| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Health Sciences Centre and University of Manitoba, Winnipeg, Manitoba, Canada
Correspondence and requests for reprints should be addressed to N. R. Anthonisen, M.D., Respiratory Hospital, 810 Sherbrook Street, Winnipeg, MB, R3A 1R8 Canada. E-mail: anthonis{at}ms.umanitoba.ca
The Lung Health Study (LHS) was a multicenter clinical trial of smoking intervention and inhaled bronchodilator in middle-aged smokers with mild to moderate chronic obstructive pulmonary disease (1) that accumulated a large cohort of high-risk individuals and has monitored them for nearly 15 years. This article reviews results of the initial study and subsequent long-term follow-up, and some of the risk factors for rapid decline of lung function in the cohort.
At 10 clinical centers in 1986–1988, the LHS recruited 5,887 smokers aged 35–59 years. Entry criteria included evidence of airway obstruction, that is, FEV1/FVC less than 0.70 and FEV1 55–90% of the predicted normal value, and a willingness to enter a smoking cessation program. People with other diseases and those being treated for lung disease were excluded. They were randomized into three groups: usual care (UC) participants were advised to stop smoking and followed; special intervention (SI) participants were enrolled in an intensive smoking cessation program (2), and half (SIA) were prescribed an inhaled bronchodilator (ipratropium bromide) and the other half (SIP) given placebo in a double-blind fashion. All participants were followed with annual spirometry for 5 years and SI participants were seen at least every 4 months to maintain compliance with the interventions. The main outcome measure, spirometry, was assessed with great care in a standardized fashion (3). Methacholine reactivity was measured at baseline and again at the end of 5 years.
On entry, participants averaged 48 years of age, and 63% were male. They were heavy smokers, averaging more than 31 cigarettes per day, with more than 40 pack-years. Their FEV1 averaged 75% of the predicted normal (2.64 L) and increased little (0.110 L) with bronchodilator treatment. They demonstrated a surprising degree of methacholine reactivity, in that about one-third had a 20% decline in FEV1 with a methacholine dose of 5 mg/ml.
Approximately 11 years after entry efforts were made to reexamine all of the original participants who were not known to be dead (4), and spirometry was repeated when possible.
 |
RESULTS OF THE INITIAL STUDY |
|---|
 TOP RESULTS OF THE INITIAL... RESULTS OF THE LATE...RISK FACTOR ANALYSESREFERENCES |
|---|
The main outcome variable was postbronchodilator FEV1. Among UC participants, this declined more or less linearly over the 5 years of follow-up, from 2.75 to 2.5 L. In the SIP group, there was little change in FEV1 during the first year of the study, and then a decline similar to that observed in the UC group occurred over the subsequent 4 years. Overall loss of FEV1 was significantly less in the SIP group than in the UC group (1). The SIA group, which had more women than the SIP group, started at a slightly lower baseline FEV1 than the other groups. In the SIA group, average FEV1 increased during the first year and then fell at a rate similar to that in the other two groups. The 5-year change in FEV1 in the SIA group was significantly less than in either of the other two groups. At the end of 5 years, the average FEV1 of the SIA and SIP groups was 50 ml greater than that of the UC group.
By intent-to-treat analysis, smoking cessation significantly reduced decline in lung function, as best exemplified by the difference between the SIP and UC groups. The actual effect of smoking cessation was much greater than the difference between treatment groups because most participants in each treatment group were still smoking. When sustained quitters, who stopped smoking in the first year and maintained abstinence throughout, were compared with participants who continued to smoke, much larger differences were observed, with continuing smokers losing 300 ml over 5 years, whereas sustained quitters lost less than 50 ml.
Again by intent-to-treat analysis, the SIA group lost slightly but significantly less lung function than the SIP group, due almost entirely to the increase in FEV1 observed in the former during the first year of the study. The clinical significance of this finding was rendered moot when participants were studied after the initial fifth year visit when inhalers had been discontinued. After discontinuation of ipratropium, there was a small decrease in FEV1 in the SIA group, which did not occur in the SIP group, and overall differences in FEV1 measured from baseline to this second fifth annual visit did not differ between the two groups. Thus, although bronchodilator appeared to increase FEV1 so long as participants were taking it, this advantage was lost when the drug was discontinued, and the drug was not shown to influence the course of the disease.
|
RESULTS OF THE LATE (11-YEAR) FOLLOW-UP |
|---|
|
TOPRESULTS OF THE INITIAL...RESULTS OF THE LATE...RISK FACTOR ANALYSESREFERENCES |
|---|
During the 6 years between the end of the initial study and the late follow-up, cross-sectional quit rates increased more in the UC group (from 22 to 49%) than in the SI group (from 39 to 52%), because quit rates were higher in the UC group. However, most of the change occurred in people who were intermittent quitters, that is, those who stopped and started smoking during the initial study. In both treatment groups about 95% of sustained quitters at 5 years had continued to abstain at 11 years, and the SI group contained a substantially larger fraction of sustained quitters (23%) than did the UC group (6%), and a smaller fraction (22 versus 31%) of participants who had smoked continuously.
In intent-to-treat analysis, the 11-year FEV1 data were essentially linear extrapolations of the original 5-year data; mean FEV1 in the SI group was 85 ml more than that in the UC group, and the difference in overall decline was still statistically significant (4). Again, differences were much larger when the data were analyzed in terms of smoking habit. The mean FEV1 in sustained quitters declined less than 27 ml/year, from 79 to 76% of the predicted normal value, whereas that of continuing smokers declined at 60 ml/year, from 79 to 64% of the predicted value. At 11 years, 18% of LHS participants who had smoked continuously had FEV1 values less than 50% of the predicted normal value as opposed to 3.3% of sustained quitters, whereas 38% of continuous smokers had FEV1 values less than 60% of the predicted normal value as compared with 10% of sustained quitters (4).
|
RISK FACTOR ANALYSES |
|---|
|
TOPRESULTS OF THE INITIAL...RESULTS OF THE LATE...RISK FACTOR ANALYSESREFERENCES |
|---|
The original LHS was designed in part to test the "Dutch hypothesis" that decline in lung function was related to preexistent airway reactivity (9). Methacholine reactivity was tested at study entry, and proved to be a powerful predictor of rate of loss of FEV1, second only to smoking habit in terms of importance (10). Reactivity was interactive with smoking status in that it had relatively little effect in people who stopped smoking but was important in those who did not. It should be noted that this was particularly true of female participants, in whom reactivity was greater than in men, for reasons that are not entirely clear but were related to the fact that women had smaller average airway size than did men on study entry. Interestingly, baseline FEV1 did not influence subsequent decline once smoking status, age, sex, and reactivity had been considered (10).
The original LHS also afforded an opportunity to test the "British hypothesis" that decline in lung function was related to repeated respiratory infections or bacterial colonization of the respiratory tract (11). At annual visits, participants were questioned regarding physician visits for bronchitis, pneumonia, influenza, or chest colds during the preceding year, and positive responses, termed respiratory infections, were related to lung function data. Respiratory infections were not common, averaging 0.2 episode per participant per year. They were more common in participants with chronic cough and sputum than in those without, and increased with time in participants who continued to smoke, but did not in those who quit. In continuing smokers, one respiratory infection per year was associated with an excess loss of FEV1 of 7 ml, which was statistically significant, whereas there was no significant loss of lung function in participants who quit (12). Although the effect of infections seemed small, it would not have been trivial in people who had as many as two such episodes per year over many years. This finding, although in agreement with some other data (13), contradicted the findings of Fletcher and coworkers (14), who studied a smaller group and used less careful spirometry.
Finally, the LHS studied a large cohort of female smokers, and afforded the opportunity to examine sex as a risk factor for the development of chronic obstructive pulmonary disease. Among continuing smokers, rate of loss of FEV1, when expressed as a percentage of the predicted normal value, did not differ significantly between men and women (4). Women smoked fewer cigarettes per day than did men, so an argument could be made that they were more "sensitive" than men, but this is probably less important than the similar rate of loss.
CONCLUSIONS
The original LHS showed that smoking cessation influenced rate of decline in lung function in intent-to-treat analysis of randomly allocated groups; the difference between treatment groups cannot be ascribed to anything other than smoking cessation. The effect of smoking cessation was striking; the development of clinically significant airway obstruction was largely prevented in those who quit as compared with those who did not. The importance of smoking cessation as the primary intervention in early chronic obstructive pulmonary disease was validated. Smoking cessation had a beneficial effect in both sexes, at all ages, and at all levels of baseline lung function. It is probably never "too late" to institute this intervention.
On the other hand, regular prescription of a bronchodilator did not influence the course of the disease, that is, the rate of decline of FEV1 when participants were not actively taking the medication. This finding supports the current use of bronchodilators for the purpose of symptomatic relief, and indicates that the only reliable way to assess their influence on long-term changes of lung function is to examine the latter after drug administration has been discontinued.
Other than smoking, several risk factors for rapid loss of lung function were identified. The most prominent of these was methacholine reactivity, which was strongly interactive with smoking. Although it is reasonable to argue that people who develop bronchospasm easily with the inhalation of foreign material might have more smoking-related problems than those who do not, it is not clear what methacholine reactivity represents in smokers with mild airway obstruction. We do not believe that it necessarily represents the same thing as reactivity does in asthma.
Respiratory infections also probably play a role in decline of lung function in smokers. This effect may be substantial, especially if the frequency of infections increases with disease severity. Again, this effect was evident in smokers but not in participants who quit.
Finally, sex did not seem to be an important risk factor, and a number of genetic influences are under active investigation, with some results of promise.
|
FOOTNOTES |
|---|
(Received in original form June 20, 2003; accepted in final form August 29, 2003)
|
REFERENCES |
|---|
|
TOPRESULTS OF THE INITIAL...RESULTS OF THE LATE...RISK FACTOR ANALYSESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
