Longitudinal associations between perceptions of the neighbourhood environment and physical activity in adolescents: evidence from the Olympic Regeneration in East London (ORiEL) study

Table 8 Associations of cumulative perceptions of the neighbourhood environment and trajectories of perceptions with outdoor physical activity (n = 2260)

	Unadjusted			Adjusted			Gender-interaction^c
	OR	95%CI	p-value	OR	95%CI	p-value	p-value
Cumulative perception^a
Perceived bus stop proximity	0.95	[0.86,1.06]	0.377	0.94	[0.83,1.07]	0.331	0.900
Perceived traffic safety	1.03	[0.97,1.09]	0.332	1.00	[0.93,1.08]	0.988	0.535
Perceived street connectivity	0.99	[0.93,1.06]	0.873	1.03	[0.96,1.11]	0.398	0.816
Enjoyment of neighbourhood for walking/cycling	1.03	[0.99,1.07]	0.165	1.01	[0.95,1.08]	0.730	0.510
Feeling safe (personal safety)	1.05	[1.01,1.08]	0.006	1.02	[0.98,1.07]	0.278	0.825
Trajectory of perception^b
Perceived bus stop proximity	1.11	[0.95,1.29]	0.192	1.11	[0.93,1.32]	0.248	0.095
Perceived traffic safety	0.94	[0.86,1.04]	0.230	0.94	[0.84,1.04]	0.222	0.828
Perceived street connectivity	1.08	[0.98,1.18]	0.112	1.07	[0.97,1.19]	0.172	0.091
Enjoyment of neighbourhood for walking/cycling	1.02	[0.94,1.10]	0.622	1.02	[0.93,1.11]	0.737	0.527
Feeling safe (personal safety)	1.01	[0.9501.14]	0.640	1.01	[0.96,1.07]	0.734	0.956

^a Results are from logistic regression models estimated with Generalised Estimating Equations to account for the clustering of individuals within schools (exchangeable working correlation matrix). The cumulative exposure are continuous variables constructed as the sum of scores of each exposure over the 3 waves. A higher score indicates a perception of supportive environment for the specific exposure. Adjusted models adjust for gender, ethnicity, health condition (at wave 3), family affluence (at wave 3), free school meal status at wave 1 and the other perception variables
^b Results are from logistic regression models estimated with Generalised Estimating Equations to account for the dependency across repeated measurements (unstructured working correlation matrix). Each exposure variable measures change since wave 1 on a continuous scale. Each unit represents an average change in exposure by one category between the baseline and the end of the study (+ 1 = improvement of the neighbourhood by one category on average). The coefficients represent the time*trajectory interaction, which assesses whether exposure trajectory is associated with different trajectory of change in the outcome. Adjusted models adjust for time, gender, ethnicity, health condition, family affluence and free school meal status at wave 1, the other perception variables, and their time*trajectory interactions
^c The adjusted models were replicated with the addition of interaction terms with gender

ISSN: 1471-2458