DYSFUNCTIONAL BREATHING IN ASTHMA: IS IT COMMON, IDENTIFIABLE AND CORRECTABLE?

Transcription

1 Thorax 2002;57(Suppl II):ii31 ii35 DYSFUNCTIONAL BREATHING IN ASTHMA: IS IT COMMON, IDENTIFIABLE AND CORRECTABLE? M D L Morgan *ii31 Introdutory artile Prevalene of dysfuntional breathing in patients treated for asthma in primary are: ross setional survey M Thomas, RK MKinley, E Freeman, C Foy Objetives: To estimate the prevalene of dysfuntional breathing in adults with asthma treated in the ommunity. Design: Postal questionnaire survey using Nijmegen questionnaire. Setting: One general pratie with 7033 patients. Partiipants: All adult patients aged with diagnosed asthma who were reeiving treatment. Main outome measure: Sore >23 on Nijmegen questionnaire. Results: 227/307 patients returned ompleted questionnaires; 219 (71.3%) questionnaires were suitable for analysis. 63 partiipants sored >23. Those soring >23 were more likely to be female than male (46/ 132 (35%) v 17/87 (20%), p=0.016) and were younger (mean (SD) age 44.8 (14.7) v 49.0 (13.8), p=0.05). Patients at different treatment steps of the British Thorai Soiety asthma guidelines were affeted equally. Conlusions: About a third of women and a fifth of men had sores suggestive of dysfuntional breathing. Although further studies are needed to onfirm the validity of this sreening tool and these findings, these prevalenes suggest sope for therapeuti intervention and may explain the anedotal suess of the Buteyko method of treating asthma. (BMJ 2001;322: ) M D L Morgan Institute for Lung Health, Department of Respiratory Mediine and Thorai Surgery, Glenfield Hospital, Groby Road, Leiester LE3 9QP, UK; mike.morgan@uhl-tr.nhs.uk BACKGROUND One remarkable thing about the physiology of the human body is the ability to adapt to onstantly hanging irumstanes and adverse environments. In the respiratory system arterial blood gas tensions are usually held within tight limits of normality in the fae of gravitational hanges in posture, the requirements of speeh and exerise, as well as many other human ativities. Any tendeny for ventilation to stray beyond these limits is rapidly orreted by appropriate homeostati mehanisms that operate impereptibly. In theory, abnormal ventilatory equilibrium an be produed by sustained hyperventilation or hypoventilation. However, the latter is only likely to our in serious organi disease or drug indued narosis when the ause is usually obvious. Nevertheless, the potential for variability exists and it has been reognised for deades that the pattern of ventilation that develops in advaned hroni obstrutive pulmonary disease (COPD) is independent of the airway pathology. Some patients will defend their arterial arbon dioxide tensions as pink puvers while others will passively aept the omfort and onsequenes of hyperapnia as blue bloaters. This behaviour in COPD is not voluntary or psyhogeni but represents an intrinsi possibly genetially determined reation to illness. Sine in other irumstanes the strong intrinsi physiologial defene mehanisms would make sustained voluntary hypoventilation unahievable, it has beome aepted in this instane that the phenotypi appearane of the patient in respiratory failure is due to an interation of lung pathology and a predetermined pattern of respiratory drive. As a onsequene, the presene of an altered breathing pattern in COPD is seen as an epiphenomenon rather than the root ause. This view is not neessarily aepted in asthma where the impression of linkage between breathing pattern and pathology appears less aeptable. Sustained hyperventilation, surplus to physiologial requirement, an be indued by onsious ation, behavioural hange, pharmaologial stimulation, or neurologial disease. As every hild will have learnt, deliberate onsious and temporary interferene with homeostasis by hyperventilation is relatively easy to ahieve. Aute deliberate or psyhogeni hyperventilation an be assoiated with distressing physial symptoms that are usually reognised as suh and traditionally treated by rebreathing or anxiolysis. Chroni or reurrent hanges in breathing pattern have also been assoiated with physial symptoms of breathlessness, hest tightness, hest pain, dizziness, and anxiety that are

2 Morgan ii32 * variously desribed as dysfuntional breathing or hyperventilation syndrome. 1 4 This behaviour an masquerade as asthma or our in assoiation with genuine asthma or other airways disease.this assoiation has always presented diyulties for liniians who tread a fine line between inadequately treating naturally anxious people with asthma or presribing unneessary and exessive treatment where it will be inevetive. It would therefore be helpful to have a method of distinguishing the ontribution of possible dysfuntional breathing from the inflammatory or bronhoonstritor omponents of genuine asthma. One obvious solution is to titrate preventative ortiosteroid treatment to the inflammatory response so that patients only reeive the asthma treatment that they need. This approah is promising and is urrently being explored with a variety of markers of airway inflammation. 5 The bronhoonstritor omponent of asthma is traditionally managed by a ombination of physiologial monitoring and symptom ontrol by bronhodilator. Although the latter strategy is adequate for most situations, it will not be helpful where symptoms of dysfuntional breathing are interspersed with those of airway narrowing. For that reason it would be helpful to be able to identify and distinguish the evets of dysfuntional breathing so that the true ontribution an be assessed and alternative or additional treatment overed where appropriate or possible. In this respet, the Introdutory artile by Thomas et al 6 should be onsidered alongside the aompanying editorial by Keely and Osman. 7 Thomas et al have used a simple questionnaire to identify the prevalene of dysfuntional breathing in a general pratie population that has previously been defined as asthmati. They identified a high prevalene of dysfuntional breathing aross all grades of severity of asthma, partiularly in women. Superfiially, the study is fairly innouous and appears to make a useful ontribution to the body of knowledge in this area. It seems to define the size of the problem and point the way to alternative treatment for those who are inappropriately treated. However, there are several assumptions within the methodology and the disussion whih make the onlusions diyult to generalise. As a result, the study raises almost as many questions as it answers and unsrambling the eggs from the hikens is diyult. The main issues an largely be onsidered under three headings. Can dysfuntional breathing be defined and identified? The term dysfuntional breathing is not a preise or definable entity. It asribes ausality to a pattern of abnormal breathing that may result at worst in a symptom omplex of breathlessness, hest tightness, paraesthesiae, anxiety, and dizziness. This may inlude overt hyperventilation and hypoapnia but also inludes more subtle and subjetive features that are diyult to haraterise. This new term of desription presumably inludes overt episodi hyperventilation, hroni hyperventilaton syndrome, and disproportionate breathlessness. This reates something of a diyulty sine, in the ontext of this paper, the ondition is defined by the method used to reognise it. Dysfuntional breathing is therefore just the latest impreise desription of a behaviour and symptom omplex whih remains unexplained. Physiologial alveolar hyperventilation leads to hypoapnia and respiratory alkalosis that may lead to neuronal hyperexitability. The initial desription of hyperventilation syndrome was used to identify people who developed somati symptoms and anxiety assoiated with overbreathing. This desription has sine beome orrupted to inlude those with anxiety and appropriate somati symptoms even when hypoventilation is not obvious or onstant. In fat, the relationship between the apparently reognisable symptom omplex and hypoapnia is far from lear. Although harateristi symptoms an be reprodued by voluntary hyperventilation, the demonstration of hypoapnia is not a reliable diagnosti test. 89 Hypoapnia in symptomati people may be episodi and may not relate diretly to the symptoms. Ambulatory monitoring of arterial PaCO 2 or end tidal CO 2 has not yet provided evidene of a relationship between symptoms and hypoapnia. In fat, it is evident that hyperventilation and hypoapnia an our in the omplete absene of symptoms. If there is no physiologial observation that haraterises the syndrome, it may be possible that a provoation test might reprodue the symptoms. In normal people voluntary hyperventilation will produe tetany or paraesthesia but not the wider range of symptoms. Some simple provoation like the 20 deep breath test may reprodue the symptoms and suggest the diagnosis. More formal observation of prolonged hyperapnia assoiated with short periods of hyperventilation has also been proposed. Unfortunately the validity of this approah is partly undermined by the observation that symptoms an still our when normal PaCO 2 is artifiially maintained. 10 It is now generally aknowledged that the syndrome need not neessarily be linked to hypoapnia but more to the at of overbreathing. 2 The term hyperventilation remains, but other desriptors suh as disproportionate breathlessness may be more apt. Other physiologial approahes have examined the errati ventilatory response to exerise that an dislose inappropriate ventilation. Another onept is that the syndrome may result from a disorder of physiologial ontrol or autonomi failure. So far, no speifi abnormality of CO 2 regulation has been identified, but hanges in ventilation with postural hallenge have been observed. 11 If there is no satisfatory physiologial definition of dysfuntional breathing, perhaps the answer an be found in a psyhologial explanation. The patient with obvious air hunger, sighing respiration, or episodi hyperventilation may well have a psyhiatri disorder and respond to appropriate treatment. However, the waters are very muddied and dismissal of dysfuntional breathing as a manifestation of pani attak is onsidered by both psyhiatrists and physiologists as an oversimplifiation. 1 It is established that some people with asthma may have impaired pereption of dyspnoea that ould be potentially dangerous. It is also possible that the spetrum extends to those who develop an enhaned sensitivity to airway responsiveness or airway ondutane and reat by hyperventilating. More overtly, people with anxiety may overbreathe, but then distressing asthma is a perfetly understandable soure of anxiety. In this ase the syndrome of dysfuntional breathing therefore defines itself by the self-reinforing symptom hek list whih was developed from those with a linial diagnosis of hyperventilation. The Nijmegen questionnaire is therefore diverent from other disease speifi health status measures where the primary diagnosis is predefined by independent riteria. In the original validation of the questionnaire there was some rossover with a healthy referene group. 12 No validation was performed with people who had other auses of breathlessness. It is therefore diyult to determine the disriminative properties of the questionnaire when many of the 16 relevant symptoms might appear in genuine respiratory disease. Furthermore, people with a positive

3 Dysfuntional breathing in asthma Nijmegen result may or may not exhibit spontaneous hypoapnia or response to provoation. 9 It seems therefore that a linial syndrome of symptoms exists whih may be loosely linked to overbreathing and formally desribed by the strutured linial history of the Nijmegen questionnaire. However, physiologial definition or reognition of the syndrome does not exist and lear separation of the ondition by the questionnaire from genuine organi disease has not yet been made. This does not mean that dysfuntional breathing is not an entity, but does make it diyult to quantify in the suspeted presene of asthma. How does dysfuntional breathing relate to asthma? Aute and hroni hyperventilation are the omponents of dysfuntional breathing that are most likely to relate to asthma. Aute asthma attaks are often assoiated with a redution in PaCO 2 in assoiation with developing hypoxaemia. This is expeted as the natural attak develops and reverses with evetive treatment. Similar aute hyperventilation also ours with indued bronhoonstrition through bronhial hallenge and is again reversible. The onverse is observed when aute bronhospasm is indued by deliberate hyperventilation or inreased ventilation assoiated with physial exerise. In this ase the mehanism of bronhial hallenge is thought to inlude airway ooling and drying. These aute situations are well reognised and appropriate treatment or advie an usually resolve the problem. The relationship between hroni hyperventilation and asthma is more ompliated. Several papers have reported a high inidene of overt respiratory disease in patients who have been linially diagnosed with the hyperventilation syndrome. In partiular, underlying asthma has been identified as the most ommon ause of onfounding illness. In these ases the identifiation of asthma has been made by bronhial hallenge, a bronhodilator response, or improvement with treatment. In one series 80% of subjets with the hyperventilation syndrome turned out to have asthma. 13 The reverse situation of hyperventilation ourring naturally in mild asthma does not appear to be true exept during periods of bronhial hyperresponsiveness. 11 Interestingly, there is a irumstane where the airways might be exposed to prolonged high levels of ventilation where airway ooling and drying may our. Athletes, partiularly those involved in winter sports, may be exposed to airway hallenge in this way. There is evidene that asthma is more ommon in these people and that inreased airway responsiveness is assoiated with the environmental hallenge. 14 Experimentally, repeated hyperventilation has also led diretly to peripheral airway inflammation, hyperreativity, and impaired bronhodilation in dogs. 15 It therefore appears that hyperventilation an our in assoiation with aute asthma and possibly ontribute to a hyperreative state in extreme onditions. It is more diyult to unover a relationship between asthma and other features of dysfuntional breathing. Some of the symptoms will obviously overlap. Breathlessness, hest pain, hest onstrition, and aelerated breathing are ommon to asthma and the Nijmegen questionnaire. Even some of these somati symptoms ould reasonably be experiened by any person with a worrying illness. It is possible that some of the somati symptoms ould be an expression of an enhaned pereption of dyspnoea. However, when this has been examined in severe asthma the opposite appears to be true. Some patients with severe asthma have a poor pereption of dyspnoea and therefore put themselves at risk. 16 The main emphasis of the Introdutory artile is not to suggest that hyperventilation or dysfuntional breathing is the ause or result of asthma. The suggestion is that the diagnosis of asthma may be inorret in some ases, or at least only part of the problem. This is a perfetly reasonable hypothesis but, in the absene of a speifi and disriminatory test, it is diyult to prove. Can dysfuntional breathing be treated effetively? There is an assumption in the paper by Thomas et al 6 that dysfuntional breathing an be evetively treated after it has been reognised. The evidene for this assumption needs to be explored. In spite of the apparent definition of the ondition by the linial syndrome or the Nijmegen questionnaire, there is little good quality researh to support evetive treatment. There is some limited evidene for the evet of beta blokade and breathing retraining in overt hyperventilation, but very little information about treatment in the ontext of other additional lung disease. If hyperventilation is diretly related to underlying illness suh as asthma or pulmonary embolism, then treatment of the primary ondition may be all that is required. Speifi independent treatment for dyspnoea or hyperventilation should only be neessary if these features annot be ontrolled by primary therapy. COPD is a good example of a ondition where the speifi symptom of dyspnoea an be evetively detahed from the underlying disease. In this ase, exertional dyspnoea an be treated beyond evorts to produe bronhodilation by a number of treatment modalities. Pulmonary rehabilitation modulates dyspnoea predominantly through the vehile of physial exerise training. Additional benefit may be obtained by ognitive behavioural therapy in the ontext of rehabilitation, although this does not appear to work on its own. Some patients with emphysema naturally learn to hange the pattern of breathing by pursing their lips to delay airway losure. For this reason, breathing retraining exerises have been attempted in COPD to try to orret ineyient or unomfortable breathing patterns. 19 Although there is some logi in this approah that may produe short term benefits, the evets do not seem to be sustained. One other approah in COPD is the speifi pharmaologial modifiation of dyspnoea. Several agents inluding opioids and antidepressants have been tried with some suess. In asthma the main aim of pharmaologial treatment is the redution of inflammation and relief of bronhospasm. Speifi isolated treatment of dyspnoea is seldom onduted on the grounds that amelioration of the underlying ondition will suye. However, there are two irumstanes where the treatment of dyspnoea in asthma might be onsidered potentially helpful: (1) when loss of physial fitness resulting from hroni dyspnoea might be a ontributing fator and (2) when the dyspnoea or its pereption has another ause. Individuals with hroni asthma enter the same yle of breathlessness and skeletal musle deonditioning as those with COPD. When this beomes a signifiant omponent of their disability, it may be helped by physial rehabilitation in exatly the same way. The benefits in physial performane and health status for patients with disability from hroni asthma appear to be similar to those ahieved in COPD. 17 Indeed, even in younger patients with asthma who would not regard themselves as disabled, there is a redution in exerise apaity whih is more losely related to lak of fitness than a ventilatory limit. 20 *ii33

4 Morgan Learning points ii34 * Dysfuntional breathing (also known as hyperventilation syndrome and disproportionate breathlessness) is a syndrome of exessive breathlessness and other variable somati symptoms that is sometimes but not always assoiated with demonstrable hypoapnia Dysfuntional breathing may make an important ontribution to symptoms in asthma and result in overpresription of drug treatment There is no aepted gold standard of diagnosis of dysfuntional breathing beyond the linial desription, but the Nijmegen questionnaire is a symptom heklist that an be used to disriminate dysfuntional breathers from normal individuals No tehnique has yet been validated to identify dysfuntional breathing in the presene of other respiratory disease suh as asthma Estimates of the frequeny of dysfuntional breathing in subjets with asthma must therefore remain speulative until there is an agreed definition of the ondition and validation of the investigative instrument To date the evidene of benefit for the treatment of dysfuntional breathing in the ontext of asthma remains santy The hypothesis of Thomas et al 6 is that some patients with asthma may exhibit hyperventilation or dysfuntional breathing whih may ompound or be onfused with the original ondition. It is possible that these symptoms may be helped by onventional rehabilitation and there is some evidene that this is evetive even in mild ases of asthma. 21 The opportunity for speifi treatment therefore lies in some form of physiotherapy or ognitive behavioural therapy that is aimed at breathing retraining or relaxation tehniques. In spite of the hopeful tone of the paper, there is very little evidene for benefit of any tehnique in the speifi management of hyperventilation or dysfuntional breathing in the ontext of asthma. The most well known breathing retraining tehnique is the Buteyko programme. The speifi details of this tehnique are unfortunately subjet to ommerial sensitivities but involve some form of ontrol of hyperventilation and mouth breathing. So far one published randomised ontrolled trial would suggest that, in some ases at least, there is a redution in beta agonist use and hyperventilation. 22 However, no evet was noted on onventional measures of airway alibre or inhaled steroid use. It is highly likely that this subjet will be further explored in future publiations. The use of traditional physiotherapy retraining and disease eduation has been explored by Thomas et al but so far only published as an abstrat. 23 In this situation enhaned physiotherapy appears to have improved a disease speifi health status measure (Asthma Quality of Life questionnaire) as well as the Nijmegen questionnaire for a period of up to 6 months. The full results of this study will be awaited with some interest. Conlusions It would be an attrative hypothesis to suggest that some patients with asthma had an alternative or additional explanation for their symptoms. Firstly, it overs an opportunity for treatment beyond pharmaology and, seondly, an opportunity to redue unneessary mediation. Every liniian reognises the oasional patient with obvious hyperventilation, often in assoiation with asthma, where alternative treatment might be of value. The sale of the problem has never been previously desribed but to suggest that one third of women with asthma may have a omponent of dysfuntional breathing is quite signifiant. If this is true, it will have a onsiderable impat on the provision of servies and move treatment away from reliane on pharmaotherapy alone. However, before there is a wholesale hange in pratie, these results need to be verified. This is not just a matter of repeating the study or examining a wider population. In this ase the definition of dysfuntional breathing and the methods for identifying it require areful validation. The appliation of the Nijmegen questionnaire in the ontext of asthma or other lung disease needs further exploration. Even if the results of this study are reprodued, there is still an outstanding diyulty. One of the guiding priniples of any sreening programme is that evetive treatment should be available to those unovered by the proess. In this ase it is far from lear whether there is yet good evidene for evetive treatment of dysfuntional breathing. The authors have been brave enough to put their toes in the water but further immersion will need areful onsideration. Referenes 1 Gardner WN. The pathophysiology of hyperventilation disorders. Chest 1996;109: Howell JB. The hyperventilation syndrome: a syndrome under threat? Thorax 1997;52(Suppl 3):S Lewis RA, Howell JB. Definition of the hyperventilation syndrome. Bull Eur Physiopathol Respir 1986;22: Howell JB. Behavioural breathlessness. Thorax 1990;45: Sont JK, Willems LN, Bel EH, et al. Clinial ontrol and histopathologi outome of asthma when using airway hyperresponsiveness as an additional guide to long-term treatment. The AMPUL Study Group. Am J Respir Crit Care Med 1999;159: Thomas M, MKinley RK, Freeman E, et al. Prevalene of dysfuntional breathing in patients treated for asthma in primary are: ross setional survey. BMJ 2001;322: Keeley D, Osman L. Dysfuntional breathing and asthma. It is important to tell the differene. BMJ 2001;322: Hornsveld H, Garssen B. Hyperventilation syndrome: an elegant but sientifially untenable onept. Neth J Med 1997;50:13 20.

5 Dysfuntional breathing in asthma 9 Vansteenkiste J, Rohette F, Demedts M. Diagnosti tests of hyperventilation syndrome. Eur Respir J 1991;4: Hornsveld HK, Garssen B, Dop MJ, et al. Double-blind plaebo-ontrolled study of the hyperventilation provoation test and the validity of the hyperventilation syndrome. Lanet 1996;348: Osborne CA, O Connor BJ, Lewis A, et al. Hyperventilation and asymptomati hroni asthma. Thorax 2000;55: van Dixhoorn J, Duivenvoorden HJ. Effiay of Nijmegen Questionnaire in reognition of the hyperventilation syndrome. J Psyhosom Res 1985;29: Demeter SL, Cordaso EM. Hyperventilation syndrome and asthma. Am J Med 1986;81: Langdeau JB, Boulet LP. Prevalene and mehanisms of development of asthma and airway hyperresponsiveness in athletes. Sports Med 2001;31: Davis MS, Freed AN. Repeated hyperventilation auses peripheral airways inflammation, hyperreativity, and impaired bronhodilation in dogs. Am J Respir Crit Care Med 2001;164: Veen JC, Smits HH, Ravensberg AJ, et al. Impaired pereption of dyspnea in patients with severe asthma. Relation to sputum eosinophils. Am J Respir Crit Care Med 1998;158: Morgan MD, Singh S, Calverley PM, et al. British Thorai Soiety statement on pulmonary rehabilitation. Thorax 2001;56: Sassi-Dambron DE, Eakin EG, Ries AL, et al. Treatment of dyspnea in COPD. A ontrolled linial trial of dyspnea management strategies. Chest 1995;107: Cahalin LP, Braga M, Matsuo Y, et al. Effiay of diaphragmati breathing in persons with hroni obstrutive pulmonary disease: a review of the literature. J Cardiopulm Rehabil 2002;22: Garfinkel SK, Kesten S, Chapman KR, et al. Physiologi and nonphysiologi determinants of aerobi fitness in mild to moderate asthma. Am Rev Respir Dis 1992;145: Hallstrand TS, Bates PW, Shoene RB. Aerobi onditioning in mild asthma dereases the hyperpnea of exerise and improves exerise and ventilatory apaity. Chest 2000;118: Bowler SD, Green A, Mithell CA. Buteyko breathing tehniques in asthma: a blinded randomised ontrolled trial. Med J Aust 1998;169: Thomas M, MKinley RK, Freeman E, et al. Breathing retraining for dysfuntional breathing in asthma: a randomised ontrolled trial. Thorax 2001;56(Suppl III):iii17. *ii35