What Can Heart Rate Variability Reveal about the Horses’ Inner State?

Veterinary

What can Heart Rate Variability reveal about the Horses’ Inner State?" This was one of the first topics of discussion at the 2018 International ISES conference in Rome, Italy, last September.

Emerging technologies and methods for a non-invasive acquisition and analysis of physiological signals in horses have become a field of study in equine and equitation science. Researchers Marlin, Lanatà, Nardelli, and Baragli from the universities in Cambridge and Pisa presented their study at the 2018 ISES conference.

The Study

A horses’ heart rate at rest typically ranges from 15 to 40 bpm, depending on many factors, including how the heart rate is recorded, the external environment, the time of day, arousal, hydration status and when taken in relation to time of feeding. When measured with a stethoscope or a heart rate monitor it may appear that there is relatively little variation in heart rate over time. However, ECG recordings which show individual beats and allow calculation of beat to beat heart rate (inter-beat intervals (IBI) or RR intervals), reveal a subtle pattern of variation in instantaneous heart rate over time. This is referred to as heart rate variability (HRV).

Techniques to measure pulse rate variation were developed with advances in electronics and digital signal processing in the 1960’s and lead to an explosion of interest in HRV in health and disease.

Heart rate is controlled by the sino-atrial (SA) node. The primary inputs to the SA node that induce beat to beat variations in rate are respiration, blood pressu- re and thermoregulation but stress, hormones, electrolytes, acid-base balance, activity, eating, sleep, arousal and disease are also modifiers. The primary inputs originate from activity of the sympathetic (SNS) and parasympathetic nervous systems (PNS). In simple terms, incre- ased SNS activity and or decreased PNS activity leads to decreased HRV. The main methods for analysing HRV are time domain, frequency domain and non-linear techniques. Whilst on one level collecting and analysing data to generate HRV indices is simple, there are many requirements that if not met can lead to a high risk of artefacts and unreliable conclusions.

In horses HRV has been used to study behaviour/equitation (including stress, human-horse interaction, massage, nosebands, hippotherapy, transport and head and neck position), pregnancy, the effects of exercise and disease. Current technology offers exciting new oppor- tunities for monitoring vital signs in horses. Moreover, commercial systems are emerging for research studies, primarily in laboratory settings, for improving performance as well as monitoring of patients suffering from specific diseases.

Nowadays research and free-living environment applications require a more robust and reliable monitoring of subjects both in natural and laboratory conditions, either for pathological or behavioural investigation. Many current physiological monitoring systems measure conventional vital signs (e.g., heart rate, activity) with standard Ag/AgCl, or plastic conductive electrodes, but these are limited to a restricted set of applications. Moreover, there is a gap in developing and validating wearable physiological monitoring systems and the algorithms that convert data into useful and actionable information for medical management, welfare monitoring and performance optimization.

Recent technological advances in miniaturized sensors, material sciences, robust embedded computing, signal processing and artificial intelligence are empowering state-of- the-art systems to leap forward beyond measurement and telemetry of conventional vital signs to provide “smart” and personalized decision aids to monitor and improve health and performance. Such “smart” physiological monitoring systems will use real-time physiological data fusion to predict changing status (improving or worsening) and will eventually allow ubiquitous use in every-day activities, health care monitoring, as well as use in controlled environmental conditions.

HRV to Monitor Stress in Horses

Heart Rate Variability (HRV) analysis has received a growing attention from clinicians and physiologists, as a powerful tool to assess autonomic nervous system (ANS) activity. Standard methodologies used to investigate HRV consist of time and spectral analysis methods. Their effectiveness in understanding the autonomic control of R-R interval fluctuations and in the prediction of cardiovascular diseases, as acute myocardial infarction, is already known.

However, if spectral analysis of HRV is a powerful tool to assess the regulatory mechanisms of the sympathetic and parasympathetic subsystems over the heart, this technique requires the stationarity of the signals. A wide range of physiological studies imply the processing of non-stationary signals, e.g., tilt tables, stress tasks, emotional stimulation. For that reason several time-frequency and time-varying methodologies of the HRV analysis have been introduced.

Research studies have revealed non-linear aspects of the dynamic interaction between sympathetic and parasympathetic nervous systems. Currently, non-linear methodologies applied to cardiovascular signals are considered to be a strong complement to standard time-frequency analysis. The prominent non-linear techniques can be grouped into four main families: fractal measures, entropy and complexity metrics, symbolic dynamics methods and Poincaré maps. Moreover, multiscale approaches are used, based on the rationale that complex systems generally reveal long-range correlation structures over multiple temporal scales. Methodological advances in estimating the complex behaviour of physiological systems have been directed towards the study of multichannel recordings (e.g., cardiovascular, respiration, blood pressure signals). Novel research has demonstrated that multivariate approaches for the investigation of non-linear autonomic interactions often outperform univariate methodologies.

Conclusion

From observing a heart rate monitor or listening to the heart with a stethoscope we get the impression that heart rate is quite regular. Analysis of ECG recordings shows that there are subtle patterns of variation in rate referred to as heart rate variability (HRV). This variation or sometimes lack of variation in human medicine reveals important diagnostic information about health and disease. Analysis of this variation may also provide an insight into behavioural states but this is still a controversial area. This type of analysis is complex and whilst commercial systems do exist which provide an estimate of HRV they are open to artefact and being misleading.