Hi All

Francis Burton (thank you!) made me aware of this.

Recent developments in technology suggest that it may be possible to factor out heart rate responses to the exercise component, leaving the emotional component.

Below is a copy of an abstract as the full paper isn't yet published.


Detection of emotionality in horses during physical activity

F Jansen1, J Van der Krogt1, K Van Loon1, V Avezzù2, M Guarino2, S Quanten1, D
Berckmans1

1 Division Monitor, Model & Manage Bioresponses (M3-BIORES), K.U. Leuven, Belgium
2 Department of Veterinary and Technological Sciences for Food Safety, University of Milano, Italy

In equine sports, the horses performance depends on both its physical abilities as well as its emotional status. The objective of this research was to measure in a continuous and non-invasive way emotionality, specifically fear, in horses during training activity.

In the experiment carried out, a horse walked for 5 minutes, then trotted for 10 minutes and finally walked again for 5 minutes. After 5 minutes of trotting stress was applied to the horse by suddenly opening an umbrella, as used by other authors. The experiment was repeated 37 times using two different horses.

During the experiment the horses heart rate and physical activity were continuously measured using the Polar Equine RS800 and the ActiGraph GT1M respectively.

The relationship between the horses physical activity and its heart rate was described in real-time by a single input single output (SISO) mathematical model. This real-time input-output modelling allows determination of the component of the horses heart rate, which is solely related to physical activity.

During the first part of the experiment, before stress is applied, on average 93% of the variance of the measured heart rate could be related to physical
activity (RT2avg = 0.93). When stress is applied during trotting by suddenly opening an umbrella, the physical model accounts for only 58%, on average over 37 experiments, of the variance of the horses heart rate (RT2avg = 0.58).

This increase in the model error can then be used to asses the horses fearfulness. In 33 out of the 37 experiments, the opening of the umbrella could be detected from the physical model error. In the 4 remaining experiments, no stable SISO model could be found. This shows that model-based decomposition of heart rate can be used to monitor stress levels and corresponding welfare levels during training, i.e. while improving physical performance.

Thanks for that Jan. I'd really like to see the whole paper if anyone finds it when it's published. I'd like to know whether they account for habituation to the umbrella as only using 2 horses and 37 umbrella-opening events might well show a drop-off in response. If they don't account for habituation and they can still distinguish between physical and fear responses then it's even more impressive

Using heart rate as an indicator of stress is a nice idea but is potentially problemmatic when the horse is working significantly above resting level, as in Join Up. I don't know what controls and standards were used (or are going to be used) in the Monty Roberts study. If I were a reviewer of the submitted paper, that would probably be my primary concern. I wonder whether it would be possible to monitor heart rate in horses that are running of their own volition, e.g. with friends in a paddock, at a similar speed to that produced by "sending away" - and to do this before the Join Up trial. That would give an indication, albeit rather crude and without using a swanky mathematical model, of the magnitude of HR increase one would expect due to simply exercising. Any excess could then be attributed to stress or emotional arousal. That leads to the question of whether non-stressful emotional arousal can increase heart rate, and if it can (which I think is almost certain) how does one discriminate between that and stress?

Finding out whether or not Join Up is stressful for horses is one of the stated aims. I hope there are others, or else the study may end up rather inconclusively. For example, it would be useful if measurable effects of Join Up on subsequent behaviour were also found. Then the reasons for these effects could be examined, i.e. the mechanism of Join Up.

As comparison HRs weren't mentioned in the 'results' in the Listening Post, I looked them up (results from some HR measurements taken from race horses in training, so these were very fit horses):

Resting = 30-40 bpm
Walk = 70-80 bpm
Trot - 110-120 bpm
Canter = 160-200 bpm
Gallop (racing) = 210-220 bpm

Some of these ranges seem quite broad, not sure how many horses were measured and the various differences in them all, but maybe there's an element of emotional stress in these measurements, ie, resting HR of 40 bpm is 30% more than the lower end of that scale, with the canter range being 25%. But then at trot it's 1%, walk 1.5% and gallop 0.5%, so maybe I'm talking a load of old baloney

...Horses React to Human Heart Rates, Study Finds:

An increase in a human's heart rate affects the heart rate of the horse they are leading or riding, researchers at the Swedish University of Agricultural Sciences recently reported.
Linda Keeling, PhD, and colleagues tested horses and riders to see if humans inadvertently communicate fear and anxiety to horses. Using heart rate as a fear indicator, the researchers asked 20 people with varying levels of horse experience to walk and ride 10 horses from Point A to Point B four times. The researchers told participants an umbrella would open as they rode or led the horse on the fourth pass. The umbrella never opened, but heart rates in both horses and humans increased during the fourth trip between the points, when the human expected the umbrella to open.

"The increase in the horses' heart rates probably means that they are more alert and prepared to react to any potential danger," Keeling said. "In the wild, horses are adapted to respond to other animals in their group. A startle reaction is more likely when the horse is very alert."

If you are a nervous person leading or riding a horse, your nervousness might increase the likelihood of the "spook" that you are anxious to avoid.

The study, "Investigating horse-human interactions: the effect of a nervous human," was published in the July 2009 issue of The Veterinary Journal. The abstract is available on PubMed.

http://www.thehorse.com/ViewArticle.aspx?ID=14464