Abstract
Severe alpha-1-antitrypsin deficiency is the only proven genetic risk factor for chronic obstructive pulmonary disease (COPD). We have assembled a cohort of 44 probands with severe, early-onset COPD, who do not have severe alpha-1-antitrypsin deficiency. A surprisingly high prevalence of females (79.6%) was found. Assessment of the risk to relatives of these early-onset COPD probands for airflow obstruction and chronic bronchitis was performed to determine whether significant familial aggregation for COPD, independent of alpha-1-antitrypsin deficiency, could be demonstrated. First- degree relatives of early-onset COPD probands had significantly lower FEV1 and FEV1/FVC values than control subjects (p < 0.01), despite similar pack-years of smoking. Reduced spirometric values in first-degree relatives of early-onset COPD probands were found only in current or ex-cigarette smokers. The mean FEV1 in current or ex-smoking first-degree relatives was 76.1 ± 20.9% predicted compared to 89.2 ± 14.4% predicted in current or ex-smoking control subjects (p < 0.01); in lifelong nonsmokers, the mean FEV1 was 93.4% predicted for both control subjects and first-degree relatives of early-onset COPD probands. Generalized estimating equations, adjusting for age and pack-years of smoking, demonstrated increased odds of reduced FEV1 and chronic bronchitis in current or ex-smoking first-degree relatives of early-onset COPD probands. Using a new method to estimate relative risk from relative odds, we estimate that the relative risks for FEV1 below 60%, FEV1 below 80%, and chronic bronchitis are each approximately three in current or ex-smoking first-degree relatives of early-onset COPD probands. The increased risk to relatives of early-onset COPD probands for reduced FEV1 and chronic bronchitis, limited to current or ex-smokers, suggests genetic risk factor(s) for COPD that are expressed in response to cigarette smoking.
Chronic obstructive pulmonary disease (COPD), which includes chronic bronchitis and emphysema, is the fourth leading cause of death in the United States and a major cause of morbidity (1, 2). Cigarette smoking is the major known environmental risk factor for the development of COPD (3). Although the dose-response relationship between cigarette smoking and pulmonary function is well established, there is considerable variability in the degree of airflow obstruction that occurs in response to smoking (4). The low percentage of variation in pulmonary function explained by smoking (approximately 15%) and the presence of persons with early- onset, severely reduced pulmonary function suggest that individuals may vary in their genetic susceptibility to the effects of smoking (5).
Alpha-1-antitrypsin deficiency is the only proven genetic risk factor for COPD. Patients with severe alpha-1-antitrypsin deficiency—most commonly, protease inhibitor (PI) type Z— are at increased risk for severe, early-onset COPD (6). In our previous study of the genetic and environmental factors influencing the variable development of pulmonary function impairment in alpha-1-antitrypsin deficiency, we observed striking variability in pulmonary function impairment among PI Z subjects (7). Genetic modeling of these data suggested that additional genetic factors, other than PI type, may influence the variable development of severe COPD in PI Z subjects (8).
Several types of studies have suggested that genetic factors other than severe alpha-1-antitrypsin deficiency may be involved in the development of COPD. In the general population, studies of families have found statistical evidence for a significant genetic contribution to variation in pulmonary function (9). Studies of pulmonary function in twins have provided additional evidence for the role of genetic factors in the determination of pulmonary function in the general population (10-12). In the 1970s, several studies reported an increased prevalence of airflow obstruction among first-degree relatives of subjects with COPD compared with relatives of control subjects (13-15). However, none of these previous studies focused on subjects with severe, early-onset COPD unrelated to alpha-1-antitrypsin deficiency. Identification of genetic factors influencing the development of severe, early- onset COPD unrelated to alpha-1-antitrypsin deficiency could clarify the biochemical mechanisms causing COPD, allow identification of highly susceptible persons, and lead to new therapeutic interventions for COPD.
This report summarizes our initial findings in a cohort of patients with severe early-onset COPD, without severe alpha-1-antitrypsin deficiency, and appropriate age and sex-matched control subjects. Assessment of risk to relatives provides a useful measure of the likelihood that genetic or common familial environmental factors influence a phenotype; higher values of risk to relatives have been shown to increase the likelihood of finding genetic linkage to a phenotype of interest (16, 17). Therefore, as an initial step in determining if genetic factors other than alpha-1-antitrypsin deficiency are involved in severe, early-onset COPD, we studied the risk for airflow obstruction and chronic bronchitis, adjusted for the effects of age and pack-years of cigarette smoking, among 204 first-degree relatives and 45 second-degree relatives of the 44 early-onset COPD probands compared with appropriate control subjects.
Forty-four probands with severe, early-onset COPD were identified primarily from Lung Transplant Programs and Lung Volume Reduction Surgery Programs at Brigham and Women's Hospital and Massachusetts General Hospital. Pulmonary Clinics at these hospitals and at the Brockton/West Roxbury VA Hospital served as additional sources of probands with early-onset COPD. Enrollment criteria for probands with severe early-onset COPD included: FEV1 less than 40% predicted, age younger than 53 yr, and absence of severe alpha-1-antitrypsin deficiency (e.g., PI Z, PI null-null). Subjects were eligible only if they had not yet undergone lung transplantation or lung volume reduction surgery.
After medical record review, 55 subjects who appeared to meet the enrollment criteria were contacted by letter; subsequently, 44 subjects with severe, early-onset COPD were enrolled. Eleven of the contacted subjects were not enrolled: five subjects refused to participate, five subjects had undergone lung volume reduction surgery or lung transplantation before enrollment, and one subject died before she could participate. For the 44 probands with early-onset COPD, all available first-degree relatives, second-degree relatives (restricted to aunts, uncles, and grandparents), and spouses were invited to enroll. In this report, we will discuss 204 first-degree relatives (28 parents, 91 siblings, and 85 children) and 45 second-degree relatives of probands with early-onset COPD.
Control probands were recruited from previous population-based studies in Watertown and East Boston, Massachusetts (18, 19); 169 letters were mailed to subjects with similar age and sex as the probands with early-onset COPD, with a response card asking if they would be willing to participate. Forty responses were obtained; 33 subjects were willing to participate and seven refused to participate. Only participants in the Watertown and East Boston studies who were current or ex-smokers were included as control probands. Of the 33 potential control probands, 11 were excluded because of minimal smoking history, and two subjects could not be scheduled. Thus, 20 control probands were included. Spouses and first-degree relatives of the control probands were asked to participate. We have included 20 control probands, 54 first-degree relatives, and nine spouses. Control probands were not selected on the basis of their pulmonary function; therefore, we have included all 83 control family members (probands, first-degree relatives, and spouses) for comparison with first- and second-degree relatives of probands with early-onset COPD.
Participants gave written informed consent and completed a protocol that included a questionnaire, spirometry, and a blood sample. The protocol was completed at the subjects' homes (> 90% of cases) or at the Outpatient General Clinical Research Center at Brigham and Women's Hospital. The protocol was approved by the Human Research Committee of Brigham and Women's Hospital in Boston.
Each participant completed a modified version of the 1978 ATS-DLD Epidemiology Questionnaire, which was expanded to include questions about passive tobacco smoke exposure and diet (20). The questionnaire was used in two forms, one for adults (> 12 yr of age) and one for children (⩽ 12 yr of age) which was completed by parents. Pack-years of cigarette smoking was calculated as the product of the duration of smoking (in years) and the average number of cigarettes smoked per day, which was divided by 20 to convert to packs. Chronic bronchitis was defined from affirmative responses to questionnaire items for both chronic cough and chronic phlegm production for at least 3 mo per year for at least 2 yr.
Although many items on the children's questionnaire were comparable to the adults' questionnaire, several items were not part of the children's questionnaire. For the 11 children included in this report, data are treated as missing for: chest illnesses within the previous 3 yr, “lung trouble” before 16 yr of age, usual cough, usual phlegm production, and dyspnea. Questions regarding smoking behavior were not included in the children's questionnaire; the 11 children were assumed to be lifelong nonsmokers.
Spirometry was performed with a Survey Tach Spirometer (Warren E. Collins, Braintree, MA). The maneuvers were performed in a standardized manner with the subject seated and wearing a noseclip. To obtain three acceptable measures, the technician asked the subject to perform as many as eight attempts. Subjects were instructed to abstain from inhaled bronchodilator use for 4 h before testing, unless significant respiratory symptoms necessitated bronchodilator use. Spirometry was performed in accordance with ATS specifications (21); we report the best FEV1 value and the FEV1/FVC value from the best-test effort. Height was measured in stocking feet to the nearest 0.5 inch.
Pulmonary function test results are expressed as percent of predicted using predicted equations from Crapo and colleagues (22); for participants younger than 18 yr of age, predicted values were determined from Hsu and colleagues (23).
The PI type of each proband with early-onset COPD was determined by isoelectric focusing of dithioerythritol-treated serum at pH 4.2 to 4.9 in polyacrylamide gels (24). In addition, immunoreactive and functional alpha-1-antitrypsin levels were measured for the probands with early-onset COPD. Immunoreactive alpha-1-antitrypsin levels were measured in serum by ELISA with a mouse monoclonal antibody to human alpha-1-antitrypsin (25). Functional alpha-1-antitrypsin levels were assessed by measurements of serum elastase inhibitory capacity against active site-titrated human leukocyte elastase, essentially as previously described, using the elastase substrate methoxysuccinyl-Ala-Ala-Pro-Val para-nitroanilide (26-28).
The PI types of all first-degree and second-degree relatives of the three probands with early-onset COPD who were Z allele heterozygotes (2 PI MZ, 1 PI FZ) were determined by isoelectric focusing of dithioerythritol-treated serum at pH 4.2 to 4.9 in polyacrylamide gels (24).
Fisher's exact tests, Student's t tests, and univariate odds ratios were calculated with the SAS statistical package (SAS Statistical Institute, Cary, NC) on a SUN Microsystem.
For the estimation of the risk to relatives for airflow obstruction and chronic bronchitis, several statistical issues needed to be addressed. Because multiple subjects were included from each family, the phenotypic values for all participants cannot be assumed to represent independent observations. Therefore, standard logistic regression analysis was not an appropriate method to assess the influence of relationship to a proband with early-onset COPD on the development of airflow obstruction and chronic bronchitis. Generalized estimating equations allow the inclusion of a common familial correlation, as well as covariates such as age and pack-years of smoking (29, 30). We employed generalized estimating equations with SAS (GEE Version 2.03) to assess the odds ratios for abnormal values of FEV1, for chronic bronchitis, and for physician-diagnosed asthma in relatives of probands with early-onset COPD compared with control subjects.
Although the odds ratio provides insight into risk factors for relevant phenotypes, the odds ratio is not equivalent to the relative risk ratio, which is of greater interest to further genetics research (16, 17). If we allow p1 to represent the probability that a first-degree relative has a relevant phenotype (e.g., chronic bronchitis, FEV1 < 60%), and p2 to represent the probability that a control subject has this phenotype, then the relative risk is p1/p2, and the odds ratio is [p1/(1 − p1)]/ [p2/(1 − p2)]. We used an indicator variable (1 for first-degree relatives of probands with early-onset COPD, 0 for control subjects) in logistic regression analysis with generalized estimating equations to find an equation for the logit of p1. Using this equation, we calculated the estimated value of p1 for each first-degree relative. Subsequently, we switched the value of the indicator variable from 1 to 0, to calculate an estimate of p2 for each first-degree relative. The ratio of p1/p2 is calculated for each first-degree relative; the mean value of p1/p2 for all first-degree relatives provides a point estimate for the relative risk.
The probands with early-onset COPD were selected to have severe airflow obstruction at an early age; nonetheless, the severity of the reduction in FEV1 (mean, 16.9% predicted) was striking. An extremely high prevalence of women, 79.6%, was found among the probands with early-onset COPD. The 44 probands with early-onset COPD were compared with 20 control probands with respect to lung function, age, sex, and pack-years of smoking (Table 1). Although the control probands had similar age and sex distributions compared with the probands with early-onset COPD, the control probands had significantly lower mean pack-years of cigarette smoking (p < 0.05). Therefore, a second control group was created for comparison to the probands with early-onset COPD. The 18 control probands and 12 spouses of the probands with early-onset COPD with greater than 10 pack-years of smoking were combined to form a group of 30 smoking-matched control subjects. Probands with early-onset COPD had greatly reduced FEV1 and FEV1/FVC values compared with control probands and smoking-matched control subjects (p < 0.01). All of the probands with early-onset COPD and the control subjects were Caucasian.
| Group | Age (yr) | Pack-years | FEV1/FVC (% Pred ) | FEV1(% Pred ) | Sex (% female) | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Early COPD probands, n = 44 | 47.2 ± 5.6 | 38.8 ± 22.5 | 37.2 ± 11.8 | 16.9 ± 6.1 | 79.6% | |||||
| Control probands, n = 20 | 49.3 ± 6.8 | 24.6 ± 16.7† | 91.4 ± 11.0‡ | 89.3 ± 15.9‡ | 75% | |||||
| Smoking-matched control subjects, n = 30 | 49.6 ± 6.2 | 36.2 ± 26.1 | 90.1 ± 11.0‡ | 86.5 ± 16.8‡ | 50%† |
The PI types of the 44 probands with early-onset COPD included 38 PI M, 2 PI MZ, 1 PI FZ, and 3 PI MS. The functional and immunologic levels of alpha-1-antitrypsin were consistent with these PI types. The functional alpha-1-antitrypsin levels were 35.9 ± 9.4 μM (mean ± SD), with a range from 19.0 to 55.6 μM; immunologic levels were similar. No evidence for severe alpha-1-antitrypsin deficiency based on low immunologic level or functional activity was present for any of the 44 probands with early-onset COPD.
To determine that a diagnosis of COPD was correct in our probands with early-onset COPD, available information regarding chest CT scans and lung pathologic samples were reviewed. Chest CT scan reports from Brigham and Women's Hospital or Massachusetts General Hospital were available for 28 subjects; in 27 of 28 reports, gross emphysema was noted; for the remaining subject without gross emphysema, attenuated pulmonary vasculature consistent with emphysema was noted. Eight other subjects had chest CT reports available for review from outside hospitals; for seven of eight subjects, emphysema was noted. Thus, chest CT scan evidence for emphysema was available for 35 of our 44 probands with early-onset COPD. Seventeen subjects had lung pathologic specimens; all specimens showed emphysema. Thus, the chest CT scan and pathology data, at least one of which was available for most of the probands with early-onset COPD, revealed that the COPD probands had emphysema associated with their severe airflow obstruction.
Preliminary evidence that familial factors influence the development of early-onset COPD was provided by questionnaire responses of the probands with early-onset COPD regarding chronic bronchitis and emphysema in their parents (Figure 1). Compared with both control groups, probands with early- onset COPD reported higher rates of paternal chronic bronchitis and maternal emphysema (p < 0.05). Of note, when the prevalence of parental COPD (chronic bronchitis and/or emphysema) was compared between probands with early-onset COPD and control subjects, a higher rate of parental COPD was reported by probands with early-onset COPD (p < 0.01); 58% of probands with early-onset COPD indicated that at least one parent had COPD. The univariate odds ratio (with logit confidence intervals) for probands with early-onset COPD compared with control probands for parental COPD was 7.9 (95% CI, 2.0 to 30.9); for probands with early-onset COPD compared with smoking-matched control subjects, the odds ratio for parental COPD of 4.4 (95% CI, 1.5 to 12.4) was also significantly elevated.
Fig. 1. Questionnaire reports of parental chronic bronchitis (CBR), emphysema, and COPD (chronic bronchitis and/or emphysema) by probands with early-onset COPD, control probands, and smoking-matched control subjects. *p < 0.05 compared with probands with early-onset COPD; **p < 0.01 compared with probands with early-onset COPD by Exact Test. For all measures, probands with early-onset COPD reported a higher prevalence of airway disease, which was statistically significant for maternal emphysema, paternal chronic bronchitis, and either parent with COPD.
[More] [Minimize]The 204 first-degree relatives and 45 second-degree relatives of probands with severe, early-onset COPD were compared with 83 control family members for spirometry, age, and pack-years of smoking (Table 2). Despite similar values for age and pack-years of smoking, first-degree relatives of probands with early-onset COPD had significantly lower mean values for FEV1 (percent predicted) and FEV1/FVC (percent predicted) than did control subjects (p < 0.01). In recognition of the increased likelihood of airflow obstruction with increasing age, second-degree relatives were selected to include older subjects— all available aunts, uncles, and grandparents of probands with early-onset COPD. Not surprisingly, these second-degree relatives had higher mean values for age and pack-years of smoking than did control subjects. Although the second-degree relatives had lower FEV1 and FEV1/FVC than did control subjects, this initial analysis does not indicate whether increased smoking intensity or another factor (e.g., genetic susceptibility) is involved in the lower spirometric values for second-degree relatives.
| Group | FEV1/FVC (% Pred ) | FEV1(% Pred ) | Age (yr) | Pack-yr | ||||
|---|---|---|---|---|---|---|---|---|
| First-degree relatives, n = 204 | 87.6 ± 13.8‡ | 83.9 ± 19.7‡ | 40.7 ± 18.9 | 15.6 ± 24.2 | ||||
| Second-degree relatives, n = 45 | 89.8 ± 10.5‡ | 83.3 ± 19.5† | 62.9 ± 14.7‡ | 28.7 ± 30.2‡ | ||||
| Control family members, n = 83 | 94.8 ± 9.1 | 91.0 ± 14.4 | 45.0 ± 16.3 | 12.8 ± 20.0 |
Although first-degree relatives of probands with early- onset COPD had statistically significant reductions in mean FEV1 values compared with control subjects, the magnitude of the difference was not large (7.1%). However, current or ex-smoking first-degree relatives of probands with early-onset COPD had substantially lower FEV1 and FEV1/FVC than did current or ex-smoking control subjects (p < 0.01), despite similar values for age and pack-years of smoking (Table 3). The distribution of FEV1 values for current or ex-smoking first- degree relatives is clearly shifted toward lower FEV1 values when compared with current or ex-smoking control subjects (Figure 2). No significant differences in FEV1 or FEV1/FVC were found between lifelong nonsmoking first-degree relatives and nonsmoking control subjects; in fact, the mean FEV1 values (93.4% predicted) were identical in each group.
| Group | FEV1/FVC (% Pred ) | FEV1(% Pred ) | Age (yr) | Pack-yr | ||||
|---|---|---|---|---|---|---|---|---|
| Smoking first-degree relatives, n = 112 | 83.5 ± 16.1† | 76.1 ± 20.9† | 45.9 ± 17.3 | 28.5 ± 26.6 | ||||
| Smoking control subjects, n = 48 | 94.3 ± 10.3 | 89.2 ± 14.4 | 48.6 ± 13.9 | 22.1 ± 22.1 | ||||
| Nonsmoking first-degree relatives, n = 92 | 92.7 ± 7.6 | 93.4 ± 12.9 | 34.4 ± 18.9 | 0 | ||||
| Nonsmoking control subjects, n = 35 | 95.5 ± 7.2 | 93.4 ± 14.2 | 39.9 ± 18.2 | 0 |
Fig. 2. FEV1 (percent predicted) in current or ex-smoking first-degree relatives of probands with early-onset COPD compared with current or ex-smoking control subjects. Solid bars correspond to first-degree relatives of probands with early-onset COPD; open bars correspond to control subjects. The distribution for first-degree relatives is clearly shifted toward low FEV1 values. Although 7% of first-degree relatives had FEV1 values below 40% predicted, this degree of severe impairment was not observed in the control group.
[More] [Minimize]PI types were assessed for first-degree and second-degree relatives of probands with early-onset COPD who were carriers of the Z allele. Four first-degree relatives were found to be PI MZ; the mean FEV1 for these four (82.7%) was similar to the mean FEV1 for the 199 other first-degree relatives (83.9%). Four second-degree relatives (3 PI MZ, 1 PI SZ) were Z allele heterozygotes; although there was a statistically nonsignificant trend for lower FEV1 values in these four (69.0%) than in the 41 other second-degree relatives (84.6%), there was also a trend for greater pack-years among the Z allele-carrying second-degree relatives (50.3 pack-years) than the other second-degree relatives (26.6 pack-years).
From the standpoint of future genetic studies, the critical issue is whether relatives of probands with early-onset COPD have increased risk for reduced FEV1 and chronic bronchitis beyond the predicted effects of smoking. To account for potential familial correlations as well as the effects of age and pack-years of smoking, generalized estimating equations were used to calculate odds ratios for various levels of reduction in FEV1 (Table 4). When all first-degree relatives were compared with all control subjects, increased risk of FEV1 below 80% (odds ratio, 3.4; 95% CI, 1.5 to 7.8) and a trend toward increased risk of FEV1 below 60% (odds ratio, 2.5; 95% CI, 0.8 to 7.7) were demonstrated. Stratification by smoking status revealed that this risk was exclusively found in current or ex-smokers; among cigarette smokers, the odds ratio was 4.5 (95% CI, 1.8 to 11.5) for FEV1 below 80% and 3.5 (95% CI, 1.0 to 12.9) for FEV1 below 60%. Lifelong nonsmokers had no increased risk for reduced FEV1.
| Relative Odds of FEV1 Below Designated % Predicted (95% Confidence Intervals) | ||||||
|---|---|---|---|---|---|---|
| Group | FEV1 < 100% | FEV1 < 80% | FEV1 < 60% | |||
| All | 1.14 | 3.42 | 2.45 | |||
| First-degree relatives | (0.54 to 2.43) | (1.50 to 7.76) | (0.78 to 7.65) | |||
| Smoking | 1.77 | 4.50 | 3.53 | |||
| First-degree relatives | (0.60 to 5.24) | (1.77 to 11.47) | (0.97 to 12.85) | |||
| Nonsmoking | 0.82 | 1.85 | 0.45 | |||
| First-degree relatives | (0.35 to 1.90) | (0.53 to 6.44) | (0.04 to 5.16) | |||
No significantly increased risk for reduced FEV1 was found for second-degree relatives when age and pack-years of cigarette smoking were included as covariates. For FEV1 below 60%, the odds ratio for second-degree relatives was 0.5 (95% CI, 0.1 to 3.5), and for FEV1 below 80%, the odds ratio for second-degree relatives was 1.5 (95% CI, 0.4 to 5.3).
To estimate the actual relative risk, rather than the odds ratio, for reduced FEV1 among first-degree relatives of probands with early-onset COPD, the logistic regression equations generated with the generalized estimating equation approach were analyzed (see Methods for methodology). For current or ex-smoking first-degree relatives of probands with early-onset COPD, the relative risk of FEV1 < 80% was 2.7 and the relative risk of FEV1 < 60% was 3.1.
To determine if environmental exposure variables associated with the reduced spirometric values in first-degree relatives of probands with early-onset COPD could be identified, we assessed a number of potential risk factors such as smoking and respiratory infections. No differences in smoking behavior, including age that smoking was started, average number of cigarettes smoked per day, or smoking other than cigarettes (pipe, cigar, nontobacco products), were found between first-degree relatives of probands with early-onset COPD and control subjects. Similarly, no differences in acute respiratory illnesses (assessed with questions regarding pneumonia, acute bronchitis, and chest illnesses within the previous 3 yr), or “lung trouble” before 16 yr of age were found. Elevated rates of childhood home passive smoke exposure and dust exposure at work, of borderline statistical significance, were reported by first-degree relatives of probands with early-onset COPD (p = 0.04 for each comparison).
Although a trend toward higher rates of cough was noted in first-degree relatives of probands with early onset COPD (p = 0.07), increased rates of phlegm production or chronic bronchitis were not found. To determine if smoking status influenced the development of chronic bronchitis and related symptoms, separate analyses were performed for current or ex-smokers and lifelong nonsmokers. Among smokers, significantly higher rates of usual cough and chronic bronchitis were found in the first-degree relatives of probands with early-onset COPD compared with control subjects; no differences were found in the low rates of chronic bronchitis and related symptoms in lifelong nonsmokers (Table 5).
| Group | Proportion of Affirmative Responses (%) | |||||
|---|---|---|---|---|---|---|
| Cough | Phlegm | Chronic Bronchitis | ||||
| Smoking first- degree relatives | 49/112 (44)* | 44/112 (39) | 34/112 (30)* | |||
| Smoking control subjects | 11/48 (23) | 13/48 (27) | 6/48 (13) | |||
| Nonsmoking first- degree relatives | 7/81 (9) | 6/81 (7) | 2/92 (2) | |||
| Nonsmoking control subjects | 4/35 (11) | 5/35 (14) | 2/35 (6) | |||
Although a trend toward more attacks of wheezing was reported in first-degree relatives of probands with early-onset COPD compared with control subjects (p = 0.054), no differences in chest wheeziness apart from colds or physician-diagnosed asthma were noted. However, a high rate of physician-diagnosed asthma was reported among the control subjects (21.7%).
As in the analysis of reduced FEV1, generalized estimating equations were used to perform logistic regression analysis of chronic bronchitis and physician-diagnosed asthma for various groups of relatives, with adjustment for age and pack-years of smoking (Table 6). For physician-diagnosed asthma, the odds ratios for first-degree or second-degree relatives of probands with early-onset COPD were not significantly elevated. For chronic bronchitis, the group of all first-degree relatives showed a trend toward higher risk for chronic bronchitis, with an odds ratio of 2.0 (95% CI, 0.7 to 5.6). As with reduced FEV1, this increased risk was found exclusively in current or ex-smokers, who had a significantly elevated odds ratio of 3.6 (95% CI, 1.1 to 11.5). Using our approach to estimate relative risk from logistic regression equations, the estimated relative risk for chronic bronchitis in all first-degree relatives was 1.8. In current or ex-smoking first-degree relatives of probands with early-onset COPD, the estimated relative risk was 2.7, whereas the estimated relative risk in lifelong nonsmoking first-degree relatives was 0.3. Second-degree relatives did not have a significantly increased risk for chronic bronchitis.
| Group | Relative Odds of Chronic Bronchitis or Physician-diagnosed Asthma (95% Confidence Intervals) | |||
|---|---|---|---|---|
| Chronic Bronchitis | Asthma | |||
| All | 2.04 | 0.72 | ||
| First-degree relatives | (0.73 to 5.64) | (0.37 to 1.40) | ||
| Smoking | 3.63 | 1.36 | ||
| First-degree relatives | (1.14 to 11.51) | (0.53 to 3.50) | ||
| Nonsmoking | 0.27 | 0.38 | ||
| First-degree relatives | (0.02 to 3.88) | (0.14 to 1.04) | ||
| All | 1.63 | 0.56 | ||
| Second-degree relatives | (0.53 to 5.06) | (0.15 to 2.13) | ||
Chronic obstructive pulmonary disease is likely a complex condition, influenced by multiple genetic and/or environmental risk factors. One critical environmental risk factor, cigarette smoking, is well-established; one genetic risk factor, severe alpha-1-antitrypsin deficiency, increases the risk of developing COPD in a small percentage of the population (approximately 1 in 3,000 in the United States) (31). In an effort to identify additional genetic risk factors for COPD, we have assembled a group of 44 probands with severe, early-onset COPD as well as 204 of their first-degree relatives and 45 of their older second-degree relatives. By focusing our efforts on subjects with severe COPD at an early age, the likelihood of a significant genetic contribution to disease etiology may be increased, analogous to other complex conditions, including breast cancer and glaucoma (32, 33).
Since the discovery of alpha-1-antitrypsin deficiency, studies regarding the epidemiology and clinical characteristics of subjects with severe, early-onset COPD have been limited. Although the radiologic characteristics of giant bullous emphysema in nine young male subjects were described (34), epidemiologic analysis focusing on patients with severe COPD at an early age has not been performed. In this study, we identified a cohort of patients, mostly female, with COPD out of proportion to their age and smoking histories.
In our group of 44 unrelated probands with severe, early-onset COPD, most subjects had the common PI M type of alpha-1-antitrypsin; three subjects had type PI MS and three were heterozygotes for the Z allele. None of our probands with early-onset COPD had severely reduced functional or immunologic levels of alpha-1-antitrypsin. Available information regarding chest CT scans and lung pathologic samples from probands with early-onset COPD confirmed that the probands with early-onset COPD had emphysema associated with their severe airflow obstruction.
We found a strikingly elevated prevalence (79.6%) of women within our early-onset COPD group. Our finding contrasts with the historically higher rates of COPD in men at later ages, traditionally attributed to greater prevalence of cigarette smoking among men (1, 35). The female predominance in our sample clearly differs from previous studies of severe COPD. O'Donnell and Webb (36) studied factors influencing breathlessness in 37 subjects with COPD and a mean FEV1 38% of predicted. Their sample included 30 men and seven women, with a mean age of approximately 67 years. Wegner and coworkers (37) studied the relationships between exercise capacity, dyspnea, and pulmonary function in 62 subjects with severe COPD and a mean FEV1 39% of predicted. Their sample included 51 men and 11 women; the mean age of the patients was 66 yr. Postma and coworkers (38) studied the natural history of severe COPD in 129 subjects with a mean FEV1 of 0.61 L (percent predicted not reported). The mean age of their sample was 54 yr; only 19% of their sample were women. Damsgaard and Kok-Jensen (39) studied 187 patients with severe COPD and a mean FEV1 of 0.9 L (percent predicted not reported). Their sample included 54 women and 133 men; the mean age of their subjects was 57 yr. Of interest, among the nine subjects younger than 40 yr of age in the Damsgaard and Kok-Jensen study, six were women and only three were men.
The etiology of the female predominance in our sample compared with previous studies is unclear. However, our sample included, on average, younger subjects with lower FEV1 values than the other series listed. The mean FEV1 in our series was 0.49 L, with a mean percent predicted FEV1 of 16.9%; the mean age of our patients was 47.2 yr. In addition, we rigorously excluded severe alpha-1-antitrypsin deficiency, which the other series did not report.
Although we contacted 55 potential probands with early-onset COPD, only 44 were enrolled. Among the 11 subjects who were not enrolled, a similarly high prevalence of women (nine of 11) was noted; therefore, reluctance of male subjects to participate in our study was not an explanation for the female predominance. Other explanations for our observed female predominance among probands with early-onset COPD include the possibility that men with similar degrees of severe airflow obstruction may be more likely to die from coronary artery disease, potentially related to the effects of cigarette smoking. In addition, relatively young women with very severe COPD may be more likely to be referred to lung transplant/lung volume reduction surgery centers by their local health care providers. Thus, the elevated prevalence of women could represent a survivor effect or referral bias. In summary, although our results differ from previous studies, we contend that by carefully excluding severe alpha-1-antitrypsin deficiency and by focusing on very severe disease at an early age, we have identified a distinct and interesting subset of subjects with COPD. The mechanism responsible for the loss of lung function in this group, however, remains to be defined.
The questionnaire responses of the probands with early-onset COPD (Figure 1) suggested an elevated prevalence of COPD in their parents. Among first-degree relatives of probands with early-onset COPD, we found reduced FEV1 and FEV1/FVC when compared with control subjects with similar age and pack-years of smoking (Table 2). Of interest, significant reductions in FEV1 and FEV1/FVC were limited to current or ex-smokers. For current or ex-smoking first-degree relatives, significantly increased odds of FEV1 below 80% predicted (odds ratio, 4.5), independent of smoking intensity, were demonstrated; a trend toward increased odds of FEV1 below 60% predicted (odds ratio, 3.5) was also found. No increased risk for reduced FEV1 was demonstrated for nonsmoking relatives. Although we attempted to address the nonindependence of phenotypic values within a family with generalized estimating equations, our application of this approach included only one common familial correlation. Thus, no distinction was made between different familial correlations (e.g., sibling-sibling, spouse-spouse). The inclusion of the spouse-spouse correlation is the most intuitively unsatisfying; however, only six early COPD families and one control family included both the father and the mother of the proband, so inclusion of spouse-spouse pairs in the estimation of the familial correlation effect was infrequent.
We have developed a method to estimate relative risk from the generalized estimating equation logistic regression model. Point estimates for relative risk in current or ex-smoking first-degree relatives of probands with early-onset COPD were 2.7 for FEV1 below 80% and 3.1 for FEV1 below 60%. Although these relative risk values are modest in comparison with classic Mendelian traits (e.g., cystic fibrosis sibling risk of 500) (16), they are greater than the estimated risk to first-degree female relatives of probands with breast cancer of approximately 2.1 (40).
Although a high rate of participation (44 of 55) was found among probands with early-onset COPD, a much lower recruitment rate was noted among control probands. Only 20 control probands were recruited from 169 letters to previous participants in population-based respiratory studies. However, among 40 responders to our initial contact letter, 33 (82.5%) indicated that they were willing to participate. The low rate of response to our initial mailing likely relates, at least in part, to the long interval since the performance of the East Boston and Watertown studies—at least 10 yr in each case. Twenty-one of our initial letters were returned undelivered to us. Thus, the low recruitment rate among control probands is a potential problem in interpreting our results; however, it is unclear if any bias that resulted from the low recruitment rate would lead to falsely low or high rates of pulmonary problems among the control population.
Several previous studies have reported reduced spirometric values in first-degree relatives of patients with COPD without early-onset disease. In 1970, Larson and coworkers (13) found airflow obstruction in 23% of first-degree relatives of patients with COPD but in only 9% of spouse control subjects. Although this study did show familial aggregation for airflow obstruction, alpha-1-antitrypsin deficiency was not rigorously excluded as a potential contributor to airflow obstruction. In addition, a higher percentage of first-degree relatives of patients with COPD than control subjects were smokers, so at least some of the observed differences in airflow obstruction may have related to differences in smoking behavior. Kueppers and colleagues (14) reported that the FEV1 levels among siblings of subjects with COPD were significantly lower than the siblings of control subjects; significant differences in pulmonary function between sibs of subjects with COPD and control subjects remained after adjustment for smoking history. A large study of COPD in families was performed by Cohen and colleagues (15). Variance components analysis of their data suggested a significant genetic contribution to FEV1 and FEV1/FVC (41).
In addition to increased risk for reduced FEV1, we found an increased risk for chronic bronchitis among current or ex-smoking first-degree relatives of probands with early-onset COPD. As with FEV1, no differences were found in lifelong nonsmoking subjects. The estimated relative risk for chronic bronchitis in current or ex-smoking first-degree relatives of probands with early-onset COPD was 2.7.
Higgins and Keller (42) also found familial aggregation for chronic bronchitis in a sample of 9,226 subjects from the general population. When at least one parent had chronic bronchitis, they found that the prevalence of chronic bronchitis among offspring was 3-fold higher than if neither parent had chronic bronchitis; this elevated risk for chronic bronchitis was limited to subjects 16 to 39 yr of age. They did not adjust for smoking status; they speculated that their observations regarding chronic bronchitis reflected familial aggregation of smoking habits. Our data are more suggestive of an influence by genetic factor(s) or by a common familial environmental factor other than smoking.
The increased risk to first-degree relatives of probands with early-onset COPD for reduced FEV1 and chronic bronchitis could be caused by genetic or shared environmental factors. To determine if environmental factors were responsible for the familial aggregation of reduced FEV1 and chronic bronchitis in our families with early-onset COPD, we examined a variety of smoking-related and other environmental exposures. No significant differences in smoking behavior or acute respiratory illnesses were found between first-degree relatives of probands with early-onset COPD and control subjects. Borderline increased rates of childhood home passive smoke exposure and dust exposure at work were found, but the significance of these findings in light of the multiple statistical comparisons performed is uncertain. We did not attempt to quantify the extent of childhood passive smoke exposure or dust exposure at work. Further study of childhood passive smoke exposure and dust exposure at work in early-onset COPD is warranted; however, we have not found compelling evidence that they are critical determinants of the development of COPD in our pedigrees. In addition, we did not assess for every potential environmental risk factor; for example, we did not assess for the effects of injecting drug use, which has been suggested as a potential environmental risk factor for COPD (43).
The absence of identified environmental factors to account for the observed increased risk to relatives of probands with early-onset COPD for airflow obstruction and chronic bronchitis is consistent with a role for genetic factors in the development of early-onset COPD. The observed pattern of increased risk for reduced FEV1 and chronic bronchitis found only in current or ex-smoking first-degree relatives of probands with early-onset COPD may relate to genotype-by-environment interaction between cigarette smoking and one or more unidentified genetic factors.
We did not find an increased risk for asthma among first-degree relatives of probands with early-onset COPD. However, an unexpectedly high prevalence of asthma was noted among our control subjects (21.7%). Although only two of 20 control probands reported physician-diagnosed asthma, 16 of 63 (25.4%) other control family members reported physician-diagnosed asthma. Thus, self-selection by other control family members to participate in our study likely contributed to the high rate of asthma observed. Therefore, we cannot exclude the possibility that common genetic factors influence the development of COPD and asthma based on our sample.
Our 44 probands with severe COPD at an early age included 80% women. However, among the first-degree relatives of probands with early-onset COPD, a similar percentage of men (11.5%) and women (10.3%) had FEV1 values below 60% predicted. Although our families were selected through early-onset probands, a much higher rate of reduced FEV1 was found among older relatives. Only 5.2% of first- degree relatives younger than 60 yr of age had FEV1 < 60% predicted, whereas 41.9% of subjects older than 60 yr of age had FEV1 values < 60%. However, nine first-degree relatives with FEV1 < 60% predicted at < 60 yr of age have been enrolled; three subjects would have met our stringent criteria to be probands for the study (FEV1 < 40% predicted and < 53 yr of age). Of interest, seven of these nine relatives with early-onset COPD were women. Thus, some evidence for female predominance and for early-onset disease is present in the first-degree relatives of probands with severe, early-onset COPD.
In summary, we have studied 204 first-degree relatives and 45 second-degree relatives of a series of predominantly female probands with severe, early-onset COPD. Increased risk for reduced FEV1 and chronic bronchitis has been found for the first-degree relatives, exclusively among current or ex-smoking subjects. These data suggest that familial factors other than PI type, likely genetic, influence the development of airflow obstruction and chronic bronchitis.
The writers wish to thank Dana Mandel and Krista McQueeney for assisting with data entry and Vincent Carey for helpful discussions. They are especially thankful for the enthusiastic support for this study from the members of the early-onset COPD and control families.
Supported by National Institutes of Health Grants NCRR GCRC MO1 RR02635 to the Brigham and Women's Hospital General Clinical Research Center, P50 HL56383 and HL48621 to the Brigham and Women's Hospital Division of Pulmonary and Critical Care Medicine, and HL46440 to the University of Utah Health Sciences Center.
| 1. | Feinleib M., Rosenberg H. M., Collins J. G., Delozier J. E., Pokras R., Chevarley F. M.Trends in COPD morbidity and mortality in the United States. Am. Rev. Respir. Dis.140(Suppl.)1989S9S18 |
| 2. | Kochanek K. D., Hudson B. L.Advance report of final mortality statistics, 1992. Monthly Vital Statist. Rep.431995173 |
| 3. | U.S. Department of Health and Human Services. 1984. The health consequences of smoking: chronic obstructive lung disease. A report of the Surgeon General. U.S. Department of Health and Human Services, Office on Smoking and Health, Bethesda, MD. (DHHS Publication No. (PHS)84-50205. |
| 4. | Burrows B., Knudson R. J., Cline M. G., Lebowitz M. D.Quantitative relationships between cigarette smoking and ventilatory function. Am. Rev. Respir. Dis.1151977195205 |
| 5. | Silverman E. K., Speizer F. E.Risk factors for the development of chronic obstructive pulmonary disease. Med. Clin. North Am.801996501522 |
| 6. | Larsson C.Natural history and life expectancy in severe alpha 1-antitrypsin deficiency, Pi Z. Acta Med. Scand.2041978345351 |
| 7. | Silverman E. K., Pierce J. A., Province M. A., Rao D. C., Campbell E. J.Variability of pulmonary function in alpha 1-antitrypsin deficiency: clinical correlates. Ann. Intern. Med.1111989982991 |
| 8. | Silverman E. K., Province M. A., Campbell E. J., Pierce J. A., Rao D. C.Variability of pulmonary function in alpha-1-antitrypsin deficiency: residual family resemblance beyond the effect of the Pi locus. Hum. Hered.401990340355 |
| 9. | Lewitter F. I., Tager I. B., McGue M., Tishler P. V., Speizer F. E.Genetic and environmental determinants of level of pulmonary function. Am. J. Epidemiol.1201984518529 |
| 10. | Redline S., Tishler P. V., Rosner B., Lewitter F. I., Vandenburgh M., Weiss S. T., Speizer F. E.Genotypic and phenotypic similarities in pulmonary function among family members of adult monozygotic and dizygotic twins. Am. J. Epidemiol.1291989827836 |
| 11. | Hankins D., Drage C., Zamel N., Kronenberg R.Pulmonary function in identical twins raised apart. Am. Rev. Respir. Dis.1251982119121 |
| 12. | Webster P. M., Lorimer E. G., Man S. F. P., Woolf C. R., Zamel N.Pulmonary function in identical twins: comparison of nonsmokers and smokers. Am. Rev. Respir. Dis.1191979223228 |
| 13. | Larson R. K., Barman M. L., Kueppers F., Fudenberg H. H.Genetic and environmental determinants of chronic obstructive pulmonary disease. Ann. Intern. Med.721970627632 |
| 14. | Kueppers F., Miller R. D., Gordon H., Hepper N. G., Offord K.Familial prevalence of chronic obstructive pulmonary disease in a matched pair study. Am. J. Med.631977336342 |
| 15. | Cohen B. H., Ball W. C., Brashears S., Diamond E. L., Kreiss P., Levy D. A., Menkes H. A., Permutt S., Tockman M. S.Risk factors in chronic obstructive pulmonary disease (COPD). Am. J. Epidemiol.1051977223232 |
| 16. | Lander E. S., Schork N. J.Genetic dissection of complex traits. Science265199420372048 |
| 17. | Risch N.Linkage strategies for genetically complex traits: II. The power of affected relative pairs. Am. J. Hum. Genet.461990229241 |
| 18. | Dockery D. W., Ware J. H., Ferris B. G., Glicksberg D. S., Fay M. E., Spiro A. I., Speizer F. E.Distribution of forced expiratory volume in one second and forced vital capacity in healthy, white, adult never-smokers in six U.S. cities. Am. Rev. Respir. Dis.1311985511520 |
| 19. | Tager I. B., Rosner B., Tishler P. V., Speizer F. E., Kass E. H.Household aggregation of pulmonary function and chronic bronchitis. Am. Rev. Respir. Dis.1141976485492 |
| 20. | Ferris B. G.Epidemiology standardization project. Am. Rev. Respir. Dis.118(Suppl.)19781120 |
| 21. | Gardner R. M., Hankinson J. L., Clausen J. L., Crapo R. O., Johnson R. L., Epler G. R.Standardization of Spirometry: 1987 update. Official statement of the American Thoracic Society. Am. Rev. Respir. Dis.136198712851298 |
| 22. | Crapo R. O., Morris A. H., Gardner R. M.Reference spirometric values using techniques and equipment that meet ATS recommendations. Am. Rev. Respir. Dis.1231981659664 |
| 23. | Hsu K. H. K., Jenkins D. E., Hsi B. P., Bourhofer E., Thompson V., Tanakawa N., Hsieh G. S. J.Ventilatory functions of normal children and young adults: Mexican-American, white, and black. I. Spirometry. J. Pediatr.9519791423 |
| 24. | Pierce J. A., Eradio B. G.Improved identification of antitrypsin phenotypes through isoelectric focusing with dithioerythritol. J. Lab. Clin. Med.941979826831 |
| 25. | Silverman E. K., Province M. A., Rao D. C., Pierce J. A., Campbell E. J.A family study of the variability of pulmonary function in alpha-1-antitrypsin deficiency: quantitative phenotypes. Am. Rev. Respir. Dis.142199010151021 |
| 26. | Liou T. G., Campbell E. J.Quantum proteolysis resulting from release of single granules by human neutrophils. J. Immunol.157199626242631 |
| 27. | Boudier C., Pelletier A., Pauli G., Bieth J. G.The functional activity of alpha-1-proteinase inhibitor in bronchoalveolar lavage fluids from healthy human smokers and non-smokers. Clin. Chim. Acta1321983309315 |
| 28. | Powers J. C., Gupton B. F.Reaction of serine proteases with aza-amino acid and aza-peptide derivatives. Methods Enzymol.461977208216 |
| 29. | Zeger S. L., Liang K.-Y.Longitudinal data analysis for discrete and continuous outcomes. Biometrics421986121130 |
| 30. | Liang K.-Y., Zeger S. L.Regression analysis for correlated data. Annu. Rev. Pub. Health1419934368 |
| 31. | Silverman E. K., Miletich J. P., Pierce J. A., Sherman L. A., Endicott S. K., Broze G. J., Campbell E. J.Alpha-1-antitrypsin deficiency: high prevalence in the St. Louis area determined by direct population screening. Am. Rev. Respir. Dis.1401989961966 |
| 32. | Hall J. M., Lee M. K., Newman B., Morrow J. E., Anderson L. A., Huey B., King M. C.Linkage of early-onset familial breast cancer to chromosome 17q21. Science250199016841689 |
| 33. | Stone E. M., Fingert J. H., Alward W. L. M., Nguyen T. D., Polansky J. R., Sunden S. L. F., Nishimura D., Clark A. F., Nystuen A., Nichols B. E., Mackey D. A., Ritch R., Kalenak J. W., Craven E. R., Sheffield V. C.Identification of a gene that causes primary open angle glaucoma. Science2751997668670 |
| 34. | Stern E. J., Webb W. R., Weinacker A., Muller N. L.Idiopathic giant bullous emphysema (vanishing lung syndrome): imaging findings in nine patients. A.J.R.1621994279282 |
| 35. | Davis R. M., Novotny T. E.Changes in risk factors: the epidemiology of cigarette smoking and its impact on chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.1401989S82S84 |
| 36. | O'Donnell D. E., Webb K. A.Breathlessness in patients with severe chronic airflow limitation: physiologic correlations. Chest1021992824831 |
| 37. | Wegner R. E., Jorres R. A., Kirsten D. K., Magnussen H.Factor analysis of exercise capacity, dyspnoea ratings and lung function in patients with severe COPD. Eur. Respir. J.71994725729 |
| 38. | Postma D. S., Burema J., Gimeno F., May J. F., Smit J. M., Steenhuis E. J., Weele L. T. V. D., Sluiter H. J.Prognosis in severe chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.1191979357367 |
| 39. | Damsgaard T., Kok-Jensen A.Prognosis in severe chronic obstructive pulmonary disease. Acta Med. Scand.1961974103108 |
| 40. | Schildkraut J. M., Risch N., Thompson W. D.Evaluating genetic association among ovarian, breast, and endometrial cancer: evidence for a breast/ovarian cancer relationship. Am. J. Hum. Genet.451989521529 |
| 41. | Beaty T. H., Liang K. Y., Seerey S., Cohen B. H.Robust inference for variance components models in families ascertained through probands: II. Analysis of spirometric measures. Genet. Epidemiol.41987211221 |
| 42. | Higgins M., Keller J.Familial occurrence of chronic respiratory disease and familial resemblance in ventilatory capacity. J. Chronic Dis.281975239251 |
| 43. | Schmidt R. A., Glenny R. W., Godwin J. D., Hampson N. B., Cantino M. E., Reichenbach D. D.Panlobular emphysema in young intravenous ritalin abusers. Am. Rev. Respir. Dis.1431991649656 |
