Evaluation of a long day care intervention targeting the mealtime environment and curriculum to increase children's vegetable intake: a cluster randomised controlled trial using the multiphase optimisation strategy framework

Objective: To determine the reach, adoption, implementation and effectiveness of an intervention to increase children's vegetable intake in long day care (LDC). Design: A 12-week pragmatic cluster randomised controlled trial, informed by the multiphase optimisation strategy (MOST), targeting the mealtime environment and curriculum. Children's vegetable intake and variety was measured at follow-up using a modified Short Food Survey for early childhood education and care and analysed using a two-part mixed model for non-vegetable and vegetable consumers. Outcome measures were based on the RE-AIM framework. Setting: Australian LDC centres. Participants: Thirty-nine centres, 120 educators and 719 children at follow-up. Results: There was no difference between intervention and waitlist control groups in the likelihood of consuming any vegetables when compared with non-vegetable consumers for intake (OR = 0<middle dot>70, (95 % CI 0<middle dot>34-1<middle dot>43), P = 0<middle dot>32) or variety (OR = 0<middle dot>73 (95 % CI 0<middle dot>40-1<middle dot>32), P = 0<middle dot>29). Among vegetable consumers (n 652), there was no difference between groups in vegetable variety (exp(b): 1<middle dot>07 (95 % CI:0<middle dot>88-1<middle dot>32, P = 0<middle dot>49) or vegetable intake (exp(b): 1<middle dot>06 (95 % CI: 0<middle dot>78, 1<middle dot>43)), P = 0<middle dot>71) with an average of 1<middle dot>51 (95 % CI 1<middle dot>20-1<middle dot>82) and 1<middle dot>40 (95 % CI 1<middle dot>08-1<middle dot>72) serves of vegetables per day in the intervention and control group, respectively. Intervention educators reported higher skills for promoting vegetables at mealtimes, and knowledge and skills for teaching the curriculum, than control (all P < 0<middle dot>001). Intervention fidelity was moderate (n 16/20 and n 15/16 centres used the Mealtime environment and Curriculum, respectively) with good acceptability among educators. The intervention reached 307/8556 centres nationally and was adopted by 22 % eligible centres. Conclusions: The pragmatic self-delivered online intervention positively impacted educator's knowledge and skills and was considered acceptable and feasible. Intervention adaptations, using the MOST cyclic approach, could improve intervention impact on children' vegetable intake.