LAUSR

BACKGROUND The overall impact of interventions targeting multiple behaviours remains largely unexplored. OBJECTIVES This study adopted an integrative lifestyle pattern analysis approach to assess the overall effectiveness of an early childhood intervention on change across multiple behaviours. METHODS The Melbourne INFANT program was a 15-month cluster-randomized controlled trial involving 4-month-old infants and their parents at baseline in 2008 (n = 542). The intervention included six education sessions helping parents to promote a healthy diet, physical activity and limit sedentary behaviour in their infants. Participants were followed-up twice post-intervention, at ages 3.6 (2011) and 5 years (2013), to assess sustained effects of the intervention. Previous principal component analyses identified two lifestyle patterns from dietary intake, outdoor time and television viewing time. Random effect linear regression models were conducted to assess the impact of the intervention on lifestyle patterns. RESULTS The intervention group had a lower 'Discretionary consumption and TV' lifestyle pattern score than the control group at all time points with adjusted mean difference: -0.29, 95% CI -0.49, -0.09, p = 0.004 post-intervention at age 1.5 years; -0.29, 95% CI -0.54, -0.04, p = 0.02 at the first follow-up (age 3.6 years); and -0.21, 95% CI -0.43, 0.01, p = 0.06 at the second follow-up (age 5.0 years). No evidence of between-group differences was found for the 'Fruit, vegetables and outdoor' lifestyle pattern score. CONCLUSION This early childhood intervention designed to promote change in more than one obesity-related behaviour was effective in improving correlated unhealthy lifestyle behaviours. Lifestyle pattern analysis is a useful and interpretable approach for evaluating multi-behavioural interventions.