Does peak inspiratory flow contribute to setting V̇O2max?

109

Respiration Physiology, 95 (1994) 109-118 © 1994 Elsevier Science Publishers B.V. All rights reserved. 0034-5687/94/$07.00

RESP 02090

Does peak inspiratory flow contribute to setting V o 2 m a x 9. A test of symmorphosis S t a n L. L i n d s t e d t a, R i c k e y

G. Thomas a a n d

David

E.

Leith

aPhysiology and Functional Morphology Group, Department of Biological Sciences, Northern Arizona, University Flagstaff~ A Z 86011-5640 USA; b 5025 Lakewood Drive Manhattan, K S 66502, USA (Accepted 7 September 1993) Abstract. Symmorphosis predicts that animal design is optimized in such a way that structure 'satisfies but does not exceed' functional requirements. To provide one test of this hypothesis, we examined peak inspiratory flow and its relation to maximum oxygen uptake in humans. We measured maximal forced (peak) inspiratory flow ('¢! ..... ) and maximum oxygen uptake ('~o2m~x) via cycle ergometry in well trained (.Vo2..... > 6 5 ml O 2 . k g - l . m i n -I) and untrained (~'o2 .... < 4 5 ml O 2 . k g - l . m i n 1) male subjects. Tests of VIm~~ and peak oxygen uptake ('qO2peak) were made while the subjects were breathing through inspiratory orifices differing in area. "¢I..... varied as an identical function of orifice diameter in both groups of subjects. However, "qo_,pCakwas more sensitive to decreasing orifice diameter in trained endurance athletes than it was in untrained individuals. The diameter of the largest orifice that caused a reduction in oxygen uptake was over two times larger for trained than for untrained subjects, corresponding to about a four-fold difference in resistance at any flow rate. These results suggest that the structures setting "qI...... (airway resistance and inspiratory muscle strength) are not matched to oxygen demand (Vo_,..... ) in humans. While these structures seem to be 'over-built' and hence do not likely contribute to setting the limits to aerobic performance in most humans, they may be among the primary limiting factors in the most elite endurance athletes.

Flow, maximum inspiratory; Inspiration, maximal flow vs maximal 02 uptake; Mammals, humans; Symmorphosis, peak inspiratory flow, maximum 02 uptake

Symmorphosis, the concept of optimal design. Galileo, Cuvier, and others suggested long ago that animals are built with a fundamental balance between structure and function (as reviewed in Lindstedt and Jones 1987). As articulated under the term 'symmorphosis', the idea is that animal structures are built and maintained with capacity sufficient to 'satisfy but not exceed' functional requirements (Weibel and Taylor, 1981). Because structure and function are independently quantifiable, symmorphosis is a testable prediction of animal design. As a null hypothesis, it predicts that the structure available to support a given function is just the amount needed, never more nor less. Symmorphosis predicts proportionality between the amount of a structure and the

Correspondence to." Stan L. Lindstedt, Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640. Tel.: (602)523-7524; Fax:-7500.

110 function it supports (dashed line in Fig. 1). More structure is found in association with proportionally greater function; available structure always and only satisfies functional demand. Of course there is another possibility: the available structure may exceed the minimum needed to satisfy demand. The region of excess structure lies below the line of symmorphosis in Fig. 1. Because a structure in this region would be able to support more function, up to the dashed line, such structures must not be rate limiting. Since function can never exceed the capacity of the structure available to support it, the region above the symmorphosis line is not accessible. We have proposed that structures with the greatest phenotypic plasticity should best conform to symmorphosis (Lindstedt et al., 1988; Lindstedt, 1993). In animals adapted to low demands, the 'fixed' structures will appear to be over-built, but in animals adapted to high demands, both sets of structures will approach conformity with symmorphosis. Such relationships might be revealed by comparisons between sedentary and athletic humans. The hypothesis of symmorphosis has been tested in studies of the mammalian respiratory system. Among wild African ungulates spanning two orders of magnitude in body mass and aerobic capacity, there is a close match between maximum oxygen uptake (Vo~max) and quantitative measures of capillary and mitochondrial structure in

1 O0

J s i 1 s /

.....

Symmorphosis

,"

~" 75

i J

,B ¢#d

I / i

Phenotypically "plastic " ," structure e.g., m i t o c h o n d r / i

Im

=

50

O

,

C

u.

"Fixed" structure e.g., airway resistance

,, 25

i i i i i

0

I

I

J .

0

I

.

.

.

i

.

.

.

.

t

,

,

,

,

I

25 50 75 S t r u c t u r e (relative)

,

,

,

,

100

Fig. 1. Symmorphosis predicts an optimal structure-function match. Thus, the magnitude of any structure present should fall on a single line (the dashed line) perfectly matched to the function that it supports. Above this line is an area of structural impossibility; function can never exceed the minimum amount of structure necessary to support it. However, if the measure of structure falls below the line, the structure is present in excess of functional requirements. Those structures with the greatest phenotypic plasticity likely remain on the line as they adapt up or down acutely as demands dictate. In contrast those structures with the least phenotypic plasticity must be built with enough apparent excess capacity to accommodate changes in those structures that can adapt. As demand state increases these structures will approach the line vertically.

111 skeletal muscle (Hoppeler et al., 1980). So convincing were these correlations that an apparent mismatch in the scaling of pulmonary diffusing capacity and Vo2ma× was presented as a paradox (Weibel et al., 1981); why does the former increase more than the latter as body size increases? In comparative studies of animals with the same body size but different aerobic capacities (allometric pairs), differences in "~O2maxagain were matched by differences in oxygen delivery and muscle ultrastructure (Taylor et al., 1987). Both studies suggest that in these respects the design of the mammalian respiratory system is consistent with symmorphosis. Ventilation and the limits to iZo,ma x. The total volume of mitochondria, and thus presumably the aerobic power, of the diaphragm varies allometrically in proportion to ~'o . . . . in mammals from shrews to cattle (Hoppeler et al., 1980; Mathieu et al., 1981). Likewise, airway conductance and tracheal dimensions appear to scale in proportion to metabolic rate (Stahl, 1967; Leith, 1983). Thus there is evidence that both inspiratory muscles and airways are 'tuned' to oxygen demands across mammalian species, as predicted by symmorphosis. But within a species, the situation may be different. While Vozma x is not thought ordinarily to be limited by ventilation, maximum exercise intensities exceed those at which "~o. . . . is reached, and some humans reach the upper limits to ventilation (i.e., the envelope of the maximal volitional flow-volume loop) 'coincident with the termination of maximum exercise' (Dempsey and Johnson, 1990; Johnson et al., 1992); thus the question remains open whether the structures used for ventilation play a role in limiting human exercise tolerance. The upper limits to ventilation depend, in part, on maximum inspiratory flows ('~I.... ), and these in turn depend on both airway size (resistance) and the force-velocity characteristics of the inspiratory muscles (IM) (Agostoni and Fenn, 1960). We know of no evidence that whole-body exercise training elicits airway adaptations, and have found no direct studies of its effects on IM force-velocity behavior. IM are capable of substantial increases in strength and endurance in response to intense specific ventilation training (Leith and Bradley, 1976), but whole-body endurance training likely provides less intense ventilatory muscle training and therefore elicits equivocal responses (Robinson and Kjeldgaard, 1982; Moore and Gollnick, 1982; Metzger and Fitts, 1986; Fregosi et al., 1987; Dempsey et al., 1988; Powers et al., 1992; Uribe et al., 1992).

Trained endurance athletes (A) differ from inactive sedentary subjects (S) in their much higher rates of oxygen uptake and the exercise intensities they are able to maintain. These performance differences are associated with greater ventilation. To support their substantial ventilatory demands, A may be expected to have significant adaptations of the more plastic structures used for inspiration. Thus, they should be capable of greater ~/I . . . . and peak inspiratory power, and likely will demonstrate different IM force-velocity behavior than S. If, on the other hand, S can achieve the same IM power and VI..... as A, then according to symmorphosis, the Experimental rationale.

112 relevant structures are over-built in S. To test the hypothesis of inspiratory muscle symmorphosis, we m e a s u r e d "QImax and 9O2max in 6 A and 6 S while their breathing was restricted through a series of graded inspiratory resistances (orifices differing in diameter).

Materials and methods

Subjects. The study was approved by the institutional review board and all subjects gave written consent. After a progressive exercise test on a cycle ergometer, 12 healthy male subjects were selected to participate in this study based on their levels of maximum oxygen uptake. The subjects (19-28 years old) were divided into two groups: Vo:n~a×>65 ml O2"kg l.min-I in A and <45 ml O2.kg-l.min -I in S. All of the subjects in A were competitive cyclists on a local semi-professional team. Each had been cycling regularly for years and all maintained training programs during the course of this study. The subjects in S were students with no history of or current participation in regularly scheduled physical activities. Graded resistors. Seven orifices with diameters = 2.12, 1.69, 1.30, 0.87, 0.60, 0.42, and 0.32 cm (resistors R 1_ 7, respectively) were constructed to fit airtight in the inspiratory airstream during ~/Ima x a n d Vo,-..... tests. These orifices were made from plastic discs 0.2 m m thick. During use, each was sealed within a cylinder (3.2 cm ID, 22 cm long) with pressure taps (15 gauge Luer stub adapters) fitted flush with the inside wall upstream and downstream from the orifice. When there is flow through the orifice, the pressure across each orifice will vary as a 2nd order geometric function of the flow rate. We measured the pressure drop across each R through a broad range of air flow rates with a differential pressure transducer calibrated daily with a water manometer, and a mass flow meter calibrated by timing the displacement of a known volume of water at a constant flow rate. The system with no added R was designated R o. Peak inspiratory flow (i/Zm~,.,).

~/Ima x w a s taken as the tangent slope of a record of volume in time at the volume midpoint of a maximal forced inspiratory vital capacity maneuver, measured with a Collins 13.5 liter respirometer and corrected to body temperature, saturated with water vapor (BTPS). No correction was made for the departure of the intrathoracic pressure from atmospheric, because maximum inspiratory flow varies little with thoracic gas volume in the mid vital capacity range. Subjects repeated this maneuver at least 3 times with no added resistance (R0) and with each of the orifices, R1-RT, in place, and the peak values at each R were averaged. Peak pressures at the airway opening (Pao) with a given R were determined from measurements of peak inspiratory flow and that resistor's specific pressure-flow curve. The data were used to construct isovolume maximum inspiratory pressure-flow (MIPF) curves for the respiratory system (Fig. 2), and to calculate peak instantaneous power ('VVo) applied to the orifice by the respiratory system ('~¢o = Pao'VI ..... ). Maximum inspira-

113

175

I

I

o

I

I

I

I

I

Pao=-17.1(Flow)

150

I

I

+ 136.6r=0.99

16.3(Flow) + 1 2 9 . 2 r = 0 . 9 9

125 0 N

100

£ o

75

"I-

o

~"

50 25

O

,

0

I

1

~

I

2

3

4

i

I

5

,

I

,

6

I

7

,

I~",

8

I

9

,

1C

F l o w (L. s 1 ) Fig. 2. Maximum peak inspiratory pressure-flow (MIPF) pairs are shown for A (open symbols) and their S (filled symbols) during maximum forced inspiratory maneuvers through graded resistors. Neither the slope nor intercept of these two regression lines are different (see text). Pao is pressure measured at the airway opening and flows are BTPS. Data shown are group means + 1 SEM.

tory pressure ( M I P ) was calculated as the pressure intercept of straight lines fitted to the data by least squares regression. Maximum oxygen uptake. "~'O,m,× was taken as the plateau value of ~'o2 observed during progressive cycle exercise to exhaustion at workloads that increased by 40 watts every 2 rain. The greatest external power output (~¢~×t) that each subject maintained for two minutes (with enthusiastic encouragement) is W,× t. Subjects breathed through a mouthpiece and non-rebreathing valve, and "~'o2 was measured with an open-flow system (Rayfield metabolic cart, operated according to the manufacturer's instructions). The average of two determinations is reported. Exercise with added resistance. After ~'o . . . . and "£Vext had been established, subjects began exercise tests with each of the R in the inspiratory path, to quantify the effects of added resistance on peak Vo~ (Vo.peak)' After 5 rain warm up at no more than 80 watts, subjects cycled for 2 min each at 40 watts below "VVe~t, at W~xt, and if possible at 40 watts above "v;gext, again with vigorous encouragement to finish the test. The order by which the different sized orifices were presented to the subjects was systematically varied to control for potential bias due to familiarity with the apparatus. The

I14

,6

I

I

I

I

1.4 I,i,,I

N +1

I

[~~

A

1.2 1.0 0.8

_= >

0.6

> ',~

0.4

""

0.2 0.0

'

'

'

.'

Wext ~/02rnaxVl max W0

'

NiP

Fig. 3. Relative values of key variables for inspiratory flow and oxygen uptake are shown for A (open bars) and S (filled bars). Peak inspiratory flow (VI..... ), calculated maximum inspiratory pressure (MIP, see methods) and maximal instantaneous power exerted on the orifice (Xi/'o,.the product of peak Pao and 91.... ) are not different between the two groups. However, Vo2..... and W~ t differ significantly. Error bars represent _+2 SEM.

subjects were unaware of the size of the orifice before starting exercise. Results were compared using common statistics: ~/L~m~with R 0 was compared with a Student t-test, and regression of Pao and V~..... were established by least squares and the slopes were analyzed for homogeneity by analysis of covariance with :t< 0.05.

Results M a x i m u m external p o w e r o u t p u t (t~ext) was 43~o greater in A t h a n in S (399 + 6 W a t t s [~+SEM] in A v s 2 7 9 + 2 0 w a t t s in S, P = 0 . 0 1 3 ) a n d Vo:m~x was 49°/2, greater ( 6 7 . 6 + 0 . 9 , v s 4 5 . 3 + 1.l ml O 2 " k g - l - m i n 1, P = 0.018). W e failed to find any differe n c e s a m o n g any o f the subjects in their flow c h a r a c t e r i s t i c s d u r i n g f o r c e d i n s p i r a t i o n

Fig. 4. (A) Peak minute v o l u m e s ~Q~peak during aerobic workload tests fell as a function of orifice diameter in all subjects. (B) Peak oxygen uptake is shown as a function of orifice diameter during aerobic workload tests. Maximum oxygen uptake (~'o2,,~,,) was 49°; greater in A (open symbols) than S (filled symbols). The diameter of the orifice that caused a reduction in Vo2r,e~k was much greater in A (13 ram) than

115

200 A .

0

150

{

Athletes Sedentary

= .c m

E ~-- 100 0.

.> 50

R8 R 5 R4 I

0

,

5

,

R3 I

~

,

R2 I

~

R1 I

Ro r

10 15 20 Orifice Diameter (mm)

25

80 B.

70

E.

o

60

...:i"..

'~

~ ..'1'.'" Region o f Critical Resistance

"N 0

50

.-i.. "'" ./.

,,,

40

..~.."~

..;.

",'~

~0

0.

.>o

Region of Critical Resistance

"t

t"

30

R 6 R5 R 4 O

i

0

I,

5

.

.

,

R3 ~

R2 i

,

,

R1

R0

i

10 15 20 Orifice Diameter (mm)

25

in S (7.5 mm). We plotted all points below these critical resistor areas on a single regression line (Vo2pe~k (ml O2'kg -1"rain I)= 4.8'(Orifice diameter)+ 5.4, r = 0.894). This line is plotted with its 95% confidence limits.

116 through inspiratory resistors. Neither peak f l o w s (VImax) nor peak inspiratory power ('~o) differed between A and S. Comparing A and S: '¢1..... ~ with R o was 9.1 _+0.58 L's -~ in A, 9.0 + 0.30 L's -1 in S; P = 0.367); peak Wo in A, 23.5 + 1.8 watts in S, 21.6 _+2.7 watts (P = 0.354); and pressure intercepts of the MIPF curves: A and S, 137 cm H 2 0 and 129 cm H20, respectively (P = 0.434) (Fig. 3). However, unlike the response of the subjects to the ~'hm',~× maneuver, there was a significant difference between the two groups in their response to cycle ergometry when the inspiratory resistors were in place. Adding RI or R 2 did not reduce "qo..... in any of the subjects, but increasing resistance above R 3 reduced Vo:p~k significantly in A. Neither R 3 n o r R 4 reduced the V%v~, k of S, however R s and R 6 did. None of the subjects was able to complete the exercise test when breathing through R 6 or R7 (Fig. 4). Ventilation (~'E) dropped during exercise in all subjects when they exercised while breathing through orifices of critical sizes. When ventilation-limited "¢%p~,k (i.e., ;qo. while breathing through R 3 and above in A and R5 and R6 in S) is plotted against orifice diameter, data from all subjects can be fit by a single line ('qO, p~k = 4.8"(orifice diameter) + 5.4, Fig. 4). Extrapolating this "qo;/(orifice diameter) line to the diameter of R 2

10

.

.

.

.

i

.

.

.

.

i

.

.

.

.

i

.

.

.

.

i

.

.

.

90

.

i

A

80 v

8 o "

"~" 70

6

"

o

E

D.

'T,

60

0

50

"~

4

E

_= 2

40

0 0

........................ 5 10 15

•

o~

30 20

25

Orifice D i a m e t e r (ram) Fig. 5. When peak flows and Vo,pe.k are plotted together as a function of orifice diameter three inflection points, or critical orifice diameters, become apparent. First, VI.... of all subjects falls while breathing through an orifice smaller than 16.9 mm. Second, Vo2veak of the trained subjects falls if the orifice diameter is below ., 13 mm and finally, ¥O,.v~-k of the untrained subjects drops when the orifice diameter is below 7.5 mm. If the regression line describing "~o2p~.,k as a function of orifice diameter is extrapolated to the size of the orifice necessary to cause a reduction in VInla~, 16.9 mm, this equation predicts a ~'o2 ...... of 86 ml O 2 k g ~-min-1 close to the highest recorded ~'o_~.... values among elite human endurance athletes.

117 (the largest orifice that resulted in a detectable reduction in ~QImax in all subjects), the Y intercept of this line (Vo,ma~ at a diameter of R 2 is about 86 ml O2"kg-l'min -~ (Fig. 5). Discussion

Peak inspiratory flow performance was indistinguishable comparing our trained endurance athletes and their sedentary controls, despite the great differences in their "~o_,m~x and exercise habits. Specifically, VIm~X, the extrapolated pressure at zero flow (MIP), maximal inspiratory pressure-flow curves and the peak power applied to orifices ('ViJo) were not different between the groups. Thus we found no evidence of structural differences between the two groups, or that chronic whole-body exercise endurance training elicits adaptations of either airways or inspiratory muscles. Likewise, we found no evidence for ventilatory limitation of exercise capacity, even among the most highly trained athletes. In all subjects, dropped when breathing through an orifice of critical diameter. We interpret the fall in "~O,_peakwith decreasing orifice diameter to be attributable to an inability to sufficiently widen the a-v 0 2 content to compensate for the apparent arterial hypoxemia. Each subject was able to achieve his baseline "~O2ma× during exercise with added resistances small enough to reduce voluntary maximum inspiratory flow (R 2 in A and R 4 in S). Thus all our subjects had inspiratory structures in excess of the functional demands of Vo21~..... the excess being far greater in sedentary than in athletic subjects. Our working hypothesis of symmorphosis can be rejected for the structures setting inspiratory flow in humans. Vo:m~,x in our athletic subjects was well below the highest values reported in humans, ~85 ml O2'kg-l'min -1 (Saltin and Astrand, 1967). However, extrapolation of the V%pe~,k/orifice diameter line suggests that collectively, without adaptation, the inspiratory structures could support ventilation sufficient for a "~o,,..... of 85 ml 02' kg 1.min- 1 in all the subjects included in this study. Perhaps only in rare circumstances (i.e., the most elite of human endurance athletes) do inspiratory structure-function relationships approach symmorphosis. Possibly here one could demonstrate training effects in the inspiratory muscles, but not in the phenotypically less plastic airways. For the remainder of our species, inspiratory structures' capacities exceed the demands of aerobic exercise. We suggest that less plastic structures are built to accommodate the greatest functional demands likely to be achieved by extreme adaptations of the more plastic structures in a system; but, if the adaptive capacities of the latter are great enough, the less adaptable (and, under ordinary circumstances, seemingly over-built) structures may finally set the overall limits of adaptation and performance.

~QO,peak

Acknowledgements.We are grateful for collaboration with Stanley Rasmussen, J ames and William Hokanson, and Micke Eliasson and for stimulating discussions with C. Richard Taylor and Ewald R. Weibel. Supported in part by N I H RO1HL41986, Flinn Foundation Fellowship FG-30389 from the American Heart Association.

118 References Agostoni, E. and W. O. Fenn (1960). Velocity of muscle shortening as a limiting factor in respiratory air flow. J. Appl. Physiol. 15: 349-353. Dempsey, J.A., E. Aaron and B. Martin (1988). Pulmonary function and prolonged exercise. In: Perspectives in Exercise Science and Sports Medicine: Prolonged Exercise, edited by D. R. Lamb and R. Murray Benchmark Press. Indianapolis, In: Benchmark Press, 1: 75-124. Dempsey, J.A. and B.D. Johnson (1990). Adaptations and limitations in the puhnonary system during exercise. Chest 97:81 s-87s. Fregosi, R.F., M. Sanjak and D.J. Paulson (1987). Endurance training does not affect diaphragm mitochondrial respiration. Respir. Physiol. 67: 225-237. Hoppcler, H.O. Mathieu and S.L. Lindstedt (1980). Scaling structural parameters of oxygen consumption in muscle against Vo2 ...... and body mass. In: Exercise Bioenergetics and Gas Exchange: edited by' P. Cerretelli and B.J. Whipp. Amsterdam: Elsevier, 129-136. Johnson, B. D., K.W. Saupc and J.A. Dempsey (1992). Mechanical constraints on exercise hyperpnea in endurance athletes. J. Appl. Physkd. 73: 874-886. Leith, D.E. and M. Bradley' (1976). Ventilatory muscle strength and endurance training. J. Appl. Physiol. 4l: 508-516. Leith, D.E. (1983). Mammalian tracheal dimensions: scaling and physiology. J. Appl. Ph~wiol. 55: 196-2(/0. Lindstedt, S.L. and J. H. Jones (1987). Symmorphosis: the concept of optimal design. In: New Directions in Ecological Physiology, edited by M. Feder, A.F. Bennett, W. Burrgren and R. Huey. Cambridge: Cambridge University Press, 289-309. Lindstedt, S,L., D.J. Wells, J.H. Jones, H. Hoppler and H.A. Thronson Jr. (1988). Limitations to aerobic performance in mammals: interaction of structure and demand. Int. J. Sports Med. 9: 210-217. Lindstedt, S.L., 1993. Oxygen delivery to the tissues in animal athletes: keeping up with demand. In: The Vertebrate Gas Transport Cascade: Adaptations to Environment and Mode of Life. edited by' J. Eduardo, P.W. Bicudo. Boca Raton, FL: CRC Press, pp. 293-3(/3. Mathicu, O., R. Krauer, H. Hoppeler, P. Gehr, S. L. Lindstedt, R. Alexander, C. R. Taylor and E. R. \Veibcl (1981). Design of the mammalian respiratory system: VII. Scaling mitochondrial volume in skeletal nmscle to body mass. Respir. Ptowiol. 44:113-128. Metzger, J.M. and R.H. Fitts (1986). Contractile and biochemical properties of diaphragm: effects of exercise training and fatiguc. J. Appl. Physiol. 60: 1752-1758. Moore, R.L. and P.D. Gollnick (1982). Response of vcntilatory muscles of the rat to endurance training. t~ftiigers Arch. 392: 268-271. Powers, S. K., D. Criswell, F-K. Lieu, S. Dodd and H. Silverman (1992). Diaphragmatic fiber type specific adaptation to endurance exercise. Re.spir. Physiol. 89: 195-207. Robinson. E.P. and J.M. Kjeldgaard (1982). Improvement in ventilatory nmscle function with running. J. Appl. Physiol. 52: 1400-1406. Saltin, B. and P-O. /i~strand (1967). Maximal ox).gen uptake in athletes. J. Appl. Physiol. 23: 353-358. Stahl, W.R. (1967). Scaling of respiratory variables in mammals. J. Appl. Physiol. 22: 453-4611. Taylor, C.R., R.H. Karas, E.W. Weibel and H. Hoppeler (1987). Adaptive wariation in the mammalian respiratory system in relation to energetic demand. ResT~ir. Physiol. 69: 1-28. Uribe, J.M., C.S. Stump, C.M. Tipton and R.F. Fregosi (1992). Influence of exercise training on the oxidative capacity of rat abdominal muscles. Re.v)ir. Physiol. 88: 171-180. Weibel, E.R.. C.R. Taylor, P. Gehr, H. Hoppeler, O. Mathieu and G , M . O . Maloiy (1981). Design of the mammalian respiratory system: IX. Functional and structural limits for oxygen flow. Re.g~ir. Physiol. 44: 151-164. Weibel, E.R. and C.R. Taylor {1981). Design of the mammalian respiratory system. Re.g)ir. Physiol. 44: 1-164.