total of 13 focus groups, stratified by gender and school level, but not weight status, were conducted with 108 participants. For the current study, only OW adolescents and matching NW adolescents were selected from the focus groups. Since the groups were not weight status specific, there were varying numbers of OW participants from 11 of the 13 groups, with 4 OW participants in 2 groups and only 1 OW participant in 3 groups. For the matching of NW to OW participants, NW adolescents were not required to be from the same focus group but did have to be from the same school level. Each focus group lasted approximately 45 to 60 minutes and was conducted by a trained moderator. All focus groups were audiorecorded using a digital recorder, and an assistant moderator took notes of any salient events. The focus groups had a semistructured design with a follow-up process on key topics of interest. Questions were based on PA and ST, friendship groups, and the influence of friends on PA and ST. Participants were also asked questions on the psychological factors associated with activity behaviors, with the moderator focusing on feelings of SE, enjoyment of PA and ST, and perceived barriers to PA. After each question, the assistant moderator recorded the number of positive and negative responses, as well as the nonresponses. The conversations from these sessions were audio recorded, transcribed, and coded based on content response.

PA/ST Measures

The ActiGraph GT3X+ (ActiGraph, LLC, Pensacola, Florida) was used to assess MVPA. The ActiGraph has the ability to detect normal human motion while filtering out high-frequency vibrations that would artificially increase movement data and has been validated for use in children and adolescents in laboratory and field studies.²⁷ Movement was captured in 3 axes and postprocessed to counts per minute (CPM). Participants were instructed to wear the accelerometers for 7 days, except during sleep, swimming, or bathing. Times where ≥60 minutes of accelerometer CPM values were equal to 0 were considered times when the device was not worn²⁸ and were excluded from the analysis. To be included in the analysis, participants were required to have at least 3 days of at least 600 minutes of data per day, with at least 2 weekdays and 1 weekend.^3,29 ActiGraph CPM were categorized using the count cutoff points of ≤100, 101 to 2996, and >2296 to identify sedentary, light, and MVPA, respectively.^28,30 Minutes per day of MVPA were averaged across the number of full days that participants wore the device.

Screen Time

Participants responded to an ST questionnaire^21,31 that asks, “In your free time on an average weekday (Monday-Friday), how many hours do you spend doing the following activities?” These activities included watching television/DVDs/videos, using a computer, and playing videogames (Xbox/PlayStation/other electronic games). Participants were asked to rate the amount of time engaged in these activities on an average weekday and weekend day. The 7 response options were as follows: (1) 0 hours, (2) ½ hour, (3) 1 hour, (4) 2 hours, (5) 3 hours, (6) 4 hours, and (7) 5+ hours. The weighted mean was calculated based on responses to these 6 questions to obtain the weekly hours spent on ST. Test–retest reliabilities for both weekday and weekend items were r = 0.63 and 0.64 for television viewing, r = 0.76 and 0.77 for computer use, and r = 0.72 and 0.84 for electronic games.³¹

Psychosocial Questionnaires

The SE for PA was measured using a previously validated scale³² consisting of 3 questions that measure children’s confidence in their ability to overcome barriers and engage in PA. The stem for each question is, “I can be physically active during my free time on most days…” followed by the statements “no matter how busy my day is,” “even if it is very hot or cold outside,” and “even if I have to stay home.” Response options for this 5-point scale range from 1 (“Disagree a lot”) to 5 (“Agree a lot”). Internal consistency for this scale in a school-based sample of 100 adolescents was α = .76.³²

The SE to reduce ST was measured using a 7-item scale that assesses confidence to reduce ST (eg, plan ahead of time what TV shows you will watch during the week).¹⁴ The 5-point Likert scale contains responses that range from 1 (“I’m sure I can’t”) to 5 (“I’m sure I can”). A higher score indicated that the participant was more confident that he or she could refrain from ST. Test–retest reliability for this scale was r = 0.81, and internal consistency was α = .81.^14,33

The PA enjoyment was measured using a validated modified version of the original scale³⁴ that asks 3 questions that started, “When I am active…” followed by the items “I feel bored,” “I dislike it,” and “it frustrates me.” The 5-point scale contains responses that range from 1 (Agree a lot) to 5 (Disagree a lot), with a higher score indicative of more enjoyment related to PA.

Enjoyment of ST was measured using a validated scale^14,35 that consisted of 10 items that started, “I enjoy doing the following activities…” followed by ST items such as “computer use” and “television viewing.” The 5-point scale contains responses that range from 1 (“Strongly disagree”) to 5 (Strongly agree”).^14,35

Perceived barriers to PA were measured with 4 items that asked, “How often do these things keep you from being physically active?” Items included “The weather is bad,” “I don’t have time to do PA,” “It would take time away from my school work,” and “I’m embarrassed about how I look when I’m active.”³⁶ The 5-point scale ranges from 1 (“Very often”) to 5 (“Never”), with a higher score indicative of fewer barriers. Internal consistency for this scale was α = .49.³⁶

Nominated Friends

In order to compare adolescents’ PA and ST with their friends’ PA and ST, participants invited up to 5 friends to participate in this study.¹⁵ These friends did not have to be in the same grade or attend the same school. Participants were given an envelope for each nominated friend containing a parental consent form, an assent form, a PA questionnaire (G-S Recall; 37, 21), and an ST questionnaire. The PA questionnaire asked nominated friends to record the amount of time spent in MVPA during a typical week.³⁷ The ST was assessed using the same questionnaire participants completed.²¹ An average MVPA minutes per week and ST hours per day score was calculated across each participant’s nominated friend.

Statistical Analysis

No significant differences existed between matching variables (school-level, sex, and race); therefore, independent samples t tests were conducted to examine the difference in objective MVPA and reported ST between NW and OW adolescents. Separate multivariable regression models, for NW and OW, examined the association of psychosocial variables and nominated friends’ behaviors on adolescent MVPA and ST.

Additionally, exploratory analyses were conducted to examine whether there were differences among gender and weight status. One-way analysis of variances were conducted to examine differences among MVPA and ST in the following 4 categories: (1) NW males (n = 13), (2) MW females (n = 8), (3) OW males (n = 13), and (4) OW females (n = 8). Due to the small sample size of each group, Pearson correlations were first conducted to examine significant associations among MVPA, ST, psychosocial factors, and nominated friends’ behaviors. Significant associations were then added into regression models to further examine the association of psychosocial factors and friends’ behaviors on MVPA and ST in NW and OW males and females. All analyses were conducted using SAS version 9.3 with a significance level set a priori at P < .05.

Qualitative Analysis

All focus groups were audiorecorded and transcribed verbatim. Consistent with content analysis,³⁸ transcripts were read line by line and marked with independent codes that described the content response. Two trained researchers coded the transcriptions from the focus groups independently. The researchers subsequently met to refine code definitions and address any inconsistencies. Cohen κ statistic was calculated to assess inter-rater reliability. Cohen κ statistic was 0.85, suggesting that there was good inter-rater reliability. The data and coding structure were then entered into a database in NVivo qualitative analysis software, version 10.0. After the coding was applied in NVivo, the software was used to search the data for patterns of codes, yielding matrices of codes that allowed for the identification of hierarchical codes (categories that describe a broader group of themes). Text retrievals were then performed on the hierarchical codes, and content was analyzed for patterns, interpreted for meaning, and summarized. Selected quotes (deidentified), which represented the broader themes, were selected and provided verbatim.

Results

Quantitative Results

Comparison of OW versus NW adolescent health behaviors

There were no significant differences in minutes per day of MVPA (mean [standard deviation]) between NW (52.8 [22.6] min/d) and OW adolescents (45.9 [18.6] min/d, P = .28). The NW adolescents, compared to OW, reported significantly lower levels of ST (8.9 [7.7] h/wk vs 13.1 [6.6] h/wk, P = .04). The NW adolescents had significantly higher scores of PA enjoyment (P = .03), and SE to reduce ST (P = .03) than OW adolescents. The ST enjoyment and nominated friends’ ST were not significantly different between NW and OW adolescents (P = .09 and P = .08, respectively; Table 1).

Comparison of health behaviors by weight status and gender

When the sample was stratified by gender and weight status, results indicated that OW males (15.61 [7.87]) had significantly greater hours per week of ST (P = .004) compared to NW females (6.13 [4.22]) and OW females (8.14 [3.75]). Additionally, post hoc analysis revealed that NW males (61.79 [22.5]) had significantly greater minutes of MVPA (P = .04) compared to OW females (38.18 [22.45]).

Regression analysis results

In NW participants, friends’ MVPA was positively associated with greater adolescent MVPA (P = .04). In OW participants, PA SE was positively associated with adolescent MVPA (P = .04). For NW participants, neither friends’ behavior nor psychosocial variables were significantly associated with reported ST. For OW participants, self-reported ST was positively associated with friends’ ST (P = .002) and negatively associated with SE to reduce ST (P = .05; Table 2).

Table 2.

Regression Results for MVPA and ST for NW and OW Participants.

	NW Participants, n = 21		OW Participants, n = 21
	β	95% CI	β	95% CI
MVPA
PA self-efficacy	0.33	(−5.78 to 6.42)	3.23	(0.12 to 6.4)^a
PA barriers	−1.22	(−6.71 to 4.26)	−2.34	(−5.8 to 1.17)
No enjoyment of PA	3.63	(−1.67 to 8.93)	1.03	(−1.92 to 3.98)
Friends’ MVPA (h/wk)	6.77	(0.19 to 13.3)^a	1.65	(−2.47 to 5.78)
ST
ST self-efficacy	−0.39	0.13	−0.51	0.05^a
ST enjoyment	0.19	0.3	0.08	0.14
Friends’ ST (h/d)	0.14	0.57	0.83	0.002^a

Abbreviations: CI, confidence interval; MVPA, moderate-to-vigorous physical activity; NW, normal weight; OW, overweight; PA, physical activity; ST, screen time.

^aP ≤ .05.

Exploratory analyses by weight status and gender

Correlational analyses revealed that associations between adolescents’ MVPA, psychosocial factors, and nominated friends’ MVPA differed by gender and weight status. For NW males (n = 13), only friends’ MVPA was associated with adolescents’ MVPA (r = 0.57, P = .03), while for NW females (n = 8), nominated friends’ ST was inversely related to adolescents’ MVPA (r = −0.73, P = .04). For OW females (n = 8), self-reported MVPA was positively associated with PA SE (r = 0.74, P = .05), while neither friends’ MVPA nor psychosocial factors correlated with self-reported MVPA in OW males (n = 13). In contrast, self-reported ST was correlated with friends’ ST for both NW (r = 0.63, P = .01) and OW males (r = .89, P < .0001). For OW females, enjoyment of sedentary behaviors (r = 0.83, P = .02) and SE to reduce ST (r = −0.76, P = .04) were positively associated with self-reported ST, while neither nominated friends nor psychosocial factors were associated with self-reported ST in NW females.

Qualitative results

In all, 20 of the 21 NW and 18 of the 21 OW adolescents participated in the focus group discussions. Tables 3 and 4 display the frequencies of responses and summarize the key themes between NW and OW adolescents. Several notable differences existed between the NW and OW participants, including their behavior during the focus group sessions. Overall, NW participants were more talkative than the OW participants, particularly the girls, and were more likely to expand on the answers that they did provide.

Table 3.

Friends’ Influence on the PA and ST of NW and OW Participants From Focus Group Responses.^a

	NW Participants, n = 20			OW Participants, n = 18
Theme	Yes	No	Did not Respond	Yes	No	Did not Respond
Do friends’ influence PA^b	18 (100)	0 (0)	2	10 (77)	3 (23)	5
Do friends’ influence ST	4 (25)	12 (75)	4	9 (69)	4 (31)	5
Prefer PA with friends	18 (100)	0 (0)	2	15 (79)	4 (21)^c	3
Prefer ST with friends	9 (53)	8 (47)	3	8 (73)	3 (27)	7

Abbreviations: MVPA, moderate-to-vigorous physical activity; NW, normal weight; OW, overweight; PA, physical activity; ST, screen time.

^an (% out of those who responded).

^bValues presented as number of responses from NW (out of 20) or OW (out of 18) groups (%).

^cFour OW participants answered yes and no and were counted twice (% out of 19).

Table 4.

Key Summary of Differences Between NW and OW Participants Based on Analysis of Focus Group Discussions.

Themes	NW Participants	OW Participants
Types of PA	Sports	Sports
Types of ST	Television	TelevisionVideo games
Friends’ influence on PA	CoparticipationVerbal encouragement	CoparticipationVerbal encouragement
Friend’s influence on ST	Little to no influence	Coparticipation
PA with friends or solo	With friends	Mostly with friendsSpecific activities solo
ST with friends or solo	Majority solo	Mixed resultsVideo games with friends
Benefits of PA	CompetitionStress reliefEnjoymentSocial interaction	CompetitionStress reliefEnjoymentWeight control/loss

Abbreviations: NW, normal weight; OW, overweight; PA, physical activity; ST, screen time.

Physical activity and ST behaviors

Both NW and OW adolescents reported sports as the most popular form of PA. The NW adolescents were more likely to report playing on sports teams as their main form of PA. While OW adolescents also reported engagement in sports, they reported sports participation with their friends and not on teams. Also, OW adolescents reported a larger variety of activities, such as hiking and walking. Both NW and OW boys reported playing video games as a main ST activity. The OW boys, however, reported playing video games for longer durations than NW adolescents. All participants reported watching television; however, OW participants were more likely to report greater than 2 hours per day of television watching, whereas NW participants reported only watching television when they had extra time after sports and homework. An NW boy reported that he watched television on a weekday evening only if he was tired from sports or needed a break from homework, while an OW boy reported that he normally watched television for at least 2 hours each weekday evening.

Friendship Groups

Participants were asked to describe their various friendship networks. Both NW and OW reported having family friends, while NW participants were more likely to report spending time with friends from their neighborhood than OW adolescents. Both NW and OW participants reported that they were most active with their sports team friends and least active with their school friends. As 1 NW participant reported:

If you don’t have sports friends, you guys end up watching TV or Netflix and just hanging out. With sports friends, you guys want to do the sports together. The only video game I play is Just Dance which is more fun to do with a friend and it’s kind of active. (NW, high school girl)

However, 2 of the OW boy participants reported only having school friends, while noting that they did not have a large network of friends. As 1 OW participant reported:

I don’t have a ton of friends. Just a couple of really good friends from school. (OW, middle school girl)

Influence of Friends on PA and ST

When asked about their friends’ influence on PA, all of the NW and a majority of the OW adolescents reported that their friends were a positive influence on their PA. For all participants, coparticipation in PA was the most popular form of influence by friends. Verbal encouragement was also a form of influence, although not as strong as coparticipation. As 1 NW participant reported:

They [my friends] actually want to do active things with me. So even if I’m not in the mood and feeling lazy, they’ll come over and grab the soccer ball and we’ll go out. (NW, high school boy)

Two of the NW adolescents reported that there was an equal influence between themselves and their friends on PA; sometimes their friends influenced them to be active, but other times it was they who influenced their friends. As 1 NW reported:

Yeah, I have a friend who I’ll say to her, let’s go outside and play, and other times, I’ll want to watch TV, and she’ll say let’s go outside. So we both kind of influence each other to do active stuff. (NW, middle school girl)

Although the majority of OW participants reported that their friends influenced their PA, in contrast to NW participants, a few of the participants reported no influence of friends on PA. As 1 middle school OW participant reported:

The friends I’m hanging with now prefer to play videogames, so I’d definitely say they don’t influence me to be active. (OW, middle school boy)

For the 3 OW participants who reported no friend influence on PA, 2 of the participants reported that their friends did not enjoy PA and preferred to play video games, while the remaining participant reported that he felt he influenced his friends to be more active. He commented:

I think I influence my friends more. I’m the most active one of my friends, and I always say let’s play football or basketball. (OW, high school boy)

In contrast to PA, there were mixed results concerning the influence of friends on ST. The majority of NW adolescents reported that their friends did not influence them to engage in ST behaviors. Only 1 NW girl reported having a friend from school who always wanted to watch television when they would get together. In fact, 11 of the 16 NW adolescents who responded to this question reported that friends actually influenced them to engage in less ST behaviors. As 1 NW boy reported:

Not really (asked if friends influence ST). Most of my friends don’t watch much TV or play video games. They always want to play sports or go out and do something. (NW, high school boy)

In contrast, OW participants (mainly boys) reported that friends influenced them to increase their ST. Of the 9 OW boys who responded to this question, 7 participants reported that their friends would influence them to play video games, and that they were more likely to play video games when they were with their friends than when by themselves. One OW boy responded:

I’ll only play video games if my friends want to play. It’s no fun to play by yourself. (OW, middle school boy)

The OW girls responded similarly to NW girls, indicating that their friends did not have a major influence on their ST behavior. Two OW girls reported that sometimes their friends want to watch movies or television, but sitting and socializing (not considered ST) were the sedentary behavior their friends were most likely to engage in with participants.

All participants responded that they preferred to engage in PA with friends rather than alone. Interestingly, 4 of the OW participants also reported that they preferred to do certain types of PA alone.

For me it depends on the activity. Like, I prefer to go running by myself. (OW, high school girl)

Reasons for preferring to engage in PA with friends were similar across participants, although some sex differences were noted. Boys engaged in sports and PA with friends because they enjoyed the competitive aspect, while girls responded that socializing and enjoyment were their primary reasons for engaging in PA with friends.

The OW participants reported that they preferred to engage in ST with friends rather than alone. Of the 8 OW participants who reported a preference to engage in ST with friends, 6 of those participants were boys who reported mainly playing video games with friends. In contrast, nearly half of the NW participants who responded to this question reported that they preferred to engage in ST alone rather than with friends. In addition, they responded that when they were with friends, they would rather play sports and spend their time outside, as opposed to stay inside and watch television or play video games.

Motivations to Engage in PA

Both NW and OW participants reported similar physical and psychological health benefits for why PA was important to them, although these reasons differed between boys and girls. Boys reported competition between friends and peers as the main reason they engage in PA and why it is important to them:

It’s really about the competition and being part of the team. [why PA is important] (OW, high school boy)

…like when you’re with a group of friends playing basketball or something, you’ll compete against each other. (NW, high school boy)

For both NW and OW girls, enjoyment of PA was the most popular reason for PA engagement, although both stress-relief and overall feelings of well-being were mentioned by both NW and OW girls as well. The NW girls were more likely to report that they enjoyed PA for the social aspect and that it provided a gateway for meeting new people. As 1 NW girl reported:

You can branch out by joining a sport and doing something active like going to a class. It can be a way to meet new people [on the benefits of PA]. (NW, high school girl)

However, OW girls were more likely to suggest additional psychological benefits such as a feeling of accomplishment, and 3 OW girls reported weight loss as an additional benefit:

It keeps you from getting too fat [on benefits of PA]. (OW, middle school girl)

Additionally, 2 of the OW girls reported that parents pushed them to be more active and emphasized the importance of PA to them. These 2 OW girls reported that this actually made them less motivated to be active.

My dad pushed me into volleyball and I remember crying and they [my parents] signed me up for camps that I didn’t even know about. They always want me to be active, and that makes me want to not do anything but watch TV. (OW, high school girl)

Discussion

The purpose of this study was to determine whether adolescent obesity status modifies the relationships among nominated friends’ behaviors, psychological correlates, and individual MVPA and ST behavior. Nominated friend MVPA was a factor associated with NW individual’s MVPA but not for OW participants. In contrast, nominated friend ST was a factor associated with OW participant’s ST but not for NW participants. The qualitative results generally supported the quantitative results and further indicate that different intervention strategies may be needed for MVPA and ST that are also tailored by sex and an individual’s weight status.

Moderate-To-Vigorous Physical Activity

In contrast to previous research suggesting that OW adolescents engage in lower levels of MVPA than NW adolescents,^39,40 there were no significant differences in MVPA levels between NW and OW participants; however, the limited sample size made it difficult to detect significant differences (Table 1). Additionally, when the sample was split by both gender and weight status, NW males had significantly greater levels of MVPA than OW females, suggesting there may be an interaction effect between gender and weight status. Due to the inadequate sample size when split according to gender and weight status (n = 13 for males; n = 8 for females), these results must be interpreted with caution. We did observe that friends’ MVPA was associated with NW individual’s MVPA, which supports the findings from our focus group analyses, and is supported by previous studies.^19,21 In analyses that controlled for weight status, Sirard et al²¹ examined the association of friends’ MVPA with adolescents’ self-reported MVPA in a sample of 2126 adolescents, finding that MVPA behavior in adolescents was positively associated with their friends’ MVPA. Similar to the lack of quantitative findings regarding the association between friends’ MVPA and self-reported MVPA in OW adolescents, our focus group results indicated that friends were not a factor in MVPA behaviors in a subsample of OW adolescents. These findings are in contrast with those of Salvy et al²⁴ who reported that friends significantly influenced PA in both OW and NW youth; however, that study differs from the present study in that it undercontrolled laboratory settings. Furthermore, another study by Salvy et al⁴¹ found that OW youth engaged in higher intensity activity in the presence of friends than NW youth, which is in contrast to the findings of the current study, although interpretation of PA intensity was not a focus of the qualitative discussions and warrants further investigation. Interestingly, Salvy et al⁴¹ found that OW adolescents reported spending more time alone than NW adolescents, which is partially supported in the current qualitative findings (comments from OW participants about engaging in PA alone). There are a few differences between the study by Salvy et al⁴¹ and the present study that should be mentioned. Salvy et al⁴¹ objectively explored participants’ behaviors when in the presence of friends, peers, and family over a 7-day period, while the present study is based on qualitative findings from the focus group discussions and correlations between participants’ and friends’ behavior. Further exploration into the reasons for OW adolescents’ preference to engage in PA alone is necessary. For example, do OW adolescents prefer to engage in PA alone due to fear of peer rejection or lack of friend support? Additionally, the qualitative findings, with support from the exploratory associations regarding gender and weight status, indicate that gender differences exist among NW and OW adolescents, suggesting that investigators may want to consider both gender and weight status when designing optimal interventions to promote PA.

During the focus groups, several participants commented on the bidirectionality of the influence on MVPA; friends influenced their MVPA but sometimes participants influenced their friends to be more active. With cross-sectional data, we cannot determine in which direction the causal pathway flows. Sophisticated social network analyses^19,20 on longitudinal cohorts support the possibility of reverse causality and social selection (ie, we choose friends who are already similar to ourselves). Such complex social interactions require multiple data points from longitudinal cohorts and complex modeling techniques to fully understand these underlying mechanisms. However, the focus groups were able to capture the essence of that bidirectional influence, without the need for complex quantitative modeling.

From the quantitative data, the psychosocial variables were not strongly associated with MVPA in NW adolescents, although these data should be interpreted with caution as our sample was relatively small and homogeneous, which may have prevented the ability to detect significant associations with individual’s MVPA and ST in the NW sample (Table 2).^13,42 In contrast to the quantitative results, findings from the focus group discussions revealed that NW adolescents reported psychosocial factors, such as enjoyment of PA, were important to promote continued participation in PA. This finding links back to social cognitive theory and how enjoyment, as a psychological state, can be a valuable source of SE. Similar to previous research,^13,43 PA SE was found to be positively associated with MVPA but only in OW adolescents (Table 2).

The OW adolescents reported greater amounts of ST than NW adolescents and, based on focus group comments, placed more importance on their ST behaviors than NW adolescents. Few studies have examined the association of participant’s ST with friends’ ST, and findings have been null or mixed. Sirard et al²¹ reported significant associations between girl participant’s ST and boy friends’ ST, while Ali et al¹⁹ found no association between friends’ and participants’ reported ST in a sample of 3839 adolescents. Both of these studies controlled for weight status rather than performing weight status–specific analyses. Although research on the influence of friends and psychosocial factors on ST in adolescents is limited, existing evidence suggests friends’ ST and SE to reduce ST are associated with ST levels.¹⁴ Our findings in OW adolescents are in agreement with these studies, with both nominated friends’ ST and SE to reduce ST positively associated with adolescents’ ST. From the regression results, both friends’ ST and psychosocial variables were not associated with ST levels in NW adolescents, although further exploration among weight status and gender revealed that friends’ ST may be a factor for NW males. These findings should be interpreted with caution due to the contrast of results from the qualitative analysis and the small sample size of the exploratory correlations. Although the current study did not assess weight status in nominated friends, this may be a factor in the association among ST in NW and OW males and warrants further investigation. Additionally, NW adolescents reported relatively low levels of ST, limiting the ability to detect significant associations in the quantitative analyses; however, focus group findings provide further explanations for low amounts of ST, particularly in NW females, who reported socializing with friends to be their main source of sedentary behavior rather than ST. Further investigation into the types of sedentary behaviors reported by NW and OW adolescents is necessary, given the health risks associated with specific forms of sedentary behaviors.⁴ It should be noted that the current sample of both NW and OW adolescents reported lower amounts of ST (∼1-2 hours per day) compared to the ST levels (∼8 hours per day) of a national sample of youth between the ages of 8 and 18 years.⁵ This discrepancy in ST levels suggests that caution should be taken with generalizing the results of the current sample to other youth populations.

Several limitations should be noted. Although objective measures were utilized to measure MVPA in participants, nominated friends’ MVPA data were subjective, and therefore, it could be prone to bias. Also, demographic characteristics and BMI were not collected from nominated friends, and therefore, the weight status of the nominated friends was unknown. Finally, this was a small, relatively homogeneous sample. Therefore, the current findings cannot be extrapolated to all NW and OW adolescents. Also, not all participants answered every focus group question. Although responses from all participants were encouraged, participants could choose not to answer any question. For example, only 8 of 18 OW participants responded to the question probing about the motivations/benefits of PA. It cannot be assumed that the 10 who did not respond did not find PA to be important or have any benefits. However, it is interesting to note that NW adolescents were more talkative throughout the discussions than the OW participants. This was especially true for girls. Although it is unclear why OW participants were quieter, previous research has suggested that OW adolescents are more self-conscious and afraid of criticism and rejection from NW peers,^23,44 which may account for the differences in response rates during the focus groups. Since the focus group discussions were focused around PA, OW participants may have refrained from discussing their enjoyment of ST and lack of enjoyment of PA due to social desirability, since PA was described by all of the NW participants as “important” and “fun.” Due to the potential bias, it is recommended that future qualitative studies either conduct one-on-one interviews with participants or categorize focus groups according to weight status when discussing weight and weight-related behaviors such as PA and ST.

There were several strengths of the present study. This is the first study, to our knowledge, that examined the influence of friends on MVPA and ST in a matched sample of NW and OW adolescents using a mixed-methods analysis. This is also one of the first studies where participants were able to nominate friends who were not part of the study and, therefore, were not limited to school-based friends,^19
-21 allowing investigators to measure various friendship groups that may have influenced both MVPA and ST behavior. The objective measure of MVPA also was considered a strength of the current study, as the majority of previous investigations relied on subjective measures of MVPA.^19

-22 Additionally, the in-depth nature of a qualitative or mixed-methods approach provided the opportunity to evaluate any discrepancies in findings, which may not be feasible with quantitative approaches. For example, findings from the ST questionnaire indicated that participants reported very low levels of ST; however, investigators were able to probe participants for further details during the focus group discussions, finding that there were discrepancies in the interpretation of the ST items on the questionnaire (eg, reporting 0 hours per day of television watching on questionnaire vs reporting several hours per day watching movies on their iPads during focus group discussions). Further investigation revealed that several participants had interpreted the questionnaire to include only watching movies or shows in front of a television set, which then informed investigators that clarification would be needed to ensure accurate interpretation of the ST questionnaire. Ultimately, although the discussion group findings also indicated that participants engaged in lower amounts of ST compared to national averages,⁵ the mixed-methods approach enabled investigators to obtain a more accurate estimate of current ST levels and provided input on further limitations using quantitative ST questionnaires.

The present data suggest that correlates of MVPA and ST may differ according to weight status. Although results from the qualitative analysis suggest that friends influence PA in both NW and OW adolescents, it appears that friends may have a greater influence in promoting PA in NW adolescents, while other factors such as PA SE may be more important to OW adolescents. Although direct causality cannot be determined from the current study, these findings provide initial support for PA interventions to consider weight status when developing a PA program. Additionally, both the qualitative and quantitative results provide support for further investigations into the possible interaction among weight status and gender with health behaviors. Future research should also examine whether weight status of an adolescents’ social network (eg, family and friends) may moderate the association of ST and PA behaviors in adolescents and their social networks. Such findings may provide support for the consideration of not only adolescents’ weight status and gender but also that of close social ties when developing health behavior interventions.

So What?

As pediatric obesity is a public health issue in the United States, it is necessary to understand the psychosocial factors that drive MVPA and ST in OW adolescents. The results from the present study indicate that effective health interventions for NW adolescents may not be as effective for OW adolescents, suggesting that weight status may need to be considered when developing health behavior interventions. Additionally, these current findings from the qualitative analysis suggest that, although PA and ST have been negatively correlated with each other in previous studies, the influences on these behaviors in adolescents may stem from different mechanisms (eg, PA SE vs friends’ PA) that may differ according to both weight status and gender. Thus, future recommendations may need to consider not only both psychological and social factors when targeting reductions in ST and increases in PA behaviors but also demographic characteristics of the targeted sample.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Curry School of Education, University of Virginia.

References

Chaput

Saunders

T J

Mathieu

MÈ

. Combined associations between moderate to vigorous physical activity and sedentary behaviour with cardiometabolic risk factors in children. Appl Physiol Nutr Metab. 2013;38(5):477–483.

U.S. Department of Health and Human Services. 2008 physical activity guidelines for Americans. October 2018. http://www.health.gov/paguidelines/pdf/paguide.pdf. Accessed March 30, 2017.

Troiano

Berrigan

Dodd

Masse

Tilert

McDowell

. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008;40(1):181–188.

Costigan

Barnett

Plotnikoff

Lubans

. The health indicators associated with screen-based sedentary behavior among adolescent girls: a systematic review. J Adolesc Health. 2013;52(4):382–392.

The Henry J. Kaiser Family Foundation. Generation M2: media in the lives of 8- to 18-year-olds. 20 January 2010. http://kff.org/other/report/generation-m2-media-in-the-lives-of-8-to-18-year-olds/. Accessed March 30, 2017.

Brodersen

Steptoe

Boniface

Wardle

. Trends in physical activity and sedentary behaviour in adolescence: ethnic and socioeconomic differences. Br J Sports Med. 2007;41(3):140–144.

Salmon

Owen

Crawford

Bauman

Sallis

. Physical activity and sedentary behavior: a population-based study of barriers, enjoyment, and preference. Health Psychol. 2003;22(2):178–188.

Epstein

Beecher

Graf

Roemmich

. Choice of interactive dance and bicycle games in overweight and nonoverweight youth. Ann Behav Med. 2007;33(2):124–131.

Tremblay

Colley

Saunders

Healy

Owen

. Physiological and health implications of a sedentary lifestyle, Appl Physiol Nutr Metab. 2010;35(6):725–740.

10.

Bandura

. Self-Efficacy: The Exercise of Control. New York, NY: Macmillan; 1997.

11.

Trost

Pate

Dowda

Ward

Felton

Saunders

. Psychosocial correlated of physical activity in White and African-American girls. J Adolesc Health, 2002;31(3):226–233.

12.

Puhl

Luedicke

. Weight-based victimization among adolescents in the school setting: emotional reactions and coping behaviors. J Youth Adolesc. 2012;41(2):27–40.

13.

Craggs

Corder

van Sluijs

EMF

Griffin

. Determinants of change in physical activity in children and adolescents: a systematic review. Am J Prev Med. 2011;40(6):645–658.

14.

Norman

Schmid

Sallis

Calfas

Patrick

. Psychosocial and environmental correlates of adolescent sedentary behaviors. Pediatrics. 2005;116(4):908–916.

15.

Leatherdale

Wong

. Association between sedentary behavior, physical activity, and obesity: Inactivity among active kids. Prev Chronic Dis. 2009;6(1):A26.

16.

Thompson

Rehman

Humbert

. Factors influencing the physically active leisure of children and youth: a qualitative study. Leisure Sci. 2005;27(5):421–438.

17.

Bahr

Hoffmann

Yang

. Parental and peer influences on the risk of adolescent drug use. J Prim Prev. 2005;26(6):529–551.

18.

O’loughlin

Karp

Koulis

Paradis

DiFranza

. Determinants of first puff and daily cigarette smoking in adolescents. Am J Epidemiol. 2009;170(5):585–597.

19.

Ali

Amialchuk

Heiland

. Weight-related behavior among adolescents: the role of peer effects. PLoS One. 2011;6(6):e21179.

20.

de la Haye

Robins

Mohr

Wilson

. How physical activity shapes, and is shaped by, adolescent friendships. Soc Sci Med. 2011;73(5):719–728.

21.

Sirard

Bruening

Wall

Eisenberg

Kin

Neumark-Sztainer

. Physical activity and screen time in adolescents and their friends. Am J Prev Med. 2013;44(1):48–56.

22.

Jago

Brockman

Fox

Cartwright

Page

Thompson

. Friendship groups and physical activity: qualitative findings on how physical activity is initiated and maintained among 10-11 year old children. Int J Behav Nutr Phys Act. 2009;6:4.

23.

Salvy

Bowker

Germeroth

Barkley

. Influence of peers and friends on overweight/obese youths’ physical activity. Exerc Sport Sci Rev. 2012;40(3):127–132.

24.

Salvy

Roemmich

Bowker

Romero

Stadler

Epstein

. Effect of peers and friends on youth physical activity and motivation to be physically active. J Pediatr Psychol. 2009;34(2):217–225.

25.

Rittenhouse

Salvy

Barkley

. The effect of peer influence on the amount of physical activity performed in 8-to-12-year-old boys. Pediatric Exerc Sci. 2011;23(1):49–60.

26.

Horner

. Using focus group methods with middle school children. Res Nurs Health. 2000;23:510–517.

27.

Puyau

Adolph

Vohra

Butte

. Validation and calibration of physical activity monitors in children. Obes Res. 2002;10(3):150–157.

28.

Trost

Loprinzi

Moore

Pfeiffer

. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sports Exerc. 2011;43(7):1360–1368.

29.

Matthews

Chen

Freedson

. Amount of time spent in sedentary behaviors in the United States, 2003–2004. Am J Epidemiol. 2008;167(7):875–881.

30.

Evenson

Catellier

Gill

Ondrak

McMurray

. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26(14):1557–1565.

31.

Utter

Neumark-Sztainer

Jeffery

Story

. Couch potatoes or French fries: are sedentary behaviors associated with body mass index, physical activity, and dietary behaviors among adolescents? J Am Dietetic Assoc. 2003;103(10):1298–1305.

32.

Motl

Dishman

Trost

. Factorial validity and invariance of questionnaires measuring social-cognitive determinants of physical activity among adolescent girls. Prev.Med. 2000;31(5):584–594.

33.

Norman

Vaughn

Roesch

Sallis

Calfas

Patrick

. Development of decisional balance and self-efficacy measures for adolescent sedentary behaviors. Psychol Health. 2004;19(5):561–575.

34.

Motl

Dishman

Saunders

Dowda

Felton

Pate

. Measuring enjoyment of physical activity in adolescent girls. Am J Prev Med. 2001;21(2):110–117.

35.

Salmon

Owen

Crawford

Bauman

Sallis

. Physical activity and sedentary behavior: a population-based study of barriers, enjoyment, and preference. Health Psychol. 2003;22(2):178.

36.

Elder

Lytle

Sallis

. A description of the social-ecological framework used in the Trial of Activity for Adolescent Girls (TAAG). Health Educ Res. 2007;22(2):155–165.

37.

Godin

Shephard

. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci. 1985;10:141–146.

38.

Patton

. Qualitative Research and Evaluative Methods. Thousand Oaks, CA: Sage; 2002.

39.

Eisenmann

Bartee

Smith

Welk

. Combined influence of physical activity and television viewing on the risk of overweight in US youth. Int J Obes. 2008;32:613–618.

40.

Sallis

Alcaraz

McKenzie

Hovell

. Predictors of change in children’s physical activity over 20 months: variations by gender and level of adiposity. Am J Prev Med. 1999;16(3):222–229.

41.

Salvy

Bowker

Roemmich

. Peer influence on children’s physical activity: an experience sampling study. J Pediatr Psychol. 2008;33(1):39–49.

42.

Spiridon

. Psychological correlates of physical activity in children and adolescents in a cluster analytical approach. Int J Sports Sci. 2011;1(11):9–19.

43.

Barr-Anderson

Young

Sallis

. Structured physical activity and psychosocial correlates in middle-school girls. Prev Med. 2007;44(5):404–409.

44.

Barkley

Salvy

Roemmich

. The effect of simulated ostracism on physical activity behavior in children. Pediatrics. 2012;129(3):e659–e666.