Early indicators of child obesity to aid future clinical trials for lifecycle obesity prevention

Background: 39M children worldwide are overweight or have obesity, accelerating risk for adult non-communicable diseases. Presently, interventions to prevent obesity have had limited success due to poor timing and lack of personalisation. Objective: We aimed to identify early-life predictors of childhood obesity (ChOB) that could aid targeting specific population subsets for obesity prevention interventional studies. Methods: Data were from the Raine Study Gen2 participants (n=1494). Anthropometric and genetic predictors evaluated included birthweight (BW), early-life BMI (1-3 years), and three polygenic scores (PGS) [two BW-PGSs (BW-PGS2016 and BW-PGS2019) and a ChOB-PGS], developed from BW and ChOB genome-wide-association-studies, respectively. Multivariate analyses were performed to investigate associations between predictors and child-BMI (5-, 8-, 10-years). Results: BW-PGS2019 associate with child-BMI at 5-years. BW-PGS2016 was not associated with child-BMI. Remaining predictors positively associate with child-BMI at 5-, 8- and 10-years (p<0.001). Early-life BMI, ChOB-PGS and BW accounted for up to 38.7%, 5.8% and 3.4% of the variability in child-BMI, respectively. Conclusions: Our data suggest early-life BMI is a better predictor of child-BMI than ChOB-PGS, and BW, accounting for up to ten-fold more variance in child-BMI. Future interventional studies to mitigate obesity could target early-life BMI as a marker to identify children at the highest risk.