Application Of Longitudinal Data Analysis Allows To Detect Differences In Pre-Breeding Growing Curves Of 24-Month Calving Angus Heifers Under Two Pasture-Based Systems With Differential Puberty Onset

BACKGROUND Longitudinal data analysis contributes to detect differences in the growing curve by exploiting all the information involved in repeated measurements, allowing to distinguish changes over time within individuals, from differences in the baseline levels among groups. In this research, longitudinal and cross-sectional analysis were compared to evaluate differences in growth in Angus heifers under two different grazing conditions, ad libitum (AG) and controlled (CG) to gain 0.5 kg day(-1). RESULTS Longitudinal mixed models show differences in growing curve parameters between grazing conditions, that were not detected by cross-sectional analysis. Differences (P < 0.05) in first derivative of growth curves (daily gain) until 289 days were observed between treatments, AG being higher than CG. Correspondingly, pubertal heifer proportion was also higher in AG at the end of rearing (AG, 0.94; CG, 0.67). CONCLUSION In longitudinal studies, the power to detect differences between groups increases by exploiting the whole information of repeated measures, modelling the relation between measurements performed on the same individual. Under a proper analysis, valid conclusion can be drawn with fewer animals in the trial, improving animal welfare and reducing investigation costs. (c) 2019 Society of Chemical Industry