Published ahead of print on July 14, 2005, doi:10.1164/rccm.200503-323OC
© 2005 American Thoracic Society doi: 10.1164/rccm.200503-323OC An Increase in Bronchial Responsiveness Is Associated with Continuing or Restarting SmokingDepartment of Public Health Sciences, King's College London, London, United Kingdom; Institute of Social and Preventive Medicine, University of Basel, Basel, Switzerland; Medical Research Institute (IMIM) and Universitat Pompeu Fabra (UPF), Barcelona, Spain; Division of Respiratory Diseases, IRCCS "San Matteo" Hospital, University of Pavia, Pavia, Italy; University of Verona, Department of Medicine and Public Health, Division of Epidemiology and Statistics, Verona, Italy; Department Allergy, Respiratory Medicine and Sleep, University Hospital (E7), Reykjavik, Iceland; Institute of Epidemiology, GSF—National Research Center for Environment and Health, Neuherberg, Germany; Respiratory Medicine and Allergology, University of Uppsala, Uppsala, Sweden; Keck School of Medicine, University of Southern California, Los Angeles, California; INSERM—The French Institute of Health and Medical Research, Epidemiology, Faculty of Medicine, Paris, France; University of Groningen, Department Epidemiology and Bioinformatics, University Medical Center Groningen, Groningen, The Netherlands; and Department of Thoracic Medicine, Haukeland Hospital, Bergen, Norway Correspondence and requests for reprints should be addressed to Professor S. Chinn, Department of Public Health Sciences, 5th floor Capital House, 42 Weston Street, London SE1 3QD, UK. E-mail: sue.chinn{at}kcl.ac.uk
Rationale: Bronchial responsiveness (BHR) has been found to be associated with smoking, atopy, and lower lung function in cross-sectional studies, but there is little information on determinants of change in adults. Objectives: To analyze change in bronchial responsiveness in an international longitudinal community study. Methods: The study was performed in 3,993 participants in the European Community Respiratory Health Survey who had bronchial responsiveness measured in 1991–1993, when aged 20 to 44 yr, and in 1998–2002. Measurements: Bronchial responsiveness was assessed by methacholine challenge. Serum samples were tested for total IgE, and for specific IgE to four common allergens. Smoking information was obtained from detailed administered questionnaires. Change in bronchial responsiveness was analyzed by change in IgE sensitization, smoking, and lung function, with tests of interaction terms with age and sex. Main Results: Continuing and restarting smokers had increasing bronchial responsiveness, approximately equivalent to a mean reduction in PD20 of 0.68 and 0.75 doubling doses, respectively, over 10 yr, in addition to a small increase explained by decline in FEV1. No other risk factor for change in bronchial responsiveness was identified. Conclusions: Smoking is a risk factor for increasing bronchial responsiveness over and above the effect of decreasing lung function. Neither baseline IgE sensitization nor change in sensitization was shown to be a risk factor for increasing BHR, the latter possibly due to little overall increase or decrease in sensitization.
Key Words: asthma • atopy • bronchial hyperreactivity • immunoglobulin E • pulmonary disease, chronic obstructive Bronchial responsiveness (BHR) is well established as an outcome measure in epidemiologic studies on asthma, particularly in international studies in which comparisons of symptom prevalence may be affected by translation difficulties and cultural differences (1). Factors associated with increased BHR in adults have been identified in cross-sectional community studies, with atopy, smoking, and low lung function being the major predictors (2–6). However, not all authors have found smoking and low lung function to be independent predictors (7). The effects of changes in these factors on change in BHR in the general adult population have been less systematically studied. In a review of the impact of smoking cessation, Willemse and coworkers found only three studies in smokers without chronic symptoms that measured BHR before and after smoking cessation (8). None showed a change in BHR, but this could be attributed to small numbers and low BHR at baseline. Other studies have been performed on patients with chronic obstructive pulmonary disease (COPD) (9–11), asthma (12), or included only smokers (13). Trigg and colleagues reported minor seasonal variation over 2 yr in a study of 122 general practice patients. Grass pollen skin-prick sensitivity was associated with a reduction in BHR in April to June, a season in which grass sensitization might be expected to increase BHR. Atopy, as defined as one or more positive tests, was related to a reduction in BHR during February to April, in only 1 yr of the study (14). Britton and coworkers reported seasonal variation over 1 yr in 60 participants in a community study, and that change in BHR was related to change in FEV1/FVC, but not to symptoms, age, or smoking status; changes in smoking and atopy were not reported (15). Two studies reported long-term variability in BHR in the general population (16, 17), but not factors associated with change. The Normative Aging Study, which studied 435 men and excluded those with symptoms, found no relation of change in BHR over 3 yr to smoking group, or to change in lung function, total IgE, or skin test wheal diameter (18). Hence there is very little information on factors associated with change in BHR in the general population, and none on whether there are differences between men and women. The European Community Respiratory Health Survey II (ECRHS II) (19) was a multicenter follow-up study of participants in ECRHS I, which selected adults aged 20 to 44 yr (20). Based on the above we hypothesized that change in BHR would be related to smoking groups, change in lung function, sensitization to common allergens, and total serum IgE. In addition, we included change in body mass index (BMI), due to the continuing interest in a reported relation between obesity and asthma (21). Interaction terms were tested, and variation between centers examined across much of western Europe.
Subjects Protocols for the ECRHS I and II have been published elsewhere (19, 20) (see the online supplement). Only data for the random sample selected for ECRHS I stage 2, performed in 1991 to 1993 (20), were included in this paper. ECRHS II was performed from 1998 to 2002. Median length of follow-up was 8.9 yr, with an interquartile range from 8.4 to 9.3 yr.
Lung Function
Methacholine Challenge
Total and Specific IgE
Smoking
BMI
Statistical Analysis
A total of 12,698 people in the ECRHS I random sample in the 23 centers were eligible to take part in ECRHS II, of whom 5,309 (41.8%) were untraced or declined any follow-up (Figure 1). A further 1,197 (9.4%) responded to the questionnaire but had no lung function measurement, 1,069 (8.4%) had no BHR measured at ECRHS I, and another 775 (6.1%) had lung function but were either ineligible or declined methacholine challenge at ECRHS II. BHR was measured at the first survey in 5,202 of these 7,647 excluded individuals; mean slope did not differ significantly between those excluded on grounds of no questionnaire, no lung function or no BHR at ECRHS II (p = 0.28), but was lower (greater BHR) than that of those included in the analysis (p < 0.001). Those who smoked at ECRHS I were more likely to be untraced or decline follow-up, 45.6% compared with 38.9% of nonsmokers and 40.5% of ex-smokers, and a little more likely to be atopic (see online supplement). Of the 3,993 with BHR at each survey and in one of the included smoking groups, there were 1,741 (43.6%) never-smokers, 679 (17.0%) sustained quitters, 1,048 (26.2%) smokers, 387 (9.7%) quitters, and 138 (3.5%) restarters. Mean age and FEV1 at ECRHS I were close to those previously reported for all participants with change in lung function in 27 centers (24) (see on-line supplement).
Change in BHR Mean slope was lower (i.e., BHR greater) in women than in men at ECRHS I (Table 1). There was a small, but highly significant (p < 0.0001) overall mean decrease in slope (i.e., increase in BHR) of –0.016 per year (95% confidence interval [CI], –0.024 to –0.008), which was not significantly different between men and women. There was a highly significant difference in change in BHR between smoking groups (p < 0.0001), with smokers having the greatest decline in slope except for the small group of male restarters.
Change in Lung Function and BMI The mean decrease in FEV1 and mean increase in BMI were close to the values previously reported for 27 centers (24) (see on-line supplement).
Change in Sensitization
Effect of Change in Smoking The interactions of smoking group with sex and with age were not statistically significant. Table 3 shows the results of analyzing interactions of smoking with  FEV1 and with FEV1/FVC, and all main effects. The interactions of smoking with change in lung function were statistically significant (p = 0.038 and 0.006). However, when the nonsignificant main effects were dropped from the model (i.e., only center, smoking groups, change in lung function, and the interactions were included), the evidence for the interaction terms was weaker (smoking group by FEV1, p = 0.034; smoking group by FEV1/FVC, p = 0.077). Omitting the latter interaction, the main effect of FEV1/FVC was not significant (–0.010; 95% CI, –0.027 to 0.007), and the interaction with FEV1 of borderline statistical significance (p = 0.058). The simple model with center, smoking group, and FEV1 was therefore considered descriptive of the results. There was no evidence for a relation of change in BHR with change in BMI.
Results estimated by center and combined across centers by meta-analysis are shown in Table 4. The effect of continuing smoking over 10 yr, compared with never-smokers, was approximately equivalent to a decrease of 0.68 PD20 doubling doses (23). Restarters also had an increase in BHR similar to that of smokers, compared with nonsmokers, although only marginally statistically significant as the sample size was smaller (Figure 2). The estimates for smoking group differences did not show any significant heterogeneity across centers. However, the positive regression coefficient of slope on FEV1, which should be interpreted as a mean decline in slope, or increase in BHR, with decline in FEV1, did show weak heterogeneity (p = 0.044). No center had a significantly negative coefficient.
Results in Table 4 were similar to those calculated with center adjustment. Center-adjusted differences in change in BHR from never-smokers are shown in Table 5, with (overall differences between smoking groups p = 0.0001) and without adjustment (p < 0.0001) for change in FEV1. The greater increase in BHR (decline in slope) for smokers and restarters compared with nonsmokers was only slightly greater when unadjusted for FEV1 than when adjusted, so that differences in decline in lung function account for little of the effect of smoking on BHR.
In smokers, quitters, and restarters, there was no additional effect of average between survey smoking intensity (p = 0.781). The relation of change in BHR to smoking intensity did not differ between the three smoking groups (p = 0.303), and was not statistically significant when unadjusted for smoking group (p = 0.613).
In this large population study of the determinants of change in BHR, we found an overall increase in BHR over time. As on average nonsmokers showed no change, the increase was found only in continuing smokers and in those restarting smoking, while recent and long-term quitters showed little change, independent of change in lung function. This pattern was found in all centers even though the mean change in BHR differed between centers. Smokers had a mean decline equivalent to about two-thirds of a doubling dose in PD20 over 10 yr, and restarters about three quarters of a doubling dose. A change of one doubling dose or concentration has been considered clinically important (26), based on early studies which achieved reproducibility within one doubling concentration. The mean increase in BHR associated with continuing or restarting smoking is substantial. It implies that a considerable proportion of smokers or restarters experienced a clinically important change over 10 yr. Loss to follow-up is a problem in any longitudinal study; our response rates are comparable to several recent population studies (27, 28), and greater than that of the one population study of change in BHR (18). Those not followed up had greater BHR at ECRHS I than participants, and were more likely to be smokers, the latter a finding in other studies (29). It can be argued either that those not followed up could include a very responsive group who would not experience further decline, or that deterioration in health might influence participation at ECRHS II, and hence nonparticipants might have experienced greater decline. However, to affect the relationships nonparticipation would need to be differential between smoking groups according to BHR. Among those excluded there was no significant difference in mean BHR at ECRHS I between those excluded as complete nonresponders, those with questionnaire data, and those with lung function but no BHR data at ECRHS II. Multiple imputation of missing data suggested less difference between the smoking groups, but the greater mean increase in BHR in smokers and restarters persisted. IgE sensitization has previously been shown to be strongly associated with BHR in cross-sectional analyses of data from ECRHS I (5, 30), but we could find no relation of sensitization, nor change in sensitization, to change in BHR. Nor was any interaction found between smoking group and baseline sensitization to any allergen. The lack of relation of change in BHR to change in sensitization may be due to lack of power, as there was little change in sensitization between ECRHS I and II. Numbers of "losers" and "gainers" were small, and may be little more than expected from natural variation around the limit of detection. There is evidence that both sensitized and nonsensitized individuals with asthma do show an increase in BHR when exposed to airborne allergens (31). In our study exposure to allergens was not considered, and changes in IgE sensitization may not adequately reflect changes in exposure to allergens. As expected smokers had a greater decline in lung function than other groups (24), but our results show directly that little of the increase in BHR can be ascribed to the decline in lung function. Furthermore, restarters did not have a greater decline in lung function compared with nonsmokers (24), but have here been shown to have a greater increase in BHR. Lim and colleagues found a greater decline in height-adjusted FEV1 and greater increase in BHR in smokers than in ex-smokers, but did not analyze change in BHR adjusted for change in FEV1 (13). Our results differ from those of the Lung Health Study (10), in which quitters showed an increase in BHR and changes in BHR were largely explained by changes in FEV1. However, the Lung Health Study was a randomized controlled trial in which one group was prescribed inhaled ipratropium bromide, and included only patients with mild COPD. It has been shown that BHR is related to FEV1 in individuals with COPD but not in individuals with asthma, suggesting that the association is due to the influence of the airway caliber on BHR. In our study in middle-aged adults, lung function was relatively well preserved, which may account for why the association between change in smoking and BHR was not explained by change in lung function. The greater increase in BHR in continuing smokers compared with never-smokers is consistent with greater differences between smokers and nonsmokers at older ages compared with differences at younger ages in cross-sectional studies (2, 5). It may also explain an overall increase with age, which was more pronounced in participants with symptoms, reported by Rijcken and coworkers (7). However, their results are difficult to interpret, as they analyzed level, not change, in BHR. In a 3-yr follow-up of middle-aged and older men in the Normative Aging Study, Sparrow and colleagues found no significant increase in BHR in smokers or differences between smoking groups (18). Only annual change in basophil count was associated with change in BHR. No basophil counts were performed in the ECRHS. The difference in findings for smoking is likely to be due to the very small numbers of continuing smokers and restarters and the shorter follow-up period in the Normative Aging Study. For smoking our results are supportive of those of Renkema and coworkers (11), who found in a study of nonallergic patients with COPD that smokers had a greater rate of increase in BHR than ex-smokers, but differ in that we did not find a relation of change in BHR with baseline serum total IgE. Willemse and colleagues found a decrease in BHR in a small group of patients with COPD who quit smoking (9), but our results do not support this, as no significant increase in BHR slope was found in the quitters or sustained quitters (Table 1). No differences in change in BHR were found between men and women, nor in the relation of change in BHR with smoking group by sex. The lack of relation of change in BHR to smoking intensity over the period of follow-up may be due to error on recall of amount smoked, to a low threshold of exposure to smoking for an increase in BHR, or to BHR reflecting current inflammation rather than any cumulative effect of smoking. Only smoking has been shown to have an unequivocal effect on change in BHR. Estimates were similar regardless of other variables in the model. The relation of change in BHR to smoking group was unchanged when those who said they ever had asthma at ECRHS II were excluded, although there was a small reduction in the relation to change in FEV1. A recent paper has reported an association between asthma incidence and current or past smoking (32). Change in BHR may precede changes in symptoms, and smoking-related changes in symptoms are beyond the scope of this paper. As in all studies, it is difficult to differentiate between symptoms of asthma and those of COPD. Increasing BHR in continuing and restarting smokers may be an indication of development of either disease.
The co-ordination of ECRHS II was supported by the European Commission, as part of their Quality of Life program. Funding for the individual centers is listed at www.ecrhs.org Personnel and centers taking part in the study are listed at the end of the article. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: S.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. D.J. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.M.L. do not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. U.A.-L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.M.A's unit received $40,000 from Novartis, Spain, for the period 2004–2005 to support a collaborative project unrelated to the present study. I.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. R.d.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. T.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.J. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. N.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. F.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.P.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. P.G.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Coordinating Centre Project Leader: P. Burney; Statistician: S. Chinn; Principal Investigator: D. Jarvis; Project Co-ordinator: J. Knox; Principal Investigator: C. Luczynska; Assistant Statistician: J. Potts; Data Manager: S. Arinze. Steering Committee for ECRHS II Professor Josep M. Antó, Institut Municipal d'Investigació Mèdica (IMIM-IMAS), Universitat Pompeu Fabra (UPF); Professor Peter Burney, King's College London (Project Leader); Dr. Isa Cerveri, University of Pavia; Professor Susan Chinn, King's College London; Professor Roberto de Marco, University of Verona; Dr. Thorarinn Gislason, Iceland University Hospital; Dr. Joachim Heinrich, GSF–Institute of Epidemiology; Assoc Professor Christer Janson, Uppsala University; Dr. Deborah Jarvis, King's College London; Miss Jill Knox, King's College London; Dr. Nino Künzli, University of Basel and University of Southern California Los Angeles; Dr. Bénédicte Leynaert, Institut National de la Santé et de la Recherche Médicale (INSERM); Dr. Christina Luczynska, King's College London; Dr. Françoise Neukirch, Institut National de la Santé et de la Recherche Médicale (INSERM); Dr. J. Schouten, University of Groningen; Dr. Jordi Sunyer, Institut Municipal d'Investigació Mèdica (IMIM-IMAS), Universitat Pompeu Fabra (UPF); Dr. Cecilie Svanes, University of Bergen; Professor Paul Vermeire, University of Antwerp; Dr. Matthias Wjst, GSF–Institute of Epidemiology. List of Principal Investigators and Senior Scientific Team Belgium: South Antwerp & Antwerp City (P. Vermeire, J. Weyler, M. Van Sprundel, V. Nelen). Estonia: Tartu (R. Jogi, A. Soon). France: Paris (F. Neukirch, B. Leynaert, R. Liard, M. Zureik), Grenoble (I. Pin, J. Ferran-Quentin), Pessac (A. Taytard, C. Raherison). Germany: Erfurt (J. Heinrich, M. Wjst, C. Frye, I. Meyer). Iceland: Reykjavik (T. Gislason, E. Bjornsson, D. Gislason, T. Blondal, K. B. Jorundsdottir). Italy: Turin (M. Bugiani, P. Piccioni, E. Caria, A. Carosso, E. Migliore, G. Castiglioni), Verona (R. de Marco, G. Verlato, E. Zanolin, S. Accordini, A. Poli, V. Lo Cascio, M. Ferrari), Pavia (A. Marinoni, S. Villani, M. Ponzio, F. Frigerio, M. Comelli, M. Grassi, I. Cerveri, A. Corsico). Netherlands: Groningen & Geleen (J. Schouten, M. Kerkhof). Norway: Bergen (A. Gulsvik, E. Omenaas, C. Svanes, B. Laerum). Spain: Barcelona (J. M. Antó, J. Sunyer, M. Kogevinas, J. P. Zock, X. Basagana, A. Jaen, F. Burgos), Huelva (J. Maldonado, A. Pereira, J. L. Sanchez), Albacete (J. Martinez-Moratalla Rovira, E. Almar), Galdakao (N. Muniozguren, I. Urritia), Oviedo (F. Payo). Sweden: Uppsala (C. Janson, G. Boman, D. Norback, M. Gunnbjornsdottir), Goteborg (K. Toren, L. Lillienberg, A. C. Olin, B. Balder, A. Pfeifer-Nilsson, R. Sundberg), Umea (E. Norrman, M. Soderberg, K. Franklin, B. Lundback, B. Forsberg, L. Nystrom). Switzerland: Basel (N. Künzli, B. Dibbert, M. Hazenkamp, M. Brutsche, U. Ackermann-Liebrich). UK: Norwich (D. Jarvis, B. Harrison), Ipswich (D. Jarvis, R. Hall, D. Seaton). Centers Taking Part at Their Own Expense France: Bordeaux (A. Taytard, C. Raherison), Montpellier (J. Bousquet, P. Demoly). Germany: Hamburg (K. Richter). USA: Portland (M. Osborne, S. Buist, W. Vollmer, L. Johnson). Received in original form March 1, 2005; accepted in final form July 7, 2005
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
