Respiratory Muscle Weakness and Dyspnea in Thyrotoxic Patients1-3t Z !"
G. N. MCELVANEY,4 P. G. WILCOX,5 M. S. FAIRBARN, C. HILLIAM, G. E. WILKINS, P. D. PARE, and R. L. PARDY
Dyspnea at rest and in exerciseis reported as a common symptom of thyrotoxicosis (1, 2). Occasionally, the cause of dyspnea may be readily apparent as in the case of tracheal compression by a large goiter. An association has also been suggested between thyroid disease and asthma with worsening of asthma upon development of thyrotoxicosis and improvement after successful antithyroid treatment (3). However, in the majority of cases, there is no obvious reason for dyspnea in thyrotoxic patients. The mechanisms underlying dyspnea are complex but recent work suggests that dyspnea results at least in part from a perception of respiratory muscle effort (4,5). Skeletal muscle weakness has been reported in as many as 82010 of thyrotoxic patients and there is electro myographic evidence of myopathy in 93% (6). Increased dyspnea in thyrotoxic patients could therefore be due to respiratory muscle weakness secondary to myopathy. Although respiratory muscle weakness has been documented in patients with thyrotoxicosis (2, 7), in only one study were category scales used to rate breathlessness in exercise (8). These authors (8) did not measure respiratory muscle strength but suggested that seven thyrotoxic patients were not more breathless than six control subjects at a given minute ventilation (VI) or level of carbon dioxide production in exercise. We therefore sought to determine whether thyrotoxic patients are in fact more dyspneic on exertion than age- and sex-matched controls, and if so, whether the increased dyspnea is associated with respiratory muscle weakness. Methods The study group consisted of 12 thyrotoxic patients and 12 control subjects who were matched for age and sex (table 1). All hyperthyroid patients were assessed prior to antithyroid treatment by clinical examination, iodine-131 thyroid scan, and serum thyroxine levels (table 2). All control subjects were
Dyspnea on exertion is a frequently reported symptom of thyrotoxicosis. In the malori-
ty of cases, there is no obvious cause of dyspnea, but as skeletal myopathy is also common in thyrotoxic patients, it has been postulated that increased dyspnea could be secondary to respiratory muscle weakness. We sought to determine whether thyrotoxic patients were in fact more dyspneic on exertion than age- and sex-matched controls, and if so, whether the increased dyspnea was secondary to respiratory muscle weakness. The stUdy group consisted of 12 thyrotoxic patients and 12 control subjects matched for age and gender. We measured lung volumes, compliance, elastic recoil, respiratory muscle strength, maximal exercise performance, and the intensity of breathlessness (modified Borg scale) at various levels of exercise in all subjects. The respiratory muscles were weaker in patients than controls. This weakness improved in treated patients (p < 0.05) with concomitant increases in VC, IC, and TLC (all p < 0.05). Despite this, we found no differences in breathlessness intensity scores between patients and controls or in patients before and after successful antithyroid therapy. AM REV RESPIR DIS 1990; 141:1221-1227
assessed by clinical examination. Subjects with clinical evidence of cardiac failure or arrhythmia wereexcluded from the study as were those currently on beta-adrenergic receptor blocking medications. We measured lung volumes, compliance, elastic recoil, respiratory muscle strength, maximal exercise performance, and the intensity of breathlessness at various levels of exercise in both patients and subjects. Measurements were repeated in eight of the hyperthyroid patients after treatment. Four of the previously studied hyperthyroid patients were not restudied. Twolived in remote areas and could not attend tests (Patients 9 and 10). One subject refused treatment (Patient 11) and another was lost to follow-up (Patient 9). We also restudied eight control subjects about 3 months after their initial tests. These tests were timed to coincide approximately with the time interval between tests in the treated patients. Pulmonary function tests were performed in a volume displacement body plethysmograph. Volume was measured with a Krogh spirometer coupled to a linear displacement transducer (Shaevitz, Type 33; Pennsauken, NJ), and flow was measured with a pneumotachometer (Fleisch No.3) coupled to a differential pressure transducer (Sanborn 270). Transpulmonary pressure was monitored with an esophageal balloon catheter, which was positioned about 10 em from the gastroesophageal junction. Pressure was measured with a ± 100 em H 2 0 differential pressure transducer (Model 45 MP; Validyne Co., Northridge, CA). Subdivisions of lung volume were measured using the Boyles' law tech-
nique to determine thoracic gas volume. A minimum of 25 static pressure and volume points were measured during five static expiratory pressure-volume (P-V)maneuvers and the data points were fitted to the single exponential equation V = A - Be-K P as described by Colebatch (9). The exponential constant K, which describes the shape of P-V curve, and maximal elastic recoil pressure (PLmax) and recoil pressures at 90070 and 60% of total lung capacity (TLC) were calculated and also expressed as percent predicted according to the formula of Colebatch and colleagues (10). Maximal static respiratory pressures were measured as described previously (11). Each subject was seated and wearing noseclips. The apparatus for measuring the pressures consisted of a stiff rubber tube into which were inserted two cannulas. One of these cannulas was connected to a ± 300em H 2 0 differential
(Received in original form June 6, 1989 and in revised form November 10, 1989) 1 From the University of British Columbia Pulmonary Research Laboratory and Department of Medicine, S1. Paul's Hospital, Vancouver, BC, Canada. 2 Supported by the British Columbia Lung Association. 3 Correspondence and requests for reprints should be addressed to Dr. G. N. McEivaney, UBC Pulmonary Research Laboratory, S1. Paul's Hospital, 1081 Burrard Street, Vancouver, BC V6Z lY6 Canada. 4 British Columbia Lung Association Fellow. S Canadian Lung Association Fellow.
MCELVANEY. WILCOX. FAIRBARN, HILL/AM, WILKINS, PARE, AND PARDY
TABLE 1 ANTHROPOMETRIC DATA IN 12 HYPERTHYROID PATIENTS AND 12 CONTROL SUBJECTS
Number Male:female Age. yr Height. cm Weight. kg
12 3:9 33 ± 9 163 ± 7 60 ± 9
12 3:9 30 ± 6 167 ± 7 66 ± 11
Values are mean :!: standard deviation. Patients and controls were not different.
pressure transducer (Model 45-32, Validyne Co.) and the other was open to the atmosphere. This second orifice (internal diameter = 0.6 mm) permitted an air leak to prevent pressure artifact secondary to facial muscle contraction. Maximal expiratory pressure (PEmax) was measured near TLC after a maximal inspiration. Maximal inspiratory pressure (PImax) was measured near residual volume (RV)following a maximal expiration. The determinations wererepeated until three measurements with less than 5% variability and sustained for at least 1 s were recorded. The highest value obtained was reported. Exercisewas performed on an electronically braked cycleergometer (Model 844; Quinton Instruments, Seattle, WA) calibrated by dynamometer. The initial power was 33 watts and the power was increased by 16 watts at I-min intervals. Subjects exercised to their symptom limited maximal power output. Heart rate was measured by electrocardiogram. Subjects breathed through a one-way sliding valve, connected on the inspiratory port to a turbine flow transducer (Alpha Technologies, Inc., Laguna Hills, CA) for measurement of inspired ventilation. On the expiratory side, gas passed into a IO-L mixing chamber. Expired gas was sampled for measurement of mixed expired CO, concentration (Ametek Carbon Dioxide Analyzer, Model CD-34; Applied Electrochemistry, Pittsburgh, PA) and mixed expired oxygen
concentration (Beckman Oxygen Analyzer OM-ll; Beckman Instruments, Anaheim, CA). The CO, and 0, analyzers were each calibrated with three gas mixtures. Pleural pressure was measured during exercise using a differential pressure transducer (Model 4532; Validyne Co.) attached to the esophageal balloon-catheter. Measurements made during exercise were workload (W), inspiratory pleural pressure (Pipl) swings, cardiac frequency (fc), minute ventilation (VI), tidal volume (VT), breathing frequency (fb), oxygen uptake (Vo,), carbon dioxide production (Vco-), inspiratory time (TI), total respiratory time (1'1'), and breathlessness intensity during the last five breaths at each workload. Breathlessnessintensity was derived using a category scale (modified after Borg) (12). The scale (graded from 0 to 10)was shown to the subjects before exercise. They weretold that we were only interested in assessing "shortness of breath" and at the end of every workload they wereasked to point to that expression on the scale that best described how short of breath they felt at that time. Wealso calculated "breathlessness"during the last five breaths of each workload by applying the multiple linear regression equation derived by Leblanc and coworkers (13). We called this a dyspnea index. Dyspnea Index = 3.0 (PpIlPImax) + 1.2 (VI) + 4.5 (VT/VC) + 0.13 fb + 5.6 TI/TT - 6.2. Compared to Leblanc and coworkers, we used VT/TI as a measure of inspiratory flow rate (instead of VImeasured by pneumotachometer) and we also used only Pipl swings. In this equation (13) the intensity of breathlessness was correlated with the Pipl as a fraction of PImax, reflecting the pressure required to expand the lungs; inspiratory flow rate (VT/TI), reflecting velocity of muscle shortening; VT as a fraction of VC, reflecting the extent of muscle shortening; fb, reflecting frequency of respiratory muscle contraction; and duty cycle (TI/TT), reflecting the fraction of total respiratory time spent on inspiration.
TABLE 2 SERUM THYROXINE AND RADIOACTIVE IODINE UPTAKE IN PATIENTS
Patient Number 1 2 3 4 5 6 7 8 9 10 11 12
Serum Thyroxine (T4) (nmo/IL)
Radioactive Iodine (' 31 1) Uptake (%) Pretreatment
Statistical Analysis An unpaired t test was used to compare anthropometric, respiratory muscle strength, pulmonary function, pulmonary mechanics and maximal exercisedata between the original 12 hyperthyroid patients and the 12 ageand sex-matched controls. Degrees of significance were calculated using a modified sequential rejective Bonferroni procedure. Correlation coefficients were used to examine whether there were any relationships between PImax, PEmax, and serum thyroxine levels (14, 15). Paired t tests were used to compare anthropometric, respiratory muscle strength, pulmonary function, pulmonary mechanics, and maximal exercisedata pre- and posttreatment in hyperthyroid subjects. Similar testing was applied to the two sets of data in 8 normal subjects. A modified sequential rejective Bonferroni procedure was used to calculate degrees of significance. Incremental exercise data were compared in 12 hyperthyroid patients and 12 controls, in 8 hyperthyroid patients pre- and posttreatment, and in the 8 control subjects who had 2 sets of measurements. The relationships between variables were calculated by linear regression analysis (16). This permitted calculation of group relationships and confidence intervals. Group differences in intercepts, slopes, and the entire regression line were analyzed using a chi-square test (16). Results
Maximal Static Respiratory Pressures, Pulmonary Function, and Maximal Exercise Measurements in Hyperthyroid Patients and Control Subjects Maximal static respiratory pressures (PI max and PEmax) were lower in patients than in controls (figure I, table 5). There was no correlation between maximal static respiratory pressure and serum thyroxine levels. There were no differences in static lung volumes, static compliance, elastic recoil, or K values between the two groups (table 3). Maximal power (Wmax) and ViI max as a fraction of maximal oxygen uptake (V0 2max) were lower in patients than controls (table 4). There were no differences in other variables at maximal exercise (table 4).
Measurements during Exercise in Patients and Control Subjects
254 195 296 286 245 303 372 198 816 185 185 256
68 50 16 81 65 100 130 91
60 45 76 85 69 40 86 30
78 62 82 24 81 40 85 37
Radioactive iodine Radioactive iodine Radioactive iodine Radioactive iodine Radioactive iodine Propylthiouracil Propylthiouracil Propylthiouracil Radioactive iodine Radioactive iodine Refused treatment Radioactive iodine
The slope of the Vo 2- wo r k rate relationship was not different in the two .groups (p = 0.20) although the resting V02 was higher in patients (p = 0.007, figure 2A). Resting heart rate and the slope of the heart rate-work rate relationship was greater in patients (p ;= 0.006 figure fB). Minute ventilation (VI) at a given Ve0 2 in exercise was greater in patients (p < 0.005, figure 2e). This was due solely to a
THYROTOXICOSIS AND DYSPNEA
v v 200
Fig. 1. Maximum inspiratory (Plmax) and maximum expiratory (PEmax)pressures at the mouth in 12 patients and 12 controt subjects. Horizontal lines are means and interrupted vertical lines SD. Both Plmax and PEmaxwere less in patients than in controls (both p < 0.05).
Plmax Patients Controls
greater fb (I.' < 0.04, figure 3A) since VT at a given VI was less (p < O.Q1, figure 3A) in patients. However, when VT was corrected for VC (figure 3C) there was no difference between the groups (p = 0.09). Breathlessness intensity in exercise (scored on the category scale) was not different in the two groups (figure 4A). The same was true when breathlessness intensity was plotted against VI in exercise. However, a dyspnea index [calculated using the equation of Leblanc and
PEmax Patients Controls
colleagues (13)] was higher in exercise in patients (p = 0.003, figure 4B) due to a higher VI(figure 2C), higher fb (figure 3B), and PipllPImax (p = 0.001, figure 4C). Other components of the equation of Leblanc and colleagues (VT/TI, VT/VC, and TI/TT) were not different in the two groups.
Measurements in Eight Treated Hyperthyroid Patients and Eight Control Subjects Eight of the 12 patients were restudied
TABLE 3 PULMONARY FUNCTION AND MECHANICS DATA IN 12 HYPERTHYROID PATIENTS AND 12 CONTROL SUBJECTS
Control Mean SD Hyperthyroid Mean SD
PLmax (em H2O)
PL•• (em H2O)
PL•• (em H2O)
Definition of abbreviations: TLC = total lung capacity;VC = vital capacity; FRC = functional residualcapacity; IC = inspiratory capacity; RV = residual volume; Cst = static compliance; PLmax = maximal elastic recoil pressure;PL.. = elastic recoil pressure a190%of TLC; PL.. = elastic recoil pressureat 60% of TLC; K = exponentialconstantdescribingshapeof pressurevolume curve (see text). NOTE: No significant differences between control and hyperthyroid for any variable.
when their serum thyroxine (T4) levels returned to normal (table 2). Five (Patients 1 to 5) had been treated with radioactive iodine and three (Patients 6 to 8) with propylthiouracil. Three had been rendered hypothyroid following radioactive iodine treatment (Patients 3, 4, 5, table 2) but were on replacement therapy and wereclinically euthyroid with serum thyroxine levelswithin the normal range. One patient (Patient 2, table 2) still felt clinically hyperthyroid despite the fact that her T4 and thyroid-stimulating hormone levels were normal. On examination, she was still clinically hyperthyroid and a repeat radioactive iodine scan showed residual uptake over the left lobe of the thyroid. Nevertheless, she was included in the study population. In control subjects there was no change in pulmonary function, respiratory muscle strength, maximal exercise performance (table 5), or variables (including breathlessness intensity) throughout exercise on repeat tests. In patients VC, IC, TLC, PI max and PEmax (table 5) increased posttreatment. There were no changes in functional residual capacity (FRC), residual volume, the position or shape of the pressurevolume curves, or in elastic recoil following treatment in patients. Exercise Wmax (table ~) and maximal values for V02, VC02, VI, breathlessness intensity, and dyspnea index did not change. The relationship between breathlessness intensity and work rate did not change in patients' posttreatment (figure 5A), although there was a change in the line (p = 0.08) and decrease in intercept (p = 0.08) of the dyspnea index-work rate relationship posttreatment (figure 5B). None of the individual components of the dyspnea index changed posttreatment (suggesting that the decrease in this index was multifactorial in origin).
TABLE 4 MAXIMAL EXERCISE DATA (CYCLE ERGOMETRY) IN 12 HYPERTHYROID PATIENTS AND 12 CONTROL SUBJECTS
VT Breathlessness Dyspnea Pipl/MIP VTNC VTITi Intensity Index (%) (%) (Lis)
(watts) (Llmin) (wattS/L·min-') (bpm) (Llmin) (breaths/min) (L) Control Mean SD Hyperthyroid Mean SD
Definition of abbreviations: IN = work rate; vo, = oxygen uptake; fc = heart rate: fb = breathing rate; VT = tidal volume; Breathlessness intensity = dyspneascore on modified Borg scale; Dyspneaindex = dyspneascore calculated from equation of Leblanc and coworkers(seetext); Pipl = peak inspiratory pleural pressure; MIP = maximal static inspiratory pressure; VC = vital capacity; TI = inspiratory time.
• p = 0.004.
t p = 0.001.
Wort Rate ( wat U >
Car bo n D ioxide Productio n (J/m'n)
~ / . ..
Fig. 2. Changes in oxygen uptake (A) and heart rate (B) with increasing work rate in 12 patients and 12 control subjects. The change In minute ventilation with increasing CO, production is shown in (C). The solid line is the regression line for the patient data. and the interrupted line the regression line for the control subject data. (A) The resting Vo, was greater in patients and the slope of the Vo,-work rate relationsh ip was not different in patients and controls. (B) The slope of the heart rate-work rate relationsh ip was greater in patients (p = 0.006). (C) Minute ventilation was greater in patients than controls (p = 0.006).
+++ . . . ..
.fj~ljl +~....c o
IlLcP+ o ++ +P l5' () () o lDl +0 0 0 Do
• Co n t ro l.
o PaUc nU
'!1>l'" +~t" 28
r- ~ 'i'
Vent ila ti o n
THYROTOXICOSIS AND DYSPNEA
1225 TABLE 5
RESPIRATORY MUSCLE STRENGTH, MAXIMUM POWER , AND LUNG VOLUMES PRE· AND POSTIREATMENT IN 12 HYPERTHYROID PATIENTS AND 12 CONTROL SUBJECTS W max (Kpm)
101 " 39
2.4 " 0.7
5.3 " 1.2
3.9 " 1.2
4 .2 0.6
Plmax (em H2O)
PEmax (em H2O)
Patients Pretreatment Mean SD Posttreatment Mean SD Control Pretreatment Mean SD Posttreatment Mean SD
De/inirlon 0/ abbreviations: Plmo. = maximal inspiratory pressure; PErno. = maximal expiratory pressure; W mo. = maximalworkrate; IC = inspiratory capacity; TLC = tolallung capacity; VC ~ vital capacity. • p < 0.05 pre- versus posttreatment.
6 .• 0
- 2.00 ' - - - - - - - - , - - - - - . - - - - - - , - - - - - . - - - - - - , o 2 4 6 8 10 12.00
9 .20 X
6 .• 0
- 2.00 L . - - - - , - - - -- , - - - - - - - - , r - - - - - . - - - - - - - , 4 6 2 o 8 10 BREA THLESSNESS INTENSITY Fig. 5. The breathlessness intensity (A) and dyspnea index (B) changes in exercise in 8 patients with thyrotoxicosis pre- and posttreatment. There was no change in breathlessness intensity posttreatment. Dyspnea index was less (p = 0.08) posttreatment (interrupted regression line) .
Discussion In this study, we demonstrated respiratory muscle weakness in hyperthyroid patients relative to age- and sex-matched control subjects. This weakness improved significantly in treated hyperthyroid patients with concomitant increases in VC, IC, and TLC. Despite this, we found no differences in breathlessness intensity in exercise (assessed by category scaling) between patients and control subjects or in patients before and after successful antithyroid treatment. This suggests that (1) breathlessness in exercise is not greater in hyperthyroid patients than in age- and sex-matched controls, and (2) respiratory muscle weakness does not cause increased breathlessness in hyperthyroid patients. Respirologists have long reported that thyrotoxic patients complain of increased dyspnea. Peabody and Wentworth (17), as early as 1917, showed a low VC and a high RV in thyrotoxic patients who complained of dyspnea, and these findings were confirmed by other investigators (18-20) and presumed to be due to respiratory muscle weakness. Both Mier and colleagues (2) and Stein and coworkers (7) showed decreased PImax and PEmax in hyperthyroid patients, and they also showed significant increases in these pressures following treatment. Stein and coworkers also noted that while only three of thei r 13thyrotoxic patients had VC significantly less than predicted, there were large increases in VC in five out of seven patients restudied after treatment. Like Stein and colleagues, we found that mean VC in our study population was not significantly different from that in controls (table 3), but that there were significant increases in VC, IC , and TLC following treatment for thyrotoxicosis (table 5). Unlike Stein and colleagues, we did not find reduced compliance in our thyrotoxic patients, and compliance and static lung recoil did not change after treatment. The decreased lung volumes in thyrotoxic patients prior to treatment may be related to respiratory muscle weakness. It has been shown that in subjects with respiratory muscle weakness, the decrease in lung volume is often less than would be suggested by the loss of muscle force (21,22). Although Stein and colleagues (7) found decreased compliance in their thyrotoxic patients, they also noted that the TLC was not significantly altered following treatment and suggested that this was evidence against lung compliance being a major determinant of the observed
changes in ve. Another possible cause of the decreased ve in thyrotoxic patients is congestion of the lungs with blood, but the application of tourniquets to hyperthyroid patients does not elevate ve to predicted or posttherapy levels (7). The widespread microatelectasis reported in patients with longstanding respiratory muscle weakness (21)probably contributed little to the low lung volumes in our patients, as elastic recoil and compliance remained the same after treatment despite changes in lung volumes. Webelieve that the changes in lung volume following treatment of thyrotoxicosis in this study resulted from changes in muscle strength and also perhaps from (unmeasured) changes in chest wall stiffness. Longstanding muscle weakness results in increased chest wall stiffness (23), and after treatment these changes may regress and add to the increased lung volumes noted. Despite the contention that respiratory muscle weakness may account for increased dyspnea in exercise in thyrotoxicosis, formal exercise testing has infrequently been applied in these patients, and only Gibbons and colleagues (8) have previously used category scales to rate dyspnea in exercise. In seven thyrotoxic patients, they found that breathlessness was less than in six control subjects at all levels of ventilation in exercise and that there was no difference in the slope of the breathlessness score- Veoz relationship in these two groups. Stein and colleagues studied thyrotoxic subjects walking on a treadmill at a set slope and speed (7). The subjects were instructed to signal by hand if they experienced any dyspnea and four of them did so (unlike euthyroid controls, none of whom felt dyspneic). Massey and coworkers (20) studied 19patients in exercise. They noted that in the majority of patients VOz values were greater than normal at all workloads although the slope of the VO r workload relationship was not different from normal. In our study, wefound, like Massey and coworkers, that the VOzworkload slope was similar in thyrotoxic patients and controls and that VOz was higher at rest and in exercise in patients (figure 2A). Interestingly, there was a decrease in the slope and intercept of the Vorworkload relationship in our patients following successful antithyroid therapy. These changes, along with the decreased maximal workload but similar VOzmax seen in thyrotoxic patients compared to control subjects, suggest less efficient work in the thyrotoxic patients (table 4). We found an increase in ventilation for
MCELVANEY, WILCOX, FAIRBARN, HILLIAM, WILKINS, PARE, AND PARDY
a given Veoz (figure 2C), with decreased VT (figure 3A) and increased fb (figure 3B) at any minute ventilation in patients. This is in agreement with the findings of others (7, 20, 24). Massey and colleagues (20) suggested that if VT were limited by muscle weakness, the subject with thyrotoxicosis could maintain alveolar ventilation only by increasing respiratory frequency and that dyspnea would result from the patient's awareness that for a given level of external work she needs a greater VIthan she did when euthyroid. In our study, Pipl swings were higher for a given VIin patients compared to controls and with this further reflection of increased respiratory drive in thyrotoxic patients one might expect increased dyspnea. Despite these findings, the breathlessness intensity-work rate (figure 4A) and breathlessness intensity-ventilation relationships were the same in patients and controls and in patients before and after successful treatment (figure 5A). We also measured Pipl as a fraction of PImax, VT as a function of Tr, VT as a fraction of ve, frequency of breathing and duty cycleduring the last fivebreaths at each workload, and applied them to the formula of Leblanc and colleagues (13) to derive a dyspnea index. Leblanc and colleagues (13) showed by using a multiple linear regression technique that knowledge of these variables explained 690/0 of the variance in the relationship between work and breathlessness. Their study was done on 20 subjects and patients with a range of exercise capacity and pulmonary function. To the extent that the relationship between these variables and the sensation of dyspnea holds true, one would expect to be able to predict the degree of breathlessness for a given work rate in an individual subject or a group of subjects. By applying Leblanc and colleagues' equation for breathlessness intensity to the results of our control and hyperthyroid patients and to the hyperthyroid patients before and after treatment, we predicted that the patients would have increased dyspnea for any given work rate and that the dyspnea should be improved posttreatment (figures 4B and 5B). Neither of these predictions was borne out. A possible explanation for the failure of the predictions with the equation of Leblanc and colleagues includes differences in patient selection. Leblanc and colleagues studied two normal subjects and 18 patients with a wide variety of clinical disorders, whereas we studied 12 normal subjects and 12 thyrotoxic patients. Leblanc and colleagues studied pa-
tients with an age range of 25 to 65 yr, FEV, 41 to 106% pred, FVe 46 to 107% pred, and PI max 25 to 180 em HzO, whereas the range of values in the present study for the same variables was age 20 to 49 yr, FEV, 88 to 130% pred, FVe 81 to 131% pred, and PImax 44 to 160 cm HzO. The more homogeneous population in the present study may have led to a decreased power of the equation of Leblanc and colleagues in the prediction of breathlessness. Another possibility is that the equation of Leblanc and colleagues does not contain all of the variables that predict the sensation of breathlessness. This possibility is supported by the data in figure 6, where dyspnea index (calculated from the equation of Leblanc and colleagues) is plotted against breathlessness intensity (modified Borg score). Since the equation of Leblanc and colleagues was proposed as a representation of the physiological correlates of the subjective sensation of shortness of breath, the relationship of dyspnea index to breathlessness intensity should be the same in patients and controls and in patients pre- and posttreatment. In fact, the dyspnea index at a given level of breathlessness intensity was less in control subjects than in patients (figure 6B) and was also less in patients posttreatment than pretreatment (figure 6A). This suggests that although the equation of Leblanc and colleagues incorporates factors representing respiratory muscle tension, velocity and length of shortening, the frequency of respiratory muscle contraction, and duration of these contractions, there are other unmeasured physiological variables that contribute to the sensation of shortness of breath. These variables decrease after treatment of thyrotoxicosis (figure 6A). As measured by category scaling, we, like Gibbons and coworkers (8), did not find increased dyspnea on exertion in thyrotoxic patients compared to age- and sex-matched controls. Despite the increased PImax and lung volumes that occurred in thyrotoxic patients, breathlessness scores in exercise remained unchanged following treatment (figure 5A). These findings seem to argue against the conventional wisdom, which suggests that dyspnea on exertion is a common symptom of thyrotoxicosis. The differences between our results and conventional wisdom might result from our small population sample or could stem from problems with application of the category scale. It could be argued that as sensations are perceived in the light of previous
THYROTOXICOSIS AND DYSPNEA
o Pre Treatment o POlt Treatment
o o PaUenti
BREATHLESSNESS INTENSITY Fig. 6. The relationship of dyspnea indexto breathlessness intensityin exercise in 8 patients withthyrotoxicosis pre- and posttreatment (A) and in 12 patients and 12 control subjects (B). At a given breathlessness intensity, the dyspnea index was less in patients' posttreatment (p < 0.05, Panel A, dashed line) and was less in control subjects (dashed line, Panel B) than in patients (p < 0.01).
experience, it is difficult to compare breathlessness scores between two distinct groups, and indeed, although the intensity of breathlessness increases progressively during graded exercise, there is considerable variability in the response from subject to subject. The response, however, is quite reproducible in a given individual (25, 26). This was confirmed by our retested controls and supports the conclusion that there was in fact no difference in the intensity of breathlessness on exercise pre- and posttreatment in thyrotoxic patients. It is also possible that we may have obtained different results had we used a different method to score dyspnea (12, 25-28). However, dyspnea is perceived in the light of previous experience. Since thyrotoxicosis is an insidious disease, it is possible that thyrotoxic patients slowly accustom themselves to their decreased muscle power and lung volumes result-
ing in a rating of dyspnea on exertion that is not different from that of age- and sexmatched controls no matter which category scale is used. In summary, we did not find an increased intensity of breathlessness in thyrotoxic patients relative to matched controls nor did we find any change in breathlessness dyspnea scores in thyrotoxic patients posttreatment despite significant improvements in respiratory muscle strength and lung volumes. References I. Wayne EJ. The diagnosis of thyrotoxicosis. Br Med J 1954; 1:411-9. 2. Mier A, Brophy C, Wass JAH, Besser GM, Green M. Reversible respiratory muscle weakness in hyperthyroidism. Am Rev Respir Dis 1989; 139:529-33. 3. Settipane GA, Schoenfeld E, Hamolsky MW, Asthma and hyperthyroidism. J Allergy Clin Immunol 1972; 49:348-55. 4. Killian KJ, Campbell EJM. Dyspnea and exercise. Annu Rev Physiol 1983; 45:465-79.
5. Killian KJ, Jones NL. The use of exercise testing and other methods in the investigation of dyspnea. Clin Chest Med 1984; 5:99-108. 6. Ramsay I. Thyrotoxic muscle disease. Postgr Med J 1968; 44:385-97. 7. Stein M, Kinmbel P, Johnson RL. Pulmonary function in hyperthyroidism. J Clin Invest 1961; 49:348-63. 8. Gibbons WJ, Levy RD, Deschamps A, et al. Exertional breathlessness in hyperthyroidism. Am Rev Respir Dis 1987; 135:A294. 9. Colebatch HJH, Ng CKY, Nikov N. Use of an exponential function for elastic recoil. J Appl Physiol 1979; 46:387-93. 10. Colebatch HJH, Greaves lA, Ng CKY. Exponential analysis of elastic recoil and aging in healthy males and females. J Appl Physiol 1979; 47:683-91. II. McElvaney G, Blackie S, Morrison NJ, Wilcox PG, Fairbarn MS, Pardy RL. Maximal static respiratory pressures in the normal elderly. Am Rev Respir Dis 1989; 139:277-81. 12. Killian KJ, Jones NL. The use of exercisetesting and other methods in the investigation of dyspnea. Clin Chest Med 1984; 5:99-108. 13. Leblanc P, Bowie DM, Summers E, Jones NL, Killian KJ. Breathlessness and exercise in patients with cardiorespiratory disease. Am Rev Respir Dis 1986; 133:21-5. 14. Fleiss JL. The design and analysis of clinical experiments. New York: John Wileyand Sons, 1986; 220-40. 15. Miller RG. Simultaneous statistical inference. 2nd ed. New York: Springer-Verlag, 1981; 67. 16. Feldman H. Families of lines: random effects in linear regression analysis. J Appl Physiol 1988; 64:1721-32. 17. Peabody FW, Wentworth JA. Clinical studies of the respiration l.V. The vital capacity of the lungs and its relation to dyspnea. Arch Intern Med 1917; 20:443-67. 18. Rabinowitch 1M. The vital capacity in hyperthyroidism with a study of the influence of posture. Arch Intern Med 1923; 31:910-5. 19. Lemon WS, Moersch HJ. Basal metabolism and vital capacity. Arch Intern Med 1924;33:130-3. 20. Massey DG, Becklake MR, McKenzie JM, Bates DV.Circulatory and ventilatory response to exercise in thyrotoxicosis. N Engl J Med 1976; 276:1104-12. 21. Gibson GJ, Pride NB, Newsom-Davis J, Loh LC. Pulmonary mechanics in patients with respiratory muscle weakness. Am Rev Respir Dis 1977; 115:389-95. 22. Braun NM, Rochester DF. Muscular weakness and respiratory failure. Am Rev Respir Dis 1979; 119(2 part 2:123-5). 23. Estenne M, Heilporn A, Delhez L, Yernault JC, Delroyer A. Chest wall stiffness in patients with chronic respiratory muscle weakness. Am Rev Respir Dis 1983; 128:1002-7. 24. Small D, de Lucas P, Levy RD, Gregory W, Cosio G. Respiratory drive during exercise in hyperthyroid patients. Clin Invest Med 1988; II (Suppl:CIII). 25. Stark RD, Gambles SA, Chattergee SS. An exercise test to assess clinical dyspnea. Estimation of reproducibility and sensitivity. Br J Dis Chest 1982; 76:269-78. 26. Altose MD. Current pulmonology. Chicago: Year Book Medical Publishers, 1986; 199-226. 27. Allose MD. Assessment and management of breathlessness. Chest 1985; 88(Suppl:77-83). 28. Mahler DA, Rosiello RA, Harver RA, Lentuie T, McGovern JF, Dauberspeck JA. Comparison of clinical dyspnea ratings and psychophysical measurements of respiratory sensation in obstructive airway disease. Am Rev Respir Dis 1987; 135:1229-33.