UPPER AIRWAY ASSESSMENT IN EXCERISING HORSES: FUTURE ROLE
FOR RESPIRATORY SOUND ANALYSIS
Frederik J. Derksen, DVM, PhD, ACVIM Diplomate
Department of Large Animal Clinical Sciences
College of Veterinary Medicine
Michigan State University, East Lansing, MI 48824-1314
Introduction
The upper airway is most often evaluated in standing horses by use of palpation,
radiography, ultrasonography, scintigraphy, or endoscopy. However, in many cases, upper airway
evaluation in standing horses fails to yield a definitive diagnosis. In such cases, upper airway
evaluation during exercise is warranted. At present, videoendoscopy during exercise and flow
mechanics measurements are the most common methods of evaluation. In this paper, I will suggest
that evaluation of upper airway sound may become useful in the assessment of upper airway
obstructions in exercising horses.
Videoendoscopy during Exercise
The availability of the fiberoptic endoscopy in the early 1970s was a major breakthrough in
our ability to assess the upper airway. Prior to this time, little was known about upper airway
function in health and disease. Only the most common and obvious lesions, such as idiopathic
laryngeal hemiplegia had been described. With the availability of the fiberoptic endoscope came
descriptions of a variety of upper airway lesions, including lymphoid hyperplasia, entrapment of the
epiglottis, and dorsal displacement of the soft palate (DDSP).
In some cases, endoscopic findings in resting horses are easily interpretable and yield a
definitive diagnosis. In other cases, it is difficult to evaluate the functional significance of lesions.
For example, lymphoid hyperplasia was first thought to be an important cause of airway obstruction
and of exercise intolerance in performance horses. Later it was realized that lymphoid hyperplasia is
probably a normal development in young racehorses and does not cause significant airway
obstruction.
When horses exercise, large pressure swings necessary to produce the high inspiratory and
expiratory air flows cause dynamic changes in the upper airway. In horses with dynamic lesions,
these large pressure swings can actually cause airway obstruction not appreciable in the resting
animal. Therefore, we adapted fiberoptic endoscopy for use in exercising horses. The first paper on
upper airway endoscopy in exercising horses appeared in 1989 and was rapidly followed by others.
It is now clear that videoendoscopy during exercise is an excellent method for evaluation of upper
airway function. In a recent paper, Martin et al. reported on the clinical evaluation of poor training or
racing performance in 384 horses. Significant upper respiratory tract abnormalities were the cause of
poor performance in 148 horses (42.6%) undergoing examination on a high-speed treadmill. The
most common abnormality was DDSP in 43 (29%) cases, followed by laryngeal hemiplegia (LH) in
33 (22.3%) cases. Of these horses, 39 (26%) had more than one upper airway problem.
Dorsal displacement of the soft palate is usually an intermittent condition occurring only
during exercise. Therefore, at present, videoendoscopy during exercise is required to make a
diagnosis. However, some horses with a history indicative of DDSP do not displace the soft palate
when examined on the treadmill, suggesting that treadmill exercise differs from racing. This fact
must be considered when interpreting the results of all treadmill exercise-related examinations.
Laryngeal hemiplegia has been classified using a four-step grading system. Grade 1 and
Grade 2 are thought to be normal, while Grade 4 LH results in severe airway obstruction. Horses
with Grade 3 LH present the greatest diagnostic challenge. In a recent study, 20 of 26 horses with
Grade 3 LH had severe airway obstruction when examined during intense exercise. These data
demonstrate the use of videoendoscopy during exercise in the evaluation of these cases.
Videoendoscopy during exercise, while a valuable diagnostic tool, is still subjective in
nature. In some cases, the functional consequences of lesions are readily apparent. In other cases,
quantitative evaluation is needed to interpret videoendoscopic findings.
Quantitative Evaluation of Upper Airway Function
Quantitative evaluation of upper airway function is most easily accomplished on a highspeed treadmill. Evaluation of air flow mechanics requires the measurement of air flow and
transupper airway pressures. To measure air flow, horses are fitted with a face mask and a flow
measurement device. Pressures necessary to move the air flow through the upper airway portion of
interest are measured using a differential pressure transducer. The ratio of pressure and flow in a
rigid tube is called resistance. However, when the walls of the tube are not rigid, like in the upper
airway, and when air flow rates accelerate and decelerate, as during inhalation and exhalation, the
ratio of pressure and flow is properly called impedance. Measurement of air flow mechanics has
been used to determine the best surgical treatment for a variety of upper airway obstructive disorders.
Another way to quantitate upper airway function is by the evaluation of tidal breathing flowvolume loops. Tidal breathing flow-volume loops are formed when air flow rate is continuously
plotted against volume during successive respirations. Tidal breathing flow-volume loops are most
sensitive when obtained during high-speed exercise. A variety of indices can be used to quantify the
effect of airway obstructive lesions on the flow-volume loop. We have demonstrated that flowvolume loop analysis is a more sensitive test of airway obstruction in horses with LH than
conventional flow mechanics techniques.
Evaluation of Respiratory Sounds in Exercising Horses
A disadvantage of videoendoscopy during exercise and measurements of flow mechanics is
that a high-speed treadmill is required. A high-speed treadmill is not universally available and
despite its usefulness, high-speed treadmill exercise is not equivalent to racing. Racing speeds are
generally higher than treadmill speeds and track surface, wind resistance, and the weight of jocky,
sulkey, or driver may all affect performance.
It has long been reported that upper airway abnormalities are associated with distinctive
respiratory sounds. Presently, we are examining the possibility that respiratory sounds associated
with each upper airway obstructive condition are unique to that condition. If this is the case,
evaluation of upper airway sound could yield a specific diagnosis. Recording of respiratory sounds
in exercising horses has several challenges, including placement of the microphone and overcoming
interference by extraneous noises associated with exercise, such as hoof beats and wind noise. For
this reason, we have developed the following technique for the recording of respiratory sounds. A
unidirectional recording microphone is placed approximately 4 cm from the tip of the horse's nose.
The microphone is connected to a cassette recorder containing a compression circuit. The
compression circuit is activated by the horse's respiratory sound and squelches environmental sounds
associated with exercise. Respiratory sounds are then analyzed using a computer-based spectrogram
program. In human medicine, spectrogram analysis of speech is a large field of study and practical
applications of this field include speech therapy and voice recognition. Spectrogram analysis of
recorded respiratory sounds of horses is easily accomplished using computer software and a personal
computer. Using this technique, respiratory sounds in strenuously exercising horses can be measured
easily and inexpensively under field conditions.
In a recent study, we tested the hypothesis that LH and DDSP are associated with unique
respiratory sounds and that spectrogram analysis of these sounds allows differentiation of these two
conditions. We found that in normal horses galloping or trotting at a speed corresponding to
maximum heart rate, expiratory sounds predominated. Expiratory sound was present throughout
exhalation. Expiratory sound frequency range ranged up to a mean of 4.2 KHz. Because the human
hearing frequency range is from 20 to 20,000 Hz, expiratory sounds of exercising horses are well
within the audible range. Most of the sound energy was contained in frequencies up to a mean of 0.8
KHz. Using spectrogram analysis, these timing, frequency, and amplitude characteristics formed an
easily recognizable pattern.
Laryngeal hemiplegia was induced in our experimental horses using a local anesthetic
technique. As verified by endoscopy before and after exercise, all treated horses had grade 4 LH. In
grade 4 LH, exercising horses experience dynamic collapse of the paralyzed arytenoid cartilage,
creating a very narrow rima glottis during inhalation. During exhalation, positive pressure in the
larynx relative to atmospheric pressure reopens the rima glottis. Dynamic collapse of the affected
arytenoid cartilage causes inspiratory upper airway obstruction and inspiratory noise production. In
LH-affected horses, expiratory sounds did not differ from those observed in normal horses.
Similarly, expiratory flows and pressures in horses with LH are also unaffected. In contrast,
inspiratory sounds differed markedly. In every affected horse, inspiratory sounds occurred
throughout inhalation. This sound featured three distinct frequency bands centered at approximately
0.3, 1.6, and 3.8 KHz. In human speech analysis, these frequency bands are called formants and are
characteristic of certain sounds such as whispered vowels. These formants occur because the upper
airway has its own resonant frequencies and preferentially transmit the formant frequencies.
Dorsal displacement of the soft palate causes an expiratory upper airway noise. Most upper
airway conditions cause inspiratory airway obstruction. Dorsal displacement of the soft palate is an
exception to this rule, as the condition does not affect inspiratory flow mechanics. In DDSP, the
noise occurred throughout exhalation and was characterized by a rapid periodicity or rattling sound.
Most likely, this sound is generated by the vibrations of the dorsally displaced soft palate and
modified by the nasal chambers. Dorsal displacement of the soft palate was not associated with
abnormal respiratory sounds in all horses. In one animal, no abnormal sounds were observed
throughout the exercise period at maximum heart rate. Instead, abnormal sounds were evident as the
horse decelerated after the exercise period. In two other horses, some breaths taken during highspeed exercise were associated with a typical rattling noise, while other breaths were not. We and
others have previously reported that in some horses with confirmed DDSP, there is no history of
abnormal noise production. Factors that may influence whether or not a horse with DDSP has an
abnormal respiratory noise include air flow velocity, the compliance of the dorsally displaced soft
palate, and the geometry of the individual horse's airway.
Spectrogram analysis of respiratory sounds in exercising horses may have important
applications. It is possible that all of the upper respiratory conditions of horses are associated with
unique sounds and spectrogram patterns. If this is the case, simple recording of respiratory sound
under field conditions could yield a diagnosis of specific upper airway lesions, thereby avoiding the
need for endoscopic examinations of a high-speed treadmill. It is now also possible to evaluate the
efficacy of surgical techniques in reducing the upper airway noises associated with many of the
conditions.
References
Derksen, F.J., Holcombe, S.J., Hartmann, W., Robinson, N.E. and Stick, J.A. (1999) Spectrogram
analysis of respiratory sounds in exercising horses. AAEP Proceedings 45, 314-315.
Hackett, R.P., Ducharme, N.G., Fubini, S.L., Erb, H.N. (1991) The reliability of endoscopic
examination in assessment of arytenoid cartilage movement in horses. Part I: subjective and
objective laryngeal evaluation. Vet. Surg. 20, 174-179
Hammer, E.J., Tulleners, E.P., Parente, E.J. and Martin, B.B. (1999) Videoendoscopic assessment of
dynamic laryngeal function during exercise in horses with grade-III left laryngeal
hemiparesis at rest: 26 cases (1992?1995). J. Am. Vet. Med. Assoc. 212, 399-403.
Holcombe, S.J., Derksen, F.J., Stick, J.A., Robinson, N.E. and Boehler, D.A. (1996) Effect of nasal
occlusion on tracheal and pharyngeal pressures in horses. Am. J. Vet. Res. 57, 1258-1260.
Holcombe, S.J., Derksen, F.J., Stick, J.A. and Robinson, N.E. (1998) Effect of bilateral blockade of
the pharyngeal branch of the vagus nerve on soft palate function in horses. Am. J. Vet. Res.
59, 504-508.
Lumsden, J.M., Derksen, F.J., Stick, J.A., Robinson, N.E. and Nickels, F.A. (1994) Evaluation of
partial arytenoidectomy as a treatment of equine laryngeal hemiplegia. Equine Vet. J. 26,
125-129.
Lumsden, J.M., Derksen, F.J., Stick, J.A. and Robinson, N.E. (1993) Use of flow-volume loops to
evaluate upper airway obstruction in exercising Standardbreds. Am. J. Vet. Res. 54, 766-775.
Lumsden, J.M., Stick, J.A., Caron, J., Nickels, F.A., Brown, C.M., Godber, L.M. and Derksen, F.J.
(1995) Upper airway function in performance horses: Videoendoscopy during high-speed
treadmill exercise. Comp. Cont. Ed. 17, 1134-1143.
Martin, B.B., Reef, V.B., Parente, E.J. and Sage, A.D. Clinical evaluation of poor training or racing
performance in 348 horses (1992?1996). AAEP Proceedings 45, 322-324.
Morris, E.A. and Seeherman, H.J. (1990) Evaluation of upper respiratory tract function during
strenuous exercise in racehorses. J. Am. Vet. Med. Assoc. 196, 431-438, 1990.
Petsche, V.M., Derksen, F.J., Berney, C.E. and Robinson N.E. (1995) Effect of head position on
upper airway function in exercising horses. Equine Vet. J. Suppl. 18, 18-22.
Stick, J.A. and Derksen, F.J. (1989) Videoendoscopy during exercise determines surgical treatment
of laryngeal hemiplegia in a Standardbred colt. J. Am. Vet. Med. Assoc. 195, 619?622.
Tetens, J., Derksen, F.J., Stick, J.A., Lloyd, J.W., and Robinson, N.E. (1996) Efficacy of prosthetic
laryngoplasty with and without bilateral ventriculocordectomy as treatments for laryngeal
hemiplegia in horses. Am. J. Vet. Res. 57, 1668-1673.