Extracellular pH defense against lactic acid in untrained and trained altitude residents

Eur J Appl Physiol (2008) 103:127–137 DOI 10.1007/s00421-008-0675-0

ORIGINAL ARTICLE

Extracellular pH defense against lactic acid in untrained and trained altitude residents D. Bo¨ning Æ J. Rojas Æ M. Serrato Æ O. Reyes Æ L. Coy Æ M. Mora

Accepted: 31 December 2007 / Published online: 15 January 2008 Ó Springer-Verlag 2008

Abstract The assumption that buffering at altitude is deteriorated by bicarbonate (bi) reduction was investigated. Extracellular pH defense against lactic acidosis was estimated from changes (D) in lactic acid ([La]), [HCO3-], pH and PCO2 in plasma, which equilibrates with interstitial fluid. These quantities were measured in earlobe blood during and after incremental bicycle exercise in 10 untrained (UT) and 11 endurance-trained (TR) highlanders (2,600 m). During exercise the capacity of non-bicarbonate buffers (bnbi = -D[La]
DpH-1 - D[HCO3-]
DpH-1) amounted to 40 ± 2 (SEM) and 28 ± 2 mmol l-1 in UT and TR, respectively (P \ 0.01). During recovery bnbi decreased to 20 (UT) and 16 (TR) mmol l-1 (P \ 0.001) corresponding to values expected from hemoglobin, dissolved protein and phosphate concentrations related to

extracellular fluid (ecf). This was accompanied by a larger decrease of base excess after than during exercise for a given D[La]. bbi amounted to 37–41 mmol l-1 being lower than at sea level. The large exercise bnbi was mainly caused by increasing concentrations of buffers due to temporary shrinking of ecf. Tr has lower bnbi in spite of an increased Hb mass mainly because of an expanded ecf compared to UT. In highlanders bnbi is higher than in lowlanders because of larger Hb mass and reduced ecf and counteracts the decrease in [HCO3-]. The amount of bicarbonate is probably reduced by reduction of the ecf at altitude but this is compensated by lower maximal [La] and more effective hyperventilation resulting in attenuated exercise acidosis at exhaustion. Keywords

D. Bo¨ning (&) Institute of Sports Medicine, Charite´-Universita¨tsmedizin Berlin, Arnimallee 22, 14195 Berlin, Germany e-mail: [email protected] J. Rojas
L. Coy Centro de Fisiologia de Ejercicio, Universidad Nacional de Colombia, Bogota´, Colombia e-mail: [email protected] M. Serrato
O. Reyes Centro de Servicios Biomedico, Coldeportes, Bogota´, Colombia e-mail: [email protected] O. Reyes e-mail: [email protected] M. Mora Departamento de Nutricio´n, Universidad Nacional de Colombia, Bogota´, Colombia e-mail: [email protected]

Base excess
Buffers
Exercise
Hypoxia

Introduction It is general belief that the defense of extracellular pH constancy against exercise-induced lactic acidosis is attenuated at altitude because of the reduced bicarbonate (bi) concentration (reviewed by Cerretelli and Samaja 2003). The increased mass of hemoglobin (Hb), the most important non-bicarbonate (nbi) buffer for the extracellular space (accessible by the Cl-–HCO3- exchange with the red cells, the so-called Hamburger shift) and the exaggerated hyperventilation during exercise seem not to compensate for. However, there have been occasional observations that buffering is better than expected. Before and after stays at moderate or extreme altitude, Bo¨ning et al. (1980, 2001a) calculated total extracellular buffer capacity (b) from changes (D) of lactic acid [La] concentration and pH (b = -D[La] 9 DpH-1) in arterialized

123

128

blood during ergometer tests and found a rise of b within the first week after descent. The cause was a large increase in non-bicarbonate b (bnbi). Astonishingly bnbi measured by in vivo CO2 titration at rest using hypo- and hypercapnia showed no significant rise. Other indications for hypoxia-induced improvements of buffering are unexpectedly small decreases in base excess (BE) compared to increases in blood [La] in highlanders (Favier et al. 1995; Wagner et al. 2001). The following factors play a role for the discrepancies. First, hematocrit increases at altitude in Caucasians and Amerindians; since erythrocytic [La] is smaller than plasma [La], whole blood [La] is reduced. If [La]blood and plasma pH are used for calculation of buffering, b in the extracellular space is underestimated. Second, the results also depend on the time and conditions of blood sampling. Often blood is taken after stopping exercise. But with interruption of work rapid changes of buffering occur which we have analyzed in recent sea level experiments. Total and bi b during and after exercise were calculated from D[La], D[HCO3-] and DpH in plasma (Bo¨ning et al. 2007b). Whereas bbi remained rather constant, bnbi (btotal - bbi) during recovery corresponded to values previously found by in vivo CO2 titration at rest (10–17 mmol l-1), but increased to 32 mmol l-1 in untrained (UT) and to 20 mmol l-1 in endurance trained (TR) subjects during exercise. Main causes seem to be increasing buffer concentrations due to temporary shrinking of the extracellular fluid (ecf) by water shift to the working muscle fibres, exchange of small amounts of HCO3- or H+ with cells and a delayed equilibration of HCO3- between blood and interstitial fluid. A related phenomenon is the rapid decrease of BE at rather constant [La]blood within the first minutes after exercise (Bo¨ning et al. 2007a). Additionally the change in BE is consistently larger than that of [La]blood. For actual BE (BE of blood) we could show that this is not caused by a larger entrance of H+ than La- into the extracellular space as previously suggested (e.g. Bangsbo et al. 1997; Juel et al. 2004), but by the Donnan effect causing an additional interchange of Cl- versus HCO3- across the erythrocytic membrane. The difference became insignificant when comparing standard base excess (SBE) which is defined for the extracellular space plus erythrocytes with the mean estimated [La] in this space. Thus the classical idea of lactic acid as cause of exercise acidosis remains valid. In the present paper these new aspects of extracellular buffering against La have been investigated in UT and TR residents of moderate altitude. To get a more complete synopsis of the participating factors, we performed additionally in vitro titrations of blood. The following questions have been addressed:

123

Eur J Appl Physiol (2008) 103:127–137

1.

2.

3.

Are differences in buffering during and after exercise and between UT and TR also detectable in highlanders? Is the occasionally observed improved buffering at altitude a general phenomenon and which might be the possible mechanisms? Are discrepancies between BE and [La] changes also observable at altitude?

Methods Measurements were performed in 15 untrained (UT) and 14 trained (TR) non-smoking male subjects in Bogota´/ Columbia (2,600 m above sea level), but technical problems reduced the number of experiments with sufficient measurements of acid–base status during exercise to 10 and 11, respectively. Additional data (Hb mass, blood volume) are published elsewhere (Bo¨ning et al. 2001b). Informed consent was obtained from all participants, the study protocol was approved by the ethics committee of the faculty. The subjects were born in Bogota´ except four living there since at least 5 years. The UT (age 22.7 ± 1.2 (SE) years, body mass 63.2 ± 1.7 kg, height 172.6 ± 1.6 cm, body _ 2 peak mass index (BMI) 21.2 ± 0.4 kg m-2, VO -1 -1 41.7 ± 1.7 l min kg , peak power 3.0 ± 0.2 W kg-1) were university students. The TR group consisted of runners (36–80 km training distance per week) and one Karate fighter performing regular endurance training (age 25.4 ± 1.7 years, body mass 59.6 ± 1.3 kg, height _ 2 peak 169.2 ± 1.5 cm, BMI 20.7 ± 0.5 kg m-2, VO -1 -1 55.2 + 1.2 l min kg ). All subjects were clinically healthy. The subjects performed an incremental test (+40 W every 3 min, beginning with 40 W in untrained and 80 W in trained subjects) until subjective exhaustion on a bicycle ergometer (ER 900, Jaeger Wuerzburg, Germany), which was preceded by a control period (10 min) and followed by 21 min of recovery (beginning with 5 min cycling at 25 W _ 2 was measured using to prevent orthostatic problems). VO a Quinton Metabolic Cart QMC-TM. Blood gases and acid–base status (ABL 50, Radiometer Copenhagen, Denmark), hematocrit (Hct, microcentrifugation method, 23,000g, 5 min) and [La] (amperometric lactate oxidase method, Biosen 5030 L, Envitec-Wismar, Germany) were measured in heparinized blood samples taken from hyperemized earlobes 6 and 3 min before the start of cycling, every 3 min during and after exercise, and additionally at exhaustion. One Hct capillary (75 ll, maximum sampling time 5 s) was filled simultaneously with a blood gas capillary (100 ll). All blood measurements were performed immediately or after a short delay with intermittent storage

Eur J Appl Physiol (2008) 103:127–137

on ice. [La] was determined in plasma (all samples) and in erythrocytes (control, exhaustion, after 6, 12 and 18 min of recovery). A more extended description of methods has been included in the preceding papers (Bo¨ning et al. 2007a, b). Venous blood (5 ml) sampled in the morning was equilibrated in syringes with air and then with varying amounts of CO2. After measurement of oxygen saturation (SO2), [Hb] (both with a hemoxymeter OSM 3, Radiometer Copenhagen, Denmark), pH and PCO2, CO2 dissociation curves (SO2 [ 90%) were constructed to obtain in vitro bnbi. CO2 dissociation curves were also obtained for oxygenated hemolysed cells diluted 1:4 with a salt solution (135 mmol l-1 KCl and 15 mmol l-1 NaHCO3) which had been stored. Because of storage at -30°C the methemoglobin percentage was high (mean 40%, up to 63%); this is without relevance for buffering properties because the quaternary globin structure is equal to that of oxygenated Hb.

Calculations La is the sum of undissociated (HLa, \1%) and dissociated (La-,[99%) lactic acid. [La]ery was corrected for 2% trapped plasma. The calculation of ABE was performed using SO2 (calculated from measured PO2 and pH) and [Hb], that of SBE using a standardized [Hb] of 5 g dl-1 (Christiansen 1981). The underlying assumption for the determination of SBE is that a mixture of erythrocytes, plasma and interstitial fluid possesses one-third of non-bicarbonate buffer capacity of blood corresponding to a 1:3 dilution with an unbuffered solution. This is slightly different from the real volume ratio (1:3.2, Rowlands 2005) at sea level but the applied value considers the small amount of buffering proteins and phosphates in the interstitial fluid. Small increases of [Hb] like at moderate altitude or during exercise produce only negligible changes for both ABE and BE. Since we had originally not planned to analyse the relation between BE and [La], [La]blood was not measured and [La]ery only in part of the samples for other purposes. [La]blood was therefore calculated from [La]plasma, [La]ery and Hct where possible. Corresponding assumptions about red cell, plasma and interstitial volumes as for SBE have to be made for the calculation of mean [La]ecf + ery from [La]plasma and [La]ery. Mean [La]ecf was assumed to be equal to [La]plasma in arterialized blood which cannot differ markedly from the mixed venous value. For a measured Hct value of 50% with 2% (relative) trapped plasma and a relation of 0.91 between whole body and central hematocrit (Gregersen and Rawson 1959) the fractions of red cell and extracellular volumes are 0.14 and 0.86 (possible volume

129

changes at altitude are considered in ‘‘Discussion’’). The resulting equation is: ½Laecfþery ¼ 0:14
½Laery þ 0:86
½Laplasma In the sea level study (Bo¨ning et al. 2007a) we used a similar equation for [La]blood and [La]plasma since [La]blood shows less scattering than [La]ery. Deviations of [La], [HCO3-], PCO2 and pH in plasma from baseline were calculated for each sample during and after exercise. Mean values of -D[La] plasma
DpH-1 and the other ratios at each time as well as the slope of the titration curves for measurements at subsequent sampling times yielded information about the different components of pH defense. All components of a buffer capacity add up algebraically. -D[La]
DpH-1 contains total buffering; if the PCO2 is reduced by hyperventilation, it also includes respiratory (r) compensation. D[HCO3-]
DpH-1 contains bicarbonate buffering and eventually respiratory compensation but not non-bicarbonate buffering, thus the difference between -D[La]
DpH-1 and D[HCO3-]
DpH-1 should yield bnbi. Correcting for the effects of varying PCO2 on pH and [HCO3-] excludes respiratory and leaves only non-respiratory (nr) effects; we obtain total buffering (btot = -D[La]
DpHnr-1) and bicarbonate buffering (bbi = D[HCO3-]nr
DpHnr-1), if no additional reactions play a role in pH defense. Consumption or cotransport of both components of lactic acid, i.e. La- and H+, to other compartments induces a movement along the titration curve but does not influence the magnitude of -D(La)
DpH-1. However, if only one component is concerned, this is visible as an apparent change of buffering. bnbi is theoretically equal to -D[HCO3-]r
DpHr-1, the slope of the CO2-equilibration line of the ecf. Using this relation and the equation DPCO2
DpHr-1 = 1.04 + bnbi [HCO3-] (Bo¨ning et al. 2007b) we obtained [HCO3-] and pH changes without respiratory compensation (D[HCO3-]nr and DpHnr). Since the large bnbi found during exercise does not fully correspond to D[HCO3-]r
DpHr-1 (see ‘‘Discussion’’) and individual variation exerted only a negligible influence, we applied mean recovery values of bnbi, allowing for extracellular volume changes during exercise estimated from Hct changes.

Statistics The data is presented as mean ± standard errors (SEM). Dependent on the number of comparisons, t tests or analysis of variance with following Bonferroni corrected t tests were used for significance calculations.

123

130

Eur J Appl Physiol (2008) 103:127–137

Results

. -1

Acid–base status and hematocrit Figure 1 shows the time course of acid–base changes in plasma during and after the incremental test for the UT. At rest acid–base status presents values typical for respiratory alkalosis compensated by non-respiratory acidosis. After exhaustion acidosis is initially more marked than at peak exercise despite only small additional changes in [La]plasma and an exaggerated hyperventilation effect. PCO2 is reduced at high workloads and even more after stopping exercise with low values throughout the recovery period. In the TR (not shown) changes are similar at equal relative intensity; DpH at exhaustion is 0.025 units larger. Hct rises until 6 min of recovery and does not reach initial values again (initial, maximum, 6 and 21 min of recovery: UT 0.498 ± 0.006, 0.536 ± 0.007, 0.547 ± 0.006, 0.536 ± 0.008; TR 0.488 ± 0.006, 0.530 ± 0.011, 0.537 ± 0.011, 0.516 ± 0.009).

Changes of [La] and BE The relation between [La]plasma and [La]ery during exercise and recovery is shown in Fig. 2 for the UT, the reaction of the TR (not shown) is equal. [La]ery is consistently smaller than [La]plasma and, as visible from the decreasing ratio [La]ery/[La]plasma, rises with delay until the first minutes of recovery. The initial value of the ratio (0.8) is not reached again during recovery but remains below 0.6 (P \ 0.01). Changes of [La]blood as well as ABE are presented in Fig. 3. In the UT the decrease of ABE which reaches its lowest values in the sixth minute of recovery is as large as

20.0 ∆pCO2 (mmHg)

Untrained

0.20

15.0 ∆[La ]

0.10

10.0

0.05

5.0

0.00 ∆pH

-1

∆[HCO3 ] (mmol l ) -1 (mmol l )

∆[La]

0.15

0.0

-0.05

∆PCO2

-0.10

∆[HCO3 ]

-0.15

-5.0 -10.0 -15.0

∆pH

-0.20

-20.0

-0.25

-25.0 40

80 120 160 Max Watt

3

6

9

12

15

18

21 min

Fig. 1 Temporal course of changes (D) relevant for acid–base status in blood of untrained subjects. Lactate concentrations in plasma. Mean and SEM are omitted for the sake of clarity. Initial values: [La] 1.5 ± 0.1 mmol l-1, PCO2 32.0 ± 0.6 mmHg, [HCO3-] 21.1 ± 0.6 mmol l-1, pH 7.433 ± 0.005

123

[La]ery /[La] plasma

[La] (mmol l ) 20.0 18.0 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0

[La]plasma

[La]ery/[La]plasma

[La] ery

0

40

80 120 160 Max Watt

3

6

9

1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00

12 15 18 21 min

Fig. 2 Temporal course of lactate concentrations in erythrocytes and plasma and their ratio during and after the incremental test for the untrained subjects (mean and 1 SEM)

the increase of [La]blood during exercise. After exhaustion, however, the difference -DABE - D[La]blood suddenly increases to +4.3 mmol l-1 after 6 min of recovery and remains high for the rest of the experiment. In the TR the quantities vary equally during recovery but -DABE is slightly larger than D[La]blood already during exercise. In contrast, the decrease of SBE (Fig. 4) is smaller than D[La]ecf + ery during exercise (3 mmol l-1 in UT, 2 mmol l-1 in TR) and approximately equal afterwards.

Extracellular buffering Figure 5 shows the time course of -D[La]plasma
DpH-1 and D[HCO3-]plasma
DpH-1. Both ratios increase initially during exercise and again towards the end of the recovery period in both UT and TR. After rather high values at heavy work loads, however, there is a marked reduction of -D[La]plasma
DpH-1 shortly after exercise. The difference between this ratio and D[HCO3-]
DpH-1, corresponding to bnbi, is decreased during the recovery period in all subjects; additionally bnbi is lower in TR than UT during the whole experiment. The rise in total pH defense towards the end of exercise and with progressing recovery is caused by a relatively higher contribution of hyperventilation as it subsides after correcting the values to constant initial PCO2 (Fig. 6): bnbi is again reduced during recovery compared to exercise. In contrast bicarbonate buffering (D[HCO3-]nr
DpHnr-1) remains fairly constant after an initial decrease during exercise. In Fig. 7 titration curves using single values are presented exemplarily for the UT in a D[La]plasma (upper part) or D[HCO3-] (lower part) versus DpH diagram. The distribution of points is approximately linear for both relations if considered separately during and after exercise. During recovery the slope of the regression curves is reduced. This

Eur J Appl Physiol (2008) 103:127–137

131

Untrained

20 18

#

16

*

-∆ΑΒΕ

18

*

16

#

*

12 10 8

∆[La]blood

6 4

∆[La]ecf+ery

* *

14 . -1 ∆[Acid] (mmol l )

. -1 ∆[Acid] (mmol l )

14

12

#

10 8

-∆SΒΕ

*

6 4

2

2



0



0

160

Max Watt

6

9

12

15

18 min

160

Max

6

9

12

15

Watt

Trained

20

18 min

Trained

20

18

18 #

16

*

14

-∆ΑΒΕ

*

*

12 10

*

8 6

∆[La]blood

4

. -1 ∆[Acid] (mmol l )

16 . -1 ∆[Acid] (mmol l )

Untrained

20

*

14

-∆SΒΕ

#

12 # #

10 8

*

*

6

∆[La]ecf+ery

4

2



2



0 200

Max Watt

6

9

12

15

18 min

Fig. 3 Changes (D) in actual base excess (ABE, diamonds) and [La]blood (squares) in untrained and trained subjects. To show the uptake of acid, DABE is multiplied by -1. Initial values (mean and 1 SEM): Untrained ABE -1.4 ± 0.6 mmol l-1, [La]blood 1.3 ± 0.1 mmol l-1; trained ABE -2.1 ± 0.5 mmol l-1, [La]blood 1.4 ± 0.1 mmol l-1. * Significant differences between -DABS and D[La]blood, # significant changes of -DABS - D[La]blood compared to maximal exercise (P \ 0.05 or better)

is partly an apparent effect resulting from the increasing relative contribution of respiratory compensation. However, if the respiratory compensation is mathematically excluded, an attenuation of total buffering after exercise is visible (Fig. 7, lower panel). DpHnr is markedly more negative for a given D[La]plasma than during work. As in Fig. 6, this seems to be due to the reduction of the bnbi, since bbi behaves rather uniform throughout the entire experiment with little scattering and may be approximated also by one common regression line. In reality the relation is slightly curved because it is determined by the logarithmic Henderson–Hasselbalch equation for constant

0 200

Max Watt

6

9

12

15

18 min

Fig. 4 Changes (D) in standard base excess (SBE, diamonds) and [La]ecf + ery (squares) in untrained and trained subjects. To show the uptake of acid, DSBE is multiplied by -1. Initial values (mean and 1 SEM): untrained SBE -2.3 ± 0.6 mmol l-1, [La]ecf + ery 1.4 ± 0.1 mmol l-1; trained SBE -3.0 ± 0.5 mmol l-1, [La]ecf + ery 1.5 ± 0.1 mmol l-1. * Significant differences between -DSBS and D[La]blood, # significant changes of -DSBS - D[La]ecf + ery compared to maximal exercise (P \ 0.05 or better)

PCO2 (Bo¨ning et al. 2007b), but for the sake of simplicity we have applied only linear regression. This calculation yields the mean bbi for the part of the titration curve under consideration. In Table 1 a synopsis of the different components of pH defense calculated from means of individual regression coefficients is shown. All defense mechanisms tend to lower values in the TR than in the UT, the decreases in bnbi (-30%) and in the sums are significant after cycling. Comparing exercise and recovery, bbi shows only small differences, if the recovery regression includes not only measurements after exercise (used for calculation of bnbi)

123

132

Eur J Appl Physiol (2008) 103:127–137 Untrained

Untrained

140

140 . -1

80 60 .

-

∆[HCO3 ] ∆pH

40

-1

20

80 60 40

.

-∆[La] ∆pHnr

-1

- .

-1

∆[HCO3 ] ∆pHnr

20 0

0 -20

100

-1

− ∆[La] ∆pH

100

-1

.

.

-1

. -1

∆[Acid] ∆pH (mmol l )

.

∆[Acid] ∆pH (mmol l )

120 120

40 80 40

80 120 160 Max Watt

3

6

9

12

15

18

-20

21 min

120 160 Max Watt

3

6

9

12

15

18

21 min

Time (min)

Time (min)

Trained

140 Trained

140

120 .

∆[Acid] ∆pH (mmol l )

.

. -1

− ∆[La] ∆pH-1

100

-1

80 60

.

-1

. -1

∆[Acid] ∆pH (mmol l )

120

-

.

-1

∆[HCO3 ] ∆pH

40 20

100 80

.

∆[La] ∆pHnr-1 -∆

60 40 - .

∆[HCO3 ] ∆pHnr

20

-1

0 0 -20

80 120 160 200 Max

3 6 Time (min)

9

12

15

18

21 min

Fig. 5 Temporal course of pH defense in plasma for untrained and trained subjects (mean and 1 SEM). Values calculated with measured PCO2. The large scattering in the 21st min of the TR is caused by a reduced number of samples and small pH differences

-20

80 120 160 200 Max Watt

3

6

9

12

15

18

21 min

Time (min)

Fig. 6 Temporal course of pH defense in plasma for untrained and trained subjects (mean and 1 SEM). Values calculated with constant PCO2

Discussion but all values during the whole experiment (corresponding to back-titration to initial values). The lower bbi during recovery is only caused by the more acid pH values shortly after stopping exercise falling on the flattening part of the buffer curve (Fig. 7b). bnbi during exercise is approximately two times greater than during recovery in both groups. No realistic average value of respiratory compensation can be calculated during recovery from the regression coefficients, since it increases with decreasing acid load. Total pH defense is markedly higher during exercise in the UT than the TR.

In vitro buffering The nbi buffer value in blood drawn before exercise (Table 2) is relatively high for the UT (values in lowlanders approximately 28–30 mmol l-1, Bo¨ning 1974; Siggaard-Andersen 1974) and tends to lower values in the TR. In contrast there is a slight tendency for increased values (related to 1 g Hb) in the stored hemolysate of the TR.

123

General The results of this investigation show markedly higher bnbi during exercise in altitude residents than in lowlanders, which compensate the decrease in bbi as shown later in the discussion. This is in contrast to the traditional view but confirms earlier studies of our group in altitude sojourners after descent (Bo¨ning et al. 1980, 2001a). bnbi during recovery tends to higher values than in lowlanders (Bo¨ning et al. 2007b) as well. Differences in pH defense between exercise and rest as well as between UT and TR observed at sea level are also found at altitude. Finally differences between -DBE and +D[La] resemble findings in lowlanders (Bo¨ning et al. 2007a).

Lactate concentrations The changes of [La]plasma and [La]ery are similar to those observed at sea level (Bo¨ning et al. 2007b), but [La]plasma reach 3 mmol l-1 lower maximal values (UT: P \ 0.01, TR: P \ 0.05). Since [La] is practically equal to [La-],

Eur J Appl Physiol (2008) 103:127–137

133 25

A

20

∆[La] . -1 (mmol l )

Table 1 Synopsis of extracellular pH defense during and after exercise in untrained (UT) and trained (TR) subjects

15

Exercise

Recovery

10

Bicarbonate buffers (bbi)

5 0 -5

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

41 ± 1

38 ± 2***

38 ± 1

37 ± 1**

Non-bicarbonate buffers (bnbi)

-10 -15

UT TR ∆[HCO3 ] . -1

-20 (mmol l ) 0.00 -0.05 0.05

∆pH

UT

40 ± 2

20 ± 3***

TR

28 ± 2




16 ± 2**

81 ± 2

58 ± 3***

66 ± 2



53 ± 2***

Total buffers (btot) UT TR

25

B

20

∆[La ] . -1 (mmol l )

15 10

Respiratory compensation UT

27 ± 4

TR

21 ± 3

Total pH defense

5

UT

0

TR

108 ± 5 87 ± 3




-5 -10 -15

∆[HCO3 ]nr . -1

(mmol l )

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

-20 0.00 -0.05

0.05

∆pHnr

Fig. 7 Relation between changes in [La]plasma (squares) and [HCO3-]plasma (diamonds), and pH changes in earlobe blood during (closed symbols continuous regression lines) and after (open symbols dashed regression lines) exercise for untrained subjects. The 3- and 6min recovery values are not included in the regression analysis to exclude after-effects of exercise. a Measured values. b Values after correction of [HCO3-] and pH for respiratory compensation. All correlation coefficients (0.80 or better) are highly significant

Mean and SEM of individual regression coefficients and derived differences (mmol l-1 per pH unit) for untrained (UT) and trained (TR) subjects * P \ 0.05, ** P \ 0.02, *** P \ 0.01 between exercise and recovery;
P \ 0.05,

P \ 0.02,


P \ 0.01 between groups. Since respiratory compensation increases with decreasing acid load during recovery, no realistic mean value can be calculated for this period. Bicarbonate buffers for recovery are calculated including all measured values during the whole experiment. Total pH defense contains all defense mechanisms except the common shift of La or H+ to other compartments. Not measured acids like pyruvic acid are not included in the calculation

Table 2 In vitro non-bicarbonate buffer capacities (b) UT

[La]ery/[La]plasma should be governed by a Donnan distribution valid for all diffusible anions (An-) like Cl-, HCO3- and La- caused by the large amount of non-diffusible ions of Hb and organic phosphates in the red cells. The Donnan ratio r ([An-]ery/[An-]plasma) is dependent on pH: the higher the pH the lower is r because of enforced dissociation of non-diffusible buffers. The expected value of r (about 0.50 at rest and 0.56 after exercise for concentrations per l red cells) is not different from predicted values at sea level (for calculation see Table 1 in Bo¨ning et al. 2007a); deviations of pH and arterial SO2 (changing [Hb-] by the Haldane effect) are too small to exert a remarkable effect. Only the rise of [2,3-diphosphoglycerate] in highlanders (+5 lmol g Hb-1 in Bogota´, Schmidt et al. 1999) might cause an appreciable change (reduction of r by 0.05). The expected value of r is observed only after exercise when La concentrations are increased and changing slowly. Before exercise it is higher ([0.8) than theoretically

TR

bblood (mmol l-1)

34.2 ± 3.8 (11)

27.6 ± 1.2 (13)*

bhl (mmol gHb-1)

0.180 ± 0.005 (15)

0.190 ± 0.005 (13)*

Mean and SEM, number of untrained (UT) and trained (TR) subjects in parenthesis hl hemolyzed stored red cells diluted to [Hb] of 7 g/dl Significance * P \ 0.1, calculations were performed with one-tailed t test for the hemolysate, since increased hemolysate values because of reduced red cell age in the TR may be expected

expected probably because of continuous production of La by the erythrocytic metabolism. During exercise the increase of [La]plasma is too rapid since the La- - H+ cotransporters (mainly MCT 1) need several minutes to establish the equilibrium; therefore r falls to between 0.46 and 0.48 at exhaustion. Possibly an increased amount of MCT 1 as in subjects acclimatized to 4,100 m (Juel et al. 2003) plays a role since the corresponding r in Berlin were only 0.40 (UT) and 0.43 (TR).

123

134

Eur J Appl Physiol (2008) 103:127–137

Relation between BE and [La] changes

Hamburger shift between plasma and red cells (Bo¨ning et al. 2007a).

The larger changes of [La]ecf + ery than -DSBE during exercise indicate either entrance of more La- than H+ into the extracellular space (alternatively also exit of more H+ than La- is possible) or additional buffering (e.g. increase of [HCO3-] by shrinking of the ecf during muscle contractions, see further discussion). The lacking difference between D[La]ecf + ery and -DSBE during recovery after rapid return of the extracellular volume to approximately initial values (Bo¨ning et al. 2007b) suggests an equal transport of H+ and La- across cell membranes. The idea that more H+ than La- leave the muscle during or shortly after exercise (e.g. Bangsbo et al. 1997; Juel et al. 2004) is not supported by these experiments as well as by our corresponding investigation at sea level. Obviously the moderate increases of various muscle membrane proteins important for H+ transport after altitude acclimation (Na+–H+ exchanger NHE1, Na+–HCO3- cotransporter, carboanhydrase CA IV, Juel et al. 2003) either do not exist in our subjects or do not markedly influence the relation between H+ and La- movements across the sarcolemma. One might suggest that the calculation of SBE and [La]ecf + ery is influenced by the fact that the [Hb] is increased and the extracellular volume is reduced at altitude (see further discussion). But if one assumes rather large changes (50%, extracellular values of 7.5 instead of 5 g/dl for [Hb] and 0.21 instead of 0.14 for Hct) DSBE decreases by maximally 0.2 mmol l-1 and D[La]ecf + ery increases by maximally 0.9 mmol l-1. Thus the effect in our experiments should be negligible. A decrease of ABE is a correct indicator for the uptake of a specific acid only for blood in vitro but biased in vivo by an interchange of HCO3- and Cl- across the capillary wall. The positive difference between -DABE and D[La]blood during recovery is also found at sea level and explained by a concomitant uptake of Cl- in exchange for HCO3- from the interstitial fluid as a consequence of the

pH defense Considering the titration curves for constant pCO2 in Fig. 7b, D[HCO3-]nr versus DpHnr behaves as expected with slightly decreasing slope (bbi) at lower pH. bnbi is constant according to in vivo titrations with CO2 (e.g. Bo¨ning et al. 1999). Consequently the lines for the sum of all buffers (D[La]plasma vs. DpHnr-1) should be similarly curved as the bicarbonate buffer lines. This is not visible probably because of scattering and has been neglected over the short DpH ranges (usually 0.2 units) which were evaluated for calculations. A synopsis of pH defense during exercise with and without altitude acclimation is presented in Table 3. bbi is generally reduced by altitude effects. In the highlanders of the present study bnbi is clearly increased but not as much as in mountaineers after a Himalayan expedition (Bo¨ning et al. 2001a); whether the high value in the latter is caused by more extreme altitude, the enormous rise in the proportion of young red cells or large scattering remains open. In spite of the reduced bbi the maximal pH decrease during exercise is 0.05 (UT) and 0.03 (TR) pH units, respectively, smaller in Bogota´ than in Berlin caused by higher bnbi but also by a reduced D[La] and a more effective PCO2 decrease because of the lower initial values (a given change in PCO2 exerts a more marked effect on pH between 30 and 20 than between 40 and 30 mmHg, Siggaard-Andersen 1974). Also bnbi during recovery is higher by 33–43% than in lowlanders (16 ± 2 vs. 12 ± 1 mmol l-1 in TR, 20 ± 3 vs. 14 ± 3 mmol l-1 in UT, sea level values from Bo¨ning et al. 2007b), the difference is significant for the pooled values of UT and TR in analysis of variance (P \ 0.01). All bnbi are in reality approximately 1 mmol l-1 larger because of appearance of other acids (approximately 0.5 mmol l-1, Bo¨ning et al. 2007a).

Table 3 Influence of altitude acclimatization on pH defense during exercise

Bicarbonate buffers (bbi) Non-bicarbonate buffers (bnbi) Respiratory compensation Total pH defense

UT-Berlin

TR-Berlin

UT-Bogota´

46 ± 1 32 ± 2

48 ± 1 20 ± 2




41 ± 1** 40 ± 2*

TR-Bogota´

Mountaineers before expedition

38 ± 1*** 28 ± 2**,




48 ± 1 25 ± 3 19 ± 4

22 ± 3

23 ± 3

27 ± 4

21 ± 3

100 ± 4

91 ± 3

108 ± 5

87 ± 3






92 ± 6

Mountaineers after expedition 44 ± 1# 61 ± 16# 23 ± 5 128 ± 12##

Mean and SEM (unit mmol l-1 per pH unit) for untrained (UT) and trained (TR) subjects. Comparison Berlin (32 m above sea level, Bo¨ning et al. 2007b) versus Bogota´: * P \ 0,05, ** P \ 0,02, *** P \ 0,01;
for corresponding comparisons untrained versus trained; #Comparisons before and after the expedition. Mountaineers (data from Bo¨ning et al. 2001a): measurements performed in Ulm/Germany (700 m above sea level), six subjects who had stayed 37 days in the Himalaya 7–8 days after descent below 2,800 m

123

Eur J Appl Physiol (2008) 103:127–137

The rise of bnbi during exercise compared to recovery at sea level is mainly caused by the temporary shrinking of the extracellular volume resulting from water shift into the muscle fibres leading to a rise of bi as well as nbi buffer concentrations (Bo¨ning et al. 2007b). This is an osmotic effect caused by rising metabolite concentrations in the cells (Lundvall et al. 1972). Since this shrinkage occurs during the titration after measurement of the control values, the increase of the steepness of the titration curve is exaggerated. This can be detected, if the control values of the acid–base status (except PCO2) are correspondingly corrected for the volume reduction which leads to higher initial pH and [HCO3-] (Bo¨ning et al. 2007b). A similar but contrary effect is dilutional acidosis (acidosis caused by infusion of electrolyte solutions without buffers if PCO2 remains constant, e.g. Lang and Zander 2005). In addition retarded equilibration of HCO3- between plasma and interstitial fluid because of reduced contact time and of Lacompared to H+ between plasma and red cells during rapid changes with incremental exercise might play a role. All HCO3- movements during titration cause more marked changes of the steepness of the titration curve for all buffers than for that of bicarbonate alone thus increasing bnbi (Bo¨ning et al. 2007b). The smaller effects in TR seem to be caused by their larger extracellular volume, which shrinks relatively less during exercise as well as by enhanced equilibration of HCO3- and La-. But since the total amount of HCO3- and the solvent volume for La are increased, the puzzling reduction of b in endurance-trained subjects is mainly compensated for. There might be various causes for the additionally improved nbi buffering in the highlanders. From the increase in Hb mass compared to sea level (approximately 12% in UT and 9% in TR, Bo¨ning et al. 2001b; Robinson et al. 2007) one might expect a corresponding increase in bnbi by only 2 mmol l-1 at rest. The effect should be larger because of two effects: extracellular volume and total body water seem to be reduced in highlanders (Ramirez et al. 1998) as well as in sojourners (rev. by Milledge 1992) and the buffer properties of erythrocytes might be enforced by a rise in [2,3-diphospoglycerate] (Schmidt et al. 1999) as well as by a reduced mean red cell age in trained residents (Bo¨ning et al. 2001b). Unfortunately quantitative data for the ecf of altitude residents seems not to exist, whereas a decrease by 6–14% has been described in sojourners after ascent (Malapartida and Moncloa 1967; Jain et al. 1980). But a marked reduction of total body water in Colombians from Pasto (2,600 m of altitude) compared to lowlanders in the United States (426 vs. 519 g kg-1 body mass, -18%, Ramirez et al. 1998) suggests a comparable effect in extracellular volume. Also a reduction in plasma volume, which is approximately 10 and 8% in our UT and TR, respectively,

135

lower than in comparable groups in Germany (Bo¨ning et al. 2001b; Robinson et al. 2007) may contribute to changes of the ecf. A rough estimate of the extracellular volume for our subjects can be obtained from in vitro bnbi of blood and blood volume applying the indicator dilution method. The product of both quantities divided by the extracellular bnbi corrected for interstitial buffers (dissolved proteins approximately 20 g l-1 and phosphates approximately 1.5 mmol l-1 corresponding to b = 3 mmol l-1) and additional acids (b = 1 mmol l-1) yields the extracellular volume after subtraction of the red cell volume. The values obtained with bnbi during recovery are 0.12 l kg-1 for UT and 0.16 l kg-1 for TR in Bogota´ compared to 0.16 l kg-1 for UT in Berlin, the latter falling into the range of sulphate spaces at rest (Hamadeh et al. 1999). Thus one might suggest a decrease of the extracellular volume in Bogota´ for UT (-25%) and a corresponding change in TR. In any case the increase of the recovery values of bnbi by 33–43% at altitude suggests a larger effect than caused by the Hb mass change alone (+9 to 12%). The constriction of the extracellular volume at rest might be the basis also for the clearly larger bnbi during exercise. Since the ratio of bnbi during exercise and rest is similar in Bogota´ and Berlin (approximately 2 in UT and 1.7 in TR) other mechanisms like a relatively larger exercise-induced volume shrinkage are not necessary for the explanation of these results in the highlanders.

Blood buffers The possible effect of increased nbi buffer properties of blood constituents is detectable by in vitro titration of blood with CO2. In former studies we found increased values (up to 44 vs. 31 mmol l-1 before ascent) during and shortly after an altitude stay (Bo¨ning et al. 1980, 1999). This was explainable by a large proportion of young red cells possessing improved buffering properties (Bo¨ning et al. 1999) since Hb concentrations had not markedly changed. In the present investigation only the in vitro bnbi of the UT tends to an increase in spite of a lowered mean cell age in the TR (Bo¨ning et al. 2001b). However, the difference in [Hb] between these groups (UT 17.3, TR 16.0 g/dl, bnbi of 1 g Hb approximately 0.19 mmol according to Siggaard-Andersen 1974) can explain 2.5 mmol l-1 of the observed difference in bnbi (+6.6 mmol l-1 for the UT). Thus there remain approximately 4 mmol l-1 unexplained. If this is a real effect caused by unknown erythrocytic buffers, in vivo bnbi in the UT would rise maximally by 1.5 mmol l-1. Plasma protein concentrations were in the normal range for both groups

123

136

(Bo¨ning et al. 2001b), other plasma buffers like phosphates contribute only little. Measurements in hemolyzed red cells and calculation of bnbi per gram Hb exclude effects of different [Hb] and plasma buffer concentrations. The values correspond to previous results in intact non-stored cells of different age from sea level residents (Bo¨ning et al. 1999). The tendency for an increase in the TR compared to the UT coincides with the slightly higher amount of young erythrocytes (Bo¨ning et al. 2001b). We did not measure in vitro bnbi at the end of exercise. According to calculations in the preceding paper (Bo¨ning et al. 2007b) an increase may be expected. It is, however, mainly caused by the rise of [Hb], [plasma protein] and Hct the latter influencing HCO3- distribution between red cells and plasma and therefore no proof for additional buffers. Thus one may conclude that in altitude residents the high in vivo bnbi during exercise is not caused by increased intraerythrocytic or plasma buffer concentrations whereas in sojourners the partly enormous rise in the proportion of young red cells (Bo¨ning et al. 1999) might contribute to the high in vivo and in vitro nbi buffering.

Eur J Appl Physiol (2008) 103:127–137

increased Hb mass and the reduced extracellular volume. Consequently the additional but temporary shrinking of the extracellular volume by the water shift into the working muscle fibres leads also to larger exercise bnbi at altitude. The smaller effects in TR seem to be caused by an increased extracellular volume as well as by enhanced equilibration of HCO3- and La- during exercise. But since the extracellular volume decrease at rest also reduces the total amount of HCO3-, the old ideas on reduced buffering at altitude apparently cannot be fully rejected. However, the concomitant reduction of maximal [La] and the larger effect of hyperventilation on pH at low PCO2 finally lead to a slightly less marked exercise acidosis in the extracellular space of highlanders. Acknowledgments The investigation was performed during a stay of the first author as visiting professor at the Universidad Nacional de Colombia in Bogota´ supported by this institution and the German Academic Exchange Service. The authors thank J. Gomez, B. Himmelsbach, J. Nadol and I. De Velandia for excellent technical assistance and all the subjects for their willing cooperation.

References Importance of velocity of reactions Juel et al. (2003) have recently shown dramatic increases of MCT1 and anion exchanger AE1 in erythrocytes as well as moderate elevations in muscular Na+–H+ exchanger NHE1, Na+–HCO3- cotransporter, and carboanhydrase CA IV in sojourners, residents or both groups at 4,100 m of altitude. They speculate that these modifications might facilitate the transport of La-, H+ and CO2 from muscle to blood. A more rapid entering of La- into the red cells should not only increase [La]ery during exercise as mentioned above, but also reduce D[La]plasma
DpH-1. Thus the formerly suggested small contribution of a retarded entrance of La into the red cells (Bo¨ning et al. 2007b) to the high bnbi (approximately 1–2 mmol l-1) plays possibly an even less important role in the highlanders. A more rapid increase of HCO3- in plasma without complete equilibration with the interstitial fluid, however, is conceivable and might contribute to the high measured bnbi and the small difference between -DABE and D[La]blood during exercise in our study.

Conclusions The results show that pH defense on a per liter basis is not weakened in highlanders in spite of the decreased bicarbonate concentration. The modest rise of nbi buffer capacity at rest compared to lowlanders results from the

123

Bangsbo J, Juel C, Hellsten Y, Saltin B (1997) Dissociation between lactate and proton exchange in muscle during intense exercise in man. J Physiol 504:489–499 Bo¨ning D (1974) The ‘‘in vivo’’ and ‘‘in vitro’’ CO2-equilibration curves of blood during acute hypercapnia and hypocapnia. II. Theoretical considerations. Pflu¨gers Arch 350:213–222 Bo¨ning D, Trost F, Braumann K-M, Bitter H, Bender K, v. z. Mu¨hlen A, Schweigart U (1980) Altitude acclimatization in skiing lowlanders. Int J Sports Med 1:191–198 Bo¨ning D, Maassen N, Steinacker JM, Seifert C, Thomas A, Schmidt W (1999) Carbon dioxide storage and nonbicarbonate buffering in the human body before and after an Himalayan expedition. Eur J Appl Physiol 79:457–466 Bo¨ning D, Maassen N, Thomas A, Steinacker JM (2001a) Extracellular pH defense against lactic acid in normoxia and hypoxia before and after a Himalayan expedition. Eur J Appl Physiol 84:78–86 Bo¨ning D, Rojas J, Serrato M, Ulloa C, Coy L, Mora M, Gomez J, Hu¨tler M (2001b) Hemoglobin mass and peak oxygen uptake in untrained and trained residents of moderate altitude. Int J Sports Med 22:572–578 Bo¨ning D, Klarholz C, Himmelsbach B, Hu¨tler M, Maassen N (2007a) Causes of differences in exercise-induced changes of base excess and blood lactate. Eur J Appl Physiol 99:163–171 Bo¨ning D, Klarholz C, Himmelsbach B, Hu¨tler M, Maassen N (2007b) Extracellular bicarbonate and non-bicarbonate buffering against lactic acid during and after exercise. Eur J Appl Physiol 100:457–467 Cerretelli P, Samaja M (2003) Acid–base balance at exercise in normoxia and in chronic hypoxia. Revisiting the ‘‘lactate paradox’’. Eur J Appl Physiol 90:431–448 Christiansen TF (1981) An algorithm for calculating the concentration of the base excess of blood. In: Siggaard-Andersen O (ed) Proceedings of the IFCC Expert Panel on pH and blood gases held at Herved Hospital 1980. Radiometer A/S, Copenhagen, pp 77–81

Eur J Appl Physiol (2008) 103:127–137 Favier R et al (1995) Training in hypoxia vs. training in normoxia in high-altitude natives J Appl Physiol 78:2286–2293 Gregersen MI, Rawson RA (1959) Blood volume. Physiol Rev 39:307–342 Hamadeh MJ, Robitaille L, Boismenu D, Hongsprabhas P, Mamer OA, Hoffer LJ (1999) Human extracellular water volume can be measured using the stable isotope Na342SO4. J Nutr 129:722–727 Jain SC, Bardhan J, Swamy YV, Krishna B, Nayar HS (1980) Body fluid compartments in humans during acute high-altitude exposure. Aviat Space Environ Med 51:234–236 Juel C, Lundby C, Sander M, Calbet JA, Hall G (2003) Human skeletal muscle and erythrocyte proteins involved in acid–base homeostasis: adaptations to chronic hypoxia. J Physiol 548:639– 648 Juel C, Klarskov C, Nielsen JJ, Krustrup P, Mohr M, Bangsbo J (2004) Effect of high-intensity intermittent training on lactate and H+ release from human skeletal muscle. Am J Physiol Endocrinol Metab 286:E245–E251 Lang W, Zander R (2005) Prediction of dilutional acidosis based on the revised classical dilution concept for bicarbonate. J Appl Physiol 98:62–71 Lundvall J, Mellander S, Westling H, White T (1972) Fluid transfer between blood and tissues during exercise. Acta Physiol Scand 85:258–269

137 Malapartida M, Moncloa F (1967) Radiosulfate space in humans at high altitude. Proc Soc Exp Biol Med 125:1328–1330 Milledge JS (1992) Salt and water control at altitude. Int J Sports Med 13:S61–S63 Ramirez G, Pineda D, Bittle PA, Rabb H, Rosen R, Vesely D, Sasaki S (1998) Partial renal resistance to arginine vasopressin as an adaptation to high altitude living. Aviat Space Environ Med 69:58–65 Robinson Y, Cristancho E, Bo¨ning D (2007) Erythrocyte aspartate aminotransferase as an indirect marker for stimulated erythropoiesis in athletes. Lab Hematol 13:49–55 Rowlands DS (2005) Model for the behaviour of compartmental CO2 stores during incremental exercise. Eur J Appl Physiol 93:555– 568 Schmidt W, Rojas J, Bo¨ning D, Bernal H, Garcia S, Garcia O (1999) Plasma electrolytes in natives to hypoxia after marathon races at different altitudes. Med Sci Sports Exerc 31:1406–1413 Siggaard-Andersen O (1974) The acid–base status of the blood. Munksgaard, Copenhagen Wagner PD et al (2001) Pulmonary gas exchange and acid–base state at 5,260 m in high-altitude Bolivians and acclimatized lowlanders. J Appl Physiol 92:1393–1400

123