ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1300

Exercise-induced laryngealobstruction

Prevalence, laryngeal findings and evaluation oftreatment

KATARINA NORLANDER

ISSN 1651-6206ISBN 978-91-554-9819-1urn:nbn:se:uu:diva-315123

Dissertation presented at Uppsala University to be publicly examined in Skoogsalen,Akademiska sjukhuset, Uppsala, Friday, 31 March 2017 at 09:00 for the degree of Doctor ofPhilosophy (Faculty of Medicine). The examination will be conducted in Swedish. Facultyexaminer: Docent Riitta Möller (Institutionen för medicinsk epidemiologi och biostatistik,Karolinska institutet).

AbstractNorlander, K. 2017. Exercise-induced laryngeal obstruction. Prevalence, laryngeal findingsand evaluation of treatment. Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1300. 61 pp. Uppsala: Acta Universitatis Upsaliensis.ISBN 978-91-554-9819-1.

Exercise-induced laryngeal obstruction (EILO) is one of many different causes for adolescentsto experience dyspnoea during exercise. Objective exercise-testing with continuous videolaryngoscopy is crucial for a correct diagnosis since it is difficult to differentiate EILO fromother exercise related conditions in the airways only on the symptomatology. The main symptomin EILO is inspiratory stridor arising from an obstruction at the laryngeal level during ongoingexercise which quickly resolves after the exercise has stopped. EILO is often misdiagnosed asexercise-induced bronchoconstriction (EIB), which is obstruction in the peripheral airways thattypically arises after cessation of exercise.

From a previous survey investigating self-reported exercise-induced dyspnoea in all 12-13-year-old adolescents in Uppsala (n=3,838, response rate 60.2%) a subset of 150 randomlyselected adolescents (103 with dyspnoea and 47 controls) performed standardized treadmillexercise-tests for EIB and EILO.

During the exercise-test for EIB the subjects breathed dry air according to the currentrecommended guidelines. EIB was defined as a decrease in FEV1 ≥10% from baseline. EILOwas diagnosed during a continuous laryngoscopy exercise (CLE) test by use of the CLE-score method and was defined as an obstruction of grade 2 at either glottic or/and supraglotticlaryngeal level. The estimated prevalence of EIB in the general population was 19.2% and theestimated prevalence of EILO was 5.7%. No gender differences were detected.

A diagnostic software program for EILO, EILOMEA, was compared with the CLE-scoreand the methods were found to be compatible. EILOMEA was used to map and comparelaryngeal response patterns in adolescents with exercise-induced dyspnoea (EIB and/or EILO),in adolescents with dyspnoea but without a diagnosis of EIB or EILO, and in healthy controls,all of whom had performed the CLE-test. No differences were seen between the healthy controlsand the adolescents with dyspnoea without a diagnosis. Only adolescents diagnosed with EILOshowed a significant different laryngeal response pattern which strongly suggests that thediagnostic procedure is reliable.

In a follow-up study of patients referred for investigation of exercise-induced dyspnoea, weinvestigated the outcome of surgical vs. conservative treatment of EILO-positive subjects andsubjects tested negative for the diagnosis, regarding the level of exercise-induced dyspnoeaand physical activity. Surgically treated patients had less breathing problems and were morephysically active than both conservatively treated patients and patients who were testednegative.

Keywords: Exercise, laryngeal obstruction, bronchoconstriction, CLE-score, EILOMEA

Katarina Norlander, Department of Surgical Sciences, Otolaryngology and Head and NeckSurgery, Akademiska sjukhuset, Uppsala University, SE-75185 Uppsala, Sweden.

© Katarina Norlander 2017

ISSN 1651-6206ISBN 978-91-554-9819-1urn:nbn:se:uu:diva-315123 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-315123)

To my family ___

128√e980 = ;) 2

Cover picture, photos and illustrations are made by the author and co-workers.

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Norlander K, Christensen PM, Maat RC, Halvorsen T, Heimdal

JH, Morén S, Rasmussen N, Nordang L. Comparison between two assessment methods for exercise-induced laryngeal obstruc-tions. Eur Arch Otorhinolaryngol. 2016;273(2):425-30.

II Johansson H*, Norlander K*, Berglund L, Janson C, Mali-novschi A, Nordvall L, Nordang L, Emtner M. Prevalence of exercise-induced bronchoconstriction and exercise-induced laryngeal obstruction in a general adolescent population. Thorax. 2015;70(1):57-63. * Shared first authorship

III Norlander K, Johansson H, Janson C, Malinovschi A, Nordvall L, Emtner M, Nordang L. Laryngeal findings during exercise in dyspneic and non-dyspneic adolescents. Manuscript.

IV Norlander K, Johansson H, Janson C, Nordvall L, Nordang L. Surgical treatment is effective in severe cases of exercise-in-duced laryngeal obstruction. A follow-up study. Acta Otolaryn-gol. 2015;135(11):1152-9

Reprints were made with permission from the respective publishers.

Additional publications during my PhD studies which are not included in this thesis

Johansson H, Norlander K, Hedenström H, Janson C, Nordang L, Nordvall L, Emtner M. Exercise-induced dyspnea is a problem among the general adolescent population. Respir Med. 2014;108(6):852-8 Johansson H, Norlander K, Alving K, Hedenström H, Janson C, Malinovschi A, Nordang L, Emtner M. Exercise test using dry air in random adolescents: temporal profile and predictors of bronchoconstriction. Respiology. 2016;21(2):289-96

Johansson H, Norlander K, Janson C, Malinovschi A, Nordang L, Emtner M. The relationship between exercise induced bron-chial obstruction and health related quality of life in female and male adolescents from a general population. BMC Pulm Med. 2016;(1):63.

Contents

Introduction ................................................................................................... 11 

Background ................................................................................................... 12 Exercise-induced laryngeal obstruction, EILO .................................... 12 Exercise-induced bronchoconstriction, EIB ........................................ 12 

Brief historical summary of the diagnosis of laryngeal obstruction ......... 13 The complexity of the dyspnoea-problem ................................................ 14 

Prevalence of exercise-induced dyspnoea ........................................... 14 Prevalence of EIB ................................................................................ 14 Prevalence of EILO ............................................................................. 14 The importance of correctly diagnosing exercise-induced dyspnoea and its different causes ......................................................................... 15 

The airways .............................................................................................. 16 Anatomy of the larynx ......................................................................... 16 Airway physiology at rest .................................................................... 17 Normal laryngeal function at rest ........................................................ 17 Changes in lower airway physiology during increasing exercise ........ 18 Changes in the larynx during exercise ................................................. 18 

Clinical features of EIB and EILO ........................................................... 19 EIB ....................................................................................................... 19 EILO .................................................................................................... 19 

Diagnostic methods for EIB and EILO .................................................... 19 EIB ....................................................................................................... 19 EILO .................................................................................................... 20 

Aetiological aspects of EILO ................................................................... 22 Treatment for EILO .................................................................................. 23 

Glottic EILO ........................................................................................ 23 Supraglottic EILO ................................................................................ 23 

Aims of present studies ................................................................................. 24 

Methods ........................................................................................................ 25 Ethics ........................................................................................................ 25 

Diagnostic procedures .............................................................................. 26 EILOMEA (paper I and III) ................................................................. 26 EIB-test (paper II) ................................................................................ 27 CLE-test (paper II) ............................................................................... 27 CLE-score method (paper I-IV) ........................................................... 28 

Study populations/Material and Design ................................................... 28 Statistical analyses .................................................................................... 34 

Summary of results ....................................................................................... 35 Comparison of CLE-score and EILOMEA (paper I) ............................... 35 Prevalence of EIB and EILO (paper II) .................................................... 36 Laryngeal findings during exercise (paper III) ......................................... 38 

Laryngeal findings assessed with CLE-score ...................................... 38 Laryngeal findings assessed with EILOMEA ..................................... 38 

Surgical vs conservative treatment for EILO in comparison with patients tested negative (paper IV) ........................................................... 38 

Impact on exercise-induced dyspnoea ................................................. 39 Impact on physical activity .................................................................. 39 

Discussion ..................................................................................................... 41 CLE-score vs EILOMEA ......................................................................... 41 Prevalence of EIB and EILO .................................................................... 42 

Our findings compared to others ......................................................... 42 Gender differences ............................................................................... 42 

Can we trust the prevalence numbers? ..................................................... 43 Factors in study II that might have influenced the prevalence numbers ............................................................................................... 44 

Larynx during exercise in adolescents with and without dyspnoea ......... 47 Surgical vs conservative treatment ........................................................... 47 

Who should be offered surgical treatment? ......................................... 48 Future perspectives on EILO .................................................................... 48 

Conclusion .................................................................................................... 50 

Populärvetenskaplig sammanfattning ........................................................... 51 

Acknowledgements ....................................................................................... 53 

References ..................................................................................................... 55 

Abbreviations

CI Confidence interval CLE-test Continuous laryngoscopy exercise test CT Conservatively treated EIA Exercise-induced asthma EIB Exercise-induced bronchoconstriction EILO Exercise-induced laryngeal obstruction EVH Eucapnic voluntary hyperventilation FEV1 Forced expiratory volume in one second PCA Posterior cricoarytenoid muscles ST Surgically treated VCD Vocal cord dysfunction

11

Introduction

When I started to work in the Ear-, Nose and Throat department in Uppsala in 2006 the first investigations of a new, somewhat mysterious group of patients, often young women, had just begun. These patients had been referred to our clinic for breathing difficulties during exercise when previous investigations had not led to any diagnosis, or prescribed asthma medication therapy had not resulted in expected symptom relief. That same year, 2006, Heimdal et al. published a paper describing the continuous laryngoscopy exercise test. It en-ables visualization of the larynx through flexible transnasal fibre-optic laryn-goscopy so that we can actually see what happens in the larynx during ongoing exercise (1). Often observed in the patients was an obstruction of the laryngeal inlet during exercise that disappeared shortly after cessation, a condition now known as exercise-induced laryngeal obstruction, EILO.

During the last decade, the recognition and knowledge about EILO has in-creased but there is still very much we don’t know. As an extension of a pre-vious study on the prevalence of exercise-induced dyspnoea in a general ado-lescent population, we investigated the prevalence of EILO as well as the prevalence of exercise-induced bronchoconstriction (EIB), by letting adoles-cents perform standardized exercise tests. Further, two different methods to diagnose EILO were compared and subsequently used to assess laryngeal re-sponse patterns during exercise in adolescents with and without dyspnoea. Fi-nally, the outcome of surgical treatment of patients referred to our clinic has been analysed.

The studies in this thesis is an effort to add further scientific knowledge about EILO. I hope that it will inspire to further exploring of this, in many ways, young research field.

12

Background

Inducible laryngeal obstruction is a new consensus term which includes dis-orders known in the literature as vocal cord dysfunction, paradoxical vocal fold movement, induced laryngomalacia and many others. The consensus term underlines that an external stimulus is involved (2) and with the use of a prefix, a specific inducer to the laryngeal obstruction is pointed out. Hence, when exercise is the inducer the condition is called exercise-induced laryngeal ob-struction.

Exercise-induced laryngeal obstruction, EILO Laryngeal obstruction during exercise is not one, single diagnosis. Instead, EILO consist of a group of conditions in which airflow is hampered at the laryngeal level only during exercise, never at rest. The obstruction can be ob-served by performing continuous laryngoscopy during exercise (1). There are two main subtypes of EILO. The glottic type involves abnormal adduction or paradoxical movements of the vocal cords. The supraglottic type comprises an inward collapse of anatomical structures above the vocal cords; aryepiglot-tic folds, cuneiform tubercles and sometimes the epiglottis (3). The supraglot-tic type, often combined with a secondary glottic obstruction, is the most com-mon type (4) (figure 1). This is in contrary to what has long been expressed in the literature, where the glottic type has been considered more prevalent and often been referred to as vocal cord dysfunction (VCD) (5, 6).

Exercise-induced bronchoconstriction, EIB In the literature, dyspnoea during exercise is often referred to as exercise-in-duced asthma (EIA). In this thesis, we have investigated exercise-induced bronchoconstriction (EIB). Distinguishing between EIA and EIB is a complex task. When exercise triggers bronchoconstriction and exacerbation of asthma symptoms in a patient with asthma, it is called EIA. EIB on the other hand, can be diagnosed if the patient has no evidence of asthma except during or after exercise (7). This thesis will focus on EIB in the context used by Weiler et al. as a transient narrowing of the lower airways following exercise in the presence or absence of clinically recognized asthma (8).

13

Figure 1. Figure 1. Larynx during exercice in a healthy adolescent and an adolescent with EILO with obstruction grade 2 at both glottic and supraglottic level.

Brief historical summary of the diagnosis of laryngeal obstruction The term “inducible laryngeal obstruction” was formally accepted only a few years ago (1), but inspiratory wheeze as a possible symptom of obstruction has been described and discussed for more than a century. Many different terms have been presented trying to explain the trigger or the mechanism be-hind the obstruction, but for a long time it was considered to be of psycholog-ical origin only. Dunglison reported hysterical croup in women 1842 (9), Mac-Kenzie described paradoxical vocal cord movements in hysterical adults 1869 (10), and in the beginning of the twentieth century, William Osler wrote about spasm of the laryngeal muscles due to hysteria (11). MacKenzie was first to describe direct laryngoscopy as an examination method (10). With the devel-opment of the fibre endoscopic technique, starting with Shigeto Ikeda in 1966, the larynx could easily be studied during symptoms. However, the main theory during the 1970s and 1980s was still that the trigger was psychological (12, 13). In 1983 Christopher et al wrote about vocal cord dysfunction presenting as asthma (5). A supraglottic obstruction instead of a glottic obstruction was first described by Lakin et al in 1984. By letting the subject run on a treadmill before laryngoscopy they identified exercise as a trigger (14). Laryngoscopy during exercise was performed on a stationary bicycle in 1996 (15) and four years later Naito let twelve persons perform various exercises, like playing tennis against a wall, with a fibre-optic laryngoscope inserted in the nose and fixated to an ordinary motorcycle helmet (16). Today we investigate EILO with the continuous laryngoscopy exercise test developed by Heimdal et al (1). The test is described in detail later in this thesis.

14

The complexity of the dyspnoea-problem Prevalence of exercise-induced dyspnoea Exercise-induced dyspnoea is common among children and adolescents (17-19) and a correct diagnosis is a challenge for the clinician (14, 20). EIB and EIA are common causes of exercise-induced dyspnoea and these diagnoses should be investigated (8). EILO must also be considered early in the diag-nostic process since symptoms of EIB/EIA and EILO can be very similar.

Sometimes respiratory symptoms are due to poor physical fitness. In a pro-spective survey Seear et al found poor physical condition among 12 out of 52 children (23%) referred for poorly controlled asthma (21). In a retrospective study, Abu-Hasan et al found about the same percentage (22).

Prevalence of EIB The American Academy of Allergy, Asthma & Immunology Work Group re-port from 2007 states that up to 90% of patients with asthma (varying ages) have EIB (23). The prevalence of asthma also tends to increase globally (18). The prevalence of EIB has been investigated in several studies and ranges be-tween 7-16% depending on examination method and study population (24-28). The diagnosis is however too often based solely on observed symptoms (29) a strategy which has been shown to have poor sensitivity and specificity (27). Perceived dyspnoea symptoms do not correlate well with objective measures (30). Among athletes, the prevalence of EIB has been found to vary a lot in different studies. When eucapnic voluntary hyperventilation (EVH) has been the diagnostic method, numbers between 3-25% have been reported (31, 32). Gender differences towards a predominance of females have been reported in some studies (24, 27).

Prevalence of EILO The prevalence of EILO is not well known. In an unselected cohort of 556 persons in Copenhagen aged 14-24 years, whereof 97 performed an exercise test with simultaneous laryngoscopy, the estimated prevalence in the general population was 7.5% (33). Morris et al. did a study including 40 militaries with exertional dyspnoea and twelve healthy controls. Fifteen percent of the patients had laryngeal findings consistent with vocal cord dysfunction. The assessment, however, was made with laryngoscopy before and after exercise but not during the actual effort (34). EILO prevalence has been found to be as high as 35% among athletes but the study population was highly selected (35).

Most patients referred to our clinic for investigation are young females, median age 16.5 years. Many studies report that EILO is more common in young females than males (4, 33, 36-38).

15

Complicating factors There are several factors that complicate the interpretation of previous studies regarding the prevalence of EILO, in analogy with what was previously de-scribed for EIB. The heterogeneity regarding inclusion criteria, diagnostic methods, diagnostic criteria and terminology used to describe laryngeal find-ings makes it almost impossible to get an overview. Referral bias is a problem since most studies take place in tertiary settings.

It is important to remember that stridor, high-pitched noisy breathing, is a symptom, not a diagnosis. Røksund et al examined 166 patients with continu-ous laryngoscopy during exercise and were not able to link inspiratory stridor to a single causal factor, or to a specific structure in the larynx (4). More than 80 different terms can be found in the literature that describe respiratory prob-lems at the laryngeal level (39). In many studies where the term vocal cord dysfunction has been used, we know today that what is described in many of the cases probably is what we now call EILO.

The importance of correctly diagnosing exercise-induced dyspnoea and its different causes EIB is common among asthmatics, and EIB can coexist with EILO (40). Also, asthma and EILO can coexist (37, 41). In line with earlier described problems dealing with terminology and diagnostic methods, it is difficult to know how common coexistence between asthma or EIB and EILO is. Many studies de-scribe inspiratory stridor and label it as VCD. Sometimes laryngoscopy has been performed during symptoms that have not been triggered by exercise (42), sometimes after exercise has stopped (6) and sometimes laryngoscopy has not been done at all (43). To investigate coexistence, strict and accurate diagnostic procedures must be used since the symptoms, at least for the ob-server, are very similar (5). From the patient´s history alone it is very difficult to tell if it is EILO or EIB since patients seldom can pinpoint if their respira-tory distress originates in the central or the peripheral airways. Most of them just describe “dyspnoea”. Unfortunately, the recommended diagnostic re-quirements are often not met (32).

EIB appears to be over-diagnosed frequently which is very unfortunate (44). A correct diagnosis of asthma, EIB or EILO provides the possibility to offer targeted treatment, but equally important is to cease unnecessary medical treatment for asthma if the person has EILO (45). The symptom similarity of EIB and EILO often leads to the misunderstanding that the patient suffers from “refractory asthma” and the EILO-diagnosis is delayed (46, 47).

Children and adolescents with exercise-induced dyspnoea are reported to have reduced health related quality of life, lower physical activity level, poor sleep and higher school absenteeism (17, 48). Furthermore, participating in sport activities can be important from a social point of view (49).

16

The airways The airways are anatomically divided into the upper, extrathoracic airways and the lower, intrathoracic airways. The larynx forms the boundary between the upper and lower airways and the narrowest part of the upper airways is at the level of the vocal cords.

Anatomy of the larynx The larynx is built up by several cartilages, joints, small muscles and liga-ments. It extends from the tip of the epiglottis to the inferior border of the cricoid cartilage, which is a circular cartilage anchored to the trachea. The larynx can be described as a narrow funnel. The anterior border consists of the thyroid cartilage and the epiglottis. Epiglottis is, as is the whole interior of the larynx, covered by mucosa. From the lateral borders of the epiglottis, the mu-cosa extends posteriorly forming the aryepiglottic folds. The posterior border is in the intraarytenoid notch. The small triangular arytenoid cartilages are placed upon the cricoid plate and regulated by the arytenoid muscles, enabling abduction (via posterior cricoarytenoid muscles) and adduction (via lateral cricoarytenoid muscles) of the vocal cords. Other cartilages in the larynx are the cuneiform and the corniculate cartilages (50) (figure 2).

The laryngeal interior is divided into three levels: the supraglottic, the glot-tic and the subglottic level. This is important to keep in mind for the under-standing of EILO.

Figure 2. Larynx from above.

17

Airway physiology at rest Inside the chest there is normally a negative pressure at rest, i.e. negative in-trathoracic pressure. Inhalation, which is an active process, is initiated by the diaphragm and the external intercostal muscles. When the diaphragm con-tracts it is pulled downwards. Simultaneous contraction of outer intercostal muscles pull the sternum and ribs upward and out. This combined actions cause the thorax to widen and, due to the negative intrathoracic pressure, air will flow in through the nose, and mouth, if it is open. The air is conducted through the pharynx, larynx, trachea and bronchi to the alveoli where gas ex-change occurs. At the end of inhalation, the lungs are filled with air and ex-panded. The elasticity of the lung tissue then causes the air to flow out pas-sively during expiration when the diaphragm and outer intercostal muscles return to their resting position. If there is an obstruction in the lower airways, the passive airflow on expiration will be hampered consequently expiratory symptoms will occur (51, 52).

Normal laryngeal function at rest The larynx is a complex organ with vitally important and highly evolved bio-logical functions. It must serve as an open valve during respiration, as a par-tially closed valve during phonation and as a completely closed valve when it protects the lower respiratory tract against aspiration during swallowing. The vocal cords are controlled both voluntarily and by the autonomic nervous sys-tem, and therefore breathing can be affected through several different mecha-nisms: physiological, neurological and psychological.

During normal quiet breathing the glottis opens more during inhalation than exhalation (53). This movement of the vocal cords is due to a coordinated reflex between the posterior cricoarytenoid muscles (PCA) with other respir-atory muscles, foremost the diaphragm. During inhalation, the PCA muscles are activated and the vocal cords are abducted shortly before the diaphragm starts to contract (53, 54). Due to this mechanism, but also because the vocal cords can be controlled voluntarily, the larynx functions as a valve regulating both pressure and airflow through the glottis. Since the PCA and vocal cords are ahead of other actions, and the glottis usually is narrower during expira-tion, a positive pressure will be maintained in the airways for a longer time during expiration which optimizes gas exchange (53). This neuronal reflex to narrow the glottic opening is augmented during bronchoconstriction (55).

18

Changes in lower airway physiology during increasing exercise Exercise increases the body´s metabolic demands which will require a greater amount of oxygen. This requirement is met by several adaptations that in-crease the minute ventilation. Normal minute ventilation is 6-8 litres per mi-nute at rest. The minute ventilation can increase by taking deeper breaths and/or increase the respiratory rate (56). Normal breathing rate at rest is 12-16 breaths per minute. During exercise, it can be as high as 30-40 breaths per minute and the minute ventilation subsequently increases 20 to 30 times (57). In elite athletes, minute ventilation can be over 200 litres per minute (57). To be able to breathe in more oxygen we change from nasal breathing to mouth breathing when minute ventilation reaches 25-30 litres per minute (58). Inha-lation is also facilitated by activation of accessory respiratory muscles such as the sternomastoid and scalene muscles. Exhalation changes from a passive to an active process by activation of intercostal and abdominal muscles (59).

Changes in the larynx during exercise When the respiratory rate increases, air passes through the larynx with a higher flow rate than at rest. At inhalation during hyperventilation the vocal cords are more abducted than during quiet breathing. During heavy hyperventilation, they are also abducted during exhalation which is beneficial as it reduces the flow resistance (60, 61).

The airflow exerts forces on the laryngeal tissues resulting mainly in a neg-ative transmural pressure gradient which will cause the aryepiglottic folds to move medially during inhalation. These forces will be greater the more the airflow is increased (62). The larynx can be kept open because of its construc-tion of rigid cartilages, strong ligaments and many small intrinsic and extrinsic muscles. At the supraglottic level, the epiglottis tilts anteriorly when these muscles are activated, and this tightens the aryepiglottic folds (15). At the glottic level, the vocal cords will lengthen and abduct due to increased activity in PCA (58, 59). While exercising, the glottic opening is increased mainly posteriorly during exhalation, leading to a reduction in airflow resistance (54).

When the airflow is increased, the relative contribution of upper airway resistance to the total airway resistance is increased (58). Gradually stronger forces acting on the laryngeal tissues result in correspondingly increased movements of the aryepiglottic folds. A mild adduction of the aryepiglottic folds is considered a normal response since it has been observed in healthy subjects (3, 15). At the glottic level, it has been suggested that a 50% narrow-ing of the glottis is needed for it to be pathological (42) but it is not yet thor-oughly investigated. When extrathoracic airflow is hampered, as in laryngeal obstruction, inspiratory symptoms will occur (63).

19

Clinical features of EIB and EILO EIB Symptoms of bronchoconstriction include cough, expiratory wheeze, dysp-noea, chest tightness and increased mucus production. Severity of symptoms depends on intensity and effort and is modified by temperature and humidity in the inhaled air (8, 42, 64). Symptoms evolve shortly after the exercise ses-sion and reaches maximum within 3-15 minutes (65). Symptoms generally last for 30-90 minutes in the absence of treatment (32). At physical examina-tion, expiratory wheeze and a prolonged expirium is noted. Laryngoscopy dur-ing ongoing symptoms shows normal findings or a mild glottic adduction (55, 66).

EILO Clinical examination of the larynx at rest is normal. Symptom onset is typi-cally triggered by exercise requiring high ventilation (1). Respiratory distress can be observed.

In glottic EILO a very sudden onset is typical, due to an abrupt adduction of the vocal cords. Symptoms occur during the exercise. A prolonged inspir-ium is noted, often accompanied by inspiratory stridor of high frequency. Lar-yngoscopy during symptoms shows adducted vocal cords or paradoxical vocal cord motion (3).

In supraglottic EILO symptoms arise quickly, yet to some extent gradually. Symptoms increase with the workload and reach peak level at the end of the exercise session or during the first 2-3 minutes after stopping. It is the symp-toms that forces the person to cease exercising and then symptoms disappear in a few minutes. At examination, a prolonged inspirium is noted, sometimes accompanied by an inspiratory wheeze of low frequency (4, 67). Laryngos-copy in a symptomatic person reveals collapse of supraglottic structures.

Diagnostic methods for EIB and EILO EIB Direct or indirect methods can be used to diagnose EIB. A direct provocation by inhalation of a provoking agent such as methacholine is used for detecting or ruling out bronchial hyperreactivity. A positive test strengthens the suspi-cion of underlying asthma. The recommended indirect method is an exercise test on treadmill or bicycle while the person is breathing dry air (32). The EIB-diagnosis is suggested if a decrease in FEV1 ≥ 10% on spirometry is noted after the test (32, 68). Another indirect method is EVH. It causes the airway mucosa

20

to dry out and creates a hypertonic environment, which is thought to be the aetiology behind EIB (69).

EILO Patients history Patients with EILO describe symptoms like shortness of breath, tightness in the throat and sometimes a sensation of suffocation and panic. Exercise must be recognized as the inducer to the breathing problems. The clinician should focus on distinguishing EILO from EIB and ask key questions about symptom onset, duration and how quickly the breathing problems subside after cessa-tion of exercise. A sudden onset of respiratory distress during exercise in com-bination with a relatively short duration and absence of symptoms at rest points toward an obstruction at the laryngeal level. Whether symptoms occur during inhalation or exhalation is important information but often difficult for the patient to answer. During recent years, several questionnaires have been developed with the aim to facilitate clinical evaluation of dyspnoea and to dif-ferentiate VCD from asthma (70-72).

Exercise test Treadmill is most commonly used for the exercise test but other modalities are possible (73, 74) on condition that the test subject can reach maximum ex-haustion (75). The larynx is filmed continuously and the dynamic develop-ment of obstruction is observed. If breathing is hampered during inhalation and/or exhalation and how long it takes before the larynx normalizes after ex-ercise has stopped is also noted (figure 3). Filming with sound enables a later, thorough analysis. Laryngeal findings at glottic and supraglottic level are as-sessed and scored according to the criteria by Maat et al (3). Since it is the effort that triggers the symptoms and the symptoms subside very quickly after the effort has ceased, it is not meaningful to perform a laryngoscopy either before or after exertion. Laryngoscopy during ongoing exercise is considered gold standard for diagnosing EILO (2, 4, 15, 61, 67, 73).

The scoring method by Maat et al is a validated method and the most widely used. This scoring method is described in detail later in this thesis. However, neither this method, nor any other proposed method for quantification of the laryngeal findings (76-78) are considered gold standard for setting the diag-nosis.

21

Figure 3. Continuous laryngoscopy exercise test on treadmill. A fibre-optic laryngo-scope is inserted through the nose, fixated in a specially designed helmet and con-nected to a camera. The tip of the laryngoscope is placed just above the epiglottis to provide a detailed view of the larynx.

Lung function test Spirometry is described as a diagnostic method of investigation. A truncated inspiratory flow-volume curve has been observed in studies on patients with stridor (14, 42) but since stridor can have other causes this information is of little value. The method is also highly dependent on the patient´s cooperation (37). Another study evaluated flow-volume loops before and after exercise and it showed that spirometry cannot be used as a diagnostic tool for EILO (79). However, lung function tests are valuable in the investigation for co-morbidities like asthma.

Other provocation tests EVH has been shown to induce both glottic and supraglottic obstruction (80). Methacholine and mannitol (81, 82) as well as histamine (76, 83) have been tested in the same way as for bronchoconstriction with no significant results. On the contrary, it is not possible to draw any conclusions of a positive meth-acholine challenge test since it is also used to investigate asthma.

A great disadvantage with all methods except direct visualization of the larynx is that no information is obtained about the location of the obstruction, i.e. if it is a glottic or a supraglottic obstruction.

22

Aetiological aspects of EILO The aetiology for EILO has not been established. Since laryngeal obstruction can occur at different levels, several theories have been suggested. A review of the literature reveals a complexity that is more complicated than that there is just one, single trigger (exercise), that would cause the typical breathing problems and laryngeal findings in EILO. The presence of a co-morbidity could lead to a laryngeal hypersensitivity, leading to that a lower amount of stimulus, which in the normal case would not trigger breathing problems, will do so in that context (84). Breathing problems can be further aggravated if the response to the stimulus is disproportionately large (85). Asthma and gastro-oesophageal reflux are the most studied conditions that have been associated with EILO. It is therefore possible that EILO in some cases are caused by one mechanism alone, and in other cases by a combination of aerodynamic mech-anisms, laryngeal pathophysiology and co-morbidity.

Aerodynamic forces The larynx is, as earlier described, the narrowest part in the upper airways. When airflow increases, so does the forces acting on the laryngeal tissues. A possible explanation for supraglottic EILO could be that the supporting struc-tures no longer manage to keep the arytenoid cartilages and the plicae aryepi-glottica in an upright position, which results in a supraglottic collapse. When the opening for the airflow gets smaller the velocity of the airflow increases according to Poiseuille´s law. This leads to an even greater negative pressure that may affect the vocal cords and give rise to a secondary glottic obstruction (4, 38).

Impaired neurological function The glottic closure reflex protects the lower airways against aspiration and can be triggered by chemical stimuli or air puff stimuli (81, 86). Hyperventilation during exercise could theoretically elicit the reflex. Gastro-oesophageal reflux could possibly act as a chemical trigger but studies are inconclusive whether reflux leads to a more easy-triggered reflex or in fact desensitizes the laryngeal receptors (87).

Psychological factors There are different views on whether psychological factors, for instance a panic attack, causes breathing problems in EILO, or if the panic attack is a response to severe breathing problems during exercise. Since EILO for a long time was considered to be only a vocal cord dysfunction problem, and because diagnosis in most cases was made without proper endoscopic examination during exercise, the predominant view has been that psychological problems are either the main cause or a strong contributor (42, 46, 88-92). When endo-scopic examination of the larynx during exercise has been performed, panic

23

attacks or stress reactions has been noted after the obstruction occurs, which strengthens the reverse opinion (4, 38). However, stress as a trigger cannot be completely ignored, as psychological factors always can play a role in all sit-uations.

Treatment for EILO The starting point for treatment is a correct diagnosis where untreated co-mor-bidities have been excluded. For both glottic and supraglottic EILO it has been reported to be of great value to the patient to receive thorough information about the origin of symptoms and to have an opportunity to see the recording of the laryngeal findings obtained during the CLE-test (38, 75).

Glottic EILO The psychological view has greatly influenced treatment. The pressure “to perform” has been stressed as a contributing factor among athletes. Individual support and counselling have dominated (93). Psychotherapy (5, 94) and hyp-nosis (95) have been suggested. Speech therapy with the aim of learning to relax the laryngeal muscles (96, 97) and to master different breathing tech-niques (5, 98-100) have been reported to be effective but no controlled studies have been done. Neither the placebo effect nor how different breathing tech-niques should be performed during exercise have been taken into proper con-sideration.

Doshi reported that Atrovent (Ipratropium) oral inhaler before exercise could prevent exercise-triggered VCD in six patients (101). Inspiratory mus-cle training (IMT) is believed to strengthen PCA (102, 103) in subjects with laryngeal obstruction but it remains to be proven. In healthy subjects, IMT has been shown to have effect on abduction of the vocal cords (104). Anti-reflux therapy has been tried, however not specifically for glottic obstruction occur-ring during exercise (105).

Supraglottic EILO Both speech therapy and breathing techniques are tried at some centres (108) but no long-time follow-up results exists. Surgery (laryngoplasty) involves re-duction of the aryepiglottic folds to prevent them from covering the laryngeal inlet (38, 106, 107) (figure 1 and 2). The supraglottic subtype must be verified endoscopically during an exercise-test before surgery.

24

Aims of present studies

To analyse the consistency between CLE-score and EILOMEA, two diagnostic methods for laryngeal obstructions (I) To investigate the prevalence of EIB and EILO in adolescents with and without self-reported exercise- induced dyspnoea (II)

To map and compare movements of glottic and supraglottic anatomi-cal structures in healthy adolescents and adolescents experiencing dyspnoea during strenuous exercise (III) To investigate the robustness of diagnostic methods for EILO (I-III) To evaluate surgery as a treatment for severe cases of supraglottic EILO regarding self-reported exercise-induced breathing problems (IV)

To investigate how symptoms and physical activity change over time in patients with EILO who have undergone surgery, in patients with EILO treated conservatively and in patients tested negative for EILO at CLE-test (IV)

25

Methods

Ethics All studies were approved by the Regional Ethics Board at the Medical Fac-ulty at Uppsala University, Uppsala Sweden (Dnr 2010/107 (I), Dnr 2011/413 (II-III), Dnr 2011/208 (IV)). Informed written consent was obtained from all participants and their guardians in study II-IV. According to the Ethics Rewiev Board, no written consent was needed for study I since it was a methodological study of unidentifiable endoscopic laryngeal recordings where no personal data was exposed.

Table 1. Overview of studies included in this thesis.

Study I n=60

Study II n=150

Study III n=125

Study IV n=84

Objective

Comparison of CLE-score and EILOMEA methods

Prevalence of EIB and EILO

Laryngeal move-ments during ex-ercise in controls and dyspneic subjects

Impact of ST, CT and a nega-tive diagnosis on exercise-induced dyspnoea and physical activity

Design

Method compar-ison

Observational Descriptive Cross sectional

Observational Descriptive Case-control

Observational Follow-up

Data used for analyses

CLE-scores and measurements from laryngeal recordings

Outcome of EIB- and CLE-tests

CLE-scores and measurements from laryngeal recordings

Self-reported ex-ercise-induced dyspnoea and physical activity

Dropouts or exluded

7 25 46 25

Total analysed

53 Anonymous

125 Controls n=42 Dyspnoea n=83

79 Controls n=28 Dyspnoea, no diagnosis n=31 EIB n=10 EILO n=10

59 ST n=14 CT n=17 Tested negative n=28

ST = surgically treated, CT = conservatively treated

26

Diagnostic procedures EILOMEA (paper I and III) EILOMEA is a validated, diagnostic, software measuring tool developed to calculate the severity of an obstruction at the glottic or the supraglottic level (78). The rater chooses a still frame obtained from a laryngeal recording at a CLE-test which shows the maximum obstruction during inhalation. A set of points are plotted on the image which mark anatomical landmarks. A baseline is chosen where the plicae aryepiglottica turns into the epiglottis and hereafter both tuberculum cuneiforme, the anterior commissure, the vocal cords and the posterior wall between the arytenoid cartilages are marked (figure 4). The soft-ware calculates three parameters: the glottic angle at the anterior commissure between the true vocal cords, the glottic obstruction (obtained by dividing the area between the vocal cords by the anterior-posterior laryngeal distance) and the supraglottic obstruction (ratio between the actual lumen and the lumen at the level of the arytenoid region). Glottic obstruction is referred to as P-factor and supraglottic obstruction as C-factor.

Figure 4. Anatomical structures marked according to guidelines for EILOMEA

27

EIB-test (paper II) The test for exercise-induced bronchoconstriction was performed according to current guidelines presented by the American Thoracic Society (32). All participants were instructed to cease short-acting β2-agonists 8 h before the test, long-acting β2-agonists 24 h before the test and leukotriene receptor an-tagonists 72 h before the test. On the day of the test no inhaled corticosteroids should be used. Within four hours before the test they should not perform vig-orous exercise, eat heavy meals or use nicotine or caffeine containing food or beverages.

The test was performed by letting the adolescent run for 6 minutes on a treadmill while wearing a nose clip and breathing dry air through a tube. The treadmill speed and slope was adjusted individually so that all participants reached 90% of predicted maximum heart rate (109) within the first 1.5 minutes and maintained this exercise load throughout the test. Baseline spi-rometry was measured following recommended guidelines (110) both before running and repeatedly after the test. EIB was defined as a decrease of ≥10% in FEV1 from baseline in at least one measurement during a 30-minutes time span after the test.

CLE-test (paper II) Prior to the test all participants were informed that there were no restrictions regarding asthma medication, meaning that those who normally use asthma medication before physical exercise should do so also before the CLE-test. They were instructed to avoid vigorous exercise on the day of the test and heavy meals within four hours before the test.

The test was performed according to the procedure described by Heimdal et al (1). Before the test, naphazoline-lidocaine was sprayed twice into each nostril to achieve local anaesthesia and dilatation. After warming up for two minutes, the treadmill was stopped and a fibre-optic laryngoscope was in-serted in one nostril so that the tip of the laryngoscope was placed just above the epiglottis, providing a detailed view of the larynx. The laryngoscope was fastened in a specially designed helmet so that it stayed in position, and then it was connected to a camera. The participant started to run at a speed of 9 km/h with no inclination and the speed was increased by 1 km/h every minute until 13 km/h. Hereafter the inclination was increased by 2% every minute. The participants were instructed to run until breathing problems forced them to stop, or if no breathing problems occurred, until exhaustion. For the test to be considered successful, the heart rate had to reach ≥90% of predicted maxi-mum (110). The larynx was filmed continuously and laryngeal findings were graded at moderate and maximum effort according to the CLE-score method (3). The upper body was recorded using an external camera that also filmed the monitor showing the larynx (figure 3). An obstruction of grade 2 or more

28

at either the glottic or supraglottic level was considered a positive response and was defined as EILO.

CLE-score method (paper I-IV) The CLE scoring method is an assessment tool developed for assessing the dynamic behavior of anatomical structures in the larynx during exercise, from moderate effort to exhaustion (3). When scoring is made during inhalation in a representative breathing cycle it allows evaluation of presence, type and se-verity of both glottic and supraglottic obstruction. The degree of glottic ad-duction and medial rotation of the aryepiglottic folds are assessed separately, each on a four-graded scale ranging from 0 to 3, where 0 represents no ob-struction, 1 is mild, 2 is moderate and 3 is severe obstruction (figure 1). An obstruction score of 2 or more at either level is considered diagnostic for EILO. Assessments are made at two separate moments during the CLE-test; at moderate effort within 1 minute from start, and at maximum effort shortly before the test person stops due to exhaustion or respiratory distress.

In study I, II and IV, only the CLE-scores at maximum effort have been used. In study III, CLE scores from both time points have been used. Scoring has been made in real time during CLE-tests (study II and IV), and from video recordings obtained at earlier CLE-tests (study I and III).

Study populations/Material and Design

Paper I Study material Sixty anonymous laryngeal recordings obtained from previously performed CLE-tests in three Scandinavian centres were used. Each centre submitted 20 laryngoscopic recordings, purposely chosen to reflect cases with varying com-binations and severity of glottic and/or supraglottic obstruction and varying quality of the recordings, including compressed file format (i.e. lower resolu-tion of the picture).

Study design Using their respective method, the developers of the CLE score method (3) and the EILOMEA method (79) evaluated the recordings. The developers as-sessed each recording once and were blinded both to the history of the patients and to each other´s assessments. Correlation between assessments made with the two methods were investigated.

29

Papers II-III Background data All 12-13-year-old adolescents (born 1997-1998) in the city of Uppsala (n=3,838) were invited to participate in a survey investigating exercise-in-duced dyspnoea (19). The response rate was 60.2% (n=2,309) whereof 1,136 were females. No differences were seen regarding age, gender, body mass in-dex and postal area codes between the adolescents who responded and those who did not. An affirmative answer to the question “Have you had an attack of shortness of breath that came following strenuous activity at any time in the last 12 months?” (111) classified the adolescent as dyspneic (n=330), and a negative response as a healthy control (n=1,979).

Study population In study II the study population consisted of 150 adolescents invited to per-form exercise tests. A total of 146 adolescents completed an EIB-test and 125 completed both an EIB- and a CLE-test, 83 with dyspnoea and 42 controls.

In study III the study material consisted of laryngeal recordings from the 125 adolescents who had completed both tests.

Study design Paper II Results from the survey regarding dyspnoea showed a female-male ratio of 3:2. A stratification for these gender differences was made and two randomly ordered lists were created, one with adolescents with dyspnoea and one with healthy controls. The exclusion criteria were pulmonary disease apart from asthma, cardiac co-morbidity or a functional disability that resulted in inability to perform exercise tests. By random selection, 199 adolescents with dyspnoea and 123 controls were invited to perform two separate exercise tests to diag-nose or exclude EIB and/or EILO.

A total of 103 adolescents with dyspnoea and 47 controls accepted to par-ticipate. In both groups, there were more adolescents that had reported physi-cian diagnosed asthma who chose to participate, compared to adolescents who declined participation. No other differences were detected between the groups. Group characteristics are presented in table 2.

At a first clinical visit, a research nurse informed all participants individu-ally about forthcoming exercise-tests for EIB and EILO and how to prepare for them.

At a second visit the adolescents performed a standardized EIB-test on treadmill with inhalation of dry air (8). It was performed on average a median (IQR) of 12 (2-19) days after the pre-test visit and the investigator who col-lected the data was blinded to whether the participants belonged to the exer-cise-induced dyspnoea group or the control group. Four adolescents with dyspnoea declined participation in the EIB-test.

30

Table 2. Characteristics of participating adolescents and adolescents declining par-ticipation in study II. Results presented as mean ± SD and n (%).

Exercise-induced dyspnoea Controls

Answers from background questionnaire

Participat-ing adoles-cents n=103

Declining participa-tion n=96 p-value

Participating adolescents n=47

Declining participation n =76 p-value

Age 13.7±0.66 13.7±0.66 0.92 13.6±0.64 13.8±0.71 0.29 BMI (kg/m2) 19.4±2.6 19.4±2.8 0.96 19.1±2.2 19.3±3.1 0.65 Asthma 49 (47.6) 31 (32.2) 0.028 10 (21.3) 3 (3.9) 0.002 Rhinitis 29 (28.3) 30 (31.3) 0.63 6 (12.8) 10 (13.2) 0.95 Current wheeze 57 (55.3) 47 (49) 0.37 3 (6.4) 3 (3.9) 0.54 Day time dyspnoea 17 (16.5) 9 (9.4) 0.13 1 (2.1) 2 (2.6) 0.86 Nocturnal dyspnoea 27 (26.2) 20 (20.8) 0.37 1 (2.1) 2 (2.6) 0.86

At a third visit the adolescent performed a CLE-test on treadmill (1). The test was performed on average a median (IQR) of 38 (22-66) days after the EIB-test. The investigator was not the same person who had conducted the EIB-test and was blinded to whether the participants belonged to the exercise-in-duced dyspnoea group or the control group, as well as to the EIB test result. Out of 146 adolescents who had performed the EIB-test, 19 (14 with dysp-noea) declined participation in the CLE-test. Another two adolescents from the dyspnoea group was excluded while doing the test, one due to equipment failure of the recording and one due to fainting shortly before the test when the laryngoscope was inserted. An overview of the design and participants in each step are presented in figure 5.

31

Figure 5. Participants´inclusion and exclusion in study II.

32

Study design Paper III Not all tested adolescents were found to have EIB or EILO. Scoring results from adolescents in the control group and the exercise-induced dyspnoea group without a diagnosis was examined further. The laryngeal films recorded during the CLE-tests were anonymized and coded so that the investigator was blinded to whether the adolescent had stated breathing problems or not. Films were excluded due to poor visualization of laryngeal structures or if there was no endoscopic recording at all, only an external film with the larynx visible on the monitor beside the test subject (figure 3).

A still frame was chosen from the film for each subject both at moderate effort and at maximum effort, i.e. at the same time points as when the CLE-scoring had been done. The still frames were assessed in random order and measurements were made three times on each still frame using the software measuring tool EILOMEA (78). Mean values for the glottic angle, glottic obstruction (P-factor), and supraglottic obstruction (C-factor) were calculated, as well as the difference between the mean values at moderate and maximum effort.

This procedure was repeated with films from adolescents who had been diagnosed with EIB and EILO. From the EIB group, only films from subjects that had not taken any asthma medication before the CLE-test were included. Study population and group characteristics are presented in table 3.

Table 3. Population and group characteristics of participants in study III. All data ex-cept for gender are presented as mean ± SD.

Controls n=28

Exercise-induced dyspnoea n= 31

EIB n= 10

EILO n= 10

Age; years 14.9±0.7 15.0±0.7 14.9±0.5 14.7±0.5 Females, n (%) 17 (61) 16 (52) 7 (70) 6 (60) Body mass index 20.3±2.4 20.9±3.4 20.5±2.5 21.5±3.7 Height, cm 170.4±8.8 169.1±7.5 166.8±10.4 169.3±9.3 Weight, kg 59.3±10.9 60.0±11.4 59.9±5.8 61.9±13.3

Paper IV Study population The study population consisted of a cohort of 84 consecutive patients (69 fe-males) referred for exercise-induced dyspnoea and CLE-tested during a four-year period. Forty-seven of the subjects were diagnosed with EILO and 37 had a negative test result. Out of the 47 EILO positive, 15 underwent surgical treatment during this period.

33

Study design All patients answered a questionnaire at diagnostic CLE-test regarding exer-cise-induced breathing problems, current physical activity level and medical history of asthma and perennial allergy. To rate the symptom severity of the exercise-induced breathing problems, a visual analogue score (VAS) ranging from 0 (no breathing problems) to 10 (worst possible breathing problems) was used. Patients diagnosed with EILO received information about breathing ex-ercises that might reduce symptoms and were invited to contact the investi-gating doctors by phone two months after the CLE-test and give a follow-up report on the effect. If breathing exercises were insufficient, the possibility of improvement by surgery were discussed with subjects having severe supra-glottic EILO. Out of 27 subjects, 15 underwent laryngoplasty.

A follow-up questionnaire was sent to all subject 1-3 years after the diag-nostic CLE-test (Appendix 1). The original questionnaire was supplemented with questions about whether the subject had changed physical activity level and/or type of sport. If that was the case, they were asked to specify if these changes were because of breathing problems or not. Patients were again asked to grade the severity of the breathing problems on a visual analogue scale. The responders were categorized as subjects with EILO who had undergone sur-gery, patients with EILO treated conservatively and patients who were tested negative for laryngeal obstruction (figure 6).

Figure 6. Study IV. Participants and dropouts including EILO subtypes and number of patient who were offered and accepted or refrained surgery in each group.

34

Statistical analyses For all analyses, only two-sided tests were used and the selected critical sta-tistical significance level was 95% (p<0.05).

Paper I The EILOMEA method results were controlled for normal distribution and equal variances for each matching CLE-score. Differences in means of the EILOMEA measures were tested between each CLE-score grade using t-test. Analyses were performed using JMP statistical software version 11.1.1.

Paper II Anthropometric data, lung function, asthma medication and exercise test re-sults were summarized as means, standard deviations and minimum and max-imum for continuous variables, and as numbers and percentages for categori-cal variables. Age, BMI and FEV1 were compared between subjects using un-paired Student´s t-test. For all categorical variables, a cross tabulation vs sub-ject groups were performed and subject groups were compared using Fisher´s exact test. The EIB and EILO prevalence estimates were based on the assump-tions that the population strata with and without exercise-induced dyspnoea differed. Assuming that there was no selection bias, the dyspnoea stratum with exercise-induced dyspnoea was calculated from the rate of exercise-induced dyspnoea among responders multiplied by the population size, and the stratum without exercise-induced dyspnoea was assumed to represent the rest of the population. Tested adolescents were assumed to be random samples from the two strata. The prevalence estimate and 95% confidence interval were calcu-lated according to the theory for stratified samples (112). Data were analyzed using Statistical Package for Social Science software v.21.

Paper III All analyses were made at group level. Non-parametric Mann-Whitney U-tests were performed to compare groups two-by-two. Continuous variables were summarized as means and standard deviations. A regression analysis (95% confidence interval) was performed with change in percent from value at moderate effort to value at maximum effort as dependent variable and group, gender and value at moderate effort as independent variables.

Paper IV Kruskal-Wallis test (non-parametrical ANOVA) was used to analyze if there was any difference between groups regarding symptom severity. If results were significant, Mann-Whitney U-tests were performed to compare groups two-by-two. Fisher´s exact test was used to compare adjusted physical activity level between surgically and conservatively treated EILO subjects. SAS soft-ware version 9.3 was used for the analyses.

35

Summary of results

Comparison of CLE-score and EILOMEA (paper I) The CLE-score and EILOMEA method were compatible, showing significant differences in the means for all EILOMEA measures corresponding to CLE-score 1 vs 2 (mild vs moderate obstruction) and 2 vs 3 (moderate vs severe obstruction). EILOMEA also distinguishes between CLE-score 0 and 1 (no obstruction vs mild obstruction), regarding the glottis area (P-factor), but not regarding supraglottic area (C-factor) or the glottic angle (table 4).

Table 4. Means, standard deviations and p-values for EILOMEA glottic angle and glottic area for each glottic CLE-score, and for EILOMEA C-factor for each supra-glottic CLE-score

CLE-score EILOMEA p-value

Glottic obstruction Glottic angle0 48 (9) -1 45 (14) >0.052 25* (9) <0.00013 13* (6) 0.0023Glottic obstruction P-factor0 64 (20) -1 48* (17) 0.01392 25* (8) <0.00013 16* (8) 0.012Supraglottic obstruction C-factor0 0.51 -1 0.47 (0.09) >0.052 0.34* (0.08) <0.00013 0.21* (0.12) 0.0006

*Differed significantly from the mean of the CLE-score of the row above

36

Prevalence of EIB and EILO (paper II) In the dyspnoea-group 42 adolescents (42.4%) were diagnosed with EIB and among the controls seven (14.9%) met the criteria for an EIB diagnosis. The estimated prevalence of EIB in the total population of 3,838 adolescents was 19.2% (95% CI 9.9;28.4), with no gender differences (18.6% in females and 19.7% in males).

EILO was diagnosed in nine (10.8%) adolescents in the dyspnoea group and two (4.8%) in the control group. The estimated prevalence of EILO in the population was 5.7% (95% CI 0.01;11.4) with no gender differences (5,7% in both females and males).

Among the dyspneic adolescents (n=83), EIB were found in 39.8%, and among the nine adolescents diagnosed with EILO, four (4.8%) were found to have both diagnoses, all were females. In 49.4% of dyspneic adolescents, no diagnosis was found (figure 7). Males were significantly overrepresented among adolescents where no diagnosis was found; 64.7% of males and 38.8% of females remained undiagnosed (p=0.026) (figure 8).

Participants with dyspnoea and a diagnosis of EIB and/or EILO where no different from participants with dyspnoea who did not get diagnosed regarding body mass index, lung function (FEV1), self-reported ever diagnosed asthma or self-reported use of asthma medication (table 5).

Table 5. Characteristics of subjects with exercise-induced dyspnoea without EIB and EILO versus subjects with exercise-induced dyspnoea with EIB and/or EILO.

Exercise-induced dyspnoea without EIB or EILO

Exercise-induced dyspnoea with EIB and/or EILO

p-value

Subjects, n (%) 41 (49.4) 42 (50.6)Age (years) mean (min, max) 14.29 (13,15) 14.24 (13,15) 0.67 Girls, n (%) 19 (46.3) 30 (71.4) 0.026 BMI, mean (SD) 20.8 (3.1) 21.1 (2.2) 0.53 FEV1 % predicted*, mean (SD) 91.9 (12.0) 91.1 (9.5) 0.94 Ever asthma, n (%) 16 (39.0) 19 (45.2) 0.66 Inhaled corticosteroids, n (%) 7 (17.1) 12 (28.6) 0.30 Short-acting beta agonists, n (%) 19 (46.3) 19 (45.2) >0.99 Long-acting beta agonists, n (%) 2 (4.9) 0 0.24 Leukotriene receptor antagonists, n (%) 2 (4.9) 3 (7.1) >0.99

37

Figure 7. EIB and EILO in adolescents with self-reported exercise-induced dysp-noea. No diagnosis was found in almost half of the tested adolescents.

Figure 8. EIB and EILO in males and females with self-reported exercise-induced dyspnoea. All adolescents with both EIB and EILO were females.

38

Laryngeal findings during exercise (paper III) Laryngeal findings assessed with CLE-score Movements of the laryngeal structures corresponding to a CLE-score 1 at glot-tic and/or supraglottic level was seen in 74% of controls, in 83% of dyspneic adolescents without a diagnosis and in 66% of adolescents with EIB. Most common in all groups was an isolated supraglottic CLE score 1. The combi-nation of CLE score 1 at both glottic and supraglottic level was more common among controls than in the dyspnoea group (5 vs 1). In subjects with EILO, all but one had supraglottic obstruction only.

Laryngeal findings assessed with EILOMEA All groups tended to reduce the glottic angle and develop some degree of su-praglottic movement that reduced the laryngeal inlet area during increasing work effort. In all three parameters that where investigated, no significant dif-ference could be detected from moderate to maximum effort between controls, the dyspnoea group without a diagnosis of EIB or EILO, and the EIB group. All three parameters differed significantly between subjects diagnosed with EILO compared to the other groups (p=<0.001) (table 6).

Table 6. Change in percent of glottic angle, glottic area (P-factor) and supraglottic area (C-factor) from moderate to maximum effort, presented as mean (SD).

Controls

n=28

Dyspnoea, no EIB or EILO

n=31

EIB n=10

EILO n=10

Angle change (%) -11.84 (12.56) -11.46 (12.07) -13.73 (9.83) -36.8 (24.24) P-factor change (%) -15.32 (10.41) -14.89 (11.07) -9.76 (13.56) -44.92 (20.73) C-factor change (%) -19.67 (17.46) -19.93 (17.69) -17.06 (12.39) -46.78 (25.37)

Surgical vs conservative treatment for EILO in comparison with patients tested negative (paper IV) Time between the first questionnaire at diagnostic CLE-test and the follow-up questionnaire was on average 17.6 months. Out of 84 invited subjects, 59 (70%) answered both questionnaires. Results from follow-up questionnaire are presented in table 7.

39

Impact on exercise-induced dyspnoea All surgically treated subjects reported less breathing problems when assessed by change in VAS between diagnosis and follow-up. There was a significant difference compared with both conservatively treated subjects and subjects who were tested negative (p<0.001).

Figure 9. Change in severity of exercise-induced dyspnoea between diagnostic CLE-test and follow-up as given by visual analogue scores.

Impact on physical activity Among surgically treated none were less physically active and none had changed sport due to exercise-induced dyspnoea. One third of surgically treated subjects did report a lower physical activity level at follow-up but ex-ercise-induced dyspnoea was not stated to be the cause for this in any of the cases. Instead different sorts of changes in life were mentioned. In contrast, 41.2% of the conservatively treated had adjusted their activity level or started to practice another sport than the sport they practiced at the time for the diag-nostic CLE-test. This difference between EILO-groups was found to be statis-tically significant (p<0.001).

Among all EILO positive subjects, surgically treated as well as conserva-tively treated, practice of endurance sports was more common than practice of interval sports or other sports containing both endurance and interval train-ing. Among subjects tested negative, practice of endurance sport and interval sports were equally common.

40

Table 7. Data from follow-up questionnaire presenting symptom severity, activity level and practiced sports. All data presented as n (%).

EILO ST n=14

EILO CT n=17

Tested negative n=28

Symptom severity at follow-up Less symptoms 11 (78.6) 3 (17.6) 12 (42.9) Symptoms unchanged 3 (21.4) 12 (70.6) 14 (50)More symptoms 0 2 (11.8) 2 (7.1) Activity level at follow-up Less active or adjusted activity because of breathing problems 0 7 (41.2) 7 (25)Less active, other causes 5 (35.7) 2 (11.8) 7 (25)As active as before 5 (35.7) 3 (17.6) 8 (28.6) More active 4 (28.6) 5 (29.4) 6 (21.4) Practiced sports Endurance sports 8 (57.1) 10 (58.8) 10 (35.7) Interval sports 2 (14.3) 3 (17.6) 8 (28.6) Other sports 3 (21.4) 2 (11.8) 6 (21.4) Not active 1 (7.1) 2 (11.8) 3 (10.7) Changed sport because of breathing problems 0 4 (23.5) 2 (7.1)

Missing answers 0 0 1 (3.4)

41

Discussion

The work behind the present thesis aims to increase and deepen the knowledge of exercise-induced dyspnoea with special focus on exercise-induced laryn-geal obstruction.

CLE-score vs EILOMEA In this method comparison study, we investigated if assessments of laryngeal obstructions with EILOMEA (78) were equivalent to assessments made ac-cording to CLE-score (3).

Both CLE-score and EILOMEA were developed as diagnostic tools for EILO. CLE-score is the method used in the clinic since it allows evaluation of the dynamic process in the larynx from rest to exhaustion. EILOMEA evalu-ates measures in a still frame, i.e. a specific moment in this dynamic process. None of the methods can assess all different subtypes of EILO. EILOMEA is highly dependent on a clear view of defined anatomical structures. If supra-glottic structures cover the vocal cords, as in CLE-score 3, the glottic values are not possible to obtain. With the CLE-score method it is not possible to score an obstruction where the epiglottis tilts backwards and covers the laryn-geal inlet.

Analyses of laryngeal recordings were performed by the two developers of the respective method. By this, user manual misunderstandings as a source of error was avoided, leaving the actual application and technical differences of the methods to account for any lack of correlation. To examine the robustness of the two methods, the raters were given recordings of various subtypes of laryngeal obstructions.

When CLE-score was developed, both patients and healthy controls were tested. Some controls were found to have mild inward movements of laryngeal structures during exercise and thus a CLE-score 1 is not considered to be clin-ically important. That EILOMEA did not detect differences between CLE-score 0 and 1 for the glottic angle and the supraglottic obstruction is therefore not crucial for a correct medical evaluation.

42

Prevalence of EIB and EILO Our findings compared to others The present cross-sectional study demonstrated that EIB and EILO, assessed using standardized exercise tests, are common in a general population of ado-lescents. The estimated prevalence of EIB was 19.2% and the estimated prev-alence of EILO was 5.7% with no gender differences. In subjects with exer-cise-induced dyspnoea, EILO alone and EIB in combination with EILO were found to a similar extent. In approximately half of the subjects with exercise-induced dyspnoea, neither EIB nor EILO could be demonstrated and this un-explained self-reported exercise-induced dyspnoea was more common among males than females.

Prevalence numbers for EIB in older children and adolescents ranges be-tween 7-16% in the few previous studies that have been published (24-27). Our prevalence rate was higher, probably due to the use of a strict method (dry air ventilation) and setting the diagnosis at a fall in FEV1 ≥ 10 %.

Comparing results in EILO studies are even more difficult since study pop-ulations, methods and diagnostic criteria vary (35, 73, 97). Additionally, many different terms have been used which adds further confusion (5, 15, 113) and no difference is made between inducers; exercise-induced laryngeal obstruc-tion is often not considered to be any different from laryngeal obstructions induced by other triggers.

There is only one previous study that have investigated the prevalence of EILO in a general population (33). EILO was found in 7.6% in a population with mean age 18.6 years. Dropouts during this previous study was very high, only 17.6% of participants entering the study performed a CLE-test. In our study 83.3% of adolescents entering the study completed both an EIB and a CLE-test and mean age was 14.2 years.

Gender differences No gender differences were found neither for EIB nor EILO. For EIB this is consistent with results from other studies (114, 115). However, there are stud-ies who do report gender differences for EIB (24, 27) and interestingly the prevalence is higher in females. Females are also overrepresented in most studies and case reports on EILO (33, 38, 113, 116). The main difference be-tween our study and other studies was that participants in the studies where a gender difference was found were older. This might indicate that gender dif-ference depends on larynx size. The size of the laryngeal opening does not differ between males and females in pre-pubertal age. During puberty, the lar-ynx grows more in males than in females and it has reached adult size by the age of 16-17 years (117, 118). In our study, CLE-tests were conducted during a period of 1.5 years and some of the participants were still pre-pubertal when

43

tested. We found four males with supraglottic EILO, all of them born in 1998, and three of them born at the end of the year, thus being among the youngest participants in the study. All males with EILO, as a group, were younger than the females, mean age being 14.2 and 15.0 respectively.

Can we trust the prevalence numbers? The estimated prevalence of EIB was 19.2% and the estimated prevalence of EILO was 5.7%. Since study II was an extension of a previous survey inves-tigating the prevalence of exercise-induced dyspnoea (19), it is necessary to take the methodology and results from that survey into account when we in-terpret the prevalence results.

The survey was sent to adolescents born in 1997-1998, who at the time were 12-13 years old. The prevalence of exercise-induced dyspnoea in this study was 14.3%. Exercise tests in study II started approximately one year later and went on for 1.5 years, meaning that the oldest tested adolescent was 15 years old. Mean age in the total tested population was 14.2 years.

The response rate to the survey was 60.2%. No differences were seen among responders and non-responders regarding age, gender or postal code. Although the response rate is acceptable, there is a risk of overrepresentation of adolescents that experience exercise-induced dyspnoea.

The survey contained over 30 questions comprising questions about exer-cise-induced dyspnoea, asthma, allergy and physical activity. We used the question “Have you had an attack of shortness of breath that happened after strenuous activity at any time during the last 12 months?” (111) as a screening question for exercise-induced dyspnoea. This question has been proven to be useful in detecting asthma (119) but have not been tested for other diagnoses. The wording “…after strenuous exercise” might have played a role. In EIB, symptoms typically occur after exercise but EILO symptoms typically occur during the exercise. If the question was interpreted literally, adolescents that experienced dyspnoea during exercise might not have answered affirmative to this particular question, and wrongly been classified as a control in study II. Our intent with the question, however, was the opposite: we wanted adoles-cents not only with typical asthma to answer yes to this question. This is also more likely to be the case since 67% of adolescents who reported exercise-induced dyspnoea had no asthma diagnosis. It is possible that we got affirma-tive answers from adolescents with poor physical condition, which might be an explanation for the high proportion of adolescents reporting dyspnoea who remained undiagnosed. Poor physical condition has been confirmed to be common in retrospective studies investigating the accuracy of an earlier asthma diagnosis (21, 22, 120). As exclusion criteria for the study, we asked about heart conditions and lung conditions apart from asthma, but we did not investigate any other diagnoses that can cause exercise-induced dyspnoea.

44

Recall bias must always be taken into consideration (121). Also, this ques-tionnaire was not validated as an entity but many of the questions had previ-ously been used in other population studies.

Factors in study II that might have influenced the prevalence numbers There is a possibility of a selection bias in the process where adolescents ac-cepted to enter study II. People with a disease or symptom tend to be more prone to participate in studies about this particular condition. The survey pre-ceding study II included a question about “ever being diagnosed by a physi-cian as having asthma”. Both among adolescents reporting exercise-induced dyspnoea and among controls there were more adolescents who had answered yes to this question and chose to participate, than among adolescents who de-clined participation. This, however, does not confirm for certain that they had asthma. Since EILO can be mistaken for asthma, some of the adolescents with an asthma diagnosis could in fact have EILO instead.

EIB-test All adolescents were tested for EIB in the same way following a standardized protocol and recommended guidelines (32). The investigator was blinded to whether the test subject had stated exercise-induced dyspnoea or not and could not influence the test result. The subjects breathed dry air with a nose clip during the EIB-test. This has not been done in other studies (24, 25, 27, 28), and might have increased the sensitivity of the test, thereby increasing the prevalence. The sensitivity of the test might also have been increased by en-suring high effort exercise throughout the test (64).

Guidelines for the EIB-test includes instructions for the person to be tested. The adolescents were instructed to cease asthma medications before the test but if that was followed by everybody is uncertain. If effect of asthma medi-cation were present at the time of the test this may have led to false negative results.

Spirometry was performed several times. Reliable results depend on a mo-tivated patient performing the procedure correctly. In this study, the investi-gator instructed all adolescents in the same way and made certain that the test subject had a correct technique. It is possible that symptoms can be affected by other factors and vary over time, especially if the adolescent also have asthma. It would therefore have been interesting to test the adolescents twice, even if the EIB-test has been shown to be reproducible in the majority of sub-jects tested in a study by Anderson et al (122).

45

CLE-test All adolescents were tested for EILO in the same way following a standard-ized protocol which was a modified version of the CLE-test (1). Electrocardi-ogram or oxygen uptake was not measured during the test, only heart rate, but this has been reported to be sufficient (108). All tested adolescents reached 90% of predicted maximum heart rate. The investigator was blinded to whether the test subject belonged to the dyspnoea group or the control group, and to the result of the EIB-test. The investigator could not influence the test result.

Performing laryngoscopy continuously during exercise is gold standard for investigating EILO but the test has some weaknesses. For the test to be suc-cessful the test subject must reach maximum exercise capacity. If the person stops too early symptoms might not be triggered. The result is therefore highly dependent on the motivation and expectations of the subject tested, and he or she must feel safe in the test situation. For a person with EILO it can be very stressful to expose themselves to a situation where breathing problems might occur. Thus, this anticipating anxiety can hinder the person from trying their best which might lead to a false negative test result.

The CLE-test is performed on treadmill in a laboratory setting which can-not be compared with situations during exercise in “real life”, like a stressful competitive situation for the athlete. This can also have influenced the preva-lence numbers. Also with the CLE-test it would have been desirable to test the adolescents twice. No studies have yet examined the reproducibility of the CLE-test.

Before the fibre-optic laryngoscope was inserted, naphazoline-lidocaine was sprayed twice into each nostril to achieve local anaesthesia and dilatation in the nasal cavity. A small amount of the anaesthetics is transported to the laryngopharynx. There are no studies on how, or if, local anaesthetics applied in this way affect the laryngeal reflexes, and more specifically, the glottic clo-sure reflex. If the aetiology for glottic EILO is in fact an impaired neurological function, then locally administered anaesthetics could have reduced the prev-alence numbers for glottic EILO in our study. However, if aetiology for glottic EILO are caused by aerodynamic forces acting on laryngeal tissue, this is less likely, even though the hyperventilation itself can trigger a glottic response (81, 86). In the clinic, we have seen two types of glottic obstruction. One where the vocal cords adduct abruptly or gradually during the exercise session, and one where the glottis closes like a zipper, starting in the anterior commis-sure. The latter might be caused by aerodynamic forces. In study II we found one adolescent with glottic EILO (adduction type), the remaining ten had su-praglottic EILO.

46

CLE-scoring CLE-score allows assessment of obstructions at glottic and supraglottic level on an ordinal scale graded 0-3. A higher number means more severe obstruc-tion. The CLE-score method has been shown to correlate well with the symp-toms as reported by patients and assessed by observers (3). EILO is defined as a CLE-score 2 or more in conjunction with exercise-induced dyspnoea. In the present study, we observed two adolescents in the control group, one male and one female, with supraglottic obstruction grade 2. Even though they had obvious supraglottic obstructions, both stated exhaustion and not dyspnoea as the reason for them to stop running on the treadmill.

The normal laryngeal response to exercise is not completely known, i.e. we don´t know exactly what findings should be considered normal and what should be considered pathological. To put up criteria that differs healthy sub-jects from subjects with EILO is difficult. Mc Fadden et al proposed that vocal cord adduction exceeding 50% should be regarded as consistent with vocal cord dysfunction (42), but if, and when, a person experiences dyspnoea is sub-jective. Many other factors contribute to whether a subject will be diagnosed with the “disease” EILO (symptomatic EILO), or diagnosed with the “laryn-geal finding” EILO (asymptomatic EILO). The size of the laryngeal inlet at rest compared to at a high workload, body size, physical condition, motivation and other psychological factors can all influence the laryngeal response we observe. The motivation to exercise hard was very strong in the healthy sub-ject with supraglottic obstruction grade 2 in our study. One might speculate if the supraglottic obstruction in these asymptomatic EILO subjects is an effect on the laryngeal structures of high intensity exercise, and that the laryngeal obstruction occurs in symptomatic EILO subjects before they had the chance to exercise as hard as they wanted.

The scoring we did in this study resulted in eleven EILO positive adoles-cents and from this a prevalence of 5.7% in the general population wasesti-mated. However, this prevalence number is in fact not the prevalence of symp-tomatic EILO, but the prevalence of both symptomatic and asymptomatic EILO, i.e. the prevalence of laryngeal findings consistent with an EILO diag-nosis. Thus, the CLE-score as a diagnostic tool for EILO not always diagnose the actual condition that causes exercise-induced dyspnoea. Regardless, symptomatic EILO is not uncommon and should be diagnosed and treated.

Dropouts A total of 25 participants did not complete both exercise tests. However, as a group these adolescents did not differ from the adolescents who completed both tests regarding baseline characteristics. It is therefore unlikely that these dropouts have affected the prevalence estimate of EILO.

47

Larynx during exercise in adolescents with and without dyspnoea Consensus in the literature is that the larynx is maximally opened during ex-ercise to facilitate airflow. Just before exhaustion, at maximum minute venti-lation, a discrete adduction of the aryepiglottic folds can be noted (4, 15, 62). EILOMEA analyses showed that all adolescents tended to reduce the glottic angle and develop some degree of supraglottic obstruction during maximum exercise. Additionally, we found that we tended to underestimate the glottic obstruction when we used the CLE-score.

The pattern of mild obstruction at both glottic and supraglottic level was seen in five of the controls, but only in one of the adolescents with exercise-induced dyspnoea. This indicates that the additive effect of mild obstruction at two levels is not clinically important, at least not if no feeling of respiratory distress is present.

The typical dynamic development described for EILO is that the supraglot-tic adduction occurs first and the glottic involvement follows (4, 33). One might speculate if the controls could work harder for a longer time so that the glottic engagement was starting to emerge before they finally stopped exer-cising. Among adolescents with EILO the majority had supraglottic obstruc-tion grade 2 but the span of what we have assessed as a grade 2 is rather wide. This highlights the difficulties of the CLE-score being assessed on an ordinal grading scale. A grade 1 or 3 is more clearly defined than a grade 2 which might lead to that a wider range of obstruction severity will be assessed as grade 2.

Only one adolescent had glottic obstruction only. When analysing the re-sults, we included this adolescent in the total group of EILO subjects. This might have increased the difference seen between EILO subjects and the other groups regarding the glottic angle and area. In EIB, expiratory glottic obstruc-tion has been described (55, 67). In the present study, we only present meas-urements on glottic angle and area during inhalation at the same time as the maximum supraglottic obstruction occurred. We therefore do not know if the EIB-group, or other groups, kept the vocal cords adducted also during exha-lation. The fact that we performed within subject comparison and adjusted for the values obtained at moderate effort, minimizes the impact of individual lar-ynx size.

Surgical vs conservative treatment This follow-up study investigated the outcome of supraglottic surgery for EILO compared to the outcome of conservatively treated EILO. Additionally, the study included the patients tested for EILO but who did not get an EILO-diagnosis. Questionnaires were sent to all patients that, at that time, had been

48

tested on treadmill at our clinic (n=84). Conservative treatment implies breath-ing exercises (treatment for all EILO subtypes), and speech therapy (treatment for glottic EILO). Conservatively treated patients in this study with the supra-glottic subtype did, however, not follow a structured treatment program, which means that they can be regarded as untreated.

Prior to the present study, only one follow-up study had been published in which both surgically and conservatively treated EILO patients were investi-gated (38). Our questionnaires were similar but none of them validated. As earlier discussed (paper II), many types of biases can affect the results when a questionnaire is used and one important bias is recall bias. We also have a selection bias since the response rate was considerably higher in surgically treated than in conservatively treated (93% vs 53%).

Based on self-reported information, surgically treated EILO patients had less exercise-induced dyspnoea, were more physically active and had not to the same extent as the other groups adjusted their physical activity level by changing what sport they practiced. It would have been interesting to have re-tested all surgically treated patients to obtain objective data. As a routine, all operated patients are invited to perform a second CLE-test six months after the operation but too few in this study had performed such a test.

Potential biases in terms of positive and negative expectations must be taken into consideration. Surgically treated might be grateful for having some-thing done at all and overestimates the effect. On the other hand, if their ex-pectations are too high, they will be disappointed even if they manage to per-form better on a CLE-test after surgery. Conservatively treated might be dis-appointed over having an inoperable subtype of EILO. All these biases may have influenced the results.

Who should be offered surgical treatment? The decision to operate must be taken on a solid ground. Surgery can be con-sidered when severe supraglottic EILO is seen in combination with a strong subjective sensation of dyspnoea in a motivated patient with expectations on a realistic level. Additionally, surgery should preferably be offered only to adolescents and adults where the larynx has reached adult size.

Future perspectives on EILO Our prevalence study in adolescents shows that EILO is not an uncommon diagnosis. Despite progress in recent years, EILO is unfortunately too often misdiagnosed as exercise-induced asthma or EIB. We need to increase the awareness of EILO in the medical field so that persons with symptomatic EILO are diagnosed correctly and can be offered adequate treatment. .

49

EILO is most likely not a single disease with a single cause. We need to learn more about aetiology and pathophysiology, and try to identify the factors that underlie why some persons present with exercise-induced symptoms when other do not, even though they have similar obstructions in the larynx. By developing better diagnostic methods which can be used in a more “real-life”-exercise situation, we can learn a lot. We must also conduct randomized controlled trials where we investigate the impact of co-morbidities and objec-tively evaluate the effect of treatment. We also need to develop guidelines for treatment of the different subtypes of EILO. The terminology recently sug-gested by a Task force needs to be implemented in both clinical practice and in research.

50

Conclusion

In this thesis, we have investigated the prevalence of EIB and EILO in a gen-eral population of adolescents with two separate, standardized exercise-tests. Main findings are that EIB and EILO are important diagnoses in adolescents with exercise-induced dyspnoea. EILO is not uncommon and should be inves-tigated early in the diagnostic process if EIB and asthma is not diagnosed or if medical therapy, despite a diagnosis, does not provide symptom relief. We also confirmed that EIB and EILO can coexist.

For EILO the investigation was conducted according to gold standard with CLE-test and the use of CLE-score as diagnostic tool. By comparing CLE-score and EILOMEA we found the methods compatible. EILOMEA offers a possibility to study laryngeal differences and changes in detail. By use of these two validated methods we could further establish that there were no dif-ferences in laryngeal findings between healthy controls and adolescents that remained undiagnosed after EIB-and CLE-tests. Only adolescents with EILO differed from other participants regarding glottic and supraglottic measure-ments obtained with EILOMEA. This strengthens the use of current diagnostic procedures as it indicates that, when typical laryngeal findings are seen during exercise in a person with dyspnoea, and who recognizes these symptoms as their main problem, we diagnose EILO correctly.

A correct diagnosis of EILO is mandatory when it comes to offering treat-ment to EILO-patients. Surgical treatment should exclusively be offered to highly motivated patients with severe supraglottic EILO, where the diagnosis has been established during a CLE-test. Surgically treated patients in our study reported reduced exercise-induced dyspnoea and a positive effect on physical activity levels.

51

Populärvetenskaplig sammanfattning

Ansträngningsutlöst laryngeal obstruktion, på engelska “exercise-induced la-ryngeal obstruction”, EILO, är en av flera orsaker till att ungdomar kan upp-leva andnöd under träning. När behovet av syre ökar under kraftig ansträng-ning andas vi snabbare och djupare, och för att få in så mycket luft som möjligt övergår vi till att andas mer genom munnen än genom näsan. Struphuvudet och stämbanden är den trängsta delen av våra övre luftvägar. När andnings-frekvensen ökar går stämbanden isär så att luften lättare ska passera ner genom luftstrupen till lungorna. EILO innebär att det uppstår en förträngning i strup-huvudet i samband med ansträngning. Förträngningen består oftast av två en-skilda, eller tillsammans uppkomna rörelser: att stämbanden går ihop istället för att gå isär, och att mjuka slemhinneveck som sitter en bit ovanför stäm-banden i struphuvudet faller in mot mitten vid inandning och stör luftflödet. Vi vet ännu inte varför det här uppstår hos vissa när de anstränger sig hårt. Tillståndet är vanligast hos kvinnor i tonåren.

För att ta reda på om någon har EILO måste struphuvudet undersökas sam-tidigt som personen anstränger sig kraftigt. Om man undersöker struphuvudet i vila ser allt normalt ut. Utan en korrekt undersökning kan EILO lätt misstas för att vara ansträngningsastma och det är vanligt att personer med EILO har behandlats med astmamedicinering utan att det haft någon effekt. Vid under-sökning förs en tunn fiberoptisk kikare in via näsan så att spetsen på kikaren kan placeras strax ovanför struphuvudet. Den fiberoptiska kikaren är kopplad till en kamera så att det går att följa vad som händer i struphuvudet medan personen som testas springer på ett löpband (på foto s. 21 kan man se ungefär hur långt ner kikaren når eftersom ljuset från kikaren syns på halsen).

I det första arbetet jämfördes två olika metoder som kan användas för att dia-gnostisera EILO. Den ena metoden (CLE-score) används i samband med att ansträngningstestet görs och innebär att man graderar hur kraftigt förträng-ningen som uppstår i struphuvudet är på en skala från 0-3. Den andra metoden (EILOMEA) innebär att man tittar på en filminspelning av vad som hänt i struphuvudet, tar ut en bild från filmen som visar den maximala förträng-ningen som uppstått vid inandning och sedan, med hjälp av ett dataprogram, räknar ut hur kraftig förträngningen är. Vid jämförelse av metoderna fann vi att bedömningar med EILOMEA och CLE-score överensstämde för de för-trängningar i struphuvudet som kan ha klinisk betydelse.

52

I det andra arbetet undersöka vi hur vanligt EILO är bland ungdomar i Upp-sala. Före den aktuella studien hade vi skickat ut en enkät till alla ungdomar födda 1997-1998, de var vid den tidpunkten 12-13 år gamla, och frågat om de hade ansträngningsutlösta andningsbesvär. Undersökningen visade att 14,3% upplevde sådana andningsbesvär när de tränade. Av de som svarat på enkäten valdes 150 ungdomar ut slumpmässigt som först fick göra ett ansträngnings-test för att undersöka förekomst av förträngning i de små perifera luftvägarna, och därefter gjordes vårt löpbandstest med filmning av struphuvudet. Av de 150 var en tredjedel friska ungdomar inte hade några besvär alls. Vi som un-dersökte ungdomarna fick dock inte veta vem som hade andningsbesvär eller inte förrän de hade testats klart. Det var 125 ungdomar som genomförde båda testerna. Utifrån fynden vi observerat vid de två testerna beräknades före-komsten av perifer förträngning (ofta benämnt ansträngningsastma) vara 19,2% och förekomsten av EILO vara 5,7%. Eftersom EILO är så pass fre-kvent förekommande hos de som har ansträngningsutlösta andningsbesvär drar vi slutsatsen att det är viktigt att undersöka struphuvudet under ansträng-ning om inte astmamediciner hjälper.

Det tredje arbetet var en fortsättning på den föregående studien. Det var inte alla som hade uppgett ansträngningsutlösta besvär som fick någon av de dia-gnoser vi undersökte. Vi undrade då om förklaringen till deras upplevda and-ningsbesvär ändå kunde vara relaterad till struphuvudet. Genom att undersöka filmerna som spelats in under ansträngningstesterna togs stillbilder ut från bör-jan och slutet av testet och undersöktes med EILOMEA. Vi kunde då mäta hur mycket trängre det hade blivit i struphuvudet under tiden som de ansträngt sig. Vi kunde inte hitta någon skillnad mellan de som hade angett besvär och de friska kontrollerna och drog slutsatsen att orsaken till de upplevda and-ningsbesvären inte sitter i struphuvudet.

Om ansträngningen ger upphov till en kraftig förträngning på grund av att slemhinnevecken lägger sig framför ingången till luftstrupen så kan dessa veck minskas ner genom operation. I det fjärde arbetet gjorde vi en enkätun-dersökning där vi ställde frågor till patienter som tidigare remitterats till vår klinik för utredning av andningsbesvär under ansträngning. Vi jämförde sva-ren vi fick med de svar som patienterna angett då de genomförde ansträng-ningstestet. Vi var intresserade av att se hur de som opererats mådde jämfört med de som inte opererats. Undersökningen visade att de som opererats hade mindre andningsbesvär och att de kunde anstränga sig mer än tidigare. De som inte hade opererats uppgav till stor del oförändrade besvär och många hade försökt anpassa sig till besvären genom att hitta andra former att träna än de gjort tidigare. Resultaten tyder på att operation är effektivt hos de som har en kraftig förträngning i struphuvudet och att det är viktigt att göra en noggrann undersökning så att rätt patienter får rätt behandling.

53

Acknowledgements

Först vill jag tacka alla fantastiska ungdomar och föräldrar som deltagit i våra undersökningar! Sedan vill jag rikta ett varmt tack till

Leif Nordang, min huvudhandledare och stora inspirationskälla under mina doktorandstudier. Tack för att du frågade mig om jag ville forska inom det här området och för all stöttning under åren. Utan din energi och enorma idérike-dom skulle den här avhandlingen inte blivit av och EILO skulle förmodligen fortfarande vara ganska okänt. Malin Svensson, min bihandledare. Du har inte hunnit vara min handledare i det här projektet så länge men väl i det kliniska arbetet. Tack för att du alltid delar med av dina kunskaper och för sättet som du gör det på. Henrik Johansson, medförfattare och medarbetare i ANDAS-projektet, för ditt stora engagemang och fantastiskt positiva inställning under hela projektet. Övriga otroligt kunniga kollegor och medarbetare i ANDAS-projektet, Chris-ter Janson, Margareta Emtner, Lennart Nordvall, Andrei Malinovschi och Hans Hedenström. Det har varit enormt kul och lärorikt att genomföra det här projektet tillsammans med er. Övriga inblandade i ANDAS-projektet som gjorde förarbetet med datain-samling och träffade alla ungdomar innan de kom till oss. Medförfattare i Norge, Jon-Helge Heimdal, Robert Maat, Tomas Halvorsen och Ola Drange Røksund. Medförfattare i Danmark, Pernille Christensen och Niels Rasmussen. Alla fina kollegor på Öron- Näs och Halskliniken för att ni alla bidrar till den goda stämning vi har. Jag trivs verkligen att jobba med er! Övrig personal på kliniken för att ni alltid får mig att känna mig så välkommen var jag än är.

54

Det finns även många personer som inte varit direkt involverade i min forsk-ning men som har spelat en stor roll för mig under min doktorandtid och som i högsta grad har bidragit till att jag utvecklat ett vetenskapligt förhållnings-sätt. Tack Ulf, Göran, Lina, Mats, Gunnel, Viola, Andrea och Jonas. Ett särskilt tack till Henrik, Madeleine och Daniel. För er positiva grundsyn, styrka och ihärdighet. För att ni visar vad gemenskap och gemensamma an-strängningar kan leda till. Sedan vill jag tacka min underbara familj Min mamma Monica och min pappa Björn för att ni lärt mig att stå upp för det jag tror på och för allt stöd i stort som smått genom åren. Tack också för ert enorma tålamod när jag behöver hjälp med något i sista minuten. Mina svärföräldrar Lilian och Sture för er värme och alla trevliga år, och för att jag från första stund känt mig som hemma hos er.

Mina barn, Tuss, Johanna, Mirja och Simona. Jag tycker så oändligt mycket om er, ni är min lyckofyrklöver. Min make och bästa vän Peter. Jag har nu skrivit en hel bok och tycker att jag kunnat få fram det budskap jag vill ganska bra. När jag ska försöka formulera vad jag känner för dig och vad du betyder för mig så finns det inga ord som räcker till att förmedla det jag vill säga. Jag vet att du förstår vad menar.

55

References

1. Heimdal JH, Røksund OD, Halvorsen T, Olofsson J. Continuous laryngoscopy exercise test: a method for visualizing laryngeal dysfunction during exercise. Laryngoscope. 2006;116(1):52-7.

2. Christensen PM, Heimdal JH, Christopher KL, Bucca C, Cantarella G, Friedrich G, et al; ERS/ELS/ACCP Task Force on Inducible Laryngeal obstructions. ERS/ELS/ACCP 2013 international consensus conference nomenclature on in-ducible laryngeal obstructions. Eur Respir Rev. 2015;24(137):445-50.

3. Maat RC, Røksund OD, Halvorsen T, Skadberg BT, Olofsson J, Ellingsen TA, et al. Audiovisual assessment and validity of observations. Eur Arch Otorhinolaryn-gol. 2009;266(12):1929-36.

4. Røksund OD, Maat RC, Heimdal JH, Olofsson J, Skadberg BT, Halvorsen T. Exercise induced dyspnea in the young. Larynx as the bottleneck of the airways. Respir Med. 2009;(12):1911-8.

5. Christopher KL, Wood RP 2nd, Eckert RC, Blager FB, Raney RA, Souhandra JF. Vocal-cord dysfunction presenting as asthma. N Engl J Med. 1983;308(26):1566-70.

6. Landwehr LP, Wood RP 2nd, Blager FB, Milgrom H. Vocal cord dysfunction mimicking exercise-induced bronchospasm in adolescents. Pediatrics. 1996;98(5):971-4.

7. Fitch KD, Sue-Chu M, Anderson SD, Boulet LP, Hancox RJ, McKenzie DC, et al. Asthma and the elite athlete: summary of the International Olympic Commit-tee´s consensus conference, Lausanne, Switzerland, January 22-24, 2008. J Al-lergy Clin Immunol. 2008;122(2):254-60, 260.e1-7.

8. Weiler JM, Anderson SD, Randolph C, Bonini S, Craig TJ, Pearlman DS, et al. American Academy of Allergy, Asthma and Immunology; American College of Allergy, Asthma and Immunology. Pathogenesis, prevalence, diagnosis, and management of exercise-induced bronchoconstriction: a practice parameter. Ann Allergy Asthma Immunol. 2010;105(6 Suppl):S1-47.

9. Dunglison R. In The Practice of Medicine vol II. 1842. Philadelphia PA: Lea and Blanchard page 258.

10. MacKenzie M. In Use of laryngoscopy in diseases of the throat. 1869. Philadel-phia, Lindsey and Blackeston pp 248-249.

11. Osler W. In The principles and practice of medicine, 9th edition: New York and London: D Appleton and company page 607.

12. Patterson R, Schantz M, Horton M. Munchausens´s stridor: non-organic laryn-geal obstruction. Clin Allergy. 1974;4(3):307-10.

13. Downing ET, Braman SS, Fox MJ, Corrao WM. Factitious asthma. Physiological approach to diagnosis. JAMA. 1982;248(21):2878-81.

14. Lakin RC, Metzger WJ, Haughey BH. Upper airway obstruction presenting as exercise-induced asthma. Chest. 1984;86(3):499-501.

56

15. Bent JP 3rd, Miller DA, Kim JW, Bauman NM, Wilson JS, Smith RJ. Pediatric exercise-induced laryngomalacia. Ann Otol Rhinol Laryngol. 1996;105(3):169-75.

16. Naito A, Niimi S. The larynx during exercise. Laryngoscope. 2000;110(7):1147-50.

17. Yeatts K, Davis KJ, Sotir M, Herget C, Shy C. Who gets diagnosed with asthma? Frequent wheeze among adolescents with and without a diagnosis of asthma. Pe-diatrics. 2003;111(5 Pt 1):1046-54.

18. Pearce N, Aït-Khaled N, Beasley R, Mallol J, Keil U, Mitchell E, et al. ISAAC Phase Three Study Group. Worldwide trends in the prevalence of asthma symp-toms: phase III of the International Study of Asthma and Allergy in Childhood (ISAAC). Thorax. 2007;62(9):758-66.

19. Johansson H, Norlander K, Hedenström H, Janson C, Nordang L, Nordvall L, Emtner M. Exercise-induced dyspnea is a problem among the general adolescent population. Respir Med. 2014;108(6):852-8.

20. Weinberger M, Abu-Hasan M. Pseudo-asthma: when cough, wheezing, and dysp-nea are not asthma. Pediatrics. 2007;120(4):855-64.

21. Seear M, Wensley D, West N. How accurate is the diagnosis of exercise-induced asthma among Vancouver Schoolchildren? Arch Dis Child. 2005;90(9):898-902.

22. Abu-Hasan M, Tannous B, Weinberger M. Exercise-induced dyspnea in children and adolescents: if not asthma then what? Ann Allergy Asthma Immunol. 2005;94(3)366-71.

23. Weiler JM, Bonini S, Coifman R, Craig T, Delgado L, Capao-Felipe M, et al. Ad Hoc Committee of Sports Medicine Committee of American Academy of Al-lergy, Asthma & Immunology. American Academy of Allergy, Asthma & Immu-nology Work Group report: exercise-induced asthma. J Allergy Clin Immunol. 2007;119(6):1349-58.

24. De Baets F, Bodart E, Dramaix-Wilmet M, Van Daele S, de Bildering G, Masset S, et al. Exercise-induced respiratory symptoms are poor predictors of broncho-constriction. Pediatr Pulmonol. 2005;39(4):301-5.

25. Hallstrand TS, Curtis JR, Koepsell TD, Martin DP, Schoene RB, Sullivan SD, et al. Effectiveness of screening examinations to detect unrecognized exercise-in-duced bronchoconstriction. J Pediatr. 2002;141(3):343-8.

26. Anthracopoulos MB, Fouzas S, Papadopoulos M, Antonogeorgos G, Papadi-mitriou A, Panagiotakos DB, et al. Physical activity and exercise-induced bron-choconstriction in Greek schoolchildren. Pediatr Pulmonol. 2012;47(11):1080-7.

27. Busquets RM, Antó JM, Sunyer J, Sancho N, Vall O. Prevalence of asthma-re-lated symptoms and bronchial responsiveness to exercise in children aged 13-14 years in Barcelona, Spain. Eur Respir J. 1996;9(10):2094-8.

28. Lødrup Carlsen KC, Håland G, Devulapalli CS, Munthe-Kaas M, Pettersen M, Granum B, et al. Asthma in every fifth child in Oslo, Norway: a 10-year follow up of a birth cohort study. Allergy. 2006;61(4):454-60.

29. Bush A, Fleming L. Is asthma overdiagnosed? Arch Dis Child. 2016;101(8):688-9.

30. Melani AS, Ciarleglio G, Pirelli M, Sestini P. Perception of dyspnea during ex-ercise-induced bronchoconstriction. Respir Med. 2003;97(3):221-7.

31. Teixeira RN, Teixeira LR, Costa LA, Martins MA, Mickleborough TD, Carcalho CR. Exercise-induced bronchoconstrictions in elite long-distance runners in Bra-zil. J Bras Pneumol. 2012;38(3);292-8.

57

32. Parsons JP, Hallstrand TS, Mastronarde JG, Kaminsky DA, Rundell KW, Hull JH, et al. American Thoracic Society Subcommittee on Exercise-induced Bron-choconstriction. An official American Thoracic Society clinical practice guide-line: exercise-induced bronchoconstriction. Am J Respir Crit Care Med. 2013;187(9):1016-27.

33. Christensen PM, Thomsen SF, Rasmussen N, Backer V. Exercise-induced laryn-geal obstructions: prevalence and symptoms in the general public. Eur Arch Oto-rhinolaryngol. 2011;268(9):1313-9.

34. Morris MJ, Deal LE, Bean DR, Grbach VX, Morgan JA. Vocal cord dysfunction in patients with exertional dyspnea. Chest. 1999;116(6):1676-82.

35. Nielsen EW, Hull JH, Backer V. High prevalence of exercise-induced laryngeal obstruction in athletes. Med Sci Sports Exerc. 2013;45(11):2030-5.

36. Olin JT, Clary MS, Deasdorff EH, Johnston K, Morris MJ, Sokoya M, et al. In-ducible laryngeal obstruction during exercise: moving beyond the vocal cords with new insights. Phys Sportsmed. 2015;43(1):13-21.

37. Christopher KL, Morris MJ. Vocal cord dysfunction, paradoxic vocal fold mo-tion, or laryngomalacia? Our understanding requires an interdisciplinary ap-proach. Otolaryngol Clin North Am. 2010;43(1):43-66. viii.

38. Maat RC, Hilland M, Røksund OD, Halvorsen T, Olofsson J, Aarstad HJ, Heim-dal JH. Exercise-induced laryngeal obstruction: natural history and effect of sur-gical treatment. Eur Arch Otorhinolaryngol. 2011;268(10):1485-92.

39. Maat RC. Exercise-induced laryngeal obstruction. Diagnostic procedures and therapy. University of Bergen, Norway 2011.

40. Hanks CD, Parsons J, Benninger C, Kaeding C, Best TM, Phillips G, et al. Etiol-ogy of dyspnea in elite and recreational athletes. Phys Sportsmed. 2012;40(2):28-33.

41. Brugman SM, Simons SM. Vocal cord dysfunction: don´t mistake it for asthma. Phys Sportsmed. 1998;26(5):63-85.

42. McFadden ER Jr, Zawadski DK. Vocal cord dysfunction masquerading as exer-cise-induced asthma. A physiologic cause for “choking” during athletic activities. Am J Respir Crit Care Med. 1996;153(3):942-7.

43. Rundell KW, Spiering BA. Inspiratory stridor in elite athletes. Chest. 2003;123(2):468-74.

44. Hull JH. Not all wheeze is asthma: time for patients to exercise their rights. Thorax. 2015;70(1):7-8.

45. Bush A, Fleming L. Diagnosis and management of asthma in children. BMJ. 2015;350:h996.

46. Andrianopoulos MV, Gallivan GJ, Gallivan KH. PVCM, PVCD, EPL, and irri-table larynx syndrome: what are we talking about and how do we treat it? J Voice. 2000;14(4):607-18.

47. Traister RS, Fajt ML, Petrov AA. The morbidity and cost of vocal cord dysfunc-tion misdiagnosed as asthma. Allergy Asthma Proc. 2016;37(2):25-31.

48. van Gent R, van Essen LE, Rovers MM, Kimpen JL, van der Ent CK, de Meer G. Quality of life in children with undiagnosed and diagnosed asthma. Eur J Pe-diatr. 2007;166(8):843-8.

49. Corder K, Ogilvie D, van Sluijs EM. Invited commentary: Physical activity over the life course – whose behaviour changes, when, and why? Am J Epidemiol. 2009;170(9):1078-81; discussion 1082-3.

50. Reidenbach MM. Aryepiglottic fold: normal topography and clinical implica-tions. Clin Anat. 1998;11(4):223-35.

51. Baughman RP, Loudon RG. Stridor: differentiation from asthma or upper airway noise. Am Rev Respir Dis. 1989;139(6)1407-9.

58

52. Meslier N, Charbonneau G, Racineaux JL. Wheezes. Eur Respir J. 1995:8(11):1942-8.

53. Bartlett D Jr. Respiratory functions of the larynx. Physiol Rev. 1989;69(1):33-57. 54. Brancatisano TP, Dodd DS, Engel LA. Respiratory activity of posterior cricoary-

tenoid muscle and vocal cords in humans. J Appl Physiol Respir Environ Exerc Physiol. 1984;57(4):1143-9.

55. Baz M, Haji GS, Menzies-Gow A, Tanner RJ, Hopkinson NS, Polkey MI, et al. Dynamic laryngeal narrowing during exercise: a mechanism for generating in-trinsic PEEP in COPD? Thorax. 2015;70(3):251-7.

56. Halldin M, Lindahl S. Anestesi. Liber 2000. 57. Hurbis CG, Schild JA. Laryngeal changes during exercise and exercise-induced

asthma. Ann Otol Rhinol Laryngol. 1991;100(1):34-7. 58. Ferris BG Jr, Mead J, Opie LH. Partitioning of respiratory flow resistance in man.

J Appl Physiol. 1964;19:653-8. 59. Hudson AL, Gandevia SC, Butler JE. The effect of lung volume on the co-ordi-

nated recruitment of scalene and sternomastoid muscles in humans. J Physiol. 2007;584(Pt 1):261-70.

60. England SJ, Bartlett D Jr. Changes in respiratory movements of the human vocal cords during hyperpnea. J Appl Physiol Respir Environ Exerc Physiol. 1982;52(3):780-5.

61. Beaty MM, Wilson JS, Smith RJ. Laryngeal motion during exercise. Laryn-gocope. 1999;109(1):136-9.

62. Strohl KP, Butler JP, Malhotra A. Mechanical properties of the upper airway. Compr Physiol. 2012;2(3):1853-72.

63. Bent J. Pediatric laryngotracheal obstruction: current perspectives on stridor. Laryngoscope. 2006;116(7):1059-70.

64. Carlsen KH, Engh G, Mørk M. Exercise-induced bronchoconstriction depends on exercise load. Respir Med. 2000;94(8):750-5.

65. Carlsen KH. The breathless adolescent asthmatic athlete. Eur Respir J. 2011;38(3):713-20.

66. Collett PW, Brancatisano T, Engel LA. Changes in the glottic aperture during bronchial asthma. Am Rev Respir Dis. 1983;128(4):719-23.

67. Morris MJ, Christopher KL. Diagnostic criteria for the classification of vocal cord dysfunction. Chest. 2010;138(5):1213-23.

68. Joos GF, O´Connor B, Anderson SD, Chung F, Cockcroft DW, Dahlén B, et al; ERS Task Force. Indirect airway challenges. Eur Respir J. 2003;21(6):1050-68.

69. Rundell KW, Anderson SD, Sue-Chu M, Bougault V, Boulet LP. Air quality and temperature effects on exercise-induced bronchoconstriction. Compr Physiol. 2015;5(2):579-610.

70. Fowler SJ, Thurston A, Chesworth B, Cheng V, Constantinou P, Vyas A, et al. The VCDQ – a Questionnaire for symptom monitoring in vocal cord dysfunction. Clin Exp Allergy. 2015;45(9):1406-11.

71. Gartner-Schmidt JL, Shembel AC, Zullo TG, Rosen CA. Development and vali-dation of the Dyspnea Index (DI): a severity index for upper airway-related dysp-nea. J Voice. 2014;28(6):775-82.

72. Traister RS, Fajt ML, Landsittel D, Petrov AA. A novel scoring system to distin-guish vocal cord dysfunction from asthma. J Allergy Clin Immunol Pract. 2014;2(1):65-9.

73. Tervonen H, Niskanen MM, Sovijärvi AR, Hakulinen AS, Vilkman EA, Aalto-nen LM. Fiberoptic videolaryngoscopy during bicycle ergometry: a diagnostic tool for exercise-induced vocal cord dysfunction. Laryngoscope. 2009;119(9):1776-80.

59

74. Panchasara B, Nelson C, Niven R, Ward S, Hull JH. Lesson of the month: Row-ing-induced laryngeal obstruction: a novel cause of exertional dyspnea: charac-terised by direct laryngoscopy. Thorax. 2015;70(1):95-7.

75. Olin JT, Clary MS, Fan EM, Johnston KL, State CM, Strand M, Christopher KL. Continuous laryngoscopy quantitates laryngeal behaviour in exercise and recov-ery. Eur Respir J. 2016;48(4):1192-1200.

76. Higenbottam T. Narrowing of glottis opening in humans associated with experi-mentally induced bronchoconstriction. J Appl Physiol Respir Environ Exerc Physiol. 1980;49(3):403-7.

77. Jain S, Bandi V, Officer T, Wolley M, Guntpalli KK. Role of vocal cord function and dysfunction in patients presenting with symptoms of acute asthma exacerba-tion. J Asthma. 2006;43(3):207-12.

78. Christensen PM, Thomsen SF, Rasmussen N, Backer V. Exercise-induced laryn-geal obstructions objectively assessed using EILOMEA. Eur Arch Otorhino-laryngol. 2010;267(3):401-7.

79. Christensen PM, Maltbæk N, Jørgensen IM, Nielsen KG. Can flow-volume loops be used to diagnose exercise-induced laryngeal obstructions? A comparison study examining the accuracy and inter-rater agreement of flow volume loops as a di-agnostic tool. Prim Care Respir J. 2013;22(3)306-11.

80. Christensen PM, Rasmussen N. Eucapnic voluntary hyperventilation in diagnos-ing exercise-induced laryngeal obstructions. Eur Arch Otorhinolaryngol. 2013;270(12):3107-13.

81. Perkins PJ, Morris MJ. Vocal cord dysfunction induced by methacholine chal-lenge testing. Chest. 2002;122(6):1988-93.

82. Walsted ES, Hull JH, Sverrild A, Porsbjerg C, Backer V. Bronchial provocation testing does not detect exercise-induced laryngeal obstruction. J Asthma. 2017;54(1):77-83.

83. Bucca C, Rolla G, Brussino L, De Rose V, Bugiani M. Are asthma-like symptoms due to bronchial or extrathoracic airway dysfunction? Lancet. 1995;346(8978):791-5.

84. Vertigan AE, Bone SL, Gibson PG. Laryngeal sensory dysfunction in laryngeal hypersensitivity syndrome. Respirology. 2013;18(6):948-56.

85. Hull JH, Backer V, Gibson PG, Fowler SJ. Laryngeal Dysfunction: Assessment and Management for the Clinician. Am J Respir Crit Care Med. 2016;194(9):1062-1072.

86. Aviv JE, Martin JH, Kim T, Sacco RL, Thomson JE, Diamond B, et al. Laryn-gopharyngeal sensory discrimination testing and the laryngeal adductor reflex. Ann Otol Rhinol Laryngol. 1999;108(8):725-30.

87. Cukier-Blaj S, Bewley A, Aviv JE, Murry T. Paradoxical vocal fold motion: a sensory-motor laryngeal disorder. Laryngoscope. 2008;118(2):367-70.

88. Husein OF, Husein TN, Gardner R, Chiang T, Larson DG, Obert K, et al. Formal Psychological testing in patients with paradoxical vocal fold dysfunction. Laryn-goscope. 2008;118(4):740-7.

89. Powell DM, Karanfilov BI, Beechler KB, Treole K, Treudeau MD, Forrest LA. Paradoxical vocal cord dysfunction in juveniles. Arch Otolaryngol Head Neck Surg. 2000;126(1):29-34.

90. Maturo S, Hill C, Bunting G, Baliff C, Ramakrishna J, Scirica C, et al. Pediatric paradoxical vocaö-fold motion: presentation and natural history. Pediatrics. 2011;128(6):e1443-9.

91. Giglani L, Atkinson S, Hosanagar A, Guglani L. A systematic review of psycho-logical interventions for adult and pediatric patients with vocal cord dysfunction. Front Pediatr. 2014;2:82.

60

92. Leo RJ, Konakanchi R. Psychogenic Respiratory Distress: A case of Paradoxical Vocal Cord Dysfunction and Literature Review. Prim Care Companion J Clin Psychiatry. 1999;1(2):39-46.

93. McQuaid EL, Spieth LE, Spirito A. The pediatric psychologist´s role in differen-tial diagnosis: vocal cord dysfunction presenting as asthma. J Pediatr Psychol. 1997;22(5):739-48.

94. Elshami AA, Tino G. Coexistent asthma and functional upper airway obstruction. Case reports and review of the literature. Chest. 1996:110(5):1358-61.

95. Anbar RD. Hypnosis in pediatrics: applications at a pediatric pulmonary center. BMC Pediatr. 2002;3;2:11.

96. Sullivan MD, Heywood BM, Beukelman DR. A treatment for vocal cord dys-function in female athletes: an outcome study. Laryngoscope. 2001;111(10):1751-5.

97. Kayani S, Shannon DC. Vocal cord dysfunction associated with exercise in ado-lescent girls. Chest. 1998;113(2):540-2.

98. Rameau A, Foltz RS, Wagner K, Zur KB. Multidisciplinary approach to vocal cord dysfunction diagnosis and treatment in one session: a single institutional outcome study. Int J Pediatr Otorhinolaryngol. 2012;76(1):31-5.

99. Newsham KR, Klaben BK, Miller VJ, Saunders JE. Paradoxical vocal-cord dys-function: Management in athletes. J Athl Train. 2002;37(3):352-328.

100. Chiang T, Marcinow AM, de Silva BW, Ence BN, Lindsey SE, Forrest LA. Ex-ercise-induced paradoxical vocal fold motion disorder: diagnosis and manage-ment. Laryngoscope. 2013;123(3):727-31.

101. Doshi DR, Weinberger MM. Long-term outcome of vocal cord dysfunction. Ann Allergy Asthma Immunol. 2006;96(6):794-9.

102. Archer GJ, Hoyle JL, McCuskey A, Macdonald J. Inspiratory vocal cord dys-function, a new approach in treatment. Eur Respir J. 2000;15(3):617-8.

103. Mathers-Smith BA, Brilla LR. Inspiratory muscle training in exercise-induced paradoxical vocal fold motion. J Voice. 2005;19(4):635-44.

104. Sandnes A, Andersen T, Hilland M, Ellingsen TA, Halvorsen T, Heimdal JH, Røksund OD. Laryngeal movements during inspiratory muscle training in healthy subjects. J Voice. 2013;27(4):448-53.

105. Woolnough K, Blakey J, Pargeter N, Mansur A. Acid suppression does not re-duce symptoms from vocal cord dysfunction, where gastro-laryngeal reflux is a known trigger. Respirology. 2013;18(3):553-4.

106. Maat RC, Røksund OD, Olofsson J, Halvorsen T, Skadberg BT, Heimdal JH. Surgical treatment of execise-induced laryngeal dysfunction. Eur Arch Otorhino-laryngol. 2007;264(4):401-7.

107. Melhum CS, Walsted ES, Godballe C, Backer V. Supraglottoplasty as treatment of exercise-induced laryngeal obstruction (EILO). Eur Arch Otorhinolaryngol. 2016;273(4):945-51.

108. Røksund OD, Heimdal JH, Clemm H, Vollsæter M, Halvorsen T. Exercise-in-duced laryngeal obstruction: diagnostics and management. Pediatr Respir Rev. 2016. pii:S1526-0542(16)30063-X (Epub ahead of print).

109. Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol. 2001;37(1):153-6.

110. Miller MR, Hankison J, Brusasco V, Burgos F, Casaburi R, Coates A, et al. Standardisation of spirometry. Eur Respir J. 2005;26(2):319-38.

111. The European Community Respiratory Health Survey II. Main Questionnaire htpp://www.ecrhs.org/Quests.htm: European Community Respiratory Health Survey; 2002. Updated 25/09/07.

61

112. Murthy MN. Sampling: theory and methods. Calcutta: Statistical Publishing So-ciety; 1967;253-254.

113. Björnsdottir US, Gudmunsson K, Hjartarson H, Magnusson B, Juliusson S. Ex-ercise-induced laryngochalacia: an imitator of exercise-induced bronchospasm. Ann Allergy Asthma Immunol. 2000;85(5):387-91.

114. Benarab-Boucherit Y, Mehdioui H, Nejdar F, Delpierre S, Bouchair N, Aberkane A. Prevalence rate of exercise-induced bronchoconstriction in Annaba (Algeria) schoolchildren. J Asthma. 2011;48(5):511-6.

115. Bardagi S, Agudo A, Gonzales CA, Romero PV. Prevalence of exercise-induced airway narrowing in schoolchildren from a Mediaterranean town. Am Rev Respir Dis. 1993;147(5):1112-5.

116. Fahey JT, Bryant NJ, Karas D, Goldberg B, Destefano R, Gracco LC. Exercise-induced stridor due to abnormal movement of the arytenoid area: videoscopic diagnosis and characterization of the ”at risk” group. Pediatr Pulmonol. 2005; 39(1):51-5.

117. Castelli WA, Ramirez PC, Nasjleti CE. Linear growth study of the pharyngeal cavity. J Dent Res. 1973;52(6):1245-8.

118. Wysocki J, Kielska E, Orszulak P, Reymond J. Measurements of pre- and post-pubertal human larynx: a cadaver study. Surg Radiol Anat. 2008;30(3):191-9.

119. Yurdakul AS, Dursun B, Canbakan S, Cakaloglu A, Capan N. The assessment of validity of different asthma diagnostic tools in adults. J Asthma. 2005;42(10):843-6.

120. Mahut B, Fuchs-Climent D, Plantier L, Karila C, Refabert L, Chevalier-Bidaud B, et al. Cross-sectional assessment of exertional dyspnea in otherwise healthy children. Pediatr Pulmonol. 2014;49(8):772-81.

121. Rothman K. Epidemiology. An introduction. Oxford University Press 2nd Edi-tion. 2012.

122. Anderson SD, Pearlman DS, Rundell KW, Perry CP, Boushey H Sorkness CA, et al. Reproducibility of the airway response to an exercise protocol standardized for intensity, duration and inspired air conditions, in subjects with symptoms sug-gestive of asthma. Respir Res. 2010;11(1):120.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1300

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-315123

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017