Evaluation of a mask for breath-by-breath respirometry during exercise in horses


email: wmb@wsu.edu


Reasons for performing study: The ability to obtain breath-by-breath measures of ventilatory mechanics for the entirety of an exercise test, regardless of speed(s) or duration enables evaluations of equine ventilation during exercise that are necessary for assessments of performance.

Objective: Evaluation of a new ergospirometer (Quadflow; QF) system's accuracy and repeatability for measuring pulmonary variables in contrast to the established pneumotachometer-based system (control) and assessment of its effects, if any, on exercise capacity at high speeds.

Materials and methods: Five Thoroughbred horses each performed 10 incremental exercise tests to fatigue, 5 with the QF system and 5 with an open-circuit flow system. Measures of pulmonary variables were evaluated to determine repeatability. Heart rate, pulmonary variables, arterial blood gases, distance run and time to fatigue measured with each system were compared to assess similarity of results and effect on performance.

Results: Results from both systems had high repeatability with low coefficients of variation. The QF was associated with greater resistance to airflow, higher breathing rate at submaximal speeds, lower minute ventilation and peak inspiratory and expiratory airflows, greater acidaemia, hypoxaemia and hypercapnoea, and decreased total run time and total distance run when compared to control system results.

Conclusion: The greater resistance of the QF was responsible for altered blood gases, respiratory parameters and performance when compared to the control mask. The QF system reliably measured equine pulmonary airflows and volumes and is suitable for research and clinical use provided optimal gas exchange and best possible physical performance are not required.


The ability to evaluate ventilatory mechanics of equine athletes is an important tool for practitioners as well as clients and researchers. Evaluation of these parameters may help to determine causes of poor performance, or other respiratory limitations of individual horses. The current method for evaluating ventilatory variables in a scientific context and on a clinical level at many institutions is through an open or biased flow system on a high speed equine treadmill (Bayly et al. 1989; Art et al. 1990; Art and Lekeux 1993; Erickson et al. 1994; Langsetsmo et al. 1997; Hopkins et al. 1998).

The biased flow system that we have used in our laboratory for many years allows for measurements of ventilatory variables such as tidal volume (VT), peak inspiratory (VI) and expiratory (VE) airflows, breathing frequency (fb), and minute ventilation (Vmin). Although not the case with all biased flow systems (Langsetsmo et al. 1997; Padilla et al. 2004), our system requires temporary closure of the airflow outlets on the mask that the horse wears, briefly occluding the biased air flow through the system (Bayly et al. 1999a; Katz et al. 1999). This modification occurs while the horse is exercising, causing the horse to re-breathe some of the exhaled air that remains in the mask. Consequently, these measurements can only be made for a limited time (usually <10 s) in order to avoid equipment-induced aberrations in blood gases and ventilation. This configuration limits the breath-by-breath determinations of ventilatory mechanics as they cannot be obtained for the whole exercise period.

The ability to obtain breath-by-breath measures of ventilatory variables for the entirety of an exercise test, regardless of speed(s) or duration will enable evaluations of equine ventilation during exercise necessary for assessment of performance in both research and clinical settings. The recently described quadflow (QF) system utilises pitot technology to determine pulmonary variables, eliminating the need to incorporate valves and/or a biased flow evacuation system into the mask (Curtis et al. 2005, 2006; Kusano et al. 2007). If it can be shown that the results generated with this new type of mask are no different to those produced with the existing or traditional mask systems, then a new dimension of equine exercise research and clinical testing becomes feasible. It was the objective of this study to compare blood gases and basic measures of ventilatory variables generated with both the traditional biased flow and the newly described QF system at varying intensities of exercise, and to assess the repeatability of the measures of ventilation.

Materials and methods

The Animal Care and Use Committee at our university approved this study.


Five, fit Thoroughbred horses (geldings aged 8–13 years) previously acclimated to a high speed equine treadmill through a combination of incremental exercise tests and 30 min intervals 4 m/s 5 days a week for 2 months prior to this study were used. All horses were examined and determined to be sound and free of respiratory disease.

Experimental protocol

Each horse followed the same protocol, wearing the traditional open flow (OF) system (which served as the control system) 5 times and the QF mask 5 times, for a total of 10 tested runs, with at least one day between each series of test runs. On a given test day, one test was conducted with each of the mask systems. The order in which the masks were worn on a test day varied in a random fashion. This allowed repeatability of measurements for the QF system to be compared with that of the established mask, as well as enabling comparison of actual results generated with each system.

On exercise test days, each horse completed one incremental exercise test with each of the 2 different mask systems, with a minimum of 4 h between each test. During this time horses were allowed access to alfalfa for the first hour and water for all 4 h. All tests were run between 07.00 and 15.00 h. The treadmill was set at a 7% incline for all tests. The protocol for data collection was as follows: immediately upon placement of either face mask each horse began to exercise with a warm-up of 4 min at 4 m/s. The speed was then increased to 6 m/s for 1 min and continued to increase in speed by 1 m/s each minute until the animal would no longer continue, despite enthusiastic vocal encouragement. Once the animal had finished the test, the treadmill was stopped and the mask removed.

Prior to beginning each exercise test, each horse was assessed for lameness and determined to be sound. For 3 horses whose blood was to be collected for arterial blood gas analysis, an introducer was placed in the right jugular vein and a Swan-Ganz thermodilution catheter was passed through the introducer catheter into the right ventricle for blood temperature measurements. A 5 cm, 18 gauge catheter was placed in a previously translocated carotid loop on the left side for collection of arterial samples.

Ventilatory variables were measured on all 5 horses during the entirety of the run with the QF system and for 10 s windows during the last 15 s of each speed with the control (OF) system. Tidal volume, Vmin, fb, VI, VE and the ratio of inspiratory time to total breath time (tI/ttot) were determined using a customised software analysis programme (Katz et al. 1999; Bayly et al. 1999b) for the control system and QF software version 1.4 (Curtis et al. 2005, 2006; Kusano et al. 2007) for the QF system. Calibration of the QF was performed per manufacturer's instructions with an orifice plate provided by the manufacturer (Curtis et al. 2005). Quad flow internal markers included barometric pressure, ambient temperature, relative humidity and an orifice plate pressure (410–430 Pa advised) to be applied from any flow driven system. Calibration of the OF system was performed before exercise using previously described methods (Katz et al. 1999; Bayly et al. 1999a). Transducers were calibrated from measured open system driven flow rates (TSI 8360 hot-film anenometer)1 up to 92 l/s. All flow calibrations and software integrations were verified with a customised 12.35 l syringe (Kastner et al. 2000; Katz et al. 2006) at rates up to 86 strokes/min and 65 l/s ATPS peak flows. Heart rates for exercise with each mask were recorded during the last 15 s at each speed using a Polar cardiotachometer2. Arterial blood samples were collected into heparinised syringes for determination of blood gases from the 3 horses with carotid loops. The other 2 horses did not have carotid loops and were not utilised for this procedure. Blood samples were collected at rest and 45 s after exercise at 4, 7, 9, 11 and 12 m/s, respectively and immediately placed on ice. Blood temperature was recorded when samples were collected. Temperature-corrected blood gases were determined with an automated blood gas analyser3.

Subsequent to assessing the effects of the QF on gas exchange and because of those results, a separate follow-up experiment was performed in order to assess the effect of mask dead space on gas exchange with the QF mask. Three horses from the previous study (geldings aged 9–14 years) were used as well as one new horse (a mare, age 9 years). All 4 horses had left-sided carotid loops. Padding was packed in the open area between the surface of the mask and nose and mandibles of the horse to eliminate ∼1.9 l of the dead space. The residual dead space of 1.1 l around the transducers plus that in front of and immediately surrounding the nares (estimated total ∼2.5 l) could not be removed without changing the design of the mask. All 4 horses performed incremental treadmill exercise tests to fatigue. Each horse performed the test twice on a single day with at least 4 h between each test. On one occasion the QF was worn in its original configuration and on the other with the reduced dead space. The order of runs was randomised. Arterial samples were collected and blood temperatures recorded at 4 and 6–12 m/s and analysed as described above.

Statistical analysis

Six to 8 consecutive normal breaths were analysed during the last 15 s of exercise at each speed, for each mask system. Results were expressed as the mean ± standard error of the means (s.e.). Differences between results using the QF and control systems were evaluated with a 3 way repeated measures analysis of variance (ANOVA) with repeated factors of speed, run order and mask system. Repeatability was assessed by determining the coefficient of variation from the standard deviation of the means of the different replicates divided by the mean across all replicated conditions. Differences between the original QF mask and that with reduced dead space were also evaluated with a 2 way repeated measures ANOVA for speed and dead space volume. Run time and distance were analysed with a 2 way repeated measures ANOVA with run order and mask system as the repeated factors. Individual means were compared with the Bonferroni post hoc test when the F statistic was significant. A P value of <0.05 was considered significant for all results.


Wearing the QF was associated with greater acidaemia (P = 0.006), hypoxaemia (P = 0.014) and hypercapnoea (P = 0.005). Percent saturation of haemoglobin with oxygen (O2sat.) was significantly lower for the QF (P = 0.002). Blood gases, blood temperature and heart rate results are presented in Table 1. Temperatures did not differ with the mask system. Although heart rate while wearing the QF appeared higher when compared to rates with the OF, these differences only approached significance (P = 0.087).

Table 1. Heart rate (HR), blood gases and temperature in horses completing an incremental exercise test while wearing either an open flow-through gas collection mask (OF) or the Quadflow ergospirometer (QF)
Speed (m/s)SystemHR (beats/min)PaO2 (Torr)PaCO2 (Torr)O2sat (%)Blood temp (°C)
  • *

    Values are expressed as mean ± s.e; n = 5 horses. Blood gases (PaO2, PaCO2) were only collected at rest and speeds of 4, 7, 9, 11 and 12 m/s. O2 sat = saturation of haemoglobin with oxygen. denotes results different at that speed for the mask systems (P<0.05).

 0OF37.0 ± 0.9100.1 ± 2.340.9 ± 0.898.3 ± 0.137.5 ± 0.07
QF42.6 ± 4.597.5 ± 1.941.9 ± 1.298.1 ± 0.137.5 ± 0.08
 4OF143.1 ± 4.0108.4 ± 1.736.3 ± 0.6*99.1 ± 0.0938.2 ± 0.06
QF144.6 ± 4.6101.1 ± 1.040.8 ± 0.6*98.5 ± 0.0738.3 ± 0.07
 7OF171.4 ± 3.0100.0 ± 1.7*36.5 ± 0.7*98.4 ± 0.1*38.8 ± 0.05
QF176.2 ± 3.283.5 ± 1.0*46.4 ± 0.7*95.7 ± 0.2*38.8 ± 0.06
 9OF195.1 ± 2.790.8 ± 1.2*40.0 ± 0.5*96.3 ± 0.2*39.8 ± 0.07
QF198.5 ± 2.070.4 ± 1.9*54.1 ± 0.1*89.3 ± 0.8*39.9 ± 0.1
11OF215.7 ± 1.682.1 ± 1.2*46.4 ± 0.7*91.1 ± 0.5*41.3 ± 0.1
QF216.3 ± 1.769.4 ± 1.2*63.5 ± 1.3*81.6 ± 0.8*41.4 ± 0.1
12OF220.5 ± 2.180.0 ± 0.6*49.5 ± 0.7*86.8 ± 0.3*42.1 ± 0.08
QF219.3 ± 1.867.5 ± 3.0*68.5 ± 1.6*77.3 ± 0.5*42.2 ± 0.1

Wearing the QF had a significant effect on all measures of ventilation except VT (P = 0.091; fb: P = 0.037; Vmin: P = 0.004; VI: P = 0.041; VE: P = 0.038). However, VT was greater at 4 m/s but lower at 10, 11 and 12 m/s with the QF (Fig 1). Rate of breathing was slower with the QF at the slower exercise speeds (4, 6, 7 and 8 m/s; Table 2). Minute ventilation was lower with the QF at speeds of 10 m/s and faster (Fig 2). The highest mean VI with the QF was 84.2 ± 1.4 l/s, reached at 11 m/s while with the control mask it was 86.4 ± 1.9 at 12 m/s (Fig 3). For the QF, the highest VE was reached at 11 m/s (97.5 ± 1.7) l/s whereas it was 98.3 ± 1.98 l/s at 12 m/s with the control mask (Fig 4). Mask system had no overall effect on tI, tI/ttot or tE despite the differences in fb (Table 2). Total run time (P = 0.002) and distance (P = 0.001) were reduced with the QF (Table 3).

Figure 1.

Tidal volume (VT) for 5 sets of incremental exercise tests for 5 Thoroughbreds wearing either the open flow or Quadflow facemask systems. Values are expressed as mean±s.e. Speed is expressed m/s.*denotes results that were different for the 2 mask systems at a given speed (P<0.05).

Table 2. Breathing frequency (fb), inspiratory (tI) and expiratory times (tE) and ratio of tI to total breath duration (tI/ttot) for 5 sets of incremental exercise tests for 5 Thoroughbreds while wearing either the open flow (OF) or Quadflow (QF) facemask systems
Speed (m/s)Systemfb (/min)tI/ttottI (ms)tE (ms)
  • *

    Values are expressed as mean ± s.e. denotes results that were different for the 2 mask systems at a given speed (P<0.05).

 4OF77.0 ± 0.23*52.4 ± 1.38*418.7 ± 0.013*379.1 ± 0.022*
QF68.3 ± 2.43*50.1 ± 0.56*451.6 ± 0.012*448.5 ± 0.018*
 6OF107.5 ± 0.95*48.2 ± 0.49271.4 ± 0.003291.3 ± 0.004
QF100.9 ± 2.04*49.4 ± 0.47303.1 ± 0.010310.1 ± 0.006
 7OF108.4 ± 0.75*47.7 ± 0.30266.4 ± 0.002291.3 ± 0.003
QF104.4 ± 1.27*48.9 ± 0.33284.1 ± 0.006296.4 ± 0.005
 8OF110.6 ± 0.65*47.9 ± 0.30262.2 ± 0.002284.9 ± 0.002
QF106.8 ± 1.25*48.8 ± 0.31277.2 ± 0.006290.7 ± 0.004
 9OF112.5 ± 0.4647.7 ± 0.30257.4 ± 0.002281.3 ± 0.002
QF109.6 ± 1.2148.5 ± 0.27267.9 ± 0.006284.5 ± 0.004
10OF114.0 ± 0.3947.2 ± 0.32251.1 ± 0.002280.0 ± 0.002
QF112.2 ± 0.5848.4 ± 0.33260.0 ± 0.003277.0 ± 0.002
11OF116.4 ± 0.4347.3 ± 0.33246.0 ± 0.002274.0 ± 0.002
QF114.2 ± 0.4448.1 ± 0.33253.7 ± 0.003273.1 ± 0.003
12OF118.9 ± 0.6346.9 ± 0.38239.3 ± 0.002270.1 ± 0.002
QF117.8 ± 0.7248.0 ± 0.37245.1 ± 0.002265.0 ± 0.003
Figure 2.

Minute ventilation (Vmin) for 5 sets of incremental exercise tests for 5 Thoroughbreds wearing either the open flow or Quadflow facemask systems. Values are expressed as mean±s.e. Speed is expressed in m/s.*denotes results different for the 2 mask systems at a given speed (P<0.05).

Figure 3.

Peak inspiratory flow (PIF) for 5 sets of incremental exercise tests for 5 Thoroughbreds wearing either the open flow or Quadflow facemask systems. Values are expressed as mean±s.e. Speed is expressed in m/s.*denotes results different for the 2 mask systems at a given speed (P<0.05).

Figure 4.

Peak expiratory flow (PEF) for 5 sets of incremental exercise tests for 5 Thoroughbreds wearing either the open flow or Quadflow facemask systems. Values are expressed as mean±s.e. Speed is expressed in m/s.*denotes results different for the 2 mask systems at a given speed (P<0.05).

Table 3. Run time (seconds; [s]) and distance run (metres; [m]) for 5 Thoroughbred horses for each of 5 consecutive incremental exercise tests (IET) while wearing either the open flow (OF) or Quadflow (QF) mask systems
IETSystemRun time (s)Distance (m)
  • *

    Values are expressed as mean ± s.e. denotes results that were different (P<0.05) for the QF and OF systems for the specified IET.

1OF650.4 ± 13.8*4631.6 ± 169.4*
QF622.6 ± 14.5*4305.0 ± 183.7*
2OF671.2 ± 11.9*4890.0 ± 152.5*
QF634.6 ± 15.0*4440.6 ± 184.2*
3OF696.0 ± 10.7*5052.6 ± 141.8*
QF659.2 ± 17.4*4599.0 ± 288.9*
4OF681.2 ± 10.1*5015.6 ± 131.0*
QF645.0 ± 23.3*4790.2 ± 150.7*
5OF684.6 ± 10.7*4955.0 ± 140.2*
QF675.0 ± 17.2*4660.4 ± 206.6*

Results had high repeatability with coefficients of variation ranging from 6.7% for VE to 8.4% for VI with a value of less than 10% considered acceptable. The order in which the runs were conducted on each day had no effect on any results. Reducing QF dead space had no effect on blood gases at any speed (Fig 5). Quadflow airflow resistance was higher (ranging from 0.0424 cm H2O/l/s at flows of 25 l/s to 0.1425 cm H2O/l/s when the airflow was 90 l/s) than for the control mask (0.016 cm H2O/l/s while the biased flow outlets were closed and 0.009 cm H2O/l/s when the outlets were open).

Figure 5.

Blood gas data for the separate follow up experiment in which 4 Thoroughbreds performed 2 incremental exercise tests each, once wearing the QF mask in the original configuration and once with a 1.9 l reduction in dead space. QFA represents PaO2 for the original mask, QFB represents PaCO2 for the original mask; QFC represents PaO2 with the dead space reduced; QFD represents PaCO2 with the dead space reduced. Values are expressed as mean±s.e. Speed is expressed in m/s. Results at a given speed were not significantly different (P>0.05).


It has been established that facemasks worn to collect expired gases and assess pulmonary mechanics of exercising horses can considerably impair gas exchange during exercise (Bayly et al. 1987, 1989) if the aperture(s) through which they breathe is/are too small. The total surface area of the tubes that make up the QF apertures was about 50 cm2, similar to that which was previously shown to compromise gas exchange in horses exercising on a treadmill at high intensity (Bayly et al. 1987). The use of a biased flow system is common in laboratory settings in order to minimise these effects. Differences between mask systems can vary in their effect on gas exchange as well as the overall performance of the horse and consideration of these effects is important in the development of new mask systems, especially those that might be suitable for field use.

Results for both the QF and OF mask systems had the high repeatability which is essential for any system contemplated for application in laboratory or field settings. The fact that the QF can be used for breath-by-breath analysis of ventilatory variables for the entirety of an exercise bout, regardless of its duration and does not require access to an electrical supply as it is battery powered, makes it potentially ideal for field testing. As such, the QF represents a major advance in the spirometric evaluation of exercising horses, providing that the capacity to perform at their physiological peaks is not part of the assessment, as the results of this study leave little doubt that horses wearing the QF cannot run as far or fast as those using an OF system for analysis of expired gas. As with any system or measurement technique the comparative value of the method (in this case the QF) is dependent upon the existence of a sizeable bank of data from horses exercising under identical conditions. If such a databank could be generated for the QF then its use will represent some advances in the ease with which ventilation can be assessed in exercising horses provided optimal gas exchange and best possible physical performance are not required. The use of a face mask like the QF for spirometry will be associated with increased resistance to breathing and some adverse effect on blood gases when compared to the ‘ideal’ or OF mask situation.

The QF system resulted in higher PCO2 at all speeds, while exhibiting decreased PO2, %O2sat and pH when compared to control system results for speeds of 7 m/s and faster. Each of these findings is indicative of suboptimal gas exchange due to inadequate ventilation. At near maximal speeds (10 m/s and faster), Vmin with the QF was significantly less than with the control system. The origins of these differences would appear to be the lower peak VI and VE that were observed at 10 m/s and faster and, in turn, were manifest as lower VT at the same speeds. These changes were reflected in the horse's performance by a decreased total run time and decreased total distance run. These results compare with those reported by Curtis et al. (2006) in which the QF was shown to result in significantly elevated blood lactate concentrations when compared to an OF system during submaximal exercise. Lactate concentrations were not determined in the present study, but the blood gas, pH and performance data suggest that, had they been, the results would have been the same as those of Curtis et al. (2006). Curtis et al. (2005) suggested that the higher HR and blood lactate concentrations during submaximal exercise with the QF mask could be due to the weight of the QF mask (0.6 kg) as compared to that of an OF mask. Our OF mask weighs considerably more than the QF (6.3 kg) and, although it is suspended from the ceiling to reduce the dead weight of the mask on the horse, the difference in the weights make it unlikely that apparent reduction in performance associated with the QF is due to its weight.

The possibility that the blood gases would be altered with the QF system due to re-breathing of mask dead space air could not be ruled out initially but was not substantiated by the results of the follow-up study. Reducing the dead space of the QF mask by 1.9 l had no effect on HR, pH or blood gases at any speed. Kastner et al. (2000) suggested that resistance to airflow has a larger influence on rebreathing than the shape or size of the mask and associated dead space per se. It has also been shown that the total volume of dead space within a facemask is comprised of a static or noncontributory volume and a functional dead space volume. This functional dead space is the portion of the total dead space within the mask from which the horses actually re-breathes air (Art and Lekeux 1988). With smaller masks the static volume portion is larger than that of the functional dead space volume. Therefore, it is possible that reducing the dead space by 1.9 l had no real impact on the functional dead space and that the volume of the pitot tubes (∼1.1 l) plus that around and in front of the nares was sufficient, when combined with the high airflow resistance of the pitot tubes, to significantly interfere with ventilation and gas exchange. Due to the lack of any assisted evacuation of exhaled air from the QF system, we estimate that at the fastest speeds, 10–15% of each tidal breath was from the functional dead space. The FIO2 and FICO2 of this air would have approximated that of end-tidal expiration and as such would have been very different to that associated with inspiration of ambient air.

The QF system had a resistance of 0.0424 cmH2O/l/s at a flow rate of 25 l/s (Curtis et al. 2005) and we found it to be considerably higher at higher flows with a resistance of 0.1425 cmH2O/l/s at flow rates of 90 l/s. The control system had a resistance of 0.016 cmH2O/l/s when the biased flow outlets were closed, which occurred for 10 s or less at each speed, and 0.009 cmH2O/l/s with the outlets open. Chapman et al. (1998) found that human athletes experiencing expiratory flow limitations because of increased resistance were unable to increase their ventilation during maximal exercise, as compared to athletes with no expiratory flow limitations, and we postulate that the same thing happened in this investigation. The QF has 2 resistors, each with a functional internal diameter of 56 mm which is much less than that of the OF system which consist of 2, each with an internal diameter of 160 mm while in closed configuration (Katz et al. 2006). As a result, the surface area through which air flowed with the control system was approximately 8 times that of the QF. The blower used in our OF system was unable to achieve higher flow rates than 90 l/s through the QF pitot tubes, despite its ability to generate 140 l/s in the OF configuration. If it was possible to produce a QF with a similar air exchange area to that of the OF, the associated reductions in gas exchange and performance that we observed might be eliminated.

Geor et al. (1995) found that at submaximal speeds their closed system mask produced a lower fb than that of an open system. This finding was attributed to an increased work of breathing caused by mask-related increased resistance to airflow and/or increased mask dead space. Our study found that fb was lower with the QF at each speed up to 8 m/s suggesting that horses needed more time to inspire the same tidal volumes at these speeds (Fig 1) and that they were able to alter their breathing patterns to take this time. At near maximal speed (9 m/s) and faster, fb was the same with each mask system and VT was greater with the OF, presumably because horses wearing the QF were not able to control breathing to the same extent to compensate for lower peak VI and VE at these higher treadmill speeds. With lower fb it would be expected that stride frequency would also be lower, although stride frequency was not evaluated in this study. Under these conditions, horses would maintain speed by increasing stride length as has been demonstrated previously (Geor et al. 1995). However, as exercise intensity increased this compensatory ability with stride length would ultimately disappear and there would be no difference in fb, as was the case in our study. Flow limitations due to higher resistance secondary to the smaller surface area of the pitot tubes resulted in smaller VT with the QF at these high intensities. The Vmin was lower for the QF at speeds of 9 m/s and faster and this was reflected in the severity of the oxyhaemoglobin desaturation and hypercapnoea at these speeds. This version of the QF was not capable of measuring oxygen consumption but it can be reasonably assumed that its peak value was lower with the QF and that the horses had to increase their anaerobic contribution to energy production in order to maintain their speed. The inevitable consequence is that fatigue develops sooner and the horse is unable to maintain its speed. The ultimate manifestation of this is decreased run time and distance run as was observed in this study.

The highest peak VE which the horses could generate was less with the QF. Consequently, although there was no difference in the absolute VE at speeds of 6–10 m/s, the horses were actually breathing at a higher relative per cent of their greatest peak VE when wearing the QF. For example, at 6 m/s the mean peak expiratory flows were 75.2 and 66.8% of the highest peak VE recorded for the QF and OF, respectively. At 8 m/s, the respective figures were 89.4 and 82.8% of the highest attainable peak VE. With the QF, maximum peak VE was reached at 11 m/s while horses wearing the OF mask continued to increase their VE up to 12 m/s. With the QF at 12 m/s, VE was lower than at 11 m/s possibly due to respiratory muscle fatigue from working harder against the greater resistance of the QF mask.

It was concluded that the greater resistance of the QF mask was responsible for the altered blood gases, ventilatory parameters and decrease in performance as compared to the control mask. An over-arching issue that equine exercise scientists and practitioners of equine sports medicine must address is whether or not conduct of accurate and reliable pulmonary function tests during exercise should only be conducted using spirometric face masks that do not impose any extra resistance or functional dead space. The benefit of this study is that the effect of the QF mask (with 8 pitot tubes) on blood gases is now quantified, pulmonary function tests with QF or similar technology should be conducted accordingly. Blood gases should ideally be measured separately to the spirometric measurements if high intensity exercise is used as the hyperpnoeic stimulus. If refinement of the technology was to be considered, perhaps adding more pitot tubes would be helpful for studies during high intensity exercise. However, in view of the differences in surface area through which air flows, it is unlikely that sufficient pitot tubes could be added in order to completely obviate the possibility of some imposed alveolar hypoventilation. Despite this and given that the QF system reliably measured equine spirometric indices for the entire exercise test, the QF could be suitable for use provided optimal gas exchange and best possible performance are not required.


This work was supported in part by NCRR Grant RR07049, Students in Health Professional Schools and the Washington State College of Veterinary Medicine Equine Research Programme. The technical assistance of Abbey Burgess, Anna George, Ashley Culp, Belinda Buchholz, Ben Crosland, Eric Renner, Heidi Talbott, Jack Fillerup, Marni Hamack, Megan Kinner, Mike Clayton, Nicole Meyer, Kerry Quarry and Tess Young is greatly appreciated.

Manufacturers' addresses

1 TSI Incorporated, Shoreview, Minnesota, USA.

2 Polar Electro Oy, Kempele, Finland.

3 Nova Biomedical 200 Prospect St. Waltham, Massachusetts 02454, USA.