Veterinary Research Communications, 26 (2002) 467^478 # 2002 Kluwer Academic Publishers. Printed in the Netherlands

Di¡erences in Pulmonary Functional Indices Derived from the Single-Breath Diagram for CO2 (SBD-CO2) in Horses Related to Age, Sex and Usage C. Herholz1*, R. Straub1, S. Lu«thi1, A. Imhof1 and A. Busato2 1 Department of Equine Internal Medicine, Faculty of Veterinary Medicine, University of Berne, La«nggasstr. 124, CH-3012 Berne; 2M.E. Mu«ller Center of Evaluative Research in Orthopedic Surgery, University of Berne, Berne, Switzerland *Correspondence Herholz, C., Straub, R., Lu«thi, S., Imhof, A. and Busato, A., 2002. Di¡erences in pulmonary functional indices derived from the single-breath diagram for CO2 (SBD-CO2) in horses related to age, sex and usage. Veterinary Research Communications, 26(6), 467^478 ABSTRACT Although pulmonary function tests have been used to evaluate horses with clinically normal lungs and those with recurrent airway obstruction (RAO), comprehensive studies of equine respiratory function, considering factors such as sex, age and usage are not available. Studies on the in£uence of these factors on pulmonary function in healthy horses are required for interpretation of measurements made in those with respiratory disease. The study was performed with 63 warmblood horses and the status of their pulmonary health was evaluated by clinical examination. The functional aspect of pulmonary health was assessed by indices derived from the single-breath diagram for CO2 (SBD-CO2). Di¡erences in pulmonary functional indices between the sexes, age groups and usage groups adjusted for pulmonary health were analysed with hierarchical linear models. It was shown that the e¡ects on pulmonary function of age, sex and usage of a horse were signi¢cant and have to be considered when establishing the true pulmonary functional status of a horse su¡ering from RAO. The multivariate statistical model used included the covariates age, sex and sporting discipline of the horses and accounted for the complex, non-linear interactions between the di¡erent covariates. Keywords: activity, age, horses, multivariate statistical model, recurrent airway obstruction, SBD-CO2, sex, sport, use Abbreviations: AaPO2, alveolar to arterial oxygen tension di¡erence; A1/A2, lung function index, which describes the alveolar e¤ciency; Alve¡, alveolar e¤ciency; LSQM, least-square means; PaCO2, arterial carbon dioxide tension; PaO2, arterial oxygen tension; PIE, pre interface expirate; RAO, recurrent airway obstruction; RR, respiratory rate (breaths/minute); SBD-CO2, single-breath diagram for CO2; VA/Q, alveolar ventilation to perfusion ratio; VDalv/VTalv, ratio of the alveolar dead space according to Bohr to the expiratory tidal volume; VDBohr/VT, ratio of the dead space according to Bohr to the expiratory tidal volume; VCO2, expired CO2 volume; VDphys/VT, ratio of the physiological dead space to the expiratory tidal volume; VT, expiratory tidal volume; VT(L), VT in litres

INTRODUCTION The volumetric capnogram is a plot of the expired carbon dioxide concentration versus the expired volume measured at the airway opening, and the graphical presentation is called the single-breath diagram for CO2 (SBD-CO2). The concentration^volume curve is divided into three distinct phases. Phase I is the relatively carbon dioxide-free 467

468

washout of the proximal conducting airways; phase II is the transitional portion of the curve, characterized by a rapid upswing as the alveolar gas surges mouthward and mixes within the conducting airway; phase III represents the gases from the peripheral alveolated airways and is known as the alveolar plateau. On the basis of anatomical and gas washout studies, it has been suggested that there exists an impairment of di¡usional gas transport in the acinar airways in humans with chronic obstructive pulmonary disease (COPD), which is re£ected in a typical change in the CO2 washout pro¢le. These typical changes in the SBD-CO2 have also been reported in horses su¡ering from recurrent airway obstruction (RAO) by Herholz and colleagues (2001a). A number of variables can be obtained from the SBD-CO2 in a non-invasive manner (e.g. anatomical dead space fraction, dead space fraction according to Bohr's equation). Calculation of other variables, such as the physiological and alveolar dead space fractions, requires arterial blood samples. Moens (1992) considered that the SBD-CO2 provides useful information on pulmonary function during anaesthesia, as did Herholz and colleagues (2001a,b,c) in respect of the conscious horse. Although pulmonary functional tests have been used to evaluate horses with clinically normal lungs and those with chronic lung disease, few comprehensive studies of equine respiratory function, considering dead space ventilation and alveolar e¤ciency for CO2 elimination, are available. The range of reported `normal' values of dead space ventilation is large (Littlejohn and Bowles, 1982; Ohnesorge et al., 1998) and some authors have also reported negative dead space ratios (Davis et al., 1998). Studies of the in£uence of factors, such as sex, age and use on pulmonary functional indices derived from the SBD-CO2 in healthy horses are required to allow interpretation of measurements made in those with respiratory disease. In the present study, we used a linear regression model to determine signi¢cant di¡erences in pulmonary functional indices derived from the SBD-CO2. The objectives were to investigate sexand age-related changes in pulmonary function and the relationship between the usage and pulmonary health in horses with RAO. MATERIALS AND METHODS Study design and animals The study was performed on 63 warmblood horses, 3^22 years old, with a mean age of 10+4 years. Horses in Switzerland mainly compete at between 6 and 12 years of age, whereas horses younger than 6 years are still in training to achieve their optimal performance level. Although horses aged 13 years and older may still compete, many of these horses have been retired owing to accidents or other health problems. We therefore divided the horses into three age categories to investigate the di¡erences in pulmonary function: age category I (n = 11), horses 56 years of age; age category II (n = 37), horses between 6 and 12 years of age; age category III (n = 15), horses 412 years of age. The 5 stallions, 42 geldings and 16 mares, who constituted the study population, were also divided into groups of female (n = 16) and male (n = 47) individuals for study on the e¡ect of sex on pulmonary functional indices.

469

All the horses were sent to the Equine Clinic of the University of Berne for pulmonary function testing; 30 of them had no history of respiratory disorders and 33 had coughs and/or nasal discharges. During three days of hospitalization at the clinic, the horses were kept on straw and were fed with hay and a grain mixture. The status of their pulmonary health was evaluated by a clinical examination and supplementary diagnostic methods, including arterial blood gas analysis, tracheobronchoscopy and cytological examination of tracheobronchial aspirates. All the horses were lunged (5 min each of walking, trotting and cantering) to detect coughing during or after exercise. The respiratory rate (RR), cough at rest or during exercise, the arterial oxygen tenstion at rest (PaO2), the amount and apparent viscosity of tracheobronchial mucus (A-TBM; V-TBM) endoscopically visible by one observer, and the number of polynuclear neutrophils in the cytological examination of tracheobronchial aspirates (PNTBA) were summarized using a scoring system (Table I). TABLE I Scoring system Signs

Result

Respiratory rate (RR) Cough at rest Cough during/after exercise Amount of mucus (A-TBM) Viscosity of mucus (V-TBM) Cytology (PN-TBA) Arterial PaO2

8^16 No No 0^1 0^1 0^1 95^100

Score Result 0 0 0 0 0 0 0

416 Yes Yes 1^2 1^2 2^4 90^94

Score

Result

Score

2 4 2 0.5 0.5 0.5 1

420

3

43 43 5^6 80^89

1 1 1 2

Result

Score

7^8 55^79

2 3

A-TBM, amount of tracheobronchial mucus: 0^1, isolated small spots of mucus; 1^2, isolated moderate sized spots of mucus; 43, large spots and ventral accumulations of mucus up to a continuous ventral sheet of mucus V-TBM, viscosity of tracheobronchial mucus: 0^1, liquid mucus, easy aspiration; 1^2, serious-viscous mucus, more di¤cult aspiration; 43, viscous mucus, di¤cult to very di¤cult aspiration PN-TBA, semiquantitative estimation of polynuclear neutrophils (PN) in the cytological examinations of tracheobronchial aspirates, Pappenheim staining percentages PN's of total cells in a visual ¢eld of the microscope: 0^1, 0^10 % PN; 2^4, 20^40 % PN; 5^6, 50^80 % PN; 7^8, 90^95 % PN

Depending on their state of pulmonary health, the horses were divided into three groups: group I (score 0^1.5 points, n = 20), horses without evidence of pulmonary disease; group II (score 2^4 points, n = 24), horses with either subclinical or minor RAO; group III (score 4.5^12 points, n = 19), horses with moderate to severe RAO. The alveolar to arterial oxygen gradient for PO2 (AaPO2) was calculated for groups I to III according to Crapo and colleagues (1999) using the equation

470

AaPO2 = 0.20936(PB ^ 47) ^ PaCO26[0.2093 + (1.0 ^ 0.2093/R)] ^ PaO2 where PB was the atmospheric pressure in mmHg; the arterial oxygen tension (PaO2) and the arterial carbon dioxide tension (PaCO2) were the mean values obtained for each horse; and the expiratory exchange ratio (R) was assumed to be 0.8 (Willoughby and McDonell, 1979). Atmospheric pressure at the time of arterial blood gas analysis ranged from 705 to 714 mmHg (mean 711+3.5 mmHg). To investigate di¡erences in pulmonary functional indices between clinical groups related to their usage, they were also grouped to three categories: sporting category I (n = 13), leisure; sporting category II (n = 27), sports with an event duration of less than 5 min (dressage and show jumping); sporting category III (n = 23), sports with an event duration of more than 5 min (eventing, coaching, endurance). Measuring equipment Expiratory £ow^volume curves were recorded as a function of time with a computerized ultrasonic spirometer (Spiroson Scienti¢c, Isler Bioengineering AG, Zu«rich, Switzerland). This was calibrated each day with a 7 L calibration syringe and the £ow and CO2 sensors were warmed for 10 min before the measurements were initiated. Capnographic measurements were performed with an IR1507 Miniature Fast Response CO2 Infrared Transducer (Servomex, Jarvis Brook, Crowborough, UK; response time 90 ms; daily calibration with a reference gas mixture (5% CO2)). The sample collection catheter (length 30 cm; diameter 1.2 mm; aspiration rate 200 ml/ min; delay 240 ms) was ¢xed to the ultrasonic £ow sensor, which was 10^40 mm from the nostrils, depending on the size of the horse's head. The corresponding dead space of the facemask was estimated as between 50 and 150 ml. Prior to the pulmonary function tests, the horses were allowed to breathe through the facemask for 5 min to permit adaptation. The protocol then consisted of three or four measurement sessions for each horse during tidal breathing, with an average of 18 breath cycles being recorded during each session. The results of volumetric capnography were corrected for body temperature, pressure and saturation (BTPS) and the CO2 volume (VCO2) was corrected for the sampling delay of the CO2 analyser before being plotted against the expiratory tidal volume (VT). At least 60 breath cycles of each horse were analysed (Herholz et al., 2001b). SBD-CO2-derived indices of pulmonary function The data for the recorded CO2 and volume curves were exported for o¡-line plotting of SBD-CO2 curves and the calculation of the following SBD-CO2-derived pulmonary functional indices (Oberli-Engineering, Hasle-Ruegsau, Switzerland).

471

(1) Ratio of the dead space according to Bohr's equation to the expiratory tidal volume (VDBohr/VT); VD(Bohr) = VT ^ VT 6 (PECO2/PETCO2) (ml), where PETCO2 is the end-tidal CO2 pressure and PECO2 is the mixed expired CO2 pressure, calculated from PECO2 = VCO2/VT (Bohr, 1891). (2) Ratio of the physiological dead space to the expiratory tidal volume (VDphys/ VT); Bohr's equation was modi¢ed by replacing PETCO2 by PaCO2, the partial pressure of carbon dioxide in arterial blood. VDphys/VT = 1 ^ (PECO2/PaCO2) (ml) (Engho¡, 1938). (3) Ratio of the alveolar dead space to the alveolar part of the expiratory tidal volume (VDalv/VTalv); VDalv = VDphys ^ Vds, where Vds is the series dead space volume. The volume of Vds re£ects the volume that is contained in all pathways of the bronchial tree down to the interfaces between fresh gas and alveolar gas calculated by the pre-interface expirate (PIE) method (Wol¡ et al., 1989). The PIE method describes series dead space as a distribution function. It focuses on phase II and considers the slope of phase III of the SBD-CO2. The mean values of PIE were reproducible and plausible in healthy humans and humans with COPD. VTalv is the alveolar part of the expiratory tidal volume, calculated from: VTalv = VT ^ Vds. (4) Alveolar e¤ciency (Alve¡%), which describes the ratio of the CO2 volume eliminated in the breath to the e¡ective or alveolar part of the expiratory tidal volume, calculated from Alve¡ = A1/(A2 + A1) 6 100 (Fletcher, 1980) (Figure 1). (5) Ratio A1/A2 (%): A1/A2 is another index of e¡ective CO2 elimination from the lung; it compares the measured expired volume of CO2 with the volume of CO2 that would have been expired in an ideal lung (Figure 1). Statistical analysis Di¡erences in PaO2, PaCO2 and AaPO2 between the clinical groups, adjusted for sex, age group and usage group, were calculated using linear models. Di¡erences in pulmonary functional indices between sex, age group and usage group, adjusted for pulmonary health, were analysed with hierarchical linear models. The pulmonary functional indices listed in Tables II^V were de¢ned as outcome variables. The explanatory variables included sex, age (age categories I^III), usage (sporting categories I^III) and severity of pulmonary disease (health groups I^III). It was assumed that the e¡ects of pulmonary disease di¡er between the di¡erent uses of the horses. The health groups were therefore analysed hierarchically as nested within a usage category. Additionally, ¢rst-order interaction terms of sex by clinical group, as well as age by clinical group, were included in the model. A preliminary analysis had indicated non-linear e¡ects of age for some variables under study. It was therefore decided to categorize age into the three groups described above. Additional explana-

472

Figure 1. Derivation of pulmonary functional indices of e¡ective CO2 elimination from the lung from the SBD-CO2. Alveolar e¤ciency (Alve¡ ) % = A1/(A2 + A1) 6100, and A1/A2 (%) = A1/A2

tory variables included the number of the measurement sessions and the number of breaths within the session, in order to adjust the analysis for the respective e¡ects. Agreement between the model and the data was assessed using residual analysis and interpretation of the amount of variance of the outcome variable accounted for by the models (R2 values). Power calculations of statistical tests were performed for various comparisons and a type I error rate of 0.05. Data are given throughout the paper as least-square means (LSQM), derived from the models described above, and di¡erences between groups were assessed as di¡erences between LSQMs. The level of signi¢cance for all statistical analyses was set at 0.05. RESULTS The PaO2, given as the LSQM, was 100.7 mmHg for group I, 93.5 mmHg for group II and 72.1 mmHg for group III. The di¡erences in PaO2 were signi¢cant (p50.05) between groups I, II and III. Least-square means of PaCO2 were 39.0 mmHg in group I, 42.0 mmHg in group II and 45.2 mmHg in group III. The di¡erences in PaCO2 were also signi¢cant (p50.05) between groups I, II and III. The least-square mean for AaPO2 was 2.1 mmHg in group I, 6.4 mmHg in group II and 23.6 in group III. The di¡erences in AaPO2 were signi¢cant (p50.05) between groups I and III. Power calculations for the various comparisons revealed values above 0.95 for all the indices under study. Di¡erences related to the di¡erent age categories in the LSQM of

473

TABLE II Di¡erences in least-square means (LSQM) of pulmonary functional indices derived from the SBD-CO2, as related to di¡erent age categories Index (LSQM) VT(L) VDBohr/VT VDphys/VT VDalv/VTalv Alve¡ (%) A1/A2 (%)

56 years

6 to 512 years

412 years

4.74a 0.50a 0.56a 0.37a 63.08a 54.38a

5.36b 0.58b 0.66b 0.46b 54.13b 48.80b

5.71c 0.55c 0.61c 0.36a 64.17a 50.47c

See Abbreviations for meaning and symbols a,b,c

Di¡erent superscript letters indicate signi¢cant di¡erences (p50.05) between age groups

TABLE III Di¡erences in least-square means (LSQM) of pulmonary functional indices derived from the SBD-CO2, as related to sex Index (LSQM) VT(L) VDBohr/VT VDphys/VT VDalv/VTalv Alve¡ (%) A1/A2 (%)

Female

Male

5.29a 0.53a 0.59a 0.39a 60.03a 51.64a

5.25a 0.55b 0.62b 0.39a 60.04a 50.52b

See Abbreviations for meaning and symbols a,b,c

Di¡erent superscript letters indicate signi¢cant di¡erences (p50.05) between male and female animals

the pulmonary functional indices derived from the SBD-CO2 are presented in Table II. The relationship between pulmonary function and age appeared to be complex and non-linear, since the pulmonary functional indices deteriorated from age category I to II and improved from II to III. Signi¢cant (p50.05) age group by clinical group interactions were found for all the SBD-CO2 indices listed in Table II. Table III shows the di¡erences related to sex in the LSQMs of the pulmonary functional indices derived from the SBD-CO2. The sample population consisted of 25% females and 75% males. There were no remarkable sex-related trends in pulmonary function suggested by the indices VT, VDalv/VTalv or Alve¡, with the

474

exception of a signi¢cant deterioration in VDBohr/VT, VDphys/VT and A1/A2 in male compared to female horses. The interactions between sex and clinical group were signi¢cant (p50.05) for the indices VT, VDBohr/VT, VDphys/VT and A1/A2. Di¡erences in the LSQMs of the pulmonary functional indices derived from the SBD-CO2 between healthy horses (group I), horses su¡ering from subclinical or minor RAO (group II) and horses with moderate or severe RAO (group III) are presented in Table IV. The di¡erences in VDBohr/VT, VDphys/VT and A1/A2 were signi¢cant between groups I, II and III and VDalv/VTalv and Alve¡ di¡ered signi¢cantly between groups I and III. The di¡erences in VT were signi¢cant between groups I and II. No consistent clinical group-related trends in pulmonary functional indices that suggested improving or deteriorating pulmonary function were observed when investigating di¡erences of SBD-CO2 indices within a distinct sporting category (Table V), except that in the sporting category I, leisure, the e¡ect of the di¡erent sporting disciplines on pulmonary function di¡ered and was not linear. DISCUSSION The model yields estimates of pulmonary functional indices derived from the SBDCO2 related to age categories, sex and the additional covariables. This method of analysis allows subjects to have an unequal number of observations. The explanatory variables entered into the model were the number of breaths, number of measurement sessions, the clinical group of pulmonary health and the sporting category. The number of breaths and measurement sessions that were included in the data analysis had proved to be signi¢cant in previous studies (Herholz et al., 2001b,c). TABLE IV Di¡erences in least-square means (LSQM) of SBD-CO2 indices between clinical groups Index (LSQM) VT(L) VDBohr/VT VDphys/VT VDalv/VTalv Alve¡ (%) A1/A2 (%)

Group I

Group II

Group III

5.17a 0.53a 0.60a 0.39a 61.25a 51.51a

5.57b 0.50b 0.57b 0.38a 62.08a 54.05b

5.07a 0.60c 0.66c 0.43b 57.23b 47.69c

See Abbreviations for meaning and symbols Group I (n = 20), healthy horses; group II (n = 24), horses with either subclinical or minor RAO; group III, (n = 19), horses with moderate to severe RAO a,b,c

Di¡erent superscript letters indicate signi¢cant di¡erences (p50.05) between health groups

5.59a 0.37a 0.36a 0.12a 88.17a 62.64a

5.25a 0.34a 0.36a 0.25b 75.39b 66.74b

4.21b 0.61a 0.68b 0.40c 60.32c 47.35c

5.33a 0.57a 0.62a 0.43a 56.18a 47.71a

6.42b 0.52b 0.57b 0.41b 59.09b 51.52b

5.67c 0.58a 0.61a 0.45c 55.81a 48.40a

Sport category II ööööööööööööööö Group I Group II Group III 6.07a 0.49a 0.53a 0.36a 63.92a 54.78a

5.97a 0.49a 0.54a 0.38a 62.41a 55.07a

6.17a 0.49a 0.50a 0.20b 79.79b 57.19b

Sport category III ööööööööööööööö Group I Group II Group III

a,b,c

Di¡erent superscript letters indicate signi¢cant di¡erences (p50.05) between health groups within a sporting category

Sporting category I, leisure; sporting category II, sports with an event duration of less than 5 min (dressage and show jumping); sporting category III, sports with an event duration of more than 5 min (eventing, coaching, endurance)

See Abbreviations for meaning and symbols

VT(L) VDBohr/VT VDphys/VT VDalv/VTalv Alve¡ (%) A1/A2 (%)

Index (LSQM)

Sport category I ööööööööööööööö Group I Group II Group III

TABLE V Di¡erences in least-square means (LSQM) of SBD-CO2 indices between di¡erent clinical groups within the sporting categories I, II and III

475

476

Di¡erences in pulmonary functional indices derived from the SBD-CO2 as related to age category From the evidence of SBD-CO 2 indices, the present study indicated that the pulmonary function of horses was worse between 6 and 12 years of age than it was before or after that age. We consider that the observed di¡erences between the age categories of the horses in the pulmonary functional indices may re£ect di¡erences in the stabling and environmental conditions during the course of life. During the ¢rst period of life, up to 5 years of age, horses are most commonly kept on pasture and are trained in the associated stable. Between 6 and 12 years of age they are mainly kept indoors and most of them travel from event to event, stay in di¡erent barns and are much more exposed to dust and respiratory tract infections. When the horses are retired (at 13 years of age or older), they are again predominantly kept on pasture and do not travel very frequently from barn to barn. The study has also shown that there exists a non-linear and complex relationship between pulmonary function and age, an observation that was previously made in humans (Berhane et al., 2000; Gri¤th et al., 2001). We found signi¢cant (p50.05) interactions between age group and pulmonary health for all SBD-CO2 indices under study, indicating that the e¡ect of age group has to be considered in evaluating pulmonary function within clinical groups of horses. Di¡erences in pulmonary functional indices derived from the SBD-CO2 as related to sex Bohr's ratio and the physiological dead space tidal volume ration, and the index A1/A2, were signi¢cantly di¡erent between male and female horses, with the males having larger dead space ratios and a smaller value for alveolar e¤ciency. Two recent studies in humans have reported con£icting results on sex di¡erences in the e¡ect of asthma on the level of pulmonary function (Weiss et al., 1992; Gold et al., 1994). In one of these studies, a larger percentage de¢cit in FEV1 was seen in males with asthma than in females with asthma. The interactions between sex and pulmonary health were signi¢cant (p50.05) for VT, VDBohr/VT, VDphys/VT and A1/A2 and should be considered if clinical groups of horses with RAO are investigated. Di¡erences in SBD-CO2 derived pulmonary functional indices between clinical groups of COPD within di¡erent sporting categories As expected, the SBD-CO2 indices between the clinical groups, regardless of the sporting categories of the horses, di¡ered ^ with increasing dead space ratios and decreasing alveolar e¤ciency ^ as the degree of RAO progressed. Dead space and alveolar e¤ciency indices have been used to quantify non-uniformities in the distribution of pulmonary blood £ow relative to ventilation (Fletcher, 1980). Traditionally, the AaPO2 is another important index in terms of assessing alveolar ventilation/perfusion

477

(VA/Q) mismatching (Willoughby and McDonell, 1979). Di¡erences (p50.05) in the AaPO2 between groups I and III corresponded to the di¡erences observed in Alve¡ and VDalv/VTalv (Table IV) and, in particular, VDalv/VTalv represents the VA/Q matching of the lungs (Littlejohn and Bowles, 1982). The interactive e¡ects of clinical status within the usage groups of the horses were signi¢cant but not consistent between the three sporting categories. We were not able to o¡er a plausible explanation for the di¡erences between the sporting categories. However, as with the more intense training in sporting category III in our study, it has been shown in human patients with COPD that, with more intense training, the CO2 kinetics in the body ^ and therefore alveolar e¤ciency ^ is improved (PuenteMaestu et al., 2000). Although, in human patients with COPD, alveolar e¤ciency at rest was improved by training, it was not better than in healthy individuals (PuenteMaestu et al., 2000). REFERENCES Berhane, K., McConnell, R., Gilliland, F., Islam, T., Avol, E., London, S.J., Rappaport, E., Margolis, H.G. and Peters, J.M., 2000. Sex-speci¢c e¡ects of asthma on pulmonary function in children. American Journal of Respiratory Critical Care Medicine, 162, 1723^1730 Bohr, C., 1891. Ueber die Lungenathmung. Skandinavisches Archiv fu«r Physiologie, 2, 236 Crapo, R.O., Jensen, R.L., Hehewald, M. and Taskin, D.P., 1999. Arterial blood gas reference values for sea level and an altitude of 1,400 meters. American Journal of Respiratory Critical Care Medicine, 5, 1525^1531 Davis, M.S., Murray, M.J. and Donaldson, L.L, 1998. Clinical assessment of gas exchange in mature horses. Equine Veterinary Journal, 30, 396^400 Engho¡, H., 1938. Volumen ine¤cax. Uppsala La«karefend Fo«rhand, 44, 191 Fletcher, R., 1980. The single breath test for carbon dioxide, (PhD thesis, Berlings, Arlo«v, Schweden) Gold, W.R., Wypij, D., Wang, X., Speizer, F.E., Pugh, M., Ware, J.H., Ferris, B.G., Jr. and Dockery, D.W., 1994. Gender- and race speci¢c e¡ects of asthma and wheeze on level and growth of lung function in children in six U.S. cities. American Journal of Respiratory Critical Care Medicine, 149, 1198^1208 Gri¤th, K.A., Sherill, D.L., Sigel, E.M., Manolio, T.A., Bonekat, H.W. and Enright, P.L., 2001. Predictors of loss of lung function in the elderly. American Journal of Respiratory Critical Care Medicine, 163, 61^ 68 Herholz, C., Straub, R., Moens, Y. and Busato, A., 2001a. Statistical shape analysis of volumetric capnograms: evaluation of a new approach for the assessment of pulmonary function in horses with chronic obstructive pulmonary disease (COPD). Journal of Veterinary Medicine, 48, 75^84 Herholz, C., Straub, R. and Busato, A., 2001b. Ultrasound^spirometry and capnography in horses: analysis of measurements reliability. Veterinary Research Communications, 25, 137^147 Herholz, C., Straub, R. and Busato, A., 2001c. The variability and repeatability of indices derived from the single breath diagram for CO2 (SBD-CO2) in horses with COPD and the e¡ect of lobelin hydrochloride on these indices. Veterinary Research Communications, 25, 401^412 Littlejohn, A. and Bowles, F., 1982. Studies on the physiopathology of chronic obstructive pulmonary disease in the horse. VI. The alveolar dead space. Onderstepoort Journal of Veterinary Research, 49, 71^ 72 Moens, Y.P.S., 1992. Ventilation and gas exchange in each lung of the anaesthetised horse. The in£uence of body position and mechanical ventilation, Proefschrift (University of Utrecht) Ohnesorge, B., Tro«tschel, Ch. and Deegen, E., 1998. Bestimmung von Totraum und exspiratorischem Mischluftvolumen zur Diagnostik chronischer Lungenerkrankungen beim Pferd. Pferdeheilkunde, 14, 450^455 Puente-Maestu, L., Sanz, M.L., Sanz, P., Ruiz de Ona, J.M., Rodriguez-Hermosa, J.L. and Whipp, B.J., 2000. E¡ects of two types of training on pulmonary and cardiac responses to moderate exercise in patients with COPD. European Respiratory Journal, 15, 1026^1032

478 Weiss, S.T., Tosteson, T.D., Segal, M.R., Tager, I.B., Redline, S. and Speizer, F.E., 1992. E¡ects of asthma on pulmonary function in children: a longitudinal population based study. American Review of Respiratory Disease, 145, 58^64 Willoughby, R.A. and McDonell, W.N., 1979. Pulmonary function testing in horses. Veterinary Clinics of North America, Large Animal Practice, 1, 171^197 Wol¡, G., Brunner, J.X., Bowes, C.L., Muchenberger, R. and Bertschmann, W., 1989. Anatomical and series dead space volume: concept and measurement in clinical praxis. Applied Cardiopulmonary Pathophysiology, 2, 299^307 (Accepted: 17 March 2002)