| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh, India.
Abstract
Study objective: To evaluate if direct substitution of arm span for height during interpretation of spirometry data leads to any significant statistical or clinical differences in Indian adults, and to compare this method with the use of height estimated indirectly from arm span.
Design: Cross-sectional.
Setting: Respiratory laboratory of a tertiary referral hospital in North India.
Participants: Two hundred twenty-eight subjects referred for spirometry.
Measurements and results: Standing height and arm span were measured for all subjects. Spirometry measurements included FVC, FEV1, FEV1/FVC, peak expiratory flow, and maximal midexpiratory flow. Predicted values for each parameter were calculated separately for height, arm span, and height estimated from fixed height:arm span ratio. Results were classified into normal, obstructive, and restrictive defects for each height, arm span, and estimated height measurement, and any abnormality was categorized as mild, moderate, or severe. Arm span exceeded height in 182 (79.82%) subjects. Thirty-seven (16.2%) and 32 (14.0%) results were classified or categorized incorrectly when arm span and estimated height were substituted respectively, for actual height, with a kappa estimate of agreement 0.779 and 0.808, respectively; 17.4% and 11.0% normal results were classified, respectively, as restrictive defects using arm span and estimated height. Limits of agreement, which were almost equally wide for both sets of data, were more than the permissible intraindividual variability for FVC and FEV1.
Conclusions: The substitution of arm span for height introduces statistically significant changes in spirometry results. Use of height estimated from arm span using fixed ratio also leads to misclassification of data, though less than that caused by use of arm span alone. Height estimated from arm span can be substituted for actual height in patients in whom height cannot be measured reliably. Where racial/ethnic norms for height and arm span correlation are not available, arm span is a reasonable surrogate for standing height.
Key Words: arm span • height • spirometry
Of all the physical measurements of the human body, lung function best correlates with standing height. Almost all equations for prediction of lung function in various populations include standing height, age, and sex as the only variables.1 However, several patients referred for pulmonary function testing are unable to stand as a result of extreme debility, structural defects, or neuromuscular weakness. Others have deformities of the axial skeleton that make measurement of height both difficult and inaccurate. Under such circumstances, arm span has been proposed as a surrogate measurement for height.2 ,3 ,4 ,5 ,6 Height may be estimated from arm span either by application of a fixed correction factor or by using regression equations. Although not absolutely accurate, height obtained by such calculations closely approximates actual height.2 ,3 ,4 ,5 ,6
Previous studies have indicated that normal individuals have only a small difference between height and arm span measurements.7 ,8 We hypothesized that direct substitution of arm span for height is likely to introduce only a minuscule error in the interpretation of pulmonary function results, which may not be of any statistical or clinical significance. The present study was carried out to test this hypothesis.
Materials and Methods
The study was conducted on a sample population of adult subjects of either sex referred to the respiratory laboratory for spirometry. All adult subjects seen consecutively over a 2-month period were enrolled prospectively into the study. Subjects in whom height could not be correctly measured due to debility, structural, or neuromuscular defects were excluded. Similarly, subjects with chest or upper limb deformities that made arm span measurements inaccurate were also excluded. We perform spirometry on subjects with a variety of medical and surgical problems, many of whom are sick, bedridden, or have skeletal deformities. Accurate measurement of both height and arm span in the same subject was therefore not possible in 204 of the 474 subjects on whom spirometry was performed over the period of study. Another 42 subjects refused to give consent and were excluded; 228 subjects were thus finally evaluated.
Standing height was measured on a barefooted subject using a stadiometer. Arm span was measured from the tips of middle fingers of maximally outstretched hands with the subjects standing and facing the wall. Both measurements were taken to the nearest centimeter. Arm span:height ratio was calculated separately for both male and female subjects, and used to estimate height from arm span measurements.
All subjects underwent evaluation using a rolling seal spirometer (Spiroflow; P. K. Morgan Ltd; Kent, UK). FVC, FEV1, FEV1/FVC ratio, peak expiratory flow (PEF), and maximal midexpiratory flow (FEF25–75%) were measured for all subjects using standard guidelines,9 and they were expressed at body temperature and pressure saturated with water vapor. The predicted value for each of the measured parameters was calculated using regression equations for healthy north Indian adults of either sex, previously derived by us (Table 1 ).10 Three sets of values based on height, arm span, and height estimated from arm span measurements were generated for each subject. The lower limits of normal were calculated as the difference between the predicted value and 1.645 times the standard error of estimate.1 FEV1, FVC, and FEV1/FVC ratio were used as basic parameters to interpret spirometry data.1 A FEV1/FVC ratio less than the lower limit of normal for that subject was taken to indicate an obstructive defect. A FVC value less than the lower limit of normal, associated with a normal FEV1/FVC ratio, was taken to indicate a restrictive defect. Severity of the obstructive or restrictive defect was calculated by using FEV1 or FVC values expressed as a percentage of the predicted normal. In the absence of any clear guidelines,1 arbitrary cutoff limits, which have been in use at our respiratory laboratory for the past several years, were used to categorize abnormal results. Mild obstruction (or restriction) was defined as FEV1 (or FVC) value > 60% of predicted but less than the lower limit of normal. Moderate obstruction (or restriction) was defined as FEV1 (or FVC) value between 40% and 60% of predicted, and severe obstruction (or restriction) as FEV1 (or FVC) value < 40% of predicted.
|
2 statistics or paired t test as required.
Agreement between various sets of measurements was evaluated by
calculating the kappa measure11
(for nominal variables)
and limits of agreement12
(for continuous variables). Results
There were 228 subjects (118 male and 110 female) with an age range of 13 to 77 years (Fig 1 ). Both height and arm span were significantly greater in male as compared with female subjects (Table 2 ). However, there was no significant difference in the absolute difference between arm span and height measurements between the two sexes. Arm span exceeded height by a mean of 4.0 (±4.4) cm; the mean of difference between arm span and height was only 2.43% of height. Arm span measurements were larger than height in 182 (79.82%), less than height in 37 (16.22%), and equal to height in 9 (3.95%) subjects. The arm span to height ratios were similar in both groups and were used to estimate height from arm span for either sex (Table 2 ). Height thus estimated was larger than actual height in 101 (44.30%), less than actual height in 102 (44.74%), and equal to height in 25 (10.96%) subjects. All the spirometric values (FVC, FEV1, PEF, and FEF25–75%) were significantly greater in male subjects as compared with female subjects. Only the calculated FEV1/FVC ratios were similar in both groups (Table 2 ).
|
|
|
Fifty-one subjects were classified as having restrictive defects using actual height. When arm span was used, only one subject was incorrectly classified as having normal spirometry. Among the other 50 subjects, 1 was placed in a worse category (from moderate to severe restriction) and 3 were placed in a better category (2 from moderate to mild obstruction and 1 from severe to moderate obstruction). When estimated height was used, six subjects were incorrectly classified as having normal spirometry (Table 3 ). Of the 45 subjects correctly classified, 2 were placed in a better category (one each from moderate to mild and severe to moderate obstruction).
Overall, 37 (16.22%) and 32 (14.04%) subjects were wrongly classified or categorized when actual height measurements were replaced by arm span and estimated height measurements, respectively. These differences in misclassification rates by the two methods were not statistically significant. Kappa measure was calculated for agreement in classification and categorization, and a kappa value of 0.779 and 0.808 was obtained for methods using arm span and estimated height measurements, respectively. Limits of agreement were also calculated for differences in various spirometric indexes using height and arm span measurements as well as height and estimated height measurements. Even though the absolute numerical values were different, the limits were equally wide for FEV1 and FEV1/FVC ratio for the two groups, and marginally narrower for the other variables (Table 4 ).
|
Discussion
Arm span and height correlation have been shown to be different in different racial and ethnic groups,6 ,13 ,14 but there is a paucity of information on this relationship in the Indian population. The arm span to height ratios obtained in this study are in close agreement with the previously recorded figures for the adult white population.2 ,3 ,4 ,6 In the vast majority of our subjects, arm span exceeded height.7 ,8 ,15
Arm span has been proposed as a surrogate for standing height in the prediction of lung volumes in subjects who are unable to stand or who have skeletal deformities. Considering the fact that the absolute difference between height and arm span in any individual is usually small (mean of 4.0 cm in the present study), and to avoid an additional calculation in estimating height from arm span, we proposed that arm span may be directly substituted for height in prediction equations, as also suggested by few previous studies.2 ,5 ,16 ,17 This is likely to introduce a small error in the final results, the clinical relevance of which has not been studied so far (to our knowledge). Although Hepper and colleagues2 had found that predicted vital capacity based on height and arm span measurements were almost equal, no further analysis was provided. Allen17 had also showed that standardized FEV1 values using height and arm span were not statistically different in elderly women.
Each subject in our study had three sets of predicted values based on calculations from height, arm span, and height estimated from arm span using a fixed ratio. The regression equations for prediction of these normal values in either sex use only height and age as variables (Table 1 ),10 and any change in the value of one variable will result in a corresponding change in the final calculated result. The magnitude of this change will depend on the magnitude of change in the concerned variable, and direction of resulting change will depend on the actual regression equation. For example, the results of the regression equation for FVC used by us are positively correlated with height, and any addition (or subtraction) in height will consistently increase (or decrease) the predicted FVC. The consistency in the direction of change in these values is responsible for the statistical significance of results, when arm span is substituted for height (as 79.82% subjects had arm span greater than height), even though the absolute differences between predicted values based on height and arm span measurements are small (Table 4 ).
Of the tests in the study, 83.78% and 85.96% were classified and categorized correctly when arm span and estimated height, respectively, were substituted for actual height. The kappa measure of agreement takes values between zero and unity depending on the level of agreement between two such sets of data. A value > 0.75 is indicative of excellent overall agreement.11 The kappa estimates for the two sets of readings in this study were 0.779 and 0.808, which suggest that both arm span and height estimated from arm span can be used in place of actual height during classification and categorization of spirometric data, with good results. Results may be better, however, when estimated height values are used.
Agreement can also be estimated from the mean and SD of the differences between two sets of data by calculating limits of agreement, and evaluating if differences within this range are clinically important.12 We have calculated limits of agreement for various spirometric indexes using height and arm span as well as height and height estimated from arm span. The limits are almost equally wide in both instances (Table 4 ), and much wider than the 0.15-L normal variability permitted for both FEV1 and FVC measurements.9 Thus, both these methods are probably equally unacceptable (or acceptable) for clinical purposes. The use of arm span alone instead of height estimated from arm span may therefore be equally appropriate in a given clinical situation.
A significant finding of this study is that a sizable proportion of normal spirometric data may be misinterpreted as restrictive when arm span is substituted for height. Similar misinterpretation, though less, also occurs when height estimated from arm span is used. Both methods also wrongly classify and categorize obstructive and restrictive defects. However, these differences in misclassification rates using the two methods are statistically insignificant. It would appear that the use of height estimated from arm span is only a marginally better alternative. However, height and arm span correlation are not available for several racial/ethnic groups, and under such circumstances, arm span can also be directly used.
Conclusions
The substitution of arm span for height in prediction equations for lung function introduces a small error in the calculation of predicted values, which is statistically significant. The use of both arm span as well as height estimated from arm span is associated with misclassification of spirometric data. Notably, several normal spirometric results may be misinterpreted as restrictive. Although the use of height estimated from arm span may be preferred in subjects in whom height cannot be measured reliably, arm span is a reasonable surrogate for height in populations in whom established norms for correlation between height and arm span are not available.
Footnotes
Correspondence to: Surinder K. Jindal, MD, FCCP, Head, Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh 160012, India
Abbreviations: FEF25–75% = maximal midexpiratory flow; PEF = peak expiratory flow
Received for publication March 4, 1998. Accepted for publication September 3, 1998.
References
This article has been cited by other articles:
|
|
 |
|
  R. Pellegrino, G. Viegi, V. Brusasco, R. O. Crapo, F. Burgos, R. Casaburi, A. Coates, C. P. M. van der Grinten, P. Gustafsson, J. Hankinson, et al. Interpretative strategies for lung function tests Eur. Respir. J., November 1, 2005; 26(5): 948 - 968. [Full Text] [PDF]
|
|
|
|
|
|
M. R. Miller, R. Crapo, J. Hankinson, V. Brusasco, F. Burgos, R. Casaburi, A. Coates, P. Enright, C. P. M v. d. Grinten, P. Gustafsson, et al. General considerations for lung function testing Eur. Respir. J., July 1, 2005; 26(1): 153 - 161. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Teramoto, T. Matsuse, Y. Ouchi, A. N. Aggarwal, D. Gupta, and S. K. Jindal Substitution of Arm Span for Standing Height Is Important for the Assessment of Predicted Value of Lung Volumes in Elderly People With Osteoporosis Chest, December 1, 1999; 116(6): 1837 - 1838. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
