Pulmonary diffusion limitation after prolonged strenuous exercise

Respiration Physiology. 83 (1991) 143-154

143

Elsevier RESP 0 ! 746

Pulmonary diffusion limitation after prolonged strenuous exercise G6rard Manier, Jean Moinard, Pierre T6choueyres, Nicole Var6ne and Herv6 Gu6nard Laboratory of Physiology. Faculty of Medicine, University of Bordeaux, Bordeaux, France (Accepted 15 September 1990) Abstract. To determine the effect of strenuous prolonged exercise on alveolo-capillary membrane diffusing capacity, i l marathon r ~.nners aged 37 + 7 years (mean + SD) were studied before and during early recovery (28 + 14 min) from a marathon race. Lung capillary blood volume (Vc) and the alveolo-capillary diffusing capacity (Dm) were determined in a one-step maneuver by simultaneous measurements of CO and NO lung transfer (DLco and DLNo, respectively) using the single breath, breath-holding method. After the race, both DLco and DLNo were significantly decreased in all subjects ( - 10.9 + 4.8~, P < 10 - 4 and -29.0 + 11.1%, P < 10 -4, respectively). The mean value of the derived Dmco decreased by 29.3 + I 1.1 ~/o,whereas Vc had not entirely returned to control resting value. Although these results do not indicate the detailed mechanism involved, interstitial lung fluid was suspected to accumulate, particularly in alveoli, during the race. We concluded that the high overall work load and the extended duration of the exercise both contributed to a transient change in the structure of the alveolo-capillary membrane thereby affecting the diffusing capacity of the alveolo-capillary membrane. -

Animal, men; Diffusing capacity, pulmonary, for CO, for NO; Exercise, and pulmonary diffusing capacity

There is considerable circumstantial evidence for a role of pulmonary diffusion limitation in the widening of the alveolar arterial Po.~ difference during strenuous exercise, both in humans (Dempsey et ai., 1984; Hammond et al., 1986) and horses (Wagner et al., 1989). Pulmonary diffusion limitation may occur via two different mechanisms, although the end result in highly trained athletes at sustained high metabolic rate is the inability for mixed venous blood to equilibrate with alveolar gas by the time red cells reach the end of the pulmonary capillary: (i) the pulmonary capillary transit time can be shortened because blood flow continues to increase after the pulmonary capillary blood volume (Vc) has reached its maximum value. (ii) Accumulation of extravascular lung water is also thought to account for the widened alveolar arterial Po, difference, although whether there is sufficient edema to actually increase alveolo-capillary Correspondence to: G. Manier, Laboratoire de Physiologic, Facult6 de M/~decine Victor Pachon, Universit/: de Bordeaux I1, 33076 Bordeaux C~dex, France 0034-5687/91/$03.50 © 1991 Elsevier Science Publishers B.V.

144

G. MANIER et al.

diffusion is not known. However, if present, pulmonary edema should persist in the minutes after completion of the exercise, and continue to depress the diffusing capacity of the lung (DL). DL can be partitioned (Roughton and Forster, 1957) into the diffusing capacity of the blood gas barrier (Din) and the gas uptake capacity ofcapillaries, which is a function of the product of the specific blood gas conductance (0) and Vc. There is considerable dispersion in the reported values for DLco both with respect to the absolute values and the relative changes with work load during exercise (Harris and Heath, 1987; Piiper and Scheid, 1980). Moreover, relatively little is known about the effect of exercise duration on Dm and Vc. While measuring DLco by the breath-holding method in athletes breathing air and pure 02 in turn, Miles et al. (1983) found that Dm decreased significantly while Vc remained close to control values during the I to 2 hours after a marathon race. These results are in apparent disagreement with those of Maron et al. (1979) who found no significant change in DLco after a marathon. Nevertheless, in this study there was a significant correlation between the change in DLco and heart rate at the time of n.easurements after the race, indicating that persisting increased cardiac output may have led to an overestimation of the post-race values of DLco (Lewis et al., 1978). In the present study, we measured lung transfer for both carbon monoxide (DLco) and nitric oxide (DLNo) by the single breath breath-holding method in trained endurance athletes before and in the very early phase of recovery after a marathon race. Vc and Dm were derived according to the method described by Gu~nard et al. (1987) and Moinard and Ou~nard (1990). The post-exercise-induced alterations in DL were also related to cardiac function evaluated at the same time by echocardiography. Individual cardiac data are presented and discussed elsewhere (Manier et al., 1990). The main result of the present study is the consistent decrease in both DLco and DLNo after completiol~ of the marathon.

Methods Subjects. The subjects for this investigation were one female and ten male long distance runners who were training for a marathon race. They had participated in competitive long distance running for the past 2 to 20 years. They volunteered for the study, and gave their informed consent according to the guidelines of the human subject Institutional Review Committee. The subjects' physical characteristics are presented in table 1. They had normal screening physical examinations, ECG, and spirometry. There were no smokers, and none reported active asthma or the use of any medication including b~onchodilators and vitamins. Protocol. Measurements were taken in each athlete a few hours before the race and on the same day immediately after finishing the race. Although the equipment was close to the finish line, a short delay (28 + 14 min) was unavoidable as a single system was used for the lung transfer measurements. In seven subjects, a heart rate monitor (Sport

DIFFUSION LIMITATION IN MARATHON RUNNERS

145

i= 0 o,m ¢,1 U I= III

(',l

-t-I

~

+i~

,_,,

L. I,-. o°

°°

e,i ~

°o

~

°,

~

oo

m

°o

m

°°

~

°o

°°

~

°°

~

~

°o

~

ill o,,,~

o,,.~

f.,g "H

-,H

,,<

,,,.=,

Ill I= Ill

I= o,m

+I ~.

I= III

•~

~ I~

v-~ ~

v-~ I~

~ ~

,~

,.~

,.~

,,.,~

I

I

I

,.~

,~

--'

~r~ i~ •

I

~ ~.~

'-' V

~r~ ~

tli .

.

.

.

,,~

+I

~

o

+,= ~,-,~ + I 4-I " " eq v'~ V

III I= tll

I ~

~

~

~

~

~

~

,.=,

,~

,,,,,

,,,.,

,..,

,~

.

~

~

~

~

+I

,=~

+I

I,,) ,~

.

.

.

~

0 U o,,., °~ L. I,,,I

? 4-I I".-

I

146

G. MANIER et al.

Tester PE 3000) was installed, and the subjects were instructed in its use. Heart frequency was stored every 15 sec throughout the race, and the time of the start and finish were also recorded. All values were uploaded to a micro computer to determine the change in heart frequency and mean running velocity during the race. Individual mean heart rate was compared to individual maximal heart rate recorded previously on a treadmill in the laboratory. Experimental set up. Measurements of gas volume were made in a bag in box system (P. K. Morgan, U.K.) connected to a dry spirometer. A rotating system ofelectronically controlled valves allowed the subject to complete the maneuver in a single operation. Helium was used as insoluble gas. Helium and CO were measured with Hewlett Packard analyzers (47 305A DCO S.B. Controller) and NO was measured by chemiluminescence (Thermoelectron, U.S.A.). This technique is the standard physical reference method for measurement of low concentrations of nitrogen oxides in polluted atmospheres. Procedure. The experiments were performed in the sitting position. The subject was connected to the apparatus. After a full exhalation, he inspired a gas mixture up to a volume close to his predetermined vital capacity, held his breath for 3 sec and finally performed a rapid expiration. During this expiration a 900 ml 'alveolar' sample was withdrawn af,~r the washout of the first 900 ml. Inspired volume, inspired and alveolar fractions of NO, CO, He (FINo, FANo(t), Flco, FAco(t), FIrle, FAHe(t), respectively) were measured. The effective breath holding time (t) was taken as the sum of the inspired time, the true breath-holding time and the expired time to the start of the sampling period. It has been shown in normal subjects (Gu6nard et al., 1987) that a true breathholding time of about 3 sec enables determination of DLNO with a low coefficient of variation. Because of the high affinity of haemoglobin for NO (see re£ in Meyer and Piiper, 1989) u true breath-holding of more than 3 sec in normal subjects (homogeneous lung) would lead to an unmeasurable fraction of NO in the alveolar gas after addition of 8 ppm NO to the inspired gas. The sensitivity ofthe apparatus was 1~o for a 10 ppm full scale reading. A true breath-holding time of 3 sec does not affect the measurement of DLc.o in normal subjects (Graham et al., 1985). DL and Dm are expressed in ml STPD rain- ' .ram Hg- ~, and Vc in ml. The inspired fractions were 0.21 02, 0.10 He, and 2.8 × 10-3 CO in N 2. 8 ppm NO were added to this mixture just before the determinations. Calculations. The conventional equation describing the transfer for CO in a homogeneous lung model is I/DLco = l/Dmco + l/Vc x 0co), where 0co is the specific blood transfer conductance for CO (l/0co - 0.73 + 4.4 Po2 (Roughton and Forster, 1957)). Dm and Vc are two unknowns. Two principles can be used to solve this equation. The usual way is based on the competitive reactions of 02 and CO for Hb and the associated change in 0co (competitive principle). Another way, used in the present study, is to measure the transfer capacities for two different gases, and then solve

DIFFUSION LIMITATION IN MARATHON RUNNERS

147

the set oftwo equations. NO and CO can be used as test gases, DLNo and DLco being determined simultaneously (Borland and Higenbottam, 1989). As the rate of uptake of nitric oxide with hemoglobin is extremely rapid l/(Vc × 0NO) Can be neglected, and the equation for DLNo Can be written: I / D L N o ffi l / D m N o =

l/aDmco

(1)

where a = (~/Mwco/~/MWNo) × (~so/aCco)then Vc ffi 1/0co x [ 1/(1/DLco - a/DLNo)]

(2)

Mw are molecular weights, 30 and 28 for NO and CO, respectively, 0~are solubilities, 0.0439 and 0.0215 for NO and CO in plasma at 37 °C, respectively (Gibson and Roughton, 1957), coefficient 'a' is 1.97. Statistical analysis. Pre-race and post-race results were compared by Student's t-test for paired observations. The level of significance was set at P < 0.05. The relationships of the changes in the functional variables were determined by linear regression.

Results

Monitoring the heart rate in seven out of the eleven subjects during the race showed that a steady state was reached rapidly which was then maintained during a period of around three hours. Moreover, the mean heart frequency during the race related to the maximal heart rate previously measured in our laboratory (87 + 4%) demonstrated the high relative rate in the runners during the race (table 1). Although the resting heart rate before the race (66 + 13. min - i ) was lower than after the race (85 + 13. min- ! ), the echocardiographic measurements (Manier et aL, 1990) showed that the post-race mean cardiac index value returned to control values within the short interval between the ~rrival of the runners and the gas transkr measurements (4.2 + 1.2 and 3.8 + 1.2 L. min- I. m - 2 respectively, P = N S ) . Before the race, values of DLco (31.2 +_ 3.0 ml' rain- t" mm Hg- i (table 2)) were close to mean results obtained in healthy subjects at rest using the breath-holding method (Piiper and Scheid, 1980). Values of DLNo were close to the values reported in the original study (Gu6nard et al., 1987), and the mean value of the DLNo/DLco ratio was close to the value given by Borland and Higenbottam (1989) who measured DLco and DLNo simultaneously in healthy subjects at rest using the same technique. After the race both DLco and DLNo decreased significantly in all subjects (10.8 and 29.3~o, respectively; P < 10-4). Since DLNo = DmNo = a D m c o , and DLNo decreased much more than DLco, we concluded that prolonged exercise did affect the membrane diffusing capacity. The mean value of Dmco decreased by 29~o (P < 10 -4) after exercise, while Vc increased by only 10~o.

148

G. MANIER et ai.

< -H

L.

Z i=.

+1

m

Oo _I

L.

~7

< -!-I

+~ -~

-I-I

~.~ ~

4"I ~ I

~. r-. V

I

U

E

IJ

~

~e~ ~e~ ~

~1"

~

~1- ~1- ~e~ ~

r'~

r~l

¢',1 ¢~1 ¢',1 ¢ ~

~

~e~

+1

r~l

+l

i_

~"

u °'~

<

¢~

¢~

r,4

¢',1 ¢',1

,.! + 1 -I-I - - ' ¢',1 ,'.'~ V

¢"1

I

I

t~ ¢~

+1

< ~.-----_ ~I '-"

..=. 0

~

+I

I--i 1

~~ _

--'

. •

+I

o,o.

.

.

.

.

.

.

.

.

I

_

V I

+1

.

.

_

+,

o

~,=,, -I-I

4-I C~__

~.~.v

:'E= =

!

--

--'

+1

o,i. Z

Z ~J =

--

e.)

I

DIFFUSION LIMITATION IN MARATHON RUNNERS

149

Discussion The main result obtained in this study, i.e. the 29~, decrease in Dm, strongly suggests that prolonged exercise alters the diffusing capacity of the alveolo-capillary membrane, probably as a result of interstitial edema. However, there are potential methodological errors in the calculation of Vc and Dm, especially with respect to core temperature and the regional distribution of the diffusing capacity.

M~thodology. During exercise, core temperature may rise to 39 °C or more, depending on both time, relative workload and environmental conditions (Saltin and Hermansen, 1966; Adams et al., 1975). However, it has also been shown that core temperature rapidly returns to normal after exercise (Saltin and Hermansen, 1966). During the marathon, athletes ran at a high relative workload, and measurements were performed in the very early phase of recovery after the race. The rise in blood temperature could alter the value of 0co and hence the calculated values of Vc and Dm. We therefore calculated the potential error in both Vc and Dm, which would have occurred ifthe core temperature had not returned to control values. From the results of Holland (1969) l/0co decreases by 5% for each °C above 37 °C. That is, ifcore temperature remained close to the value observed during exercise, the post-race value of Vc would be 5 % below that of the control value for each °C rise in temperature. However, the difference between the measured Vc (using a value of 0co for a normal core temperature of 37 °C) and a corrected Vc value for the actual core temperature at the time of measurement should be correlated with the interval between the end of the race and the time of the measurement. Since changes in both DLco and DLNo were not correlated with this interval, and since the mean time delay was above that required for core temperature to return to 37 °C (Saltin and Hermansen, 1966), it is unlikely that the mean Vc value was significantly affected by the lack of correction of 0co for core temperature at the time of measurement. For the calculation of Dm, the relationship is Dmco -- DLr~o/a -- DmNo/a, where a is a direct function of the ratio of the blood solubilities (0c)of NO and CO. ~co and ~NO fall by 1% and 1.6~, respectively, for each °C rise in temperature (Monteil, 1980). Thus, even if core temperature had not entirely returned to resting values, the systematic error in Dm would have been negligible. It is also of interest to note that using the NO/CO method, the derived Dm¢o is quite independent of 0co, and therefore of the potential effect of temperature and hemoglobin concentration on 0co. Finally, knowing that Hb concentration can increase slightly in thermal conditions similar to those of the present experiments, 0co should be corrected correspondingly (Holland 1969) in the calculation of Vc. Assuming that the Hb concentration increased by 2~o post-race (a value slightly above the value observed in a series of 90 marathon runners (Davidson et ai., 1987)), the post-race values for Vc were still higher than the pre-race values (A - 5.2 + 6.8 ml. m - 2). In the single breath breath-holding method it has been demonstrated that DLco is affected by the regional distribution of the diffusing capacity of the membrane with

150

G. MANIERetal.

respect to alveolar volume. It will thus depend on the breath-holding time (Graham et al., 1985). In the present study, a relatively short breath-hold time was selected for the reasons detailed above in Methods. The decreases in DLco and DLNO after the race might have been due to an enhanced heterogeneity in the regional distribution of alveolar volume (VA). Although we have no data on the regional distribution of VA after the race, other workers (Mahler and Loke, 1981; Miles et ai., 1983) have failed to demonstrate airway obstruction in runners after a marathon race at above zero ambient temperatures. It thus seems reasonable to assume that: (i)there was no airway obstruction in our subjects, and that there was no increase in the heteroge~eity of VA after the race, and (ii) the breath-hold time we used did not induce, per se, a decrease in DL as has been found in patients suffering from airway obstruction. Vc after prolonged exercise. The values obtained after the marathon race, which were little different from control values, are in agreement with data on pulmonary circulation in healthy subjects at rest (Reeves et ai., 1988). During upright exercise, pulmonary vascular recruitment is mainly due to an increase in pulmonary vascular pressure as a direct consequence of the increase in cardiac output. In the present study, the post-race cardiac output had returned to near control values (Manier et al., 1990). Hence, it is likely that pulmonary vascular pressure had also returned to pre-race values, inducing a return of Vc to the pre-race control value. Our values of Vc are in agreement with those of Miles et al. (1983), although these authors found no difference between pre- and post-race values, probably because they measured DLco 1-2 h after the end ofthe race. Decrease in Dm after prolonged exercise. Values of DL after the race could also have been decreased by a reduction in gas exchange area. It is known that this parameter increases during exercise via two main mechanisms: (i) increase in the diameter of capillary vessels by increased transmurai pressure, (ii)recruitment of capillaries which are not perfused under resting conditions. However, our measurements were performed after exercise and we have shown that cardiac output returned to control values, and that post-race Vc was slightly above control values by the time ofrneasurement. Alveolar volume was also the same in the pre- and post-race conditions (table 1). It is thus unlikely that gas exchange area was reduced after the race. Our results indicate that after prolonged strenuous exercise the decrease in DLco and DLNo is due to a modification in membrane diffusing capacity. These results are in agreement with those obtained by Miles et al. (1983) who found that Dm fell by 21 ~o in 8 men after running a marathon. These values are comparable to the value we measured sooner after the race using a different technique. It is of interest to compare our results with those of Hammond et al. (1986) who studied pulmonary gas exchange by the inert gas method in normal subjects during an increasing work load exercise at sea level. They showed that the widening of the alveolar arterial Po2 difference with increasing work load was not only due to a deterioration in ventilation to perfusion (QA/0) relationships, but also to a limitation in pulmonary diffusion of 02 when the work load reached a value corresponding to high 02 uptake

DIFFUSION LIMITATION IN MARATHON RUNNERS

i.e.

151

'V'o, > 3 L. min- ~ (_~ 40 ml. min - ~. k g - ~ in subjects weighing 76 + 4.4 kg).

Although the exercise test performed in this latter study was of relatively short duration, these authors suggested that interstitial edema could have accounted for their results. In our study, we estimated the mean metabolic rate of the athletes from the mean cost ofrunning. The general equation for Vo, (ml. min - ~) in a runner depends on body mass and running velocity (Van der Walt and Wyndham, 1973). If we take a mean velocity of 14 km/h (3 h marathon), and a mean body mass of 62 kg, we calculated that the mean Vo, was 3 L. min- ~ (-~ 48 ml. min- i. kg- ~). This value is similar to the absolute threshold value observed by Hammond et al. (1986), but for our subjects weighing only 62 + 4 kg, it corresponds to a greater relative work load. If the limitation in lung diffusing capacity due to accumulation of water in the lung depends on work load, the diffusing capacity would be also affected by the duration of the exercise, especially for exercise performed at a high metabolic rate such as during a marathon race. Interstitial pulmonary edema during exercise. The persisting decrease in the diffusing capacity of the alveolar capillary membrane when the subjects had returned to a resting condition indicated that there was a structural alteration in the membrane, probably due to interstitial pulmonary edema. This structural alteration could partially account for the persisting resting VA/(~ inequality following heavy exercise (Hammond et al., 1986). During exercise, both pulmonary arterial and left atrial pressure increase (Reeves et al., 1988). At submaximal levels corresponding to 80~o of Vo,max, it has been shown that mean pulmonary arterial pressure in healthy young subjects in the upright position increases by up to 38 mm Hg (28.6 + 5.5), and left atrial pressure by up to 25 mm Hg (17.5 + 4.9). We are not aware of any data on pulmonary hemodynamics during prolonged exercise. Nevertheless in the present study, heart rate was monitored during the race and mean metabolic rate was computed, which confirmed the high work load performed over the 3 hours of the race. Thus it seems likely that lung capillary pressure during such prolonged exercise was at, or was near to the values reported at a high level of submaximal exercise (Reeves et al., 1988). Younes et al. (1987) using an in situ perfused canine lobe preparation, have recently shown that when pulmonary blood flow is increased to levels observed during heavy exercise in healthy subjects, the hydrostatic pressure at the filtration site is well above that required for progressive water accumulation. Local stress on endothelial surfaces could also lead to a change in membrane permeability (Reeves et al., 1988). In summary, we found evidence for a decrease in alveolar capillary membrane diffusion in healthy subjects after very strenuous prolonged exercise. Although the results give no indication of a possible mechanism, accumulation of interstitial fluid was suspected. The overall work load and the duration of exercise appeared to be two main factors affecting the diffusion limitation. Nevertheless, further studies will be required to establish whether such structural alterations in the lung may be a limiting factor in the performance of heavy and/or prolonged exercise.

152

G. MANIER etal.

Acknowledgements. The authors wish to thank Mr. P. Roux for permission to use his facilities and the equipment of Sport Athl6tique M~rignacais. We thank Mrs. N. CapdeviUe and Mrs. G. Flamand-Gracia for their secretarial assistance.

References Adams, W. C., R. H. Fox, A. J. Fry and I.C. MacDonald (1975). Thermoregulation during marathon running in cool, moderate, and hot environment. J. Appl. Physiol. 38: 1030-1037. Borland, C. D. R. and T.W. Higenbottam (19S9). A simultaneous single breath measurement of pulmonary diffusing capacity with nitric oxide and carbon monoxide. Eur. Respir. J. 2: 56-63. Davidson, R. J. L., J. D. Robertson, G. Galea and R.J. Maughan (1987). Hematological changes associated with marathon running. Int. J. Sports Med. 8: 19-25. Dempsey, J.A., P.G. Hanson and K. S. Henderson (1984). Exercise-induced arterial hypoxaemia in healthy human subjects at sea level. J. Physiol.(London) 355: 161-175. Gibson, Q.H. and F.J.W. Roughton (1957). The kinetics and equilibrium of the reactions of nitric oxide with sheep hemoglobin. J. Physiol.(London) 136: 507-526. Graham, B. L., J.T. Mink and D.J. Cotton (1985). Effect of breath-hold time on DLco (SB) in patients with airway obstruction. J. Appl. PhysioL 58: 1319-1325. Gu~nard H., N. Var~ne and P. Vaida (1987). Determination of lung capillary blood volume and membrane diffusing capacity in man by the measurements of NO and CO transfer. Respir. Physiol. 70: ! 13-120. Hammond, M. D., G. E. Gale, K. S. Kapitan, A. Ries and P.D. Wagner (1986). Pulmonary gas exchange in humans during exercise at sea level. J. Appl. Physiol. 60: 1590-1598. Harris, P. and D. Heath (1987). The human pulmonary circulation, 3rd Edn. Edinburgh, Churchill Livingstone. Holland, R. A. B. (1969), Rate at which CO replaces 0 , from O,Hb in red cells of different species. Respir. Physiol. 7: 43-63. Lewis, B.M., T. Lin, F.E. Nee and R. Komisaruk (1978). The measurement of pulmonary capillary blood volume and pulmonary membrane diffusing capacity in normal subjects; the effects of exercise and position. J. Cfln. Invest. 37: 1061-1070. Mahler, D.A. and J. Loke (1981). Lung function after marathon ru~ning at warm and cold ambient temperature. Am. Rev. Respir. Dis. 124: 154-157. Manicr, G., F. Wickers, A.M. Lomenech, G. Cazorla and R. Roudaut (1990). Echocardiographic assessment of myocardial performance after prolonged strenuous exercise. Submitted to European Heart Journal. Maron, N. B., L.H. Hamilton and M.G. Maksud (1979). Alterations in pulmonary function consequent to competitive marathon running. Med. Sci. Sports I I: 244-249. Meyer, M. and J. Piiper (1989). Nitric oxide (NO), a new test gas for study of alveolar-capillary diffusion. Eur. Respir. J. 2: 494-496. Miles, D.S., C.E. Doerr, S.A. Schonfeld, D.E. Sinks and R.W. Gotshall (1983). Changes in pulmonary diffusing capacity and closing volume after running a marathon. Respir. Physiol. 52: 349-359. Moinard, J. and H. Gu~nard (1990). Determination of lung capillary blood volume and membrane diffusing capacity in patients with COLD using the CO-NO method. Eur. Respir. J. 3: 318-322. Montcil, C. (1980). Solubilitds des gaz dans les liquides. In: Techniques de ring6nieur. Paris, K2 Constantes. pp. 630-639. Piiper, J. and P. Scheid (1980). In: Pulmonary Gas Exchange, edited by J.B. West, Vol. i, Chap. 5, New York, Academic Press, pp. 132-173. Reeves, J.T., J.A. Dempsey and R.F. Grover (1988). Pulmonary circulation during exercise. In: Lung Biology in Health and Disease, Vol. 38. Pulmonary vascular physiology and pathophysiology, edited by K. Weir and J.T. Reeves, New York, Marcel Dekker.

DIFFUSION LIMITATION IN MARATHON RUNNERS

153

Roughton, F.J. and R.E. Forster (1957). Relative importance of diffusion and chemical reaction rates in determining rate of exchange of gases in the human lung, with special reference to true diffusing capacity of pulmonary membrane and volume of blood in the lung capillaries. J. Appi. Physiol. 1i: 290-302. Saltin, B. and L. Hermansen (1966). Esophageal, rectal and muscle temperature during exercise. J. AppL Physiol. 21: 1757-1762. Van der Walt, W. H. and C. H. Wyndham (1973). An equation for prediction ofenergy expenditure of walking and running. J. Appl. Physiol. 34: 559-563. Wagner, P.D., J. R. Gillespie, G.L. Landren, M.R. Fedde, B.W. Jones, R.M. Debowes, R.L. Pieschl and H.H. Erikson (1999). Mechanism of exercise-induced hypoxemia in horses. J. Appl. Physiol. 66: 1227-1232. Younes, M., Z. Bshouty and J. All (1987). Longitudinal distribution of pulmonary vascular resistance with very high pulmonary blood flow. J. AppL Physiol. 62: 344-358.