Table 1.

Descriptive characteristics of potential predictors of change in physical activity

Factor	Description	Device	Mean (SD) or %
Biological and socio-demographic
Age (years)	Child-reported	Child Q	9.8 (0.4)
Sex (% male)	Child-reported	Child Q	41.5%
BMI z-score	Height was measured to the nearest millimetre (Leicester height measure, Chasmors Ltd., Leicester, UK). A non-segmental bio-impedance scale was used to measure weight (to the nearest 0.1 kilogram) and impedance in light clothing (Tanita, type TBF-300A. Tokyo, Japan). Height and weight were used to calculate body mass index (BMI, kg/m2). British reference data was used to calculate z scores [31]	Measured	0.3 (1.1)
Parental education (% >16 years)	Parents self-reported their age at leaving full time education (reported as 3 categories and dichotomized into ≤16 years, and >16 years of age)	Parent Q	52.0%
Index of Multiple Deprivation	Computed using parent-reported home postcode. This measure combines information on deprivation from seven domains: income, employment, health, education, barriers to housing and services, living environment and crime. The derivation of this measure is described in detail elsewhere [32] and quartiles were derived for use in this analysis.	GIS	15.4 (9.5)
Psychological factors
PA preference (% with preference for PA)	Sum of 4 questions asking: I would prefer to: play indoors or outdoors; walk or watch TV, run or walk; watch TV or run. Active and inactive options coded 1 and 0 respectively. Dichotomised with ≤2 as preference for inactivity; ≥3 as preference for PA recoded as 0 (reference category) and 1 respectively. [33]	Child Q	69.3%
Self efficacy in overcoming barriers (% high)	Child answered Yes (1) or No (0) to: I can do something active even: if it is hot or cold outside; if I have a lot of homework; no matter how tired I feel. Dichotomised as low self-efficacy if score ≤2, high self-efficacy if scoring 3; recoded as 0 (reference category) and 1 respectively. [34]	Child Q	42.7%
Self-efficacy in support seeking (% high)	Child answered Yes (1) or No (0) to: I can ask my parent to: sign me up for PA; my parent to do PA with me; my best friend to do something active with me. Dichotomised as low self-efficacy if score ≤2, high self-efficacy if scoring 3; recoded as 0 (reference category) and 1 respectively. [34]	Child Q	77.8%
Personal barriers (% with barriers)	Child answered Yes (1) or No (0) to: Are you ever stopped from doing PA because: there you want to watch TV; you don’t think you’re good at PA; you don’t like PA; and you might get hurt. Responses were summed and dichotomised to 0 (reference category) reporting no barriers and those reporting ≥1 barriers as having personal barriers (1). [35,36]	Child Q	52.8%
Lack of PA equipment	Child answered Yes (1) or No (0) to: Are you ever stopped from doing PA because you don’t have the equipment you need? [35,36]	Child Q	25.0%
Socio-cultural factors
Peer support (% reporting peer support)	Child answered Yes (1) or No (0) to: During a typical week, do the following things happen: my friends do PA with me; I ask friends to do PA with me; My friends ask me to do PA with them. Responses were summed and then recoded as 0 no peer support and ≤1 as peer support. [35]	Child Q	69.5%
Parental encouragement	Child answered 1 (Never), 2 (once/twice/week), 3 (nearly daily), 4 (everyday) to: During a normal week, someone in my family: encourages me to do PA, tells me I am doing well at PA, tells me PA is good for my health. Responses were summed range 3–12. [37]	Child Q	8.3 (2.4)
Parental logistic support	Child answered 1 (Never), 2 (once/twice/week), 3 (nearly daily), 4 (everyday) to: During a normal week, someone in my family: does PA with me, takes me somewhere to do PA, watches me do PA. Responses were summed range 3–12. [18,37]	Child Q	6.5 (2.0)
Sedentary restrictions	Parents reported how often they restrict their child watching TV, playing computer games and using the computer with response categories N/A (0), never (0), rarely (1), sometimes (1), often (3) and very often (4). Responses were summed with range 0–12. [36]	Parent Q	6.3 (2.7)
PA restrictions	Parents reported how often they restrict their child playing outside or walking/cycling to a friends house with response categories N/A (0), never (0), rarely (1), sometimes (1), often (3) and very often (4). Responses were summed with range 0–8. [36]	Parent Q	1.7 (1.6)
Family cohesiveness (times/week)	Parents reported the number of times/week they do the following activities together as a family: eating meals, reading, sport, visiting family/friends, going to the park, swimming, cycling, watching TV, cooking with response categories as 0, 1–4 and >4 recoded as 0, 2.5 and 4.5 respectively and then summed with range 5–40.5. [36]	Parent Q	20.5 (5.1)
Freedom to play (% allowed)	Parents reported how often they allow their child to play outside anywhere within the neighbourhood with response categories never (0), rarely (1), sometimes (1), often (3) and very often (4). Dichotomised as ‘Not allowed’ (never & rarely) and ‘Allowed’ (sometimes, often and very often).	Parent Q	44.5%
Games console at home (% yes)	Children reported whether they had a games console at home with responses as Yes (1) and No (0) [35]	Child Q	88.5%
Electronic equipment in the bedroom (% yes)	Children reported whether they had a TV, PC or games console in their bedroom with responses as Yes (1) and No (0) which were summed (range 0–3) and dichotomised into (0 reference category) no media in bedroom and (1) media in bedroom. [35]	Child Q	75.0%
Behavioural factors
Travel mode to school (% active)	Children reported how they usually travel to school. Car and Bus/train were recoded as 0 (passive) and walk and cycle as 1 (active). [38]	Child Q	49.8%
Environmental factors
Living in a cul-de-sac (% yes)	GIS was used to determine whether the home address was a cul-de-sac, coded as 0 no (reference category) and 1 yes.	GIS	31.1%
Effective walkable area	Measure of neighbourhood connectivity. Calculated by dividing the total neighbourhood area (the area reached via the street network within 800 m from the home) by the potential walkable area (the area generated using a circular buffer with a radius of 800 m from the home). Higher values indicate higher walkability [38] (GIS-derived)	GIS	0.42 (0.16)
Woodland percentage (log%)	The percentage of participant’s neighbourhood covered by woodland or green space. Logged due to skewness.	GIS	1.26 (0.95)
Herfindahl-Hirschmann Index (HHI)	Land use mix in the participant’s neighbourhood, calculated as ∑(landusepercentage)2. A high HHI indicates low variation in land cover (presented higher and lower split by median) [38].	GIS
- Higher land use mix			46.1%
- Lower land use mix			53.9%
Distance to green space (log km)	Distances (in km) to nearest green space (excluding parks), calculated as the shortest route from home address via the street network, assuming use of the nearest entrance. Logged due to skewness.	GIS	7.2 (1.0)
Distance to sport venue (log km)	Distances (in km) to nearest sport venue, calculated as the shortest route from home address via the street network, assuming use of the nearest entrance. Logged due to skewness.	GIS	6.3 (0.9)
Distance to school (log km)	Distances (in km) to school, calculated as the shortest route from home address via the street network, assuming use of the nearest entrance. Logged due to skewness.	GIS	7.2 (1.2)
Availability of parks (% yes)	Child reported perceived availability of parks. Children answered yes (1) or no (0) to: There are playgrounds, parks, or sports halls close to my home that I can use. [35]	Child Q	79.1%