Bronchial Responsiveness in Children Exposed to Atmospheric Pollution in Hong Kong: Discussion
The crude prevalence of BHR was 25.5 percent (95 percent CI=21.4 to 29.7 percent). A comparison of the distribution of BHR, categorized by degree of severity between the two districts, is shown in Table 4. There was a higher prevalence of BHR in children from KDT (x2=7.74, df=3, p=0.052) mainly attributable to more mild hyperreactivity in KDT. The district difference was accounted for by a higher prevalence of BHR in boys (x2= 11.28, df=3, p=0.010) but not in girls (x2=6.24, df=3, p=0.100). When those children who had a history of wheezing and/or who were diagnosed asthmatic were excluded, a difference in prevalence of BHR between the two districts was still observed (x2=8.93, df=3, p=0.030) (Table 4).
Mean (SD) PD20 among those cases with a PD20<7.80 дто1 was 4.65 дто1 (2.24) for SDT (n=50) and 3.93 fimol (2.33) for KDT (n=58) (t=1.63, df=105, p=0.107).
The results of the GEE on the BR dose-response slopes for the nonwheezing and nonasthmatic children revealed that sex, age, session of school attended, education of father, living quarter type, and passive smoking did not have any effect on BR (p>0.05), but there was a significantly higher degree of BR in children from KDT (z=1.97, p=0.Q49). The effect due to clustering was small (intraclass correla-tion= —0.026). There was no association between a history of cold within 10 days of the test and either PD20 or slope (p=0.2). online antibiotics
The results of logistic regression on the BHR showed that the differences between districts were significant when PD20<3.2 /imol was used as a cutoff point (z=4.452, p<0.001) but was not when PD20<7.8 jumol was used (z=1.179, p=0.238). One major concern about the effects of atmospheric pollution on the respiratory system of a child is the potential effect on lung growth and pulmonary disease in adulthood. Most studies using respiratory questionnaires have shown a high prevalence of symptoms in children exposed to atmospheric pollution. The results of the first year of our study carried out in 1989 showed a consistently higher prevalence of respiratory symptoms in children from KDT who were exposed to a higher level of pollution compared with those from SDT. This was confirmed in the second year follow-up of the same cohort, in whom we studied BR for objective evidence of a health effect of pollution. We consider
such a difference to be real as the BR effect remains after adjustment for gender, session of school attended, education of father, type of living quarters, age, and exposure to passive smoking. In studies of this kind, the design effects of cluster sampling and their interpretation must also be taken into account. One approach is to introduce within-cluster correlation in the model to take account of the effect due to the use of clusters. The results using this approach supported the original findings.
Table 4—Prevalence of Bronchial Hyperreactivity in All Children and in Nonasthmatic, Nonwheezing Children in SDT and KDT Districts
BHR | All Children | Nonasthmatic, Nonwheezing Children | ||||
SDT, n (%)* | KTD, n (%) | Total, n {%) | SDT, n (%) | KDT, n {%) | Total, n (%) | |
Moderate | 4(2) | 5(3) | 9(2) | 3(1) | 2(1) | 5(1) |
Mild | 9(4) | 22(11) | 31 (7) | 6(3) | 18(10) | 24(6) |
Slight | 37(17) | 31 (15) | 68 (16) | 35 (17) | 28 (16) | 63 (16) |
None | 172(78) | 143 (71) | 315(75) | 168 (79) | 130 (73) | 298(76) |
Total | 222 (100) | 201 (100) | 423 (100) | 212 (100) | 178 (100) | 390 (100) |