Long-term outcomes (2 and 3.5 years post-intervention) of the INFANT early childhood intervention to improve health behaviors and reduce obesity: cluster randomised controlled trial follow-up

International audienceAbstract Background The few health behavior interventions commencing in infancy have shown promising effects. Greater insight into their longer-term benefits is required. This study aimed to assess post-intervention effects of the Melbourne INFANT Program to child age 5y on diet, movement and adiposity. Methods Two and 3.5y post-intervention follow-up (2011–13; analyses completed 2019) of participants retained in the Melbourne INFANT Program at its conclusion (child age ~ 19 m; 2008–10) was conducted. The Melbourne INFANT Program is a 15-month, six session program delivered within first-time parent groups in Melbourne, Australia, between child age 4-19 m. It involves strategies to help parents promote healthy diet, physical activity and reduced sedentary behavior in their infants. No intervention was delivered during the follow-up period reported in this paper. At all time points height, weight and waist circumference were measured by researchers, children wore Actigraph and activPAL accelerometers for 8-days, mothers reported children’s television viewing and use of health services. Children’s dietary intake was reported by mothers in three unscheduled telephone-administered 24-h recalls. Results Of those retained at program conclusion (child age 18 m, n = 480; 89%), 361 families (75% retention) participated in the first follow-up (2y post-intervention; age 3.6y) and 337 (70% retention) in the second follow-up (3.5y post-intervention; age 5y). At 3.6y children in the intervention group had higher fruit (adjusted mean difference [MD] = 25.34 g; CI 95 :1.68,48.99), vegetable (MD = 19.41; CI 95 :3.15,35.67) and water intake (MD = 113.33; CI 95 :40.42,186.25), than controls. At 5y they consumed less non-core drinks (MD = -27.60; CI 95 :-54.58,-0.62). Sweet snack intake was lower for intervention children at both 3.6y (MD = -5.70; CI 95 :-9.75,-1.65) and 5y (MD = -6.84; CI 95 :-12.47,-1.21). Intervention group children viewed approximately 10 min/day less television than controls at both follow-ups, although the confidence intervals spanned zero (MD = -9.63; CI 95 :-30.79,11.53; MD = -11.34; CI 95 :-25.02,2.34, respectively). There was no evidence for effect on zBMI, waist circumference z-score or physical activity. Conclusions The impact of this low-dose intervention delivered during infancy was still evident up to school commencement age for several targeted health behaviors but not adiposity. Some of these effects were only observed after the conclusion of the intervention, demonstrating the importance of long-term follow-up of interventions delivered during early childhood. Trial registration ISRCTN Register ISRCTN81847050 , registered 7th November 2007