Bronchial Responsiveness in a Norwegian Community1-3
PER
s.
BAKKE, VALBORG BASTE, and AMUND GULSVIK
Introduction
In recent years there has been an increased interest in the role of bronchial responsiveness as a host characteristic that may be important in the development of obstructive lung disease (1). Studies of bronchial responsiveness in general population samples have been performed to reduce selection bias. In a rural population sample of 916 subjects 30 to 60 yr of age in Australia, bronchial responsiveness was observed in 11070 of the sample (2). Bronchial responsiveness was defined as a dose of ~ 3.9 umol histamine provoking a 20070 fall in FEV 1. In a community study in England tests of bronchial responsivenesswerecarried out in 511 subjects 18to 64 yr of age (3). Bronchial responsiveness to ~ 8 umol histamine was present in 14070 of the sample. A Dutch population study of 2,156 subjects 15to 64 yr of age showed that 25070 of the sample had bronchial responsiveness defined as PC 10 ~ 16 mg/ml histamine (4). In the Nordic countries limited community-based information is available on the distribution of bronchial responsiveness (5). The objectives of the present crosssectional study of a Norwegian general population were (1) to examine the prevalence of bronchial responsiveness to methacholine by sex and age, (2) to examine the association of bronchial responsiveness to airborne exposures as assessed by smoking habits, urban-rural area of residence, and occupational airborne exposure, and (3) to assessthe relationship of bronchial responsiveness to levelof lung function, airway caliber, and allergy. Methods Population The study is a two phased cross-sectionalcommunity survey. The first phase was conducted from September to December 1985. A questionnaire was mailed to a random sample of 4,992 subjects of the general population between 15and 70 yr of age in Hordaland county on the southwest coast of Norway. The response rate was 90010. In the second phase, conducted from April 1987 to July 1988, a random sample of 595 subjects of those re-
SUMMARY Bronchial responsiveness to methacholine was examined In a Norwegian general popu490) 18 to 73 yr of age. Altogether, 20 and 8% of the sample had PC20 4it 32 lation sample (n mg/ml and PC20 ~ 8 mglml, respectively. The relationship of bronchial responsiveness to the followIng potential predictors were examined: sex, age, smoking habits, airway caliber (FEV1 ) , FEV1 percent predicted (%FEV 1) , urban-rural area of residence, occupational airborne exposure In present job, and total serum IgE. After adjusting for age and FEV1, the odds ratio for PC20 ~ 32 mglml was higher for men than for women In smokers and In ex-smokers, but did not vary by sex In nonsmokers, the adjusted odds ratio for PC20 ~ 32 mg/mlln male compared with female smokers being 8.4 (95% CI: 2.5-37.4). Irrespective of smoking status the sex- and FEV1-adjusted odds ratio for PC20 ~ 32 mg/ml fell with Increasing age. For every 1o-yr Increase In age the adjusted odds ratio for PC20 < 32 mg/ml methacholine In nonsmokers decreased by 2.0 (95% CI: 1.3-3.3). Also FEV1 and %FEV1 were predictors of PC20 ~ 32 mg/ml after adjusting for sex and age Irrespective of smoking status. Bronchial responsiveness (PC20 ~ 8 mg/ml) was more prevalent In rural than In urban areas, the adjusted odds ratio being 2.5 (95% CI: 1.1-5.9) for bronchial responsiveness In rural compared with urban residents after adjusting for sex, age, smoking habits, and FEV1. Bronch..1responsiveness was not associated with present airborne occupational exposure. No relationship was found between sex- and age-adjusted level of serum IgE and bronchial responsiveness. In this general population the following variables were Independent predictors of bronchial responsiveness: male sex, younger age, smoking, level of airway caliber (FEV1), level of pulmonary function (%FEV 1), and rural area of residence. All REV RESPIR DIS 1911; 143:317-322
=
sponding in the first phase and living in the city of Bergen and 11 surrounding municipalities was invited to attend a clinical examination. The examination included completion of a questionnaire on smoking habits and on all jobs held for more than 6 months since leaving school, spirometric measurements, blood sampling, and a test of bronchial responsiveness. The attendance rate was 86010 of those invited. The bronchial responsiveness test was performed in 490 subjects of those attending. 1\venty-two subjects wereexcluded for the following reasons: pretest FEV 1 was below 60010 of predicted in five persons, 15 subjects did not want to do the test, and in two subjects the test was stopped because of nausea. Nonsmokers were defined as subjects who had never smoked daily (6). Ex-smokers were subjects who had smoked daily but had given it up. Subjects were classified as smokers if they were smoking daily at the time of the study. The occupational titles were coded according to the three digit numbers of the Nordic Classification of Occupations, which followsthe recommendations of the International Standard Classification of Occupations (7). The occupational titles were allocated into three exposure groups based on the estimated degree of airborne exposure (AE) from dusts, fumes, mists, or gases experienced in a typical job of that particular occupation: the AE + + group included occupations with a high degreeof airborne pollutants, i.e., foun-
dry workers, painters, and insulation workers. The AE + group included occupations with a moderate degreeof airborne pollutants, i.e., farmers, greasers, and typographers. The AE- group contained occupations with no airborne exposure, i.e., teachers, clerks, fishermen. This allocation was performed by three experts in occupational medicine and hygiene. The complete list of occupations allocated into the three exposure groups may be obtained from the authors. Spirometry was performed using a Gould 2100 spirometer (Gould Instruments, Cleveland, OH). The variables recorded included FVC and FEV 1. Three technically satisfactory measurements wereobtained in which FVC was reproducible within 300 ml. A technically satisfactory test met the lung function test-
(Received in original form January 11, 1990 and in revised form August 9, 1990) 1 From the Department of Thoracic Medicine, University of Bergen, Bergen, Norway. 2 Supported by the Royal Norwegian Council for Scientific and Industrial Research, the Norwegian Research Council for Science and the Humanities, and the Norwegian Asthma and Allergy Association. 3 Correspondence and requests for reprints should be addressed to Per Bakke, M.D., Department of Thoracic Medicine, 5021 Haukeland Hospital, University of Bergen, Bergen, Norway.
317
318
BAKKE, BASTE, AND GULSVIK
ing criteria of the European Coal and Steel Community (8). Airway caliber was examined in terms of FEV 1, and level of lung function was examined in terms of FEV 1 percent of predicted (OJoFEV 1). The reference values of FEV 1 were taken from a previous Norwegian population study (9).
Markers of Allergy Serum total IgE was used as a marker of allergy. IgE was measured by the PRIST method and reported in terms of international units (IV) per milliliter. In 22OJo of the subjects the laboratory could only state that the serum IgE was less than 5 IV/ml. These subjects were given the value 3 IV/ml. The distribution of serum IgE in this general population sample was positively skewed. Hence, serum IgE was log-transformed with base 10 (lOgl0), to obtain a more normal distribution. Mean log IgE was higher in men than in women (table 1) and decreased by rising age in women, whereas no age trend was observed in men. Smokers had higher mean log IgE than did nonsmokers. For each sex and age group (18-34, 35-54, 55-73 yr), the mean and standard deviation of log 19B was estimated. Each subject was then characterized by the number of standard deviations by which he or she differed from the mean of his or her specific sex and age group (Z-score) (10). The number of standard deviations was divided into three groups; < -0.5, -0.5 to < 0.5, and ~ 0.5, the negative values indicating the number of standard deviations below the mean and the positive values indicating the number of standard deviations above the mean.
Bronchial Responsiveness 'lest A Wright nebulizer was driven by air at a flow rate of 8 L/min and calibrated to give an output of 14 ml/min. The nebulizer was primed with a 3-ml solution of methacholine chloride. The subject wore a noseclip and inhaled the aerosol via a mouthpiece, breathing ti-
dally for 2 min at 5-min intervals. Pretest FEV 1 was taken as the highest of three measurements within 0.3 L/s. Subjects with a pretest FEV 1 ~ 80OJo of predicted and without a diagnosis of obstructive lung disease started with a methacholine concentration of 2 mg/ml, whereas subjects with a pretest FEV 1 of 60 to 800/0 of predicted or with a diagnosis of obstructive lung disease started with a methacholine concentration of 0.5 mg/ml. After each inhalation one technically satisfactory FEV 1 was obtained at 30 and at 90 s, of which the lowest value was recorded. If FEV 1 fell less than 10OJo of the pretest value after an inhalation, a fourfold higher concentration was given next. If FEV 1 fell by 10 to 20OJo of the pretest value, a twofold higher concentration was given. The test was stopped if FEV 1 fell by 20OJo of the pretest value or a concentration of 32 mg/ml had been given. The degree of responsiveness was expressed as the concentration of methacholine provoking a 20OJo fall in FEV 1 from pretest value (PC10). PC10 was calculated by interpolation between the results of the last two inhaled concentrations of methacholine. Each subject was then given five inhalations of salbutamol (500 ug), and FEV 1 was measured until it rose to at least 95OJo of the pretest value. To examine the reproducibility of the bronchial responsiveness test, 20 subjects with PC10 ~ 32 mg/ml were tested on a second occasion within 3 to 7 days at the same time of the day. The subjects included 10 with obstructive lung disease and 10 without disease. The PC10 of the 10 healthy subjects were reproducible within one doubling concentration of methacholine (figure 1). For the 10 subjects with obstructive lung disease, seven were within one doubling concentration and all within two doubling concentrations (figure 1).
Data Analysis Contingency tables were analyzed for statistical significance using chi-square tests or, for
TABLE 1 CHARACTERISTICS OF THE STUOY POPULATION BY SEX
Variables Age, mean (SO) yr* Smoking habits t Smokers, % Cigarettes/day, mean (SO) Ex-smokers, % Nonsmokers, % Urban/rural area of residence Urban, %* Past or present physician's diagnosis of asthma, % t Log serum IgE, mean (SO) FEV1, mean (SO) Us %FEV1, mean (SO)§ FEV 1/FVC, mean (SO) • Age at December 31,1987.
t Information from postal questionnaire.
*liVing In the municipality of Bergen.
§ Percent of predicted value from reference 9.
Men (n == 251)
Women (n == 239)
41 (15.7)
43 (15.6)
40 15 (6.7) 26
34
39 12 (5.6) 18 42
70
70
3
4
1.38 (0.48)
1.22 (0.23)
4.1 (0.9) 95 (13.3) 0.81 (0.07)
3.0 (0.7) 96 (14.8) 0.82 (0.06)
-
•
32.0
0 0
/ /
:: 16.0
/
/
0
/
E
/
/
/
/
en
/
.§ 8.0
•/
/ /
0
/
•
/
/
/
/
/
/
/
/
/
/
/
/
1.0
/
/
/
2.0
/ /
/ /
4.0
/
0
/
C>
LJ
o,
/
/
I/)
~
/
/ /
./
/
/
/
//.
0.5
/ /
/
0:5
1:0
2:0
4:0
8:0
16:0
32:0
P(20 (mg/ml) ttest
Fig. 1. Reproducibility of PC 20 measurements within 1 wk in 10 healthy subjects and in 10 subjects with obstructive lung disease. The solid line is the line of identity, and the dashed lines indicate one doubling concentration of methacholine. Open circles = healthy; closed circles = obstructive lung disease.
stratified data, the Mantel-Haenszel extension of the test (11). Possible predictors of bronchial responsiveness, defined as PC10 ~ 32 giL methacholine, were examined in smokers, ex-smokers, and nonsmokers separately through logistic regression analyses. The numbers of smokers, ex-smokers and nonsmokers with PC10 ~ 8 giL methacholine were too small to give stable regression coefficients. Also, the small number of subjects with a high degree of airborne occupational exposure prevented analysis in separate smoking groups when examining the association of bronchial responsiveness to airborne occupational exposure. The logistic regression models were backwards, stepwise. Only statistically significant variables were kept in the final models. Entry and exit levels were set at p = 0.01. All analyses were performed using the BMDP statistical software package (12).
Results Altogether, 20070 of the population had a PC 20 ~ 32 mg/ml, whereas 6070 had a PC l O ~ 8 mg/ml (table 2). Bronchial responsiveness was more prevalent in women than in men (p < 0.01). This tendency was evident in all age groups and regardless of level of bronchial responsiveness (table 3). The sex difference in prevalence of bronchial responsiveness defined as PC20 ~ 32 mg/ml was twice as high in the age group below 50 yr as in that above 50 yr. In both men and women the prevalence of bronchial responsiveness increased by age, being twice as common in the oldest age group (54-73 yr) as' in the youngest age group (18-34 yr). Cigarette smoking was associated with an increased (p < 0.05) prevalence of PC20 ~ 8 mg/ml (table 2), whereas no association was evident between smoking habits
319
BRONCHIAL RESPONSIVENESS IN A NORWEGIAN COMMUNITY
TABLE 2 CRUDE PREVALENCES OF BRONCHIAL RESPONSIVENESS TO METHACHOLINE BY SEX, AGE, SMOKING HABITS, AREA OF RESIDENCE, AND OCCUPATIONAL AIRBORNE EXPOSURE IN A NORWEGIAN GENERAL POPULATION 1987/1988 (n ~ 490) PC20 (giL) Variable
Subjects (n)
8 to ~ 32 mg/ml it was 2.9 Lis (0.8 Lis), and at PC20 > 32 mglml it was 3.8 Lis (0.9 Lis). Seventy percent of all subjects had
an FEV. greater than 90070 predicted (table 4). The mean FEV. was higher for men than for women for all categories of OJoFEVI. In both sexes the prevalence of bronchial responsiveness fell with increasing pulmonary function level. For all levelsof pulmonary function the prevalence of bronchial responsiveness was two to three times higher in women than in men (table 4). The relationship of bronchial responsiveness to potential determinants was examined in logistic regression analyses to achieve simultaneous adjustment for several variables. Sex, age, and pretest air-
TABLE 3 DISTRIBUTION OF BRONCHIAL RESPONSIVENESS TO METHACHOLINE BY SEX AND AGE IN A NORWEGIAN GENERAL POPULATION, 1987/1988 (n = 490)
PC2()(g/L) Subjects (n)
2to 8 to < 32 (0/0)
> 32 (0/0)
Men Women
102
80
0 3
3 6
7 11
90 80
35-54
Men Women
82 90
0 1
0 8
9 16
91 76
55-73
Men Women
67 69
1 9
4 3
21 22
73 87
18-73
Men Women
251 239
0* 4
2 6
11 16
86 74
Age (yr) 18-34
* 100
TABLE 5 ADJUSTED ODDS RATIOS (OR) WITH 95% CONFIDENCE INTERVALS (CI) FOR BRONCHIAL RESPONSIVENESS (PC.... 32 giL METHACHOLINE) BY SEX, AGE. FEV" AND FEV, PERCENT OF PREDICTED (%FEV,) IN SMOKERS, EX·SMOKERS, AND NONSMOKERS OF A NORWEGIAN GENERAL POPULATION 1987/88 (n = 490) Ex-smokers (n = 109)
Nonsmokers (n = 186)
Sex· Women Men Age, yrt X X + 10 yr FEV,* X X-o.5 Us %FEV,* X X·l0 percentage points
Smokers (n = 195)
OR
95% CI
OR
95% CI
OR
95% CI
1 1.4
0.4-4.9
1 4.0
2.3-21 .4
1 8.4
2.5-37.4
1 0 .6
0.4-0.9
1 0.7
0.4-1 .3
1 0.5
0.3-0.8
1 2.2
1.4-3.5
1 3.8
1.9-7.8
1 4.0
2.4-6.7
1 1.9
1.2-2.7
1 2.5
1.5-4.0
1 3.6
2.3-5.8
• Adjusted lor age and FEV, . t Adjusted lor sex and FEV,. f Adjusted lor sex and age.
Discussion
IgE level (figure 2). The relationship of sex- and age-adjusted serum IgE level to bronchial responsiveness,defined as pew ~ 32 giL methacholine, was further analyzed in smokers, ex-smokers, and nonsmokers separately. Irrespective of smoking status, there was no significant association between sex- and age-adjusted serum IgE level and bronchial responsiveness in logistic regression analyses after adjusting for llJoFEV 1 (table 6).
pc,.
pc,. ,,6 9"
40
In this adult, general population study both female sex and older age were associated with higher crude rates of bronchial responsiveness to methacholine. However, after adjustment for airway caliber, in terms of pretest FEV 10 the sex and age relationship to bronchial responsiveness was reversed, the adjusted odds for bronchial responsiveness becoming higher in men than in women and increas-
,,329"
%
30
Fig. 2. Prevalence of bronchial responsiveness to methacholine (PC.... 8 giL and PC.... 32 giL) by serum IgE Z-score in men and women of a Norwegian general populat ion 1987-1988.Two subjects refused blood sampling. Hatched bars = women; closed bars = men .
20
10
PER
s.
BAKKE, VALBORG BASTE, and AMUND GULSVIK
Introduction
In recent years there has been an increased interest in the role of bronchial responsiveness as a host characteristic that may be important in the development of obstructive lung disease (1). Studies of bronchial responsiveness in general population samples have been performed to reduce selection bias. In a rural population sample of 916 subjects 30 to 60 yr of age in Australia, bronchial responsiveness was observed in 11070 of the sample (2). Bronchial responsiveness was defined as a dose of ~ 3.9 umol histamine provoking a 20070 fall in FEV 1. In a community study in England tests of bronchial responsivenesswerecarried out in 511 subjects 18to 64 yr of age (3). Bronchial responsiveness to ~ 8 umol histamine was present in 14070 of the sample. A Dutch population study of 2,156 subjects 15to 64 yr of age showed that 25070 of the sample had bronchial responsiveness defined as PC 10 ~ 16 mg/ml histamine (4). In the Nordic countries limited community-based information is available on the distribution of bronchial responsiveness (5). The objectives of the present crosssectional study of a Norwegian general population were (1) to examine the prevalence of bronchial responsiveness to methacholine by sex and age, (2) to examine the association of bronchial responsiveness to airborne exposures as assessed by smoking habits, urban-rural area of residence, and occupational airborne exposure, and (3) to assessthe relationship of bronchial responsiveness to levelof lung function, airway caliber, and allergy. Methods Population The study is a two phased cross-sectionalcommunity survey. The first phase was conducted from September to December 1985. A questionnaire was mailed to a random sample of 4,992 subjects of the general population between 15and 70 yr of age in Hordaland county on the southwest coast of Norway. The response rate was 90010. In the second phase, conducted from April 1987 to July 1988, a random sample of 595 subjects of those re-
SUMMARY Bronchial responsiveness to methacholine was examined In a Norwegian general popu490) 18 to 73 yr of age. Altogether, 20 and 8% of the sample had PC20 4it 32 lation sample (n mg/ml and PC20 ~ 8 mglml, respectively. The relationship of bronchial responsiveness to the followIng potential predictors were examined: sex, age, smoking habits, airway caliber (FEV1 ) , FEV1 percent predicted (%FEV 1) , urban-rural area of residence, occupational airborne exposure In present job, and total serum IgE. After adjusting for age and FEV1, the odds ratio for PC20 ~ 32 mglml was higher for men than for women In smokers and In ex-smokers, but did not vary by sex In nonsmokers, the adjusted odds ratio for PC20 ~ 32 mg/mlln male compared with female smokers being 8.4 (95% CI: 2.5-37.4). Irrespective of smoking status the sex- and FEV1-adjusted odds ratio for PC20 ~ 32 mg/ml fell with Increasing age. For every 1o-yr Increase In age the adjusted odds ratio for PC20 < 32 mg/ml methacholine In nonsmokers decreased by 2.0 (95% CI: 1.3-3.3). Also FEV1 and %FEV1 were predictors of PC20 ~ 32 mg/ml after adjusting for sex and age Irrespective of smoking status. Bronchial responsiveness (PC20 ~ 8 mg/ml) was more prevalent In rural than In urban areas, the adjusted odds ratio being 2.5 (95% CI: 1.1-5.9) for bronchial responsiveness In rural compared with urban residents after adjusting for sex, age, smoking habits, and FEV1. Bronch..1responsiveness was not associated with present airborne occupational exposure. No relationship was found between sex- and age-adjusted level of serum IgE and bronchial responsiveness. In this general population the following variables were Independent predictors of bronchial responsiveness: male sex, younger age, smoking, level of airway caliber (FEV1), level of pulmonary function (%FEV 1), and rural area of residence. All REV RESPIR DIS 1911; 143:317-322
=
sponding in the first phase and living in the city of Bergen and 11 surrounding municipalities was invited to attend a clinical examination. The examination included completion of a questionnaire on smoking habits and on all jobs held for more than 6 months since leaving school, spirometric measurements, blood sampling, and a test of bronchial responsiveness. The attendance rate was 86010 of those invited. The bronchial responsiveness test was performed in 490 subjects of those attending. 1\venty-two subjects wereexcluded for the following reasons: pretest FEV 1 was below 60010 of predicted in five persons, 15 subjects did not want to do the test, and in two subjects the test was stopped because of nausea. Nonsmokers were defined as subjects who had never smoked daily (6). Ex-smokers were subjects who had smoked daily but had given it up. Subjects were classified as smokers if they were smoking daily at the time of the study. The occupational titles were coded according to the three digit numbers of the Nordic Classification of Occupations, which followsthe recommendations of the International Standard Classification of Occupations (7). The occupational titles were allocated into three exposure groups based on the estimated degree of airborne exposure (AE) from dusts, fumes, mists, or gases experienced in a typical job of that particular occupation: the AE + + group included occupations with a high degreeof airborne pollutants, i.e., foun-
dry workers, painters, and insulation workers. The AE + group included occupations with a moderate degreeof airborne pollutants, i.e., farmers, greasers, and typographers. The AE- group contained occupations with no airborne exposure, i.e., teachers, clerks, fishermen. This allocation was performed by three experts in occupational medicine and hygiene. The complete list of occupations allocated into the three exposure groups may be obtained from the authors. Spirometry was performed using a Gould 2100 spirometer (Gould Instruments, Cleveland, OH). The variables recorded included FVC and FEV 1. Three technically satisfactory measurements wereobtained in which FVC was reproducible within 300 ml. A technically satisfactory test met the lung function test-
(Received in original form January 11, 1990 and in revised form August 9, 1990) 1 From the Department of Thoracic Medicine, University of Bergen, Bergen, Norway. 2 Supported by the Royal Norwegian Council for Scientific and Industrial Research, the Norwegian Research Council for Science and the Humanities, and the Norwegian Asthma and Allergy Association. 3 Correspondence and requests for reprints should be addressed to Per Bakke, M.D., Department of Thoracic Medicine, 5021 Haukeland Hospital, University of Bergen, Bergen, Norway.
317
318
BAKKE, BASTE, AND GULSVIK
ing criteria of the European Coal and Steel Community (8). Airway caliber was examined in terms of FEV 1, and level of lung function was examined in terms of FEV 1 percent of predicted (OJoFEV 1). The reference values of FEV 1 were taken from a previous Norwegian population study (9).
Markers of Allergy Serum total IgE was used as a marker of allergy. IgE was measured by the PRIST method and reported in terms of international units (IV) per milliliter. In 22OJo of the subjects the laboratory could only state that the serum IgE was less than 5 IV/ml. These subjects were given the value 3 IV/ml. The distribution of serum IgE in this general population sample was positively skewed. Hence, serum IgE was log-transformed with base 10 (lOgl0), to obtain a more normal distribution. Mean log IgE was higher in men than in women (table 1) and decreased by rising age in women, whereas no age trend was observed in men. Smokers had higher mean log IgE than did nonsmokers. For each sex and age group (18-34, 35-54, 55-73 yr), the mean and standard deviation of log 19B was estimated. Each subject was then characterized by the number of standard deviations by which he or she differed from the mean of his or her specific sex and age group (Z-score) (10). The number of standard deviations was divided into three groups; < -0.5, -0.5 to < 0.5, and ~ 0.5, the negative values indicating the number of standard deviations below the mean and the positive values indicating the number of standard deviations above the mean.
Bronchial Responsiveness 'lest A Wright nebulizer was driven by air at a flow rate of 8 L/min and calibrated to give an output of 14 ml/min. The nebulizer was primed with a 3-ml solution of methacholine chloride. The subject wore a noseclip and inhaled the aerosol via a mouthpiece, breathing ti-
dally for 2 min at 5-min intervals. Pretest FEV 1 was taken as the highest of three measurements within 0.3 L/s. Subjects with a pretest FEV 1 ~ 80OJo of predicted and without a diagnosis of obstructive lung disease started with a methacholine concentration of 2 mg/ml, whereas subjects with a pretest FEV 1 of 60 to 800/0 of predicted or with a diagnosis of obstructive lung disease started with a methacholine concentration of 0.5 mg/ml. After each inhalation one technically satisfactory FEV 1 was obtained at 30 and at 90 s, of which the lowest value was recorded. If FEV 1 fell less than 10OJo of the pretest value after an inhalation, a fourfold higher concentration was given next. If FEV 1 fell by 10 to 20OJo of the pretest value, a twofold higher concentration was given. The test was stopped if FEV 1 fell by 20OJo of the pretest value or a concentration of 32 mg/ml had been given. The degree of responsiveness was expressed as the concentration of methacholine provoking a 20OJo fall in FEV 1 from pretest value (PC10). PC10 was calculated by interpolation between the results of the last two inhaled concentrations of methacholine. Each subject was then given five inhalations of salbutamol (500 ug), and FEV 1 was measured until it rose to at least 95OJo of the pretest value. To examine the reproducibility of the bronchial responsiveness test, 20 subjects with PC10 ~ 32 mg/ml were tested on a second occasion within 3 to 7 days at the same time of the day. The subjects included 10 with obstructive lung disease and 10 without disease. The PC10 of the 10 healthy subjects were reproducible within one doubling concentration of methacholine (figure 1). For the 10 subjects with obstructive lung disease, seven were within one doubling concentration and all within two doubling concentrations (figure 1).
Data Analysis Contingency tables were analyzed for statistical significance using chi-square tests or, for
TABLE 1 CHARACTERISTICS OF THE STUOY POPULATION BY SEX
Variables Age, mean (SO) yr* Smoking habits t Smokers, % Cigarettes/day, mean (SO) Ex-smokers, % Nonsmokers, % Urban/rural area of residence Urban, %* Past or present physician's diagnosis of asthma, % t Log serum IgE, mean (SO) FEV1, mean (SO) Us %FEV1, mean (SO)§ FEV 1/FVC, mean (SO) • Age at December 31,1987.
t Information from postal questionnaire.
*liVing In the municipality of Bergen.
§ Percent of predicted value from reference 9.
Men (n == 251)
Women (n == 239)
41 (15.7)
43 (15.6)
40 15 (6.7) 26
34
39 12 (5.6) 18 42
70
70
3
4
1.38 (0.48)
1.22 (0.23)
4.1 (0.9) 95 (13.3) 0.81 (0.07)
3.0 (0.7) 96 (14.8) 0.82 (0.06)
-
•
32.0
0 0
/ /
:: 16.0
/
/
0
/
E
/
/
/
/
en
/
.§ 8.0
•/
/ /
0
/
•
/
/
/
/
/
/
/
/
/
/
/
/
1.0
/
/
/
2.0
/ /
/ /
4.0
/
0
/
C>
LJ
o,
/
/
I/)
~
/
/ /
./
/
/
/
//.
0.5
/ /
/
0:5
1:0
2:0
4:0
8:0
16:0
32:0
P(20 (mg/ml) ttest
Fig. 1. Reproducibility of PC 20 measurements within 1 wk in 10 healthy subjects and in 10 subjects with obstructive lung disease. The solid line is the line of identity, and the dashed lines indicate one doubling concentration of methacholine. Open circles = healthy; closed circles = obstructive lung disease.
stratified data, the Mantel-Haenszel extension of the test (11). Possible predictors of bronchial responsiveness, defined as PC10 ~ 32 giL methacholine, were examined in smokers, ex-smokers, and nonsmokers separately through logistic regression analyses. The numbers of smokers, ex-smokers and nonsmokers with PC10 ~ 8 giL methacholine were too small to give stable regression coefficients. Also, the small number of subjects with a high degree of airborne occupational exposure prevented analysis in separate smoking groups when examining the association of bronchial responsiveness to airborne occupational exposure. The logistic regression models were backwards, stepwise. Only statistically significant variables were kept in the final models. Entry and exit levels were set at p = 0.01. All analyses were performed using the BMDP statistical software package (12).
Results Altogether, 20070 of the population had a PC 20 ~ 32 mg/ml, whereas 6070 had a PC l O ~ 8 mg/ml (table 2). Bronchial responsiveness was more prevalent in women than in men (p < 0.01). This tendency was evident in all age groups and regardless of level of bronchial responsiveness (table 3). The sex difference in prevalence of bronchial responsiveness defined as PC20 ~ 32 mg/ml was twice as high in the age group below 50 yr as in that above 50 yr. In both men and women the prevalence of bronchial responsiveness increased by age, being twice as common in the oldest age group (54-73 yr) as' in the youngest age group (18-34 yr). Cigarette smoking was associated with an increased (p < 0.05) prevalence of PC20 ~ 8 mg/ml (table 2), whereas no association was evident between smoking habits
319
BRONCHIAL RESPONSIVENESS IN A NORWEGIAN COMMUNITY
TABLE 2 CRUDE PREVALENCES OF BRONCHIAL RESPONSIVENESS TO METHACHOLINE BY SEX, AGE, SMOKING HABITS, AREA OF RESIDENCE, AND OCCUPATIONAL AIRBORNE EXPOSURE IN A NORWEGIAN GENERAL POPULATION 1987/1988 (n ~ 490) PC20 (giL) Variable
Subjects (n)
8 to ~ 32 mg/ml it was 2.9 Lis (0.8 Lis), and at PC20 > 32 mglml it was 3.8 Lis (0.9 Lis). Seventy percent of all subjects had
an FEV. greater than 90070 predicted (table 4). The mean FEV. was higher for men than for women for all categories of OJoFEVI. In both sexes the prevalence of bronchial responsiveness fell with increasing pulmonary function level. For all levelsof pulmonary function the prevalence of bronchial responsiveness was two to three times higher in women than in men (table 4). The relationship of bronchial responsiveness to potential determinants was examined in logistic regression analyses to achieve simultaneous adjustment for several variables. Sex, age, and pretest air-
TABLE 3 DISTRIBUTION OF BRONCHIAL RESPONSIVENESS TO METHACHOLINE BY SEX AND AGE IN A NORWEGIAN GENERAL POPULATION, 1987/1988 (n = 490)
PC2()(g/L) Subjects (n)
2to 8 to < 32 (0/0)
> 32 (0/0)
Men Women
102
80
0 3
3 6
7 11
90 80
35-54
Men Women
82 90
0 1
0 8
9 16
91 76
55-73
Men Women
67 69
1 9
4 3
21 22
73 87
18-73
Men Women
251 239
0* 4
2 6
11 16
86 74
Age (yr) 18-34
* 100
TABLE 5 ADJUSTED ODDS RATIOS (OR) WITH 95% CONFIDENCE INTERVALS (CI) FOR BRONCHIAL RESPONSIVENESS (PC.... 32 giL METHACHOLINE) BY SEX, AGE. FEV" AND FEV, PERCENT OF PREDICTED (%FEV,) IN SMOKERS, EX·SMOKERS, AND NONSMOKERS OF A NORWEGIAN GENERAL POPULATION 1987/88 (n = 490) Ex-smokers (n = 109)
Nonsmokers (n = 186)
Sex· Women Men Age, yrt X X + 10 yr FEV,* X X-o.5 Us %FEV,* X X·l0 percentage points
Smokers (n = 195)
OR
95% CI
OR
95% CI
OR
95% CI
1 1.4
0.4-4.9
1 4.0
2.3-21 .4
1 8.4
2.5-37.4
1 0 .6
0.4-0.9
1 0.7
0.4-1 .3
1 0.5
0.3-0.8
1 2.2
1.4-3.5
1 3.8
1.9-7.8
1 4.0
2.4-6.7
1 1.9
1.2-2.7
1 2.5
1.5-4.0
1 3.6
2.3-5.8
• Adjusted lor age and FEV, . t Adjusted lor sex and FEV,. f Adjusted lor sex and age.
Discussion
IgE level (figure 2). The relationship of sex- and age-adjusted serum IgE level to bronchial responsiveness,defined as pew ~ 32 giL methacholine, was further analyzed in smokers, ex-smokers, and nonsmokers separately. Irrespective of smoking status, there was no significant association between sex- and age-adjusted serum IgE level and bronchial responsiveness in logistic regression analyses after adjusting for llJoFEV 1 (table 6).
pc,.
pc,. ,,6 9"
40
In this adult, general population study both female sex and older age were associated with higher crude rates of bronchial responsiveness to methacholine. However, after adjustment for airway caliber, in terms of pretest FEV 10 the sex and age relationship to bronchial responsiveness was reversed, the adjusted odds for bronchial responsiveness becoming higher in men than in women and increas-
,,329"
%
30
Fig. 2. Prevalence of bronchial responsiveness to methacholine (PC.... 8 giL and PC.... 32 giL) by serum IgE Z-score in men and women of a Norwegian general populat ion 1987-1988.Two subjects refused blood sampling. Hatched bars = women; closed bars = men .
20
10
