Journal of Applied Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Appl Physiol 94: 1279-1287, 2003. First published November 27, 2002; doi:10.1152/japplphysiol.00859.2002
8750-7587/03 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
94/3/1279    most recent
00859.2002v1
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fatemian, M.
Right arrow Articles by Robbins, P. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fatemian, M.
Right arrow Articles by Robbins, P. A.
Vol. 94, Issue 3, 1279-1287, March 2003

HIGHLIGHTED TOPICS
Plasticity in Respiratory Motor Control
Selected Contribution: Ventilatory response to CO2 in high-altitude natives and patients with chronic mountain sickness

Marzieh Fatemian1, Alfredo Gamboa2, Fabiola León-Velarde2, Maria Rivera-Ch2, Jose-Antonio Palacios2, and Peter A. Robbins1

1 University Laboratory of Physiology, University of Oxford, Oxford OX1 3PT, United Kingdom; and 2 Department De Ciencias Biologicas y Fisiologicas/IIA, Universidad Peruana Cayetano Heredia, Lima 100, Peru


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

The ventilatory responses to CO2 of high-altitude (HA) natives and patients with chronic mountain sickness (CMS) were studied and compared with sea-level (SL) natives living at SL. A multifrequency binary sequence (MFBS) in end-tidal PCO2 was employed to separate the fast (peripheral) and slow (central) components of the chemoreflex response. MFBS was imposed against a background of both euoxia (end-tidal PO2 of 100 Torr) and hypoxia (52.5 Torr). Both total and central chemoreflex sensitivity to CO2 in euoxia were higher in HA and CMS subjects compared with SL subjects. Peripheral chemoreflex sensitivity to CO2 in euoxia was higher in HA subjects than in SL subjects. Hypoxia induced a greater increase in total chemoreflex sensitivity to CO2 in SL subjects than in HA and CMS subjects, but peripheral chemoreflex sensitivity to CO2 in hypoxia was no greater in SL subjects than in HA and CMS subjects. Values for the slow (central) time constant were significantly greater for HA and CMS subjects than for SL subjects.

regulation of ventilation; hypercapnic ventilatory response; human; Andean natives; blunting


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

PULMONARY VENTILATION can be stimulated by CO2 through both the central and peripheral chemoreflex pathways. To assess central chemoreflex sensitivity, investigators have commonly employed a background of high O2, which minimizes, but probably does not abolish, the component of the response to CO2 that arises from the peripheral chemoreceptors. An alternative, more recent approach has been to use the difference in response speeds of the peripheral chemoreflex (fast) and central chemoreflex (slow) to separate the contributions of the central and peripheral chemoreflexes to the overall ventilatory response to CO2 (1, 6, 20).

There have been a number of studies of the ventilatory response to CO2 in both high-altitude (HA) natives and patients with chronic mountain sickness (CMS). Some studies have reported that the ventilatory sensitivities to CO2 of HA natives are similar in magnitude to those of sea-level (SL) natives at SL and lower than those of newly acclimatized SL natives at HA (3, 9), whereas others have found that they are similar to those of SL natives newly acclimatized to HA (18, 23). In general, these studies have been conducted on relatively few subjects, and the comparisons lack a rigorous statistical basis. Studies of patients with CMS have generally found that their ventilatory sensitivities to CO2 are similar to those of healthy HA natives (12, 23). However, there are the same drawbacks to these comparisons as with those for normal HA subjects. Furthermore, to the best of our knowledge, there have been no attempts to separate the peripheral and central chemoreflex components on the basis of their relatively different speeds of response.

The purpose of the present study was primarily descriptive and was to improve our understanding of the ventilatory chemoreflex response to CO2 in HA natives and patients with CMS. In particular, the present study used a dynamic technique to separate the peripheral and central components of the chemoreflex response to CO2. The dynamic variation in end-tidal PCO2 (PETCO2) followed a multifrequency binary sequence (MFBS) that had been optimized for separating the central and peripheral components of the response (20). These measurements were made against a background end-tidal PO2 (PETO2) of 100 and 52.5 Torr.


    METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Subjects. Originally, it was planned to recruit 25 SL natives, 25 healthy HA natives (who had spent their life at an altitude of >3,500 m) and 15 patients with CMS (who had an excessive degree of erythrocytosis, defined as a hematocrit of >63%). However, not all subjects completed both protocols, and so additional subjects were recruited as necessary to ensure that that the total number of repeats of each protocol was adhered to, except in the case of the patients with CMS, where only 14 patients were successfully recruited. SL natives for the experiments at SL were recruited in Lima, Peru, and subjects for the experiments at HA (healthy HA natives and patients with CMS) were recruited from among the residents of Cerro de Pasco, Peru (altitude 4,300 m; barometric pressure of laboratory 450 Torr, personal communication from C. Monge).

Protocols. Two protocols were employed. They differed from one another in relation to the background level of PO2; in one protocol, PETO2 was held at 100 Torr, and in the other protocol PETO2 was held at 52.5 Torr. Both protocols used a MFBS in PETCO2 that lasted for 1,408 s and had been optimized to separate the peripheral and central chemoreflex contributions to the overall ventilatory sensitivity to CO2 (20). Before the start of the MFBS, the subjects' PETCO2 was held at +2 Torr above their normal air-breathing value for 5 min. The lower value of PETCO2 for the MFBS was always the same as for the lead-in period of 5 min. For SL natives and patients with CMS, the upper value for the PETCO2 of the MFBS was always +10 Torr above the air-breathing value. For the first few experiments on the healthy HA natives, we employed a value of +8 Torr above the air-breathing value, but this was changed to +10 Torr to match the other subjects once it was clear that this value did not cause discomfort.

Apparatus and techniques. During the experiments, subjects sat upright in a chair and breathed through a mouthpiece with the nose of the subject occluded with a clip. Respired volumes were measured with a turbine volume-measuring device (VMM 400, Interface Associates, Laguna Niguel, CA), and respired gases were analyzed with a fast gas analyzer (Datex Ohmeda, Hatfield, UK). Data were logged to computer by using National Instruments interface cards (types DAQCard-1200 and DAQCard-AO-2DC).

PETCO2 and PETO2 were regulated breath by breath by using the technique of dynamic end-tidal forcing (22). Before the experiment started, values for inspiratory PCO2 and PO2 that were likely to produce the desired PETCO2 and PETO2 were calculated by using a model of the cardiorespiratory system. At the start of the experiment, the inspiratory gas was mixed to the correct composition by using a fast gas-mixing system constructed from commercially available mass flow controllers (type 1559A, MKS Instruments, Altringham, UK). Deviations in PETCO2 and PETO2 away from their desired values were calculated breath by breath during the experiment, and these deviations were used as feedback through an integral proportional controller to modify the inspiratory PCO2 and PO2 as required.

Data analysis. To separate the fast (peripheral) and slow (central) components of the ventilatory reponse to hypoxia, a mathematical model was fitted to the data. This model is model 2 of Pedersen et al. (20). Each component of the model has a gain term (G), a time constant (T), and pure delay (d); for the peripheral component of the model, these are denoted as Gp, Tp, and dp, respectively, and for the central component of the model, these are denoted as Gc, Tc, and dc, respectively. In addition to these terms, there was a single bias term, B, denoting the extrapolated PETCO2 for which minute ventilation (VE) = 0, and a linear trend term, C. To fit these equations to the data, we assumed that PETCO2 was constant within each breath, which enabled the differential equations to be solved to provide a set of difference equations that could be used in the parameter-estimation process. This process is described in more detail in Ref. 20.

The deterministic model was fitted in conjunction with a parallel noise model that was used to model the correlation that exists between breaths. The two parameters associated with this model are the system gain (f) and the ratio of the variances between the process and measurement noise (Rv/Rw). Further details of this process are given elsewhere (17, 20).

These two models were fitted concurrently to the data by using a standard algorithm for minimizing a sum of squared residuals (subroutine E04FDF, Numerical Algorithms Group, Oxford, UK).

For the purposes of displaying an average response, breath-by-breath data were first interpolated over 1-s intervals and then averaged across subjects. The same procedure was employed for the breath-by-breath model output for comparison.

Statistical comparisons were generally undertaken by using ANOVA. For the comparisons between PETO2 = 100 Torr and PETO2 =52.5 Torr, subjects were used as a random factor; for comparisons across groups, this was, of course, not possible. Post hoc comparisons were drawn in cases where the overall result from ANOVA was significant by using the least significant difference technique. Where correlations were drawn, these were undertaken by using the product-moment correlation coefficient. Statistical significance was accepted at P < 0.05.


    RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Subjects. The physical characteristics of the subjects are given in Table 1, and the numbers of subjects common to the protocols are given in Table 2. The HA natives were ~4 yr older than the SL controls, and the CMS subjects were ~10 yr older than the SL controls. The HA and CMS subjects were, on average, lighter and shorter than the SL controls, although this only reached significance for the HA group of subjects. As expected, the patients with CMS had higher hematocrits than the HA subjects. All HA and CMS subjects were scored for symptoms of CMS (15), where a score of >= 12 is taken as evidence of significant symptomatology. HA subjects scored an average of 6.8 ± 5.6 (mean ± SD), and CMS subjects scored an average of 15.6 ± 3.7. The difference between these was highly significant (P < 0.001). Patients with CMS had higher values for ambient, air-breathing PETCO2 and lower values for ambient, air-breathing PETO2.

                              
View this table:
[in this window]
[in a new window]
  		Table 1.   Physical characteristics of subjects


                              
View this table:
[in this window]
[in a new window]
  		Table 2.   Number of subjects common to any pair of protocols

Values for the acute ventilatory responses to hypoxia were available for almost all subjects (Table 2) from previous studies (10, 16), and the average values are listed in Table 1. Values for the ventilatory sensitivities to acute hypoxia were substantially lower for HA and CMS subjects compared with SL subjects, and, for the data for more severe hypoxia, values for CMS subjects were significantly below those for healthy HA subjects.

Responses to MFBS in PETCO2. Example data for individual subjects are shown for the MFBS in PETCO2 against a background PETO2 of 100 Torr in Fig. 1 and against a background PETO2 of 52.5 Torr in Fig. 2. PETCO2 followed the desired pattern of the MFBS reasonably in all cases. PETO2 was well controlled in all cases, with the exception of a glitch at ~8 min in the SL subject under conditions of hypoxia. Such problems are usually associated with disturbances of respiratory rhythm, e.g., coughing. VE for all subjects and protocols can be seen to be following the pattern of stimulation by the MFBS in all cases, although the variation in sensitivity between different subjects is clearly wide, as is the general breath-by-breath variability.


View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1.   Examples of data and model fits for the multifrequency binary sequence (MFBS) in end-tidal PCO2 (PETCO2) at a constant end-tidal PO2 (PETO2) of 100 Torr. Left, sea-level (SL) native; middle, high-altitude (HA) native; right, patient with chronic mountain sickness (CMS). PETCO2 (A), PETO2 (B), ventilatory response () and deterministic component of the model fit (line) together with associated residuals (C), and residuals from overall model fitting process (deterministic model plus noise model; D).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2.   Examples of data and model fits for the MFBS in PETCO2 against a background PETO2 of 52.5 Torr. Left, sea- level (SL) native; middle, high-altitude (HA) native; right, patient with chronic mountain sickness (CMS). PETCO2 (A), PETO2 (B), ventilatory response () and deterministic component of the model fit (line) together with associated residuals (C), and residuals from overall model fitting process (deterministic model plus noise model; D).

Also shown in Figs. 1 and 2 are example fits of the respiratory model to the data. The deterministic component of the model appears to describe the pattern of response within the data reasonably well, although the individual data points are quite variable. The process of fitting the deterministic and noise models simultaneously to the data provides residuals that are close to white (they lack longer term trends) compared with the residuals when just the deterministic component of the model is considered. For each fit, cross-correlations were calculated between PETCO2 and the residuals from the fitting process and averaged for each subject group and protocol. Very few of these exceeded the 95% confidence interval (CI) of ±0.02, which if they had might have indicated an inadequacy of the model. Average parameter values from the fitting process are given in Table 3.

                              
View this table:
[in this window]
[in a new window]
  		Table 3.   Parameter values for the model fits to the MFBS data under euoxic and hypoxic conditions for all subjects in the SL, HA, and CMS subject groups

Not all subjects undertook both euoxic and hypoxic MFBS protocols, and some of the earlier HA subjects had amplitudes for PETCO2 in the MFBS of 6 Torr rather than the 8 Torr used elsewhere. Figure 3 shows averaged records for the subset of subjects for whom there were MFBS sequences at an amplitude for PETCO2 of 8 Torr at both a PETO2 of 100 and 52.5 Torr. The parameter values for this subset of subjects are given in Table 4.


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 3.   Averaged MFBS in PETCO2, ventilatory response to MFBS, and model fits for a group of 15 SL residents (left), 15 HA residents (middle), and 13 patients with CMS (right). This subset of subjects has undertaken both MFBS protocols with an MFBS amplitude of 8 Torr. A: averaged PETCO2 at a PETO2 of 100 Torr (solid line) and at a PETO2 of 52.5 Torr (broken line). B: averaged ventilatory response to MFBS at a PETO2 of 100 Torr (open circle ) and a PETO2 of 52.5 Torr (). Solid lines are model fits. C: same as B, except data and model fits have all been aligned to start at a ventilation of 10 l/min and have had the fitted trend removed from the responses.


                              
View this table:
[in this window]
[in a new window]
  		Table 4.   Parameter values for the model fits to the MFBS data under euoxic and hypoxic conditions for the subset of SL, HA, and CMS subjects whose responses are plotted in Fig. 3

Figure 3A shows that the MFBS in PETCO2 was well matched between the data gathered at a PETO2 of 100 Torr and the data gathered at a PETO2 of 52.5 Torr, although for the SL level subjects, the absolute values for PETCO2 were ~1 Torr higher at a PETO2 of 52.5 Torr than at a PETO2 of 100 Torr.

Figure 3B shows the averaged VE for the subjects for both the euoxic and hypoxic protocols. There are a number of features of note. First, in marked contrast to SL subjects, there is no perceptible effect of the two different levels of PETO2 on VE at the start of the MFBS sequence for either healthy HA natives or subjects with CMS. This suggests that, in HA and CMS groups, any reduction in peripheral chemoreflex discharge at the onset of the higher level of PETO2 has been matched by an equivalent increase in central stimulation of VE through the 5-min period of euoxia that precedes the start of the MFBS record in the figure. Second, VE at the start of the MFBS record was higher for SL subjects than for healthy HA natives and patients with CMS. Third, for all groups and protocols except the SL group at a PETO2 of 52.5 Torr, there is a trend toward an increase in VE over time. For the SL group at a PETO2 of 52.5 Torr, this trend is absent with the result that VE is broadly similar between the two protocols at the lower PETCO2 at the end of the MFBS sequence. These findings are consistent with the known slow progressive stimulatory effect of hypercapnia on VE and, in the case of SL subjects at a PETO2 of 52.5 Torr, the progressive development of hypoxic ventilatory decline (see Parameter B and trend term C in DISCUSSION). Fourth, for healthy HA natives and subjects with CMS, the progressive rise in VE appears slightly greater for the MFBS protocol at a PETO2 of 100 Torr than for the protocol at a PETO2 of 52.5 Torr, such that by the end of the protocol, VE against a background PETO2 of 100 Torr slightly exceeds VE against a background PETO2 of 52.5 Torr. This is consistent with a continued, slow stimulatory effect of the higher PETO2 in HA and CMS subjects.

The mean parameter estimates for the trend term (parameter C) for the different protocols and subject groups (Tables 3 and 4) are consistent with these observations. Parameter C is negative for the SL subjects at a PETO2 of 52.5 Torr and positive for all other groups/protocols. For the HA and CMS groups, the trend term is larger at a PETO2 of 100 Torr than at a PETO2 of 52.5 Torr.

The presence of different starting values for VE and different long-term trends within the data make the visual comparison of ventilatory sensitivities to CO2 and the interactions between CO2 and hypoxia difficult. In Fig. 3C, the ventilatory data and model responses have been replotted so that all responses are aligned to start at a VE of 10 l/min, and the model trend terms have been subtracted away from both the data and model output. There are three features of note. First, the total ventilatory sensitivity to CO2 at a PETO2 of 100 Torr appears greatest for healthy HA subjects, intermediate for CMS subjects, and least for SL subjects. Second, the interaction between CO2 and hypoxia appears greatest for SL subjects, intermediate for the healthy HA natives, and least for the patients with CMS. Third, although subtraction of the model trend has removed all the apparent slow increase in VE over time for SL subjects, this is not the case for both the HA and CMS groups, where some progressive increase in VE over time remains.

Parameters for ventilatory sensitivity to CO2. At a PETO2 of 100 Torr, the total ventilatory sensitivity to CO2 in the HA native group was ~5 l · min-1 · Torr-1, which was around double that of the SL group. The sensitivity of the CMS group, at ~4.4 l · min-1 · Torr-1, was also significantly greater than for the SL group, with the difference between the HA and CMS groups not attaining significance. These differences can be attributed primarily to significant differences in Gc, although the value for Gp for the HA natives was also significantly higher than for the SL natives.

The presence of a background of hypoxia (PETO2 = 52.5 Torr) increased the total ventilatory sensitivity to CO2 compared with euoxia. This increase was greater for the SL group than for the HA and CMS groups combined (P < 0.05). Interestingly, the increments in Gp with hypoxia did not differ between groups, but there was a significant increase in Gc in the SL group that did not occur in the HA and CMS groups (P < 0.05).

Parameter B. For the HA and CMS groups, average values for parameter B were very similar for the euoxia and hypoxia protocols. For the SL group, there was a small but statistically significant difference in the values for parameter B between protocols, which possibly is explained by small differences in the control values for air-breathing PETCO2 between the two protocols. Of greater significance is that, for the SL subjects, the average difference between normal air-breathing PETCO2 and parameter B was ~5.5 Torr, whereas for HA and CMS subjects this difference was ~1.5 Torr. These differences are consistent with the higher starting level for VE in the SL group compared with the HA and CMS groups, and also the much greater effect of hypoxia at the lower value for PETCO2 in the SL group, especially when this is combined with the greater effect of hypoxia on the total ventilatory sensitivity to CO2 in the SL group of subjects.

Dynamic parameters of the peripheral chemoreflex loop. For Tp, there were no differences between groups or protocols. For dp, there were no differences between the groups, but the values for dp were lower in hypoxia than euoxia (P < 0.001, ANOVA).

Dynamic parameters of the central chemoreflex loop. A striking finding of this study is that Tc was very much slower for the HA and CMS groups compared with the SL group. This finding was present in both euoxia and hypoxia and did not differ between protocols. This finding is consistent with the observation that, after removing trend, VE was similar at the beginning and end of the MFBS for the SL subjects but remained somewhat elevated at the end for both HA and CMS subjects (Fig. 3). Values for dc did not differ significantly.

Correlations between CO2 responsiveness and measures of acute hypoxic ventilatory response and hematocrit. Correlations were undertaken for the HA and CMS groups combined. Gp under euoxic conditions exhibited a weak correlation with hematocrit (r = -0.39, P < 0.02). However, neither Gp under hypoxic conditions nor Gc in either euoxia or hypoxia correlated significantly with hematocrit.

Both the values for Gp in hypoxia and the increase in Gp from euoxia to hypoxia correlated significantly with measures of acute hypoxic ventilatory response (Gph-45 for data gathered by using a protocol lowering PETO2 to 45 Torr and Gph-34 for data gathered by using a protocol lowering PETO2 to 34 Torr) from our laboratory's previous studies (10, 16) for these subjects (for Gp in hypoxia with Gph-45, r = 0.39, P < 0.05; and with Gph-34, r = 0.56, P < 0.001; for the difference in Gp between hypoxia and euoxia with Gph-45, r = 0.47, P < 0.005; and with Gph-34, r = 0.52, P < 0.002). Neither Gp under euoxic conditions nor Gc under either euoxic or hypoxic conditions correlated significantly with either measure of acute hypoxic ventilatory response (Gph-45 or Gph-34).


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

The major findings of this study were as follows.

First, under conditions of euoxia (PETO2 = 100 Torr), total ventilatory sensitivity to CO2 in HA natives is around double that of SL natives at SL. Total ventilatory sensitivity to CO2 in patients with CMS was also greater than that of SL natives at SL and was not significantly different from healthy HA natives.

Second, under conditions of euoxia, peripheral (fast) chemoreflex sensitivity to CO2 was higher in healthy HA natives than in SL natives at SL. Fast chemoreflex sensitivity in CMS subjects was not different from SL natives at SL.

Third, hypoxia caused a greater increase in total CO2 sensitivity in SL subjects than in the HA and CMS subjects combined.

Fourth, values for parameter B, the extrapolated value for PETCO2 for which VE = 0 were higher for patients with CMS compared with healthy HA natives.

Fifth, the values for Tc were significantly higher in the HA and CMS groups combined than in the SL group.

Ventilatory response to CO2 of the SL subjects. For the SL group, the parameters of the respiratory model mostly correspond very well with those previously published using a MFBS in PETCO2 to stimulate breathing (8, 20). Under hyperoxic conditions, values of total chemoreflex sensitivity to CO2 (Gtot) have been given as 2.75 and 2.5 l · min-1 · Torr-1 (in the present study, Gtot in euoxia was 2.4 l · min-1 · Torr-1), and under hypoxic conditions (PETO2 = 50 Torr), a value for Gtot was given as 4.1 l · min-1 · Torr-1 (in the present study, Gtot in hypoxia was 4.4 l · min-1 · Torr-1). Values for parameter B from these studies have been given as 33.3 and 34.7 Torr (in the present study, parameter B = 33.5 Torr).

One area in which the correspondence between the present study and a previous study using an MFBS in PETCO2 (20) is less good relates to the changes in the individual sensitivities with hypoxia. In particular, in the study by Pedersen et al. (20), there was a nonsignificant decrease in Gc, from 2 to 1.7 l · min-1 · Torr-1, rather than the significant increase from 1.7 (95% CI, 1.40-2.05 l · min-1 · Torr-1) to 2.3 l · min-1 · Torr-1 (95% CI, 1.96-2.74 l · min-1 · Torr-1) of the present study. Studies that model the ventilatory response to step changes in PETCO2 have not found a significant increase in Gc in hypoxia, with a small, nonsignificant fall in Gc occurring in one (6) and a small, nonsignificant rise in Gc occurring in another (1). Given the existing literature, we do not feel confident that this result is one that can be generalized beyond our particular sample of SL natives without it first being repeated.

Ventilatory sensitivity to CO2 of the HA and CMS subject groups. Chiodi (3) gave data for four HA subjects showing a mean total sensitivity to CO2 of ~2 l · min-1 · Torr-1, which was close to our values for SL natives at SL. Forster et al. (9) gave data on 10 HA natives with a mean total sensitivity to CO2 of ~2.4 l · min-1 · Torr-1, which was again close to our values for SL natives. On the other hand, Severinghaus et al. (23) gave a mean value for six HA natives of 4.0 l · min-1 · Torr-1 and for six patients with CMS of 3.8 l · min-1 · Torr-1, both of which are well in excess of both his and our values for SL natives. Similarly, Milledge and Lahiri (18) gave a value of 4.2 l · min-1 · Torr-1 for four Sherpas. These differences between studies in the total ventilatory sensitivity to CO2 in euoxia/hypoxia are difficult to reconcile. However, the present study provides a considerably larger sample of values from HA natives; indeed, the total number of HA and CMS subjects in the present study exceeds the combined total number of subjects from all four of the previous studies.

The effect of hypoxia on the total ventilatory sensitivity to CO2 was greater in the SL subjects than in the HA subjects and patients with CMS. This is very much in keeping with the observations of both Severinghaus et al. (23) and Milledge and Lahiri (18). However, the values obtained for Gp in euoxia and hypoxia provide no evidence to support the notion that peripheral chemoreflex sensitivity to CO2 is reduced in HA natives compared with SL natives at SL. Indeed, the values for Gp in euoxia for the HA natives were clearly above those for SL natives. Values for Gp for the patients with CMS appeared to be lower than those for the healthy HA natives, but this did not reach significance (except for the subset of subjects in Fig. 3 under conditions of hypoxia). The one note of caution in these observations is that the increment in sensitivity in Gp in the SL natives with hypoxia may have been underestimated because of the increase in Gc with hypoxia in these subjects (see Ventilatory response to CO2 of the SL subjects). Even so, if, in SL subjects, the entire increment in Gc with hypoxia is attributed instead to an increase in the value of Gp, then the value for Gp in these subjects in hypoxia would still not be much greater than the value for HA natives.

Parameter B and trend term C. In each of the protocols, a trend term was present that reflected the known slow effects of altering a subject's PETCO2 and PETO2 away from their ambient values. In the case of lowering the PETO2 of the SL subjects, there was a trend associated with ventilatory depression by hypoxia (7, 11). In the case of raising the PETO2 of HA natives and patients with CMS, there was a trend associated with the slow rise in VE under these conditions (13, 14, 16). In the case of raising PETCO2 in all subject groups, there is a slow trend toward increased VE (25). These factors necessarily mean that values of parameter B will depend on how long these trends have been continuing; for example, the longer PETO2 has been raised in HA natives before a ventilatory sensitivity to CO2 is measured, the higher the value of parameter B will be. This implies that drawing any meaningful comparison for values of parameter B across studies where the protocols differ is difficult.

Within our study, it was clear that parameter B for patients with CMS was significantly above parameter B for healthy HA natives. This correlated tightly with the significant difference in ambient air-breathing PETCO2 between these two subject groups. A perhaps more surprising observation was that the difference between ambient air-breathing PETCO2 and parameter B was very much greater for SL subjects than for HA natives. For SL natives, this implies that the back projection of the VE-PETCO2 response line will pass above the value for normal VE and PETCO2 breathing room air. This observation is consistent with data from other studies (8, 20), and it would seem to relate to another observation made on SL subjects that there is a degree of discontinuity in the VE-PETCO2 response relation just above a subject's normal value for PETCO2 (2, 4).

Tc in HA subjects and patients with CMS. Reported values for Tc can differ substantially (5). Nevertheless, mean values for Tc for both the HA and CMS groups were slower than reported from a collection of other studies of SL natives at SL (5, 8, 20), as well as strikingly slower than in the control SL group of the present study.

One possible explanation for the increase in Tc in the HA and CMS groups is that there is some new, unidentified slow component of the ventilatory response to CO2 that is not present in the SL group. The presence of such a slow component would imply that the "two-compartment plus trend" model of the ventilatory response to CO2 is not appropriate in these subjects and that an apparent increase in Tc arises because an inappropriate model has been fitted to the data.

Assuming the two-compartment plus trend model is appropriate, the relatively slow time constant of the central component is normally attributed to the well-buffered environment within which the chemoreceptors reside. The time constant of such a system will depend on the cerebral blood flow per unit volume of brain tissue together with the ratio between blood and brain for the buffering capacity of CO2 (21). Of these factors, it is well recognized that CBF is decreased in natives of HA (19, 24). Within this, the oxygen delivery is relatively well protected because of the increased hematocrit (24). However, it is not immediately apparent whether CO2 removal is similarly protected. On one hand, the increase in hematocrit will increase the proton buffering and carbamino compound formation, but, on the other hand, the increase in hematocrit will compound the reduction in plasma flow (and hence bicarbonate space in the blood). It is possible that a quantitative analysis of these factors could reveal a relative slowing of rate of change of PCO2 within the brain in response to changes in PCO2 of the arterial blood.


    ACKNOWLEDGEMENTS

This study was supported by the Wellcome Trust.


    FOOTNOTES

Address for reprint requests and other correspondence: P. A. Robbins, Univ. Laboratory of Physiology, Parks Rd., Oxford OX1 3PT, UK (E-mail: peter.robbins{at}physiol.ox.ac.uk).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

First published November 27, 2002;10.1152/japplphysiol.00859.2002

Received 19 September 2002; accepted in final form 25 November 2002.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

1.   Bellville, JW, Whipp BJ, Kaufman RD, Swanson GD, Aqleh KA, and Wiberg DM. Central and peripheral chemoreflex loop gain in normal and carotid body-resected subjects. J Appl Physiol 46: 843-853, 1979.

2.   Bertholon, JF, Eugene M, Labeyrie E, and Teillac A. A dynamic analysis of the ventilatory response to hypoxia in man. J Physiol 408: 473-492, 1989.

3.   Chiodi, H. Respiratory adaptations to chronic high altitude hypoxia. J Appl Physiol 10: 81-87, 1957.

4.   Cummin, AR, Alison J, Jacobi MS, Iyawe VI, and Saunders KB. Ventilatory sensitivity to inhaled carbon dioxide around the control point during exercise. Clin Sci (Lond) 17: 17-22, 1986.

5.   Cunningham, DJC, Robbins PA, and Wolff CB. Integration of respiratory responses to changes in alveolar partial pressures of CO2 and O2 and in arterial pH. In: Handbook of Physiology. The Respiratory System. Control of Breathing. Bethesda, MD: Am. Physiol. Soc, 1986, vol. II, p. 475-528, sect. 3, pt. 2, chapt. 15.

6.   Dahan, A, DeGoede J, Berkenbosch A, and Olievier ICW The influence of oxygen on the ventilatory response to carbon dioxide in man. J Physiol 428: 485-499, 1990.

7.   Easton, PA, Slykerman LJ, and Anthonisen NR. Ventilatory response to sustained hypoxia in normal adults. J Appl Physiol 61: 906-911, 1986.

8.   Fatemian, M, and Robbins PA. Selected Contribution: chemoreflex responses to CO2 before and after an 8-h exposure to hypoxia in humans. J Appl Physiol 90: 1607-1614, 2001.

9.   Forster, HV, Dempsey JA, Birnbaum ML, Reddan WG, Thoden J, Grover RF, and Rankin J. Effect of chronic exposure to hypoxia on ventilatory response to CO2 and hypoxia. J Appl Physiol 31: 586-592, 1971.

10.   Gamboa, A, León-Velarde F, Rivera-Ch M, Palacios JA, Pragnell TR, O'Connor DF, and Robbins PA. Acute and sustained ventilatory responses to hypoxia in high-altitude natives living at sea level. J Appl Physiol 94: 1255-1262, 2003.

11.   Khamnei, S, and Robbins PA. Hypoxic depression of ventilation in humans: alternative models for the chemoreflexes. Respir Physiol 81: 117-134, 1990.

12.   Kryger, M, McCullough R, Doekel R, Collins D, Weil JV, and Grover RF. Excessive polycythemia of high altitude: role of ventilatory drive and lung disease. Am Rev Respir Dis 118: 659-666, 1978.

13.   Lahiri, S, and Edelman NH. Peripheral chemoreflexes in the regulation of breathing of high altitude natives. Respir Physiol 6: 375-385, 1969.

14.   Lahiri, S, Milledge JS, Chattopadhyay HP, Bhattacharyya AK, and Sinha AK. Respiration and heart rate of Sherpa highlanders during exercise. J Appl Physiol 23: 545-554, 1967.

15.   León-Velarde, F, and Arregui A. La Desadaptacion a la Vida en las Grandes Alturas. Lima, Peru: IFEA/UPCH, 1994.

16.   León-Velarde, F, Gamboa A, Rivera-Ch M, Palacios JA, and Robbins PA. Peripheral chemoreflex function in high-altitude natives and patients with chronic mountain sickness. J Appl Physiol 94: 1269-1278, 2003.

17.   Liang, PJ, Pandit JJ, and Robbins PA. Statistical properties of breath-to-breath variations in ventilation at constant end-tidal PCO2 and PO2 in humans. J Appl Physiol 81: 2274-2286, 1996.

18.   Milledge, JS, and Lahiri S. Respiratory control in lowlanders and Sherpa highlanders at altitude. Respir Physiol 2: 310-322, 1967.

19.   Milledge, JS, and Sorensen SC. Cerebral arteriovenous oxygen difference in man native to high altitude. J Appl Physiol 32: 687-689, 1972.

20.   Pedersen, MEF, Fatemian M, and Robbins PA. Identification of fast and slow ventilatory responses to carbon dioxide under hypoxic and hyperoxic conditions in humans. J Physiol 521: 273-287, 1999.

21.   Robbins, PA. The ventilatory response of the human respiratory system to sine waves of alveolar carbon dioxide and hypoxia. J Physiol 350: 461-474, 1984.

22.   Robbins, PA, Swanson GD, and Howson MG. A prediction-correction scheme for forcing alveolar gases along certain time courses. J Appl Physiol 52: 1353-1357, 1982.

23.   Severinghaus, JW, Bainton CR, and Carcelen A. Respiratory insensitivity to hypoxia in chronically hypoxic man. Respir Physiol 1: 308-334, 1966.

24.   Sorensen, SC, Lassen NA, Severinghaus JW, Coudert J, and Zamora MP. Cerebral glucose metabolism and cerebral blood flow in high-altitude residents. J Appl Physiol 37: 305-310, 1974.

25.   Tansley, JG, Pedersen MEF, Clar C, and Robbins PA. Human ventilatory response to 8 h of euoxic hypercapnia. J Appl Physiol 84: 431-434, 1998.

J APPL PHYSIOL 94(3):1279-1287
8750-7587/03 $5.00 Copyright © 2003 the American Physiological Society


This article has been cited by other articles:


Home page
 Eur Respir JHome page
L. Spicuzza, N. Casiraghi, A. Gamboa, C. Keyl, A. Schneider, A. Mori, F. Leon-Velarde, G.U. Di Maria, and L. Bernardi
Sleep-related hypoxaemia and excessive erythrocytosis in Andean high-altitude natives
Eur. Respir. J., January 1, 2004; 23(1): 41 - 46.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
94/3/1279    most recent
00859.2002v1
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fatemian, M.
Right arrow Articles by Robbins, P. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fatemian, M.
Right arrow Articles by Robbins, P. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online