Table 3.

Effects on self-reported purchase of soft drinks and sweets or cakes over the last week: proportions per group at pre-test and post-test (1= one or more times per week; 0= never); cluster-randomized controlled intervention providing lessons on healthy food and changes to school vending machines in secondary schools in the Netherlands. Results of multilevel logistic regression analyses are shown per outcome

	Intervention group			Control group
		Pre	Post		Pre	Post	Intervention v. control
	n †	%	%	n †	%	%	OR‡	95 % CI
Soft drinks
School vending machine	248	25	19	255	25	20	1·01	0·51, 1·98
School canteen or tuck shop	248	19	18	259	20	17	1·10	0·57, 2·11
Shop in the school surroundings	248	37	34	259	38	36	1·00	0·58, 1·74
Brought from home	252	47	43	259	52	56	0·55*	0·34, 0·88
Extra foods
School vending machine	253	29	20	262	33	20	1·02	0·54, 1·93
School canteen or tuck shop	253	34	25	257	48	30	0·86	0·45, 1·63
Shop in the school surroundings	258	52	45	265	51	49	0·89	0·53, 1·51
Brought from home	253	57	52	262	58	52	0·95	0·64, 1·41

*P<0·05.

^†Sample size per group of participants without missing data on post-test and covariates.

^‡Odds ratio of the group effect adjusted for the effects of pre-test, age, gender, ethnicity, education and BMI.