Table 3.

Associations between previous cycle training and children’s cycling behaviour across the study population (N = 3336)

Outcome	Whether child had done cycle training	Percentage (95 % CI)	Unadjusted analysis (risk ratio, 95 % CI)	Adjusted analysis (risk ratio, 95 % CI)
Child cycles at least once a week	Untrained	41.5 (38.9, 44.2)	1	1
	Trained	55.0 (52.8, 57.2)	1.27 (1.17, 1.39)	1.26 (1.16, 1.37)
Child ever cycles	Untrained	73.0 (70.6, 75.4)	1	1
	Trained	92.7 (91.4, 93.8)	1.23 (1.18, 1.28)	1.20 (1.15, 1.25)
Child usually travels to school by bike	Untrained	1.9 (1.2, 2.8)	1	1
	Trained	3.4 (2.7, 4.3)	1.62 (0.99, 2.67)	1.38 (0.83, 2.29)
Child makes local bike trips independentlya	Untrained	43.3 (40.6, 45.9)	1	1
	Trained	56.7 (54.5, 58.9)	1.23 (1.13, 1.34)	1.21 (1.11, 1.32)

All p < 0.001 for association. Analyses based on our study population of 3336 children, of whom 1378 were untrained, 1956 trained and 2 had missing data (imputed using multiple imputation). Adjusted analyses adjusted for all variables shown in Table 1 (with the local prevalence of cycling to work entered as a continuous variable), and also for the region of England that the child lived in and the season of data collection

CI confidence interval

^a Defined as ever making local, non-school bicycle trips without an adult, either on their own or with other children