Experimental protocol

All measures were taken at posting trot and working, 3-point position canter, which are gaits that showjumpers are familiar with.

A repeated measures, single visit study design was implemented. All measures lasted <2 h and took part on a 20 × 40 m geotextile footed indoor arena. After obtaining informed consent of the participants, thirteen inertial measuring units (IMUs) (Delsys Trigno, Natick, MA, USA) were placed on the rider, saddle, bridles and horse (Fig. 1). Five trials were performed: (1) one standardization trial, (2) one trial with short stirrups, (3) one trial with long stirrups, (4) one trial saddle. Trials 2–5 were randomized. On the standardization trial (refer to ‘Horse set-up’ for details of the saddle setup), 120 strides on the right hand and 120 strides on the left hand were recorded at trot and canter while the horse was being ridden around the arena. On the remaining trials, 60 strides on the right hand and 60 strides on the left hand at trot and canter were recorded (refer to ‘Interventions design’ for details of the saddle setup). Participants were encouraged to keep the horse balanced and straight at all times.

(1) Helmet, (2) 7th cervical vertebrae, (3) anterior sacral promontory, (4) cantle of the saddle, (5) forelimb of the horse. (C) Close-up picture of the saddle at a neutral position. Two circles indicate the most distal ends of the saddle in contact with the horse’s back. Arrow indicates the lowest part of the seat, which was aimed to be at the center between the two most distal ends, with the horse at halt and squared.

The horse was cooled down at walk on loose reins upon completion of all five trials. The horse had a >24-h rest between trials.