- Email: [email protected]
Prediction of Normal Spirometric Values for Adults Incapable of Standing
Prediction of Normal Spirometric Values for Adults Incapable of Standing* William P. Temple, B.A.;t James F. Morris, M.D., F.C.C.P.;* and Arthur Koski, D.Ed.§
A significant problem exists in predicting normal values for
regression equations using sitting height and age. Either of these relationships can be selected as a predictor of height for substitution into any standard spirometric prediction equations. Spirometric prediction equations using age and sitting height are also presented. (Chest 1988; 94:572-74)
prediCting the normal values of pulmonary function for an individual to be compared with the individuals performance on a test is an important part of the process of pulmonary evaluation. The consensus among pulmonologists is that the most reliable predictors of pulmonary function are the individuals sex, age, and standing height. 1 Prediction equations using these factors have been developed independently by different investigators; however, our patient population includes amputees, spinal cord injuries, and others who are unable or incapable of standing upright, for whom the generally available prediction equations do not apply While considering other physical characteristics that might be predictors of pulmonary function and also might be readily determined, we speculated that a person's sitting height should be proportional to his or her standing height and would therefore be a reasonable substitute when standing height could not be measured. We found the opportunity to test this hypothesis during a repeat study of pulmonary function in normal individuals after a 15-year interval. 2
pollutants. Standing height (H) was measured in inches with the subject standing with the back of the head against a wall and without shoes, comparable to the method of Ferris and Stroudt. 3 Sitting height (S), also in inches, was measured with a specially constructed gauge functionally similar to a caliper. It consisted of a base that serves as a seat for the subject, a back which is fastened perpendicularly to the base and supports the height scale, and a triangleshaped height transfer block which is held against the top of the subjects head and the height scale to measure the sitting height. Studies of pulmonary function were performed in the standing position." Standing-to-sitting height ratios (HIS) were calculated and analyzed for frequency distribution. Means and standard deviations were calculated for age (A) and for sitting and standing heights using standard methods.' The relationships between standing and sitting heights, age, sex, and spirometric values were determined using regression analysis," The study was approved by the Oregon State University Human Studies Committee.
pulmonary function in subjects unable to stand for measurement of height. We studied 196 normal men and women to determine the relationship between sitting and standing height. Two predictors of standing height are recommended: (1) standing-to-sitting height ratio; and (2) multiple
MATERIALS AND METHODS
A total of 208 of the original sample of 988 individuals studied in 1969 were still residing in the Willamette River Valley of western Oregon. Nine subjects were rejected on the basis of abnormal responses to questions on a pretest questionnaire. The remaining sample population consisted of 199 men and women. All subjects were healthy white lifelong nonsmokers with no history of cardiopulmonary disease or exposure to ambient or occupational air
*From the Pulmonary Disease Section, Veterans Administration Medical Center, Portland, and the College of Health and Physical Education, Oregon State University, Corvallis, OR. Partially funded by biomedical support grant RR-07079 from the National Institutes of Health. tComputer Specialist, VA Medical Center. +Professor of Medicine. §Assistant Dean and Professor of Health. Manuscript received November 9; revision accepted February 1. Reprint requests: Dr: Morris, VA Medical Center, PO Box 1034, Portland, OR 97207
572
RESULTS
The data on height were first analyzed as the HIS ratio. The data for one man and two women were rejected because their ratios were greater than 3.8 standard deviations from the mean. The man and one woman had sitting heights within 1 SD of the sample means but were the shortest man and woman for standing height. This could indicate incomplete development of long bones. The remaining woman's standing height was within 1 SD of the sample mean, but the sitting height was abnormally large. This would most likely result from an error in transcription. The statistics on physical characteristics were calculated for the remaining 103 men and 93 women and are summarized in Table 1. The HiS ratio was tested for age dependence by fitting age and HIS to a simple linear regression Table I-Physical Characteristics ofPopulation* Group
No.
Age, yr
Standing Height, In
Sitting Height, In
SIH Ratio
Men Women
103 93
58±11 57±lO
69±3 64±2
36±1 34±2
1.93±0.05 1.90±0.05
*Data are means ± SD. Prediction of SpirometricValuesin Incapacitated Adults (Temple, Morris, Koski)
Table 2-Comparison ofValues for SEE
Table 3-Sitting Height Spirometric PredictionEquations*
Group
Fixed HiS Ratio
Age-Variable HiS Ratio
S and Age Regression *
Men Women
1.78 1.85
1.76 1.78
1.57 1.55
*S, Sitting height in inches.
equation and evaluating the age coefficients using a one-sample t-test. The resulting equations are as follows: for men, HIS = 1.8846 + O. 00073A
(1)
where the correlation coefficient (r) = 0.1612, and the standard error of the estimate (SEE) = 0.0492; and for women, HIS = 1.8186+0.00135A
. (2)
where r = 0.2646, and SEE = 0.0518. The levels of significance (p value) for the slopes are 0.10 for men and 0.01 for women, with the upper level of significance equal to 0.05. The standing height was then fitted into a multiple linear regression equation against age and sitting height. The resulting equations are as follows: for men, H=21.1492-0.0037A+ 1.3459S
(3)
where r=O. 7874, and SEE = 1.5723; and for women, H = 19.7789 + O. 0192A+ 1.2754S
(4)
where r=0.7759, and SEE = 1.5497. The values for SEE were calculated from the residuals (differences between the observed and predicted standing heights in inches) of each method of prediction and are listed in Table 2. The predicted standing heights are calculated as follows: (1) sitting height is multiplied by the fixed ratio for the individual's sex (Table 1); (2) the HIS ratio is calculated by substituting the individuals age into the HIS regression equation (equation 1 or 2) for the individuals sex and solving for the ratio; the individuals standing height is then estimated by multiplying his or her sitting height by the HIS ratio; (3) the individuals age and sitting height are substituted into the regression equation (equation 3 or 4) appropriate for their sex, and the equation solved for standing height. The values for SEE (Table 2) indicate that the multiple regression equations are more accurate in estimating standing height and that a statistically significant age dependence in HIS ratio exists for women; however, the actual differences in the estimated values between the three methods are small (O.3-inch maximum SEE) and may be the result of the limited accuracy of measuring standing and sitting heights. Spirometric prediction equations using the subjects age, sex, observed spirometric values, and sitting
Group and Equation Men FVC (L)=0.1836S-0.0399A+0.4222 FEV! (L) = 0.11815 - 0.0401A + 1.4370 FEF200-1200 (Usee) = 0.109450.1021A + 9.4781t FEF25-75% (Usee) = O. 0182S 0.0546A+5.5467 FEV/FVC% = 95.633 - 0.27235 - 0.2503A Women FVC (L)= 0.1756S - O. 0293A- 0.9539 FEV! (L)=0.1340S-0.0239A-0.8136 FEF200-1200 (Usee) = 0.256980.0374A -1.5970 FEF25-75% (Usee) = 0.1322S O. 0250A- 0.7273 FEV/FVC% = 63.698 + 0.41098 - 0.1148A
rValue
SEE
0.6645 0.7268 0.5551
0.6644 0.4980 1.7767
0.5484
0.9274
0.3793
6.4030
0.6675 0.6872 0.4683
0.5069 0.3799 1.1683
0.4675
0.6937
0.2858
4.9346
*S, Sitting height in inches; and A, age in years. tFEF200-1200, Mean forced expiratory flow between 200 ml and 1,200 ml of FVC.
height were developed. These equations are listed in Table 3. DISCUSSION
A significant number are unable to stand for measurement of vertical height as part of pulmonary function testing. This is especially true for amputees, stroke victims, and spinal cord injured patients. We did not compare spirometric results in sitting and standing positions. Pierson et al" made such a comparison in 235 men and women. These investigators" found values obtained in the standing position to be greater than in the sitting position. The differences were statistically significant but small for the forced vital capacity (FVC) (0.04±0.01 L; p
573
correlation coefficients for sitting and standing heights for FEVl . Ferris and Stroudt" suggest that arm span be used for predicting normal values when an accurate standing height was not attainable. Their correlation coefficient for arm span was reasonable for FVC but poor for FEVl . Three methods for estimating standing height from sitting height have been presented (a fixed standingto-sitting height ratio, simple, and multiple regression equations). Although the multiple regression equations have the smallest values for SEE, the maximum difference between the methods values for SEE was 0.3 inch, which translates into a volume difference of approximately 30 ml for both FVC and FEV1 using typical prediction equations. The method to be selected depends upon the discretion of the user. We recommend the fixed ratio for manual or nonprogrammable calculator calculations. The multiple regression equations or the sitting height spirometric prediction equations are appropriate for programmable calculators or computers because they are slightly more accurate for prediction and are convenient once they have been installed in the programs. The principal value of these height prediction methods is that they
574
allow the use of customary equations to predict spirometric values for subjects incapable of standing but able to sit erect. REFERENCES 1 Buist AS. Evaluation of lung function: concepts of normality In: Simmons DH, ed. Current pulmonology New York: John Wiley and Sons, 1981:141-65 2 Morris JF, Koski A, Temple WI?, Claremont A, Thomas DR. Fifteen year interval spirometric evaluation of the Oregon predictive equations. Chest 1988; 92:123-27 3 Ferris BG Jr, Stroudt HW Correlation of anthropometry and simple tests of pulmonary function. Arch Environ Health 1971; 22:672-76 4 Chatfield C. Statistics for technology: a course in applied statistics. 3rd ed. London: Chapman and Hall, 1983:106-30 5 Neter J, Wasserman ~ Kutner M. Applied linear regression models. Homewood, IL: Richard D Irwin, Inc, 1983:60-8 6 Pierson DJ, Dick NI?, Petty TL. A comparison of spirometric values with subjects in sitting and standing positions. Chest 1976; 70:17-20 7 Townsend MD. Spirometric forced expiratory volumes measured in the standing versus the sitting posture. Am Rev Respir Dis 1984; 130:123-24 8 Hepper NGG, Black LF, Fowler WS. Relationship of lung volume to height and arm span in normal subjects with spinal deformity Am Rev Respir Dis 1965; 91:356-62
Prediction of Spirometric Values in Incapacitated Adults (Temple, Morris, Koski)
A significant problem exists in predicting normal values for
regression equations using sitting height and age. Either of these relationships can be selected as a predictor of height for substitution into any standard spirometric prediction equations. Spirometric prediction equations using age and sitting height are also presented. (Chest 1988; 94:572-74)
prediCting the normal values of pulmonary function for an individual to be compared with the individuals performance on a test is an important part of the process of pulmonary evaluation. The consensus among pulmonologists is that the most reliable predictors of pulmonary function are the individuals sex, age, and standing height. 1 Prediction equations using these factors have been developed independently by different investigators; however, our patient population includes amputees, spinal cord injuries, and others who are unable or incapable of standing upright, for whom the generally available prediction equations do not apply While considering other physical characteristics that might be predictors of pulmonary function and also might be readily determined, we speculated that a person's sitting height should be proportional to his or her standing height and would therefore be a reasonable substitute when standing height could not be measured. We found the opportunity to test this hypothesis during a repeat study of pulmonary function in normal individuals after a 15-year interval. 2
pollutants. Standing height (H) was measured in inches with the subject standing with the back of the head against a wall and without shoes, comparable to the method of Ferris and Stroudt. 3 Sitting height (S), also in inches, was measured with a specially constructed gauge functionally similar to a caliper. It consisted of a base that serves as a seat for the subject, a back which is fastened perpendicularly to the base and supports the height scale, and a triangleshaped height transfer block which is held against the top of the subjects head and the height scale to measure the sitting height. Studies of pulmonary function were performed in the standing position." Standing-to-sitting height ratios (HIS) were calculated and analyzed for frequency distribution. Means and standard deviations were calculated for age (A) and for sitting and standing heights using standard methods.' The relationships between standing and sitting heights, age, sex, and spirometric values were determined using regression analysis," The study was approved by the Oregon State University Human Studies Committee.
pulmonary function in subjects unable to stand for measurement of height. We studied 196 normal men and women to determine the relationship between sitting and standing height. Two predictors of standing height are recommended: (1) standing-to-sitting height ratio; and (2) multiple
MATERIALS AND METHODS
A total of 208 of the original sample of 988 individuals studied in 1969 were still residing in the Willamette River Valley of western Oregon. Nine subjects were rejected on the basis of abnormal responses to questions on a pretest questionnaire. The remaining sample population consisted of 199 men and women. All subjects were healthy white lifelong nonsmokers with no history of cardiopulmonary disease or exposure to ambient or occupational air
*From the Pulmonary Disease Section, Veterans Administration Medical Center, Portland, and the College of Health and Physical Education, Oregon State University, Corvallis, OR. Partially funded by biomedical support grant RR-07079 from the National Institutes of Health. tComputer Specialist, VA Medical Center. +Professor of Medicine. §Assistant Dean and Professor of Health. Manuscript received November 9; revision accepted February 1. Reprint requests: Dr: Morris, VA Medical Center, PO Box 1034, Portland, OR 97207
572
RESULTS
The data on height were first analyzed as the HIS ratio. The data for one man and two women were rejected because their ratios were greater than 3.8 standard deviations from the mean. The man and one woman had sitting heights within 1 SD of the sample means but were the shortest man and woman for standing height. This could indicate incomplete development of long bones. The remaining woman's standing height was within 1 SD of the sample mean, but the sitting height was abnormally large. This would most likely result from an error in transcription. The statistics on physical characteristics were calculated for the remaining 103 men and 93 women and are summarized in Table 1. The HiS ratio was tested for age dependence by fitting age and HIS to a simple linear regression Table I-Physical Characteristics ofPopulation* Group
No.
Age, yr
Standing Height, In
Sitting Height, In
SIH Ratio
Men Women
103 93
58±11 57±lO
69±3 64±2
36±1 34±2
1.93±0.05 1.90±0.05
*Data are means ± SD. Prediction of SpirometricValuesin Incapacitated Adults (Temple, Morris, Koski)
Table 2-Comparison ofValues for SEE
Table 3-Sitting Height Spirometric PredictionEquations*
Group
Fixed HiS Ratio
Age-Variable HiS Ratio
S and Age Regression *
Men Women
1.78 1.85
1.76 1.78
1.57 1.55
*S, Sitting height in inches.
equation and evaluating the age coefficients using a one-sample t-test. The resulting equations are as follows: for men, HIS = 1.8846 + O. 00073A
(1)
where the correlation coefficient (r) = 0.1612, and the standard error of the estimate (SEE) = 0.0492; and for women, HIS = 1.8186+0.00135A
. (2)
where r = 0.2646, and SEE = 0.0518. The levels of significance (p value) for the slopes are 0.10 for men and 0.01 for women, with the upper level of significance equal to 0.05. The standing height was then fitted into a multiple linear regression equation against age and sitting height. The resulting equations are as follows: for men, H=21.1492-0.0037A+ 1.3459S
(3)
where r=O. 7874, and SEE = 1.5723; and for women, H = 19.7789 + O. 0192A+ 1.2754S
(4)
where r=0.7759, and SEE = 1.5497. The values for SEE were calculated from the residuals (differences between the observed and predicted standing heights in inches) of each method of prediction and are listed in Table 2. The predicted standing heights are calculated as follows: (1) sitting height is multiplied by the fixed ratio for the individual's sex (Table 1); (2) the HIS ratio is calculated by substituting the individuals age into the HIS regression equation (equation 1 or 2) for the individuals sex and solving for the ratio; the individuals standing height is then estimated by multiplying his or her sitting height by the HIS ratio; (3) the individuals age and sitting height are substituted into the regression equation (equation 3 or 4) appropriate for their sex, and the equation solved for standing height. The values for SEE (Table 2) indicate that the multiple regression equations are more accurate in estimating standing height and that a statistically significant age dependence in HIS ratio exists for women; however, the actual differences in the estimated values between the three methods are small (O.3-inch maximum SEE) and may be the result of the limited accuracy of measuring standing and sitting heights. Spirometric prediction equations using the subjects age, sex, observed spirometric values, and sitting
Group and Equation Men FVC (L)=0.1836S-0.0399A+0.4222 FEV! (L) = 0.11815 - 0.0401A + 1.4370 FEF200-1200 (Usee) = 0.109450.1021A + 9.4781t FEF25-75% (Usee) = O. 0182S 0.0546A+5.5467 FEV/FVC% = 95.633 - 0.27235 - 0.2503A Women FVC (L)= 0.1756S - O. 0293A- 0.9539 FEV! (L)=0.1340S-0.0239A-0.8136 FEF200-1200 (Usee) = 0.256980.0374A -1.5970 FEF25-75% (Usee) = 0.1322S O. 0250A- 0.7273 FEV/FVC% = 63.698 + 0.41098 - 0.1148A
rValue
SEE
0.6645 0.7268 0.5551
0.6644 0.4980 1.7767
0.5484
0.9274
0.3793
6.4030
0.6675 0.6872 0.4683
0.5069 0.3799 1.1683
0.4675
0.6937
0.2858
4.9346
*S, Sitting height in inches; and A, age in years. tFEF200-1200, Mean forced expiratory flow between 200 ml and 1,200 ml of FVC.
height were developed. These equations are listed in Table 3. DISCUSSION
A significant number are unable to stand for measurement of vertical height as part of pulmonary function testing. This is especially true for amputees, stroke victims, and spinal cord injured patients. We did not compare spirometric results in sitting and standing positions. Pierson et al" made such a comparison in 235 men and women. These investigators" found values obtained in the standing position to be greater than in the sitting position. The differences were statistically significant but small for the forced vital capacity (FVC) (0.04±0.01 L; p
573
correlation coefficients for sitting and standing heights for FEVl . Ferris and Stroudt" suggest that arm span be used for predicting normal values when an accurate standing height was not attainable. Their correlation coefficient for arm span was reasonable for FVC but poor for FEVl . Three methods for estimating standing height from sitting height have been presented (a fixed standingto-sitting height ratio, simple, and multiple regression equations). Although the multiple regression equations have the smallest values for SEE, the maximum difference between the methods values for SEE was 0.3 inch, which translates into a volume difference of approximately 30 ml for both FVC and FEV1 using typical prediction equations. The method to be selected depends upon the discretion of the user. We recommend the fixed ratio for manual or nonprogrammable calculator calculations. The multiple regression equations or the sitting height spirometric prediction equations are appropriate for programmable calculators or computers because they are slightly more accurate for prediction and are convenient once they have been installed in the programs. The principal value of these height prediction methods is that they
574
allow the use of customary equations to predict spirometric values for subjects incapable of standing but able to sit erect. REFERENCES 1 Buist AS. Evaluation of lung function: concepts of normality In: Simmons DH, ed. Current pulmonology New York: John Wiley and Sons, 1981:141-65 2 Morris JF, Koski A, Temple WI?, Claremont A, Thomas DR. Fifteen year interval spirometric evaluation of the Oregon predictive equations. Chest 1988; 92:123-27 3 Ferris BG Jr, Stroudt HW Correlation of anthropometry and simple tests of pulmonary function. Arch Environ Health 1971; 22:672-76 4 Chatfield C. Statistics for technology: a course in applied statistics. 3rd ed. London: Chapman and Hall, 1983:106-30 5 Neter J, Wasserman ~ Kutner M. Applied linear regression models. Homewood, IL: Richard D Irwin, Inc, 1983:60-8 6 Pierson DJ, Dick NI?, Petty TL. A comparison of spirometric values with subjects in sitting and standing positions. Chest 1976; 70:17-20 7 Townsend MD. Spirometric forced expiratory volumes measured in the standing versus the sitting posture. Am Rev Respir Dis 1984; 130:123-24 8 Hepper NGG, Black LF, Fowler WS. Relationship of lung volume to height and arm span in normal subjects with spinal deformity Am Rev Respir Dis 1965; 91:356-62
Prediction of Spirometric Values in Incapacitated Adults (Temple, Morris, Koski)