Researchin VeterinaryScience1997,62, 205-211
I~[~'~
Airflow mechanics in models of equine obstructive airway
disease under conditions simulating exercise
W. M. BAYLY, Department of Veterinary Clinical Sciences, Washington State University, Pullman, Washington USA,
R. F. SLOCOMBE, School of Veterinary Science, University of Melbourne, Parkville, Victoria, Australia
SUMMARY
Effects of respiratory tract obstructions on ventilatory mechanics in horses exercising at high speeds were tested with a fibreglass
replica of the airways (nares to mainstem bronchi) of an adult horse. Segmental pressures were recorded at six sites along the
model at four different unidirectional flows (1300 - 4100 litre min-1), and the respective resistances (R) to airflow were calculated. The external nares and the larynx made the greatest contributions to the total resistance (RToT) when no obstruction was present. Modifying the model to simulate severe pharyngeal lymphoid hyperplasia (PLH)had no effect on R at the larynx or at any
point in the trachea under these flow conditions. Two 16 litre anaesthetic rebreathing bags were attached to the bronchial end of
the model, and tidal ventilation generated by a piston pump. Upper (nares to pharynx) and lower tract R (Ru and RL) and RTOT,
and dynamic compliance were determined for pump volumes (VD) of six and 12 litres, at pumping frequencies (fD) of 20 - 100
rain-1 while the airway was clear, and after modifying it to simul~ate either PLH or partial bronchial obstruction. IVfodelcondition
had no effect on RU. However, RL and RTOT were higher in the PLH simulated condition when f~ _> 90 and V.e-- 12 litres
(P<0.05). This suggested that severe PLH may significantly interfere with airflow distal to the site of tVhelesions during high frequency, high volume ventilation of the type seen in galloping horses. With partial bronchial obstruction R L and RTOT were
increased when fp>34 with each V~. The applicability of the model was verified by comparing results from the unobstructed state
with those from normal horses exercising on a treadmill.
•
Ij
.
,
.
AN improved knowledge of the mechanical properties of
the horse's respiratory system may be conducive to an
improved understanding of factors which limit exercise
performance in health and disease. Functional tests of
mechanical properties of airways have been well developed
for use in resting horses (Gillespie et al 1966, Willoughby
and McDonnell 1979). Some of these techniques have been
applied under exercise conditions to evaluate the effects of
laryngeal diseases and their surgical treatments (Derksen et
al 1986, Shappell et al 1988, Belknap et al 1990, Williams
et al 1990, Fulton et al 1991, Lumsden et al 1993).
However, under exercise conditions, extensive testing of
pulmonary mechanics has been largely limited to studies on
ponies (Art and Lekeux 1988a, b, c, 1989, Art et al 1988),
although reports on horses exercising at high intensities
have appeared in recent years (Art et al 1990, Slocombe et
al 1992, Lumsden et al 1993).
For horses during exercise, tidal volume (VT), airflow
(~?), airway pressures and breathing frequencies fib) may
change by at least an order of magnitude. The development
of techniques which allow accurate and repeatable measurement of respiratory tract mechanics over these large
ranges has been very desirable, as such availability facilitates clinical evaluation of the respiratory systems of poorly
performing horses. Such methodology also encourages the
development of models for the scientific study of conditions suspected to adversely affect performance capacity,
particularly those leading to upper or lower airway obstruction, or both.
We sought to develop such a model. Initially, we built a
replica of the upper respiratory system in order to test measurement systems under contrived laboratory conditions
which simulated those encountered during various levels of
exercise. The suitability of the model for the study of these
conditions was evaluated by comparing data produced in
0034-5288/97/030205 + 07 $18.00/0
the unobstructed state with those determined in healthy
horses undergoing submaximal treadmill exercise. We
report on the results of these studies here.
MATERIALS A N D M E T H O D S
An initial plaster mould of the upper respiratory tract
from the external nares to the mainstem bronchi was coated
with fibreglass and the plaster removed to form a replica of
the upper airway of a healthy 455 kg Standardbred gelding.
Prior to pouring the primary plaster mould, the head was
extended to represent the position of a normal galloping
horse. While pouting the plaster the external nares were
seen to passively dilate maximally. Examination of this cast
before application of the fibreglass indicated that a degree
of laryngeal closure had occurred. This was partially rectified by adding plaster to these sites, with the result that the
final diameter of the cast at this point was conservatively
estimated to be 85 per cent of that of the rest of the larynx
(Fig 1). Mechanical properties of the model were assessed
using both continuous and oscillating flow preparations•
Continuous flow
Air was blown through the model at four different, but
constant flow rates of between 1300 and 4100 litre rain -1,
and segmental airway resistances along the model determined by measuring the pressure drop across segments for
known flow rates. Pressures were determined at lateral holes
in the model located at the extemai nares, the nasopharynx,
pharynx, just cranial to the larynx, proximal trachea, midtrachea and distal trachea using differential pressure transducers (Validyne DP-45, Validyne Engineering, Northridge,
CA). Segmental resistances were determined between these
© 1997W. B. Saunders CompanyLtd
206
W. M. Bayly, R. F. Slocombe
Three models of lungs were evaluated by attaching them to
each of the bronchi: fresh lungs obtained from an abattoir, 16
litre rubber anaesthetic rebreathing bags (North American
Drager, Telford, PA), and rebreathing bags suspended in water
filled tanks (bag-in-tank system). Inhomogeneous filling of the
abattoir specimens and the bag-in-tank system rendered them
unsuitable for use. Pump frequency (fp) was increased incrementally from 20 rain-1 to about 100 min-1. Airflow was measured just cranial to the external nares using the pneumotachograph, and segmental pressure changes were determined
between the nares and caudal pharynx and the caudal pharynx
and rebreathing bags for each volume (Vp) and fp. Upper
(from just cranial to the nares to the caudal pharynx), lower
(larynx to the Drager bags), and trans-model pressures (Pu,
PL, PTOT, respectively) were measured with differential pressure transducers (Validyne DP45-34) via lateral holes at the
external nares and caudal pharynx immediately rostral to the
arytenoids, and an oesophageal balloon catheter that was introduced into one of the rebreathing bags (Fig 2). Corresponding
segmental resistances (Ru, RE, RTOT), and dynamic compliance Cdyn of the model were calculated from graphic analysis
of polygraph recordings of pressure, flow and volume data
using the classical methods of Mead and Whittenberger
(1953). Because inertial pressure (Pin) influences the value of
Cdyn in a simple resistance-compliance model, particularly at
high Vp, the likely contribution of inertance (I) to pressures at
points of zero flow was calculated using the equation: Pin =
4rca@V_I, where I was assumed to be 0-026 cmI-I20 litre-1 sec-s
based o~ the work of Young and Tesarowski (1994).
Three conditions were evaluated with the oscillating flow
model: (i) a clear airway (control); (ii) simulated grade 4 PLil,
which was produced as described above; and (iii) a parlial
bronchial obstruction. In the latter situation, the cross-sectional
FIG 1: Cross-sectional view of fibreglass model of equine upper airway
through nasopharynx showing simulated grade 4 pharyngeal lymphoid hyperarea of one bronchus was reduced by 50 per cent by obstructplasia. The arrow is pointing to the right arytenoid which was not completely
ing it with a fenestrated robber stopper.
abducted
For both flow preparations, signals from the transducers
were demodulated and displayed on a multichannel thermal
points for each flow rate when the airways were clear (con- array recorder (Gould TA2000, Cleveland, oil). Tidal volume
trol) and when the pharynx was partially obstructed (PLH was calculated by electronic integration of the flow signal
condition). Grade 4 PLH was simulated by attaching 1 cm which had been previously calibrated by injection of known
diameter spherical glass beads to the dorsal and lateral walls volumes of air through the pneumotachograph using a caliof the pharynx (Fig 1). Airflow was generated using two brated syringe (SRL Medical, Dayton, OH). All catheters and
variable speed electric fans connected in parallel. Flows transducers were phase matched up to 10 Hz with an oscillatwere determined using a pneumotachograph - transducer ing sinusoidal pressure signal using an amplifier/speaker-insystem which had been calibrated using a rotameter (Fisher- box system. Amplitudes remained within 1 per cent of nomiPorter 227-G10/83, Warminster PA). The pneumotacho- nal values up to this frequency.
graph was 155 mm diameter and constructed of fine stainless steel mesh (Mercury Instruments F3000 L, Glasgow).
The pressure across the pneumotachograph screen was
Arytenoids
determined using a differential pressure transducer
(Validyne DP45-26, Validyne Engineering, Northridge CA),
Nares /
Drager bag
and linearity of the screen pressure - flow characteristics
Pneumotachographpump
'~
I~ !
confirmed by measuring pressures across the pneumotachograph for known flows through the rotameter. Because this
system bore limited similarity to the in vivo situation, a system which could mimic a ventilatory cycle was developed.
/ ~~ "i~
Oscillating flow
A variable volume sinusoidal piston pump was built and
used to cycle six and 12 litre volumes of air in and out of
the airway model. The pump was positioned in front of the
model and connected to it via a bifurcated connector. The
pump was made from a machined metal cylinder (30 cm
internal diameter) and a metal piston that was sealed with
an O-ring, and driven by a piston rod and crankshaft.
Differential j ~ J / X X
pressure "~.~=_-~L V \ /
transducers
" " - I ~
Recorder
/
Oesophageal
balloon
catheter
FIG 2: Schematic representation of the piston pump and positioning of components used to obtain flow and pressure measurements in the oscillating
flow preparation
Equine obstructive upper airway disease
Exercise test
Five healthy adult Standardbreds were exercised on a
treadmill (0 per cent slope) at 0 (resting), 4-5 m s-1 (slow
trot), and 10 m s-1 (gallop) for 2 minutes at each speed.
These speeds were selected following preliminary evaluations at a number of additional speeds in order to identify
those which would produce fbs and VTS most similar to
those produced with the unobstructed model. Each horse
wore a facemask in which paired pneumotachographs of
the type used with the models, were placed over each nostril in order to measure inspiratory (V I) and expiratory
(V E) flows. The flow signal was electronically integrated
to give V T. Transpulmonary (PTP) pressure and PU were
measured via a compound balloon catheter which had been
passed through a narus to lie with one balloon in the midthoracic oesophagus and the other in the caudal pharynx as
previously described (Slocombe et al 1992). Recordings
were made continuously for determination of respiratory
mechanical properties, which were calculated from measurements performed on five successive artefact-free vendlatory cycles.
Statistical analyses
Data were expressed as means _+SEM. Analysis of repeated measures were only applicable to the oscillating flow
model data, where effects of model condition and Vp were
analysed by two-way analysis of variance. When means
were significantly different, they were compared using the
least significant difference multiple range test (P<0.05).
207
1.09+0-29 and 1-19-+0-38 cmH20 s litre -t respectively, for
the four flows generated. In the PLH model, corresponding
values of RTOT were 0-83_+0.23, 1.09+0.30, 1.07-+0.34 and
1.34-+0.40 cmH20 s litre-1 respectively. Therefore, partial
pharyngeal obstruction as induced by the PLH model may
have had an effect on RTOT at airflows above 3000 litres
rain -1. The relative contribution of the laryngeal and tracheal resistances to RTOy was unaffected by changes in airflow rates or the presence of the glass beads in the pharynx,
although the contribution of the nares to RTOT was greater
in the model of PLH (Fig 4). The arytenoids were not fully
abducted in the model (Fig 1), and with the nares, the larynx was the site of greatest resistance. Least resistance was
found in the mid- and distal trachea. Reynolds numbers for
the trachea increased with flow and rose from 28,000 to
greater than 63,000 over the flow range used in the study,
indicating that even at low flow rates, some turbulence of
airflow was created in the upper airway.
Oscillating flow preparation
Volume effects. Doubling Vp from 6 to 12 litres had minimal effect on RuI, RUE, RLI, RLE, RTOTI, or RTOTE, for any
model preparation, except at high fp. Pump frequencies at
which volume:related differences in airway resistance were
first demonstrated for a given model preparation are shown
in Table 1. Dynamic compliance of the rebreathing bags
was significantly greater with the Vp at 12 litres (compared
with 6 litres) at all fp for each model condition. Differences
were greatest at low fp (Fig 5).
Oscillating flow
Effects of model condition. Partial bronchial obstruction
RESULTS
Continuous flow preparation
The pneumotachograph - transducer system was linear
for flows ranging from 1000 to 4290 litres rain -1.
Resistances were determined at four flows ranging from
1342 to 4097 litres rain -I when the airway was clear, and to
3893 litres rain -1 with the model of PLH. Resistances at
each site and RTOT increased linearly with flow (Fig 3). In
the control condition RTOT was 0.87-+0.27, 1.04_+0.33,
resulted in increased RTOy for both 6 and 12 litre volumes
at all fp greater than 34 (Fig 6a and 6b). At a Vp of 12 litres
and fp = 91 with oscillating flow in this model preparation,
R E was 1.29+0.02 cmH20 s litre -1 and RTOy was
1.34+0-02 cmH20 s litre -1 compared with corresponding
values of 0.46_+0-01 and 0-58_+0-03 cmH20 s litre-] respectively, for R L and RTOT in the unobstructed model at fp =
95, V_k' = 12 litres. Compared with control values, RTOy
was hagher an the model of PLH when fp was 90 or more,
i:
90
80 -70 -
PTOT= 0.215 V - 11.434
r 2 = 0.96, P = 0.0
~
25
e
20
60
+a
15
~ 50
~ 4o _
30 --
~ ' / "
~
/
20-
, /
10
i
1000
"
i
i
:TOT = ~.026 V _ 18.424
r 2 = 0.97, P = 0.014
I
h
I
1 5 0 0 2 0 0 0 2 5 0 0 3 0 0 0 3 5 0 0 4 0 0 0 4500
Airflow (liters rain -1)
FIG 3: Correlation of trans-model pressure (PToT) to airflow in the continuous
flow preparation of the model for the clear airway and simulated pharyngeal
lymphoid hyperplasia (PLH) conditions Key: • clear airway; • PLH model; - regression line for clear airway; - - regression line for PLH model
10
5
Nares
Naso-
Caudal Larynx
pharynx pharynx
Mid-
Distal
trachea trachea
FIG 4: Relative contributions of resistances in different anatomical segments
in a fibreglass model of the equine extrathoracic respiratory tract for both a
clear airway and pharyngeal lymphoid hyperplasia (PHL) condition determined
using four different flows in the continuous flow preparation of the model.
* Indicates a value that was significantly different to that generated at the
same site with the other form of the model, Key: [ ] clear airway; • PLH model
208
W. M. Bayly, R. F. Slocombe
1.6
3.0
(a)
1.4
2.5
1.2
2.0
1.0
0.8
1.5
0.6
1.0
0.4
©
o
0.5
0.2
0 ~
15
0
15
25
35
45
55
65
75
85
95
Respiratory frequency (minute -1)
105
I
25
P
I
I
I
I
35
45
55
65
75
85
95
Respiratory frequency (minute -1)
105
FIG 5: Dynamic compliance in a model of the equine respiratory system with oscillating flow volumes of 6 fitres (a) and 12 litres (b), and clear airway, simulated
pharyngeal lymphoid hyperplasia (PLH) and partial bronchiar obstructive conditions Key: • clear airway; • PLH model; • bronchial obstruction
and Vp = 12 litres (Fig 6b). The elevations in RTOT seen
with both models of obstruction were due to increases in
RL, as no differences in R u were detected for any volume,
airway condition or fp. The relative contribution of R u to
RTOT was thus greatest in the unobstructed (i.e. control)
form of the model.
Exercise test
Results from the treadmill exercise tests are shown in
Table 2. Exercise at a slow trot produced a V T of 5.4+1.2
litres and a fb of 78_+5.5 min -1. When compared with airway mechanics measured with the control condition of the
oscillating flow model when Vp = 6 litres and fp = 76, peak
~? I and V E were similar for the horses and the model. At
the gallop, V.T was 12.8_+2.3 litres and fb was 102_+2.2
min -1. Peak V I and "Q E associated with these values were
greater in the exercising horses than in the model when Vp
was 12 litres and f~ was 106. For fb = 102 and V T = 12.8
litres we calculatedethat ZkPin was about 38 cmH20 at con-
TABLE 1 : Pump frequency (fp) at which differences (P<0.05) in mechanical
properties due to differences in pump volume were first detected for
each model condition, and the respective values for each variable at that
frequency
Model condition
clear
Rui
(cmH20.s litre -1)
PLH
bronchial
clear
RLI
(crnH20.s litre -1)
PLH
bronchial
clear
RTOTI
(cmH20.s litre -1)
PLH
bronchial
clear
RUE
(cmH20.s litre -1)
PLH
bronchial
clear
RLE
(cmH20.s litre -1)
PLH
bronchial
clear
RTOTE
(cmH20.s litre -1 )
PLH
bronchial
ND No difference detectable
fp
6 litres
12 litres
ND
ND
ND
75
obstruction ND
106
0,005 -+ 0.00 0-017 4- 0.00
ND
ND
0.54 _+0.02 0.70 _+0.05
obstruction
75
64
106
0.45 -+ 0.02
0.63-+ 0.02
0.56 4- 0.02
0.95 -+ 0.02
0.73_+0-01
0-72-- 0.05
obstruction
75
64
83
0.46-+0-02
0-63-+ 0.02
0.05-+0.00
0-96-+ 0.02
0-74-+ 0.01
0.07-+0.01
ND
obstruction 30
106
obstruction
obstruction
ND
ND
0.01 _+0-00 0.02 4- 0.00
0.45_+0-01 0.51 4-0-01
75
20
75
0.24_+ 0-00
0.08_+ 0.02
0.37_+ 0.01
0.71 _+0-02
0.12 -+ 0-01
0.43 _+0-02
75
20
0-26_+0.00
0-08_+0.02
0-73 -+ 0.02
0-14 + 0.01
secutive points of zero flow. Without taking this into
account, the m e a s u r e d Cdyn was negative. Correcting for
the effect of Pin, the value o f Cdyn was 0-34_+0.13 litres
(cmH20) -1, or about 25 per cent of the pre-exercise value.
Total airway resistance was only slightly greater in the
horses than in the model at both the trot and gallop,
although RuI, and RuE were considerably higher in the
horses. Therefore, R L was greater in the model.
DISCUSSION
We recognised that just determining the mechanical
properties of the unobstructed model would lend little to
our understanding of the same properties of the equine respiratory tract, as such measurements would primarily
reflect the output characteristics of the flow generators (i.e.
the electric fans or piston pump). However, the superimposition of the obstructive conditions on the model was of
interest in that it allowed us to differentiate between the
effects of obstruction and those intrinsic to the model-flow
generator system.
The studies of mechanical properties using different rates
of continuous flow through the model indicated that the
external nares and larynx were the major points of airflow
limitation in the upper airways. A previous in vivo investigation reported by Art et al (1988) indicated that extra theTABLE 2: Comparison of values of respiratory mechanics in exercising
horses, and those measured in the model (control condition) at similar
pump frequencies and volumes
Horses;
pre-exercise
VT (litres)
2.3 -+ 0-6
fb (minq)
33 -+ 7.5
peak VI
5.2 _+-*1-2
(litres.s "1)
peak~?E
5.7±1-1
(litre.s -1)
PuI (cmH20)
1.5-+0.5
PUE (cmH20)
1.7_+0.5
RUI
0.20 4- 0.07
(cmH20.s.iitre "1)
RUE
0.06-+ 0.06
(cmH20.s.litre-1)
RTOT
0.75 4- 0.13
(cmH20.s.litrel)
PTOT1 (cmH2Q) 8.0±2-1
PTOTE(CmH20 ) 7.2±2.3
Horses:
4.5 m s -1
Model
Horses;
10 m s-1
Model
5.4 + 1.2
6-0
12.8 _ 2-3 12.0
78 -+ 5.5
76
102 -+2-2
106
29.4 _+7-9 47.6 _+1,4 77.0 __.11.9 63.4 4- 2.9
38.3_+14.846.3_+0.9
83.44-15.2 56-1 4-0.69
6.2_+1.5
4-7_+0.1
7.7-+1.7 5.07-+0.10
5-6_+1.7
2-3_+0-2
9.0-+4.6 5.14_+0.19
0-16 -+ 0.06 0.01 _+0-00 0.15 -+ 0.08 0.09 _+0-01
0-23-+ 0.14 0.05_+0-000.16-+0.13 0.10_+0.00
0-65 _+0.25 0-53 _+'0.01 0.75 -+ 0.29 0.66 _+0.03
25,9+_3.3 24-9+1.1 27-0__.5-1 36.9+3.9
25,94-7.7 17.0_+1.0 30.3-+8.9 31-6-+0.2
209
Equine obstructive upper airway disease
1.4
I
1.4
1.2
I
@ 1.2-
1.0
1.0
m
r~
c= 0.8
£
I
0.8-
0.6
0.6 ~-
i
0.40.2
0
15
,X i
(a)
25
35 45 55 65 75 85 95
Respiratory~equency(minute -1)
105
0.4 L
°:t
10
20
* ~ l
30 40
50 60 70 80
Respiratory frequency (minute-1)
I
90
100
FIG 6: Relationship of total resistance to frequency in a model of the equine respiratory system with oscillating flow volumes of 6 litres (a) and 12 litres (b), and
clear airway, simulated pharyngeal lymphoid hyperplasia (PLH) and partial bronchial obstructive conditions. * Indicates points that are significantly different to
those for the clear airway at the same volume and similar frequency (P<0.05). Key: • clear airway; • PLH model; • bronchial obstruction
racic resistance to airflow constituted approximately 80 per
cent of total pulmonary resistance. They divided the
extrathoracic respiratory tract into two segments representing: (i) the nasal passages, and (ii) the larynx and extrathoracic trachea, and found that the former segment was
always responsible for more than 50 per cent of total
extrathoracic airway resistance. The section of the respiratory tract represented by the model was essentially the
upper airway as defined by Art et al (1988). We corroborate that nasal resistance (the sum of R in the nares,
nasopharynx and pharynx) constituted greater than half the
RTOT in the continuous flow preparation of the model. The
largest component of nasal passage resistance, came from
the external nares. While the larynx and trachea together
were responsible for less than 50 per cent of RTOT, the larynx was still an important single site of flow limitation
(about 29 per cent of RTOT) as suggested by Robinson et al
(1975), although its contribution may have been artificially
high, as arytenoid abduction was less than 100 per cent.
At similar fp and fb there was little discrepancy between
RTOT, for the oscillating flow model and corresponding
values calculated for the five horses performing the exercise test. These figures were also similar to those reported
for ponies by Art and co-workers (Art et al 1988, Art and
Lekeux 1989) and horses (Art et al 1990, Slocombe et al
1992). However, R U was lower in the model than in the
horses at low airflows when V_P was 6 litres. Pouring of the
plaster mould may have resulted in dynamic compression
of nasal and pharyngeal tissues and vasculature creating a
larger diameter upper airway in the model than in horses,
and a lower segmental resistance to airflow in accordance
with Poiseuille's law (Poiseuille 1840). At Vp = 12 litres,
there was no difference between R U for the model and
exercising horses with a V T, and fp and fb of similar magnitude, possibly because of vasoconstrictive and other dilatory responses to increased exercise intensity in the horses'
nasopharynges. This would have resulted in an increase in
their upper airway diameter, while the corresponding diameter in the model remained constant. The resulting similarity in peak V I and V g and recorded airway pressures in the
two situations suggested that the model represented a valid
means of evaluating the effects of partial obstruction of the
upper airway on the mechanical properties of airways during exercise at the gallop. Breathing frequencies and VTS
higher than those measured here have been recorded in
Thoroughbred horses exercising at faster speeds (Art et al
1990), and it is possible that at such intensities the influence of PLH on airway mechanics is greater than that
observed in this study.
We elected to investigate PLH and unilateral bronchial
obstruction with this model because their influence on airway mechanics is unclear. A useful model of laryngeal
obstruction was difficult to produce and the condition has
already been the subject of extensive experimental study.
The model of partial bronchial obstruction was of interest
because it effects on mechanics have not been previously
documented, although it may be commonly associated with
the production of mucopus in the lower airway. Placement
of glass beads in the pharynx to simulate severe PLH resulted in increases in R E, but not in R U, at high f (>90) and V
similar to the V T of galloping horses. ThereFore, while not
obstructing airway flow per se, large pharyngeal nodules
might disrupt laminar airflow patterns in exercising horses,
with the resulting extra turbulence causing an increase in
measured R L. This finding is notable in that it is in contrast
to the widely held, but never substantiated belief that conditions like PLH cause problems by inducing increases in airflow resistance at the site of the lesion (Raker & Boles
1978), but must be interpreted cautiously from a clinical
perspective. Horses may be able to alter the shape or
dimensions of their pharynx during exercise, whereas the
wall of the model was rigid. Consequently, in horses, the
presence of large pharyngeal nodules of similar size to
those in the model may not induce the same changes in R L.
Furthermore, the size of the glass beads was greater than
the nodules seen in most cases of PLH. No studies were
conducted with smaller beads, therefore, it is not known
whether they might have also induced significant alterations in ventilatory mechanics. A previous study using an
in vivo model of PLH found that the condition had no effect
on gas exchange in horses galloping maximally over 1600
m (Bayly et al 1984).
Partial obstruction of a mainstem bronchus resulted in
increased R L at relatively low fp. This was to be expected
given the relationship between resistance and airway radius
(Poiseuille 1840). Narrowing the lumen of a mainstem
bronchus by approximately 50 per cent would theoretically
have resulted in a 16-fold increase in resistance in that airway, and an overall three-fold increase in R L and RTOT.
The actual difference measured was approximately 260 per
cent. The results for the control condition were compatible
with those reported for the larynx plus extrathoracic trachea
• p
210
W. M. Bayly, R. F. Slocombe
by Art et al (1988), again suggesting that the model was
useful in predicting the effects of respiratory tract obstructions on airway mechanics under conditions similar to those
encountered in submaximal exercise We would predict that
any alterations in airway mechanics with submaximal exercise would be exacerbated under maximal and snpramaxireal conditions.
The oscillating flow studies indicated that when the
Drager bags were attached to the model, only 4 to 10 per
cent RTOT was attributable to the airway rostral to the larynx. This finding cannot be related to the in vivo situation
where Pu is approximately 50 per cent of PTP. The failure
of the model to duplicate the in vivo situation was thought
to be because there were major points of resistance at the
'bronchial' attachments to the bags. Also, the bags were
completely collapsed at the end of each ventilatory cycle
with the result that there was no 'end expiratory' or functional residual volume. The pressure-volume behaviour of
the Drager bags was very different to those of normal
lungs, with the result that the bags possessed a high initial
resistance to gas flow and were harder to inflate at low volumes. This meant that while the use of the model of the
upper airways, trachea and mainstem bronchi represented a
valid approach to the study of obstructive diseases of them,
limitations in the mechanical modelling of the lungs invalidated in vitro study of pulmonary mechanics by this means.
Consequently, although the oscillating flow model was
originally developed with the aim of studying changes in
C dyll
. _ under
conditions of different f~v and V_,
calculated val.
v .
ues mainly reflected the Cdxn of the rebreathmg bags, due
to the rigidity of the rest o t the model, and hence had no
relevance to the in vivo situation in which all respiratory
tract tissues possess some degree of compliance.
With the oscillating flow system, RTOT of the model
increased with rises in fp and Vp. This was not surprising
given the sinusoidal nature of the flow wave, the fact that
resistance is flow dependent (Rohrer 1915), and the well
established link between flow, tidal volume and frequency
(~r = [V .co.cos(cot)]/2, where co = 2nL; Pedley and Drazen
1986, RoPdarte and Rehder 1986). Thev.increase in RTOT is
also associated with greater airflow turbulence at high frequencies and airflows, as has been previously demonstrated
in horses (Slocombe et ai 1992). With respect to the exercising horse, this is particularly important because of the
effect of increased resistance (i.e. increased fb and V T) on
respiratory system impedance (IZl), as reflected by the
equation: IZl = "\/[R2+(coI-1/coC)2], where co = 27tf, I = inertance and C = Cdyn. It is the impedance that the horse must
work against to move air in and out of its respiratory system. As IZI rises, horses must progressively increase muscular work of breathing to overcome this mounting
mechanical force in order to breathe faster and deeper. It
has been suggested that horses may not be able to completely overcome such influences as these, or else may elect
to limit ventilatory effort in favour of locomotor muscle
work, with the resultant development of exercise-induced
hypoxaemia and hypercapnia (Art and Lekeux 1989, Bayly
et al 1989). This remains to be definitively investigated in
horses exercising at high speeds.
In horses, moderate to heavy exercise is characterised by
a high frequency, high volume ventilatory response
(Hornicke et al 1987, Bayly et al 1987). Until recently there
had been no determination of airway mechanical properties
at the fb (>100), flow rates (>70 litres sec-1), and tidal volumes (>12 litres) believed to be associated with high intensity exercise. This is partly because such studies require the
assembly of equipment especially developed for such
recording purposes. Such equipment needs to be accurate
over wide ranges to ensure veracity of results at all exercise
levels encountered between the resting state and maximum
speed. Art et al (1989) utilised two ultrasonic flow transducers (developed by Woakes et al [1987]) and an
oesophageal balloon catheter to generate the first measurements of mechanics of breathing during strenuous exercise
in horses. Critical to the determination of respiratory
mechanics during exercise is the accuracy of the pneumotachographs. In this study they were linear over the range of
15 to 70 litres sec -1. While horses galloping at 10 m sec -1
produced peaks ~? I and V E greater than this rate, placing
two such pneumotachographs in parallel allowed accurate
measurement of airflow without approaching their range
limit. Early experimentation indicated that pneumotachographs of smaller diameter were inaccurate in the upper
parts of the anticipated flow range. It must also be recognised that the pressures recorded at each site in the model
preparations reflected those at the lateral part of the
airstream rather its middle. Similar lateral placement of
catheters for pressure measurements has been undertaken in
other studies (Derksen et al 1986, Art et al 1988) and differences, if any, between lateral and midstream static pressures have not been reported. Theoretically, they should not
differ.
As the flow rates and frequencies measured with the
oscillating flow preparation of the model and with the
treadmill trial were similar to those previously reported
(Art et al 1990, Fulton et al 1991, Slocombe et al 1992), the
results obtained are of interest to the extent that they
demonstrate that a model such as that described here can be
of use in studying the effects of airway obstruction during
exercise. The continued use of these methods in future
studies of responses of the equine ventilatory system to
exercise would be enhanced by the development of a better
model of the lungs than the anaesthetic rebreathing bags
used here, and by creating a model with properties that
reproduce airflows, frequencies, and volumes commensurate with those seen at the highest exercise intensities.
ACKNOWLEDGEMENTS
The authors thank the Australian Equine Research Fund
and the Washington State Equine Research Program for
their funding support, and Lindsay Slocombe, Anna Platt
and Meredith Wilson for their technical assistance.
REFERENCES
ART, T., ANDERSON, L., WOAKES, A.J., ROBERTS. C., BUTLER, P. J., SNOW,
D. H. & LEKEUX, P. (1990) Mechanics of breathing during strenuous exercise
in thoroughbred horses. Respiration Physiology 82, 279-284
ART, T. & LEKEUX, P. (1988a) A critical assessment of pulmonary function testing
in exercising ponies. Veterinary Research Communications 12, 25-39
ART, T. & LEKEUX, P. (1988b) Pulmonary mechanics during treadmill exercise in
race ponies. Veterinary Research Communications 12, 245-258
ART, T. & LEKEUX, P. (1988c) Respiratory airflow pattern in ponies at rest and
during exercise. Canadian Journal of Veterinary Research 52, 299-303
ART, T. & LEKEUX, P. (1989) Work of breathing in exercising ponies. Research in
Veterinary Science 46, 49-53
ART, T., SERTYN, D. & LEKEUX, P. (1988) Effect of exercise on the
partitioning of equine respiratory resistance. Equine Veterinary Journal 20,
268-273
BAYLY, W. M., GRANT, B. D. & BREEZE, R. G. (1984) Arterial blood gas tension and acid-base balance doing exercise in hbrses with pharyngeal lymphoid
hyperplasia. Equine Veterinary Journal 16, 435-438
BAYLY, W. M., HODGSON, D. R., SCHULZ, D. A., DEMPSEY, J. A. &
GOLLNICK, P. D. (1989) Exercise-induced hypercapnia in the horse. Journal of
Applied Physiology 67, 1958-1966
Equine obstructive upper airway disease
BAYLY, W. M., SCHULZ, D. A., HODGSON, D. R. & GOLLNICK, P. D. (1987)
Ventilatory responses to exercise in horses with exercise-induced hypoxemia. In
Equine Exercise Physiology II. Eds J. R. Gillespie & N. E. Robinson. Davis, CA,
ICEEP Publications. pp 172-182
BELKNAP, J. K., DERKSEN, F. J., NICKELS, F. A., STICK, J. A. & ROBINSON,
N. E. (1990) Failure of subtotal arytenoidectumy to improve upper airway flow
mechanics in exercising Standardbreds with induced laryngeal hemiplegia.
American Journal of Veterinary Research 51, 1481-1487
DERKSEN, F. J., STICK, I. A., SCOTT, E. A., ROBINSON, N. E. & SLOCOMBE
R. F. (1986) Effect of laryngeal hemiplegia and laryngoplasty on airway flow
mechanics in exercising horses. American Journal of Veterinary Research 47,
16-20
FULTON, I. C., DERKSEN, F. J., STICK, J. A., ROBINSON, N. E. & WALSHAW,
R. (1991) Treatment of left laryngeal hemiplegia in Standardbreds using a nerve
pedicle graft. American Journal of Veterinary Research 52, 1461-1467
GILLESPIE, J. R., TYLER, W. S. & EBERLY, V. E. (1966) Pulmonary ventilation
and resistance in emphysematous and control horses. Journal of Applied
Physiology 21, 416-422
HORNICKE, H., WEBER, M. & SCHWEIKER, W. (1987) Pulmonary ventilation in
Thoroughbred horses at maximum performance. In Equine Exercise Physiology
II. Eds J. R. Gillespie & N. E. Robinson. Davis, ICEEP Publications. pp 216-224
LUMSDEN, J. M., DERKSEN, F. J., STICK, J. A. & ROBINSON, N. E. (1993) Use
of flow-volume loops to evaluate upper airway obstruction in exercising horses.
American Journal of Veterinary Research 54, 766-775
MEAD, J. & WHITTENBERGER, J. L. (1953) Physical properties of human lungs
measured during spontaneous respiration. Journal of Applied Physiology 5,
779-796
PEDLEY, T. L & DRAZEN, J. M. (1986) Aerodynamic theory. In Handbook of
Physiology, Section 3: The Respiratory System, Volume 3. Eds P. T. Macklem
& J. Mead. Bethesda, American Physiological Society. pp 41-54
POISEUILLE, J. L. M. (1840) Recherches experimentales sur le mouvement des liquides dans les tubes de trbs petits diam~tres. Comptes Rendus de l'Academie des
Sciences (Paris) 11, 961-967, 1041-1048
211
RAKER, C. W. & BOLES, C. L. (1978) Pharyngeal lymphoid hyperplasia in the
horse. Journal of Equine Medicine & Surgery 2, 202-207
RODARTE, J. R. & REHDER, K. (1986) Dynamics of respiration. Handbook of
Physiology, Section 3: The Respiratory System, Volume 3. Eds P. T. Macklem
& J. Mead. Bethesda, American Physiological Society. pp 131-144
ROBINSON, N. E., SORENSON, P. R. & GOBLE, D. O. (1975) Patterns of airflow
in normal horses and horses with respiratory disease. Proceedings 21st Ann Mtg
Amer Assoc Equine Practnrs pp 11-21
ROHRER, F. (1915) Der Strtmungswiderstand in den menschlichen Atemwegen
nnd der Einfluss tier unregelm~issigen Verzweigung des Bronchialsystems anf
den Atmnngsverlanf in verscheidenen Lungenbezirken. Pflugers Archivs
Gesamte Physiologie Menschen Tiere 162, 255-299
SHAPPELL, K. K., DERKSEN, F. J., STICK, J. A. & ROBINSON, N. E. (1988)
Effects of ventriculectomy, prosthetic laryngoplasty, and exercise on upper airway function in horses with induced left laryngeal hemiplegia. American Journal
of Veterinary Research 49, 1760-1765
SLOCOMBE, R. F., COVELLI, G. & BAYLY, W. M. (1992) Respiratory mechanics
of horses during stepwise treadmill exercise tests, and the effect of clenbuterol
pretreatment on them. Australian Veterinary Journal 69, 221-225
WILLIAMS, J. W., MEAGHER, D. M., PASCOE, J. R. & HORNOF, W. J. (1990)
Upper airway function during maximal exercise in horses with obstructive upper
airway lesions: effect of surgical treatment. Veterinary Surgery 19, 142-147
WILLOUGHBY, R. A. & McDONNELL, W. N. (1979) Pulmonary function testing in
horses. Veterinary Clinics of North America: Large Animal Practice 1, 171-196
WOAKES, A. J., BUTLER, P. J. & SNOW, D. H. (1987) The measurement of respiratory airflow in exercising horses. Equine Exercise Physiology II. Eds J. R.
Gillespie & N. E. Robinson. Davis. ICEEP Publications. pp 194-205
YOUNG, S. S. & TESAROWSKI, D. (1994) Respiratory mechanics of horses measured by conventional and forced oscillation techniques. Journal of Applied
Physiology 76, 2467-2472
Received December 15, 1995
Accepted October 8, 1996