Physical Activity

Physical activity and sedentary behaviour

This stream comprises experienced and emerging researchers with a shared interest in many aspects of physical activity and sedentary behaviours among young people, including: measurement issues; determinants; epidemiology; secular trends; interventions; and associations with child and adult health status.
Our group is committed to the advancement of research in this field through:

improving access to relevant information for fellow researchers and public health professionals
developing networks of current Australian and New Zealander researchers in this field, and
providing opportunities for emerging researchers to share their work and plan their future research directions.

Physical activity and sedentary behaviour assessment

The purpose of this website is to provide visitors with a simple user’s guide for selecting instruments to assess physical activity and sedentary behaviour.
The overall goal is to highlight the main decisions needed to be made based on a series of important issues such as the purpose of assessment; age of participants; number of participants or sample size; tolerance limits for respondent burden; resources available; and the type of information sought.

Physical Activity and Sedentary Behaviour Assessment in Childhood
Medical Research Council (MRC-UK) Toolkit

Physical Activity and Sedentary Behaviour Assessment in Childhood

Accurate assessment of physical activity and sedentary behaviour among young people is important for many reasons

detecting positive and negative health outcomes associated with these behaviours;
estimating population prevalence and trends;
identifying correlates;
detecting natural changes over time; and
evaluating the efficacy of interventions to alter physical activity and sedentary behaviour.

However, choosing the most appropriate instrument depends not only on the validity and reliability of the measure, but also on;

what the instrument is measuring,
how it manages to meet the intended purpose of the assessment,
what resources are needed and
the population group of interest.

A common question facing researchers, practitioners, teachers and policy makers wanting to assess physical activity and sedentary behaviour in children or adolescents is ‘what instrument should I use?’ However, the answer depends on a range of factors including the purpose of the assessment, target group, research question, resources available, time frame and the context.
We would like to start by introducing a series of case scenarios that have been prepared to illustrate how decisions might be made according to the needs of the user and the resources available.
A simple method selection guide has been constructed to help identify potential instruments to measure physical activity and sedentary behaviour in children and adolescents.

Medical Research Council (MRC-UK) Toolkit

We would like to acknowledge the extensive work undertaken by the Population Health Sciences Research Network (PHSRN), the Medical Research Council (MRC) in the United Kingdom, using a similar approach, to facilitate researchers in selecting appropriate dietary intake and physical activity methodologies for populations.
The ACAORN project, established in 2008, aimed to:

Support researchers in making decisions about appropriate physical activity methodology and
Provide access to physical activity assessment tools for children and adolescents

This project was undertaken in the context of child obesity, research and practice.
During the project ACAORN became aware of the PHSRN project being undertaken by the MRC. ACAORN consulted with researchers based at the MRC and would particularly like to acknowledge Dr Janet Warren.
In the spirit of collaboration, and not wanting to re-invent the wheel, it was acknowledged that The Diet and Physical Activity Measurement Matrices and associated web pages in the MRC toolkit would be a valuable resource for Australian researchers.
This ACAORN site also provides an iterative tool which guides researchers into selecting appropriate methods to assess physical activity and sedentary behaviour in children and adolescents. A practical considerations matrix has also been built with a specific focus in the context of obesity research in Australia.
To expand the level of information on the relevant ACAORN Physical Activity and Sedentary Behaviour web pages, please visit the MRC web pages here.

Physical activity and sedentary behaviour assessment - Case scenarios

The case scenarios provided have been selected to represent a range of circumstances.

These scenarios should be read in conjunction with the relevant method selection guide for assessing Physical Activity or Sedentary Behaviour. Doing so will provide more detail to help guide the most appropriate choice of instrument(s).

Physical activity assessment - case scenarios

Physical activity intervention among preschoolers

Prevalence survey among adolescents
Treatment program for overweight and obese adolescents
Trial of a school-based obesity prevention program
A clinical application

Physical activity intervention among preschoolers

Researchers plan to implement and evaluate the potential efficacy of a 12-week pilot program to promote physical activity in a pre-school setting (3–5-years old). They have recruited two pre-schools and plan to randomise them to either intervention or control group. Physical activity will be assessed among approximately 50 children (25 in each group) at baseline and 6 months. The researchers would like to determine if the intervention program results in greater participation in physical activity during the pre-school day and if a greater proportion of students are active during structured physical activity sessions.

In this scenario, the researchers are assessing changes in physical activity within a relatively small group of participants over time using a cluster-randomised trial. Key outcome variables of the evaluation are the duration and intensity of physical activity participation during the pre-school day and the proportion of children who are “active” during structured physical activity sessions. The number of days that should be assessed to capture habitual activity patterns in this age group and in this setting is unknown, and will depend on day-to-day variations in the pre-school program.

Given that the participants are too young to self-report, two approaches might be used in tandem: direct observation and accelerometry. Direct observation can be used to assess the number of children who are active in a particular area (in this case, an indoor or outdoor play space) during a specific period of time. As such, it could be used to assess the proportion of children who are active during a structured physical activity session.

A strength of this measure is that it can be used with little burden on the children. As it would be too onerous to use direct observation across the entire pre-school day, accelerometers could be used since the sample size is sufficiently small to bear the relatively high cost of these instruments. As data are captured in ‘real time’, it can also provide information on the activity levels of children during the structured physical activity sessions to obtain a more complete picture of these bouts of physical activity.

Prevalence survey among adolescents

There is concern in the community about decreasing levels of physical activity among adolescents. A sample of >1000 female and male students from different year levels and a variety of schools, educational sectors and geographic locations will be recruited. Physical activity will be assessed at one point in time (point prevalence design). Conducting a needs assessment prior to the development of physical activity promotion programs is essential for identifying appropriate behavioural targets that meet the most important needs of the community and are likely to make the greatest impact.

In this scenario, it would be important to establish the proportion of the population that meets young people’s physical activity guidelines and to provide a benchmark against which the effectiveness of physical activity programs in the community can be compared. It would also be important to determine the types or domains of physical activities among young people to establish target behaviours (e.g., active trans-port, organised sport, school-based physical activity).

The adolescent age group (12–18-years old) is of interest in this study and therefore self-report is an option. Given the need to describe physical activity quite broadly, and the likelihood of requiring a relatively large sample size, simultaneous cost-effective subjective and objective methods would be favoured: self-report (most likely self-administered in this large sample) and if the budget allows, pedometers.

Self-report physical activity recall instruments can assess compliance with guidelines and the types or domains of physical activity in which the respondents engage. Pedometers measure ‘steps’ and therefore can be used to establish a benchmark for how much walking is performed. While the pedometer is unable to represent other dimensions of physical activity (i.e., intensity or contextual information about where walking took place) the objective information on walking complements the self-report since incidental walking is particularly difficult to recall and therefore poorly captured by self-report.

Treatment program for overweight and obese adolescents

A researcher has received funding to test the feasibility and acceptability of a 26-week obesity treatment program among overweight and obese adolescents (aged 13–18 years) in a community setting. He/she plans to recruit 12 participants to the feasibility trial (as a single group) and assess their physical activity at pre- and post-intervention. The researcher would like to determine the potential efficacy of the program on the amount of time spent in MVPA outside of school hours.

In this scenario, the researcher is interested in examining changes in habitual physical activity over time. It is important to capture, as accurately as possible, the amount of time spent in specified intensities of physical activity (moderate and vigorous). While the sample size is relatively small, direct observation is not feasible since the time period of interest is outside of school hours. The errors associated with self-report (and factors such as social desirability bias) are likely to limit the ability to detect changed behaviour with sufficient resolution.

Therefore, an objective measure is recommended, and accelerometry provides the advantage of real-time data acquisition over pedometers. This allows periods outside of school hours to be partitioned from other time periods. Researchers can use cut-points developed for this age group to provide information about the intensity of the physical activity.

Trial of a school-based obesity prevention program

A research team has developed a 20-week school-based obesity prevention program that they would like to implement and evaluate with 6–9-year-old children. They have recruited 10 schools to the study and plan to randomize them to intervention (n = 300 children) or control (n = children) groups and assess children’s physical activity pre and post-intervention. They would like to determine if there has been an increase in school-based physical activity during physical education, recess, and lunch, and in the duration of different types of physical activity out of school hours and on weekends.

In this scenario, the researchers are interested in changes in physical activity over a 20-week period. The relatively large sample comprises children who are too young to self-report, and the study requires details of levels and types of physical activity. This limits the choices to cost-effective objective methods and proxy-reported subjective methods.

For activities performed at school, teacher proxy-reports would be preferred, although teachers’ assessments are likely to be more valid for structured classroom activity than free play at recess and lunch. For free play at recess and lunch, direct observation of participation in a physical activity setting (such as an oval or quadrangle) could be undertaken by trained assessors. For an overall assessment of school-based physical activity, pedometers could be worn from the beginning to the end of the school day, with readings taken at the beginning and end of recess and lunch breaks to isolate these periods.

It would be recommended that multiple assessment days be used to minimise the inflation in the intra-class correlation coefficient arising from assessments being conducted in a school on the same day. A parent proxy-report could be used to assess physical activity types and durations out of school hours.

A clinical application

A clinician is interested in determining the physical activity status of a 14-year-old child with Type-2 Diabetes prior to advising or prescribing physical activity / exercise. A number of potential approaches could be applied in this clinical setting, including: self-report questionnaire or diary/log; pedometer; HR monitor; or a combination of methods. In this example, the importance of using a multi-method approach should not be underestimated.

A rudimentary assessment of physical activity level (including the young person’s preferences for physical activities), may be performed using a standardised self-report questionnaire screening survey or diary, and a pedometer as a component of initial health screening. A pedometer may be used as a screening tool and as a source of motivation (self-monitoring) for the child to adopt a more physically active lifestyle.

A major advantage of establishing an objective measure of an individual’s baseline activity is the ability to provide a quantifiable and tailored dose of physical activity or exercise prescription. The starting point in the use of a pedometer should be to determine the number of steps an individual takes to cover a defined distance (for example, 1 km) and then a baseline number of steps over a minimum of 1 week (including weekend and weekdays). Consequently, the clinician is able to recommend a realistic number of ‘steps per day’, working towards the recommended number of steps for young people or if this is already met, a realistic number beyond.

Sedentary behaviour assessment - case scenarios

Screen time intervention among pre-schoolers

School-based RCT to reduce sitting during school class-time
Treatment program for overweight/obese primary school children
Primary prevention of adolescent screen time in clinical settings
Population prevalence of screen time among adolescents

Screen time intervention among preschoolers

Researchers plan to design and evaluate the potential efficacy of a 12-week intervention to reduce screen time in children aged 2-3 years in the home setting. Children will be randomised as a family unit to either an intervention or control group. Screen time will be assessed among approximately 80 children (40 in each group) at baseline and at 3-month post-intervention time points. The aims of this research are to (i) determine if the intervention results in reduced total screen time at home and, (ii) determine duration by each screen type.

In this scenario, the researchers are assessing changes in children’s screen time using a randomised controlled trial (RCT) design. Key outcomes of interest are the duration and context of the screen time participation at home. The sample comprises children who are too young to self-report, and the study requires details of types of screen time behaviour. Objective measurement using inclinometers or accelerometers are not appropriate as these methods will not provide the context of the behaviour or whether the sedentary time was screen time or other sedentary behaviours. An appropriate approach is a parent proxy questionnaire using recall over a certain number of days.

The number of days that represent an accurate estimate of habitual screen time behaviours in this age group and in this setting is unknown, and will depend on day-to-day variations in the home context. Ideally, weekday and weekend day screen time should be captured. However, despite recognised limitations, parent proxy self-report recall instruments can be used to provide an estimate of minutes spent in screen time and assess compliance with guidelines and determine the types of screen time in which children engage.

School-based RCT to reduce sitting during school class-time

Traditional classroom teaching techniques predominantly involve children being seated for sustained periods. Evidence among adults suggests that sustained sitting may be detrimental to health and that interrupting sitting time may reduce such risks.32 Researchers have planned a 6-month RCT to test strategies to reduce classroom sitting in primary school children through alternative teaching practices. The RCT involves two classes within each year level at six schools (three control and three intervention schools), and approximately 300 students.

To determine the effectiveness of the intervention, researchers must detect changes in time spent ‘sitting’ during class time between baseline and post-intervention. Key considerations in the selection of appropriate measurement instruments include: the age of the participants, the need to detect behaviour within particular periods, the need to differentiate sitting from other postures, potential burden to participants and minimisation of class disruption. In this scenario, self-report measures may be inappropriate due to participant age (cognitive limitations). Proxy-reports by teachers may also be inappropriate as they would only provide group-level information rather than data about individuals. The use of self-report logs/diaries may be burdensome during class-time, may disrupt class activities and could result in reactivity. Accelerometry may also be inappropriate because they quantify ‘movement’ and do not provide any postural information. Low movement counts are indicative of limited movement and not necessarily ‘sitting’.

In this scenario, inclinometers and direct observation may be the most appropriate measurement tools. Inclinometers enable researchers to determine time spent sitting or lying (based on postural information), can be worn over long periods (e.g., one week) and allow researchers to extract data from specific periods of interest (e.g., class times). Direct observation would provide the same ‘information’ with the additional opportunity to value-add by documenting specific behaviours undertaken in each posture (e.g., reading, writing, art, craft, etc). However, direct observation may result in reactivity as participants know they are being observed, and may be costly given the number of observations that would be required (each period across the school day for each class at each school) on multiple school days. Further, the number of days of observations required is not established.

Treatment program for overweight/obese primary school children

A researcher is seeking to evaluate the effects of a 10-week family-centred sedentary behaviour reduction intervention on adiposity in overweight/obese 8- to 12-year-old children. The feasibility study is a single-arm experiment involving 30 overweight/obese children with assessments of sedentary behaviour taken pre- and post-intervention, and the researcher wants to determine if the intervention reduced children’s sedentary time (i) overall daily and (ii) outside of school hours.

For this intervention, the researcher needs to selects an instrument that is both accurate and objective, and sensitive enough to detect the hypothesised changes in sedentary time. Although the sample size is relatively small, direct observation would not be feasible because it is likely that the children attend different schools, and because the researcher is also interested in understanding the effects of the intervention on sedentary time outside of school hours. Self-report questionnaires offer a cost-effective option, but the assessments would be vulnerable to recall-bias because of the age of the participants. Parent-proxy reports would also not be recommended because their estimates might be influenced by social desirability bias and this could result in under-reporting of the behaviour, or parents’ understanding of the desired effects of the intervention might result in under-reporting at post-test. It would also be difficult to accurately assess children’s total sedentary time, which occurs in many settings and contexts and not always in the presence of parents.

An objective measure is recommended and monitoring devices worn on the body, such as inclinometers or accelerometers, would be most suitable. The use of an inclinometer would allow the researcher to examine time spent in different postures, and from this changes in sitting/lying time as a result of the intervention could be evaluated. If accelerometers are chosen, the researcher can apply age-appropriate cut-points to determine sedentary time. The real-time data acquisition from objective monitoring devices would allow the researcher to specifically examine sedentary time that occurs outside of school hours, in addition to children’s overall or total sedentary time per day.

Primary prevention of adolescent screen time in clinical settings

A general practitioner (GP) is concerned about the metabolic profile of an obese adolescent patient presenting markedly overweight and with obvious signs of insulin resistance. During the consultation the GP ascertains from the adolescents’ parents and the adolescent that the adolescent spends most of their time sitting on the couch playing e-games, watching TV/DVDs.

Access to adolescent obesity management clinics is limited, and because the GP has a small, busy practice is unable to provide on-going long consultations to the adolescent. The GP decides that the best management strategy will be based on regular brief counselling consultations that incorporate goal setting. The adolescent’s parents are asked to help the adolescent set realistic screen time reduction goals and to help monitor progress towards reducing screen time.
Objective measures are not suitable for several reasons. Firstly, motion sensors do not capture contextual information and, the cost of motion sensors is prohibitive to the practice budget. Furthermore, the GP does not have the time and expertise to interpret the data collected by objective instruments. Rather, the most feasible line of intervention is for the GP to ask the adolescent to complete a time use diary, or suitable structured questionnaire, about their screen time.

This baseline information will identify the duration spent on screen time and the time of day spent on screen time. The GP can use this information to help the adolescent set realistic goals aimed at reducing screen time. The GP can monitor the adolescent’s progress towards reducing screen time at on-going consultations for the monitoring the obese adolescents’ progress.

Population prevalence of screen time among adolescents

Health and education professionals have concerns about non-school recreational screen time among adolescents. In order to determine whether investments should be allocated to develop a school-based intervention to encourage adolescents to reduce their screen time, the first step is to ascertain how prevalent screen time is, and whether there are sociodemographic differences in teenagers’ screen time. To determine the population prevalence, a large sample of adolescents (i.e., several hundred) from a range of high school years, across different educational sectors, and geographical and socioeconomic areas is required to determine population estimates which are generalisable.

In this scenario, objective measures such as accelerometers and inclinometers are inappropriate for several reasons. First and foremost, objective measures do not provide contextual information, so will not discriminate between screen time activities, or other sitting behaviours, therefore self-report is the most desirable method of measurement. Unlike younger children, adolescents are capable of self-report, albeit recall can be affected by social desirability, and estimates of time are subject to large error. Questionnaires have the ability to discriminate between screen time activities and to determine habitual non-school screen time on week and weekend days. In school environments, questionnaires can be administered either as pen and paper, or via computers/smart boards.

A significant issue to consider when asking students to report screen time activities is the concept of multi-tasking. For example, an adolescent may play on their computer while watching TV – so during the administration of the questionnaire it is important to instruct respondents to allocate the time proportionally spent on each screen activity. An alternative method to measure screen time behaviour is with time use diaries/log, or ecological momentary assessment (EMA), where respondents report activities undertaken during a specified time interval. A limitation of this method is that all activities are reported, generating large volumes of data from which screen time data are extracted.

Physical activity and sedentary behaviour assessment - Tools and validation studies

The Australasian Child & Adolescent Obesity Research Network (ACAORN) is interested in guiding Australian and international researchers in the process of selecting appropriate measurement tools. We are planning to include a dynamic repository of tools used for measuring physical activity and sedentary behaviour in child and adolescent populations.

The Adolescent Physical Activity Recall Questionnaire (APARQ)
The Adolescent Sedentary Activity Questionnaire (ASAQ)
The Children's Leisure Activities Survey (CLASS)
The hitch-hikers guide to assessing physical activity

Questionnaires to measure Physical Activity and Sedentary Behaviour

The Adolescent Physical Activity Recall Questionnaire (APARQ)

Below is a compendium of resources that will guide you in correctly administering the Adolescent Physical Activity Recall Questionnaire (APARQ)

The Adolescent Sedentary Activity Questionnaire (ASAQ)

Below is a compendium of resources that will guide you in correctly administering the Adolescent Sedentary Activity Questionnaire (ASAQ)

The Children's Leisure Activities Survey (CLASS)

Below is a compendium of resources that will guide you in correctly administering The Children's Leisure Activities Survey (CLASS)

The hitch-hikers guide to assessing physical activity

We have developed a guide to assist researchers in the process of choosing a measurement tool for the assessment of physical activity in school-aged children, giving several scenarios. Practical tips as to how to trouble-shoot questionnaires and analyse accelerometry data are also given. Read Abstract here

Physical activity and sedentary behaviour - Stream publications

Barnett LM, Hardy LL, Lubans DR, Cliff DP, Okely AD, Hills AP, Morgan PJ. 2013. Australian children lack the basic movement skills to be active and healthy. Health Promotion Journal of Australia Abstract
Hardy LL, Hills AP, Timperio A, Cliff D, Lubans D, Morgan PJ, Taylor BJ & Brown H. 2012. A hitchhiker's guide to assessing sedentary behaviour among young people: Deciding what method to use. Journal of Science and Medicine in Sport. Abstract
Lubans DR, Hesketh K, Cliff DP, Barnett LM, Salmon J, Dollman J, Morgan PJ, Hills AP & Hardy LL. 2011. A systematic review of the validity and reliability of sedentary behaviour measures used with children and adolescents. Obesity Reviews. Abstract
Dollman J, Okely AD, Hardy L, Timperio A, Salmon Jo, Hills AP. A hitchhiker's guide to assessing young people's physical activity: Deciding what method to use. Journal of Science and Medicine in Sport 2009; 12(5):518-525. Abstract.
Hardy LL, Booth ML, Okely AD. The reliability of the Adolescent Sedentary Activity Questionnaire (ASAQ). Prev med. 2007 Jul;45(1):71-4. Abstract.
Booth ML, Okely AD, Chey TN, Bauman A. The reliability and validity of the Adolescent Physical Activity Recall Questionnaire. Med Sci Sports Excerc. 2002 Dec;34(12):1986-95. Abstract.