INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

HIV infection is associated with a decline in CD4+ T-cell numbers, a reduction in the T-cell mediated immune responses to antigens, and a disruption of intercellular signaling within the immune system (1). Asthma, on the other hand, is associated with excessive immune reactivity, mediated by T-cells, that leads to airway inflammation and narrowing (2). Thus, on theoretical grounds, it might be postulated that asthma should be less frequent among HIV-infected persons.

The course of HIV infection is characterized by frequent and varied respiratory illnesses, including acute bronchitis, bacterial pneumonia, and opportunistic infections such as Pneumocystis carinii pneumonia (6). However, it is not clear if asthma, or its physiologic hallmark bronchial hyperresponsiveness (BHR), is more or less common among persons with HIV infection. Studies examining the prevalence of BHR in patients with HIV disease have provided contradictory results and are hampered by small sample size (7). On the other hand, other features of the asthmatic diathesis such an elevation of serum eosinophilic cationic protein or an elevation in serum IgE have been reported in HIV infected patients (11, 12).

We compared the prevalence of asthma and related conditions among men with HIV infection who were free of active pulmonary infection with the prevalence among male subjects of similar age from the general population. We then examined the determinants of BHR among seropositive subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The institutional ethics committee approved the study and the consent forms.

Study Sample and Design

Between September 1995 and June 1996, we conducted a cross-sectional study of the prevalence of asthma and asthmalike conditions among HIV-seropositive subjects in Montreal. The target population consisted of HIV-infected persons between 20 and 44 yr of age attending the immunodeficiency clinics at the Montreal Chest Institute, the Montreal General Hospital, and two private clinics in Montreal that specialized in immunodeficiency. Subjects at the four study clinics were first asked to participate by their physician who, after obtaining informed consent, used a screening questionnaire to establish eligibility. For comparison of prevalence rates of asthma-related characteristics, we used information obtained from another study that had used the same methods of measurement between July 1993 and June 1994, among a randomly selected population of adults 20 to 44 yr of age living in the Montreal area and selected by random digit dialing. HIV seropositive patients were eligible for the study if their chest radiographs were normal, they were free of any active cytomegalovirus infection, pulmonary neoplasms, respiratory infection for at least 6 wk, and Pneumocystis carinii pneumonia for at least 3 mon. Treating physicians provided the pertinent medical information to judge eligibility by questionnaire, and additional information was abstracted from medical records. The European Community Respiratory Health Survey (ECRHS) questionnaire was used to obtain further information on asthma and asthmalike conditions and smoking habits (13). Lung function testing was carried out according to the protocol developed for the European Community Respiratory Health Survey (14). Spirometry testing was carried out by trained technicians. The spirometers, calibration, and testing procedures met the criteria recommended by the ATS (15). FEV₁/FVC was used as a measure of airway caliber. Methacholine was administered using a Mefar dosimeter (Mefar, Bovezzo, Italy) to a maximum cumulative dose of 2 mg (16). The testing was stopped if the FEV₁ dropped by 20% or more of the postsaline value. The cumulative dose of methacholine required to produce a 20% fall in FEV₁ from the postdiluent level was then calculated. Allergy testing with the Phazet multiple puncture method (Pharmacia Diagnostics AB, Uppsala, Sweden) (17) was performed with 11 allergens: mite (Dermatophagoides pteronyssinus), cat, mixed molds (Alternaria alternata and Cladosporidium herbarium), timothy grass, olive, birch, mixed weeds, ragweed, parietaria, and aspergillus. In addition, a negative (diluent for allergens) and a positive (histamine) control were used.

Venous blood was collected from each subject to measure total serum IgE level and CD4 and CD8 cell counts. Total serum IgE was measured by the Pharmacia CAP system (Pharmacia Diagnostics).

Analysis

BHR was analyzed as a categorical variable defining the hyperresponsive men as those with 20% or greater fall in FEV₁ at a cumulative dose of methacholine of 2 mg or less. Atopy was defined as a skin weal of greater than 3 mm in diameter to one of 11 allergens tested in the presence of a negative control reaction. Subjects were considered to be nonsmokers if they had consumed less than one cigarette, or the equivalent in tobacco per day for 1 yr. Current smokers were those who had smoked during the last month before the survey. CD4 counts and total IgE were analyzed as categorical variables. The cut off points were a CD4 cell count less than 200 cells/mm³ and IgE greater than 100 IU/ml. Prevalence rates for asthma, asthmalike conditions, atopy, and BHR, as well as differences between subjects with and without BHR, were calculated using standard statistical methods. The independent effects of the potential determinants of BHR among seropositive subjects was examined using unmatched logistic regression. Data management and analysis were carried out using SAS version 6.1 (SAS Institute, Cary, NC).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study population consisted of 248 male HIV-seropositive subjects from four major Montreal immunodeficiency clinics. The comparison group consisted of 236 healthy male subjects identified by random digit dialing and who agreed to come for testing. Only men were included in the present analysis because of the small number of seropositive women. From the general population sample responding to a mailed questionnaire, 1,386 were invited to undergo testing in a central laboratory and 503 (237 men) accepted the invitation. The comparison group used here consists of the 236 men who were able to complete spirometry.

Seropositive men were older (36.2 yr, SD 4.9) when compared with a mean age of 31.7 yr (SD 6.9) in the general population group. As shown in Table 1, HIV-infected persons were more likely to be current cigarette smokers (61.7 versus 35.2%), report wheezing in the prior 12 mo (54.4 versus 21.2%), demonstrate bronchial hyperresponsiveness to methacholine (26.2 versus 14.4%), and have an elevated IgE (37.8 versus 25.7%). These differences were all statistically significant since the 95% confidence intervals did not overlap. A history of asthma also tended to be more frequent among seropositive subjects. By contrast, the prevalence of allergy skin test positivity to any one of the inhaled allergens tested was lower among HIV seropositive subjects than in the general population, 53.2% as compared with 64% . 

To examine whether differences in markers of asthma between HIV+ men and men in the general population were explained by differences in smoking, we compared the prevalence of wheeze and of BHR according to whether or not subjects were currently smoking (Table 2). The differences in the prevalence of current wheezing between HIV and HIV+ men remained substantial and statistically significant for both smokers and nonsmokers (p < 0.001). For BHR the differences between men infected or not with HIV was significant only among smokers (p < 0.05). The findings were also similar if categorizing subjects according to whether they had ever smoked or whether they smoked 25 or more cigarettes per day (data not shown).

A comparison between seropositive subjects with and without BHR is shown in Table 3. HIV-positive men with BHR had reduced airway caliber, as reflected by a lower mean FEV₁/FVC, were more likely to have a positive allergy skin test (p = 0.06) and an elevated IgE than those without BHR. Those with BHR were more often active smokers (p = 0.08), whereas the frequency of having smoked at any time was similar. There were no differences in the prevalence of a CD4 count less than 200 cells/mm³. When looking at the independent effect of these potential determinants of BHR among seropositive subjects using multivariate logistic regression, only a low FEV₁/FVC (p = 0.0001) and an elevated IgE (p = 0.0035) remained as significant independent predictors of BHR.

There was a weak (r = 0.25), although highly statistically significant (p = 0.0008), negative correlation between the absolute CD4 count and the serum IgE on a log scale.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have found that HIV infected men followed by specialized clinics in Montreal had a higher prevalence of wheezing in the previous 12 mo, BHR to methacholine, and an elevated serum IgE than did men of similar age selected randomly from the general population of the same city. Among seropositive subjects the principal determinants of BHR were a reduction in airway caliber (FEV₁/FVC) and an elevated IgE.

Previous studies of the association between BHR and HIV infection have provided apparently contradictory results. O'Donnell and colleagues (8) found evidence of BHR in 31% of seropositive subjects. Wallace and colleagues (10) found no statistically significant excess of BHR among 66 seropositive subjects when compared with a volunteer control group (19.3 and 12.9%, respectively). It should be noted, however, that these different estimates for the prevalence of BHR among seropositive subjects fall within the 95% confidence limits of our estimate of 26.2%. The negative findings of Moscato and colleagues (9) were based on testing of only 25 seropositive intravenous drug users. The lower prevalence of positive skin tests to common inhaled allergens, a common determinant of an increased BHR (18), among our seropositive subjects as opposed to the population control subjects, suggests that the prevalence of BHR in seropositive men may actually be an underestimate of the true difference between normal and HIV- infected men if normal men with positive allergy tests were more likely to have taken part in the study. Alternatively, skin- test sensitivity may have been less prevalent among seropositive men because of the reduced expression of aero-allergen-specific IgE responses seen with progression of HIV disease (19). Differences in the prevalence of wheeze and of BHR were more marked among current smokers.

What might such an excess of nonspecific BHR among subjects with HIV infection mean? The associated increase in the report of wheeze and the greater frequency of an elevated serum IgE among the seropositive subjects suggest an excess of asthma. Of note, a history of prior asthma was reported by 17.3% as compared with 12.3% of the general population sample. Prior reports of respiratory conditions among HIV-infected persons have not found asthma to be common; rather, upper respiratory infections and acute bronchitis were the most common syndromes encountered (6, 20). It seems plausible that a portion of the episodes classified as upper respiratory infections and acute bronchitis might rather have been exacerbations of unrecognized asthma. This appears especially likely for acute bronchitis, which was diagnosed based on the occurrence of an acute respiratory illness with productive cough but without radiographic evidence of pneumonia (20). Such a description could certainly include exacerbations of asthma. Differentiation of exacerbations of asthma from acute bronchitis would be important since the treatment of these two conditions differs significantly.

In a multivariate analysis, the only independent predictors of BHR in the seropositive subjects studied were FEV₁/FVC and serum IgE. These are also important predictors of BHR in non-HIV-infected persons (18, 21). Although both past and current smoking were more common among seropositive subjects, smoking was not an independent predictor of BHR. The excess of BHR is unlikely to be related to recent respiratory infection since subjects with ongoing or recent acute respiratory syndromes or with abnormal chest radiographs were specifically excluded. The presence of BHR was also not confined to subjects with advanced HIV disease since the prevalence of a CD4 count less than 200 cells/mm³ did not differ between those with and those without BHR.

The dysregulation in the immune system associated with HIV infection is characterized by polyclonal gammopathy, a decreased Th-1, and an increased Th-2 T helper cell activity profile (22, 23). Serum levels of IgE as well as eosinophils are frequently elevated among persons with HIV infection (12, 24) as a result of the imbalance in the Th-1 and Th-2 helper-cell activity. In our study, there was a weak association between lower CD4 counts and higher total serum IgE. It seems conceivable that this increase in total IgE is in the causal pathway for the increase in bronchial responsiveness since IgE is a major determinant of BHR in the well population (21, 25). It is postulated that IgE will bind to tissue mast cells and may nonspecifically trigger the bronchial inflammatory responses characteristic of asthma (26). Alternatively, BHR may follow the persistent fall in FEV₁/FVC, which has been described even long after Pneumocystis carinii pneumonia (27), whereas the elevation in IgE may reflect progressive HIV disease (12).

The increased occurrence of BHR may have important implications for the evolution of respiratory disease among seropositive persons. Recently, Diaz and colleagues (28, 29) have provided strong evidence of premature emphysema among seropositive patients who smoke. BHR may be an early sign of airway damage leading to emphysema as has been shown for those without HIV disease (30). In the present study, the excess in BHR was limited to smokers with HIV disease, which may suggest a particular susceptibility to the adverse effects of cigarette smoking among HIV-infected men.