You are on page 1of 7

J Appl Physiol

90: 926–932, 2001.

Influence of inhaled nitric oxide on gas exchange during


normoxic and hypoxic exercise in highly trained cyclists

A. WILLIAM SHEEL, MICHAEL R. EDWARDS,


GARTH S. HUNTE, AND DONALD C. MCKENZIE
Allan McGavin Sports Medicine Center and School of Human Kinetics, The University
of British Columbia, Vancouver, British Columbia, Canada V6T 1Z3
Received 10 December 1999; accepted in final form 13 September 2000

Sheel, A. William, Michael R. Edwards, Garth S. fusion (V̇A/Q̇) inequality, and 4) diffusion limitation.
Hunte, and Donald C. McKenzie. Influence of inhaled Venoarterial shunt has been identified as a minor
nitric oxide on gas exchange during normoxic and hypoxic contributor, and relative alveolar hypoventilation has
exercise in highly trained cyclists. J Appl Physiol 90: a controversial role in the pathophysiology of EIH (6).
926–932, 2001.—This study tested the effects of inhaled V̇A/Q̇ relationships have been shown to worsen with
nitric oxide [NO; 20 parts per million (ppm)] during normoxic
and hypoxic (fraction of inspired O2 ⫽ 14%) exercise on gas
exercise (10), and, during maximal exercise, ⬃60% of
exchange in athletes with exercise-induced hypoxemia. the widened A-aDO2 can be explained by V̇A/Q̇ mis-
Trained male cyclists (n ⫽ 7) performed two cycle tests to match (14). Pulmonary interstitial edema may explain
exhaustion to determine maximal O2 consumption (V̇O2 max) both V̇A/Q̇ inequality and diffusion limitations (11, 26).
and arterial oxyhemoglobin saturation (SaO2, Ohmeda Biox This would be expected to negatively influence gas
ear oximeter) under normoxic (V̇O2 max ⫽ 4.88 ⫾ 0.43 l/min exchange in the lung by lowering the compliance of the
and SaO2 ⫽ 90.2 ⫾ 0.9, means ⫾ SD) and hypoxic (V̇O2 max ⫽ alveoli and by compressing small blood vessels, result-
4.24 ⫾ 0.49 l/min and SaO2 ⫽ 75.5 ⫾ 4.5) conditions. On a ing in nonuniform airflow and blood flow distribution
third occasion, subjects performed four 5-min cycle tests, in the lungs (11). However, present techniques have
each separated by 1 h at their respective V̇O2 max, under failed to provide precise quantification of extravascular
randomly assigned conditions: normoxia (N), normoxia ⫹ NO lung water or identification of the specific mechanisms
(N/NO), hypoxia (H), and hypoxia ⫹ NO (H/NO). Gas ex-
responsible for the development of interstitial edema
change, heart rate, and metabolic parameters were deter-
mined during each condition. Arterial blood was drawn at (6).
rest and at each minute of the 5-min test. Arterial PO2 (PaO2), Inhaled nitric oxide (NO) is a potent and selective
arterial PCO2, and SaO2 were determined, and the alveolar- pulmonary vasodilator (8, 9). Inhalation of NO has
arterial difference for PO2 (A-aDO2) was calculated. Measure- been used effectively to treat several respiratory disor-
ments of PaO2 and SaO2 were significantly lower and A-aDO2 ders in humans (1, 15, 16) but has been shown to be
was widened during exercise compared with rest for all detrimental in chronic obstructive pulmonary disease
conditions (P ⬍ 0.05). No significant differences were de- (1). The positive effects of NO are explained by a
tected between N and N/NO or between H and H/NO for PaO2, preferential distribution of inhaled NO to well-venti-
SaO2 and A-aDO2 (P ⬎ 0.05). We conclude that inhalation of lated alveolar units, a reduction in the dispersion of
20 ppm NO during normoxic and hypoxic exercise has no ventilation distribution, lowering of pulmonary vascu-
effect on gas exchange in highly trained cyclists. lar pressures, and improved gas exchange. Inhalation
exercise-induced hypoxemia; pulmonary edema; ventilation- of NO has also been utilized to exert a beneficial effect
perfusion inequality; diffusion disequilibrium on arterial oxygenation in mountaineers with high-
altitude pulmonary edema (27). Although the mecha-
nisms of EIH remain debatable, there is some indirect
SOME HIGHLY TRAINED MALE ENDURANCE athletes experi- evidence to support the development of transient inter-
ence decreases in arterial PO2 (PaO2) and arterial oxy- stitial edema as a mechanism for a widened A-aDO2 (3,
hemoglobin saturation (SaO2) and a widened alveolar- 23). We sought to test the hypothesis that inhaled NO
arterial difference for O2 (A-aDO2) during heavy would improve oxygenation during exercise in athletes
exercise (5, 13). The cause and significance of exercise- with EIH by reducing A-aDO2.
induced arterial hypoxemia (EIH) has been the topic of
a considerable research effort; however, the mecha- METHODS
nism(s) responsible remains controversial. Four poten- Subjects. Highly trained male cyclists were recruited to
tial factors have been identified: 1) venoarterial shunt, participate in this study (n ⫽ 8). One subject was forced to
2) relative alveolar hypoventilation, 3) ventilation-per- withdraw due to difficulties with placement of the arterial

Address for reprint requests and other correspondence: D. C. The costs of publication of this article were defrayed in part by the
McKenzie, Allan McGavin Sports Medicine Center, 3055 Wesbrook payment of page charges. The article must therefore be hereby
Mall, Vancouver, BC, Canada V6T 1Z3 (E-mail: kari@interchange. marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
ubc.ca). solely to indicate this fact.

926 8750-7587/01 $5.00 Copyright © 2001 the American Physiological Society http://www.jap.org
INHALED NO DURING EXERCISE 927

Fig. 1. Overview of experimental protocol, repeated for each gas mixture condition of normoxia (N), normoxia ⫹ 20
parts per million (ppm) nitric oxide (N/NO), hypoxia (H), and hypoxia ⫹ 20 ppm nitric oxide (H/NO). S, sample for
blood and cardiorespiratory variables taken after breathing test gas mixture at rest for 5 min and during each
minute of 5-min maximal cycling; V̇O2 max, maximal O2 consumption.

catheter; therefore, all data are reported for n ⫽ 7. This fatigue. During part 1, cycle ergometry subjects inspired
investigation was divided into two parts. Subjects who met compressed air [fraction of inspired O2 (FIO2) ⫽ 20.93%]. The
the inclusion criteria in part 1 participated in part 2. Inclu- air was delivered from a large cylinder through a closed
sion criteria were 1) normal spirometry, that is, no history of container of water for humidification and then into a large
asthma or cardiorespiratory disease, 2) maximal O2 con- meteorological balloon, which acted as a reservoir for in-
sumption (V̇O2 max) ⱖ60 ml 䡠 kg⫺1 䡠 min⫺1 and/or 5 l/min, 3) spired air. Those who met the inclusion criteria returned on
development of EIH (maximal exercise SaO2ⱕ 91.0%), and 4) a separate day at least 72 h later to perform another maximal
between the ages of 18 and 40 yr. Before testing was started, cycle ergometry test under hypoxic conditions (FIO2 ⫽
subjects received a verbal description of the experiment and 14.00%). This FIO2 has previously been used to accentuate
completed a written, informed consent form. This study was decreases in SaO2 in exercising trained men (17). Expired
approved by the Clinical Screening Committee for Research gases, heart rate, and SaO2 were determined during the
and Other Studies Involving Human Subjects of the Univer- hypoxic exercise session as described for the normoxic condi-
sity of British Columbia. tions.
Preliminary screening: part 1. Subjects reported to the Inhaled NO: part 2. After completion of both V̇O2 max tests
Applied Physiology Laboratory in the Allan McGavin Sports (normoxic and hypoxic), subjects who met the inclusion cri-
Medicine Center (Univ. British Columbia), having refrained teria returned on a separate day at least 72 h later. A time
from exhaustive exercise for 24 h, abstained from ingestion of line describing the experimental protocol is shown in Fig. 1.
food or fluid for 4 h, except for water, and abstained from Subjects were randomly assigned and blinded to each of the
alcohol and caffeine for 12 h. Subjects’ weight and height four following conditions: 1) normoxia (N), 2) normoxia ⫹ NO
were measured and recorded. Both spirometry and pulmo- (N/NO), 3) hypoxia (H), and 4) hypoxia ⫹ NO (H/NO). Par-
nary diffusion measurements for carbon monoxide (DLCO) ticipants performed a 10- to 15-min cycling warm-up at a
were collected using the same commercial apparatus (Collins self-selected workload and then sat quietly on the cycle
DS/PLUS II, Braintree, MA). DLCO was determined using the ergometer for 5 min, at which point resting data were ob-
single-breath method. Before DLCO and spirometry measure- tained. Cycling intensity was then manually increased over 1
ments were made, subjects sat and rested for 30 min to min to 100% of their respective maximum normoxic or hy-
ensure a resting heart rate and pulmonary capillary blood poxic workload as determined in part 1. Subjects cycled at
volume. this intensity for 5 min, and cardiorespiratory variables were
V̇O2 max was determined using an incremental test to ex- recorded at each minute in the same fashion as for part 1.
haustion on an electronically braked cycle ergometer (Quin- Arterial blood samples were drawn at rest and at each min of
ton Excalibur, Lode, Groningen, The Netherlands). Subjects the 5-min test. After each test condition, subjects cycled
pedaled at a self-chosen cadence at a progressing workload, easily (30–50 W) for 10 min and then rested for 50 min before
which started at 0 W and increased 30 W/min. Subjects commencing the next test condition. A physician was in
inspired through an air flowmeter (Vacumetrics model attendance at all times and was responsible for the safety of
17150, Ventura, CA) using a two-way nonrebreathing valve the subjects during the study.
(Hans-Rudolph, model 2700B, Kansas City, KS). Expired air
passed into a 5-liter mixing chamber from which gas samples Table 1. Descriptive and resting pulmonary
were analyzed at a rate of 300 ml/min for O2 and CO2 (S-3A function data
O2 analyzer and CD-3A CO2 analyzer, Applied Electrochem-
istry, Pittsburgh, PA). Expired gases and minute ventilation Age, yr 28.9 ⫾ 3.9
(V̇E) were recorded using a computerized system (Rayfield, Height, cm 181.4 ⫾ 7.5
Waitsfield, VT). Gas analyzers were calibrated with gases of Mass, kg 74.7 ⫾ 6.6
known concentration, and the air flowmeter was calibrated FVC, liters 5.59 ⫾ 0.81 (102 ⫾ 14)
by passing 100 liters of air through the system. Heart rate FEV1, liters 4.58 ⫾ 0.73 (102 ⫾ 15)
FEF25–75%, l/s 4.52 ⫾ 0.91 (102 ⫾ 18)
was recorded every 15 s using a portable heart rate monitor
FEV1/FVC, % 80.54 ⫾ 5.82 (98 ⫾ 7)
(Polar Vantage XL, Kempele, Finland). SaO2 was measured FEFmax, l/s 9.95 ⫾ 1.08 (102 ⫾ 17)
by a pulse oximeter (Ohmeda Biox 3740, Louisville, CO), with DLCO, ml 䡠 min⫺1 䡠 Torr⫺1 36.99 ⫾ 5.04* (117 ⫾ 19)
values averaged and recorded every 5 s using a personal VA, liters 5.26 ⫾ 1.42
computer. Before placement of the oximeter sensor to the DLCO/VA 7.33 ⫾ 1.65
pinna of the ear, a topical vasodilator cream (Finalgon,
Boehringer/Ingeheim, Burlington, ON) was applied to in- Values are means ⫾ SD. Actual values are shown, with %predicted
in parentheses. FVC, forced vital capacity; FEV1, forced expired
crease local perfusion. Attainment of V̇O2 max was considered volume in 1 s; FEF25–75%, forced expiratory flow at 25–75% of FVC;
when at least three of the following four were observed: 1) a FEFmax, maximal forced expiratory flow rate; DLCO, pulmonary dif-
plateau in O2 consumption (V̇O2) with increasing workload, 2) fusion capacity for CO; VA, alveolar volume; DLCO/VA, pulmonary
respiratory exchange ratio (RER) ⬎ 1.15, 3) attainment of diffusion capacity for CO/VA. * Significantly different from predicted
90% of age predicted maximal heart rate, and/or 4) volitional (P ⬍ 0.05).
928 INHALED NO DURING EXERCISE

Table 2. Maximal exercise values during maximal caine) and sterile technique and was then secured to the skin.
cycle ergometer tests Adequate collateral circulation via the ulnar artery (Allen’s
test) was estimated before the cannula was inserted. A min-
Normoxia Hypoxia imum volume extension tube, connected in series with two
(FIO2 ⫽ 20.93%) (FIO2 ⫽ 14.00%) three-way stopcocks arranged at right angles, was flushed
V̇O2max, l/min 4.88 ⫾ 0.43 4.24 ⫾ 0.49* with a saline-heparin solution. A rapid response (⬍0.01 s)
V̇O2max, thermistor (18T, Physitemp Instruments, Clifton, NJ) used
ml 䡠 kg⫺1 䡠 min⫺1 65.3 ⫾ 1.6 56.6 ⫾ 5.6* to measure peak arterial blood temperature was inserted
V̇Emax, l/min 175.9 ⫾ 10.2 168.2 ⫾ 11.5 through a Touhy-Borsch heparin lock (Abbott Hospitals,
RER, V̇CO2/V̇O2 1.20 ⫾ 0.02 1.12 ⫾ 0.07* North Chicago, IL). Catheter patency was maintained with a
HRmax, beats/min 188 ⫾ 4 179 ⫾ 3* continuous heparin infusion (1 ml 1:1,000 units in 500 ml
Resting SaO2, % 97.7 ⫾ 0.6 97.0 ⫾ 0.8 normal saline at 3 ml/h). At the onset of sampling, 12 ml of
Lowest SaO2, % 90.2 ⫾ 0.9† 75.5 ⫾ 4.5*† blood was withdrawn, and the final 3 ml was collected in
Power, W 449 ⫾ 39 371 ⫾ 38*
preheparinized plastic syringes. The remaining 9 ml were
Values are means ⫾ SD. FIO2, fraction of inspired O2; V̇O2max, then slowly reinfused. Samples were withdrawn at rest and
maximal O2 consumption; V̇Emax, maximal minute ventilation; RER, at 1-min intervals for the duration of each test (4 condi-
respiratory exchange ratio; V̇CO2, CO2 production; V̇O2, O2 consump- tions ⫻ 6 samples per condition ⫽ 24 samples/subject). Blood
tion; HRmax, maximal heart rate; SaO2, arterial oxyhemoglobin sat- samples were placed on ice until analyzed for H⫹ concentra-
uration. * Significantly different from normoxia (P ⬍ 0.05). † Signif- tion, PO2, PCO2, base excess, and HCO3⫺ (CIBA-Corning 278
icantly different from resting SaO2 (P ⬍ 0.05).
blood-gas system, CIBA-Corning Diagnostics, Medfield, MA).
PaO2 was corrected for temperature and H⫹ concentration.
During N, subjects rested for 5 min and cycled while Temperature increased 0.9 ⫾ 0.1°C (mean ⫾ SD) from rest to
breathing normoxic gas. Condition N/NO consisted of nor- 5 min of exercise across all trials. SaO2 levels were calculated
moxic gas with 20 parts per million (ppm) NO delivered to the based on corrected PaO2. The alveolar gas equation was used
inspiratory tubing. During condition H, subjects rested for 5 to calculated alveolar PO2 and A-aDO2 (20).
min and cycled while inhaling hypoxic gas (FIO2 ⫽ 14.00%); Statistical analyses. Mean values and measures of vari-
condition H/NO consisted of the same hypoxic gas with 20 ability were determined for descriptive, anthropometric, and
ppm NO. During all conditions, the inspired air was deliv- lung function variables obtained during preliminary screen-
ered from a large cylinder through water for humidification ing. Maximal cycle ergometry data from part 1 were com-
and then into a large meteorological balloon, which acted as pared using t-tests for dependent samples (normoxia vs.
a reservoir before being inspired by the subject. NO was hypoxia). Experimental data were analyzed using a four
delivered at the distal end of the tubing while inspired (condition) by six (time) two-way factorial ANOVA with re-
concentrations of O2, NO, and nitrogen dioxide were moni- peated measures on both factors. When sphericity was not
tored continuously during each test condition 5 cm from the assumed, Greenhouse-Geisser P values were utilized. When
subject’s mouth using a commercial apparatus (PulmoNOx significant F ratios were observed, Scheffé’s test was applied
II, Pulmonox, Tofield, AB). The NO-delivery system was post hoc to determine where the differences occurred. The
calibrated before each experiment as per the manufacturer’s level of significance was set at P ⬍ 0.05. Statistical power
specifications. The concentration of NO used in the present calculations were performed a priori to estimate an appro-
study (20 ppm) has previously been shown to improve V̇A/Q̇ priate minimum sample size of five. A sample size of seven
distributions, PaO2, and pulmonary vascular resistance in was utilized to ensure sufficient statistical power (1 ⫺ ␤ ⫽
pigs (24), reverse hypoxic pulmonary vasoconstriction (HPV), 0.8).
and redistribute blood flow to better ventilated areas of the
lung in sheep (21). RESULTS
Arterial blood sampling. A 20-gauge arterial catheter was
inserted in the radial artery of the nondominant hand by Physical and maximal exercise data. Descriptive data
percutaneous cannulation using 1% local anesthesia (lido- and resting pulmonary function data are presented in

Table 3. Ventilatory and performance parameters during normoxic and normoxic ⫹ NO 5-min cycle ergometry
Normoxia Normoxia ⫹ NO

Minute Minute

Rest 1 2 3 4 5 Rest 1 2 3 4 5

V̇O2, l/min 0.44 3.70 4.30 4.37 4.37 4.42 0.70 3.93 4.29 4.69 4.82 4.95
⫾ 0.21 ⫾ 0.58* ⫾ 0.59* ⫾ 0.68* ⫾ 0.77* ⫾ 0.81* ⫾ 0.26 ⫾ 0.85* ⫾ 0.81* ⫾ 0.61* ⫾ 0.69* ⫾ 0.62*
V̇O2, 7.64 53.04 58.67 60.05 59.33 60.21 10.81 55.58 60.77 64.70 67.42 69.37
ml 䡠 kg⫺1 䡠 min⫺1 ⫾ 3.08 ⫾ 9.23* ⫾ 9.14* ⫾ 9.54* ⫾ 10.44* ⫾ 11.17* ⫾ 2.37 ⫾ 11.39* ⫾ 12.12* ⫾ 10.27* ⫾ 11.09* ⫾ 9.77*
V̇E, l/min 16.07 120.27 150.08 156.93 157.68 160.01 14.07 119.39 146.22 155.30 157.86 160.21
⫾ 4.98 ⫾ 29.20* ⫾ 8.73* ⫾ 12.98* ⫾ 14.59* ⫾ 15.23* ⫾ 4.77 ⫾ 23.51* ⫾ 11.21* ⫾ 12.21* ⫾ 13.37* ⫾ 16.07*
RER 0.87 0.93 1.06 1.06 1.06 1.03 0.84 0.88 1.00 1.02 0.99 0.98
⫾ 0.16 ⫾ 0.21 ⫾ 0.18* ⫾ 0.18* ⫾ 0.18* ⫾ 0.16* ⫾ 0.04 ⫾ 0.03 ⫾ 0.06* ⫾ 0.09* ⫾ 0.11* ⫾ 0.11*
HR, beats/min 89 167 174 177 179 180 93 165 173 177 179 181
⫾ 17 ⫾ 6* ⫾ 3* ⫾ 3* ⫾ 4* ⫾ 6* ⫾ 13* ⫾ 3* ⫾ 5* ⫾ 6* ⫾ 6* ⫾ 5*
Power, W 0 390 368 353 336 333 0 378 355 341 330 330
⫾ 36.1* ⫾ 32.5* ⫾ 44.6* ⫾ 61.7* ⫾ 60.9* ⫾ 44.8* ⫾ 59.0* ⫾ 66.2* ⫾ 49.0* ⫾ 67.2*
Values are means ⫾ SD. NO, nitric oxide; V̇E, minute ventilation; HR, heart rate. * Significantly different from rest (P ⬍ 0.05).
INHALED NO DURING EXERCISE 929

Table 4. Ventilatory and performance parameters during hypoxic and hypoxic ⫹ NO 5-min cycle ergometry
Hypoxia Hypoxia ⫹ NO

Minute Minute

Rest 1 2 3 4 5 Rest 1 2 3 4 5

V̇O2, l/min 0.52 3.31 3.60 3.76 3.86 4.03 0.48 3.72 3.98 4.19 4.13 4.17
⫾ 0.22 ⫾ 0.42* ⫾ 0.50* ⫾ 0.56* ⫾ 0.64* ⫾ 0.61* ⫾ 0.20 ⫾ 0.56* ⫾ 0.56* ⫾ 0.45* ⫾ 0.52* ⫾ 0.44*
V̇O2, 6.53 45.48 50.13 53.05 53.96 55.98 5.50 49.83 54.06 56.38 55.77 55.93
ml 䡠 kg⫺1 䡠 min⫺1 ⫾ 2.55 ⫾ 6.42* ⫾ 6.21* ⫾ 3.06* ⫾ 3.69* ⫾ 2.68* ⫾ 1.34 ⫾ 7.03* ⫾ 6.60* ⫾ 4.40* ⫾ 5.37* ⫾ 5.51*
V̇E, l/min 14.10 126.62 138.62 147.24 156.02 157.51 21.22 134.21 149.78 153.71 153.65 155.63
⫾ 4.15 ⫾ 16.44* ⫾ 8.74* ⫾ 17.74* ⫾ 22.94* ⫾ 20.52* ⫾ 9.29 ⫾ 17.48* ⫾ 26.20* ⫾ 23.33* ⫾ 24.10* ⫾ 21.76*
RER 0.94 1.04 1.12 1.09 1.08 1.06 0.95 1.01 1.07 1.05 1.02 0.99
⫾ 0.12 ⫾ 0.11* ⫾ 0.04* ⫾ 0.04* ⫾ 0.07* ⫾ 0.08* ⫾ 0.11 ⫾ 0.06* ⫾ 0.09* ⫾ 0.08* ⫾ 0.09* ⫾ 0.09*
HR, beats/min 89 166 171 174 176 177 92 166 171 173 175 177
⫾ 14 ⫾ 4* ⫾ 4* ⫾ 2* ⫾ 3* ⫾ 3* ⫾ 18 ⫾ 6* ⫾ 5* ⫾ 6* ⫾ 6* ⫾ 6*
Power, W 0 326.3 303.9 296.4 288.6 284.3 0 329.3 307.0 278.6 263.3 262.6
⫾ 33.4* ⫾ 40.7* ⫾ 55.4* ⫾ 55.0* ⫾ 52.3* ⫾ 30.3* ⫾ 42.7* ⫾ 29.3* ⫾ 30.8* ⫾ 35.6*
Values are means ⫾ SD. * Significantly different from rest (P ⬍ 0.05).

Table 1. Lung parameters were within normal values cantly lower than both normoxic conditions at all mea-
predicted for men of similar age, height, and weight surement periods. PaO2 values were significantly lower at
except for DLCO, which was significantly elevated. Data 1, 2, 3, 4, and 5 min of exercise compared with rest for all
from normoxic and hypoxic maximal cycle ergometry inspired gas conditions. Similar results were observed for
tests are presented in Table 2. Significant differences SaO2, except that values at 1 and 2 min were not signifi-
were observed between normoxic and hypoxic conditions cantly different from rest under conditions of N and
for V̇O2 max, RER, maximal heart rate, and power output, N/NO (see Fig. 3). A-aDO2 was significantly different at
whereas no significant differences were detected for V̇E. all time periods compared with rest for all inspired gas
From rest to maximal exercise, mean values for SaO2 mixtures, and significant differences were detected be-
dropped significantly under both normoxic (97.7 to 90.2) tween N/NO and H at rest and during all exercise mea-
and hypoxic (97.0 to 75.5) conditions. surements. Arterial PCO2 (PaCO2) was not significantly
Metabolic and power output during 5-min cycling. different between gas conditions but was lower compared
Metabolic and power output data are shown in Tables with rest throughout all exercise for H and H/NO and at
3 and 4. No significant differences were detected be- minutes 3, 4, and 5 for both N and N/NO. No statistically
tween N and N/NO or between H and H/NO for V̇O2, significant differences were detected between N vs. N/NO
V̇E, RER, heart rate, or power output. or between H vs. H/NO for pH, HCO3⫺, or base excess
Arterial blood variables during 5-min cycling. All (Tables 5 and 6).
data are reported in Tables 5 and 6 and Figs. 2 and 3. DISCUSSION
Across all time points, there were no significant differ-
ences for PaO2 between N and N/NO or between H and The present study is the first to systematically ex-
H/NO (see Fig. 2). Both hypoxic conditions were signifi- amine the effects of inhaled NO during normoxic and

Table 5. Blood variables during normoxic and normoxic ⫹ NO 5-min cycle ergometry
Normoxia Normoxia ⫹ NO

Minute Minute

Rest 1 2 3 4 5 Rest 1 2 3 4 5

PaO2, Torr 103.8 93.2 89.4 85.0 85.6 84.2 98.7 85.7 81.6 78.8 78.7 78.0
⫾ 8.2 ⫾ 4.7* ⫾ 6.7* ⫾ 7.2* ⫾ 8.8* ⫾ 9.3* ⫾ 5.2 ⫾ 8.9* ⫾ 8.1* ⫾ 8.2* ⫾ 9.7* ⫾ 10.2*
SaO2, % 98.0 97.4 96.3 95.2 94.7 94.0 97.7 96.6 95.2 94.3 93.7 93.4
⫾ 0.7 ⫾ 0.4 ⫾ 0.4 ⫾ 0.7* ⫾ 1.3* ⫾ 1.8* ⫾ 0.3 ⫾ 0.6 ⫾ 1.3 ⫾ 1.2* ⫾ 1.0* ⫾ 1.0*
PaCO2, Torr 38.8 38.2 37.2 36.2 33.8 31.2 37.2 36.5 37.0 35.4 34.3 32.3
⫾ 3.1 ⫾ 2.4 ⫾ 3.1 ⫾ 3.1* ⫾ 3.8* ⫾ 5.5* ⫾ 4.6 ⫾ 4.7 ⫾ 4.3 ⫾ 3.9* ⫾ 3.7* ⫾ 3.6*
A-aDO2, Torr 2.8 16.5 23.1 26.9 28.9 32.0 11.3 26.3 29.6 34.6 35.4 37.7
⫾ 3.0 ⫾ 8.1* ⫾ 7.0* ⫾ 5.5* ⫾ 7.6* ⫾ 8.1* ⫾ 9.2 ⫾ 14.2* ⫾ 11.9* ⫾ 10.7* ⫾ 11.6* ⫾ 12.8*
pH 7.40 7.38 7.32 7.26 7.23 7.21 7.41 7.39 7.32 7.27 7.26 7.24
⫾ 0.01 ⫾ 0.03* ⫾ 0.04* ⫾ 0.05* ⫾ 0.06* ⫾ 0.07* ⫾ 0.01 ⫾ 0.02* ⫾ 0.02* ⫾ 0.02* ⫾ 0.06* ⫾ 0.08*
HCO3⫺, mmol/l 23.8 22.0 19.2 16.0 14.0 12.2 23.3 22.3 18.8 16.4 15.3 13.8
⫾ 1.7 ⫾ 1.4 ⫾ 1.3* ⫾ 1.7* ⫾ 1.6* ⫾ 0.8* ⫾ 2.6 ⫾ 3.2 ⫾ 2.2* ⫾ 1.3* ⫾ 3.0* ⫾ 3.1*
BE, mmol/l ⫺0.5 ⫺2.4 ⫺6.2 ⫺10.2 ⫺12.5 ⫺14.8 ⫺0.7 ⫺2.0 ⫺6.4 ⫺9.5 ⫺10.6 ⫺12.5
⫾ 1.4 ⫾ 1.9* ⫾ 1.8* ⫾ 2.4* ⫾ 2.5* ⫾ 1.8* ⫾ 2.1 ⫾ 2.7* ⫾ 2.1* ⫾ 1.0* ⫾ 3.7* ⫾ 4.2*
Values are means ⫾ SD. PaO2, arterial PO2; PaCO2, arterial PCO2; A-aDO2, alveolar-arterial difference for O2; BE, base excess. * Significantly
different from rest (P ⬍ 0.05).
930 INHALED NO DURING EXERCISE

Table 6. Blood variables during hypoxic and hypoxic ⫹ NO 5-min cycle ergometry
Hypoxia Hypoxia ⫹ NO

Minute Minute

Rest 1 2 3 4 5 Rest 1 2 3 4 5

PaO2, Torr 66.7 47.4* 44.3 44.2 43.3 44.3 64.3 44.9 43.3 42.6 42.0 42.3
⫾ 5.4 ⫾ 4.9 ⫾ 5.3* ⫾ 5.2* ⫾ 4.2* ⫾ 3.4* ⫾ 6.6 ⫾ 3.3* ⫾ 3.4* ⫾ 3.6* ⫾ 2.4* ⫾ 2.2*
SaO2, % 93.7 84.0 79.0 78.2 74.9 74.7 93.5 82.1 78.1 75.6 72.2 71.2
⫾ 1.4 ⫾ 4.3* ⫾ 4.9* ⫾ 4.5* ⫾ 5.6* ⫾ 4.3* ⫾ 4.0 ⫾ 1.9* ⫾ 1.6* ⫾ 1.8* ⫾ 1.9* ⫾ 1.9*
PaCO2, Torr 37.6 34.9 33.7 31.7 30.1 29.3 35.9 33.4 32.9 31.6 29.7 29.0
⫾ 4.0 ⫾ 3.0* ⫾ 2.3* ⫾ 1.2* ⫾ 1.2* ⫾ 2.4* ⫾ 6.0 ⫾ 4.9 ⫾ 5.0 ⫾ 4.3* ⫾ 4.1* ⫾ 3.8*
A-aDO2, Torr 2.0 15.8 21.0 23.9 26.3 25.6 1.4 20.5 24.3 25.3 26.9 26.7
⫾ 2.2 ⫾ 8.6* ⫾ 11.5* ⫾ 7.2* ⫾ 6.9* ⫾ 6.8* ⫾ 1.9 ⫾ 7.0* ⫾ 6.2* ⫾ 5.7* ⫾ 6.6* ⫾ 6.8*
pH 7.42 7.41 7.36 7.33 7.29 7.26 7.41 7.41 7.37 7.32 7.28 7.25
⫾ 0.02 ⫾ 0.02 ⫾ 0.04* ⫾ 0.04* ⫾ 0.05* ⫾ 0.07* ⫾ 0.02 ⫾ 0.03 ⫾ 0.03* ⫾ 0.04* ⫾ 0.04* ⫾ 0.04*
HCO3⫺, mmol/l 24.1 22.0 19.2 16.8 15.0 13.7 22.6 21.3 18.4 15.9 13.9 12.4
⫾ 2.1 ⫾ 2.0 ⫾ 1.9* ⫾ 1.9* ⫾ 1.9* ⫾ 2.3* ⫾ 4.0 ⫾ 3.3 ⫾ 2.4* ⫾ 1.8* ⫾ 1.7* ⫾ 1.4*
BE, mmol/l 0.3 ⫺1.9 ⫺5.4 ⫺8.0 ⫺10.4 ⫺12.2 ⫺0.7 ⫺1.4 ⫺5.6 ⫺8.6 ⫺11.2 ⫺13.3
⫾ 1.6 ⫾ 1.3* ⫾ 2.0* ⫾ 2.1* ⫾ 2.7* ⫾ 3.5* ⫾ 3.3 ⫾ 2.2 ⫾ 2.0* ⫾ 1.6* ⫾ 1.7* ⫾ 1.6*
Values are means ⫾ SD. * Significantly different from rest (P ⬍ 0.05).

hypoxic exercise in athletes with EIH. Unique to this lated. As a result, pulmonary gas exchange is im-
investigation was the observation that inhalation of 20 proved. We hypothesized that inhaled NO would im-
ppm NO during normoxic and hypoxic high-intensity, proved gas exchange in athletes with EIH.
short-duration exercise did not significantly affect gas Inhaled NO during normoxia. The present investi-
exchange, V̇O2, or cycling power in highly trained ath- gation demonstrated no differences in gas exchange,
letes with EIH. V̇O2, power, or heart rate between N and N/NO at rest
The mechanism(s) of EIH remains controversial. or during exercise. These findings do not support our
However, V̇A/Q̇ inequality and diffusion limitations original hypothesis but do agree with observations in
may be causative (6, 11, 30). The mechanisms explain- resting sheep (17) and the fact that PaO2 was not
ing the increases in V̇A/Q̇ inequality and diffusion lim- altered in normal individuals who inhaled NO at rest
itation are unknown but may be related to the devel- (8). In addition, our findings are consistent with data
opment of transient pulmonary edema (6). Some that reported no effects of inhaled NO on Q̇ or DLCO in
authors have suggested that, in elite athletes, the humans (2) or PaO2 and V̇A/Q̇ in dogs (12) under nor-
integrity of the pulmonary blood-gas barrier is altered moxic conditions.
at extreme levels of exercise (29). However, identifica- Studies that have sought to examine the effects of
tion of pulmonary edema and pulmonary stress failure inhaled NO during exercise in humans have been few.
has been difficult to achieve in exercising humans and To date, only one group (7) has investigated the effects
remains controversial. Although the specific mecha- of NO inhalation on pulmonary gas exchange during
nism for V̇A/Q̇ inequality and diffusion limitations is exercise in highly trained athletes. In agreement with
debatable, the end result is a widened A-aDO2. Inhaled the present study, they observed no differences for
NO, a selective pulmonary vasodilator (8, 9), is used in ventilatory and performance parameters. However,
the treatment of diseases characterized by pulmonary they found that inhalation of NO caused PaO2 to de-
hypertension and hypoxemia (1, 15, 16). The rationale crease at rest and during exercise compared with
is based on the fact that NO, given by inhalation, only breathing room air. This is in contrast to the present
dilates those pulmonary vessels that are well venti- study and the resting data of Frostell et al. (8). It is

Fig. 2. Arterial PO2 (PaO2) during rest and at each minute of 5-min Fig. 3. Percent arterial oxyhemoglobin saturation (SaO2) during rest
cycle ergometry tests under condition N (■), N/NO (䊐), H (F), and and at each minute of 5-min cycle ergometry tests under condition N
H/NO (E). Values are means ⫾ SD. *N and N/NO are significantly (■), N/NO (䊐), H (F), and H/NO (E). Values are means ⫾ SD. *N and
different from resting condition (rest) (P ⬍ 0.05). †H and H/NO N/NO are significantly different from rest (P ⬍ 0.05). †H and H/NO
significantly different from rest (P ⬍ 0.05). significantly different from rest (P ⬍ 0.05).
INHALED NO DURING EXERCISE 931

important to note that arterial blood-gas measure- explain the lack of effect of NO on gas exchange.
ments were corrected for temperature in the present Furthermore, subjects may have been slightly hyper-
study, whereas those of Durand et al. (7) were not. This ventilating at the time of the resting sample (PaCO2 of
would likely overestimate the drop in PaO2 and SaO2 36–38 Torr), which could have affected the influence
they observed during exercise and impact on the inter- of NO.
pretation of their results. The lack of effect of NO during hypoxic exercise is
Although pulmonary pressures were not measured consistent with the observations of Koizumi et al. (17).
in this study, we can speculate on two possible expla- Inhalation of NO during exercise in sheep had no effect
nations for our normoxic results: 1) pulmonary pres- on PaO2. In addition, inhaled NO did not change the
sures were not altered or 2) pulmonary pressures were time course or magnitude of changes in pulmonary
reduced with no effect on gas exchange. We believe it is pressures in the exercising sheep.
more probable that the first scenario occurred, al- Limitations of the study. A debatable point is how
though we cannot exclude the second possibility. The much of the inhaled NO actually reached the lower
pulmonary capillary bed in these athletes may have airways? Although it was not possible to measure the
been already maximally dilated, and inhalation of a amount of inhaled NO that reaches the alveoli, we
vasodilator would have no further effect. Therefore, no assume that subjects in the present study did indeed
effect on pulmonary pressure would be expected nor inhale 20 ppm NO. We are confident that NO was
would any alteration in gas exchange occur. Our re- delivered to the respiratory system, as it was measured
sults are therefore consistent with the hypothesis of 5 cm from the point of inspiration. The experimental
Dempsey (4), who postulated that pulmonary capillary approach and NO concentration employed in our study
blood volume reaches its maximal morphological limit were similar to those employed by others (7). It is
and further dilation is not anatomically possible. How- possible that the concentration of NO used in this
ever, in light of our lack of hemodynamic measures, we study was not sufficient to induce vasodilation; how-
emphasize the speculative nature of this explanation of ever, this seems unlikely given that similar concentra-
our negative findings. tions have been used previously to show significant
Inhaled NO during hypoxia. Unique to this investi- alterations in gas exchange and pulmonary pressures
gation was the delivery of NO to individuals with EIH (19, 21, 25). Our choice of 20 ppm was based on the
during hypoxic exercise. As expected, during H and above-mentioned studies and other clinical investiga-
H/NO, PaO2 and SaO2 were significantly lower than N tions.
and N/NO at all time points (Figs. 2 and 3). The drop in It is possible that we failed to detect a statistical
oxygenation during hypoxic exercise was consistent difference between conditions due to our small sample
with that observed in previous EIH studies (22). Sim- size and insufficient statistical power. However, this
ilar to the normoxic trials, inhaled NO did not alter gas seems unlikely. Post hoc statistical power (1 ⫺ ␤)
exchange, V̇O2, or cycling power during hypoxia. Pison calculations were performed utilizing a computer sta-
et al. (21) showed that addition of 20 ppm NO to a tistical program for gas-exchange variables across nor-
hypoxic (FIO2 ⫽ 12%) gas mixture returned pulmonary moxic and hypoxic conditions. Power values ranged
gas-exchange measures to baseline values in mechan- from 0.72 to 0.84, indicating sufficient protection
ically ventilated sheep. These are in contrast to our against type II errors.
results in which no statistical difference was observed Determination of pulmonary artery pressure and
for gas-exchange variables between H and H/NO. This cardiac output in the present study, together with V̇A/Q̇
was an unexpected result; we expected that HPV would measures, would have allowed for more definitive con-
be induced by using a hypoxic inspiratory gas mixture clusions on the effect of inhaled NO during exercise.
at rest and during exercise. Because the use of NO was hypothesized to act as a
We hypothesized that inhaled NO would have re- selective pulmonary vasodilator, the lack of the mea-
versed the transient HPV and improved pulmonary surement of pulmonary artery and wedge pressures
gas exchange, as demonstrated in healthy humans represents a major limitation of this study.
breathing hypoxic gas (21). Why did we observe no In summary, we have demonstrated that inhalation
effect of NO on gas exchange during the resting hy- of 20 ppm NO during normoxic and hypoxic high-
poxic condition? Potentially, the HPV at rest in the intensity, short-duration cycle exercise did not signifi-
present study was of a small enough magnitude and cantly affect gas exchange in athletes with EIH. Car-
duration to have had a minimal effect on gas exchange. diorespiratory variables and cycling power output were
However, this seems unlikely given that the time also unaffected by NO inhalation during normoxia and
course for HPV has a distinct initial rapid constriction hypoxia. We conclude that inhalation of 20 ppm NO
that reaches a peak within minutes and shows pro- during normoxic and hypoxic exercise has no effect on
found increases in pulmonary vascular resistance gas exchange in highly trained cyclists.
within the time frame used in the present study (28).
An alternate explanation is that the timing of resting REFERENCES
samples impacted on arterial blood-gas values. Resting
1. Barberà JA, Roger N, Roca J, Rovira I, Higenbottam TW,
samples were taken after a warm-up period that and Rodriguez-Rosin R. Worsening of pulmonary gas ex-
was preceded by a 5-min resting period. Possibly, the change with nitric oxide inhalation in chronic obstructive pul-
warm-up influenced the V̇A/Q̇ relationship and may monary disease. Thorax 52: S16–S21, 1997.
932 INHALED NO DURING EXERCISE

2. Brett SJ, Chambers J, Bush A, Rosenthal M, and Evans nitric oxide (NO) synthase inhibition, L-arginine infusion and
TW. Pulmonary response of normal human subjects to inhaled NO inhalation. J Clin Invest 94: 2275–2282, 1994.
vasodilator substances. Clin Sci (Colch) 95: 621–627, 1998. 18. Lawler J, Powers SK, and Thompson D. Linear relationship
3. Caillaud C, Serre-Cousine O, Anselme F, Capdevilla X, between V̇O2 max and V̇O2 max decrement during exposure to
and Prefaut C. Computerized tomography and pulmonary dif- acute hypoxia. J Appl Physiol 64: 1486–1492, 1988.
fusing capacity in highly trained athletes after performing a 19. Moinard J, Manier G, Pillet O, and Castaing Y. Effect of
triathlon. J Appl Physiol 79: 1226–1232, 1995. inhaled nitric oxide on hemodynamics and V̇A/Q̇ inequalities in
4. Dempsey JA. Is the lung built for exercise? Med Sci Sports patients with chronic obstructive pulmonary disease. Am J Re-
Exerc 18: 143–155, 1986. spir Crit Care Med 149: 1482–1487, 1994.
5. Dempsey JA, Hanson PG, and Henderson KS. Exercise- 20. Otis AB. The work of breathing. In: Handbook of Physiology.
induced arterial hypoxemia in healthy human subjects at sea Respiration. Washington, DC: Am. Physiol. Soc., 1964, sect. 3,
level. J Physiol (Lond) 355: 161–175, 1984.
vol. 1, chapt. 17, p. 463–476.
6. Dempsey JA and Wagner PD. Exercise-induced arterial hy-
21. Pison U, Lopez FA, Heidelmeyer CF, Rossaint R, and
poxemia. J Appl Physiol 87: 1997–2006, 1999.
7. Durand F, Mucci P, Safont L, and Prefaut C. Effects of nitric Falke K. Inhaled nitric oxide reverses hypoxic pulmonary vaso-
oxide inhalation on pulmonary gas exchange during exercise in constriction without impairing gas exchange. J Appl Physiol 74:
highly trained athletes. Acta Physiol Scand 165: 169–176, 1999. 1287–1292, 1993.
8. Frostell C, Blomqvist H, Hedenstierna G, Lundberg J, and 22. Powers SK, Martin D, Cicale M, Collop N, Huang D, and
Zapol WM. Inhaled nitric oxide selectively reverses human Criswell D. Exercise-induced hypoxemia in athletes: role of
hypoxic pulmonary vasoconstriction without causing systemic inadequate hyperventilation. Eur J Appl Physiol 65: 37–42,
vasodilation. Anesthesiology 78: 427–435, 1993. 1992.
9. Frostell C, Fratacci MD, Wain JC, Jones R, and Zapol W. 23. Prefaut C, Anselme-Poujol F, and Caillaud C. Inhibition of
Inhaled nitric oxide; a selective pulmonary vasodilator reversing histamine release by nedocromil sodium reduces exercise-in-
hypoxic vasoconstriction. Circulation 83: 2038–2047, 1991. duced hypoxemia in master athletes. Med Sci Sports Exerc 29:
10. Gale GE, Torre-Bueno JR, Moon RE, Saltzman HA, and 10–16, 1997.
Wagner PD. Ventilation-perfusion inequality in normal hu- 24. Putensen C, Rasanen J, and Lopez FA. Improvement in
mans during exercise at sea level and simulated altitude. J Appl V̇A/Q̇ distributions during inhalation of nitric oxide in pigs with
Physiol 58: 978–988, 1985. methacholine-induced bronchoconstriction. Am J Respir Crit
11. Hopkins SR, Gavin TP, Siafakas NM, Haseler LJ, Olfert Care Med 151: 116–122, 1995.
IM, Wagner H, and Wagner PD. Effect of prolonged, heavy 25. Rossaint R, Falke KJ, Lopez F, Slama K, Pison U, and
exercise on pulmonary gas exchange in athletes. J Appl Physiol Zapol WM. Inhaled nitric oxide for the adult respiratory dis-
85: 1523–1532, 1998. tress syndrome. N Engl J Med 328: 399–405, 1993.
12. Hopkins SR, Johnson EC, Richardson RS, Wagner H, De 26. Schaffartzik W, Poole DC, Derion T, Tsukimoto K, Hogan
Rosa M, and Wagner PD. Effects of inhaled nitric oxide on gas MC, Arcos JP, Bebout DE, and Wagner PD. V̇A/Q̇ distribu-
exchange in lungs with shunt or poorly ventilated areas. Am J tion during exercise and recovery in humans: implications for
Respir Crit Care Med 156: 484–491, 1997.
pulmonary edema. J Appl Physiol 72: 1657–1667, 1992.
13. Hopkins SR and McKenzie DC. Hypoxic ventilatory response
27. Scherrer U, Vollenweider L, Delabays A, Savic M, Eichen-
and arterial desaturation during heavy work. J Appl Physiol 67:
1119–1124, 1989. berger U, Gian-Reto K, Fikrle A, Ballmer P, Nicod P, and
14. Hopkins SR, McKenzie DC, Schoene RB, Glenny RW, and Bartsch P. Inhaled nitric oxide for high altitude pulmonary
Robertson HT. Pulmonary gas exchange during exercise in edema. N Engl J Med 334: 624–629, 1996.
athletes. I. Ventilation-perfusion mismatch and diffusion limita- 28. Ward JPT and Aaronson PI. Mechanisms of hypoxic vasocon-
tion. J Appl Physiol 77: 912–917, 1994. striction. Respir Physiol 115: 261–271, 1999.
15. Hurford WE. The biological basis for inhaled nitric oxide. Resp 29. West JB and Mathieu-Costello O. Structure, strength, fail-
Care Clin N Am 3: 357–369, 1997. ure, and remodeling of the pulmonary blood-gas barrier. Annu
16. Kacmarek RM. Update on the use of inhaled nitric oxide. Can Rev Physiol 61: 543–572, 1999.
Resp J 3: 373–376, 1996. 30. West JB, Tsukimoto K, Mathieu-Costello O, and Prediletto
17. Koizumi T, Gupta R, Banerjee M, and Newman JH. R. Stress failure in pulmonary capillaries. J Appl Physiol 70:
Changes in pulmonary vascular tone during exercise: effects of 1731–1742, 1991.