Atopy as a Risk Factor for Chronic Obstructive Pulmonary Disease
Epidemiological Evidence

SCOTT T. WEISS

Channing Laboratory, Brigham and Women's Hospital, Boston, Massachusetts

	INTRODUCTION

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Susceptibility to cigarette smoke is not universal. Only 10- 15% of smokers go on to develop chronic obstructive pulmonary disease (COPD). The reasons for this are likely complex and relate primarily to three factors: genetic susceptibility, dose and timing of exposure, and interactions between smoking and other environmental factors of interest in causing the disease. The goal of this article is to consider the epidemiologic data that atopy defined as immediate type II hypersensitivity, and not as clinical asthma, is a risk factor for the development of chronic obstructive lung disease.

	DEFINITIONS

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

For the purposes of this review atopy is defined as skin test positivity to aeroallergens or elevations in total or specific IgE level. The methodology for assessing these intermediate phenotypes in epidemiologic studies has been standardized and is reasonably accepted. The outcome of interest is the development of COPD. COPD is defined as the presence of fixed IgE, and nonreversable airflow obstruction. Definitions differ as to the degree of obstruction to qualify as COPD, but most investigators would agree that levels below 70% of predicted would qualify if not reversed completely with a bronchodilator. Much of this review concentrates on surrogate markers of COPD, namely, accelerated decline in lung function, because COPD is the net result of that process.

	NATURAL HISTORY

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Contrary to popular opinion, COPD starts in early childhood. It is the in utero and lifelong progression of exposures interacting with genetic susceptibility that determines the ultimate clinical phenotype that will manifest itself most commonly after the age of 65 yr. Thus the natural evolution of the various allergy phenotypes and their relationship to growth and decline of lung function are all of relevance in determining the relationship of allergy markers to the development of COPD. Figure 1 provides a hypothetical picture of how allergy phenotypes could potentially influence growth and decline in lung function. Quite simply, there are three direct choices that could decrease maximal lung growth, promote early decline, or precipitate accelerated decline in lung function. On a more complex level, they could influence or interact with other intermediate phenotypes, such as airway responsiveness, and possibly with other exposures, such as smoking, to accelerate specific types of inflammation leading to clinical disease. Unfortunately, because of the technical problems with performing long-term observational studies, our ability to prove these relationships at every level of the life cycle is limited. This review focuses on the influence of skin test reactivity and total and specific IgE level on decline in lung function in adult epidemiologic populations.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Hypothetical graph of allergy phenotypes and their potential influence on lung function growth and decline.

	SKIN TEST REACTIVITY AND DECLINE IN LUNG FUNCTION

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Burrows and coworkers have examined the cross-sectional relationship between lung function level and skin test reactivity among the participants of the Tucson Epidemiologic Study. Using a three-point scale to quantify skin test reactivity to five common aeroallergens, they found that their skin test index was associated with low FEV₁/FVC ratios in current and former smokers, but not nonsmokers, among subjects between the ages of 15 and 54 yr (1). In a subsequent analysis, it appeared that the relationship of skin test reactivity to lung function was seen only in the presence of eosinophilia and was absent in individuals over the age of 55 yr (2). This raised the question as to whether these results were simply being driven by subjects with asthma, and finally a third analysis was performed in which removal of the subjects with asthma resulted in loss of statistical significance for the skin test FEV association (3). Burrows concluded that skin test reactivity was an important factor in asthmatic individuals who smoked and who had "asthmatic bronchitis."

Longitudinal studies have been equivocal as to the effect of skin test reactivity on decline in lung function and subsequent COPD risk. Taylor and coworkers monitored 227 subjects without a history of asthma for a 7.5-yr period and found no relation of skin test reactivity to decline in FEV₁ (4).

Frew and colleagues studied three occupational cohorts (730 men) and looked at the relationship between decline in FEV₁ and a wheal size > 3 mm generated in response to a series of common aeroallergens (5). After adjustment for age, methacholine airway responsiveness, and level of FEV₁, atopy was a significant predictor of FEV₁ decline only among smokers.

In another occupational cohort of 308 men, Annesi found no relationship between the 5-yr decline in FEV₁ and a wheal size of  5 mm after appropriate adjustment for confounding variables (6).

Gottlieb has investigated this question in 1,025 men participating in the Normative Aging Study. Subjects had a mean age of 61 ± 8 yr and denied any prior history of asthma (7). Subjects performed spirometry and underwent allergy prick testing at a baseline visit, and then performed repeat spirometry after a median of 3.1 yr. Skin prick tests were performed with glycerin-preserved allergens (house dust, mixed grasses, mixed trees, and ragweed). Multiple linear regression was used to examine the annual rate of change of lung function in relation to skin test reactivity, defined as a mean wheal size, in response to the four aeroallergens, of  2 mm after adjustment for age, height, smoking status, and initial lung function. Skin test reactivity was a significant predictor of the annual rates of decline in FEV₁ and in FEV₁/FVC ratio. The regression model predicted an excess decline in FEV₁ of 9.45 ml/yr for subjects with a mean wheal diameter of  2 mm (p = 0.05); the excess decline in FEV₁/FVC ratio was 0.29%/yr (p = 0.001). The magnitude of the effect of a mean wheal diameter  2 mm on rates of decline in FEV₁ and FEV/FVC ratio were, respectively, 34 and 49% of the effect of current cigarette smoking.

	ALLERGEN LEVELS AND DECLINE IN FEV1

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Weiss and coworkers investigated the relationship of home allergen levels to decline in FEV₁ in 10 asthmatic and 30 randomly elected age-matched participants of the Normative Aging Study (8). They investigated the relationship of annual FEV₁ decline to concentrations of Bla G1, Bla G2, Der p 1 and Der p 2, and Fel d 1 concentrations. No significant relationships were seen with house dust mite or with cat, but for both cockroach antigens significant declines in FEV₁ were associated with higher levels in house dust: Bla G1, 79.8 ml/yr, p = 0.0006, and Bla G2, 40.81/yr, p = 0.0004 were significant predictors of decline in FEV₁ after adjustment for age, smoking, and baseline FEV₁. These results were unchanged after eliminating the subjects with asthma from the analysis.

	RELATIONSHIP OF TOTAL AND SPECIFIC IgE TO LEVEL AND SPECIFIC DECLINE IN FEV1

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

A variety of studies detail a relationship of total serum IgE level to a decreased level of FEV₁ (9). To date, there is no study that relates total serum IgE level to accelerated decline in FEV₁ independent of cigarette smoking. The reason for these discrepancies is unclear and will await clarification in further studies.

	CONCLUSION

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Atopy appears to be an important factor in at least three ways. First, it is the primary influence on persistence of asthma in childhood and, hence, may contribute to decrease in maximal attained level of lung function. Second, it appears that, independent of airway responsiveness and level of FEV₁, it is related to accelerated decline in FEV₁ in later adult life, independent of cigarette smoke. Finally, and this is not proven, it may have relationships that enhance inflammation by interacting with cigarette smoking and airway responsiveness to produce disease.

	DISCUSSION

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

Quanjer: The plateau phase in lung function extends to about 35-40 yr. Smoking makes the decline start earlier and the decline accelerates with age. Have you taken that into account?

Weiss: In our population the mean is 60, so the plateau phase has long passed. There are three possibilities for the lower lung function: declining earlier, declining too fast, or not growing. It is clear that it occurs for smoking, it is not clear that it occurs for allergen exposure, which was my topic.

Aalberse: I was surprised by the difference between mites and cockroaches. Did you correct for race and social class?

Weiss: Yes, we did. All our subjects were Caucasian and when we adjust for income and education, our results are unchanged.

Kauffman: Do you have information on the protease content of airborne cockroach material? It has been shown that proteases can have a damaging effect on epithelium and in this way, may influence lung function in the long term.

Weiss: Yes, cockroach does contain proteases, but we do not know how relevant this is for the damaging effect with respect to lung function. In fact, in Dr. Diane Gold's study that Dr. Platts-Mills alluded to, the one exposure that seems to be strongly associated with wheeze during the first year of life is cockroach exposure. It is hard to believe that this is related to sensitization at that point in time and the proteolytic theory could be important.

Holt: I know this was not a smoking presentation, but I do want to go back to the effect of smoking on total IgE. What do you know about the dose-response relation? I was thinking of the Burrows data, where in fact one cigarette per day will quite markedly elevate total IgE, but the curve is biphasic.

Weiss: At the time that was done, we did not have IgE levels on those kids. One of the confounding factors for the effect of IgE on decline of lung function is smoking. There is a subgroup of smokers where smoking perhaps acts like a superantigen and these people get elevated IgE levels, but this is not in all smokers; there may be some genetic susceptibility. The point I was trying to make was, susceptibility to environmental exposure may be different at different ages.

Platts-Mills: The important issue is whether allergens cause effects on the lungs in the absence of IgE antibodies and your data raises that possibility in older people. Nobody has produced good data to show that high allergen exposure affects nonallergic children. De Monchy has recently published data on allergic children being followed in houses where there are seasonal variations in mites and has shown very nicely that seasonal variations in mites affect bronchial hyperreactivity in the mite-allergic children, but have no effect on non-mite-allergic children. This is strongly reassuring that mite exposure does not have a direct effect on the lung itself. But, the issues is important.

If I may make a point about the enzymatic activity of allergens. Many of the allergens are homologous to enzymes, many of them are not potent enzymes, but some are. Whether that really plays a role in their immunogenicity is a completely open question. The other point is that there are very high levels of normal proteases in the lungs and many discussions ignore that fact.

Kips: Could the endotoxin content of inhaled dust have influenced the results?

Weiss: We did not measure endotoxin, but we are now in Dr. Gold's study.

Martinez: People who are exposed to endotoxins at work have a decline in lung function, so there could be confounders here that have nothing to do with allergy.

Platts-Mills: There are reports of a correlation between endotoxin and asthma symptoms. However, this has only been studied in house dust mite-sensitive patients. There are no convincing reports showing an effect of endotoxin in nonallergic individuals.

Djukanovic: We looked at atopic individuals who could not recall ever wheezing, in other words, they were not seen as clinically having asthma. When we did biopsies in these individuals and looked at collagen thickness, we observed that the thickness of the reticularis layer was less than in the asthmatics, as one would anticipate. But in the atopic nonasthmatics, there was clear evidence of an increase. Whether it is directly related to atopy I don't know.

Weiss: How old were these people?

Djukanovic: They were medical students, so, young adults. I just wonder whether defining atopy and IgE responses by presence or absence of a wheal response is perhaps too crude, too insensitive to identify individuals who would be within this group that might suffer the consequences of long-term exposure.

Weiss: We did not see the effect on decline in lung function until we looked at mean wheal size. I think that in our hands the mean wheal size seems to be an effective marker.

Platts-Mills: Is there any change in the spirometry in your atopic nonasthma group with collagen thickening?

Djukanovic: No.

Platts-Mills: I though collagen had something to do with remodeling.

Djukanovic: This is no proof that this is going to affect the decline in lung function. But what it does indicate is that there is some remodeling, which one might accept as one of the indicators of subsequent progression.

Weiss: During a recent NIH COPD workshop, it was said that 75% of COPD was related to asthma or one of the intermediate phenotypes and that only a small percentage of COPD was independent of these. One of the other issues here is the phenomenon of recall bias. There are an awful lot of people who have symptoms early in life and then, as the lung grows, become asymptomatic. They retain an intermediate phenotype: There might be bronchial responsiveness, there might be atopy, there may be both, but maybe no clinical symptoms and then they experience another exposure in early adult life. They get an evaluation. Did you ever have these symptoms before? And they say, no, absolutely not. The provocative statement would be, no incident case of atopic asthma in young adults is really incident, they are all recrudescent cases from people who had symptoms in early childhood. You cannot tell without longitudinal data.

	Footnotes

Correspondence and requests for reprints should be addressed to S. T. Weiss, M.D., Brigham and Women's Hospital Channing Laboratory, 180 Longwood Avenue, Boston, MA 02115-5899. E-mail: scott.weiss{at}channing.harvard.edu

	References

TOP INTRODUCTION DEFINITIONS NATURAL HISTORY SKIN TEST REACTIVITY AND... ALLERGEN LEVELS AND DECLINE... RELATIONSHIP OF TOTAL AND... CONCLUSION DISCUSSION REFERENCES

1. Burrows, B., M. D. Lebowitz, and R. A. Barbee. 1976. Respiratory disorders and allergy skin-test reactions. Ann. Intern. Med. 84: 134-139 .

2. Burrows, B., R. J. Knudson, G. Cline, and M. D. Lebowitz. 1988. A reexamination of risk factors for ventilatory impairment. Am. Rev. Respir. Dis. 138: 829-836 [Medline].

3. Sherrill, D. L., M. D. Lebowitz, M. Halonen, R. A. Barbee, and B. Burrows. 1995. Longitudinal evaluation of the association between pulmonary function and total serum IgE. Am. J. Respir. Crit. Care Med. 152: 98-102 [Abstract].

4. Taylor, R. G., H. Joyce, E. Gross, F. Holland, and N. B. Pride. 1985. Bronchial reactivity to inhaled histamine and annual rate of decline in FEV₁ in male smokers and ex-smokers. Thorax 40: 9-16 [Abstract/Free Full Text].

5. Frew, A. J., S. M. Kennedy, and M. Chan-Yeung. 1992. Methacholine responsiveness, smoking, and atopy as risk factors for accelerated FEV₁ decline in male working populations. Am. Rev. Respir. Dis. 146: 878-883 [Medline].

6. Annesi, I., F. Neukirch, E. O. Frija, M. P. Oryszyn, M. Korobaeff, M. F. Dore, and F. Kauffmann. 1987. The relevance of hyperresponsiveness but not of atopy to FEV₁ decline: preliminary results in a working population. Bull. Eur. Physiopathol. Respir. 23: 397-400 [Medline].

7. Gottlieb, D. J., D. Sparrow, G. T. O'Connor, and S. T. Weiss. 1996. Skin test reactivity to common aeroallergens and decline of lung function: the Normative Aging Study. Am. J. Respir. Crit. Care Med. 153: 561-566 [Abstract].

8. Weiss, S. T., G. T. O'Connor, D. DeMolles, T. Platts-Mills, and D. Sparrow. 1998. Indoor allergens and longitudinal FEV₁ decline in older adults: the Normative Aging Study. J. Allergy Clin. Immunol. 101: 720-725 [Medline].

9. Orie, N. G. M., H. J. Sluiter, K. deVries, G. J. Tammeling, and J. Witkop. 1961. The hose factor in bronchitis. In N. G. M. Orie and H. J. Sluiter, editors. Bronchitis. Royal van Gorcum, Assen, The Netherlands. 43-59.

10. Rijcken, B., X. Xu, J. P. Schouten, B. Rosner, and S. T. Weiss. 1995. Airway hyperresponsiveness to histamine associated with accelerated decline in FEV₁. Am. J. Respir. Crit. Care Med. 151: 1377-1382 [Abstract].

11. O'Connor, G. T., D. Sparrow, and S. T. Weiss. 1995. A prospective longitudinal study of methacholine airway responsiveness as a predictor of pulmonary function decline: the Normative Aging Study. Am. J. Respir. Crit. Care Med. 152: 87-92 [Abstract].

12. Villar, M. T. A., L. Dow, D. Coggon, F. C. Lampe, and S. T. Holgate. 1995. The influence of increased bronchial responsiveness, atopy, and serum IgE on decline in FEV₁. Am. J. Respir. Crit. Care Med. 151: 656-662 [Abstract].

13. Dow, L., D. Coggon, M. J. Campbell, C. Osmond, and S. T. Holgate. 1992. The interaction between immunoglobulin E and smoking in airflow obstruction in the elderly. Am. Rev. Respir. Dis. 146: 402-407 [Medline].

14. Vollmer, W. M., A. S. Buist, L. R. Johnson, L. E. McCamant, and M. Halonen. 1986. Relationship between serum IgE and cross-sectional and longitudinal FEV₁ in two cohort studies. Chest 90: 416-423 [Abstract/Free Full Text].

15. Omenaas, E., P. Bakke, G. E. Eide, S. Elsayed, and A. Gulsvik. 1995. Serum house-dust-mite antibodies and reduced FEV₁ in adults of a Norwegian community. Am. J Respir. Crit. Care Med. 152: 1158-1163 [Abstract].