**2. The underlying concepts that frame the understanding of research of sedentary and inactive behaviors**

The first premise is derived from a need to know how to identify those children and adolescents who are sedentary and inactive. This is not easy as everyone exhibits some sedentary behavior on a daily basis, however, within the various fields involved in SB and IA research, the concern and focus are on those who exhibit excessive or prolonged sedentary behaviors that if persistent over time will then lead to adverse or at-risk health and well-being outcomes. This identification is complicated further as being sedentary or inactive are not medical conditions or diseases per se, but rather are exhibited from a variety of behavioral repertoires. As a result, there were a multitude of operational definitions of what constituted SB and inactive behaviors which made consensus and comparison between studies difficult. However, a conceptual breakthrough or watershed moment occurred in 2017 when the North America Sedentary Behavior Research Network (SBRN, [4])

*Changing Sedentary Behavior in Children and Adolescents: Understanding Research… DOI: http://dx.doi.org/10.5772/intechopen.114253*

proposed and agreed consensus conceptual definitions of sedentary and inactive behavior and a number of other related behaviors. The SBRN defined SB as physically active behaviors that only achieved energy expenditure ≤1.5 METs while awake and sitting or in a reclining posture. IA behaviors were characterized by energy expenditure from 1.5 to 2.9 METs such as standing in a line (1.8 METs) or slow walking (2.9 METs) according to the Compendium of Energy Expenditure in Youth (CEEY, [5]). Additionally, the SBRN also provided consensus, conceptual definitions, and age-related caveats for sleep, stationary behavior, standing, screen time, non-screen-based sedentary time, sitting, reclining, lying, as well as for bouts, breaks, and interruptions. These definitions are conceptually sound and are merited but also indicate the complexity of behaviors that can be included or need to be accounted for in the development of sedentariness and inactivity. These underlying concepts of the first premise provide the foundation on which this chapter progresses to the second premise of understanding the clues, derived from methodologies of research of SB and IA, which can become cues for changing children and adolescents' lifestyles by either increasing physical activity or decreasing sedentary and inactive behaviors.

#### **3. Understanding methodologies of successful interventions**

The second premise of this chapter is that successful interventions leave clues as to how to ameliorate the impact of IA and SB and lead to behavioral changes that increase physical activity and/or decrease in SB. To provide some insight into the types of interventions that have been recently conducted, four SR [6–9] were selected to examine how a plethora of variables impacted SB and IA. The choice was based on the following considerations: recency of publication, covering research that was pre- and post-COVID-19 lockdown, and allowing examination on how the intervention studies were developed. These SR covered a total of 310 research studies that examined the outcomes and indicators of adiposity (e.g., BMI), biomarkers (e.g., blood pressure), cognitive indicators (e.g., academic achievement), musculoskeletal growth (e.g., fat-free mass), risks (injury)/harm (e.g., headaches), and social-emotional indicators (e.g., classroom time on task), psychological measures (e.g., self-esteem), health-related fitness (e.g., grip muscular strength), technology (e.g., mobile phones), sociodemographic factors (e.g., SES), health risk factors (e.g., smoking), and other movement behaviors (e.g., physical activity). The studies involved over 1.3 million children and adolescents across an array of socioeconomic contexts in over 50 countries. All four SRs were published in either 2022 or 2023, but the reviewed research studies were published between 2004 and 2021. This review of reviews, presented in the alphabetical order of the first author, focused on the research outline, operational definition of IA and SB, sample selection, study aims, methodological issues, the type of intervention, and analysis of results.

Bauman et al. [6] researched individual mobile (using devices such as smartphones) health interventions to reduce IA and SB in children and adolescents. Given the recency of this phenomenon with children and adolescents, only 11 studies that met the selection criteria were reviewed. Although the SBRN definitions were identified, all sedentary behaviors for this study had to be school-related, rather than related to physical activity guidelines. The sample included healthy, overweight,

cancer survivors, and adolescents in the military, while the aims of the studies included reducing obesity, increasing PA, decreasing SB, and promoting health behaviors. The individualized approach of the method was based on the use of mobile/electronic device and therefore was time-consuming in terms of setup. As a result, there were nine unique mobile interventions across 11 studies, and this caused heterogeneity of results which made it difficult to establish the effectiveness of intervention. Bauman and her colleagues analyzed the use of a behavior change technique [10], and this resulted in clusters ranging from 2 to 7 intervention techniques with the most prevalent being goals and planning (10 of 11 studies) and feedback and monitoring (9 of 11 studies). As the interventions were individualized, the fidelity in terms of how consistently the programs were utilized at an individual level would potentially make it difficult to compare between interventions. The authors reported only moderate reductions in inactivity for adolescents, and that there was no support for these interventions working either for SB or children. The approach of involving PA interventions with screen time was novel even with the risks associated with increases in the screentime and has the potential to be more effective with the continued development of better apps and programs.

de Mello et al. [7] developed lifestyle interventions as the basis of identifying clusters and correlates of PA and SB for children and adolescents who ranged in age from 6 to 18 years in an SR of 17 studies. In this review, PA was used instead of IA, while SB was not associated with the SBRN consensus definition. The participants were from macro-projects such as Identifying Determinants of Eating and Activity in Adolescents (IDEA, [11]), while the aims of the studies included reducing obesity, increasing PA, decreasing SB, and promoting health behaviors. In the method, the analysis of lifestyle interventions focused on the development of clusters (e.g., high PA-low SB), which were identified based on cluster input variables (e.g., watching TV). This SR was not concerned with an intervention per se but rather focused on making recommendations for potential lifestyle interventions for separate clusters. The results indicated that few associations were found between sociodemographic variables and all cluster types; however, different cluster patterns of PA and SB were determined and resulted in 12 clusters for boys, 10 for girls, and 9 for boys and girls together. Children and adolescents in the 'high PA-high SB' clusters had higher BMI levels, whereas those in the 'high PA-low SB' clusters presented lower BMI levels, waist circumference, and overweight and obesity. These clusters show the complexity of the relationships between PA and SB and health correlates.

The largest of the SR was undertaken by Kurik et al. [8] in over 42 countries and was designed to determine associations between school-related SB and indicators of health and well-being among children and youth. The authors used the definitions of IA and SB from SBRN [4]. The sample included was over 1.3 million participants, and the aims of the studies included reducing obesity, increasing PA, decreasing SB, and promoting health behaviors. In the method, only SB was operationalized and has exposure to school-based behaviors from outcomes/indicators. The exposures were determined either as critical outcomes (adiposity, biomarkers, cognitive, musculoskeletal growth, risks, and socio-emotional indicators) or as important outcomes (fitness and other movement behaviors which included PA, non-school-related SB, and interestingly sleep!). The role of homework, which is a sedentary activity, was not fully discussed. The authors made no recommendations related to SB per se. The studies were grouped according to seven research designs

#### *Changing Sedentary Behavior in Children and Adolescents: Understanding Research… DOI: http://dx.doi.org/10.5772/intechopen.114253*

(e.g., clustered RCT). The combinations of critical and important outcomes resulted in 1133 associations that were analyzed but were not interventions per se. The study found favorable health associations for 13.5% of the critical outcomes with the highest levels for cognitive (33%) and socio-economic (32%) outcomes. However, favorable health outcomes were associated with only 4% of important outcomes. There were also unfavorable health associations for adiposity (21%), risk (30%), and socio-economic (26%) outcomes, and for the important outcome of other movement behaviors (35%). The null associations for critical outcomes totaled 68% of all associations and 62% for important outcomes. For the exposure categories, homework (29%) and active lessons (72%) were unfavorable for health when compared to more school-related SB. One interesting outcome was that there was a threshold of homework time greater than or equal to 2 hours per day that caused adverse or unfavorable reactions for health and well-being. This was a large study, but with small outcomes.

Whilhite et al.'s [9] SR focused on PA, SB, and sleep duration in associations with physical, psychological, and educational outcomes in children and adolescents in 141 studies with an inclusion criterion that the studies include at least two or movement behaviors. Rather than using the SBRN definition, there was a willingness to adjust according to the definitions used in the reviewed studies, and this led to screen time being used as SB when necessary. The total sample size was not reported, but one longitudinal study included 3979 participants. The aims of the studies included reducing obesity, increasing PA, decreasing SB, and promoting health behaviors. The authors noted that when screentime was used as the SB definition, there were more negative associated outcomes than when SB was defined as overall time spent being sedentary. The role of sleep, which in the SBRN consensus statement was not considered as not applicable to SB, was an interesting association but was not adequately justified. The notional two movement behaviors cut-off for study inclusion suggested that there was an insufficient conceptual basis for which movement behaviors were important. This study did not identify or include any interventions, but rather was focused on associations between variables. The authors' analysis determined that high levels of PA and low levels of SB were favorably associated with physical health, psychological health, and education-related outcomes especially when sleep was included. Adolescents reported stronger associations than children for SB when screentime was used to represent SB rather than overall time spent being sedentary.

Having examined the concepts and methodological guidelines, the sample, aims, types of interventions, and results of these four recent SRs of IA and SB, the second premise was to use this knowledge to identify subsequent clues in order to develop and increase successful IA and SB interventions. The overall conclusions from the four SRs were a mixed set of results toward increasing PA and less favorable results toward decreasing SB. Certainly, there were fine margins between favorable and unfavorable results, and therefore, it was pertinent to identify clues regardless of outcomes. Identifying deductive clues requires an implicit need to act upon them and provide inductive cues. The alliterative clues and cues that follow are derived from four areas. The first three clues and cues are from conceptual, methodological, intervention-based issues from the SR, and the fourth set of clues and cues are based on alternative intervention-based approaches not identified or addressed within the SR. **Table 1** provides a schematic representation of the clues and the resulting cues for the four areas that follow.


*Changing Sedentary Behavior in Children and Adolescents: Understanding Research… DOI: http://dx.doi.org/10.5772/intechopen.114253*


**Table 1.**

*Schematic of the clues and cues for conceptual, methodological, interventions, and alternative approaches.*

## **4. Conceptual clues and cues**

The first clue from the definitions of IA and SB is the inclusion of energy expenditure (EE) using MET equivalencies. However, none of the four SRs addressed EE in relation to the PA undertaken in the reviewed studies. However, if there is going to be a concerted effort to address behaviors which are based on extremely low levels of energy expenditure, then an initial cue is that there needs to be a willingness to identify levels of EE when determining samples and examining PA behaviors or interventions. This would align with the SBRN consensus definitions, and that determination of energy expenditure for PA completed should be addressed and can involve using CEEY [5] which provides light, moderate, and vigorous expenditures for the included PA/sport, sedentary, transport, schoolwork, self-care, chores and other activities. A second cue derived from the clue of EE components of the IA and SB definitions relates to the daily recommendations of PA and is the importance accumulative time of EE. A third cue was that time was a unit for SB only for Baumann et al. [15] who used pedometers and exercise tracking watches to propose daily step counts for minimum healthy levels reaching 5000 as well as other PA and health parameters. These levels differed from those proposed by Tudor-Locke et al. [16] who recommended boys to average 12,000 to 16,000 steps/day and girls to average 10,000 to 13,000 steps/day with a steady decrease in steps/day to approximately 8000–9000 steps/day for adolescents aged 18. From step counts, an intervention cue for children and adolescents with SB and IA would be to walk more as this is a positive starting point for PA and accumulative effect of walking at 2.9 MET for longer per day or walking slightly faster at a moderate level 3.6 MET which exceeds the MET levels of being inactive. Although not examined here, a future analysis of the SR would be to determine how many studies considered moderate walking as an immediate and effective intervention to resolving inactivity.

A second conceptual clue derived from the SR is determining how much time is undertaken for IA and SB. The amounts of time for each identified intervention in the four SRs varied greatly between weeks [6], years [9], and related to specific behavior measures in minutes, days, and weeks [7, 8]. Currently, it is difficult to know how or what adjustments in the waking hours could occur with time increases in SB or IA. One cue that would be valuable in getting a better sense of time commitments would be to use of a 24-hour movement-based terminology pie-chart [4], which despite needing a catchier name, is an instrument to determine what children and adolescents complete in each behavioral area (see **Figure 1**). This could be developed into a computer-based [6] program or app allowing children and adolescents to log their time spent in each element with the adjustments adding and subtracting to the amounts of time awake and time asleep. This would be more inclusive in terms of time on tasks and allow interventions to be targeted at times when there is excessive IA or SB. Once developed and providing data, it would be possible to get the fullest picture of how and what active and sedentary behaviors are actually interacting with each other. The importance of such a composite behavioral measure is that everyone has levels of SB in their daily routine. Some of it is enforced such as working at a desk. Some of it is a choice and yet both have the same implications for health outcomes. Indeed, there may be many forms of behavior that make up the time we are sedentary, but this does not help when making recommendations for how to decrease SB. This is especially so if we cannot discern distinct types of SB apart from less than 1.5 METs as it really does not matter what behaviors we decide to do without. However, having waking hours data will allow for targeted interventions, and awareness of good and not-so-good choices of both PA, sedentary time, and time asleep. This has further potential to provide analysis of accumulative and/or pervasive multiday issues for school-week and week-end patterns, which are where the health risks of SB begin to be problematic.

#### **Figure 1.**

*24-hour movement and nonmovement behaviors pie-chart [4]. Note: The pie-chart organizes movements that take place throughout the day into two components: 1) The inner ring (darker colors) that represent the main behavior categories using energy expenditure. 2) The outer ring (Lighter colors) representing general posture categories. The proportion of space occupied by each behavior is not prescriptive of time that should be sent in these behaviors each day. Adapted from: Tremblay MS, Aubert S, Barnes JD, Saunders TJ, Carson V, Latimer-Cheung AE, et al. Sedentary Behavior Research Network (SBRN) – Terminology Consensus Project process and outcome. Int J Behav Nutr Phys Act. 2017; 14:75. https://doi.org/10.1186/s12966-017-0525-8. BioMed Central Ltd Publishers.*

*Changing Sedentary Behavior in Children and Adolescents: Understanding Research… DOI: http://dx.doi.org/10.5772/intechopen.114253*

#### **5. Methodological clues and cues**

In this section, the first methodological clue was that the four SR completed analyses that determined it was not possible to complete a meta-analysis due to levels of heterogeneity between the reviewed studies. The irony being that, in an analysis of IA and SB, there were often too many 'moving parts' being analyzed which restricted to ability to discern meaningfulness of what was occurring in the interventions, clusters, and associations. As a consequence, a cue would be to reduce the number of moving parts and focus on specific or targeted achievable behaviors that can form SB interventions. This should become easier as a plethora of outcome behaviors were ineffective in the four SRs and therefore can be avoided or more carefully considered in the future studies.

A second methodological clue derived from the four SRs was that, although the samples ranged from 5 to 21 years, there were reported differences between children and adolescents. The resulting cue is that there are sufficiently different results to propose that research studies focus on either children or adolescents or specific age groups or bands. This leads to an additional cue of examining the changes that occur in the transition from childhood to adolescence which generally aligns with the transition from elementary to middle/junior high school. A third and related cue would be to extend this examination to all the transitions from one educational level to the next including from high school to university and beyond where IA and SB are both likely to increase and PA is likely to be reduced or discontinued completely.

A third methodological clue is that a broad array of variables and indicators were related to IA and SB with moderate effects at best and in many cases low to ineffective effects. How much of this was due to the components of the method is always part of every discussion. Some commonalities in the method were that samples were small or included a number of different subsamples such as those with obesity or different PA levels. As a result, a cue for the effectiveness of interventions would be to require consideration of practical significance which is the size of the effect and is not confounded by the sample size. In studies with small sample size, this provides the size of the difference that can be reported as intervention effectiveness via ES and 95% CI rather than statistical significance which is inferred from the precision of the estimate [17]. Practical significance also requires examination of whether the confidence intervals include zero, especially in the case of small effect sizes. Another cue would be to use ES levels reported in this literature as more relevant in power analysis calculations and for practically significant comparisons rather than traditional ES interpretations [12].

A fourth methodological clue evident in all studies is that there is no evidence of the prevalence levels of children and adolescents who can be identified as sedentary and inactive. This is quite different from the prevalence of SB that has been established across a multitude of behaviors [13]. Part of this is due to many studies not discerning energy expenditure as identified in the SBRN definitions but may also be due to the definitions not including time as a mitigating factor. This clue led to an attempt to discern prevalence for those that are sedentary and inactive from the SBRN definitions using data from the 2021 CDC's biannual Youth Risk Behavior Survey (YRBS, [15]). From this survey, it was possible to determine a four-variable form of sedentariness (see **Table 2**). Despite different units of time being used, all four questions had a zero-time answer option. From the sample of 15,997 adolescents who completed the survey, 860 (5.41%) adolescents reported the zero-time option


**Table 2.**

*Prevalence rates of sedentariness and inactivity in adolescents who completed the 2021 YRBS survey.*

for all four questions and were considered as sedentary under the SBRN definition. Additionally, there were 370 (2.31%) adolescents who reported three zero-time behaviors and one with the lowest time level on the remaining question. These lowest reported levels are above zero and would suggest from the SBRN definition that these adolescents are inactive. Interestingly, there were different prevalence levels of inactivity based on the question that provided a positive physical activity level in **Table 2**. However, there were caveats. The first is that the data for the MVPA question does not account for adolescents who may be active for less than 60 minutes per day and could well complete enough PA in minutes to be considered inactive or low active; unfortunately, the composition of this latter group is no longer determined by the CDC in the YRBS survey. A second caveat was that the prevalence of these populations based on a large sample is rarely reported and is unique. This clue leads to suggested cues of having complex or multibehavior determinants of sedentariness and inactivity, and an understanding that the prevalence of SB and IA is relatively low in this large population and could set a benchmark for prevalence rates in future research studies.

### **6. Clues and cues related to interventions**

A clue from all four SRs was that interventions were portrayed differently. Two focused on direct interventions and included Bauman et al.'s novel approach of mobile smartphone programs using behavior change techniques as the intervention [6] and Kuzik et al.'s interventions which were portrayed as exposures [8]. In contrast to these direct approaches to the intervention, the other two SRs indirectly involved interventions, as de Mello et al. [7] only involved the clustering of IA and SB with a range of modifiable correlates to better inform intervention strategies for behavior change, while Whilhite et al. [9] made no mention of interventions other than to recommend including sleep in more longitudinal and intervention research! From this clue, a cue is to determine what constitutes the intervention of IA and SB, be it

## *Changing Sedentary Behavior in Children and Adolescents: Understanding Research… DOI: http://dx.doi.org/10.5772/intechopen.114253*

from a therapeutic, behavior modification, or educational perspective and how it impacts behavioral change. Given that when the participants engaged in some form of PA intervention, the procedures are as important as the outcomes. To frame this cue, I highlight three approaches to program evaluation [14, 18, 19] from a variety of paradigmatic approaches that can function as cues in their own right when considering the importance of the intervention in impacting the program or study outcomes.


distinct levels of authority and involvement from parents to policymakers that accompany intervention programs as a result, there are different or even conflicting levels of interest and advocacy between these stakeholders. This means that recommendations may have to differ depending on the specific audience. Making such recommendations was part of both Kuzik et al. and Whilhite et al.'s conclusions [8, 9] and involved different stakeholders and requires awareness of evaluative processes, such as of social validity, and how the fidelity of the program's procedures impacted the achievement of outcomes of the program goals as this is required for continued (financial, resource, workforce, or volunteer) support for the intervention.

An original premise of this review of reviews was that successful and also less successful interventions leave clues in both methods and interventions. To summarize, some of these clues can be traced back to similar methodological issues that prompted the development of the SBRN [4] consensus statements and the operational definitions of IA and SB. However, despite the four SRs covering a plethora of outcomes and being well-conducted analyses, the impact on SB and IA was rather underwhelming. However, this is not discouraging as these SRs still delivered important clues regardless of the outcome as any results are important in developing future decisions about which method and intervention cues are worth continuation.
