We have observed major back-shape changes over time in some horses, the speed of which may be influenced by work-type, skeletal maturity, nutrition and saddle-fit. Currently, there are no longitudinal data quantifying changes in back-shape and no objective data assessing the importance of each variable. The objectives were to: quantify back-shape changes over time; describe the effects of age, breed, weight, work-discipline, saddle-management and pain elicited on palpation.
Preliminary data from 98 sport horses in regular work in an on-going one-year longitudinal study. Weight and management changes were recorded. The thoracolumbar shape was measured at T18, T13 and T8 using a Flexible Curve Ruler  every two months. Ratios were calculated for each site for the widths 3 cm and 15 cm ventral to the dorsal midline. Differences in back-shape were determined using Wilcoxon signed-rank test. Chi-square test was used to assess associations between categorical variables, Spearman rank correlation for continuous variables, Mann-Whitney-U for continuous variables with binary outcomes and Kruskal-Wallis to assess ≥3 groups.
T8 and T18-ratios were significantly different after two months (P<0.001) and T13 after four months (P = 0.004). Decreased ratio at T8 was related to ill-fitting saddles (P = 0.005), saddles tipping back (P = 0.020), pain on palpation (P = 0.032) and greater age (P = 0.004). Increased T18-ratio was associated with weight gain (P<0.001), more prevalent in stallions and geldings than mares (P = 0.035) and in dressage horses than other disciplines (P = 0.043).
Back-shape can change within two months, the speed and direction of which are influenced by saddle-fit, age and bodyweight. Pain elicited on palpation of the back may indicate that changes have occurred. Thus saddle-fit should be reassessed.