Breathing dysfunction
Relationships between measures of dysfunctional breathing in a population with concerns about their breathing

https://doi.org/10.1016/j.jbmt.2010.06.004Get rights and content

Summary

Background

Dysfunctional breathing (DB) is implicated in physical and psychological health, however evaluation is hampered by lack of rigorous definition and clearly defined measures. Screening tools for DB include biochemical measures such as end-tidal CO2, biomechanical measures such assessments of breathing pattern, breathing symptom questionnaires and tests of breathing function such as breath holding time.

Aim

This study investigates whether screening tools for dysfunctional breathing measure distinct or associated aspects of breathing functionality.

Method

84 self-referred or practitioner-referred individuals with concerns about their breathing were assessed using screening tools proposed to identify DB. Correlations between these measures were determined.

Results

Significant correlations where found within categories of measures however ccorrelations between variables in different categories were generally not significant. No measures were found to correlate with carbon dioxide levels.

Conclusion

DB cannot be simply defined. For practical purposes DB is probably best characterised as a multi-dimensional construct with at least 3 dimensions, biochemical, biomechanical and breathing related symptoms. Comprehensive evaluation of breathing dysfunction should include measures of breathing symptoms, breathing pattern, resting CO2 and also include functional measures such a breath holding time and response of breathing to physical and psychological challenges including stress testing with CO2 monitoring.

Introduction

Dysfunctional Breathing (DB) is commonly used to describe disturbances in breathing functionality that impacts on health (Dixhoorn, 1997, Dixhoorn, 2004, Morgan, 2002, Thomas et al., 2005, Prys-Picard and Niven, 2008, Stanton et al., 2008). The definition of DB however is unclear and no gold standards exist to define it. Dysfunctional breathing includes hyperventilation or breathing in excess of metabolic needs but also refers to breathing pattern abnormalities, poor breathing control and presence of breathing symptoms (Dixhoorn, 1997, Morgan, 2002, Warburton and Jack, 2006). Scientists have until recently focused their attention on hyperventilation, which is defined as breathing in excess of metabolic requirements that results in depletion of carbon dioxide (Comroe, 1974). However the importance of hyperventilation and hypocapnia in producing all symptoms associated with DB is disputed (Hornsveld et al., 1996, Hornsveld and Garsson, 1997). It has been proposed that DB symptoms may arise as a result of non biochemical breathing dysfunctions or have neurological causes (Howell, 1997). A broader definition of dysfunctional breathing, that considers the multiple functions of breathing may be a more useful way to characterise DB and to determine its prevalence and impact (Dixhoorn, 1997).

Maintenance of normal levels of blood gases such as carbon dioxide is an important, if not the key function of breathing; however breathing has other important functions. Breathing functions in posture and motor control (Lewitt, 1980, McGill et al., 1995, Hodges et al., 2001). It is a key influence on oscillating rhythms that are important for homeostasis, autonomic nervous system regulation and efficient interaction between body systems (Giardino et al., 2000, Bernardi, 2001, Song and Lehrer, 2003). Normal and precisely controlled breathing is also important for voice production and regulation of speech (MacLarnon and Hewitt, 1999, Sivasankar and Erickson, 2009).

Biomechanical, neurological and psychological aspects of breathing are not always tightly linked to biochemical parameters and their other relationships are complex and not adequately understood (Han et al., 1996a, Han et al., 1996b). Attempts to tie the symptoms associated with dysfunctional breathing to only the biochemical dimension i.e., hyperventilation and hypocapnia have not been successful (Burton, 1993, Hornsveld and Garsson, 1997). Physical and psychological causes of breathing dysfunction are often interwoven and can be difficult to separate, however DB is thought to contribute to additional symptoms not adequately explained by the main presenting complaint (Han, Zhu et al., 2004). Research has shown that symptoms associated with dysfunctional breathing are strongly influenced by anxiety and other emotional states and in some cases the psychological influences are primary (Wientjes and Grossman, 1994, Han et al., 2004). Other symptoms, particularly various qualities of dyspnea have been linked to breathing pattern abnormalities and poor neuromechanical coupling during breathing (O’Donnell, 2006, Prys-Picard and Niven, 2008). Muscular skeletal dysfunctions, speech and voice problems appear to be predominately linked with dysfunctions of breathing pattern and neural control of respiration rather than to the body’s carbon dioxide status (McGill et al., 1995, Gandevia et al., 2002).

The accumulation of studies showing the presence of breathing disturbances in highly symptomatic patients and results of research showing that patients with a range of symptoms and medical conditions improve after breathing therapy (Lum, 1975, Grossman et al., 1984; Tweedale, Rowbottom et al., 1994; Han, Stegen, et al., 1996a; Meuret, Rosenfield et al., 2009) lends weight to the importance of assessment and optimisation of breathing functionality in patient care. However, evaluation of dysfunctional breathing is currently hampered by lack of clear measurement guidelines. Measures used by practitioners as screening tools to identify dysfunctional breathing include biochemical measures such as end-tidal CO2 (Hardonk and Beumer, 1979, McLaughlin, 2009), biomechanical measures such assessments of breathing pattern (Prys-Picard, Kellett et al., 2004), breathing symptom questionnaires (Thomas et al., 2005, Courtney and Greenwood, 2009) and tests of breathing function such as breath holding time (Courtney and Cohen, 2008).

The Nijmegen is the most commonly used questionnaire used to identify DB. (see Figure 1). The 16 item NQ was originally devised to test for HVS and includes 4 questions on respiratory symptoms and the other 12 items on peripheral and central neurovascular or general tension related symptoms (Dixhoorn and Duivenvoorden, 1985a,b). A questionnaire called the Self Evaluation of Breathing Questionnaire (SEBQ) has also been devised to specifically assess respiratory symptoms and breathing behaviours reported to be associated with DB. The SEBQ includes a larger number of respiratory items than the NQ and can differentiate 2 distinct dimensions of breathing discomfort “lack of air” probably related to chemoreceptor derived sensations and “perception of inappropriate or restricted breathing” probably related to the biomechanics of breathing and breathing perception (Courtney and Greenwood, 2009). Normal values for the SEBQ have not been established as yet and this questionnaire is only useful at present for assessing change in breathing symptoms in individuals after treatment. However, it does have potential as a screening tool for DB once further studies are done to validate this instrument. Normal values for the NQ in European studies are generally around a sum score of 10 (Han et al., 1996a, Han et al., 1996b, Han et al., 1998, Thomas et al., 2005) whereas in China values are lower and average around 5 (Han, Zhu et al., 2004). In categorising individuals as having DB, cut-offs of both 20 and 22 have been found useful (Doorn et al., 1982, Dixhoorn and Duivenvoorden, 1985a, Dixhoorn and Hoefman, 1985b).

Clinicians usually assess breathing pattern using observation and palpation and historically have used a range of techniques most of which have not been validated (Pryor and Prasad, 2002, Perri, 2007). One component of breathing pattern that is considered dysfunctional is chronic thoracic dominant breathing at rest. Recently a technique called the Manual Assessment of Respiratory Motion (MARM), which can quantify extent of thoracic dominant breathing as well as other aspects of breathing pattern, has been found to have high levels of inter-examiner reliability and to agree with measures made simultaneously with Respiratory Induction Plethysmography (Courtney, van Dixhoorn et al., 2008). Normal healthy individuals appear to have balanced breathing with relatively equal motion of upper rib cage to lower rib cage abdominal motion. Perfectly balanced breathing gives a MARM value of 0. Normal values for the MARM in this study of 12 yoga teachers and breathing therapy practitioners were around 6. MARM values above 30 can be considered dysfunctional, as they are at least 2 standard deviations above the mean values found in normal healthy individuals (Courtney, van Dixhoorn et al., 2008). Another aspect of breathing pattern considered dysfunctional is the presence of paradoxical or asynchronous breathing (Prys-Picard, Kellett et al., 2004). In paradoxical breathing the belly is drawn in and lower rib cage narrows rather than expands during inhalation. Practitioners generally assess presence of paradoxical breathing simply by asking the patient to breathe in gently, slightly deeply and into the belly while they observe the respiratory phase relationship of chest and belly motion. If the belly moves inward, decreasing its dimensions during inhalation, the breathing is considered to be paradoxical. This simple observation by the practitioner of chest and belly motion sometimes called the Hi Lo breathing assessment has been found to be reasonably accurate for determining different types of simulated breathing patterns including paradoxical breathing (Courtney and Reece, 2009).

Persistent low levels of resting carbon dioxide might be expected in individuals with dysfunctional breathing as evidenced by chronic persistent hyperventilation. However there is considerable argument about what parameters constitute normal values of resting CO2 and the usefulness of resting CO2 as a means of identifying individuals with hyperventilation tendencies because the tendency to symptom producing hyperventilation can be intermittent rather than chronic and only become apparent in response to physical or psychological challenge testing (Hardonk and Beumer, 1979, Warburton and Jack, 2006). Some older texts state that levels of carbon dioxide below 37 mmHg indicate hyperventilation (Comroe, 1974) and more recent texts place normal CO2 levels as above 35 mmHg (Levitsky, 1995). Gardener found that many individuals had CO2 levels chronically below 35 mmHg with no apparent symptoms until levels were taken below 30 mmHg (Gardner, 1995). In fact he found that mean levels of CO2 in healthy individuals were around 36.2 mmHg with 2 standard deviations below this level being 32.2 (Gardner, 1995). Regardless of arguments over CO2 cut-offs it can be concluded that persistently low CO2 and low CO2 in response to challenge testing is an aspect of dysfunctional breathing worthy of measurement, particularly as end-tidal CO2 which fairly accurately represents arterial CO2 can be easily measured with modern capnometry equipment (McLaughlin, 2009).

Breath holding ability is an aspect of breathing functionality that is commonly disturbed in individuals with tendencies to hyperventilation and to dysfunctional breathing (Jack et al., 1998, Warburton and Jack, 2006). Breath holding time in individuals with chronic idiopathic hyperventilation has been reported to be only 20 s, when held at the end of inhalation, in comparison to normal individuals whose breath holding time is around 60 s when performed according to the same instructions (Jack, Rossiter et al., 2004). Breath holding time differs markedly depending on how it is performed, being affected by whether the hold occurs after inhalation or exhalation and by the size of the breath taken at the beginning of the breath hold (Mithoefer, 1965). One breath holding time protocol, which uses a somewhat standardized procedure and is used for evaluating and monitoring dysfunctional breathing, is the Buteyko Method technique of the Control Pause. The Control Pause is a post expiratory breath hold and is performed with 2 slight variations. In one variation the breath is held until the first urge to breathe and in another variation until the first involuntary motion of the respiratory muscles (Courtney and Cohen, 2008). Control Pause levels of below 20 are proposed to indicate the presence of DB and to correlate with resting carbon dioxide levels (Stalmatski, 1999, Stark and Stark, 2002).

Little research has systematically investigated the relationships between biochemical, biomechanical and symptomatic measures of dysfunctional breathing commonly used by clinicians and therapists to evaluate their patients. The small amount of research that does exist tends to suggest that biochemical, symptomatic and breathing pattern aspects of breathing dysfunction do not necessarily co-exist in individuals suspected of having some type of DB. For example disturbances in breathing pattern are not always associated with chronically decreased baseline levels of CO2 (Han et al., 1996a, Han et al., 1996b, Pine et al., 1998, Caldirola, 2004) and changes in breathing pattern and symptoms after breathing therapy may not be accompanied by changes in CO2 (Han, Stegen, et al., 1996a). In a recent study Meuret found that changes in CO2 mediate symptoms in patients with panic disorder (Meuret, Rosenfield et al., 2009), however other studies have found that general symptoms believed to be characteristic of dysfunctional breathing may be only mildly related to chronic CO2 levels or even acute changes in CO2 (Burton, 1993, Hornsveld and Garsson, 1997). It has been previously reported that Buteyko’s Control Pause does not correlate with resting CO2 levels (Courtney and Cohen, 2008). These observations could imply that breathing pattern, symptoms and carbon dioxide levels reflect distinct aspects of breathing functionality and that dysfunctional breathing might best be characterised as a complex condition with multiple dimensions.

While a combination of measurement tools is sometimes used to evaluate dysfunctional breathing and establish its prevalence in particular populations (Stanton, Vaughn et al., 2008), researchers and practitioners of different breathing therapies may for historical or convenience reasons evaluate only one aspect of breathing function. A number of studies have determined prevalence of DB on the basis of symptom questionnaires alone (Thomas et al., 2001, Humphriss et al., 2004, Thomas et al., 2005) or have emphasized the measurement of carbon dioxide levels (McLaughlin, 2009) or breathing pattern (Perri and Halford, 2004). If measurement in one dimension of breathing functionality proves not to be highly correlated with measurement in other dimension, this may result in incorrect assumptions about prevalence. Research into the relationships between measure of DB will help to clarify what range of measures are needed for comprehensive evaluation of DB. This type of research would help to determine what minimum requirements are needed for comprehensive evaluation of the various aspects breathing dysfunction in the clinical environment. It would also assist understanding of how best to characterise DB.

Aims of this study

  • 1.

    To compare prevalence of DB on the basis of a range of measures

  • 2.

    To evaluate relationships and correlations between various measures of DB

Given the difficulties and lack of consensus on gold standard definitions of DB, we have chosen a pragmatic study design and applied a range of clinically used measures of DB to a population with concerns about their breathing rather than attempting to test a population fitting any particular definition of DB. We believe this approach has greater external validity and applicability for practitioners as it more closely mimics what occurs in “real life” clinical situations.

Section snippets

Participants

Participants were recruited from general practices and complementary medicine clinics in Sydney, Australia. Flyers and brochures were placed in waiting rooms and practitioners received a letter about the study that was described as an investigation into the measurement of “incorrect” and dysfunctional breathing and that it presented an opportunity for individuals to have their breathing assessed.

This attracted individuals with a range of mild medical conditions who had concerns or curiosity

Results

As can be seen in Table 1, the mean levels for all dysfunctional breathing measures are not particularly high for this group, however means for the NQ, (18) and the MARM (19) are higher and more dysfunctional than those found in studies of normal individuals. In individuals with normal breathing, mean values for sum scores of the NQ are around 10 (Han, Stegen, et al., 1996a; Han et al., 1998, Thomas et al., 2001) and for the MARM are around 6 (Courtney, van Dixhoorn et al., 2008).

The majority

Discussion

Strict definitions of DB are not possible at present, but for practical purposes it is probably most useful not to think of DB as single entity limited to the biochemical dimension of breathing functionality (as in Hyperventilation Syndrome) but to consider breathing symptoms and breathing pattern as potentially separate aspects of DB which need to be measured in their own right. This study found that significant correlations exist only within categories of breathing measures but not between

Limitations of this study

Not all ways of measuring DB were assessed in this study. For example measurement of carbon dioxide at rest, as was undertaken during this study, will only reveal chronic persistent hyperventilation and may not be adequate for revealing which individuals are prone to intermittent hyperventilation in response to physical or psychological stress (Hardonk and Beumer, 1979).

The interpretation of these results is limited to their use as general screening tools, as the patient population used in this

Future research

Further Research is required to investigate the presence of more complex relationships between the various dimensions of breathing dysfunction. Research should also target individuals with stronger evidence of breathing dysfunction or with specific ailments.

Conclusion

For practical purposes DB is probably best characterised as multi-dimensional. DB can occur in at least 3 dimensions: biochemical, breathing pattern and breathing related symptoms and these might not co-exist. Screening for DB with measures representing only one of these dimension may not lead to realistic estimations of the prevalence and impact of the various types of breathing dysfunctions. Comprehensive evaluation of breathing dysfunction should include measures of breathing symptoms,

Conflict of interest statement

None of the authors of this manuscript shall derive any personal profit or gain, directly or indirectly, by reason of his or her authorship of this manuscript.

Acknowledgments

We would like to acknowledge the Australian Osteopathic Association for funding and administrative assistance.

References (71)

  • A. Meuret et al.

    Changes in respiration mediate changes in fear of bodily sensation in panic disorder

    Journal of Psychiatric Research

    (2009)
  • M.A. Perri et al.

    Pain and faulty breathing: a pilot study

    Journal of Bodywork and Movement Therapies

    (2004)
  • C. Prys-Picard et al.

    Respiratory inductance plethysmography for the diagnosis of dysfunctional breathing in patients with severe asthma

    Journal of Allergy and Clinical Immunology

    (2004)
  • J.G. Teeter et al.

    Relationship between airway obstruction and respiratory symptoms in adult asthmatics

    Chest

    (1998)
  • M. Thomas et al.

    The presence of dysfunctional breathing in adults with and without asthma

    Primary Care Respiratory Journal

    (2005)
  • M. Altose

    Respiratory muscles and dyspnea

    Seminars in Respiratory Medicine

    (1992)
  • C.D. Burton

    Hyperventilation in patients with recurrent functional symptoms

    British Journal of General Practice

    (1993)
  • K.P. Buteyko

    Experience of Application in Medical Practice

    (1990)
  • D. Caldirola

    Approximate entropy of respiratory patterns

    American Journal of Psychiatry

    (2004)
  • J.H. Comroe

    Physiology of Respiration

    (1974)
  • R. Courtney et al.

    Investigating the claims of Konstantin Buteyko M.D., PhD: the relationship of breath holding time to end tidal CO2 and other proposed measures of dysfunctional breathing

    Journal of Alternative and Complementary Medicine

    (2008)
  • R. Courtney et al.

    Evaluation of breathing pattern: comparison of a Manual Assessment of Respiratory Motion (MARM) and respiratory induction plethysmography

    Applied Psychophysiology and Biofeedback

    (2008)
  • J. Dixhoorn et al.

    Efficacy of Nijmegen Questionnaire in recognition of the hyperventilation syndrome

    Journal of Psychsomatic Research

    (1985)
  • J. Dixhoorn et al.

    Hyperventilatieklachten in de fysiotherapiepraktijk

    Ned T Fysiotherapie

    (1985)
  • J.J.v Dixhoorn

    De Nijmeegse vragenlijst in de evaluatie van adem- en ontspanningstherapie

  • J.v Dixhoorn

    Hyperventilation and dysfunctional breathing

    Biological Psychology

    (1997)
  • J.v Dixhoorn

    A method for assessment of one dimension of dysfunctional breathing:distribution of breathing movement

    Biological Psychology

    (2004)
  • P. Doorn et al.

    Een vragenlijst voor hyperventilatieklachten

    De Psycholoog

    (1982)
  • H. Folgering et al.

    Some anomalies in the control of PACO2 in patients with a hyperventilation syndrome

    Bulletin European Pathophysiology of Respiration

    (1978)
  • S.C. Gandevia et al.

    Balancing acts: respiratory sensations, motor control and human posture

    Clinical and Experimental Pharmacology and Physiology

    (2002)
  • W. Gardner

    The pathophysiology of hyperventilation disorders

    Chest

    (1995)
  • N.D. Giardino et al.

    The role of oscillations in self-regulation

  • P. Grossman et al.

    A controlled study of a breathing therapy for treatment of hyperventilation syndrome

    Journal of Pyschomatic Research

    (1984)
  • J. Han et al.

    Unsteadiness of breathing in patients with hyperventilation syndrome & anxiety disorders

    European Respiratory Journal

    (1996)
  • J. Han et al.

    Medically unexplained dyspnea, psychological characteristics and role of breathing therapy

    Chinese Medical Journal

    (2004)
  • Cited by (47)

    • The Thai version of the Nijmegen questionnaire

      2022, Heliyon
      Citation Excerpt :

      Many studies have reported that it is a valid and reliable tool which can help identify HVS quickly and easily [12, 13, 14]. Besides a screening tool, the NQ also has been used to assess clinical and research interventions in patients with HVS [15, 16, 17]. It was translated into several languages including Korean, Greek, and Iranian.

    • Effects of core stability training program on trunk muscle endurance in deaf children: A preliminary study

      2021, Journal of Bodywork and Movement Therapies
      Citation Excerpt :

      The results of this study indicated that core muscle endurance increased following the training program in deaf children and correlated to healthy spinal cords. Human respiratory system has an essential function and is determined by many biomechanical, biochemical and psychological factors (Courtney et al., 2011). Trunk muscles produced establishing biomechanical conditions in compliance with respiratory mechanism.

    • Respiratory therapy for the treatment of anxiety: Meta-analytic review and regression

      2021, Clinical Psychology Review
      Citation Excerpt :

      Very large, but non-significant, effects were found for panic symptoms in a likely under-powered analysis with just six studies reporting panic outcomes. Observed effect sizes for generalized anxiety and panic symptoms are not surprising given that somatic symptoms, in particular cardiopulmonary and musculoskeletal symptoms, are prominent in generalized anxiety and panic (Bekhuis, Boschloo, Rosmalen, & Schoevers, 2015), and subtypes of individuals with panic are in fact characterized by respiratory dysfunction (Boulding et al., 2016; Courtney, 2017; Courtney, Greenwood, & Cohen, 2011; Courtney, van Dixhoorn, et al., 2011). It is possible that respiratory interventions reduce arousal via correcting breathing abnormalities, thereby preventing the onset of panic distress.

    • Comparison of respiratory parameters in participants with and without chronic low back pain

      2019, Journal of Bodywork and Movement Therapies
      Citation Excerpt :

      Future research focussing on other respiratory parameters such as chest kinematics, breath holding time, psychological and postural deviations may be considered to evaluate respiratory dysfunctions associated with CLBP. ( Courtney et al., 2011). The effect of other inspiratory and lumbar paraspinal muscles on respiratory function in CLBP should also be studied.

    View all citing articles on Scopus
    View full text