hypoxia suppresses symptom perception in asthma

7
Hypoxia Suppresses Symptom Perception in Asthma Danny J. Eckert, Peter G. Catcheside, Janet H. Smith, Peter A. Frith, and R. Doug McEvoy Adelaide Institute for Sleep Health and Department of Respiratory Medicine, Repatriation General Hospital, Daw Park; School of Molecular and Biomedical Science, Discipline of Physiology, University of Adelaide, Adelaide; and Department of Medicine, Flinders University, Bedford Park, South Australia, Australia Any factor that inhibits the ability of an individual with asthma to recognize their symptoms appropriately may contribute to treat- ment delay, “near miss” events, and death during acute severe asthma. The purpose of this study was to investigate the effects of two common features of acute severe asthma—hypoxia and hyper- capnia—on respiratory sensation. Sixteen individuals with stable asthma were exposed to three gas conditions (34 minutes each): isocapnic hypoxia (arterial blood O 2 saturation of approximately 80%), hypercapnia (increase in end-tidal CO 2 of approximately 5–10 Torr), or isocapnic normoxia on 3 separate days. The perceived magnitude of externally applied resistive loads, measured during each gas condition, was reduced throughout hypoxia compared with normoxia, and there was a trend for a progressive decline during hypercapnia. Within the 15-minutes postgas inhalation pe- riod, methacholine-induced symptoms of difficult breathing, chest tightness, and breathlessness, measured using modified Borg scales, were 25–30% lower after hypoxia compared with normoxia but were not reduced after hypercapnia. We conclude that 30 min- utes of sustained hypoxia and possibly hypercapnia impair sensa- tions of respiratory load and that the effects of hypoxia persist for at least 10 minutes after returning to normoxia. Keywords: hypercapnia; bronchoconstriction; methacholine; dyspnea Delay in seeking treatment during severe asthmatic episodes has been implicated as an important factor responsible for increased morbidity and mortality. Poor perception of the severity of an asthma attack has been proposed as a mechanism underlying treatment delay (1). Since the initial observations of Rubinfeld and Pain (2), several studies have demonstrated that individuals with asthma, particularly those prone to severe or near fatal asthma (1, 3–6), demonstrate blunted sensations to increased respiratory load. Therefore, any factor that suppresses respira- tory sensation, be it an inherent persistent blunting of sensation or one that is acquired acutely, could expose patients with asthma to increased risk of morbidity or mortality (7). Mild to moderate hypoxia, which occurs during most severe acute episodes of asthma (8, 9), has been demonstrated to have central nervous system depressant effects and can lead to im- paired cognitive function (10). Central nervous system depres- sant effects of hypoxia may persist for up to an hour after removal of the stimulus (11). Recent research (12) in normal subjects demonstrated that 30 minutes of hypoxia (Sa O 2 of approximately 80%) depressed the perceived magnitude of externally applied (Received in original form May 9, 2003; accepted in final form March 11, 2004) Supported by the Foundation Daw Park/Channel 9 Telethon and National Health and Medical Research Council of Australia. Correspondence and requests for reprints should be addressed to Danny Eckert, B.Sc. (Hons), Adelaide Institute for Sleep Health, Repatriation General Hospital, Daw Park, South Australia, Australia, 5041. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents online at www.atsjournals.org Am J Respir Crit Care Med Vol 169. pp 1224–1230, 2004 Originally Published in Press as DOI: 10.1164/rccm.200305-630OC on March 12, 2004 Internet address: www.atsjournals.org resistive loads. These findings raise the possibility that individu- als with asthma who may already have blunted respiratory sensa- tions (1–6) could be further impaired in their ability to assess accurately the degree of airway obstruction during and after a period of hypoxia. This may in turn lead to treatment delays and an increased risk of asthma death. Hypercapnia is a respiratory stimulant that enhances sensa- tions of breathlessness and often accompanies life-threatening asthma (8, 9); however, there is some evidence to suggest that cognitive impairment occurs with moderate levels of hypercap- nia (13, 14), although these findings are conflicting (15). The contribution of cognitive impairment to respiratory sensations during acute sustained mild to moderate hypercapnia in asthma remains unexplored. We hypothesized that sustained moderate hypoxia would reduce respiratory sensations during bronchoconstriction in pa- tients with asthma. We also hypothesized that sustained hyper- capnia may lead to a relative depression of respiratory sensation. To test these hypotheses, the perception of dyspnea of subjects with asthma during methacholine-induced bronchoconstriction was measured immediately after 34 minutes of isocapnic hypoxia (Sa O 2 of approximately 80%) and 34 minutes of hypercapnia (change in end-tidal CO 2 of approximately 5–10 Torr) and com- pared with an equivalent period of normoxia. The perceived magnitude of externally applied resistive loads was also com- pared during the three test gas conditions. Some of the results of this study have been previously reported in abstract form (16–19). METHODS Patient Selection Nineteen stable individuals with asthma (5 males and 14 females) who satisfied the American Thoracic Society definition of asthma (20) gave informed written consent to participate in the study. Three women experienced nausea during hypoxia and were unable to complete the study. All patients were nonsmokers with a baseline FEV 1 of more than 70% predicted. None of the patients used oral corticosteroids during or 3 months before the study, and none had required emergency treatment for asthma in the previous 2 years. The study was approved by the Daw Park Repatriation General Hospital and Adelaide University Human Research and Ethics Committees. Study Design A within-subjects study design was chosen. After a preliminary visit (see online supplement), test gases (isocapnic hypoxia, normoxic hyper- capnia, and isocapnic normoxia) were administered for 34 minutes each on separate days approximately 1 week apart. Two methods were used to test for depressant effects of the test gases on respiratory sensations: First, during the gas exposures, patients rated the severity of a range of externally applied inspiratory resistive loads. Second, during room air breathing, immediately after each test gas exposure, respiratory sensations of difficult breathing, chest tightness, and breathlessness were quantified during methacholine-induced bronchoconstriction. Methacholine challenges were administered immediately after, rather than during, gas exposures to avoid potentially severe hypoxemia during bronchoconstriction coincident with hypoxic gas inhalation, combined with evidence that the central nervous system depressant effects of

Upload: r-doug

Post on 27-Jan-2017

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Hypoxia Suppresses Symptom Perception in Asthma

Hypoxia Suppresses Symptom Perception in AsthmaDanny J. Eckert, Peter G. Catcheside, Janet H. Smith, Peter A. Frith, and R. Doug McEvoy

Adelaide Institute for Sleep Health and Department of Respiratory Medicine, Repatriation General Hospital, Daw Park; School of Molecularand Biomedical Science, Discipline of Physiology, University of Adelaide, Adelaide; and Department of Medicine, Flinders University,Bedford Park, South Australia, Australia

Any factor that inhibits the ability of an individual with asthma torecognize their symptoms appropriately may contribute to treat-ment delay, “near miss” events, and death during acute severeasthma. The purpose of this study was to investigate the effects oftwo common features of acute severe asthma—hypoxia and hyper-capnia—on respiratory sensation. Sixteen individuals with stableasthma were exposed to three gas conditions (34 minutes each):isocapnic hypoxia (arterial blood O2 saturation of approximately80%), hypercapnia (increase in end-tidal CO2 of approximately 5–10Torr), or isocapnic normoxia on 3 separate days. The perceivedmagnitude of externally applied resistive loads, measured duringeach gas condition, was reduced throughout hypoxia comparedwith normoxia, and there was a trend for a progressive declineduring hypercapnia. Within the 15-minutes postgas inhalation pe-riod, methacholine-induced symptoms of difficult breathing, chesttightness, and breathlessness, measured using modified Borgscales, were 25–30% lower after hypoxia compared with normoxiabut were not reduced after hypercapnia. We conclude that 30 min-utes of sustained hypoxia and possibly hypercapnia impair sensa-tions of respiratory load and that the effects of hypoxia persist forat least 10 minutes after returning to normoxia.

Keywords: hypercapnia; bronchoconstriction; methacholine; dyspnea

Delay in seeking treatment during severe asthmatic episodes hasbeen implicated as an important factor responsible for increasedmorbidity and mortality. Poor perception of the severity of anasthma attack has been proposed as a mechanism underlyingtreatment delay (1). Since the initial observations of Rubinfeldand Pain (2), several studies have demonstrated that individualswith asthma, particularly those prone to severe or near fatalasthma (1, 3–6), demonstrate blunted sensations to increasedrespiratory load. Therefore, any factor that suppresses respira-tory sensation, be it an inherent persistent blunting of sensationor one that is acquired acutely, could expose patients with asthmato increased risk of morbidity or mortality (7).

Mild to moderate hypoxia, which occurs during most severeacute episodes of asthma (8, 9), has been demonstrated to havecentral nervous system depressant effects and can lead to im-paired cognitive function (10). Central nervous system depres-sant effects of hypoxia may persist for up to an hour after removalof the stimulus (11). Recent research (12) in normal subjectsdemonstrated that 30 minutes of hypoxia (SaO2 of approximately80%) depressed the perceived magnitude of externally applied

(Received in original form May 9, 2003; accepted in final form March 11, 2004)

Supported by the Foundation Daw Park/Channel 9 Telethon and National Healthand Medical Research Council of Australia.

Correspondence and requests for reprints should be addressed to Danny Eckert,B.Sc. (Hons), Adelaide Institute for Sleep Health, Repatriation General Hospital,Daw Park, South Australia, Australia, 5041. E-mail: [email protected]

This article has an online supplement, which is accessible from this issue’s tableof contents online at www.atsjournals.org

Am J Respir Crit Care Med Vol 169. pp 1224–1230, 2004Originally Published in Press as DOI: 10.1164/rccm.200305-630OC on March 12, 2004Internet address: www.atsjournals.org

resistive loads. These findings raise the possibility that individu-als with asthma who may already have blunted respiratory sensa-tions (1–6) could be further impaired in their ability to assessaccurately the degree of airway obstruction during and after aperiod of hypoxia. This may in turn lead to treatment delaysand an increased risk of asthma death.

Hypercapnia is a respiratory stimulant that enhances sensa-tions of breathlessness and often accompanies life-threateningasthma (8, 9); however, there is some evidence to suggest thatcognitive impairment occurs with moderate levels of hypercap-nia (13, 14), although these findings are conflicting (15). Thecontribution of cognitive impairment to respiratory sensationsduring acute sustained mild to moderate hypercapnia in asthmaremains unexplored.

We hypothesized that sustained moderate hypoxia wouldreduce respiratory sensations during bronchoconstriction in pa-tients with asthma. We also hypothesized that sustained hyper-capnia may lead to a relative depression of respiratory sensation.To test these hypotheses, the perception of dyspnea of subjectswith asthma during methacholine-induced bronchoconstrictionwas measured immediately after 34 minutes of isocapnic hypoxia(SaO2 of approximately 80%) and 34 minutes of hypercapnia(change in end-tidal CO2 of approximately 5–10 Torr) and com-pared with an equivalent period of normoxia. The perceivedmagnitude of externally applied resistive loads was also com-pared during the three test gas conditions. Some of the resultsof this study have been previously reported in abstract form(16–19).

METHODS

Patient Selection

Nineteen stable individuals with asthma (5 males and 14 females) whosatisfied the American Thoracic Society definition of asthma (20) gaveinformed written consent to participate in the study. Three womenexperienced nausea during hypoxia and were unable to complete thestudy. All patients were nonsmokers with a baseline FEV1 of morethan 70% predicted. None of the patients used oral corticosteroidsduring or 3 months before the study, and none had required emergencytreatment for asthma in the previous 2 years. The study was approved bythe Daw Park Repatriation General Hospital and Adelaide UniversityHuman Research and Ethics Committees.

Study Design

A within-subjects study design was chosen. After a preliminary visit(see online supplement), test gases (isocapnic hypoxia, normoxic hyper-capnia, and isocapnic normoxia) were administered for 34 minutes eachon separate days approximately 1 week apart. Two methods were usedto test for depressant effects of the test gases on respiratory sensations:First, during the gas exposures, patients rated the severity of a rangeof externally applied inspiratory resistive loads. Second, during roomair breathing, immediately after each test gas exposure, respiratorysensations of difficult breathing, chest tightness, and breathlessnesswere quantified during methacholine-induced bronchoconstriction.Methacholine challenges were administered immediately after, ratherthan during, gas exposures to avoid potentially severe hypoxemia duringbronchoconstriction coincident with hypoxic gas inhalation, combinedwith evidence that the central nervous system depressant effects of

Page 2: Hypoxia Suppresses Symptom Perception in Asthma

Eckert, Catcheside, Smith, et al.: Hypoxia and Asthma Symptom Perception 1225

isocapnic hypoxia persist for at least 15 minutes after removal of thehypoxic stimulus (11).

Three Main Experimental Visits

Patients attended the laboratory in the morning after abstaining fromcaffeine, alcohol, stimulants, and all asthma medications for at least 24hours. To characterize asthma stability during the preceding week,patients completed an asthma-control questionnaire (21). Pulmonaryfunction testing, including spirometry and whole-body plethysmo-graphy, was performed at each visit. After 5–10 minutes of nasal roomair breathing in single blind fashion and in random order, patients werepresented with one of three experimental gases: either medical-gradeair, hypoxic (fraction of inspired O2 of approximately 9%), or a hyper-capnic gas mixture (medical air with a variable increases in CO2, approx-imately 150–600 ml · minute�1) for 34 minutes. A manual inspiratorybleed of CO2 was employed to maintain isocapnia during normoxiaand hypoxia trials. During hypoxia trials, the inspired O2 fraction wasadjusted as necessary to maintain SaO2 at approximately 80% (12).During the hypercapnia protocol, the inspired CO2 concentration wasmanipulated to mimic the biphasic ventilatory response observed duringsustained hypoxia to allow comparisons between these test conditionscontrolled for drive (see the online supplement). SaO2, end-tidal CO2

(PetCO2), and minute ventilation were measured and carefully monitoredthroughout.

Externally Applied Inspiratory Resistive Loads

As described previously (12), five resistive loads (3.9 � 0.1, 4.2 � 0.1,21.8 � 0.3, 43.3 � 0.9, and 74.4 � 2.1 cm H2O · l�1 · second), werepresented for a single breath, 12 times each in random order acrossthe three 34-minute test gas conditions (see Figures E1 and E2 in theonline supplement for circuit diagram and an example trace). The lowerload was barely perceivable in most subjects, whereas the upper loadwould generally be rated as moderately severe to severe. The back-ground resistance of the circuit was 2.42 � 0.05 cm H2O · l�1 · second.Patients rated their perception of the severity of the loads using openmagnitude scaling (12, 22). Perceptual sensitivity to loads during eachgas condition was examined using both Steven’s power function expo-nent (23) and the slope of linear regression analyses of perceptionscores versus inspiratory resistance and peak inspiratory pressure (PIP)(12). Because of a clearly superior fit to perceptual scores, PIP ratherthan external resistance was regarded as the primary measure of therespiratory stimulus during loading. Similarly, linear regression wassuperior to Steven’s power function as a descriptive model of perceptionversus stimulus intensity. Consequently, the slope of the linear relation-ship between PIP and perception scores was regarded as the primarymeasure of perceptual sensitivity to external loads (see Discussion inthe online supplement for further detail). Recent data suggesting thatlinearity is the basic law of psychophysics (24) further support thisapproach.

Methacholine-induced Bronchoconstriction

Immediately after the 34-minute period of gas inhalation, patients un-derwent a modified Yan and colleagues (25) incremental methacholinebronchoprovocation test (see the online supplement for a detailed de-scription). Briefly, saline, followed by methacholine doses associatedwith 5–10%, 15–20%, and 25–30% reductions in FEV1, was deliveredusing a handheld nebulizer. One minute after each inhalation, patientsrated the severity of several components of dyspnea (difficult breathing,chest tightness, and breathlessness) using Borg category scales (26).Immediately after symptom perception assessment, two repeat mea-surements of FEV1 were performed. Three measurements of inspiratorycapacity were completed before gas inhalation and after the maximaldose of methacholine to provide a measure of hyperinflation. Perceptualsensitivity under each gas condition was measured from the slope ofthe linear regression line that best fit the FEV1% decrement versus Borgcategory score relationship (27). For a detailed description see theonline supplement.

Statistical Analysis

Analyses of covariance for repeated measures were used to comparedecrements in inspiratory capacity and symptom perception scores be-tween gas treatments, with the corresponding degree of airway obstruc-

tion as a varying covariate. All other variables comparing between-gastreatment effects were analyzed using analyses of variance for repeatedmeasures. Significant analysis of variance effects and other multiplecontrasts were examined using Student’s paired t tests adjusted formultiple comparisons using the Dunn-Sidak procedure. Statistical sig-nificance was inferred when p � 0.05. All data are reported as means �SEM.

RESULTS

Group baseline anthropometric, lung function, self-reportedmedication use, and methacholine cumulative provocation dosecorresponding to a 20% fall in the postsaline FEV1 characteristicsare displayed in Table 1. Baseline FEV1 and asthma control werestable and did not differ between experimental visits (Table 2).

During the hypoxic gas trials, SaO2 dropped rapidly withinthe first 5 minutes to 80.8 � 0.6% and was maintained thereafterat 81.5 � 0.6% (Figure 1A). SaO2 was unchanged during normoxicand hypercapnic gas trials (98.5 � 0.2% and 98.7 � 0.2%, respec-tively). PetCO2 was slightly above baseline levels during normoxicand hypoxic gas trials (0.9 � 0.3 and 1.5 � 0.4 mm Hg, respec-tively). Mean PetCO2 did not differ between these two gas condi-tions (38.2 � 0.7 and 38.0 � 0.9 mm Hg, respectively, p � 0.769).During hypercapnic trials, PetCO2 was increased (Figure 1B) tostimulate minute ventilation and mimic the hypoxic ventilatoryresponse (Figure 1C). Minute ventilation during these two gasconditions was higher than under normoxic control conditions(Figure 1C).

Perception of Externally Applied Resistive Loads (�) duringGas Inhalation

� Increased near linearly with PIP (Figure 2). Although thisrelationship was well described by Stevens’ power function � �kPIPn (overall r2 � 0.930 � 0.014, sum of squares � 21.8 � 3.4)and was superior to that of Stevens’ power function of resistance(SS p � 0.010), a linear model � � aPIP � b provided a signifi-cantly better fit (overall r2 � 0.956 � 0.006, sum of squares �13.3 � 1.9, vs. � � kPIPn sum of squares p � 0.006). Conse-quently, we have based our primary between-gas and time com-parisons on this model (see Discussion in the online supplementfor further justification). There was a significant effect of gas onthe slope of the linear relationship (p � 0.019), with a loweroverall slope during hypoxia compared with normoxia (see TableE1 in the online supplement) and a trend for a decline in slopeover time during hypercapnia (p � 0.056; Table E1). Apart froma significant difference in the exponent for � versus resistancebetween 1–5 minutes and 5–15 minutes during normoxia, therewere no other differences in exponents or intercepts (k, b) be-tween gas conditions or over time with perception expressed asa function of pressure or resistance (Table E1).

Inspiratory time during loaded breaths was prolonged withincreasing load (1.81 � 0.12 to 2.30 � 0.02 seconds from smallestto largest load, p � 0.003). However, there were no significantdifferences in loaded breath inspiratory time between gases (p �0.156) across time during gas inhalation (p � 0.788) and nointeraction effects between load magnitude, time, or gas (p �0.103).

Bronchoconstriction: Perception of Symptom Intensity

Bronchoprovocation testing was completed within 15 minutesafter gas inhalation. Time to complete testing did not differsignificantly between gas conditions (Table 2). There was a sig-nificant decrease in SaO2 during bronchoprovocation, althoughthe severity of desaturation was not statistically different be-tween gas treatments (Table 2). Inspiratory capacity decreasedafter bronchoconstriction, but the decrease was not different

Page 3: Hypoxia Suppresses Symptom Perception in Asthma

1226 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 169 2004

TABLE 1. GROUP ANTHROPOMETRIC, LUNG FUNCTION, MEDICATION,AND PD20 CHARACTERISTICS

PD20 range (cumulativemg methacholine) Mean � SEM No. of Patients

Baseline characteristicsAge, yr – 23.9 � 1.4 –Body mass index, kg · m�2 – 24.7 � 1.1 –FEV1, % predicted – 99.3 � 2.3 –MMEF75/25, % predicted – 76.1 � 4.0 –Total lung capacity, % predicted – 119.3 � 4.1 –

MedicationsInhaled short-acting �2 agonists only – – 7Inhaled corticosteroids only – – 1Inhaled corticosteroids plus short-acting �2 agonists – – 5Inhaled long and short-acting �2 agonists – – 1No regular medications – – 2

Asthma severity*Mild 0.263–1.521 – 4Moderate 0.031–0.162 – 11Severe 0.005 – 1

Definition of abbreviation: MMEF � maximal midexpiratory flow.* Asthma severity according to the cumulative provocation dose of methacholine corresponding to a 20% fall in the postsaline

FEV1 characterization described by Juniper and colleagues (53) (n � 16).

between gas conditions, suggesting similar levels of gas trappingduring methacholine challenges (Table 2).

With worsening levels of airway obstruction, patients re-ported more intense symptoms, and these relationships weresignificantly altered by gas treatment (analyses of covariance gasby FEV1 interaction; difficult breathing p � 0.001, Figure 3; chesttightness, p � 0.003; and breathlessness, p � 0.001).

For the symptoms of difficult breathing and chest tightness,the slope of the relationship between symptom score and FEV1

was significantly lower after hypoxia than after either hypercap-nia or normoxia (Table 2). Symptoms after hypercapnia werenot different compared with normoxia. For the symptom ofbreathlessness, the slope of the perception score–FEV1 relation-ship was significantly lower after hypoxia compared with thatafter hypercapnia, and approached statistical significance com-pared with the slope after normoxia (p � 0.076) (Table 2).

DISCUSSION

The main finding of this study was that in a group of individualswith stable asthma, symptom perception during methacholine-

TABLE 2. MAIN STUDY BASELINE AND BRONCHOPROVOCATION CHARACTERISTICS

Normoxia Hypercapnia Hypoxia Gas Effect p Value

Baseline MeasuresBaseline FEV1, % predicted 99.4 � 2.9 99.6 � 3.3 99.3 � 3.5 0.966Asthma control questionnaire* 0.75 � 0.17 0.84 � 0.15 0.80 � 0.15 0.761

Bronchoprovocation measuresTest duration, min† 10.0 � 0.5 8.9 � 0.3 9.3 � 0.6 0.214 Inspiratory capacity, L‡ �0.52 � 0.10 �0.42 � 0.05 �0.43 � 0.07 0.244 SaO2

, %§ 3.2 � 0.7 3.4 � 0.6 4.2 � 0.9 0.299Slope of symptom intensity versus FEV1

Difficult breathing 0.078 � 0.015 0.075 � 0.017 (102 � 24%) 0.049 � 0.013¶ || (71 � 13%) 0.031Chest tightness 0.071 � 0.013 0.066 � 0.014 (102 � 24%) 0.046 � 0.012¶ || (72 � 12%) 0.024Breathlessness 0.073 � 0.014 0.069 � 0.016 (110 � 29%) 0.050 � 0.013|| (74 � 14%) 0.045

Values are means � SEM. Values in parentheses are expressed as a percentage of normoxia values.*Asthma control questionnaire (21), a 7-point scale in which 0 is “well controlled” and 6 is “extremely poorly controlled.”† Time taken from end of 34 minutes of gas inhalation to the final symptom rating at maximal constriction during bronchoprovocation.‡ Change in inspiratory capacity at maximal constriction during bronchoprovocation compared with baseline.§ Maximal decrease in arterial blood O2 saturation below baseline.¶ Indicates a significant difference in the slope between normoxia and hypoxia (p � 0.05).|| Indicates a significant difference in the slope between hypercapnic and hypoxic gas trials (p � 0.05) (n � 16).

induced bronchoconstriction was suppressed for at least 10 min-utes after sustained isocapnic hypoxia when compared withequivalent periods of prior hypercapnic hyperventilation andnormoxia. This finding was consistent for the specific dyspneasymptoms of “difficult breathing” and “chest tightness” andapproached statistical significance for the symptom of “breath-lessness.” In the same subjects, perceptual sensitivity to exter-nally applied inspiratory resistive loads, expressed as the slopeof the linear relationship between perception scores and PIP, wasdecreased during hypoxia compared with normoxia. Althoughthese effects were not apparent on the basis of Steven’s powerfunction analyses, Steven’s function provided a significantlypoorer fit. Furthermore, recent data favor linearity rather thanSteven’s power function as the basic psychophysical law (24).

Hypoxia is a clinical feature of moderate to severe acuteepisodes of asthma (8, 9). Initially, hypoxia stimulates ventilationvia peripheral chemoreceptor activation. This is associated withsensations of dyspnea, either through increased intensity of re-spiratory motor command (28, 29) or heightened chemosensoractivation per se (30). There is evidence to suggest that these

Page 4: Hypoxia Suppresses Symptom Perception in Asthma

Eckert, Catcheside, Smith, et al.: Hypoxia and Asthma Symptom Perception 1227

Figure 1. Group SaO2 measured via pulse oximetry (A ), end-tidal carbondioxide pressure (PETCO2) measured at the mask (B ), and minute ventila-tion (V̇I) (C ) during the three 34-minute gas trials. Values are means �

SEM (n � 16).

responses may be impaired in asthma, particularly in patientswith a history of near fatal episodes (1, 31). When hypoxia issustained or repetitive in nature, there is a central accumulationof inhibitory neuroeffectors (32), which can result in depressiveeffects, such as centrally mediated depression of ventilation (32,33) and cognitive function (10). These findings raise the possibil-ity that hypoxia may cause a relative depression of respiratorysensation in disorders such as asthma. Research in healthy indi-viduals provides additional support for this hypothesis. Lane andcolleagues (28) reported that for equivalent levels of hyperventi-

Figure 2. Peak inspiratory pressure (PIP) versus the perceived magni-tude of externally applied resistive loads (�) during 34 minutes of isocap-nic normoxia, isocapnic hypoxia, and hypercapnia. Values are means �

SEM (n � 16).

lation, lower scores of breathlessness were obtained during exer-cise with hypoxia than with exercise alone or exercise combinedwith hypercapnia. After 20 minutes of sustained isocapnic hyp-oxia (SaO2 of approximately 80%), Chonan and colleagues (34)demonstrated a reduced perception of dyspnea during a low-intensity exercise task compared with the early phase of hypoxiaunder resting conditions. This was despite ventilation duringexercise being greater than during early hypoxia. The authorsspeculated that this sensory attenuation was due to central ner-vous system depression associated with sustained hypoxia. Morerecently and consistent with this hypothesis, Orr and colleagues(12) studying healthy subjects found that sustained isocapnichypoxia (30 minutes, SaO2 approximately 80%) depressed theperception of externally applied resistive loads after several min-utes of hypoxia compared with normoxic breathing. The slopeof subject’s respiratory sensation versus ventilation relationshiphas also recently been reported to be depressed with increasinghypoxia (35).

This study provides evidence that hypoxic depression of respi-ratory sensations also occurs in a population with asthma and thatthese depressant effects persist for at least 10 minutes after cessationof hypoxia. The persistence of central nervous system depressionafter hypoxia is consistent with earlier work in which depressedventilatory responses to repeated hypoxic challenges were notedfor up to an hour after a period of sustained hypoxia (11).

Perception of dyspnea and the underlying mechanisms re-sponsible for respiratory resistive load detection is complex.Multiple sensory pathways (30) and several brain regions (36–38)are involved. Hypoxia increases central nervous system levelsof specific neuroinhibitors such as endogenous opioids, adeno-sine, and gamma-aminobutyric acid (GABA) (32). Each has beenimplicated in blocking sensory pathways involved in the percep-tion of pain (39–43). Although respiratory sensations are dis-tressing rather than painful, similar hypoxia-sensitive pathwayscould potentially be involved. Hypoxia could disturb respiratoryafferent pathways and neural processing at more than one level.Using an adult cat model, Zimmerman and Grossie (44) foundthat systemic arterial hypoxemia (less than approximately 80mm Hg [SaO2 of approximately 92%]) had a depressive effecton afferent discharge of muscle spindles. However, the centralnervous system has a rostral to caudal vulnerability to hypoxia(32) such that higher centers may be particularly sensitive to theinhibitory effects of hypoxia. Higgs and Laszlo (45) demon-

Page 5: Hypoxia Suppresses Symptom Perception in Asthma

1228 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 169 2004

Figure 3. Perception of the intensity of “difficult breathing”versus the magnitude of methacholine-induced broncho-constriction (FEV1, % of postsaline value) immediately afterthe 34 minutes of each gas treatment. Values are means �

SEM (n � 16).

strated an increase in perception of asthma after theophylline(a higher center and muscle function stimulant) administrationin patients with asthma, providing some support for the impor-tance of higher center pathways for sensations of dyspnea inasthma. In healthy individuals, acute hypoxia slows perceptualprocessing to auditory and visual stimuli (46–48). Patients withsevere chronic obstructive pulmonary disease also appear tohave abnormal auditory evoked responses, which is suggested toresult from chronic hypercapnic–hypoxia exposure (49). Furtherstudies are required to establish the sites of action and the neuro-chemical mechanisms responsible for hypoxia-induced bluntingof dyspnea during bronchoconstriction.

Dynamic hyperinflation is a consequence of progressive flowlimitation as bronchoconstriction develops and has been demon-strated to increase the perception of dyspnea (30). In this study,although inspiratory capacity significantly decreased (a markerof dynamic hyperinflation), there was no difference in the degreeof hyperinflation after bronchoconstriction between the threegas trials. Thus, our results cannot be explained by differencesin hyperinflation.

In support of our findings during bronchoconstriction, thesubjects with asthma in this study showed a rapid and sustaineddecrement in � for a given PIP during hypoxia. Using an identicalprotocol with similar resistances, the overall slope of the relation-ship between � and PIP did not differ between normoxia andhypoxia when compared with the healthy individuals in ourprevious report (12) (see Figure E4 in the online supplement).However, in contrast to normal subjects who showed a delayedbut progressive decline (12), individuals with asthma demon-strated rapid decrements in � during sustained hypoxia. Thismay reflect an increase in susceptibility to the inhibitory effectsof hypoxia. Webster and Colrain (50) recently demonstrateddifferences in respiratory and auditory cortical-evoked responsesin adult patients with asthma compared with control subjects.Children with life-threatening asthma also appear to have alteredneural processing of inspiratory load information (51). Thesefindings suggest an underlying disturbance of cortical sensoryprocessing.

An apparent trend for a gradual relative blunting of � withhypercapnia may indicate different depressant effects on neuralfunction compared with hypoxia. One possibility is that CO2

narcosis impaired cognitive function. Cognitive impairment hasbeen shown in some previous studies (13, 14) but not others (15)at PetCO2 levels similar to the maximum PetCO2 values achieved inthis study.

Methodologic Considerations and Implications

We measured asthma-induced symptom scores after test gasexposures and cannot therefore be certain whether our findingof reduced severity of “difficult breathing” and “chest tightness”after hypoxia would also apply during such a gas exchange distur-bance if it accompanied, for example, an acute spontaneousasthma attack. Given our finding that sensation of externallyapplied loads, expressed in terms of PIP, were reduced duringhypoxia, we think it likely that asthma symptom scores wouldbe similarly affected. However, it must also be acknowledgedthat mechanical disturbances precede the onset of hypoxemia(52). Thus, any detrimental effects of hypoxia on respiratorysensation may be most evident relatively late in the evolutionof an acute severe attack. If a patient with acute asthma isaware of breathing difficulties but with the advent of hypoxemiaperceives improvement in symptoms, he or she may falselyconclude that the attack is abating. Approximately 15% of thepopulation with asthma are known to be “poor perceivers” atdetecting the mechanical changes associated with progressiveairway obstruction (2). In the case of an acute severe episodeof asthma, should the mechanical changes progress undetectedand hypoxemia develops, such patients may be vulnerable tofurther hypoxic-induced suppression of respiratory sensation.

It must also be acknowledged that by controlling ventilationduring hypercapnia to match that of hypoxia, hypercapnia de-creased with time to relatively low levels, which limit the inter-pretation of this arm of the protocol. To assess definitively theeffects of prolonged hypercapnia on respiratory sensation inasthma, a protocol consisting of sustained higher levels of PetCO2may be more clinically relevant.

A further methodologic limitation of the study was that PetCO2was not controlled precisely under the normoxia and hypoxiaprotocols. We think this is unlikely to have affected our conclu-sions. First, the rise in PetCO2 above eucapnic levels was small(1–1.5 mm Hg). Cognitive dysfunction does not appear to occuruntil a PetCO2 well above the levels encountered in the hypoxiaand normoxia trials in this study (13, 14). Second, PetCO2 was notdifferent during hypoxia and normoxia experiments. Anotherlimitation of this study was that we did not measure minuteventilation or PetCO2 during bronchoconstriction and thus cannotdetermine whether the reduction in sensation of dyspnea re-sulted in a reduction in load compensation. There was slightlymore desaturation (nonsignificant) during bronchoconstrictionafter hypoxia than after normoxia and hypercapnia; however,SaO2 is a relatively insensitive and imprecise measure of ven-

Page 6: Hypoxia Suppresses Symptom Perception in Asthma

Eckert, Catcheside, Smith, et al.: Hypoxia and Asthma Symptom Perception 1229

tilatory adaptation to increased respiratory load. It would beinteresting to know whether the reduced sensation of dyspnearesulting from prior hypoxia leads to a disproportionate fall inventilation during bronchoconstriction.

The findings of this study suggest that individuals with asthmawith apparently already blunted respiratory sensations may befurther impaired by hypoxia to recognize increased respiratoryload during times of airway obstruction. In this study, broncho-constriction was limited to 65–70% of baseline. However, blunt-ing of symptom perception could conceivably be much morepronounced at levels of obstruction typical of life-threateningepisodes of asthma. Extrapolating the relationships establishedin this study to a postsaline FEV1 of 40% (i.e., correspondingto a moderately severe attack), patients would report “somewhatsevere” to “severe” symptom intensities (approximately 4.6–5.0Borg scale units) under normoxic conditions but only “moder-ate” symptom severity (approximately 3 Borg scale units) underhypoxic conditions. In acute asthma, CO2 retention could po-tentially play an independent role, further depressing symptomperception.

These findings reinforce the need for simple, accessible objec-tive measures of lung function to assess asthma severity and tolimit the reliance on symptom reporting. Preservation of oxygensaturation during acute severe asthma might be important inpreserving symptom perception, preventing treatment delays,and reducing asthma morbidity and mortality.

Conflict of Interest Statement : D.J.E. does not have a financial relationship witha commercial entity that has an interest in the subject of this article; P.G.C. doesnot have a financial relationship with a commercial entity that has an interest inthe subject of this article; J.H.S. does not have a financial relationship with acommercial entity that has an interest in the subject of this article; P.A.F. doesnot have a financial relationship with a commercial entity that has an interest inthe subject of this article; R.D.M. does not have a financial relationship with acommercial entity that has an interest in the subject of this article.

Acknowledgment : The authors are indebted to David Schembri and the Respira-tory Function Laboratory staff at Repatriation General Hospital, Daw Park, SouthAustralia, for their assistance with lung function measurements and to TheanVlahakis and Paul Henshall of the Pharmacy Department for their assistance withpreparation with the methacholine solutions. Dr. Lata Jayaram provided valuableadvice on methods to assess asthma stability.

References

1. Kikuchi Y, Okabe S, Tamura G, Hida W, Homma M, Shirato K, Takis-hima T. Chemosensitivity and perception of dyspnea in patients witha history of near-fatal asthma. N Engl J Med 1994;330:1329–1334.

2. Rubinfeld AR, Pain MC. Perception of asthma. Lancet 1976;1:882–884.3. Kifle Y, Seng V, Davenport PW. Magnitude estimation of inspiratory

resistive loads in children with life-threatening asthma. Am J RespirCrit Care Med 1997;156:1530–1535.

4. Ruffin RE, Latimer KM, Schembri DA. Longitudinal study of near fatalasthma. Chest 1991;99:77–83.

5. Burdon JG, Juniper EF, Killian KJ, Hargreave FE, Campbell EJ. Theperception of breathlessness in asthma. Am Rev Respir Dis 1982;126:825–828.

6. Magadle R, Berar-Yanay N, Weiner P. The risk of hospitalization andnear-fatal and fatal asthma in relation to the perception of dyspnea.Chest 2002;121:329–333.

7. Barnes PJ. Poorly perceived asthma. Thorax 1992;47:408–409.8. McFadden ER Jr, Lyons HA. Arterial-blood gas tension in asthma. N

Engl J Med 1968;278:1027–1032.9. Tai E, Read J. Blood-gas tensions in bronchial asthma. Lancet 1967;1:644–

646.10. Berry DT, McConnell JW, Phillips BA, Carswell CM, Lamb DG, Prine

BC. Isocapnic hypoxemia and neuropsychological functioning. J ClinExp Neuropsychol 1989;11:241–251.

11. Easton PA, Slykerman LJ, Anthonisen NR. Recovery of the ventilatoryresponse to hypoxia in normal adults. J Appl Physiol 1988;64:521–528.

12. Orr RS, Jordan AS, Catcheside P, Saunders NA, McEvoy RD. Sustainedisocapnic hypoxia suppresses the perception of the magnitude of inspi-ratory resistive loads. J Appl Physiol 2000;89:47–55.

13. Sayers JA, Smith RE, Holland RL, Keatinge WR. Effects of carbondioxide on mental performance. J Appl Physiol 1987;63:25–30.

14. Fothergill DM, Hedges D, Morrison JB. Effects of CO2 and N2 partialpressures on cognitive and psychomotor performance. Undersea BiomedRes 1991;18:1–19.

15. Bloch-Salisbury E, Lansing R, Shea SA. Acute changes in carbon dioxidelevels alter the electroencephalogram without affecting cognitive func-tion. Psychophysiology 2000;37:418–426.

16. Catcheside P, Smith J, Eckert D, Frith P, Schembri D, McEvoy RD.Hypoxia and hypercapnic hyperventilation depress inspiratory resis-tive load magnitude perception in asthmatics [abstract]. Respirology2003;8:A29.

17. Catcheside P, Smith J, Eckert D, Frith P, Schembri D, McEvoy RD.Hypoxia and hypercapnic hyperventilation depress inspiratory re-sistive load magnitude perception in asthmatics [abstract]. Am J RespirCrit Care Med 2003;167:A794.

18. Eckert D, Smith J, Catcheside P, Frith P, Schembri D, McEvoy RD.Hypoxia suppresses symptom perception in asthma [abstract]. Respir-ology 2003;8:A5.

19. Eckert D, Smith J, Catcheside P, Frith P, Schembri D, McEvoy RD.Hypoxia suppresses symptom perception in asthma [abstract]. Am JRespir Crit Care Med 2003;167:A793.

20. American Thoracic Society. Standards for the diagnosis and care ofpatients with chronic obstructive pulmonary disease (COPD) andasthma: this official statement of the American Thoracic Society wasadopted by the ATS Board of Directors, November 1986. Am RevRespir Dis 1987;136:225–244.

21. Juniper EF, O’Byrne PM, Guyatt GH, Ferrie PJ, King DR. Developmentand validation of a questionnaire to measure asthma control. EurRespir J 1999;14:902–907.

22. Killian KJ, Mahutte CK, Campbell EJ. Magnitude scaling of externallyadded loads to breathing. Am Rev Respir Dis 1981;123:12–15.

23. Stevens SS. Neural events and the psychophysical law. Science 1970;170:1043–1050.

24. Johnson KO, Hsiao SS, Yoshioka T. Neural coding and the basic law ofpsychophysics. Neuroscientist 2002;8:111–121.

25. Yan K, Salome C, Woolcock AJ. Rapid method for measurement ofbronchial responsiveness. Thorax 1983;38:760–765.

26. Borg GA. Psychophysical bases of perceived exertion. Med Sci SportsExerc 1982;14:377–381.

27. Bijl-Hofland ID, Cloosterman SG, Folgering HT, Akkermans RP, vanden Hoogen H, van Schayck CP. Measuring breathlessness duringhistamine challenge: a simple standardized procedure in asthmaticpatients. Eur Respir J 1999;13:955–960.

28. Lane R, Adams L, Guz A. The effects of hypoxia and hypercapnia onperceived breathlessness during exercise in humans. J Physiol 1990;428:579–593.

29. Moosavi SH, Golestanian E, Binks AP, Lansing RW, Brown R, BanzettRB. Hypoxic and hypercapnic drives to breathe generate equivalentlevels of air hunger in humans. J Appl Physiol 2003;94:141–154.

30. Dyspnea: mechanisms, assessment, and management: a consensus state-ment: American Thoracic Society. Am J Respir Crit Care Med 1999;159:321–340.

31. Hudgel DW, Capehart M, Hirsch JE. Ventilation response and driveduring hypoxia in adult patients with asthma. Chest 1979;76:294–299.

32. Neubauer JA, Melton JE, Edelman NH. Modulation of respiration duringbrain hypoxia. J Appl Physiol 1990;68:441–451.

33. McEvoy RD, Popovic RM, Saunders NA, White DP. Effects of sustainedand repetitive isocapnic hypoxia on ventilation and genioglossal anddiaphragmatic EMGs. J Appl Physiol 1996;81:866–875.

34. Chonan T, Okabe S, Hida W, Satoh M, Kikuchi Y, Takishima T, ShiratoK. Influence of sustained hypoxia on the sensation of dyspnea. Jpn JPhysiol 1998;48:291–295.

35. Masuda A, Ohyabu Y, Kobayashi T, Yoshino C, Sakakibara Y, KomatsuT, Honda Y. Lack of positive interaction between CO2 and hypoxicstimulation for P(CO2)-VAS response slope in humans. Respir Physiol2001;126:173–181.

36. Gozal D, Omidvar O, Kirlew KA, Hathout GM, Hamilton R, LufkinRB, Harper RM. Identification of human brain regions underlyingresponses to resistive inspiratory loading with functional magneticresonance imaging. Proc Natl Acad Sci USA 1995;92:6607–6611.

Page 7: Hypoxia Suppresses Symptom Perception in Asthma

1230 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 169 2004

37. Isaev G, Murphy K, Guz A, Adams L. Areas of the brain concernedwith ventilatory load compensation in awake man. J Physiol 2002;539:935–945.

38. Peiffer C, Poline JB, Thivard L, Aubier M, Samson Y. Neural substratesfor the perception of acutely induced dyspnea. Am J Respir Crit CareMed 2001;163:951–957.

39. Christie MJ, Chesher GB, Bird KD. The correlation between swim-stressinduced antinociception and [3H] leu-enkephalin binding to brainhomogenates in mice. Pharmacol Biochem Behav 1981;15:853–857.

40. Chudler EH, Dong WK. The role of the basal ganglia in nociception andpain. Pain 1995;60:3–38.

41. Haier RJ, Quaid K, Mills JC. Naloxone alters pain perception afterjogging. Psychiatry Res 1981;5:231–232.

42. Segerdahl M, Ekblom A, Sollevi A. The influence of adenosine, ketamine,and morphine on experimentally induced ischemic pain in healthyvolunteers. Anesth Analg 1994;79:787–791.

43. Sollevi A. Adenosine for pain control. Acta Anaesthesiol Scand Suppl1997;110:135–136.

44. Zimmerman GW, Grossie J. Sensitivity and behavior of muscle spindlesto systemic arterial hypoxia. Proc Soc Exp Biol Med 1969;132:1114–1118.

45. Higgs CM, Laszlo G. Influence of treatment with beclomethasone, cro-moglycate and theophylline on perception of bronchoconstriction inpatients with bronchial asthma. Clin Sci (Lond) 1996;90:227–234.

46. Fowler B, Lindeis AE. The effects of hypoxia on auditory reaction timeand P300 latency. Aviat Space Environ Med 1992;63:976–981.

47. Fowler B, Kelso B. The effects of hypoxia on components of the humanevent-related potential and relationship to reaction time. Aviat SpaceEnviron Med 1992;63:510–516.

48. Fowler B, Prlic H. A comparison of visual and auditory reaction timeand P300 latency thresholds to acute hypoxia. Aviat Space EnvironMed 1995;66:645–650.

49. Atis S, Ozge A, Sevim S. The brainstem auditory evoked potential abnor-malities in severe chronic obstructive pulmonary disease. Respirology2001;6:225–229.

50. Webster KE, Colrain IM. P3-specific amplitude reductions to respiratoryand auditory stimuli in subjects with asthma. Am J Respir Crit CareMed 2002;166:47–52.

51. Davenport PW, Cruz M, Stecenko AA, Kifle Y. Respiratory-relatedevoked potentials in children with life-threatening asthma. Am J RespirCrit Care Med 2000;161:1830–1835.

52. Davenport PW. What caused the screech, whamity, bang, and thump?Am J Respir Crit Care Med 2002;165:2–3.

53. Juniper EF, Cockcroft DW, Hargreave FE. Tidal breathing vs othermethods: histamine and methacholine inhalation tests: tidal breathingmethod, laboratory procedure and standardisation. Lund, Sweden:Astra Draco AB; 1994. p. 31–32.