Abstract
As part of a comprehensive evaluation of lung function in Hong Kong–born Chinese children and adolescents, this study was conducted to determine updated prediction equations for spirometry, to evaluate the secular changes of lung function during the past decade, and to compare these results with other data sets. The results are based on 852 (392 male, 460 female) healthy students, age 7 to 19 yr, recruited from seven schools in Hong Kong. All were born and lived in Hong Kong, nonsmokers, free from past or present symptoms or diseases affecting the respiratory tract. A body plethysmograph was used to record lung function measurements. Natural logarithmic values of lung volumes and body height were used in the final regression model. Prediction equations for FVC, FEV1, and maximal expiratory flow at 50% of the FVC (MEF50) for both sexes are presented, with standing height as the dependent variable. Compared with Hong Kong data from 1985, the results show a significant increase in height-corrected FVC and FEV1 in both boys and girls, over the whole height range. Compared with recent data of whites, FVC in boys were 8 to 10% lower in the study population, and the difference increased to 12% above the 165 cm height ranges, while FVC in Chinese girls had similar or only slightly lower predicted values. FEV1 values showed a similar pattern with lesser difference between the two ethnic groups. Compared with recent data from Chinese children in Singapore, a similar pattern with overall lesser difference of the two populations was present in boys, whereas there was no significant difference between girls in the two places. Our findings support the conclusion that exogenous factors may contribute significantly to the differences in lung function values among ethnic groups and that it is important to examine normative values of various populations for secular trends.
Appropriate reference values are needed for the assessment of pulmonary function in disease states during childhood. Differences in pulmonary function among races have been known (1-3), although factors that account for the variation have not been clearly elucidated and have been reported to include differences in the size and shape of rib cage, respiratory muscle strength, and possibly parenchymal lung development (4). These factors in turn are influenced by genetic or environmental factors, including childhood health, environmental smoke and pollution, nutritional status, and exercise (5). There have been two published studies on reference values in spirometry of healthy Chinese children since 1985, one from Hong Kong (6), and the other more recently from Singapore (7). In view of the significant evolution of the socioeconomic environment in Hong Kong over recent decades, we postulated that lung function growth and development may also have changed significantly. This study of lung function in children and adolescents has therefore been conducted to establish updated normative values and to evaluate the differences, if any, from past data obtained in similar age groups in the same society, and also to compare with recent data from whites as well as other Chinese populations.
Children, 7 to 19 yr old, from seven schools on Hong Kong Island— three girls' schools, three boys' schools, and one coeducational—were recruited to perform the lung function test between May 1995 and December 1996. A questionnaire and a written informed consent form, both in Chinese, were distributed and collected by teachers. The questionnaire identified children with various respiratory problems or related diagnoses, exercise problems, heart disease, hospital admissions, common cold within the last 4 wk, or long-term medication. Children with any such history were offered the pulmonary function test in the Respiratory Clinic at Queen Mary Hospital, University of Hong Kong, although their test results were not included in the analysis. The questionnaire also included information on allergies and the level of physical exercise, but this was not used as inclusion/exclusion criteria.
All the remaining children, i.e., with a negative disease and medication history, and written parental informed consent, formed the target study population, comprising a total of 1,030 students (486 boys, 544 girls). At pulmonary function testing, the individual student was asked about smoking habits and any concurrent cold or cough symptoms. Smokers were excluded from analysis, and children with a cold were offered to be tested 4 wk later.
A SensorMedics 6200 Automated Body Plethysmograph (SensorMedics, CA) was transported to and stationed at each of the seven schools. The time spent in each school ranged from 3 to 7 wk. The field team consisted of one experienced physiotherapist who specialized in pediatrics and lung function together with a technician.
The same Automated Body Plethysmograph (SensorMedics 6200) was used in all children. The system was calibrated each morning and recalibrated at least every 3 to 4 h. Between each student a verification test was also performed. Disposable paper mouthpieces and Microgard disposable filters were used. Each student was provided with a demonstration by the research team together with practical testing until the student understood and performed well. Nose clips were not used in this study, but nose breathing during testing was avoided by manual occlusion.
The following anthropometric measurements were recorded: height, weight, arm span, sitting height, suprasternal height, head neck length, and trunk height. Height was measured in bare feet with a stadiometer to the last 0.1 cm. Weight was measured without shoes, in light school uniform and with empty pockets. The measuring technique followed the standardized technique as documented elsewhere (8).
Spirometry was performed according to the American Thoracic Society (ATS) criteria to ensure quality (9). All measurements were performed in a seated position and a rest was made between each repeated test. At least three trials were performed by each child in which the two largest FVC volumes as well as the two largest FEV1 had to be reproducible within 5% of each other. The curve with the largest FVC and FEV1 was chosen as the “best” curve, according to the ATS criteria. All the results were corrected to btps units.
All printouts from the tests were interpreted by a respiratory physician (M.S.M.I.). The test results were all sent home to the parents. In the event of a test with any spirometric parameter value below 80% predicted of the available reference values in the lung function system (Cotton Dust Standard, Knudson race-adjusted, Sensor Medics 6200 operator's manual) or a questionable pattern of the flow/volume loop, the student was called back for a repeat of the test and for a physical examination by the respiratory physician. This was to ensure that the student was free from any previously undetected symptoms. If no significant abnormality was detected, the test was included in the analysis. Otherwise, the results were excluded from the analysis and if necessary, the subject was referred to an appropriate health unit.
This study was approved by the Ethics Committee, Faculty of Medicine, University of Hong Kong.
Various regression models were applied to the series to explain the respiratory reference values over all ages examined, and the final selected one was the standard regression model as selected by many others (10). This regression model includes the natural logarithmic values of both lung volume values and body height.
The other models tested included both weight and age as well as height using various transformations. The choice of the appropriate regression model was made on the basis of two considerations: the highest explained variation of the dependent variable, the coefficient of determination (R2), and a constant residual standard deviation (SD) over the range of ages being included (as assumed in the statistical regression model). T-tests were applied to mean values using the two-tailed test. Reference values for the various values in boys and girls were produced in terms of charts, and table values in integers of height (cm) units, and they were all computed in terms of the functional mean and SD values after the usual antilogarithmic computation.
All analyses were performed using Statistical Analysis System (SAS) version 6 (11).
A total of 2,750 questionnaires were distributed, and 2,162 (79%) were returned. Of these, 1,030 underwent pulmonary function testing. The reasons for exclusion were (1) health problems according to the questionnaire (n = 422), (2) absence or moved from school (n = 325), (3) incomplete questionnaire replies and/or no signed informed consent forms (n = 280), and (4) omitted owing to limited time availability for the study team in some schools (n = 105).
Another 178 of the 1,030 tested children were excluded from analysis owing to various reasons such as smoking (n = 5), scoliosis (n = 3), non-Chinese parents (n = 3), physical problems (n = 23), only lung subdivisions performed (n = 37), technical problems (n = 105), and age outside range of analysis (n = 2). The analysis was based on 392 boys and 460 girls; all were born in Hong Kong and were of Chinese origin.
The grouping by age was made at midyear, i.e., children ⩾ 6.5 yr to < 7.5 yr were categorized as 7 yr of age, and so on. In the regression analysis, actual age of the individual was applied. The age range of the subjects included in this analysis was 6.9 to 19.2 yr in boys and 6.9 to 19.4 yr in girls.
The mean values of body measurements of the children are presented in Figure 1. When compared with the corresponding values of Hong Kong growth reference data from 1993 (12), they were generally comparable except being higher in height and weight in some age groups.
FVC (in liters). Various linear regression models were applied to the series by treating the two sexes separately and including FVC as the dependent measure. Height, weight, and age were also included in the regression models. The best fit of the data as determined by the R2 value was obtained by using a natural logarithmic transformation of both FVC and height. Including expression of either age or weight in the regression did not increase the explained variance, i.e., the R2 value, with more than 0.01 to 0.02. The estimated parameters, i.e., intercept (α), slope (β), R2, and residual SD are given for the two sexes separately in Table 1; the individual values and reference charts are depicted in Figures 2A (boys) and 2B (girls). Comparisons with other population data sets are presented in Figures 3A (boys) and 3B (girls).
Fig. 2. Predicted FVC (L) according to standing height in (A) boys and (B) girls. Curves show mean of study population and up to 3 SD about the mean.
[More] [Minimize]
Fig. 3. Comparison of FVC (L) in (A) boys and (B) girls in four different reference sets: I—Hong Kong, 1985; II—Hong Kong, 1996; III—Singapore, 1994; IV—U.K., 1993.
[More] [Minimize]FEV1 (in liters). Similarly to FVC as described previously, the estimated parameters are presented in Table 1 and the actual values in Figures 4A (boys) and 4B (girls). Comparisons with other population data sets are presented in Figures 5A (boys) and 5B (girls).
Fig. 4. Predicted FEV1 (L) according to standing height in (A) boys and (B) girls. Curves show mean of study population and up to 3 SD about the mean.
[More] [Minimize]
Fig. 5. Comparison of FEV1 (L) in (A) boys and (B) girls in four different reference sets: I—Hong Kong, 1985; II—Hong Kong, 1996; III—Singapore, 1994; IV—U.K., 1993.
[More] [Minimize]Maximal expiratory flow at 50% of FVC (MEF50 , in L/s). The estimated parameters are presented in Table 1. Comparison with the white data set is presented in Figures 6A (boys) and 6B (girls).
Fig. 6. Comparison of MEF50 (L/s) in (A) boys and (B) girls in two different reference sets: I—U.K., 1993; II—Hong Kong, 1996.
[More] [Minimize]Lung function is known to vary with ethnicity, and it is therefore important to establish normative values relevant to the ethnic group of the local population (5, 13). Different from genetic constitution, environmental factors affecting growth and development such as child health, nutritional status, and air quality may change over long periods of time (8), and therefore lung function values for a given ethnic community may change over time.
This study has generated updated spirometric values for Chinese children age 7 to 19 yr in Hong Kong. Using various linear regression models, we have been able to obtain equations that could produce updated normative lung function values for the local population. Similar to other studies, we found that the most important variable in the equations predicting spirometric values was height, whereas addition of either age or weight in the regression virtually did not increase the accuracy. The sample size for the children at 7 yr of age is very small (n = 4) and there is also some variability in the sample size over the ages. Nevertheless, this did not significantly influence the curve-fitting procedure and consequently not the functionally derived reference values. The mathematical functions used in the curve fitting are robust because they are fitted over the full range of ages in one run and because the functions are monotonic in nature (not permitting any abrupt change), and will thus not be influenced by some single outliers nor a small sample size for some selected ages.
Compared with spirometric data of Chinese children in Hong Kong reported a decade ago, using height-adjusted values, current values show an increase in FVC and FEV1 in both boys and girls across all height groups. The improvement is particularly seen in girls. This difference is unlikely due to the difference of methodology and technique used in the two studies, because the difference is not uniform throughout different height groups or sex. Although there is no specific information on chest dimensions or pubertal stage in this group of study subjects, we know that secular change in body size, height at puberty, and overall health in the pediatric population in Hong Kong have improved (8, 14, 15). We also know from recent local data (15) that the pubertal age range has been included in this study population. It is therefore highly tenable that there are accompanying changes in certain parameters of body habitus as well as muscle strength development in relation to changes in lifestyle factors, such as nutrition and exercise, hence affecting lung function development. The particular increase of lung volumes seen in the girls may be a result of the preferential care and attention given to male offspring in accordance with the cultural belief of male dominance in the family structure, which has likely balanced out with the decreasing family size with fewer children; thus, the recent generations of female offspring may have received more parental attention compared with the past, and therefore experienced a greater improvement in their physical development. Because the average height of children has increased, it is expected that children in Hong Kong have greater lung volumes now compared with peers of the same age and sex a decade ago.
Racial differences in lung function have been reported repeatedly (1-3, 5, 13). FVC and FEV1 in whites were found to be larger than Chinese and Indians (3). American blacks were also found to have consistently lower lung volumes than whites (2). These differences have been explained in terms of several factors, mostly related to characteristics of body size and shape. In a study comparing whites, Indians, and Chinese, the larger lung volumes in whites contributed to increased numbers of alveoli and physically larger chest cavities (4). A physique factor created by multiplying height by fat-free mass has been reported to explain differences in FVC between athletic and nonathletic whites (16), although this factor was not confirmed in another study (4). It is conceivable that the physique factor is an indicator of respiratory muscle strength, which is affected by exercise, nutrition, and overall health status.
A comparison of our data with a large white series reported in 1993 (17) shows that FVC in boys is generally lower and the fall-off increases to 12% at height range over 165 cm (Figure 3A), whereas that in girls is similar or only slightly lower across lower height groups, with a peak difference at 10% seen in the 155- to 160-cm group (Figure 3B). Values of FEV1, however, are mostly similar to white data in both boys and girls (Figures 5A and 5B). Generally, the normative values of Singaporean Chinese boys are intermediate between that of the Chinese in Hong Kong and the whites. The “lag” of FVC seen in taller height ranges is consistent in the two Chinese populations in Hong Kong and Singapore, although adolescent boys in Hong Kong show a much more significant fall-off in lung volumes than those in Singapore. The spirometric values between the two Chinese populations of girls are very similar. The apparent widening of the difference in FVC at the taller height ranges could be partly explained by a relatively accelerated increase in height at puberty in our study subjects, resulting in an apparent increase in height-predicted FVC, the effect of which wanes at the taller height ranges, which likely correspond to late or postpubertal age.
In contrast to the lower FVC and FEV1 values, the MEF50 rates were higher in girls than boys among the Chinese study subjects, and higher than white values (see Figure 6). Superior airflow per unit lung volume has been reported previously in younger white girls compared with boys (17, 18), and it has been postulated that the younger girls have shorter (to account for the reduced lung volumes at a given height in girls) but wider airways than boys. Our findings suggest that airways of Chinese children may be shorter but of similar wider caliber compared with whites. Alternatively, the higher expiratory flow rates may reflect a rigorous screening procedure in which children with subtle small airways problems have been excluded.
Because most lung function laboratories adopt the predicted reference sets in the commercially available lung function systems, we have compared our data with the predicted values as calculated by the reference formulas in the Cotton Dust Standard (Knudson race-adjusted) data set, provided by the lung function system (Sensor Medics 6200). According to the operator's manual, the race-adjusted predicted values applied to nonwhites are obtained by taking 15% off the white values. It is important to note that there are significant differences in all spirometric values, with the available reference data set underestimating the normative values (Table 2).
Our findings support the view that the differences seen in dynamic lung volumes between ethnic groups may be significantly contributed by exogenous factors and there may be changes in normative values as these factors evolve. Some of the differences between white and Hong Kong Chinese values have narrowed, but this has not persisted toward the topmost height ranges seen in the boys. Our study highlights the importance of obtaining updated normative values for lung function in different populations at intervals. Further update studies of lung function in Chinese adults in different communities as well as other ethnic groups would contribute to the understanding of the relative roles of genetic constitution and exogenous influence on lung function development.
The authors thank the personnel and all the students of the following participating schools in Hong Kong, without whose help this study could not have been conducted: St. Stephen's Girls' College, St. Stephen's Primary School, True Light Middle School, True Light Primary School, St. Paul's College, St. Paul's Primary School, and St. Peter's Secondary School. We also thank Mr. Chan Chi Kwong for his technical assistance during the data collection, and Dr. K. N. Chan for manuscript review.
Supported by a research grant from The Croucher Foundation (Grant 394/044/ 1141).
| 1. | Azizi B. H. O., Henry R. L.Ethnic differences in normal spirometric lung function of Malaysian children. Respir. Med.881994349356 |
| 2. | Wang X., Dockery D. W., Wypij D., Fay M. E., Ferris B. G.Pulmonary function between 6 and 18 years of age. Pediatr. Pulmonol.1519937588 |
| 3. | Yang T. S., Peat J., Keena V., Donnelly P., Unger W., Woolcock A.A review of the racial differences in the lung function of normal Caucasian, Chinese and Indian subjects. Eur. Respir. J.41991872880 |
| 4. | Donnelly P. M., Yang T. S., Peat J. K., Woolcock A. J.What factors explain racial differences in lung volumes? Eur. Respir. J.41991829838 |
| 5. | Cotes, J. E. 1993. Lung function throughout life: determinants and reference values. In J. E. Cotes. Lung Function: Assessment and Application in Medicine, 5th ed. Blackwell Scientific Publications, Oxford. 445–513. |
| 6. | Lau S. P., Fung K. P., Chow O. K. W., Chu S. Y., Tsoi N. S., Lun K. W., Chan S. C.A survey on lung functions of normal children in Hong Kong. Hong Kong J. Paediatr.219853745 |
| 7. | Connett G. J., Quak S. H., Wong M. L., Teo J., Lee B. W.Lung function reference values in Singaporean children aged 6–18 years. Thorax491994901905 |
| 8. | Leung S. S. F., Lau J. T. F., Xu Y. Y., Tse L. Y., Huen K. F., Wong G. W. K., Law W. Y., Yeung V. T. F., Yeung W. K. Y., Leung N. K.Secular changes in standing height, sitting height and sexual maturation of Chinese: the Hong Kong Growth Study, 1993. Ann. Hum. Biol.231996297306 |
| 9. | American Thoracic SocietyStandardization of spirometry. Am. J. Respir. Crit. Care Med.152199511071136 |
| 10. | Chinn S., Rona R. J.Height and age adjustment for cross sectional studies of lung function in children aged 6–11 years. Thorax471992707714 |
| 11. | SAS Institute Inc. 1994. SAS/STAT User's Guide. Version 6, 4th ed. SAS Institute, Cary, NC. |
| 12. | Leung S. S. F., Tse L. Y., Wong G. W. K., Law W. Y., Huen K. F., Yeung W. K. Y., Yeung V. T. F., Lau T. F., Oppenheimer S. J., Leung N. K.Standards for the anthropometric assessment of nutritional status of Hong Kong children. Hong Kong J. Paediatr.121995513 |
| 13. | Greenough A., Hird M. F., Everett L., Price J. F.Importance of using lung function regression equations appropriate for ethnic origin. Pediatr. Pulmonol.111991207211 |
| 14. | Leung S. S. F., Lau J. T. F., Tse L. Y., Oppenheimer S. J.Weight-for-age and weight-for-height references for Hong Kong children from birth to 18 years. J. Paediatr. Child Health321996103109 |
| 15. | Tam S. Y. M., Karlberg J. P. E., Kwan E. Y. W., Tsang A. M. C., Baber F. M., Low L. C. K.The improvement in growth, socioeconomic and health status in Hong Kong Chinese infants in the first two years of life—1967 to 1994. Hong Kong J. Paediatr.4199939 |
| 16. | Biersteker M. W. A., Biersteker P. A.Vital capacity in trained and untrained healthy young adults in the Netherlands. Eur. J. Appl. Physiol.5419854653 |
| 17. | Rosenthal M., Bain S. H., Cramer D., Helms P., Denison D., Bush A., Warner J. O.Lung function in white children aged 4 to 19 years: I—Spirometry. Thorax481993794802 |
| 18. | Hibbert M., Couriel J. M., Landau L. I.Changes in lung, airway and chest wall function in boys and girls between 8 and 12 years. J. Appl. Physiol.571984304308 |
