Relationship of potassium ions and blood lactate to ...

691

Relationship of potassium ions and blood lactate to ventilation during exercise Robert G. McMurray and Matthew S. Tenan

Abstract: Ventilatory control during exercise is a complex network of neural and humoral signals. One humoral input that has received little recent attention in the exercise literature is potassium ions [K+]. The purpose of this study was to examine the relationship between [K+] and ventilation during an incremental cycle test and to determine if the relationship between [K+] and ventilation differs when blood lactate [lac–] is manipulated. Eight experienced triathletes (4 of each sex) completed 2 incremental, progressive (5-min stages) cycle tests to volitional fatigue: 1 with normal glycogen stores and 1 with reduced glycogen. Minute ventilation was measured during the final minute of each stage, and blood [lac–] and [K+] were measured at the end of each exercise stage. Minute ventilation and [K+] increased with exercise intensity and were similar between trials (p > 0.5), despite lower [lac–] during the reduced-glycogen trial. The concordance correlations (Rc) between [lac–] and minute ventilation were stronger for both trials (Rc = ~0.88–0.96), but the slopes of the relationships were different than the relationships between [K+] and minute ventilation (Rc = ~0.76–0.89). The slope of the relationship between [lac–] and minute ventilation was not as steep during the reduced-glycogen trial, compared with the normal trial (p = 0.002). Conversely, the slope of the relationships between [K+] and minute ventilation did not change between trials (p = 0.454). The consistent relationship of minute ventilation and blood [K+] during exercise suggests a role for this ion in the control of ventilation during exercise. Conversely, the inconsistent relationship between blood lactate and ventilation brings into question the importance of the relationship between lactate and ventilation during exercise. Key words: respiration, chemoreceptors, tidal volume, respiratory frequency, glycogen. Re´sume´ : La re´gulation de la ventilation au cours de l’exercice utilise un re´seau complexe de signaux sanguins et nerveux. L’ion potassium [K+] est un signal sanguin affe´rent qui a fait l’objet de peu d’e´tudes en physiologie de l’activite´ physique. Cette e´tude se propose d’analyser la relation entre la [K+] et le de´bit ventilatoire au cours d’un test d’effort sur ve´lo et de ve´rifier si la relation entre la [K+] et le de´bit ventilatoire varie quand on manipule la concentration sanguine de lactate [lac–]. Huit triathloniens d’expe´rience (4 femmes et 4 hommes) participent a` 2 e´preuves d’effort progressif sur ve´lo par palier de 5 min jusqu’a` l’e´puisement volontaire en pre´sence de re´serves normales de glycoge`ne et de stocks re´duits. On mesure le de´bit ventilatoire durant la dernie`re minute de chaque palier et les concentrations sanguines de lactate et de potassium a` la fin de chaque palier d’exercice. Le de´bit ventilatoire et la [K+] augmentent avec l’intensite´ de l’exercice; les augmentations sont semblables d’une e´preuve a` l’autre (p > 0,5) meˆme en pre´sence de moins de lactate [lac–] dans la condition avec un moindre stock de glycoge`ne. Les corre´lations de concordance entre la [lac–] et le de´bit ventilatoire sont plus fortes dans les deux e´preuves d’effort (Rc = ~0,88–0,96), mais la relation diffe`re de celle entre la [K+] et le de´bit ventilatoire (Rc = ~0,76–0,89). La pente de la relation entre la [lac–] et le de´bit ventilatoire n’est pas autant accentue´e au cours de l’e´preuve en pre´sence de stocks re´duits de glycoge`ne, et ce, comparativement a` la condition normale (p = 0,002). En contrepartie, la pente de la relation entre la [K+] et le de´bit ventilatoire ne varie pas d’une e´preuve a` l’autre (p = 0,454). La relation maintenue entre le de´bit ventilatoire et la concentration sanguine de potassium au cours de l’exercice sugge`re un roˆle potentiel de cet ion dans la re´gulation de la ventilation au cours de l’exercice physique. Inversement, la relation variable entre la concentration sanguine de lactate et le de´bit ventilatoire remet en question l’importance de la relation entre la concentration de lactate et le de´bit ventilatoire au cours de l’exercice physique. Mots-cle´s : respiration, chimiore´cepteurs, volume courant, fre´quence respiratoire, glycoge`ne. [Traduit par la Re´daction]

Introduction The control of ventilation during exercise consists of complex signals from both neural (Flandrois et al. 1967; Eldridge et al. 1981; Turner et al. 1997) and humoral (Casa-

buri et al. 1977; Tibes et al. 1977; Whipp 1983; Schneider and Berwick 1998) sources. The importance of each has been debated for decades. The neural signals appear to be a network of feedforward and feedback mechanisms (Flandrois et al. 1967; Eldridge et al. 1981; Turner et al. 1997),

Received 30 April 2010. Accepted 19 July 2010. Published on the NRC Research Press Web site at apnm.nrc.ca on 6 October 2010. R.G. McMurray.1 Applied Physiology Laboratory, University of North Carolina at Chapel Hill, CB#8700, Fetzer Gym, Chapel Hill, NC 27713, USA. M.S. Tenan. Neuromuscular Physiology Laboratory, University of Texas at Austin, Austin, TX 19128, USA. 1Corresponding

author (e-mail: [email protected]).

Appl. Physiol. Nutr. Metab. 35: 691–698 (2010)

doi:10.1139/H10-063

Published by NRC Research Press

692

while the focus of the humoral component has been the chemoreceptor-dependent feedback system, which is known to respond to oxygen, carbon dioxide, hydrogen ions, and potassium ions (Linton and Band 1985). Studies have argued that CO2 is an important stimulus for ventilation during exercise (Casaburi et al. 1977; Schneider and Berwick 1998), but other studies suggest that may not be the case (Flandrois et al. 1967; Eldridge et al. 1981; Turner et al. 1997). A number of researchers have shown a relationship between ventilation and lactic acid [lac–] accumulation in the blood, particularly during high-intensity exercise (Wasserman et al. 1967, 1973; Tibes et al. 1976; Whipp 1983). The association is supported by research that has shown a very strong relationship between hydrogen ions and lactate (r = ~90–0.94 (Kato et al. 2005). In contrast, other studies have suggested a disassociation of blood [lac–] and ventilation (Busse et al. 1989, 1991; Paterson et al. 1990; Sabapathy et al. 2006), because any relationship between hydrogen ions and lactate is limited by buffering and erythrocytes (McKelvie et al. 1991; Bangsbo et al. 1992; Zavorsky et al. 2007). Several researchers have shown a relationship between arterial or venous potassium ion concentrations [K+] and minute ventilation (V_ E) during exercise (Tibes et al. 1977; Busse and Maassen 1989; Paterson et al. 1989b; Zavorsky et al. 2007), with correlations between [K+] and V_ E being as high as 0.85 to 0.95 (Tibes et al. 1976; Busse et al. 1989; Zavorsky et al. 2007). Furthermore, research by Linton and Band (1985) has shown that the carotid chemoreceptors are sensitive to [K+]; thus, potassium may serve as an important drive for ventilation. Once again, not all studies agree (Paterson et al. 1989a; McLoughlin et al. 1994). Most studies have used incremental or non-steady-state exercise tests, in which the differing time constants for metabolic rate and blood metabolites cause all variables to be in a constant state of flux (Tibes et al. 1976, 1977; Busse and Maassen 1989; McLoughlin et al. 1994; Paterson et al. 1989b). In support, some studies (Tibes et al. 1977; Linton and Band 1985; Bangsbo et al. 1992; Zavorsky et al. 2007) have noted that [K+] in the blood appears to plateau in about 1 to 3 min, whereas the time course for blood [lac–] to plateau appears to be slower. Thus, lactate and potassium ions may not be in equilibrium with the metabolic rate if the stages of incremental exercise are 1 min in duration. In addition, changes in hydrogen ion accumulation, which directly stimulate the chemoreceptors, may not mirror either the [K+] or [lac–] changes (Tibes et al. 1977; Bangsbo et al. 1992; Zavorsky et al. 2007). These studies have not adjusted for differences _ 2 max) at in proportion of maximal aerobic capacity (%VO each absolute workload, making interpretation of the results challenging. For example, 2 subjects can be working at the _ 2 max, same absolute workload, but at differing relative %VO resulting in different lactate accumulations and potentially different effects on ventilation. Zavorsky et al. (2007) exercised women for three 5-min sessions at 80% of peak power and found that, during exercise, the relationship between ventilation and [K+] was stronger than for ventilation and [lac–] (R2 = 0.91 vs. 0.45, respectively). The results of Zavorsky et al. (2007) lend credence to a relationship between [K+] and ventilation, but the results were only obtained at 1 workload and, thus, only represent 1 level of exertion, not the spectrum of exercise intensities. Furthermore, their re-

Appl. Physiol. Nutr. Metab. Vol. 35, 2010

gression analysis appears to have been completed using mean data at each time point, rather than all the subjects’ data, which, if true, reduces the variability and improves the relationship. Another approach has been to base the exercise intensity on ventilatory or lactate threshold (Busse et al. 1989; McLoughlin et al. 1994). Although this approach examines the simultaneous relationship of V_ E with both [K+] and [lac–], such an approach provides only a narrow view of the relationship between [K+] and ventilation. Finally, previous studies have used simple correlations to relate ventilation to [K+] or [lac–], using multiple data points from the same subject. This approach would inflate the relationships because of the correlated nature of multiple data points for the same subject. Therefore, the purpose of this study was to examine the relationship of [K+] and ventilation during an incremental exercise test in which the participants reached steady state (or as close as possible) during each stage. In addition, the subjects completed similar testing twice, once with normal glycogen stores and once with lowered glycogen stores, to examine the simultaneous relationships of [lac–] and [K+] to ventilation. Finally, the _ 2 max relationships were examined with respect to %VO (rather than absolute workload) to account for individual differences in aerobic power.

Materials and methods Participants Eight adults (4 males and 4 females), with a mean age of 28.0 ± 6.4 years (range, 21–38 years), participated in this study. Eight participants provided sufficient power (a = 0.05 and b = 0.80) to detect differences in potassium of 0.02 mmol
L–1 and in lactate of 1.3 mmol
L–1. The participants were experienced triathletes or cyclists with a history of training, and had been training for their respective sports for at least the previous 90 days. Two of the women were using oral contraceptives and 2 had normal menstrual cycles. Although potential sex-related differences in several of the maximal exercise responses existed, both men and women were included to improve the generalizability of the responses. In support, Zavorsky et al. (2007) suggested that relationships between ventilation and potassium or lactate are not influenced by sex. In addition, progesterone has been shown to change respiratory responsiveness, but no studies have been able to correlate progesterone levels with ventilation (Harms et al. 2008). Statistical analyses were adjusted for sex to account for potential sex-related differences. All participants read and signed the informed consent form to act as a human subject previously approved by the University of North Carolina at Chapel Hill’s (Chapel Hill, N.C.) Institutional Review Board. They filled out the department medical history and training history forms, and were medically screened before participation. The participants had a mean body mass of 68.4 ± 8.1 kg (range, 57.3– 80.2 kg), a mean height of 173 ± 10 cm (range, 159– 186 cm), and body mass indexes within the normal range (22.7–23.2 kg
m–2). Their mean predicted aerobic power was 62 ± 7 mL
kg–1
min–1 (range, 47–72 mL
kg–1
min–1). Instrumentation All exercise was completed on a Monark 828E cycle erPublished by NRC Research Press

McMurray and Tenan

gometer (Varberg, Sweden), which was modified to accommodate the subject’s own pedals, shoes, and saddle (seat). _ 2) were obtained using Measurements of oxygen uptake (VO the ParvoMedics TrueMax 2400 Metabolic System (ParvoMedics, Ogden, Utah). The system was calibrated before each subject’s cycling trial. Procedures All subjects attended the laboratory for a total of 4 sessions. The first session was a screening trial intended to collect demographic data, fill out paperwork, familiarize the subject with the cycle ergometer, and perform a 3-stage submaximal cycle ergometer exercise test to estimate maximal _ 2 max) (Heyward 2006). The inherent inoxygen uptake (VO _ 2 max from such tests was acceptaccuracy of predicting VO able in this case, because the submaximal prediction test was only used to estimate the workload for the 2-h reducedglycogen trial (discussed below). There were 3 other trials: a 2-h exercise session performed during the evening, which was designed to reduce the subject’s glycogen stores; a progressive cycling trial completed the morning after (~8 h) the reduced-glycogen ride (the reduced-glycogen trial); and a progressive cycling trial completed with normal glycogen stores. Half the subjects (2 men and 2 women) completed the normal-glycogen trial first, while the other half completed the reduced-glycogen trial first. Pretrial glycogen reduction To reduce muscle glycogen, subjects exercised for 120 min at an intensity of 65%–70% of their predicted _ 2 max (Gollnick et al. 1974; Glass et al. 1997; Sabapathy VO et al. 2006). The participant utilized the same cycle ergometer set-up on which they performed the submaximal exer_ 2 and heart rate were cise test in the screening trial. VO monitored at 20-min intervals throughout the trial to ensure that the proper intensity and workload was continually ad_ 2 max. justed to maintain 65%–70% of VO Experimental trials Both experimental trials followed an identical protocol. For the 2 women with normal menstrual cycles, both trials occurred in the same phase. Prior to arriving in the laboratory, participants were asked to refrain from vigorous exercise for at least 24 h for the normal-glycogen condition and for about 8 h for the reduced-glycogen condition. The participants were allowed to drink as much water as desired, but were asked to avoid any caffeinated or alcoholic beverages for at least 8 h prior to exercise. Participants performing the reduced-glycogen condition were asked to verify that they had not eaten anything since the reduction protocol. The participant then had a 20-gauge catheter inserted into his or her antecubital vein. The subject then mounted the cycle ergometer and rested for 5 min. Resting measurements _ 2 and heart rate were obtained during the fifth minute of VO of rest, and a 3-mL blood sample was drawn. The blood was centrifuged (3000 r
min–1) at 4 8C and the plasma aliquoted and frozen at –80 8C for later analysis. The participant then began an active warm-up (3–4 min) at a low self-selected intensity. After the warm-up, the exercise test commenced. The test consisted of 5-min stages of

693

cycling at progressively higher workloads (40 W per stage), _ 2 was measured throughout each stage. all at 80 r
min–1. VO Heart rate and perceived exertion were also assessed during the last minute of each exercise stage. At the end of each stage, a 3 mL blood sample was obtained. Each exercise stage was separated by 5 min of active recovery, during which the tension was released from the cycle and the participant was able to ‘‘spin’’ the pedals at a comfortable speed. The active recovery periods allowed the participant to remain comfortable on the ergometer throughout an otherwise strenuous exercise test. The participant was allowed and encouraged to consume water during the active recovery periods. The exercise test was terminated when the participant’s pedal frequency dropped below 70 r
min–1 for a 30-s period. A final blood sample was then obtained. Blood samples All blood draws from a venous catheter followed sterile and standard procedures: first, 0.5 mL of blood was drawn to clear catheter and discarded; then, 3 mL of blood was drawn; and then 1 mL of sterile saline was injected to keep the catheter patent. Hematocrits were first determined in triplicate from all samples using the microhematocrit method. Whole blood lactate [lac–] was measured in duplicate, using a calibrated Accutrend lactate analyzer (Roche, Mannheim, Germany). The remaining blood was cold centrifuged, and the plasma was extracted. All plasma samples were visually inspected and discarded if hemolyzed. Blood potassium [K+] was measured in duplicate at a later time, using the Johnson & Johnson Vitros DT60 II Chemistry System (Ortho-Clinical Diagnostics, Rochester, N.Y.). For all analyses, [K+] and [lac–] values were not corrected for changes in plasma volume. Direct measurement of pH was not possible; however, a very strong linear relationship between exercise-induced [lac–] and pH has been reported (Kato et al. 2005), although this finding is controversial (McKelvie et al. 1991; Bangsbo et al. 1992). Analytical procedures Data were normalized as a percent of peak aerobic ca_ 2 peak) because of the variation in the number of pacity (VO stages performed during the exercise tests. This was com_ 2 at 20%, 40%, 60%, and pleted by first determining VO _ 2 peak – VO _ 2 rest)  _ 80% of VO2 reserve, calculated as: [(VO _ _ percentage] + VO2 rest. Each participant’s VO2 values were then plotted against all other variables to extrapolate the V_ E, tidal volume (VT), respiratory frequency (fR), respiratory _ 2, V_ E/carbon dioxide output exchange ratio (RER), V_ E/VO +], and [lac–] for each predicted intensity. _ ), blood [K (VCO 2 Since muscle glycogen was not directly measured, the maximal [lac–] and RER results from both trials were first compared using a repeated-measures t test to determine if the reduced-glycogen protocol appeared to succeed (Glass et al. 1997). Similar t tests were used to determine if differ_ 2 peak. Two-way reences existed between all variables at VO peated-measures analyses of variance were used to examine trends between the normal- and reduced-glycogen trials and exercise intensities (rest, 20%, 40%, 60%, 80%, and _ 2 peak). The relationships between ventilatory measures VO and blood [K+] or [lac–] were initially explored using simple Published by NRC Research Press

694

correlations. These analyses allowed for comparisons of the results from this study and previous studies. Mixed-model regression analyses adjusted for sex (PROC MIXED, SAS, Cary, N.C.) were used to determine the significance of the associations for the relationships of V_ E, VT, or fR with blood [K+] or [lac–], as well as the slope and intercept values. Mixed-model regression analyses were used to account for the within-participant correlations of the intensities of exercise, and also accounted for sex. Concordance correlation coefficients (Rc) were computed to estimate the goodness of fit of the relationships (Vonesh et al. 1996). The Rc is like a correlation coefficient and has a range of –1 to 1, with values closer to unity indicating a strong fit, and values near zero indicating little or no fit. The Rc is most appropriate for within-participant mixed-model regression analyses (Vonesh et al. 1996).

Results The general responses for both exercise trials, at rest and for each of the intensities of exercise, are presented in Ta_ 2 or heart rate beble 1. There were no differences in VO tween trials at any intensity when analyses were adjusted for sex (p > 0.47 for both). The subjects exercised for a similar amount of time for both the normal-glycogen and reduced-glycogen trials (p = 0.454), and produced similar power outputs for both trials (p = 0.093). RERs were lower during the reduced-glycogen trial at all intensities (p < 0.015). Blood lactate values increased with exercise intensity, and were lower during the reduced-glycogen trial at 60%, 80%, and maximal intensities (p < 0.0003). Plasma [K+] increased with intensity, but was not different between trials (p = 0.86); sex was not a significant contributor (p = 0.23). The ventilatory responses for both exercise trials, at rest and for each intensity of exercise, are found in Table 2. V_ E increased with exercise intensity and was not significantly different between trials when adjusted for sex (p > 0.47). Similarly, VT and fR increased with intensity, as expected, but were not different between trials after adjustment for _ 2 and V_ E/VCO _ sex (p > 0.71). V_ E/VO 2 differed with exercise intensity, but were not different between trials (p > 0.22). The relationships between V_ E and [lac–] or [K+] are presented in Figs. 1 and 2, respectively. The Rc values between V_ E and [lac–] were significant (p = 0.0002) and strong for both trials (Rc = ~0.88–0.96). As seen in Fig. 1, the best fit for the relationship between [lac–] and V_ E was a linear model for both trials. However, the relationship was significantly shifted when glycogen levels were reduced, with the slope of the relationship for the reduced-glycogen trial being approximately twice as steep (184%) as the slope of the normal-glycogen trials (p = 0.002). The relationships between V_ E and [K+] were also significant (p = 0.0003), but weaker than for lactate (Rc = ~0.76–0.89). The slopes of the relationships between V_ E and [K+] for both trials were comparable and shifted minimally (p = 0.454).

Discussion The purpose of this study was to examine the relationship between ventilation and circulating levels of lactate and po-

Appl. Physiol. Nutr. Metab. Vol. 35, 2010

tassium ions. Although this topic has been the focus of previous studies, many in exercise and sport physiology support the notion that lactate is a key metabolite related to ventilation at high intensities of exercise (Bowers and Fox 1992; Bourgois and Vrijens 1998; Janssen 2001; Roels et al. 2005; Kravitz and Dalleck 2010). However, the relationship between hydrogen ions and lactate is controversial (Kato et al. 2005), and may be influenced by buffering (McKelvie et al. 1991; Bangsbo et al. 1992; Zavorsky et al. 2007). The results of this study suggest that manipulation of [lac–] has minimal impact on ventilation during exercise. Reducing [lac–] by ~50% during high-intensity exercise was not associated with any significant decrease in ventilation. Although circulating [lac–] was related to V_ E during the normal- and reduced-glycogen trials, the nature of the relationships (slopes) were different (Fig. 1). In contrast, the [K+] responses during both trials were similar, as were the relationships of [K+] to ventilation. The similar slopes of the relationships between V_ E and circulating [K+] in both glycogen conditions (Fig. 2) suggests that any relationship of [K+] and ventilation is consistent and independent of glycogen status or circulating lactate concentrations. Previous research (Tibes et al. 1976, 1977; Paterson et al. 1989b, 1990; Busse et al. 1991) has shown a relationship between ventilation and [K+], and our results agree. However, our results extend the results of previous work to include steady-state exercise at various submaximal intensities. We found that concordance correlations between V_ E and [K+] varied between the 2 conditions, and were lower than found in previous studies that used typical bivariate correlations Rc = ~0.76–0.88 vs. r = ~0.85–0.95) (Tibes et al. 1976; Busse and Maassen 1989; Zavorsky et al. 2007). The wider variation may be more appropriate, since our mixed-model analyses and Rc account for multiple observations from a single subject, rather than treating each point independently, and since our results were obtained at the end of a 5-min stage in which the metabolic demands of the exercise were better synchronized with the ventilatory responses. Furthermore, the 5-min duration of each stage allowed for the full metabolic and ionic changes to take place in the blood, which, in turn, allowed greater time for the proposed causative effects on ventilation. In support, Zavorsky et al. (2007) found that the [K+] responses to near-maximal exercise took approximately 3 min to plateau; a similar time frame was found for V_ E. Our concordance correlations also suggest that factors other than [K+] are significant contributors to the ventilatory responses to exercise (e.g., neurological factors may have a greater effect on changes in ventilation during exercise) (Flandrois et al. 1967; Eldridge et al. 1981; Turner et al. 1997). The importance of other factors makes sense when one considers that direct stimulation of the chemoreceptors with [K+], with levels similar to those found in the our study, resulted in only a doubling of the V_ E (Linton and Band 1985), yet the V_ E responses to exercise encountered in the preset study were much greater, as much as 13 times greater than rest at peak response. The reduced-glycogen trial resulted in lower RER responses at peak intensities and lower maximal [lac–] responses than the normal-glycogen trial; therefore, the previous day’s 2-h cycle trial appeared to achieve the goal of reducing muscle glycogen. We found strong concordance Published by NRC Research Press

McMurray and Tenan

695

Table 1. Mean ± SE metabolic responses at rest and for each exercise intensity for the reduced-glycogen and normal-glycogen trials. Metabolic responses for the 2 trials

Rest

20%

40%

60%

80%

Maximal

— —

222±76 185±57

599±189 543±216

956±324 895±325

1300±324 1203±397

1710±372 1680±362

0.30±0.03 0.29±0.02

0.96±0.07 0.94±0.08

1.35±0.08 1.33±0.13

2.28±0.21 2.24±0.22

2.93±0.28 2.90±0.29

3.60±0.35 3.54±0.37

0.23±0.03 0.20±0.02

0.72±0.05 0.65±0.06{

1.07±0.09 0.95±0.09{

1.93±0.19 1.71±0.18{

2.69±0.24 2.40±0.25{

3.64±0.38 3.19±0.40{

83±3 92±5

106±3 114±4

134±4 138±6

155±6 159±6

180±4 173±4

0.76±0.01 0.69±0.01{

0.80±0.01 0.71±0.01{

0.84±0.02 0.76±0.01{

0.92±0.02 0.82±0.01{

1.01±0.02 0.89±0.03{

Potassium (mmol
L–1)* Normal 4.2±0.1 Reduced 4.1±0.1

4.5±0.1 4.3±0.1

4.7±0.1 4.6±0.1

4.9±0.1 4.7±0.1

5.2±0.1 5.0±0.1

5.4±0.1 5.1±0.2

Lactate (mmol
L–1)* Normal Reduced

1.5±0.2 1.7±0.1

1.7±0.2 1.5±0.1

2.8±0.2 1.6±0.1{

4.6±0.3 2.7±0.2{

10.6±1.0 5.6±1.1{

Power (W) Normal Reduced V˙O2 (L
min–1)* Normal Reduced V˙CO2 (L
min–1)* Normal Reduced

Heart rate (beats
min–1)* Normal 65±2 Reduced 70±5 ˙ ˙ RER (VO2/VCO2)* Normal 0.77±0.03 Reduced 0.69±0.01

1.2±0.1 1.5±0.2

Note: RER, respiratory exchange ratio; V_ CO2, carbon dioxide output; V_ O2, oxygen uptake. *p < 0.0001 for overall exercise intensity. { p < 0.01 between trials at the specific intensity.

Table 2. Mean ± SE ventilation responses at rest and for each exercise intensity for the reduced-glycogen and normal-glycogen trials. Ventilation responses for the 2 trials ˙ E (L
min–1)* V Normal Reduced

Rest

20%

40%

60%

80%

Maximal

7.91±0.74 7.08±0.73

17.30±1.39 17.28±1.11

27.49±3.01 27.09±2.22

40.84±3.89 39.13±3.16

61.17±6.92 57.77±5.69

104.63±13.08 94.58±10.13

VT (L)* Normal Reduced

0.70±0.08 0.63±0.06

1.06±0.10 1.23±0.10

1.65±0.17 1.57±0.16

2.16±0.27 2.04±0.24

2.53±0.24 2.34±0.28

2.61±0.39 2.34±0.35

17.7±2.4 17.8±1.4

20.1±2.1 20.0±1.5

24.6±2.5 25.8±62.3

42.8±6.0 42.4±3.3

20.5±1.6 20.6±0.8

18.1±1.1 17.7±0.4

21.1±1.8 20.1±1.1

28.8±2.3 26.7±1.1

25.7±1.5 29.1±1.1

21.3±1.1 23.2±0.7

22.8±1.6 24.5±1.3

28.5±2.1 30.1±1.4

Respiratory frequency (breaths
min–1)* Normal 11.8±1.2 17.3±2.4 Reduced 11.4±1.0 14.5±1.3 ˙ E/V ˙ O2* V Normal 27.1±1.5 18.4±1.4 Reduced 24.6±1.5 18.9±1.5 ˙ E/V ˙ CO2* V Normal 35.5±2.3 24.2±1.4 Reduced 35.5±2.2 27.3±2.2

Note: V_ E, minute ventilation; VT, tidal volume; V_ O2, oxygen uptake; V_ CO2, carbon dioxide output. *p < 0.0001 for overall exercise intensity.

correlations between V_ E and [lac–] in both glycogen conditions; however, the nature of the relationships (slopes of their equations) differed between trials (Fig. 1). For example, at maximal exercise, V_ E was similar between the

reduced- and normal-glycogen conditions, yet the [lac–] was half as great during the reduced-glycogen trial. A closer inspection of our data indicates that all subjects experienced a reduction in [lac–] during the reduced-glycogen trial, yet Published by NRC Research Press

696 Fig. 1. Scatterplot of the relationship between ventilation and circulating lactate with normal- and reduced-glycogen trials. The dashed line is an estimate of the slope for reduced-glycogen trials and the solid line represents normal-glycogen trials. The regression equations for normal-glycogen trials are V_ E = 8.93[lac–] + 9.91; Rc = 0.96; SE = 9.94; and for reduced-glycogen trials are V_ E = 16.99[lac–] – 0.77; Rc = 0.884; SE = 14.4. Rc, concordance correlation coefficient; V_ E, minute ventilation.

Fig. 2. Scatterplot of the relationship between ventilation and circulating [K+] with normal- and reduced-glycogen trials. The dashed line is an estimate of the slope for reduced-glycogen trials and the solid line represents normal-glycogen trials. The regression equations for normal-glycogen trials are V_ E = 44.0[K+] – 169.2; Rc = 0.89; SE = 19.2; and for reduced-glycogen trials are V_ E = 45.3[K+] – 169.4; Rc = 0.77; SE = 16.0. Rc, concordance correlation coefficient; V_ E, minute ventilation.

only 5 of the 8 subjects had lower ventilations. Furthermore, the correlation between changes in V_ E and the change in [lac–] between the 2 trials was low (r = 0.215) and not significant. These results suggest that although an association between [lac–] and V_ E exists, causality cannot be implied. Our results, using more advanced statistical techniques accounting for the repeated-measures nature of the data, confirm the findings of researchers who used simple statistical

Appl. Physiol. Nutr. Metab. Vol. 35, 2010

techniques (Tibes et al. 1976,1977; Busse and Maassen 1989; Paterson et al. 1990; Zavorsky et al. 2007). The control of ventilation during exercise is complex, consisting of both neural and humoral signals. By requiring all subjects to maintain a constant 80 r
min–1 pedal rate and stopping exercise when rate declined by 10 r
min–1, we were able to minimize differences in the neural input due to pedaling rate (Flandrois et al. 1967; Eldridge et al. 1981; Turner et al. 1997). The humoral responses appear to be mediated by the chemoreceptors, which respond to blood O2, CO2, [H+], and [K+]. Arterial CO2 tends to decline as exercise intensity increases (Wasserman et al. 1967), suggesting it is not a key factor, but any influence of blood O2 cannot be eliminated in our study. Our results suggest that [K+] is more consistently related to ventilation than blood [lac–]. The reason for the closer relationship between V_ E and [K+] than for V_ E and [lac–] could be related to the fact that lactate does not directly stimulate ventilation, but has its effects only when the H+ ions are disassociated. Further, some of those disassociated H+ ions will be buffered or translocated to the erythrocyte, reducing the effects of lactate on ventilation (Busse et al. 1989; McKelvie et al. 1991; Zavorsky et al. 2007). However, Kato et al. (2005) showed that plasma lactate and pH are highly related during exercise. In support of our findings, Paterson et al. (1990) exercised a small group of McArdle syndrome patients, who do not produce lactate, and found that they displayed the typical ventilatory response to exercise. Since the [K+] in the range we obtained has been shown to increase ventilation via chemoreceptors (Linton and Band 1985), and since the time course of the [K+] response to exercise-induced V_ E is more consistent than for [lac–] or [H+] (Paterson et al. 1989a), we suggest that there is a much closer link between the chemoreceptors and [K+] than for [lac–], which is in agreement with previous studies (Tibes et al. 1976; Busse et al. 1989; Paterson et al. 1989b, 1990; Zavorsky et al. 2007). A relationship between ventilation and lactate thresholds has received considerable attention for quite some time. Our regression analyses suggest that the relationship is not strong. Furthermore, Fig. 3 shows that the ventilation responses were the same for both trials, yet the levels of lactate were different. Ventilation appears to increase similarly for both trials, and an exponential increase occurs for both trials between 60% to 80% of peak capacity. Concomitantly, the lactate levels are about half as great during the reducedglycogen trials at the same percentages. In addition, the pro_ 2 peak at which the 4 mmol
L–1 level is reached portion of VO during the reduced-glycogen trial is about 17% higher than _ 2 peak vs. in the normal-glycogen trial (90% ± 11% VO _ 73% ± 8% VO2 peak), yet the ventilation curves are clearly different at those intensities. Therefore, our data strongly suggest that the ventilatory and lactate thresholds are independent occurrences. This study had some limitations that are worth noting. The first is a small sample size that included both sexes. In defense, the sample size was powered on changes in potassium and lactate and, as noted in the Methods section, we had sufficient power to determine relatively small changes in each. A larger sample size would have been necessary to detect differences in the change in ventilation (n = 14). Also, the small sample size is similar to the majority of prePublished by NRC Research Press

McMurray and Tenan Fig. 3. Illustration of the concurrent ventilation and lactate responses as exercise intensity increases in the normal and reduced glycogen states.

697

have provided more information on chemoreceptor reactivity. Also, Kato et al. (2005) used arterialized venous blood and found a very strong relationship (r = ~0.90–0.94) between hydrogen ions and lactate, suggesting that measurements of lactate may be sufficient to make inferences about hydrogen ion status during exercise. In addition, some studies have suggested that the trends for hydrogen ions and lactate are similar at rest and during exercise (Laurell and Pernow 1966; Aurell et al. 1967). Fourth, we were able to manipulate the [lac–] responses to the exercise but not the [K+]. Manipulation of the [K+] may have provided more information on its role in ventilatory control. Finally, the RER value at the end of the trials was less than expected for a maximal effort. This is probably related to the duration of the trials; some took as long as 45 min and the reason the subjects stopped was probably not related to lactate but more related to the duration of the exercise.

Conclusions The consistent relationship of blood potassium concentrations with ventilation during exercise supports a role for this ion in the control of ventilation during exercise. Conversely, the inconsistent relationship between blood lactate and ventilation suggests that the importance of lactate in ventilatory control needs to be re-examined. These results also suggest that the ventilatory threshold may not be entirely related to lactate, as some researchers have suggested (Beaver et al. 1986; Caiozzo et al. 1982), and as is still presented in reference textbooks and used today (Bourgois and Vrijens 1998; Roels et al. 2005; McArdle et al. 2007; Powers and Howley 2009).

Acknowledgements The authors thank Peter Hosick and Kyle Leppert for their assistance with the exercise trials. The authors do not have any professional relationship with companies or manufacturers who would benefit from the results of this study. Funding for the project was obtained from the Smith Fund at the University of North Carolina.

References vious studies on the same topic (Tibes et al. 1977; McLoughlin et al. 1994; Busse et al. 1989, 1991; Paterson et al. 1989b; McKelvie et al. 1991; Bangsbo et al. 1992). The inclusion of both sexes may have added a confounding factor, but Zavorsky et al. (2007) have stated that women have similar electrolyte responses to men during highintensity exercise. Nevertheless, we accounted for sex in our statistical analyses. Another limitation was the 5 min of active rest between stages, which could have influenced some of the exercise responses, such as adenosine triphosphate–creatine phosphate replenishment or the rate of appearance or disappearance of lactate. However, we felt that the 5-min exercise stage would be sufficient to counter the rest period and cause plateaus of the metabolic, ventilatory, and ionic responses, even during near-maximal exercise intensities. This is similar to a report by Zavorsky et al. (2007). A third factor was the inability to obtain arterialized blood samples and measure hydrogen ions, which could

Aurell, M., Carlsson, M., Grimby, G., and Hood, B. 1967. Plasma concentration and urinary excretion of certain electrolytes during supine work. J. Appl. Physiol. 22(4): 633–638. PMID:6023174. Bangsbo, J., Graham, T., Johansen, L., Strange, S., Christensen, C., and Saltin, B. 1992. Elevated muscle acidity and energy production during exhaustive exercise in humans. Am. J. Physiol. 263(4 Pt. 2): R891–R899. PMID:1415803. Beaver, W.L., Wasserman, K., and Whipp, B.J. 1986. A new method for detecting anaerobic threshold by gas exchange. J. Appl. Physiol. 60(6): 2020–2027. PMID:3087938. Bourgois, J., and Vrijens, J. 1998. The Conconi test: a controversial concept for the determination of the anaerobic threshold in young rowers. Int. J. Sports Med. 19(8): 553–559. doi:10.1055/ s-2007-971959. PMID:9877147. Bowers, R.W., and Fox, E.L. 1992. Sport physiology. 3rd ed. Wm C Brown Publishers, Dubuque, Iowa. pp. 211–214. Busse, M.W., and Maassen, N. 1989. Effect of consecutive exercise bouts on plasma potassium concentration during exercise and recovery. Med. Sci. Sports Exerc. 21(5): 489–493. PMID: 2691811. Published by NRC Research Press

698 Busse, M.W., Maassen, N., Konrad, H., and Boning, D. 1989. Interrelationship between pH, plasma potassium concentration and ventilation during intense continuous exercise in man. Eur. J. Appl. Physiol. 59(4): 256–261. doi:10.1007/BF02388325. Busse, M.W., Maassen, N., and Konrad, H. 1991. Relation between plasma K+ and ventilation during incremental exercise after glycogen depletion and repletion in man. J. Physiol. 443: 469–476. PMID:1822534. Caiozzo, V.J., Davis, J.A., Ellis, J.F., Azus, J.L., Vandagriff, R., Prietto, C.A., and McMaster, W.C. 1982. A comparison of gas exchange indices used to detect the anaerobic threshold. J. Appl. Physiol. 53(5): 1184–1189. PMID:7174412. Casaburi, R., Whipp, B.J., Wasserman, K., Beaver, W.L., and Koyal, S.N. 1977. Ventilatory and gas exchange dynamics in response to sinusoidal work. J. Appl. Physiol. 42(2): 300–301. PMID:838654. Eldridge, F.L., Millhorn, D.E., and Waldrop, T.G. 1981. Exercise hyperpnea and locomotion: parallel activation from the hypothalamus. Science, 211(4484): 844–846. doi:10.1126/science. 7466362. PMID:7466362. Flandrois, R., Lacour, J.R., Islas-Maroquin, J., and Charlot, J. 1967. Limbs mechanoreceptors inducing the reflex hyperpnea of exercise. Respir. Physiol. 2(3): 335–343. doi:10.1016/0034-5687(67) 90038-2. PMID:4226707. Glass, C., Knowlton, R.G., Sanjabi, P.B., and Sullivan, J.J. 1997. The effect of exercise induced glycogen depletion on the lactate, ventilatory and electromyographic thresholds. J. Sports Med. Phys. Fitness, 37(1): 32–40. PMID:9190123. Gollnick, P.D., Piehl, K., and Saltin, B. 1974. Selective glycogen depletion pattern in human muscle fibres after exercise of varying intensity and at varying pedalling rates. J. Physiol. 241(1): 45–57. PMID:4278539. Harms, C.A., and Rosenkranz, S. 2008. Sex differences in pulmonary function during exercise. Med. Sci. Sports Exerc. 40(4): 664–668. PMID:18317379. Heyward, V.H. 2006. Advanced fitness assessment and exercise prescription. Human Kinetics Publishers, Champaign, Ill. pp. 68–71. Janssen, P.G.J.M. 2001. Lactate threshold training. Human Kinetics, Champaign, Ill. Kato, T., Tsukanaka, A., Harada, T., Kosaka, M., and Matsui, N. 2005. Effect of hypercapnia on changes in blood pH, plasma lactate and ammonia due to exercise. Eur. J. Appl. Physiol. 95(5-6): 400–408. doi:10.1007/s00421-005-0046-z. PMID: 16193339. Kravitz, L., and Dalleck, L. 2010. Lactate threshold training. [Online .] Ava ila ble f rom ht tp://www.unm .e du/~l kra vitz / Article%20folder/lactatethreshold.html. [Accessed 29 March 2010.] Laurell, H., and Pernow, B. 1966. Effect of exercise on plasma potassium in man. Acta Physiol. Scand. 66(1): 241–242. doi:10. 1111/j.1748-1716.1966.tb03190.x. PMID:5935677. Linton, R.A.F., and Band, D.M. 1985. The effect of potassium on carotid chemoreceptor activity and ventilation in the cat. Respir. Physiol. 59(1): 65–70. doi:10.1016/0034-5687(85)90019-2. PMID:3975503. McArdle, W.D., Katch, F.I., and Katch, V.L. 2007. Exercise physiology: energy, nutrition, and human performance. Lippincott, Williams & Wilkins, Philadelphia, Penn. p. 299. McKelvie, R.S., Lindinger, M.I., Heigenhauser, G.J., and Jones, N.L. 1991. Contribution of erythrocytes to the control of the electrolyte changes of exercise. Can. J. Physiol. Pharmacol. 69(7): 984–993. PMID:1954568. McLoughlin, P., Popham, P., Linton, R.A., Bruce, R.C.H., and

Appl. Physiol. Nutr. Metab. Vol. 35, 2010 Band, D.M. 1994. Exercise-induced changes in plasma potassium and the ventilatory threshold in man. J. Physiol. 479(Pt. 1): 139–147. PMID:7990030. Paterson, D.J., Friedland, J.S., Oliver, D.O., and Robbins, P.A. 1989a. The ventilatory response to lowering potassium with dextrose and insulin in subjects with hyperkalaemia. Respir. Physiol. 76(3): 393–398. doi:10.1016/0034-5687(89)90079-0. PMID:2664933. Paterson, D.J., Robbins, P.A., and Conway, J. 1989b. Changes in arterial plasma potassium and ventilation during exercise in man. Respir. Physiol. 78(3): 323–330. doi:10.1016/00345687(89)90107-2. PMID:2616927. Paterson, D.J., Friedland, J.S., Bascom, D.A., Clement, I.D., Cunningham, D.A., Painter, R., and Robbins, P.A. 1990. Changes in arterial K+ and ventilation during exercise in normal subjects and subjects with McArdle’s syndrome. J. Physiol. 429: 339–348. PMID:2277352. Powers, S.K., and Howley, E.T. 2009. Exercise physiology: theory and application to fitness and performance. McGraw Hill Higher Education, Boston, Mass. pp. 436–437. Roels, B., Schmitt, L., Libicz, S., Bentley, D., Richalet, J.-P., and Millet, G. 2005. Specificity of VO2 max and the ventilatory threshold in free swimming and cycle ergometry: comparison between triathletes and swimmers. Br. J. Sports Med. 39(12): 965–968. doi:10.1136/bjsm.2005.020404. PMID:16306508. Sabapathy, S., Morris, N.R., and Schneider, D.A. 2006. Ventilatory and gas-exchange responses to incremental exercise performed with reduced muscle glycogen content. J. Sci. Med. Sport, 9(3): 267–273. doi:10.1016/j.jsams.2006.03.024. PMID:16682251. Schneider, D.A., and Berwick, J.P. 1998. VE and VCO2 remain tightly coupled during incremental cycling performed after a bout of high-intensity exercise. Eur. J. Appl. Physiol. 77(1–2): 72–76. doi:10.1007/s004210050302. PMID:9459524. Tibes, U., Hemmer, B., Boning, D., and Schweigart, U. 1976. Relationships of femoral venous [K+], PO2, osmolality, and [orthophosphate] with heart rate, ventilation, and leg blood flow during bicycle exercise in athletes and non-athletes. Eur. J. Appl. Physiol. Occup. Physiol. 35: 201–214. doi:10.1007/ BF02336194. PMID:8310. Tibes, U., Hemmer, B., and Boning, D. 1977. Heart rate and ventilation in relation to venous [K+], osmolality, pH, PCO2, PO2, [orthophosphate], and [lactate] at transition from rest to exercise in athletes and non-athletes. Eur. J. Appl. Physiol. Occup. Physiol. 36: 127–140. doi:10.1007/BF00423120. PMID:13993. Turner, D.L., Bach, K.B., Martin, P.A., Olsen, E.B., Brownfield, M., Foley, K.T., and Mitchell, G.S. 1997. Modulation of ventilatory control during exercise. Respir. Physiol. 110(2–3): 277– 285. doi:10.1016/S0034-5687(97)00093-5. PMID:9407621. Vonesh, E.F., Chinchilli, V.M., and Pu, K. 1996. Goodness-of-fit in generalized nonlinear mixed-effects models. Biometrics, 52(2): 572–587. doi:10.2307/2532896. PMID:10766504. Wasserman, K., Van Kessel, A.L., and Burton, G.G. 1967. Interaction of physiological mechanisms during exercise. J. Appl. Physiol. 22(1): 71–85. PMID:6017656. Wasserman, K., Whipp, B.J., Koyl, S.N., and Beaver, W.L. 1973. Anaerobic threshold and respiratory gas exchange during exercise. J. Appl. Physiol. 35(2): 236–243. PMID:4723033. Whipp, B.J. 1983. Ventilatory control during exercise in humans. Annu. Rev. Physiol. 45(1): 393–413. doi:10.1146/annurev.ph.45. 030183.002141. PMID:6405676. Zavorsky, G.S., Gow, J., and Murias, J.M. 2007. Potassium kinetics and its relationship with ventilation during repeated bouts of exercise in women. Eur. J. Appl. Physiol. 99(2): 173–181. doi:10. 1007/s00421-006-0330-6. PMID:17089156. Published by NRC Research Press

Copyright of Applied Physiology, Nutrition & Metabolism is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.