INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Age- and height-adjusted levels of lung function in Australians of Aboriginal descent (AAD), for both adults (1, 2) and children (3), are lower than those predicted for Australians of European Descent (AED) (4, 5). In a rural community of AAD in New South Wales (1), the mean FEV₁ and FVC of male subjects older than 15 yr of age were approximately 1.0 L and 1.3 L, respectively, less than those of male AED. Although 20% of subjects in this survey were noted to have chronic respiratory symptoms, the influence of these symptoms on lung function was not determined. A study done in 1992 found that asymptomatic AAD over 20 yr of age from a North Queensland community had ventilatory function approximately 25% lower than age- and sex-matched AED (2). Because subjects with symptoms, signs, or history of respiratory disease were excluded from the study (20% of all subjects seen), the effect of symptoms on pulmonary function was not addressed. In both of the studies just described, an arbitrary age range was chosen for analysis, and children were not included.

The prevalence of asthma (as defined by the presence of wheezing with bronchial hyperresponsiveness [BHR]) is reported as being low in both children and adults among AAD (6, 7). However, the prevalence of respiratory symptoms in AAD communities is reported to be high (1, 8). The cause of these symptoms has not been established, and the extent to which various respiratory symptoms reflect disease and thereby influence lung function has not been previously studied.

The aims of the present study were to extend previous observations by investigating a different AAD population; to estimate the prevalence of respiratory symptoms, BHR, smoking, and atopy in this population; to determine the influence of these factors on lung function; and to compare the effects of age and height on lung function in this AAD population and AED from Busselton, Western Australia.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

The AAD subjects lived in an isolated community on the northern tip of Western Australia. All available residents 5 yr of age or older who were present during the 10-d period of the study were invited to participate. All 104 male and 121 female residents in the community were included. Informed personal or parental consent was obtained with the assistance of local community health workers who were employed in the project. Approval for the study was granted by the Human Rights Committee of the University of Western Australia and by the local community council (9). The results of the study were communicated to the community by way of a report to this council. Although isolated, the community had 24-h nursing cover, with two full-time community nurses and visits from the Royal Flying Doctor Service at least weekly.

The comparison population consisted of 1,020 AED subjects with satisfactory lung function measurements who were participants in a study of the genetics of respiratory disease conducted in the town of Busselton in Western Australia. This study has been described elsewhere (10), but the equipment and methods of data collection, apart from some translation of questionnaire items as required at the time of interview, were identical to those described subsequently.

Demographic Data

Subjects' weight and standing height without footwear were measured. Stated age was verified from community health records. "Aboriginality" was defined by two indicator variables, the first being "a member of the Aboriginal community" and the second being "living in the Aboriginal community with no known non-Aboriginal antecedent." The subjects' ancestry was considered reliable because pedigrees for all the families had been collected as part of a longstanding anthropologic study of the community.

Questionnaire

We administered the British Medical Research Council (MRC) questionnaire (11) on respiratory symptoms and smoking by interview to all adults and to the parents of young children. Because of a wide range of literacy in English, modifications to the questionnaire were made with the aid of the local medical officer (R.S.) by translation of questions and the preamble to the questionnaire into local idiom as required. Cough or sputum were taken as being present if reported for more than 3 mo in the year. Dyspnea was defined as shortness of breath during walking with people of the same age, and wheezing as any period of wheezing within the previous year. Asthma was defined by a positive response to the question: "Has a/the doctor/sister ever told you that you had asthma?"

Lung Function

Ventilatory capacity was measured in the sitting position with a dry wedge spirometer (Model S; Vitalograph, Buckinghamshire, UK). FEV₁ and FVC were obtained according to guidelines of the American Thoracic Society (12), apart from the 5% repeatability criterion, which was relaxed, as has been suggested elsewhere (13). Eighteen subjects (including 15 children under 6 yr of age) on whom spirometry was attempted were unable to perform satisfactory forced expiratory maneuvers and were excluded, leaving 207 subjects included in the data analysis.

Atopy

Atopy was measured by skin prick test reactions to a panel of allergens (house dust mite, cat and dog dander, rye grass, and molds [Hollister-Stier, Spokane, WA]) applied to the forearm. Histamine (10 mg/ ml) and saline were used as positive and negative controls. Fifteen minutes after application of the allergens, wheal size was recorded as the mean of the long axis and its perpendicular, and was regarded as positive if 3 mm or greater.

Bronchial Responsiveness

Bronchial responsiveness to methacholine was measured according to the method of Yan and colleagues (14) in slightly modified form. Aerosols of physiologic saline, followed by doubling doses of methacholine from 0.1 µmol to a cumulative dose of 12.0 µmol, were inhaled at 90-s intervals from calibrated, hand-held De Vilbiss (Somerset, PA) 40 nebulizers driven by compressed air (15). FEV₁ was measured 60 s after inhalation of each dose of methacholine. One forced expiratory maneuver was performed on each occasion unless the result of the forced expiratory maneuver was < 80% of the presaline level or was thought to be technically unsatisfactory. Subjects with a history of asthma or wheezing in the 12 mo before the study started with a dose of methacholine of 0.1 µmol. All other subjects started with a dose of 0.4 µmol. The challenge was continued until the FEV₁ fell by 20% from the postsaline value. The cumulative dose of methacholine provoking a 20% decrease in FEV₁ (PD₂₀) was calculated by interpolation of the last two points of the log dose-response curve. BHR was defined as a PD₂₀ of less than 4.0 µmol.

Statistical Analysis

The effects of age and height on FEV₁, FVC, and FEV₁/FVC% for AAD were initially modeled separately for each sex. It has been shown that, particularly for children, the log of FEV is linearly related to the log of height (16), implying that FEV is linearly related to some power of height. We therefore assumed a power relationship of FEV with height (the same for all ages), together with separate linear relationships with age before and after the age at which the decline in lung function began. This age was estimated, but was constrained within ages 12 to 30 yr. Such a changepoint age implies a sudden and immediate switch from increasing lung function to declining lung function, and although a more realistic analysis might assume a curvilinear relationship between ages on either side of this changepoint age, our aim was to determine the age at which it was necessary to partition subjects into "adults" and "children" in terms of lung function, and also to compare parameters of these relationships with similar ones obtained from the population of AED from Busselton (4). These models were fitted through nonlinear regression, using SPSS for Windows (SPSS, Inc., Chicago, IL) (17), and interaction terms were also included if significant (p < 0.10). The associations of smoking, respiratory symptoms, and bronchial responsiveness to lung function were estimated by adding extra terms for each indicator variable in turn to the residuals from the best fitting model that already included the appropriate effects of age, height, and sex. In order to compare the effects of these external variables in the AAD and AED populations, models were also fitted to all data combined; in the same way, in order to maximize precision in the estimates of these effects, the models were fitted to the combined data for adults and children and males and females from both communities collectively. The significance of any differences between the communities in the associations with age, sex, and height was assessed by inclusion of interaction terms for these variables and the indicator variables for symptoms with either or both of the two indicator variables for Aboriginality described earlier.

As a further attempt to disentangle external factors from inherited associations, we also first adjusted lung function for the external factors (smoking, asthma, wheezing, cough, atopy, and BHR), in order to examine any differences that might result in the coefficients of the model that included effects of age, sex, height, and Aboriginality. Thus, for example, if the effect of height on lung function was decreased after adjustment for external factors, it might be concluded that the external factors were reducing lung function by stunting growth.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Questionnaire Responses

Asthma (a history of ever being diagnosed) was present in more female than male children and adults, and was more frequent in adults (Table 1). Wheezing and dyspnea were more frequent in females and in adults. Cough was reported more often and sputum less often in male children than in female children, with a smaller difference in adults. Smoking was very common in male adults, and approximately half the adult women smoked (Figure 1).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Smoking category, age, and sex for all 207 AAD subjects.

Atopy and BHR

Atopy (any positive response) was more common in adults than in children, and was slightly more common in women than in men. Atopy was due predominantly to positive responses to house dust mite, with only a few positive responses to grasses, cat or dog, or molds. BHR was present in approximately half of all children, and was less frequent in adults, among whom it was more often seen in females.

Ancestry

Twenty-eight subjects were not of full Aboriginal descent. Of these, 17 subjects (six male) were children and 11 (three male) were adults.

Comparison of AAD with AED

The frequency of questionnaire responses including smoking, atopy, and bronchial hyperresponsiveness among the AED from Busselton are shown in Table 2. Asthma was present much more often in AED than in AAD (except for adult females), wheezing was more common in males, dyspnea was more common in children but less common in adults, and cough and sputum were less common in all groups.

Atopy was more present in all AED groups for all allergens tested, especially grasses, cat and dog, or molds. BHR was less often present in children and female adults among AED.

Lung Function

In the AAD community, the majority of subjects were between 10 and 30 yr of age, and changes in lung function with age showed the usual steep increase to about age 20 yr, with a steady shallow decline thereafter and generally higher levels in men than in women (Figure 2). Separate models fitted to males and females indicated similar associations of FEV₁ with height and a similar age changepoint. Similar results were found in the AED community (Figure 2). Thus, the model that best fitted the data for both AAD and AED populations of both sexes and all ages, was found to be:

(1)

View larger version (18K): [in this window] [in a new window]   Figure 2.   Age and FEV1 for all 207 AAD and 1,020 AED subjects.

where  is a constant, b₁ to b₆ are regression coefficients, cp is the changepoint at which lung function starts to decline with age (in years), height is in meters, and sex is 1 for females and 0 for males. The term (age  cp) is 0 if age is less than cp and is 1 otherwise. The model demonstrated different aging effects for males and females, but the same height effects.

The main difference between AAD and AED, as estimated by an interaction effect of each of these main effects with the Aboriginality variable, was in the power to which height was raised, which was significantly lower in AAD (Table 3). This difference, when estimated from the combined data, was 0.43, with a 95% confidence interval (CI) of 0.50 to 0.36. It was, however, impossible to distinguish statistically between differences in the age at which the decline in lung function started or differences in the rate of decline from the same age. When both these terms were included, neither was significant, so that for the combined data, it was estimated that either the decline after the changepoint was 7 ml per year faster in AAD (95% CI: 2 to 13 ml), or that the decline started 1.5 yr earlier in AAD (95% CI: 0.3 to 2.6 yr). There was no significant difference between the two groups in the change in FEV₁/FVC with age and sex.

For example, using the combined model from Table 3, a typical AED male aged 30 yr and of height 175 cm would be predicted to have an FEV₁ of: 990 + 86 × 30  112 × (30  21) + 1,000 × (1.75^2.14) = 3,894 ml. A typical AAD male of the same age and height would have a predicted FEV₁ of: 990 + 86 × 30  119 × (30  21) + 1,000 × (1.75^1.71) = 3,123 ml.

For an AED female aged 30 yr and of height 165 cm, the expected value of FEV₁ is: 990 + 43 × 30  62 × (30  21) + 1,000 × (1.65^2.14) + 333 = 2,995 ml, and for an AAD female of the same age and height the expected FEV₁ is 990 + 43 × (30  21) + 1,000 × (1.65^1.71) + 333 = 2,366 ml.

Further examination of the coefficients for AAD group members of non-Aboriginal ancestry indicated that although the increase in FEV₁ with height was not different from that for the rest of the group, subjects with some non-Aboriginal ancestry had either a later changepoint or a slower decline after the changepoint, and were therefore more similar to the AED group.

The model for FVC was similar to that for FEV₁ in both the AAD and AED groups, and the results are therefore not presented.

Associations of Respiratory Symptoms, BHR, Atopy, and Smoking with Lung Function

After adjustment for age, sex, and height as previously described, positive responses to all questionnaire-derived variables, and also atopy and BHR, were associated with reduced lung function (with the exception of smoking with FEV₁) (Table 4). There were no significant differences between AAD and AED in any of these associations.

In a stepwise analysis, wheezing, dyspnea, and BHR were significantly associated with decreased FEV₁, with decrements of 78 ml, 100 ml, and 134 ml, respectively. Similarly, asthma, smoking, cough, and BHR were significantly associated with a decreased FEV/FVC ratio, with decrements of 2.3%, 1.0%, 1.4%, and 2.3%, respectively.

Prior adjustment for these external factors had no effect on the regression coefficients of the combined model in Table 4.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung function in the adult AAD in this study was similar to that reported for Cape York (1) and Bourke (2) Aboriginal adults, but less than that reported for AED from Busselton (4). This confirms previous findings that when allowance is made for sex, age, and height, ventilatory function in AAD is less than that in AED (2, 7). The differences between AAD and AED in the present study appeared to be due first to a less strong relationship of lung function with height, indicating the effect of possible differences in body shape or some adverse external effects on lung growth in childhood, and either a faster decline of lung function in adulthood or an earlier age at which the decline starts, indicating the possible influence of adverse external effects on lung maturation in childhood or lung degradation in adulthood.

Because this was a cross-sectional study, no comparison with longitudinal studies (18) was made, and it must be recognized that continued longitudinal follow-up might lead to different conclusions for many reasons (19), and particularly because of the confounding between age and cohort effects found in cross-sectional studies (20).

The prevalence of cough and sputum in the AAD adults in our study was greater than reported in other AAD groups (1, 8), and considerably higher than in AED from Busselton (4, 21) and elsewhere (22). High cigarette consumption or increased prevalence of lower respiratory tract infection may provide some explanation for this observation. In contrast, the prevalences of physician-diagnosed asthma and wheezing in AAD adults were less than in AED adults from Busselton. One explanation for these differences may be related to problems in determining the presence of symptoms and history of disease using a questionnaire validated in populations of European descent. The most frequent problem that we perceived with the questionnaire was with the concept of duration of symptoms, whereas the terms wheezing and asthma, as well as others, seemed to be well understood. However, given the correspondence between reported symptoms and measures of lung function and the lack of a significant difference in this correspondence between AED and AAD, this study gives some external validity to the use of the MRC questionnaire in groups for whom English may not be the first language.

This study has again shown that lung function in AAD adults is reduced as compared with that in AED adults. The observed differences in FEV₁ and FVC of 20 to 25% in adult AAD versus adult AED are similar to those observed for other racial groups (23), and may be related to inherited characteristics or to the presence of disease causing impairment of lung function. The reduction appears to result from decreased lung growth with increasing height during childhood, together with either an earlier cessation of lung growth or a more rapid decline in lung growth with age. Both inherited and external factors therefore appear to be determinants of lung function, insofar as lung function tended to be greater in subjects with some non-Aboriginal ancestry and determinants of lung growth were different in AAD and AED, with respiratory symptoms more frequent in the Aboriginal subjects and dyspnea, cough, and sputum production, which probably reflect external influences, being related to lower levels of lung function.