Usefulness of Sniff Nasal Pressure in Patients with ... - ATS Journals

Usefulness of Sniff Nasal Pressure in Patients with Neuromuscular or Skeletal Disorders DANIELA STEFANUTTI, MARIE-ROSE BENOIST, PIERRE SCHEINMANN, MICHÈLE CHAUSSAIN, and JEAN-WILLIAM FITTING Laboratoire Explorations Fonctionnelles Respiratoires, Hôpital Necker Enfants Malades, and Laboratoire Explorations Fonctionnelles Respiratoires, Hôpital St-Vincent de Paul, Paris, France; and Division de Pneumologie, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland

Inspiratory muscle strength is an important variable in patients with neuromuscular or skeletal disorders. It is usually assessed by measuring maximal inspiratory pressure (PImax), but this test may prove difficult for some patients, and low values may originate from incomplete effort or air leaks. We assessed the usefulness of the novel sniff nasal pressure (Pnsn) test in 126 patients with a neuromuscular or a skeletal disorder, aged 5 to 49 yr. Pnsn was measured in an occluded nostril during maximal sniffs performed through the contralateral nostril. All patients performed the Pnsn maneuver easily, whereas 10 young and weak patients with neuromuscular disorders could not perform the PImax maneuver. Data were analyzed for the 116 patients who could perform both tests (92 patients with neuromuscular and 24 with skeletal disorders). When expressed as percents of the predicted values, Pnsn was similar to PImax in patients with neuromuscular disorders (54 ⫾ 25% predicted [mean ⫾ SD] versus 52 ⫾ 24% predicted), and was higher than PImax in patients with skeletal disorders (70 ⫾ 25% predicted versus 61 ⫾ 27% predicted, p ⬍ 0.05). Pnsn appeared to be the main determinant of VC in patients with neuromuscular disorders, whereas the Cobb angle and PImax were the main determinants of VC in patients with skeletal disorders. We conclude that inspiratory muscle strength can be easily assessed with Pnsn in children and adults with various neuromuscular and skeletal disorders. This new muscular parameter appears particularly useful in neuromuscular disorders, in which it represents a major determinant of VC.

Neuromuscular and skeletal disorders are characterized by loss of lung volume that may ultimately lead to ventilatory failure. Respiratory muscle weakness is the main determinant of lung volume restriction in neuromuscular disorders, and may be prominent when lung volumes are still near normal (1). In contrast, respiratory muscle weakness may or may not be associated with skeletal disorders such as idiopathic scoliosis. Hence it is important to assess respiratory muscle strength in children and adults with neuromuscular or skeletal disorders. This is generally done by measuring maximal inspiratory and expiratory pressures (PImax and PEmax, respectively), against a nearly complete occlusion (2). However, these tests require a level of coordination that cannot be achieved by some patients. Moreover, the interpretation of low values of PImax and PEmax is often difficult because they may result from incomplete muscle activation, or from air leaks around the mouthpiece in the case of orofacial muscle weakness (3). The measurement of sniff nasal pressure (Pnsn) is a novel test of inspiratory muscle strength that is easy to perform and noninvasive. Nasal pressure is measured in an occluded nostril

(Received in original form on October 8, 1999 and in revised form on May 26, 2000) Supported by the University of Lausanne, Société Académique Vaudoise, and by AstraZeneca, Bayer, GlaxoWellcome, Rhône-Poulenc Rorer, and SmithKline Beecham. Correspondence and requests for reprints should be addressed to Prof. J. W. Fitting, Division de Pneumologie, CHUV, 1011 Lausanne, Switzerland. Am J Respir Crit Care Med Vol 162. pp 1507–1511, 2000 Internet address: www.atsjournals.org

during a maximal sniff through the contralateral nostril (4). This test proved easy to achieve with healthy adults and children, yielding generally higher values than PImax (5, 6). In normal subjects, the reproducibility of Pnsn was similar to that of PImax, with a coefficient of variation (CV) of 6% (7). Pnsn was found useful for confirming suspected inspiratory muscle weakness in a group of patients with miscellaneous disorders (8). Because it is easy to perform and does not require a mouthpiece, the measurement of Pnsn may be particularly appropriate in neurologic diseases. Thus, we recently found that Pnsn was better suited than PImax for assessing the progression of inspiratory muscle weakness in patients with amyotrophic lateral sclerosis (9). The aim of the present study was to assess the usefulness of Pnsn in children and adults with a variety of neuromuscular and skeletal disorders. In particular, we compared Pnsn with PImax in terms of feasibility of measurement, amplitude of pressure, and relevance with regard to loss of lung volume.

METHODS Subjects A total of 131 patients with neuromuscular or skeletal disorders were studied at three hospitals (72 at the Hôpital St-Vincent de Paul, Paris; 37 at the Hôpital Necker Enfants Malades, Paris; and 22 at the Centre Hospitalier Universitaire Vaudois, Lausanne). Measurements of Pnsn, PImax, and PEmax were made when patients came for their annual physical examinations. Five patients were excluded from analysis for reasons representing relative limitations to the validity of Pnsn (4, 10): three had airflow limitation with an FEV1/FVC ratio ⬍ 85% predicted, and two had chronic rhinitis. Five patients with known asthma were included in the study, since they had normal results of spirometry and were not currently receiving antiasthma therapy. Ten patients were unable to perform the PImax maneuver and were analyzed separately. Finally, 116 patients (45 females and 71 males) were included in the study. One hundred and six were from European and Mediterranean countries, five were of African origin, and five were of Asian origin. Verbal informed consent was obtained from all patients.

Experimental Protocol All measures of respiratory pressures were made by the same investigator in Paris and by two trained technicians in Lausanne, using the same equipment in the three hospitals at these locations. The tests were conducted in a single session, after measurement of lung volumes and spirometry. All measures were recorded with patients in the sitting position. The patients received a detailed preliminary explanation of the tests, and were vigorously coached during the test maneuvers. Armspan was used instead of height in all patients because of skeletal deformities (11). Body weight was measured and body mass index (BMI) was calculated as weight/armspan2 for all patients. The Cobb angle was measured on a frontal radiograph as the angle of intersection of perpendicular lines from the endplates of the most tilted superior and inferior vertebrae (12). Pnsn was measured in an occluded nostril during a maximal sniff through the contralateral nostril (4). A plug was made of waxed ear plugs (Calmor, Neuhausen am Rheinfall, Switzerland) that were handfastened around the tip of a catheter (I.D. ⫽ 1 mm; length ⫽ 100 cm) and were customized to the subject’s nostril. The plug was inserted into one nostril and the catheter was connected to a hand-held pres-

1508

Age, yr Weight, kg Armspan, cm BMI, kg/m2

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE TABLE 1

TABLE 2

PATIENT CHARACTERISTICS

DIAGNOSES

Neuromuscular Disorders (n ⫽ 92)

Skeletal Disorders (n ⫽ 24)

15.0 ⫾ 7.8 (4.8–48.9) 42.3 ⫾ 17.2 (13–80) 152.7 ⫾ 19.6 (106–192) 17.5 ⫾ 4.6 (9.8–33.8)

14.9 ⫾ 3.7 (9.6–25.5) 44.1 ⫾ 14.4 (21–77) 161.5 ⫾ 14.7 (130–192) 16.6 ⫾ 3.9 (10.4–24.8)

Neuromuscular disorders (n ⫽ 92)

Data Analysis The agreement between Pnsn and PImax measured at the same lung volume (PImax at FRC; PImaxFRC) was assessed by the method of plotting the difference between Pnsn and PImax against the mean of these two variables according to the method of Bland and Altman (15). Pnsn and PImaxFRC were compared through two-tailed paired t tests. Linear regression analysis was used to assess the relationships between Pnsn, PImaxFRC, Cobb angle, and VC. A stepwise multiple regression was performed with VC as the dependent variable. A value of p ⬍ 0.05 was considered significant. Unless otherwise stated, all values are given as means ⫾ SD.

RESULTS The patients’ characteristics and diagnoses are presented in Tables 1 and 2. Their ages ranged from 5 to 49 yr. Ninety-two

2000

Diagnosis

n

Muscle

Duchenne muscular dystrophy Becker muscular dystrophy Facioscapulohumeral dystrophy Myotonic dystrophy Congenital muscular dystrophy Limb girdle muscular dystrophy Congenital myopathies Nonclassified congenital myopathy

28 11 2 1 6 5 7 6

Nerve

Spinal muscular atrophy II Spinal muscular atrophy III Charcot–Marie–Tooth Disease Spinal cord trauma

12 5 6 2

Junction

Myasthenia gravis

Definition of abbreviation: BMI ⫽ body mass index. * Values are means ⫾ SD (range), n ⫽ 116.

sure meter displaying peak pressure (Pmax Mouth Pressure Monitor; P.K. Morgan, Rhainham-Gillingham, Kent, UK). Pnsn was measured during 10 maximal sniffs performed from FRC, each separated from the next by 30 s, because this number of trials was previously established as easy to perform and adequate for obtaining a maximum value (5). Readings from all maneuvers were recorded, and the highest pressure was used. Maximal static respiratory pressures were measured with a standard flanged mouthpiece connected to the same hand-held pressure meter, with computation of the average pressure sustained over 1 s. With their noses occluded by a noseclip, the patients performed three maximal inspiratory efforts from FRC and three from RV to generate PImax, with each effort separated from the next by 30 to 60 s. The patients performed three maximal expiratory efforts from FRC and three from TLC to generate PEmax. This number of trials was chosen because it corresponds to current practice (3). For each test, the results of all trials were recorded and the highest pressure was used. Reference values from our laboratory were used for Pnsn, PImax, and PEmax for adults (5) and children (6). Lung volumes were measured with the helium dilution method in 122 patients and with whole-body plethysmography in four (Model 2400; Sensormedics, Anaheim, CA), in Lausanne and at the Hôpital Necker Enfants Malades in Paris. A Pulmonet III plethysmograph (Sensormedics) was used at the Hôpital St-Vincent de Paul in Paris. The reference values of Polgar and Promadhat (13) were used for patients aged ⬍ 17 yr, and those proposed by Quanjer (14) were used for older patients.

VOL 162

Skeletal disorders (n ⫽ 24)

1

Scoliosis Funnel-shaped thorax Flat thorax

19 4 1

patients presented with a neuromuscular disorder and 24 with a skeletal disorder. Comparison of Pnsn with PImax

The within-subject CV was 20.2 ⫾ 11.1% (mean ⫾ SD) for Pnsn, 11.5 ⫾ 8.9% for PImaxFRC, and 14.6 ⫾ 11.5% for PEmaxFRC. Mean Pnsn was 59 ⫾ 25 cm H2O, corresponding to 58 ⫾ 26% predicted (Table 3). The mean PImaxFRC was 47 ⫾ 22 cm H2O, corresponding to 53 ⫾ 25% predicted. When expressed in absolute value (cm H2O), Pnsn was higher than PImaxFRC in 85 of 116 patients (Figure 1). This pattern was similar in the two groups of patients, Pnsn being higher than PImaxFRC in 66 of the 92 patients with neuromuscular disorders (p ⬍ 0.0001) and in 19 of the 24 patients with skeletal disorders (p ⬍ 0.0005). The mean difference between Pnsn and PImaxFRC was 11.2 ⫾ 19.6 cm H2O, and the limits of agreement were 50.5 cm H2O and ⫺28.0 cm H2O (Figure 2). In the patients with neuromuscular disorders, Pnsn and PImaxFRC correlated with each other (Table 4) and were similar when expressed as percents of the predicted values (54 ⫾ 25% predicted and 52 ⫾ 24% predicted, respectively) (Table 3). This held true when myopathy patients were compared with neuropathy patients. Moreover, Pnsn and PImaxFRC were similar within different neuromuscular disorders, at 41 ⫾ 20% predicted and 40 ⫾ 21% predicted, respectively, in Duchenne muscular dystrophy; 66 ⫾ 25% predicted and 70 ⫾ 21% predicted, respectively, in Becker muscular dystrophy; and 52 ⫾ 27% predicted and 55 ⫾ 22% predicted, respectively, in spinal muscular atrophy. In the patients with skeletal disorders, Pnsn and PImaxFRC correlated with each other (Table 4). However, Pnsn was higher

TABLE 3 RESPIRATORY PRESSURES IN PATIENTS WITH NEUROMUSCULAR AND SKELETAL DISORDERS Patients All (n ⫽ 116) Neuromuscular disorders (n ⫽ 92) Skeletal disorders (n ⫽ 24)

Pnsn (cm H2O)

PImaxFRC (cm H2O)

PImaxRV (cm H2O)

PEmaxFRC (cm H2O)

PEmaxTLC (cm H2O)

59 ⫾ 25 (58 ⫾ 26%) 56 ⫾ 26 (54 ⫾ 25%) 69 ⫾ 22 (70 ⫾ 25%)

47 ⫾ 22 (53 ⫾ 25%) 46 ⫾ 22 (52 ⫾ 24%) 52 ⫾ 22 (61 ⫾ 27%)

52 ⫾ 26 (55 ⫾ 26%) 50 ⫾ 26 (52 ⫾ 25%) 60 ⫾ 23 (66 ⫾ 26%)

35 ⫾ 26 (41 ⫾ 24%) 32 ⫾ 21 (37 ⫾ 23%) 48 ⫾ 21 (58 ⫾ 22%)

50 ⫾ 28 (48 ⫾ 27%) 45 ⫾ 26 (42 ⫾ 25%) 69 ⫾ 27 (69 ⫾ 27%)

Definition of abbreviations: PEmaxFRC ⫽ maximum expiratory pressure at functional residual capacity; PEmaxTLC ⫽ maximum expiratory pressure at total lung capacity; PImaxFRC ⫽ maximum inspiratory pressure at functional residual capacity; PImaxRV ⫽ maximum inspiratory pressure at residual volume; Pnsn ⫽ sniff nasal pressure. Values are means ⫾ SD. Values expressed as a percent of the predicted value are presented in parentheses.

1509

Stefanutti, Benoist, Scheinmann, et al.: Sniff Nasal Pressure in Neuromuscular Disorders TABLE 4 RELATIONSHIPS BETWEEN RESPIRATORY PRESSURES, VITAL CAPACITY, AND COBB ANGLE Neuromuscular Disorders (n ⫽ 92) Pnsn versus PImaxFRC

r ⫽ 0.717 p ⬍ 0.0001

r ⫽ 0.636 p ⬍ 0.001

Pnsn versus VC

r ⫽ 0.590 p ⬍ 0.0001

r ⫽ 0.230 ns

PImaxFRC versus VC

r ⫽ 0.414 p ⬍ 0.0001

r ⫽ 0.473 p ⬍ 0.02

Cobb angle versus VC

Figure 1. Relationship between Pnsn and PImax at FRC (PImaxFRC) in 116 patients with neuromuscular or skeletal disorders. The line represents the line of identity.

than PImaxFRC (70 ⫾ 25% predicted versus 61 ⫾ 27% predicted, respectively; p ⬍ 0.05). Correlation of Maximal Pressures with VC

Among the 32 patients with a VC ⬎ 80% predicted (21 with neuromuscular and 11 with skeletal disorders), only three had both a Pnsn and PImaxFRC ⬎ 80% predicted. Twelve had either a Pnsn or PImaxFRC ⬍ 80% predicted, and 17 had both a Pnsn and PImaxFRC ⬍ 80% predicted. In the patients with neuromuscular disorders, VC correlated with both Pnsn and with PImaxFRC, but more closely with the former (Table 4 and Figure 3). The degree of correlation was not enhanced by considering the higher value of either Pnsn or PImaxFRC obtained for each patient (r ⫽ 0.557, p ⬍ 0.001). In a stepwise multiple regression analysis with VC as the dependent variable, Pnsn accounted for 35% of the variance in VC. The sequential addition of PImaxFRC, PImaxRV, PEmaxFRC, PEmaxTLC, and age as independent variables did not further improve this correlation. In the patients with skeletal muscle disorders, VC correlated with the Cobb angle and with PImaxFRC, but not with Pnsn (Table 4 and Figure 4). In a stepwise multiple regression analysis with VC as the dependent variable, the Cobb angle and PImaxFRC accounted for 70% of its variance. The sequential addition of Pnsn, PImaxRV, PEmaxFRC, PEmaxTLC, and age as independent variables did not further improve this correlation. All of the 10 patients who performed the Pnsn maneuver successfully but who were unable to perform the PImax maneuver had neuromuscular disorders. Their age was 10.5 ⫾ 6.9 yr and their Pnsn was 29.6 ⫾ 12.9% predicted.

Figure 2. Difference between Pnsn and PImaxFRC plotted against the mean of these two variables.

Skeletal Disorders (n ⫽ 24)

r ⫽ 0.747 p ⬍ 0.0003

Definition of abbreviations: PImaxFRC ⫽ maximum inspiratory pressure at functional residual capacity; Pnsn ⫽ sniff nasal pressure.

DISCUSSION Pnsn is a new, noninvasive test of inspiratory muscle strength that has proven easy to perform with healthy children and adults (4–6). This study represents the first assessment of its feasibility and usefulness in a large group of children and adults with various neuromuscular and skeletal disorders. In this patient group, the within-subject reproducibility was slightly less for Pnsn than for the maximal static pressures PImax and PEmax. Nevertheless, all patients performed the Pnsn maneuver without difficulty, whereas 10 of them were unable to perform the PImax and PEmax maneuvers because of lack of coordination or air leaks. On average, these 10 patients were younger and weaker than the other patients. We had made a similar observation in a group of patients with amyotrophic lateral sclerosis when muscle weakness was severe (9). Although this point was not formally assessed in the present study, most patients reported that the Pnsn maneuver was easier to perform than the PImax maneuver. This is probably explained by the brief and natural character of the sniff maneuver. A correlation was found between Pnsn and PImax measured at the same lung volume (i.e., at FRC). However, in accordance with our findings in previous studies of healthy subjects (5, 6), the agreement between these two tests was relatively poor, indicating that they are not interchangeable and instead complement one another. As in normal children and adults, we found that Pnsn was higher than PImax when both were expressed as absolute values. It should be noted that important differences exist between these two tests. First, Pnsn is generated during a ballistic effort, whereas PImax requires a sustained effort. The inspiratory muscles shorten more and at higher speed during the sniff maneuver. These latter conditions should contribute to some pressure loss in comparison with the maximal static inspiratory pressure. However, the easy and natural character of the sniff maneuver probably allows subjects to activate their inspiratory muscles more completely. Second, the Pnsn and PImax maneuvers are not equivalent because the pattern of muscle activation is known to differ between the sniff and static inspiratory efforts (16). Using electromyography, Nava and colleagues (16) reported that the diaphragm was activated to a greater degree during maximal sniffs than during maximal static inspiratory efforts: in nine normal subjects, the average electromyographic activity of the diaphragm recorded during a PImax maneuver amounted to only 61% of that recorded during a maximal sniff maneuver. This different pattern of activation explains why transdiaphragmatic pressure (Pdi) is most often higher during a maximal sniff than during a PImax maneuver (17). Accordingly,

1510

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 162

2000

Figure 3. Relationships between VC and Pnsn and PImaxFRC in patients with neuromuscular disorders.

diaphragm function may be better evaluated by sniff tests like the Pnsn test than by PImax. However, considering the patient group in our study, it must be recognized that the pattern of diaphragm activation during sniffs is unknown in patients with neuromuscular disorders. In summary, because of the differences in type of effort and pattern of muscle activation in the two maneuvers, the measurements obtained from the sniff and PImax maneuvers probably reflect different aspects of inspiratory muscle function. When expressed as percents of their respective predicted values, Pnsn and PImax were similar in patients with neuromuscular disorders. This held true for the three disorders most represented in our study: Duchenne muscular dystrophy, Becker muscular dystrophy, and spinal muscular atrophy. In contrast, in patients with skeletal disorders, Pnsn was higher than PImax both when expressed in absolute values and as percents of the predicted values. Respiratory muscle function is known to be altered in primary skeletal disorders. Values of PImax and PEmax are generally normal in mild scoliosis (18, 19), but are reduced in moderate to severe scoliosis (19–21). Pdi, measured in a small number of patients with severe scoliosis, was also found to be lower than normal (20, 21). This apparent weakness has been attributed to a mechanical disadvantage imposed by the chest deformity in scoliosis rather than to actual muscle weakness. A previous study by Kinnear and coworkers (22) suggests that sniff tests may be useful for assessing diaphragm strength in these patients. In their study, sniff mouth pressure and sniff Pdi were simultaneously measured and were found to agree closely in patients with idiopathic scoliosis. In comparison with the method of measuring sniff mouth pressure, the method of measuring Pnsn offers the advantage of not being plagued by artifacts caused by oral muscles (23). As expected, VC correlated with the results of tests of inspiratory muscle strength in patients with neuromuscular disorders. The degree of correlation varied, however, with Pnsn

appearing as a more important determinant of VC than PImax. Because the highest value of inspiratory muscle strength was reflected by Pnsn in some patients and by PImax in others, we examined whether the correlation with VC would be enhanced by considering the highest value obtained for each patient, irrespective of the method used. However, we did not find this to be the case, the degree of correlation between VC and Pnsn being higher than that between VC and the higher of either Pnsn or PImax. Thus Pnsn appears more relevant than PImax for assessing inspiratory muscle strength in neuromuscular disorders. In patients with skeletal disorders, VC correlated mainly with the Cobb angle, and to a lesser degree with PImax, but not with Pnsn. The loss of lung volume in skeletal disorders can be explained by two different mechanisms: (1) the deformity and the associated subnormal compliance of the chest wall; and (2) reduced respiratory muscle strength. Previous studies addressing this question reached different conclusions. In mild idiopathic scoliosis, Smyth and associates (18) reported that VC correlated with PImax but not with the Cobb angle. In contrast, Szeinberg and colleagues (19) observed that VC correlated with the Cobb angle but not with PImax in mild to moderate scoliosis. Our data were obtained from patients with moderate to severe scoliosis. These different studies indicate that with increasing disease severity, the chest wall deformity in scoliosis becomes a major determinant of loss of lung volume. As discussed earlier, Pnsn was less affected than PImax in our patients with skeletal disorders, and consequently did not appear as a significant determinant of VC. These results are in agreement with those of Lisboa and coworkers (20), who found that VC tended to better correlate with PImax than with maximal Pdi. Thus, apart from the easy nature of its measurement, Pnsn does not appear to provide a further advantage over PImax for assessing inspiratory muscle strength in primary skeletal disorders. In summary, we found that Pnsn could be easily measured in children and adults with neuromuscular and skeletal disor-

Figure 4. Relationships between VC and Pnsn and PImaxFRC in patients with skeletal disorders.

Stefanutti, Benoist, Scheinmann, et al.: Sniff Nasal Pressure in Neuromuscular Disorders

ders, whereas measurement of PImax was not practicable for a minority of young and weak patients. As in healthy subjects, Pnsn yielded higher values than PImax. When both were expressed as percents of their respective predicted values, Pnsn was similar to PImax in neuromuscular disorders, and was higher than PImax in skeletal disorders. Pnsn appeared to be the best determinant of VC in neuromuscular disorders. Being easy to measure, without need for a mouthpiece, Pnsn appears appropriate for clinical use in these conditions. The number of investigators in our study was limited in order to ensure methodologic uniformity, since the study was performed in three hospitals. However, in our experience, measurement of the new parameter of Pnsn can be easily mastered by any pulmonary function test technician. Acknowledgment : The authors thank Antigone Askitoglu, Jean-Claude Thévenaz, Serge Wearnessyckle, Robert Roche, Christian Lebeau, and Bernard Giusti for their technical assistance.

References 1. Black, L. F., and R. E. Hyatt. 1971. Maximal static respiratory pressures in generalized neuromuscular disease. Am. Rev. Respir. Dis. 103:641–650. 2. Black, L., and R. Hyatt. 1969. Maximal respiratory pressures: normal values and relationship to age and sex. Am. Rev. Respir. Dis. 99:696–702. 3. Polkey, M. I., M. Green, and J. Moxham. 1995. Measurement of respiratory muscle strength. Thorax 50:1131–1135. 4. Héritier, F., F. Rahm, P. Pasche, and J. W. Fitting. 1994. Sniff nasal inspiratory pressure: a noninvasive assessment of inspiratory muscle strength. Am. J. Respir. Crit. Care Med. 150:1678–1683. 5. Uldry, C., and J. W. Fitting. 1995. Maximal values of sniff nasal inspiratory pressure in healthy subjects. Thorax 50:371–375. 6. Stefanutti, D., and J. W. Fitting. 1999. Sniff nasal inspiratory pressure: reference values in caucasian children. Am. J. Respir. Crit. Care Med. 159:107–111. 7. Maillard, J. O., L. Burdet, G. van Melle, and J. W. Fitting. 1998. Reproducibility of twitch mouth pressure, sniff nasal inspiratory pressure, and maximal inspiratory pressure. Eur. Respir. J. 11:901–905. 8. Hughes, P. D., M. I. Polkey, D. Kyroussis, C. H. Hamnegard, J. Moxham, and M. Green. 1998. Measurement of sniff nasal and diaphragm twitch mouth pressure in patients. Thorax 53:96–100. 9. Fitting, J. W., R. Paillex, L. Hirt, P. Aebischer, and M. Schluep. 1999.

10.

11.

12.

13. 14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

1511

Sniff nasal pressure: a sensitive respiratory test to assess progression of amyotrophic lateral sclerosis. Ann. Neurol. 46:887–893. Uldry, C., J. P. Janssens, B. De Muralt, and J. W. Fitting. 1997. Sniff nasal inspiratory pressure in patients with chronic obstructive pulmonary disease. Eur. Respir. J. 10:1292–1296. Hibbert, M. E., A. Lanigan, J. Raven, and P. D. Phelan. 1998. Relation of armspan to height and the prediction of lung function. Thorax 43:657– 659. Cobb, J. R. 1948. Outline for the study of scoliosis. Instructional Course Lectures, American Academy of Orthopedic Surgery, Ann Arbor, MI. 5:261–275. Polgar, G., and V. Promadhat. 1971. Pulmonary Function Testing in Children. WB Saunders, Philadelphia. 254. Quanjer, P. H. 1983. Standardized lung function testing: report of the Working Party on Standardization of Lung Function Tests. Bull. Eur. Physiopathol. Respir. 19(Suppl. 5):45–51. Bland, J. M., and D. G. Altman. 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307–310. Nava, S., N. Ambrosino, P. Crotti, C. Fracchia, and C. Rampulla. 1993. Recruitment of some respiratory muscles during three maximal inspiratory manoeuvres. Thorax 48:702–707. Miller, J. M., J. Moxham, and M. Green. 1985. The maximal sniff in the assessment of diaphragm function in man. Clin. Sci. 69:91–96. Smyth, R. J., K. R. Chapman, T. A. Wright, J. S. Crawford, and A. S. Rebuck. 1984. Pulmonary function in adolescents with mild idiopathic scoliosis. Thorax 39:901–904. Szeinberg, A., G. J. Canny, N. Rashed, G. Veneruso, and H. Levison. 1988. Forced vital capacity and maximal respiratory pressures in patients with mild and moderate scoliosis. Pediatr. Pulmonol. 4:8–12. Lisboa, C., R. Moreno, M. Fava, R. Ferretti, and E. Cruz. 1985. Inspiratory muscle function in patients with severe kyphoscoliosis. Am. Rev. Respir. Dis. 132:48–52. Estenne, M., E. Derom, and A. De Troyer. 1998. Neck and abdominal muscle activity in patients with severe thoracic scoliosis. Am. J. Respir. Crit. Care Med. 158:452–457. Kinnear, W. J. M., G. C. Kinnear, L. Watson, J. K. Webb, and I. D. A. Johnston. 1992. Pulmonary function after spinal surgery for idiopathic scoliosis. Spine 17:708–713. Koulouris, N., D. A. Mulvey, C. M. Laroche, E. H. Sawicka, M. Green, and J. Moxham. 1989. The measurement of inspiratory muscle strength by sniff esophageal, nasopharyngeal, and mouth pressures. Am. Rev. Respir. Dis. 139:641–646.