Accurate pain assessment is fundamental to optimal pain management, representing a
major welfare concern. Pain assessment has received considerable attention in farm,
laboratory and companion animals, however, there is little objective equine pain
research. This study aimed to objectively identify behavioural indicators of pain,
examining both acute post-operative (castration) and chronic (laminitis) pain.
Male thoroughbred horses (n=T0/group) underwent castration or sham castration
(control) performed under either standing surgical sedation (SS) or general anaesthesia
(GA). Horses were monitored for 24 hours pre-operatively and 48 hours post¬
operatively. Additionally, seven acute laminitic horses and paired age, sex and breedmatched controls were monitored for up to five days. Assessments were made using
time-lapse video recording and direct observation of undisturbed spontaneous behaviour
and evoked human interaction behaviours. Data were acquired using The Observer™
and analysed using generalised mixed effects (GME) and discriminant analysis (DA).
GA and SS castrates spent more time with their ears back and displayed a higher
frequency of stepping away than controls in interactive tests (P<0.017, GME). Head
level with withers increased post-operatively in SS castrate, but not control horses
(P<0.001, GME). Additionally, sham GA resulted in increased inattentive behaviour
and hindlimb resting and reduced 'head up' and recumbency (P<0.039, GME).
Laminitic horses showed reduced hindlimb resting and walking with increased lying,
'head level' and forelimb lifting compared to controls (P<0.046, GME). Accuracy of
discrimination (DA) between 'painful' and 'pain-free' horses was >78.6% in acute and
We identified behavioural parameters indicative of pain and discomfort in acute and
chronic pain states. Acute pain may be most accurately identified through the
examination of evoked behaviour, whilst changes in spontaneous behaviour appear
more altered in chronic pain.