Importance: Testing the long-term usefulness of a childhood intervention and determining the best age of implementation are important for translation and policy change.

Objectives: To investigate among children aged 3 years the long-term effectiveness an intervention that aimed to reduce dental caries among South Australian Aboriginal children and to assess if children in the delayed intervention (DI) group had any benefit from the intervention from ages 2 to 3 years and if the intervention usefulness was greater when delivered between pregnancy and age 2 years (immediate intervention [II] vs ages 2 to 3 years [DI]).

Design, Setting, and Participants: Secondary analysis of a randomized clinical trial. The study enrolled 448 pregnant women across South Australia, Australia, at baseline (February 1, 2011, to May 30, 2012), with 223 randomly allocated to the II group and 225 to the DI group. Three-year follow-up data were collected November 2014 to February 2016.

Interventions: The intervention comprised dental treatment to mothers, fluoride varnish application to children, and motivational interviewing delivered together with anticipatory guidance. This was delivered during pregnancy and at child ages 6, 12, and 18 months for the II group and at child ages 24, 30, and 36 months for the DI group.

Main Outcomes and Measures: The mean number of decayed teeth measured at child age 3 years.

Results: There were 324 children at age 3 years (52.3% male). The mean number of decayed teeth at age 3 years was 1.44 (95% CI, 1.38-1.50) for the II group and 1.86 (95% CI, 1.89-2.03) for the DI group (mean difference, -0.41; 95% CI, -0.52 to -0.10). The predicted mean number of decayed teeth at age 3 years for the DI group was 2.15. Between ages 2 and 3 years, the caries increment for the II group was 0.82 (95% CI, 0.75-0.89), compared with 0.97 (95% CI, 0.87-1.17) for the DI group (P?=?.05).

Conclusions and Relevance: At the 3-year follow-up, II children had less dental caries than DI children, DI children developed dental caries at a lower trajectory than predicted had the intervention not been received at ages 2 to 3 years, and the caries increment was less between ages 2 to 3 years among II children compared with DI children. This study suggests that the best time to implement the intervention is earlier rather than later infancy.

Trial Registration: Australian and New Zealand Clinical Trial Registry Ideintifier: ACTRN12611000111976.