The influence of PaO2, pH and SaO2 on maximal oxygen uptake
Transcript of The influence of PaO2, pH and SaO2 on maximal oxygen uptake
The in¯uence of PaO2, pH and SaO2 on maximal oxygen
uptake
H . B . N I E L S E N , P . M A D S E N , L . B . S V E N D S E N , R . C . R O A C H and N . H . S E C H E R
The Copenhagen Muscle Research Centre, Department of Anaesthesia, Rigshospitalet, University of Copenhagen, Blegdamsvej
Copenhagen é, Denmark
ABSTRACT
Influence of arterial oxygen pressure (PaO2) and pH on haemoglobin saturation (SaO2) and in turn on O2
uptake (VO2) was evaluated during ergometer rowing (156, 276 and 376 W; VO2max, 5.0 L min)1;
n � 11). During low intensity exercise, neither pH nor SaO2 were affected signi®cantly. In response to
the higher work intensities, ventilations (VE) of 129 � 10 and 155 � 8 L min)1 enhanced the end tidal
PO2 (PETO2) to the same extent (117 � 2 mmHg), but PaO2 became reduced (from 102 � 2 to 78 � 2
and 81 � 3 mmHg, respectively). As pH decreased during maximal exercise (7.14 � 0.02 vs.
7.30 � 0.02), SaO2 also became lower (92.9 � 0.7 vs. 95.1 � 0.1%) and arterial O2 content (CaO2) was
202 � 3 mL L)1. An inspired O2 fraction (FIO2) of 0.30 (n � 8) did not affect VE, but increased PETO2
and PaO2 to 175 � 4 and 164 � 5 mmHg and the PETO2±PaO2 difference was reduced (21 � 4 vs.
36 � 4 mmHg). pH did not change when compared with normoxia and SaO2 remained within 1% of
the level at rest in hyperoxia (99 � 0.1%). Thus, CaO2 and VO2max increased to 212 � 3 mL L)1 and
5.7 � 0.2 L min)1, respectively. The reduced PaO2 became of importance for SaO2 when a low pH
inhibited the af®nity of O2 to haemoglobin. An increased FIO2 reduced the gradient over the alveolar-
arterial membrane, maintained haemoglobin saturation despite the reduction in pH and resulted in
increases of the arterial oxygen content and uptake.
Keywords arterial oxygen pressure, arterial oxygen saturation, hyperoxia, lactate, pH, rowing.
Received 24 October 1997, accepted 24 March 1998
A pulmonary limitation to maximal O2 uptake (VO2max)
is indicated by a reduced arterial O2 pressure (PaO2
55 mmHg, Dempsey et al. 1984) and saturation (SaO2
85%; Rowell et al. 1964). The SaO2 is reduced especially
in the athlete (to 87 vs. 93% in control subjects; Williams
et al. 1986) indicating that a high cardiac output and in
turn a low pulmonary transit time is important. Those
subjects who demonstrate the lowest PaO2 tend to
hyperventilate less as indicated by the arterial carbon
dioxide pressure (PaCO2 > 35 mmHg) and a low alve-
olar O2 pressure (PAO2 < 110 mmHg, Dempsey et al.
1984). Equally, the acceptance of O2 by the erythrocytes
is of importance and during exercise, haemoconcentra-
tion facilitates the pulmonary O2 diffusion, while the
associated acidosis counteracts the binding of O2 to
haemoglobin (Rasmussen et al. 1991, Hanel et al. 1994).
Also, an uneven distribution of pulmonary ventilation
and perfusion may contribute (Hammond et al. 1986,
Hopkins et al. 1994), but the fact that SaO2 is reduced
similarly during supine and upright exercise is taken to
indicate that this mismatch is not critical (Pedersen et al.
1996). The alveolar-capillary membrane is subjected to
stress failure in the racing horse (West & Mathieu-
Costello 1995) and in humans (Hopkins et al. 1997). A
vulnerable membrane is indicated by the post-exercise
reduction in pulmonary diffusion capacity (Hanel et al.
1994) as only »50% of this reduction is related to a drop
in the pulmonary capillary blood volume (Hanel et al.
1997). Thus, the exercise PaO2 increases following a
polyunsaturated fatty acid supplemented diet (74±
81 mmHg, Aguilaniu et al. 1995) and the drop in PaO2
correlates to plasma histamine during maximal exercise
(Anselme et al. 1994).
An elevation of the inspired fraction of O2 (FIO2) to
only 0.26 overcomes the reduction in oximetry deter-
mined SaO2 and VO2max becomes elevated (Powers
et al. 1989) supporting that PAO2 is critical for O2
transport. With a larger elevation of FIO2 to 0.50, the
arterial O2 content (CaO2) and VO2max are elevated by 8
and 13%, respectively (Ekblom et al. 1975). However,
Correspondence: Henning Bay Nielsen M.D. Department of Anaesthesia 2041, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen é,
Denmark.
Acta Physiol Scand 1998, 164, 89±97
Ó 1998 Scandinavian Physiological Society 89
with a markedly elevated FIO2, the calculation of VO2
tends to be an overestimate (Welch & Pedersen 1981).
Equally, it may be that the oximetry determined SaO2 is
in¯uenced by vasoconstriction of the evaluated capil-
lary bed. Powers et al. (1984) found it equally reduced
during arm cranking and cycle ergometry, while the
invasively determined SaO2 immediately after exercise
correlates to pH and in turn to the muscle mass in-
volved in exercise (Rasmussen et al. 1991).
Thus, although several studies have investigated the
exercise-induced hypoxaemia, no study has integrated
arterial blood gas and acid±base analysis, ventilatory
variables and effects of hyperoxia during and after ex-
ercise. The aim of this study was to evaluate the in-
¯uence of PaO2 on SaO2 when it is challenged by a low
pH. Arterial blood samples were obtained both during
exercise and in the recovery with simultaneous deter-
mination of heart rate (HR) and ventilatory variables. A
pilot study indicated that an FIO2 of 0.30 secures SaO2
during maximal exercise and we hypothesised that
VO2max would increase in proportion to CaO2.
METHODS
Eleven competitive male oarsmen [age 26 (20±
29) years, body weight 80 (70±94) kg, and height 1.87
(1.79±2.00) m (median with range)] participated in the
study as approved by the Ethics Committee of Co-
penhagen (KF 01±349/96). No subject had any disease
or injury 3 weeks prior to the experiment, nor were
they taking any medication. They were not allowed to
drink or eat after midnight on the day of the experi-
ment, which began at 08 00 hours.
Rowing was performed on a Concept II ergometer
(Morrisville, VT, USA) connected to a computer
(Concept II) that displays the split-time for every
`500 m' and the average work capacity. The subjects
attended the laboratory on 3 days to establish the ex-
ercise intensity that provoked a reduction in SaO2. On
the ®rst day, rowing for 20 min aimed at a work in-
tensity ~50% VO2max (low intensity). After 5 min of
recovery, it was followed by `high intensity rowing' as
the subject was instructed to exert as high an intensity
as possible for 30 min with constant speed and stroke
rate. On the second day, they warmed up at an indi-
vidually determined pace for 10 min; recovered for
2 min and then rowed `all-out' for 6 min simulating a
2000-m competitive effort (maximal intensity). All
subjects were familiar with this type of exercise.
An FIO2 of 0.26 during exercise does not maintain
the oximetry determined SaO2 at the level of rest
(Powers et al. 1989). The pilot study showed that during
an all-out row, a FIO2 of 0.30 maintained PaO2 and
SaO2 at 153 mmHg and 98%, respectively, at an arterial
pH of 7.28. Also, that FIO2 was expected not to affect
the fraction of N2 to an extent that the calculation of
VO2 would become affected signi®cantly. Therefore,
eight oarsmen repeated the 6-min all-out row within
3 weeks while breathing 30% O2 in N2.
Respiratory values
Respiratory values were recorded breath by breath
(MedGraphics 2001, St. Paul, MN, USA). During
hyperoxia an extension of the pneumotach was pro-
vided with a T-valve (model 2700, Hans Rudolph,
Kansas City, MO, USA) to humidi®ed air in Douglas
bags. After 5 min to allow stabilization, ventilation
(VE), VO2, respiratory rate (RR), expired carbon diox-
ide (VCO2), end tidal partial pressures for O2 (PETO2)
and carbon dioxide (PETCO2) were averaged for 30 s;
or during maximal exercise for every 15 s. As the `dead
space' increases with tidal volume during exercise
(Asmussen & Nielsen 1956), the PETO2 and PETCO2
were taken to re¯ect PAO2 and PACO2. During all-out
rowing, the peak VO2 corresponds to VO2max as it is
similar to or slightly higher than that obtained during
cycling (Cunningham et al. 1975). The O2 debt (O2debt)
was the sum of VO2 in the ®rst 4 min of the recovery
and the O2 de®cit (O2def) was calculated assuming
linearity between metabolism and work rate (Medbù
et al. 1988). For VO2 determination on incremental
workloads, the subjects performed rowing at 100, 150,
200 and 250 W on a separate day. We assumed the
same linearity for VO2 and power during exercise in
ambient and hyperoxic air. HR was monitored every
15th second using a `Sport Tester' PE 3000 (Polar
Electro, Kempele, Finland).
Blood samples
Blood samples for arterial bicarbonate (BC), carbon
dioxide pressure (PaCO2), PaO2, pH, base excess (BE),
haemoglobin (Hgb), potassium and lactate were ob-
tained anaerobically (QS50, Radiometer, Copenhagen,
Denmark) from a catheter in the radial artery of the
non-dominant arm. Collections were made at rest, ev-
ery ®fth minute during the 20-min row, immediately
prior to the 30-min row followed by every second
minute. During all-out rowing, blood samples were
drawn every minute, 30 s before and after the termi-
nation of exercise, and one minute into the recovery.
Thereafter, they were collected every second minute.
All samples were placed immediately on ice until
measured at 37 °C (ABL4 apparatus; Radiometer) while
SaO2 and Hgb were determined on OSM3 apparatus
(Radiometer). Lactate in whole blood was determined
by a YSI 2300 (Yellow Springs Instruments Co., Inc.,
OH, USA). The CaO2 was the sum of bound (1.39 Hgb
SaO2) and dissolved O2 (0.003 PaO2).
O2 content and uptake � H B Nielsen et al. Acta Physiol Scand 1998, 164, 89±97
90 Ó 1998 Scandinavian Physiological Society
Statistical analysis
Variables are presented as mean with SEM. Changes
were evaluated by the Friedmann test across time and
exercise intensities and differences were located by the
Wilcoxon test by rank. A P-value <0.05 was considered
statistically signi®cant.
RESULTS
Ambient air
During low intensity rowing, the power was
156 � 15 W. In the ®rst few minutes of exercise, VE
and RR increased to a steady state of ~50 L min)1 and
~30 breath min)1 (Fig. 1, Table 1). PETCO2 became
elevated, whereas PaCO2, PETO2, PaO2 and the PETO2±
PaO2 difference did not change signi®cantly. However,
at the onset of exercise PaCO2 increased to
43 � 1 mmHg and PETO2 and PaO2 were reduced to
94 � 3 and 86 � 2 mmHg. Equally, pH and SaO2 were
below the resting level (7.39 � 0.02 and 96.8 � 0.2%,
respectively). Yet, all variables recovered during exer-
cise. BE and BC did not change signi®cantly, whereas
plasma potassium was elevated. CaO2 was enhanced
and HR and VO2 established a steady state.
During high intensity rowing, power was
276 � 12 W. The VE and RR increased without es-
tablishing a steady state (Fig. 1, Table 1). The PETCO2
and PaCO2 decreased, and PETO2 became elevated. Yet,
PaO2 was reduced and the PETO2±PaO2 difference
therefore increased. Lactate increased to
6.5 � 0.9 mmol L)1 and both pH and SaO2 were re-
duced to make CaO2 not different from that established
during low intensity rowing. HR increased throughout
exercise as did VO2.
After exercise, both VE and RR decreased imme-
diately but hyperventilation was indicated by a low
PETCO2 and PaCO2 (34 � 1 and 33 � 1 mmHg, res-
pectively) (Fig. 1). Thus, PETO2 (114 � 2 mmHg) and
PaO2 (105 � 2 mmHg) were elevated and although pH
was lower than during exercise (7.28 � 0.02), SaO2 in-
creased to 97.3 � 0.2%.
During maximal exercise, power was 376 � 22 W
re¯ected in both a large VE and RR (Fig. 2, Table 1).
PETCO2 and PaCO2 were similar to `high intensity' ex-
ercise, whereas VCO2 was higher. Yet, there was an
immediate drop in PETCO2 at the onset and although it
recovered, it decreased to the same level towards the
end of exercise. The PETO2 and PaO2 changed to the
same extent as during high intensity exercise, but the
PETO2±PaO2 difference became smaller. PETO2 in-
creased at the onset of exercise; then decreased and
thereafter, it became elevated to 119 � 1 mmHg.
Lactate reached 11.0 � 0.7 mmol L)1 with a lower
range of arterial pH at 7.06. Arterial desaturation was
pronounced and ®ve subjects demonstrated an SaO2
below 90% and in one subject it was 85%. Yet, with the
increase in Hgb, CaO2 was similar to that during a
lower work intensity. With a similar increase in HR as
during high intensity rowing, VO2max was
5.0 � 0.3 L min)1.
Immediately following exercise, VE and RR re-
mained at the exercise level, but after 30 s they de-
creased, while PETCO2 and PaCO2 remained low
(Fig. 2). Accordingly, PETO2 was high and PaO2 be-
came elevated to 43 � 5 mmHg above the resting level.
SaO2 increased although pH was lower than during
exercise.
Figure 1 Pulmonary ventilation (A), respiratory rate (B), end tidal
carbon dioxide pressure (C), and end tidal oxygen pressure (D) at rest
and in response to low and high intensity ergometer rowing. Values
are mean with SEM.
Ó 1998 Scandinavian Physiological Society 91
Acta Physiol Scand 1998, 164, 89±97 H B Nielsen et al. � O2 content and uptake
FIO2 of 0.30
Power (388 � 21 W) was not signi®cantly different
from that obtained in ambient air and VE and RR were
also at the level of the control exercise (Fig. 2, Table 1).
Yet, PETCO2, PaCO2, and VCO2 were higher. PETO2 and
PaO2 were elevated and not signi®cantly changed by
exercise, making the PETO2±PaO2 difference smaller.
The PETCO2±PaCO2 difference (4.4 � 0.7 vs.
2.9 � 0.4 mmHg, normoxia vs. hyperoxia) during ex-
ercise was marginally reduced by O2 supplementation
(P � 0.1). The SaO2 at rest was elevated and only a
minimal reduction occurred during exercise. Neither
Hgb nor CaO2 were affected at rest and Hgb increased
to a similar level as during exercise in ambient air. Thus,
the CaO2 became elevated by 8.5 � 0.1% (Fig. 3). Also,
HR was similar, but VO2max was enhanced by 11 � 3%
(5.7 � 0.2 L min)1). The concentration of lactate
reached 9.8 � 0.6 mmol L)1 and it was not signi®-
cantly changed by hyperoxia. Also, pH, BE, and BC
were not affected. Equally, the O2debt remained un-
changed by hyperoxia, whereas the O2def was reduced
from 10.4 � 5.0 to 8.3 � 5.5 L.
DISCUSSION
This study demonstrates that (i) the arterial oxygen
pressure and saturation of haemoglobin are reduced
only during intense exercise and (ii) an elevated inspired
oxygen fraction reduces the gradient over the alveolar-
arterial membrane and maintains haemoglobin satura-
tion despite a reduction in pH and (iii) the pulmonary
oxygen uptake increases in proportion to the arterial
oxygen content.
During exercise, the alveolar PO2 expressed as the
PETO2 was insuf®cient to maintain the PaO2 and with
marked acidosis, SaO2 became low. Conversely, after
exercise a low arterial pH was not critical for SaO2 as
the PaO2 increased. Equally, an elevation of the frac-
tion of inspired oxygen increased the PETO2 and in
turn the PaO2 to an extent where the SaO2 became
affected little by pH with subsequent increases in the
CaO2 and VO2. In other words, the relationship be-
tween PaO2, SaO2, and pH established at rest (Bohr
et al. 1904) also applied to exercise as SaO2 was within
0.5 � 0.3% of the calculated value (Severinghaus
1958). This was the case although no corrections of
arterial gases were made for changes in body temper-
ature. During maximal exercise, the central temperature
increases by »0.5 °C (Saltin et al. 1972, Bergh & Ek-
blom 1979) with only a small effect on PaO2 and SaO2
(Severinghaus 1979). It should be considered that the
subjects performed a warm-up before the all-out bout,
which would make an effect of the temperature from
rest to exercise small.
We con®rmed that SaO2 tends to be low in subjects
with a large aerobic capacity (Williams et al. 1986)
(r � 0.48, P � 0.05), but even in the subject with a
VO2max of only 3.7 L min)1, it was reduced to 93%
associated with a pH of 7.1. In those subjects with a
Table 1 Variables at rest and during rowing in ambient air and with an FIO2 of 0.30
Room air, W F1O2 0.30, W
Rest 156 276 376 Rest 388
VE (L min)1) 11 � 2 51 � 7* 129 � 10* 155 � 8* 12 � 1 150 � 8*
RR (breath min)1) 16 � 1 31 � 3* 53 � 3* 59 � 2* 12 � 1 59 � 3*
PETCO2 (mmHg) 41 � 1 45 � 1* 38 � 1* 36 � 1* 37 � 1 41 � 1* PaCO2 (mmHg) 39 � 1 41 � 1 33 � 1* 33 � 1* 41 � 1 37 � 1* VCO2 (L min)1) 0.3 � 0.0 1.8 � 0.2* 3.8 � 0.3* 4.6 � 0.3* 0.3 � 0.0 5.0 � 0.3* PETO2 (mmHg) 108 � 2 102 � 1 117 � 2* 117 � 1* 175 � 4 183 � 2 PaO2 (mmHg) 97 � 1 93 � 2* 78 � 2* 81 � 3* 164 � 5 165 � 5 P(ET-a)O2 (mmHg) 12 � 2 12 � 2 41 � 3* 36 � 3* 15 � 3 21 � 4 K+ (mmol L)1) 4.0 � 0.1 4.5 � 0.1* 5.2 � 0.1* 6.0 � 0.1* 4.1 � 0.1 5.9 � 0.1*
SBC (mmol L)1) 25 � 1 25 � 1 19 � 1* 15 � 1* 24 � 1 16 � 0*
pH 7.42 � 0.0 7.40 � 0.0 7.34 � 0.0* 7.23 � 0.0* 7.39 � 0.0 7.24 � 0.0*
SBE (mmol L)1) 0.8 � 0.7 1.0 � 0.5 )7.1 � 1.2* )12.3 � 1.0* )0.1 � 0.0 )10.2 � 0.6*
SaO2 (%) 97.9 � 0.1 97.3 � 0.1 95.1 � 0.5* 92.9 � 0.7* 99.1 � 0.1 98.3 � 0.2* Hgb (mmol L)1) 8.7 � 0.2 9.1 � 0.1* 9.3 � 0.1* 9.6 � 0.2* 8.5 � 0.2 9.4 � 0.1*
CaO2 (mL L)1) 193 � 4 202 � 3* 200 � 3* 202 � 3* 193 � 4 212 � 3* VO2 (L min)1) 0.3 � 0.0 2.2 � 0.3* 4.3 � 0.3* 4.5 � 0.3* 0.5 � 0.1 5.1 � 0.2* HR (beats min)1) 69 � 7 127 � 3* 178 � 4* 180 � 2* 59 � 2 176 � 2*
Values are the average and SEM of variables during exercise and in the last minute of rest. HR, heart rate; FIO2, inspired fraction of O2; K+,
potassium; PaO2 and PaCO2, partial pressure of oxygen and carbon dioxide in arterial blood; SaO2, oxygen saturation of arterial haemoglobin; SBC,
concentration of bicarbonate, SBC, base excess; VE, Pulmonary ventilation; VCO2, expired carbon dioxide; VO2, pulmonary O2 uptake; W, watt. *,
different value compared to rest; P < 0.05.
92 Ó 1998 Scandinavian Physiological Society
O2 content and uptake � H B Nielsen et al. Acta Physiol Scand 1998, 164, 89±97
higher VO2 who did not demonstrate marked desatu-
ration, pH remained above 7.2 and the correlation was
improved when pH was taken into account (r � 0.76,
P < 0.01).
The effect(s) of a small increase in the FIO2 on the
arterial gases and/or SaO2 were shown by Dempsey
et al. (1984) and Powers et al. (1989). Dempsey et al.
investigated the effect of hyperoxia (FIO2, 0.24) only
during a submaximal exercise intensity although the
most marked reduction in PaO2 occurred at the maxi-
mum level of a graded treadmill test. Also, the SaO2
values were not presented during the exercise with
hyperoxia. Powers et al. (1989) used oximetry deter-
mined SaO2 with no evaluation of the role of pH. Only
Figure 2 Pulmonary ventilation (A), respiratory rate (B), end tidal carbon dioxide pressure (C), arterial carbon dioxide pressure (D), expired
carbon dioxide (E), end tidal oxygen pressure (G), PETO2±PaO2 difference (H), arterial bicarbonate (I), pH (J), oxygen saturation of arterial
haemoglobin (K), and oxygen uptake (L) at rest and in response to a 6-min ``all-out'' ergometer row in ambient air (s) and with an inspired
oxygen fraction of 0.30 (d). Values are mean with SEM.
Ó 1998 Scandinavian Physiological Society 93
Acta Physiol Scand 1998, 164, 89±97 H B Nielsen et al. � O2 content and uptake
the study by Peltonen et al. (1995) also investigates the
effect of hyperoxia (FIO2, 0.52) during an `all-out' er-
gometer row, but they did not evaluate arterial blood
gas variables.
Dempsey et al. (1984) considered that those subjects
who had a low PaO2 during maximal exercise also
tended to hyperventilate little, expressed as a relative
high PaCO2 and only a small increase in PAO2. Also,
decreased SaO2 during maximal exercise correlated with
PETO2 and the ventilatory equivalent for O2 (Miyachi &
Tabata 1992). In the present study, the subject that
desaturated the most (85%) showed both the lowest pH
(7.06) and PaO2 (67 mmHg) and although his PaCO2
was the highest (39 mmHg), a VE and PETO2 of
183 L min)1 and 123 mmHg, respectively, do not in-
dicate `hypoventilation'. Other subjects reached a
higher VE making PaCO2 lower and all subjects reached
a PETO2 above 120 mmHg. Thus, in contrast to rest
where an elevated PaCO2 increases VE, it appears that
during exercise, the correlation between VCO2 and VE
re¯ects CO2 elimination from bicarbonate. Also, in a
comparison of the different work rates, there was little
indication for PaCO2 control of VE. PETCO2 and to a
lesser extent PaCO2 were elevated during low intensity
exercise, whereas they became reduced during intense
exercise as VE increased implying that VE was more
linked to, e.g. pH than to PaCO2. In support, the ele-
vated FIO2 increased both PETCO2 and PaCO2 associ-
ated with similar deviations of VE and arterial pH.
An illustration of the role for a blood borne factor(s)
in control of breathing is that VE is reduced during
post-exercise muscle ischaemia with an increase fol-
lowing release of the cuff (Innes et al. 1989, Jùrgensen
et al. 1992). It is not established if this response relates
to, e.g. temperature, lactate, potassium, hormones or
even to venous return, and the results of this study
support an important role for pH. During light exercise,
RR and VE reached a steady state with no signi®cant
changes in pH. Conversely, at higher work rates, RR
and VE continued to increase (although PaCO2 re-
mained low) and pH continued to decrease. Also, im-
mediately following light exercise RR and VE
decreased, whereas they remained high for »30 s after
more intense exercise as arterial pH reached an even
lower level than during exercise.
A small central nervous in¯uence on VE is consid-
ered as VE and even more so, the ratio of VE to VO2, is
enhanced when exercise is made dif®cult during partial
neuromuscular blockade (Galbo et al. 1987). In this
study, such an in¯uence on VE may be re¯ected in the
temporary reduction in PETCO2 and increase in PETO2
at the onset of intense exercise. Conversely, at the onset
of low intensity exercise, the subjects tended to `hypo-
ventilate' as there was a temporary increase in PETCO2
and PaCO2 and a decrease in PETO2. CO2 retention
during light exercise is not speci®c to rowing but also
observed in response to light ergometer cycling, e.g. with
an increase in PaCO2 by 4 mmHg (Jùrgensen et al. 1992).
With hyperoxia, a high PaO2 is expected to reduce the
chemoreceptor sensitivity to CO2 (Asmussen & Nielsen
1946). Also, with increased CaO2 it could be that muscle
and splanchnic ischaemia is prevented, whereas prop-
rioceptive impulses remain unaffected by the O2 load.
During submaximal exercise VE is lower with hyperoxia
(Welch 1982) and we found VE to be slightly lower with
hyperoxia indicating only small hypoxic drive. Alterna-
tively, it may be that VE approaches a maximal level
during rowing where the respiratory muscles are used
not only for VE but also for stabilization of the upper
body. Thus, the ratio of VE to VO2 is reported to be
higher during cycling (Cunningham et al. 1975). During
rowing, oarsmen perform either one expiration during
the stroke and inspiration in the recovery or one com-
plete breath during each stroke and one complete breath
going up the slide (Steinacker et al. 1993). Also, stroke
frequency and RR increase in concert. It could be
speculated that ability to entrain breathing is of impor-
tance for suf®cient pulmonary gas exchange.
Of more relevance for the present study are the
deviations in the variables indicative of the role of the
alveolar-capillary membrane for O2 transport. The
PETO2±PaO2 difference increases during exercise
(Hesser & Matell 1965, Dempsey et al. 1984) and this
is taken to re¯ect ventilation±perfusion inequality, a
diffusion limitation and/or a pulmonary shunt (John-
son et al. 1996). During low intensity exercise, the
PETO2±PaO2 difference did not change, but it was
pronounced during intense exercise (Hopkins &
McKenzie 1989). Yet, it was smaller during maximal
exercise than during heavy exercise as PaO2 increased
Figure 3 Mean with SEM for arterial oxygen content (CaO2) at rest
and during a 6-min all-out ergometer row in ambient air (s) and with
an inspired oxygen fraction of 0.30 (d). Of note is the opposite
deviations in CaO2 and SaO2 [Fig. 2(K)] re¯ecting the increase in
haemoglobin during exercise.
94 Ó 1998 Scandinavian Physiological Society
O2 content and uptake � H B Nielsen et al. Acta Physiol Scand 1998, 164, 89±97
at the same PETO2 of 117 mmHg when VE was ele-
vated from 129 to 155 L min)1. One explanation is
that the elevated VE raised PAO2 if the enhanced VE
compensated for the increase in dead space (Asmussen
& Nielsen 1956).
With hyperoxia, the PETO2±PaO2 difference was
lowered to about half indicating that PAO2 was elevated
to an extent where pulmonary O2 diffusion does not
limit O2 transport also indicating that a VA/Q mis-
match is not critical for pulmonary O2 transport during
maximal exercise. The reduction in the PETO2±PaO2
difference during exercise with hyperoxia is in contrast
with the ®ndings by Dempsey et al. (1984) where O2
supplementation was administered during submaximal
running associated with a lower VE than reached during
rowing. This is of importance as the exercise PETO2
with a FIO2 of only 0.24 is similar to that during rowing
in ambient air. Thus, in the study by Dempsey et al.
(1984) a PETO2 of 127 mmHg made PaO2 increase to
94 mmHg and arterial haemoglobin could not have
been fully saturated.
Turcotte et al. (1997) were not able to demonstrate a
relationship between desaturation and pulmonary dif-
fusion capacity during exercise. Yet, the time available
for O2 equilibration in the pulmonary-alveolar capillary
is of importance for O2 transport as evidenced when
FIO2 is elevated. If gas equilibration with haemoglobin
takes the full time for the passage of the erythrocyte
through the capillary, then little O2 will be dissolved in
plasma. Conversely, the high PaO2 (and SaO2) estab-
lished with hyperoxia suggests that haemoglobin was
saturated before the erythrocyte reached the `end' of
the capillary.
On the other hand, O2 breathing reduces VA/Q
inequality (Hammond et al. 1986) and this could relate
to hypoxic vasoconstriction (Kronenberg et al. 1971).
Although this is unlikely during exercise at sea level
(Hammond et al. 1986), an SaO2 of 85% could induce
some vasoconstriction in the pulmonary circulation.
Thus, it is possible that with O2 supplementation such
vasoconstriction is abolished.
Alternatively, although O2 is electrochemically de-
termined, the reduced PETO2±PaO2 difference may
relate to an underestimation of PETO2 in hyperoxic
conditions as with the error in CO2 determination with
an infrared gas analyser (Hornby et al. 1995). Hornby
et al. (1995) showed this with O2 concentrations from
86.2 to 95.9% and with O2 concentrations of 40 and
50%, the percentage error of CO2 analysis would be
expected to be reduced to 1% of the respective readings
and it may be that the error is even lower with a FIO2
of 0.30.
When the alveolar PO2 was calculated using the
ideal alveolar air equation (Wasserman et al. 1987), the
PAO2±PaO2 difference became reduced (34 � 2 vs.
13 � 5 mmHg). However, the PAO2 equation assumes
that there is true equilibrium between ideal mean al-
veolar PCO2 and PaCO2 but this has not been validated
during maximal exercise. Also, we assumed that FIO2
was 0.208 and 0.30 in normoxic and hyperoxic condi-
tions for all experiments, thus, small errors in FIO2
with O2 supplementation may be serious for the cal-
culation of PAO2.
With the unchanged PaO2 and SaO2 during hype-
roxia, CaO2 became elevated and equally so VO2max
indicating that VO2 was not overestimated (Welch &
Pedersen 1981). pH, lactate, BC, BE, and O2debt did not
change suggestive of a similar anaerobic metabolism
and yet, power was the same. Also, randomized O2
supplementation increases VO2max by 11%, whereas
performance is enhanced by only 2% (Peltonen et al.
1995). Both observations support that O2 delivery to
the working muscles is almost unchanged with hype-
roxia as their blood ¯ow decreases (Welch et al., 1977).
Also, hyperoxia reduces blood ¯ow and increases re-
sistance after forearm ischaemia (Crawford et al. 1997).
This is consistent with an increase in mean arterial
pressure during maximal exercise with hyperoxia
(Ekblom et al. 1975). Thus, the elevated VO2 with
hyperoxia may re¯ect enhanced metabolic rate in `non-
exercising' tissue. The splanchnic region is markedly
hypoperfused during intense rowing (Nielsen et al.
1995) and hyperoxia could increase its O2 extraction.
An increase in liver function during exercise with an
FIO2 of 0.30 could account for an elevated PaCO2 and
in turn VCO2.
In summary, exercise-induced hypoxaemia and
desaturation were reversed by an FIO2 of 0.30 and
VO2max increased in proportion to the arterial O2 con-
tent with no effect on work capacity and anaerobic
metabolism suggesting that the increase in VO2 re¯ects
metabolism in non-working tissues. Thus, the alveolar
O2 pressure is critical for O2 transport supporting the
importance of a large ventilation during intense exercise.
We are grateful to Heidi Hansen and Anette Uhlmann for expert
technical assistance. We thank Idrñttens ForskningsraÊd (1995-1-19)
and the `Team Denmark' foundation for economical support. The
study was also supported by the Danish National Research Founda-
tion grant no. 504-4. Per Madsen and Henning Bay Nielsen were
funded by the Danish Medical Research Council (9400846). Lars Bo
Svendsen was employed in the Danish Armed Forces.
REFERENCES
Aguilaniu, B., Flore, P., Perrault, H., Page, J.E., Payan, E. &
Lacour, J.-R. 1995. Exercise-induced hypoxaemia in master
athletes: effects of a polyunsaturated fatty acid diet. Eur J
Appl Physiol 72, 44±50.
Anselme, F., Caillaud, C., Couret, I., Rossi, M. & PreÂfaut, C.
1994. Histamine and exercise-induced hypoxemia in highly
trained athletes. J Appl Physiol 76, 127±132.
Ó 1998 Scandinavian Physiological Society 95
Acta Physiol Scand 1998, 164, 89±97 H B Nielsen et al. � O2 content and uptake
Asmussen, E. & Nielsen, M. 1946. Studies on the regulation
of respiration in heavy work. Acta Physiol Scand 12, 171±188.
Asmussen, E. & Nielsen, M. 1956. Physiological dead space
and alveolar gas pressures at rest and during muscular
exercise. Acta Physiol Scand 38, 1±21.
Bergh, U. & Ekblom, B. 1979. Physical performance and peak
aerobic power at different body temperatures. J Appl Physiol
46, 885±889.
Bohr, C., Hasselbach, K. & Krogh, A. 1904. Ueber einen in
biologischer beziehung wichtigen ein¯uss, den die
kohlensaÈurespannung des blutes auf dessen
sauerstoffbindung uÈbt. Skand Arch Physiol 16, 402±412.
Crawford, P., Good, P.A., Gutierrez, E. et al. 1997. Effects of
supplemental oxygen on forearm vasodilation in humans.
J Appl Physiol 82, 1601±1606.
Cunningham, D.A., Goode, P.B. & Critz, J.B. 1975.
Cardiorespiratory response to exercise on a rowing and
bicycle ergometer. Med Sci Sports 7, 37±43.
Dempsey, J.A., Hamson, P.G. & Henderson, K.S. 1984.
Exercise- induced arterial hypoxaemia in healthy human
subjects at sea level. J Physiol Lond 355, 161±175.
Ekblom, B., Hout, R., Stein, E.M. & Thorstensson, A. T.
1975. Effect of changes in arterial oxygen content on
circulation and physical performance. J Appl Physiol 39,
71±75.
Galbo, H., Kjñr, M. & Secher, N.H. 1987. Cardiovascular,
ventilatory and catecholamine responses to maximal
dynamic exercise in partially curarized man. J Physiol 389,
557±568.
Hammond, M.D., Gale, G.E., Kapitan, K.S., Ries, A. &
Wagner, P. 1986. Pulmonary gas exchange in humans
during exercise at sea level. J Appl Physiol 60, 1590±
1598.
Hanel, B., Clifford, P.S. & Secher, N.H. 1994. Restricted
postexercise pulmonary diffusion capacity does not impair
maximal transport for O2. J Appl Physiol 77, 2408±2412.
Hanel, B., Teunissen, I., Rabùl, A., Warberg, J. & Secher,
N.H. 1997. Restricted postexercise pulmonary diffusion
capacity and central blood volume depletion. J Appl Physiol
83, 11±17.
Hesser, C.M. & Matell, G. 1965. Effect of light and moderate
exercise on alveolar-arterial O2 tension difference in man.
Acta Physiol Scand 63, 247±256.
Hopkins, S. R. & McKenzie, D. C. 1989. Hypoxic ventilatory
response and arterial desaturation during heavy work.
J Appl Physiol 67, 1119±1124.
Hopkins, S.R., McKenzie, D.C., Schoene, R.B., Glenny, R.W.
& Robertson, H.T. 1994. Pulmonary gas exchange during
exercise in athletes I. Ventilation-perfusion mismatch and
diffusion limitation. J Appl Physiol 77, 912±917.
Hopkins, S.R., Schoene, R.B., Henderson, W.R., Spragg, R.
G., Martin, T. R. & West, J. B. 1997. Intense exercise
impairs the integrity on the pulmonary blood±gas barrier in
elite athletes. Am J Respir Crit Care Med 155, 1090±1094.
Hornby, L., Coates, A.L. & Lands, L.C. 1995. Effect of
analyzer on determination of mixed venous PCO2 and
cardiac output during exercise. J Appl Physiol 79, 1032±
1038.
Innes, J.A., Solarte, I., Huszczuk, E., Yeh, E., Whipp, B.J. &
Wasserman, K. 1989. Respiration during recovery from
exercise: effects of trapping and release of femoral blood
¯ow. J Appl Physiol 67, 2608±2613.
Johnson, R.L., Heigenhauser, G.J.F., Hsia, C.C.W., Jones,
N.L. & Wagner, P.D. 1996. Determinants of gas exchange
and acid±base balance during exercise. In: L. B. Rowell &
J. T. Shepherd (eds) Handbook of Physiology (section 12).
Exercise: Regulation and Integration of Multiple Systems, pp. 515±
584. Oxford University Press, New York.
Jùrgensen, L.G., Perko, M., Hanel, B., Schroeder, T.V. &
Secher, N.H. 1992. Middle cerebral artery ¯ow velocity and
blood ¯ow during exercise and muscle ischemia in humans.
J Appl Physiol 72, 1123±1132.
Kronenberg, R.S., Safar, P., Lee, J. et al. 1971. Pulmonary
artery pressure and alveolar gas exchange in man during
acclimatization to 12,470 ft. J Clin Invest 50, 827±837.
Medbù, J.I., Mohn, A.C., Tabata, I., Bahr, R., Vaage, O. &
Sejersted, O.M. 1988. Anaerobic capacity determined by
accumulated O2 de®cit. J Appl Physiol 64, 50±60.
Miyachi, M. & Tabata, I. 1992. Relationship between arterial
oxygen desaturation and ventilation during maximal
exercise. J Appl Physiol 73, 2588±2591.
Nielsen, H.B., Svendsen, L.B., Hartvig, T. & Secher, N.H.
1995. Exercise-induced gastric mucosal acidosis. Med Sci
Sports Exerc 27, 1003±1006.
Pedersen, P.K., Mandùe, H., Jensen, K., Andersen, C. &
Madsen, K. 1996. Reduced arterial O2 saturation during
supine exercise in highly trained cyclists. Acta Physiol Scand
158, 323±331.
Peltonen, J.E., RantamaÈki, J., NiittymaÈki, S.P.T., Sweins, K.,
Viitasalo, J.T. & Rusko, H.K. 1995. Effects of oxygen
fraction in inspired air on rowing performance. Med Sci
Sports Exerc 27, 573±579.
Powers, S.K., Dodd, J.A., Woodward, J., Beadle, R.E. &
Church, G. 1984. Haemoglobin saturation during
incremental arm and leg exercise. Br J Sports Med 18,
212±126.
Powers, S.K., Lawler, L., Dempsey, J.A., Dodd, S. & Landry,
G. 1989. Effects of incomplete pulmonary gas exchange on
VO2max. J Appl Physiol 66, 2491±2495.
Rasmussen, J., Hanel, B., Diamant, B. & Secher, N.H. 1991.
Muscle mass effect on arterial desaturation after maximal
exercise. Med Sci Sports Exerc 23, 1349±1352.
Rowell, L.B., Taylor, H.L., Wang, Y. & Carlson. W.S. 1964.
Saturation of arterial blood with oxygen during maximal
exercise. J Appl Physiol 19, 284±286.
Saltin, B., Gagge, A.P., Bergh, U. & Stolwijk, J.A.J. 1972.
Body temperatures and sweating during exhaustive
exercise. J Appl Physiol 32, 284±286.
Severinghaus, J.W. 1958. Blood O2 dissociation line charts:
man. In: D. S. Dittmer & R. M. Grebe (eds) Handbook of
Respiration, pp. 72±73. W. B. Saunders Company,
Philadelphia.
Severinghaus, J.W. 1979. Simple, accurate equations for
human blood O2 dissociation computations. J Appl Physiol
46, 599±602.
Steinacker, J.M., Both, M. & Whipp, B.J. 1993. Pulmonary
mechanics and entrainment of respiration and stroke rate
during rowing. Int J Sports Med 14, S15±S19.
Turcotte, R., Kiteala, L., Marcotte, J.E. & Perrault, H. 1997.
Exercise-induced oxyhemoglobin desaturation and
96 Ó 1998 Scandinavian Physiological Society
O2 content and uptake � H B Nielsen et al. Acta Physiol Scand 1998, 164, 89±97
pulmonary diffusing capacity during high-intensity exercise.
Eur J Appl Physiol 75, 425±430.
Wasserman, K., Hansen, J.E., Sue, D.Y. & Whipp, B. J. 1987.
Principles of exercise testing and interpretation (Appendix D), p.
253±262. Lea & Febiger, Philadelphia.
Welch, H.G. 1982. Hyperoxia and human performance: a
brief review. Med Sci Sports Exerc 14, 253±261.
Welch, H.G., Bonde-Petersen, F., Graham, T., Klausen, K. &
Secher, N. 1977. Effect of hyperoxia on leg blood ¯ow and
metabolism during exercise. J Appl Physiol 42, 385±390.
Welch, H.G. & Pedersen, P.K. 1981. Measurement of
metabolic rate in hyperoxia. J Appl Physiol 51, 725±731.
West, J.B. & Mathieu-Costello, O. 1995. Stress failure of
pulmonary capillaries as a limiting factor for maximal
exercise. Eur J Appl Physiol 70, 99±108.
Williams, J.H., Powers, S.K. & Stuart, M.K. 1986.
Hemoglobin desaturation in highly trained athletes during
heavy exercise. Med Sci Sports Exerc 18, 168±173.
Ó 1998 Scandinavian Physiological Society 97
Acta Physiol Scand 1998, 164, 89±97 H B Nielsen et al. � O2 content and uptake