- Email: [email protected]
Risk Factors for Ventilatory Impairment Among Middle-Aged and Elderly Men
Risk Factors for Ventilatory Impairment Among Middle-Aged and Elderly Men* The Normative Aging Study George T. O'Connor, M.D., F.C.C.P.; David Spaf"T''OD. D.Sc.; Mark Segal, Ph.D.; and Scott T. ~s, M.D., F.C.C.P.
To evaluate the relationship of atopy and inftammation to the occurrence of ventilatory impairment, we studied 1,301 middle-aged and older men participating in the Normative Aging Study at the time of their 1984 to 1987 examination. Screening at entry to the study in the 1960s had excluded subjects with asthma and other chronic diseases at that time. After adjustment fur smoldng status and pack-years in this cross-sectional analysis, a weak inverse relationship between FEV, and blood eosinophil count was not statisticaDy signi&cant, and FEV, sbowed no relationship to blood eosinophil count as a percentage of total leukocytes. A weak inverse relationship between FEV, and serum total IgE concentration was not statistically significant. Cutaneous
C
hronic obstructive pulmonary disease (COPD) develops only in a minority of cigarette smokers. Conversely, studies of population samples have indicated that roughly 15 percent of adults with chronic airflow obstruction have no history of smoking.u Population studies have also indicated that estimated total pack-years of cigarette smoking accounts for only 14 to 15 percent of the variance of age- and heightadjusted forced expiratory volume in 1 s (FEV1) and expiratory ftow rates in the general population. 3 •4 It has been hypothesized that allergy may predispose to the development ofCOPD, but available data have not confirmed this hypothesis. Burrows and coworkers~! have recently reported population-based data indicating that markers of allergy, including blood eosinophilia. are not associated with reduced level of pulmonary function after excluding subjects with a history of asthma. These data suggest that only among
*From the Channing Laboratory, Department of Medicine, Brigham and Women$ Hospital; the Divisions of Pulmonary and Critical Care Medicine, BnJdwn and Women'S and Beth Israel Hospitals; Harvard Medical sChool; Nonnative Aging Study, Department of Veterans A8iUrs Outpatient Clinic, Boston; and the Department of Epidemiology and Biostatistics, University of California, San Francisco. Supported by grants HL-34645 and Hl.r45089 from the Division of Lung Diseases, National Heart, Lung, and Blood Institute, National Institute of Health, and by the Medical Research Service of the Department of Veterans Atrairs. Presented In part at the 1989 Annual Meeting, American Thoracic Society, Cincinnati, May 14, 1989. Dr. O'Connor is recipient of an Edward Livingston Trudeau Scholar Award from the American Lung Association. Manuscript received December 26, 1991; revision accepted June 26. Reprint requests: Dr: \\friss, Channing Laboratory, 180 LDngwood Avenue, BoBton 02115
m
immediate hypersensitivity to one or more COIDIDOD aeroallergens was not related to FEV,. A signilicant inverse relationship between FEV, and blood total leukocyte count was observed in never and furmer smokers. FEV, was significantly lower in subjects reporting usual phlegm production. These &ndings suggest that ventilatory impairment is not related to atopic status among middle-aged and older men without a history of asthma. The inverse relationship between total leukocyte count and FEV1 in this sample supports the hypothesis that nonallergic inAammation plays a role in the pathogenesis of chroaic air8ow obstruction in this group. (C,_, 1993; 103:376-8!)
asthmatics does allergy contribute to the development of chronic airftow obstruction; however, other investigators have observed that blood eosinophil count is inversely related to level ofFEV16 and directly related to rate ofdecline ofFEV17 among nonsmoking nonasthmatics. We have examined the relationship of FEV1 to markers of allergy and other covariates among men participating in the Normative Aging Study. METHODS
Sample The Nonnative Aging Study is a longitudinal study of aging located at the Department of Veterans Affairs Outpatient Clinic In Boston.• The cohort initially consisted of 2,280 men who were 21 to 80 years of age on entry to the study between 1961 and 1969. \blunteers were screened at entry, and those with chrooic medical conditions, including asthma, chronic bronchitis, and chronic sinusitis, were excluded. Subsequent to entry, volunteers have returned every 3 to 5 years for a comprehensive examination. The present cross·sectional analysis involves men who returned for examination from April1984, when assessment of atopic status was begun at the Normative Aging Study, until December 1987. 1\vo men with total blood leulcocyte counts greater than 30,
Table 1-Sample C'haracterimcs* Without Asthma No.t Age, yr FEV., 9& predicted FVC, % predicted FEV.;FVC ratio,% Smoking habits Current smoker, % Exsmoker, 9& Never smoker, %
1,245 61.0 93.7 96.3 78.3
(7.1) (15.9) (14.4) (7.5)
most days, fur at least 3 months per year, for at least 2 years' duration. Usual phlegm is defined as the report of usual phlegm from the chest.
With
Asthma 53 61.3 82.3 90.8 72.5
Spirometry
(9.4) (20.0~1
(17.0)t (10.1~1
20.8 39.6 39.6
15.4 48.5 36.1
Pack-years Current smokers 48.4 32.1 Exsmokers 9.3 Chronic cough, % 25.7 Usual phlegm, 9& <:!:1 positive allergy 21.7 skin test,% Log,. [lgE] 1.50 Geometric mean 31.6 [lgE], IU/ml 3.76 Log,. WBC Geometric mean WBC, 5,750 cellslmm• Log.. eosinophil counlt 2.25 Geometric mean EOS, 168 cellslmm"§ 0.609 Log.. EOS as % ofWBCi Geometric mean EOS 3.06 as%ofWBC§
(29.4) (27.5)
47.4 30.7 24.5•• 47.2** 30.6
(28.3) (22.1)
(0.66)
1.93 85.1
(0.65)11
(0.12)
3.77 5,890
(0.30)
2.43 259
(0.221)
o. 738
(0.11)
(0.33XI
(0.251XI
4.47
*WBC =blood total leukocyte count; EOS =blood eosinophil count; NS =not significant. Values in parentheses are SD. tNumber is smaller fur some characteristics because of missing data. iConstant added before logarithmic transformation to eliminate values of zero. §Constant added before logarithmic transformation has been subtracted. IIP<0.001 compared with subjects without asthma by Student's t test. 'p
Subjects perfOrmed forced vital capacity maneuvers in the standing position without noseclips. An 8-L water-filled spirometer with microprocessor (Eagle II, Warren E. Collins Inc, Braintree, Mass) was used to measure FVC and FEV., which were corrected to body temperature pressure saturated. A minimum of three acceptable maneuvers were perfOrmed, unless this could not be accomplished after eight attempts. Additional maneuvers were perfOrmed, up to a muimum of eight, until 5 percent reproducibility of the highest two values ofboth FVC and FEV1 were obtained if possible. The highest values of FVC and FEV, were used for analysis. The FEV.;FVC ratio from the spirogram with the highest sum of FVC plus FEV, was used fur analysis. Predicted values for FEV., FVC, and FEV.;FVC ratio were derived from regressions of these measurements on age and height among 250 asymptomatic never smokers in the Normative Aging Study cohort.
CeUCountB Total blood leukocyte was measured using an automated cell counter (either a Hemalog 8, Technicon, Inc, 18rrytown, NY, or a Coulter Model S-plus VI, Coulter Electronics, Hialeah, Fla). Eosinophil counts were perfOrmed by a trained technician using a hemacytometer after staining blood with a reagent system (Unopette, Becton Dickenson and Co, RutherfOrd, Nn.
Serum 'lbtollmmunoglobulln E (IgE) Concentration Serum totallgE concentration was determined by paper radioimmunosorbent test (Pharmacia Diagnostics, Piscataway, NJ). The mean of two determinations on each serum specimen was used for analysis.
Allergy Skin 'lllsfing Subjects were tested with fOur common aeroallergens (ragweed, house dust, mixed trees, and mixed grasses) along with a glycerin control using the prick method. Reactions were measured at 20 min. Wheal size was recorded as the longest diameter of the wheal plus the perpendicular diameter, and the mean of these measurements was used fur analysis. For the current analysis, a positive skin test was defined as a mean wheal diameter <:!:5 mm after subtracting the size of any wheal reaction to the glycerin control.
DataAnolym
T-tests and )(- tests (18ble 1), analyses of covariance (18bles 2 and 3 and Figs 1 and 2), and stepwise multiple linear regression (18ble 4) were perfOrmed using a computer software system (SAS, SAS Inc, Carey, NC). Values of FEV, presented in 18bles 2 and 3 and Figures 1 and 2 are adjusted for age, height, pack-years, and (fur the total sample) smoking status by analysis of covariance. Residual analysis was perfOrmed to assess the fit of models used in analyses
Table 2-&latioruhip ofFEV, (UtMs)* to Sldn 1Nt ~
Skin tests negative C!: 1
positive skin test
p value by analysis of covariance
Total
Never Smokers
FOrmer Smokers
Current Smokers
3.14 (0.02) n= 865 3.10 (0.03) n=247 NS
3.25 (0.03) n=323 3.17 (0.05) n=89 NS
3.15 (0.03) n=412 3.12 (0.05) n=125 NS
2.83 (0.05) n=130 2.91 (0.11) n=33 NS
*Values in ( ) are mean (SE). Adjusted by analysis of covariance for age, height, pack-years, and (total sample) smoking status. Number in each group is given below mean value. NS=not significant. CHEST I 103 I 2 I FEBRUARY, 1883
377
No cough or phlegm Chronic cough without usual phlegm Usual phlegm without chronic cough
Both cough and phlegm p value by analysis of covariance
1btal
Never Smokers
FOnner Smokers
Current Smokers
3.18 (0.02) n=900 3.16 (0.09) n=35 3.05 (0.03) n=242 2.98 (0.06) n=93 <0.001
3.26 (0.02) n=364 3.27 (0.14) n=10 3.17 (0.~ n=75 2.91 (0.12) n=14 <0.05
3.18 (0.02) n=462 3.00 (0.16) n=10 3.08 (0.05) n=ll7 3.01 (0.10) n=26 NS
3.01 (0.07) n=74 3.03 (0.15) n=15 2.77 (0.08) n=SO 2.76 (0.08) n=53 <0.05
*Values in ( ) are mean (SE). Adjusted by analysis of covariance for age, height,
pack-years, and (total sample) smoking status. Number in each
group is given below mean value. NS =not significant.
ofcovariance and of models derived by stepwise regression. Because of their approximately log-normal distributions, IgE concentration and cell counts were logarithmically transfonned (base 10) prior to statisticalllll8lysis.
Subjects reporting a history of asthma displayed signi&cantly lower levels of pulmonary function, higher levels of serum total lgE, higher blood eosinophil counts, and a greater prevalence ofchronic cough and phlegm compared with subjects denying asthma. Subjects reporting a history of asthma did not differ signi&cantly in terms of smoking history from those denying asthma. A weak inverse relationship between FEV1 and blood eosinophil count was not statistically signi&cant (Fig 1). FEV1 was unrelated to blood eosinophil count expressed as a percentage of total leukocytes (Fig 1). A weak inverse relationship between FEV1 and serum total lgE concentration was not statistically significant (Fig 1). In analyses restricted to subjects in a particular
REsuLTS Characteristics of the 1,298 subjects with complete questionnaire data concerning asthma history are given in 'lllble 1 (three additional subjects included in later analyses failed to complete the question about asthma). Fifty-three of 1,298 subjects (4.1 percent) reported ever having asthma, although only 16 subjects (1.2 percent) reported that they still have asthma. Of the 53 subjects reporting asthma, all but 2 reported that this diagnosis had been made by a physician. 3.6 3.!1
3.4
...
-
"jj)3.3 Cl)
5 3.2
~
3.S
"0 Cl)
1ii 3.0 ::l
:c < 2.9 2.1
2.7 (331)
(211)
(311)
(214)
2
3
•
Quartile of eosinophils/mm3
(310)
(311)
(311)
(310)
2
3
•
Quartile of eosinphils/WBC
(217)
(217)
(211)
(217)
2
3
•
Quartile of serum lgE
FIGURE 1. Adjusted FEV, (mean± SE) in liters according to quartile of blood eosinophil count, blood eosinophil count as a percentage of total leukocytes (eosinophilsiWBC), and serum total IgE concentration. FEV, bas been adjusted for age, height, pack-years, and smoking status by analysis of covariance. Adjusted FEV, was not significantly related to quartile of eosinophil count, eosinophiiiWBC, or IgE concentration. The number of subjects in each quartile is given in parentheses.
378
Flak Facklnl for Venlllalory lmpalrmant (O'Connor et el)
3.&
Total
Never smokers
Former smokers
Current smokers
p<.001
p<.05
p<.01
Not significant
(317) (323) (323) (324)
(144) (124) (1 13) ( 8!1)
(1!18) (172) (152) (131)
( 15) ( 27) ( !18) (tOO)
3.!1 3.4
-
Uj'3.3
....Q)
5 3.2
~
3.1
"Q Q)
1ii 3.0 :I
:0 < 2.9 2.8 2.7
23
..
2
3
..
2
3
23
..
..
Quartile of WBC F'IcuRE 2. Adjusted FEV, (mean±SE) in liten according to quartile of total leukocyte count (WBC) derived from the total sample. FEV, has been adjusted for age, height, pack-yean, and (total sample) smoking status by analysis of covariance. The following values (cellslmm") define the quartiles of WBC: ::S4,700; 4,800 to 5,700; 5,800 to 6,900; '?::.7 ,000. The number of subjects in each quartile ofWBC is given in parentheses.
smoking status group (ie, never, former, and current smokers), FEV1 was not significantly related to eosinophil count, eosinophil count as a percentage of total leukocytes, or lgE concentration in any smoking status group. FEV1 was not significantly related to cutaneous immediate hypersensitivity to one or more common aeroallergens (Thble 2). There was a significant inverse relationship between FEV1 and blood total leukocyte count that appeared to be limited to never and former smokers (Fig 2). For the total sample, the FEV1 of subjects in the highest quartile oftotalleukocyte count was on average 0.17
L lower than that of subjects in the lowest quartile of total leukocyte count. The report of usual phlegm production, with or without chronic cough, was associated with lower FEV1 even after adjusting for smoking status and packyears (Thble 3). This relationship was observed in all three smoking groups, although it was not statistically significant among former smokers. Chronic cough without phlegm was relatively unusual (2. 7 percent of total sample) and not significantly related to FEV1 • In a stepwise multiple regression analysis (Thble 4), significant predictors ofFEV1 included current smok-
Table 4-&latiotuhip ofFEV, to Predictor l&riabla in Stepcoiae Mullipk .Regn.ion Antllpa: .Regn.ion Coej}icienta (SE) in Final Modela* All Variable Intercept Age, yrt Height, incbest Pack-yearst Current smoking (0,1) Phlegm (0,1) Log,. WBC, Log,. Cells, mm"t
Subjects (N=1,274) 3.21 -0.032 0.078 -0.0039 -0.240 -0.134 -0.463
(0.002)* (0.006)* (0.0005)* (0.045)* (0.033)* (0.121)*
Never Smoken (N=462)
FOrmer Smoken (N=613)
Current Smoken (N=199)
3.32 -0.035 (0.003)* 0.084 (0.008)*
3.20 -0.032 (0.003)* 0.078 (0.008)* -0.0046 (0.0008)*
2.93 -0.030 (0.006)* 0.072 (0.016)* -0.00087 (0.0014)
0.122 (0.052)§ -0.552 (0.181~1
-0.082 -0.613
-0.260 0.072
(0.051) (0.183)*
(0.079~1
(0.324)
"The following variables failed to enter the model for all subjects or within any smoking status category: former smoking (0,1); chronic cough (0,1); log,. eosinophil count; log,. eosinophil percentage; log,. lgE concentration; skin test positivity (0,1). WBC=totalleukocyte count; NS =failed to enter model as significant predictor of FEV,. tMean-centered.
*P
378
ing status, total pack-years, usual phlegm production, and total leukocyte count. Former smoking as a dichotomous indicator variable, chronic cough, absolute eosinophil count, eosinophil count as a percentage of total leukocytes, serum total IgE concentration, and allergy skin test positivity failed to enter the model for all subjects or within any smoking category. Predictor variables that were significantly related to FEV1 in the model for all subjects were retained in the models for specific smoking status groups to facilitate comparison between groups ('Thble 4). Repeating the model for all subjects with a term for the interaction between current smoking and total leukocyte count revealed that the relationship between leukocyte count and FEV1 was significantly weaker among current smokers than among former and never smokers (p=O.Ol). A term for the interaction between current smoking and phlegm was not significant, indicating that the relationship of reported phlegm to FEV1 did not vary significantly with smoking status. Stepwise regression analyses were repeated after excluding subjects who reported a recent respiratory infection from which they had not been symptom free for at least 3 weeks. Exclusion of these 270 subjects did not reduce the relationship of FEV1 to total leukocyte count observed among never and former smokers. The only difference resulting from this exclusion was the entry of chronic cough into the model (regression coefficient= -0.214, p=0.04) among never smokers. The analyses described in 'Thbles 2 through 4 and Figures 1 through 4 were repeated after excluding the 53 subjects with a history of asthma and the 3 subjects who failed to complete the questionnaire item about asthma. Exclusion of these 56 subjects weakened the relationship between usual phlegm production and FEV1 among never smokers but did not appreciably alter any of the other relationships described. DISCUSSION
In this sample of middle-aged and elderly men who were free of chronic diseases, including asthma, when enrolled in the Normative Aging Study 20 to 25 years ago, smoking history was the most important determinant of FEV1 • After adjusting for smoking status, markers ofallergy (eosinophil count, serum total [IgE], and skin test reactivity to common aeroallergens) were not significantly related to FEV1 in this sample. Eosinophil count displayed a weak inverse relationship to FEV1 in the total sample but was not related to FEV1 after adjustment for total leukocyte count. There was a significant inverse relationship between blood total leukocyte count and FEV1 • The report of usual phlegm production was associated with significantly lower FEV1 • Restricting analysis to subjects who denied any history of asthma did not appreciably alter
-
the observed relationships. Previous reports have provided conflicting data concerning the relationship of eosinophil count to ventilatory impairment. In their examination of subjects between 40 and 74 years of age, Burrows and coworkersS observed that the presence of 5 percent or more eosinophils on the differential leukocyte count was associated with significantly lower FEV1 only among nonsmokers; however, this relationship was not significant after excluding subjects with a history of asthma. In contrast, Kauffinann et aJ8 studied 912 working men between the ages of 22 and 55 years and observed an inverse relationship between FEV1 and eosinophil count (expressed either as cells per cubic millimeter or as a percent of total leukocyte count) among nonsmokers that was present even after excluding asthmatics. Mensinga et al10 have reported an inverse relationship between FEV1 and absolute eosinophil count (total leukocyte count was not measured) in a population-based sample with an age range of 15 to 64 years, even after excluding asthmatics. Modification of this relationship by smoking status was not addressed. In a longitudinal study of nonasthmatic men initially between the ages of 25 and 61 years, 'Thylor and coworkers7 observed a direct relationship between eosinophil count, measured at the end of follow-up, and rate ofFEV1 decline among nonsmokers but not among current or former smokers. The current data reveal no significant relationship between FEV1 and eosinophil count as a percent of total leukocyte count, a finding that did not change when subjects reporting asthma were excluded. This observation is in agreement with that of Burrows et al, 5 whose sample had a similar age range, but it is in conflict with the above-cited studies of younger samples. 8.7·10 'Ihken together, these data suggest that the relationship of eosinophilia to ventilatory impairment may diminish with age. It remains possible that eosinophilia in earlier life may contribute to the development of irreversible ventilatory impairment even though a cross-sectional association is no longer apparent after eosinophilia has diminished with advancing age. The absence of a significant relationship between total serum IgE and FEV1 among subjects without a history of asthma is in agreement with most previous reports. Burrows and coworkersS observed an inverse relationship between serum IgE and FEV1 among all subjects between the ages of 40 and 74 years but found no significant relationship after excluding subjects with a history of asthma. Vollmer et al11 observed an inverse relationship between IgE and FEV1 among subjects attending an emphysema screening clinic, but this relationship was not seen in a younger random sample of municipal employees. Three longitudinal studies7 •11 ·1! have indicated that serum total IgE is not Rlllk FB1Ds lor Ventilatory lmplllr111111d (O'Connor et el)
related to rate of decline of FEV1. Thus, serum total IgE may be elevated in association with reduced FEV1 in the setting of asthma or asthmatic bronchitis, but IgE does not appear to be an important determmant of FEV1 in individuals denying asthma. The lack of association between atopy, as assessed by cutaneous immediate hypersensitivity to common aeroallergens, and ventilatory impairment is also consistent with prior reports. In a population sample with an age range from 40 to 74 years, Burrows and coworkers& observed only a weak inverse relationship between skin test reactivity and FEV" and no such relationship was present after excluding subjects reporting a history of asthma. Longitudinal studies have indicated that decline of FEV1over time is not related to atopic status assessed by skin testing. 7·12•13 An inverse relationship between total leukocyte count and FEV1 has been reported previously in this population14 and by other groups. 15•16 This finding is consistent with the hypothesis that inflammation of the small airways and/or lung parenchyma results in the chronic obstructive bronchiolitis and/or pulmonary emphysema that is responsible for chronic airHow obstruction in adults with COPD. Longitudinal studies have provided contradictory data concerning a possible relationship between total leukocyte count and the rate ofFEV1 decline over time. 7·14 The absence of a relationship between FEV1 and total leukocyte count among current smokers in the present sample may reHect selection factors. The proportion of current smokers in the cohort has declined from 37 to 16 percent during the approximately 20 years since the Normative Aging Study began. Subjects who continue to smoke may represent a select group of"survivors" that is less susceptible to the adverse effects of inflammation on pulmonary function compared to never smokers and former smokers in the current cohort. The finding of reduced FEV1 in association with reported sputum production is in agreement with cross-sectional data from other community-based studies.5·17 Chronic bronchial mucous hypersecretion may simply be a more accuate indicator of actual tobacco smoke exposure than estimated pack-years. Alternatively, mucous hypersecretion may be a marker of increased susceptibility to chronic obstructive bronchiolitis and pulmonary emphysema. Mucous hypersecretion may be associated with bronchiolar inflammation and goblet cell hypersecretion that increase resistance in the peripheral airways. ConHicting reports concerning chronic productive cough and the rate of FEV1 decline over time18·19 contribute to uncertainty about the explanation for this cross-sectional association. A potential limitation of the current analysis relates to its generalizability in light of the particular charac-
teristics of the Normative Aging Study cohort. Initial screening produced a cohort that was healthier than the average US population; however, such a screening effect tends to dissipate with the passage of time as a cohort is followed.• The health of the current sample may also be enhanced by a healthy survivor effect, possibly biasing observed relationships toward the null. A previous report from the Normative Aging Study indicated that markers of atopy were associated with greater methacholine airway responsiveness whereas blood total leukocyte count was unrelated to responsiveness.z1 The present report reveals a contrasting pattern of relationships for FEV" which is significantly related to total leukocyte count but not independently related to markers ofallergy. In this population, allergic airway inHammation may contribute to increased airway responsiveness while having little impact on baseline level of pulmonary function; in contrast, nonallergic inflammation manifested by elevation of the total leukocyte count may be more important in the development of persistent ventilatory impairment. In summary, factors that were independently related to FEV1 in this sample of mostly nonasthmatic middle-aged and older men included current cigarette smoking, total pack-years of cigarette smoking, usual phlegm production, and blood total leukocyte count. Markers of allergy including skin test reactivity, serum total IgE concentration, and eosinophil count did not display significant independent relationships to FEV1. Physiologic factors associated with ventilatory impairment may be different in younger populations and in persons with lifelong asthma. REFERENCES 1 Yan K, Salome CM, Woolcock AJ. Prevalence and nature of bronchial hyperresponsiveness in subjects with chronic obstructive pulmonary disease. Am Rev Respir Dis 1985; 132:25-9 2 Burrows B, Bloom Jw. Traver GA, Cline MG. The course and prognosis of different forms of chronic airways obstruction in a sample from the general population. N Eng! J Med 1987; 317:1309-14 3 Burrows B, Knudson RJ, Cline MG, Lebowitz MD. Quantitative relationships between cigarette smoking and ventilatory function. Am Rev Respir Dis 1977; 115:195-205 4 Beck GJ, Doyle CA, Schachter EN. Smoking and lung function. Am Rev Respir Dis 1981; 123:149-55 5 Burows B, Knudson RJ, Cline MG, Lebowitz MD. A reexamination of risk factors for ventilatory impairment. Am Rev Respir Dis 1988; 138:829-36 6 Kauffmann F, Neukirch F, Korobaell" M, Marne MJ, Claude JR, Lellouch J. Eosinophils, smoking, and lung function. Am Rev Respir Dis 1986; 134:1172-75 7 'lilylor RG, Joyce H, Gross E, Holland F, Pride NB. Bronchial reactivity to inhaled histamine and annual rate of decline in and ex-smokers. Thorax 1985; 40:9-16 FEV, in male ~~mokers 8 Bell B, Rose Cl, Damon H. The Normative Aging Study: an interdisciplinary and longitudinal study of health and aging. Aging Hum Devel1972; 3:5-17 9 Ferris BG Jr. Epidemiology standardization project. Am Rev CHEST 1103 I 2 I FEBRUARY. 1993
381
Hespir Dis 1978; 118(6, pt2):1-88 10 Mensinga TI, Schouten JP, Rijcken B, Weiss ST, Speizer FE, Van der Lende R. The relationship of eosinophilia and positive skin test reactivity to respiratory symptom prevalence in a community-based population study. J Allergy Clio Immunol 1990; 86:99-107 11 Vollmer WM, Buist AS, Johnson LR, McCamant LE, Halonen M. Relationship between serum IgE and cross-sectional and longitudinal FEV, in two cohort studies. Chest 1986; 90:416-23 12 Parker DR, O'Connor GT, Sparrow D, Segal MR, Weiss ST. The relationship of nonspecific airway responsiveness and atopy to the rate of decline of lung function. Am Rev Hespir Dis 1990; 141:589-94 13 Annesi I, Neukirch F, Orvoen-Frija E, Oryszczyn MP, Korobaeff M, Dore MF, et al. The relevance of hyperresponsiveness but not of atopy to FEV, decline: preliminary results in a working population. Bull Eur Physiopathol Hespir 1987; 23:397-400 14 Sparrow D, Glynn RJ, Cohen M, Weiss ST. The relationship of the peripheral leukocyte count and cigarette smoking to pulmonary function among adult men. Chest 1984; 86:383-86 15 Yeung MC, Buncio AD. Leukocyte count, smoking, and lung
function. Am J Med 1984; 76:31-7 16 Carel RS, 1bckman MS, Baser M. Smoking, leukocyte count, and ventilatory lung function in working men. Chest 1988; 93:1137-43 17 Fletcher CM, Tinker CM. Chronic bronchitis: a further study of simple diagnostic methods in a working population. BMJ 1961; 1:1491-98 18 Fletcher CM, Peto R, Tinker C, Speizer FE. The natural history of chronic bronchitis and emphysema: an 8 year study of early chronic obstructive lung disease in working men in London. New York: Oxford University Press, 1976 19 Higgins MW, Keller JB, Becker M, Howatt W, Landis JR, Rotman H, et al. An index of risk lOr obstructive airways disease. Am Rev Respir Dis 1982; 2:144-51 20 Doll R. Retrospective and prospective studies. In: Witts LJ, ed. Medical surveys and clinical trials. London: Oxford University Press, 1959; 64-90 21 O'Connor GT, Sparrow D, Segal MR, Weiss ST. Smoking, atopy, and methacholine airway responsiveness among middle-aged and elderly men. Am Rev Hespir Dis 1989; 140:1520-26
20th Annual Harvard Intensive Care Medicine Course This course will be held at the Marriott Hotel, Cambridge, Massachusetts, April 16-18, under the direction of Dr. Warren M. Zapol. For information, contact Dr. Zapol, Department of Anesthesia, Massachusetts General Hospital, Boston 02114 (617:726-3030).
382
Flllk Factors tor VenlllatDry Impairment (O'Connor 81111)
To evaluate the relationship of atopy and inftammation to the occurrence of ventilatory impairment, we studied 1,301 middle-aged and older men participating in the Normative Aging Study at the time of their 1984 to 1987 examination. Screening at entry to the study in the 1960s had excluded subjects with asthma and other chronic diseases at that time. After adjustment fur smoldng status and pack-years in this cross-sectional analysis, a weak inverse relationship between FEV, and blood eosinophil count was not statisticaDy signi&cant, and FEV, sbowed no relationship to blood eosinophil count as a percentage of total leukocytes. A weak inverse relationship between FEV, and serum total IgE concentration was not statistically significant. Cutaneous
C
hronic obstructive pulmonary disease (COPD) develops only in a minority of cigarette smokers. Conversely, studies of population samples have indicated that roughly 15 percent of adults with chronic airflow obstruction have no history of smoking.u Population studies have also indicated that estimated total pack-years of cigarette smoking accounts for only 14 to 15 percent of the variance of age- and heightadjusted forced expiratory volume in 1 s (FEV1) and expiratory ftow rates in the general population. 3 •4 It has been hypothesized that allergy may predispose to the development ofCOPD, but available data have not confirmed this hypothesis. Burrows and coworkers~! have recently reported population-based data indicating that markers of allergy, including blood eosinophilia. are not associated with reduced level of pulmonary function after excluding subjects with a history of asthma. These data suggest that only among
*From the Channing Laboratory, Department of Medicine, Brigham and Women$ Hospital; the Divisions of Pulmonary and Critical Care Medicine, BnJdwn and Women'S and Beth Israel Hospitals; Harvard Medical sChool; Nonnative Aging Study, Department of Veterans A8iUrs Outpatient Clinic, Boston; and the Department of Epidemiology and Biostatistics, University of California, San Francisco. Supported by grants HL-34645 and Hl.r45089 from the Division of Lung Diseases, National Heart, Lung, and Blood Institute, National Institute of Health, and by the Medical Research Service of the Department of Veterans Atrairs. Presented In part at the 1989 Annual Meeting, American Thoracic Society, Cincinnati, May 14, 1989. Dr. O'Connor is recipient of an Edward Livingston Trudeau Scholar Award from the American Lung Association. Manuscript received December 26, 1991; revision accepted June 26. Reprint requests: Dr: \\friss, Channing Laboratory, 180 LDngwood Avenue, BoBton 02115
m
immediate hypersensitivity to one or more COIDIDOD aeroallergens was not related to FEV,. A signilicant inverse relationship between FEV, and blood total leukocyte count was observed in never and furmer smokers. FEV, was significantly lower in subjects reporting usual phlegm production. These &ndings suggest that ventilatory impairment is not related to atopic status among middle-aged and older men without a history of asthma. The inverse relationship between total leukocyte count and FEV1 in this sample supports the hypothesis that nonallergic inAammation plays a role in the pathogenesis of chroaic air8ow obstruction in this group. (C,_, 1993; 103:376-8!)
asthmatics does allergy contribute to the development of chronic airftow obstruction; however, other investigators have observed that blood eosinophil count is inversely related to level ofFEV16 and directly related to rate ofdecline ofFEV17 among nonsmoking nonasthmatics. We have examined the relationship of FEV1 to markers of allergy and other covariates among men participating in the Normative Aging Study. METHODS
Sample The Nonnative Aging Study is a longitudinal study of aging located at the Department of Veterans Affairs Outpatient Clinic In Boston.• The cohort initially consisted of 2,280 men who were 21 to 80 years of age on entry to the study between 1961 and 1969. \blunteers were screened at entry, and those with chrooic medical conditions, including asthma, chronic bronchitis, and chronic sinusitis, were excluded. Subsequent to entry, volunteers have returned every 3 to 5 years for a comprehensive examination. The present cross·sectional analysis involves men who returned for examination from April1984, when assessment of atopic status was begun at the Normative Aging Study, until December 1987. 1\vo men with total blood leulcocyte counts greater than 30,
Table 1-Sample C'haracterimcs* Without Asthma No.t Age, yr FEV., 9& predicted FVC, % predicted FEV.;FVC ratio,% Smoking habits Current smoker, % Exsmoker, 9& Never smoker, %
1,245 61.0 93.7 96.3 78.3
(7.1) (15.9) (14.4) (7.5)
most days, fur at least 3 months per year, for at least 2 years' duration. Usual phlegm is defined as the report of usual phlegm from the chest.
With
Asthma 53 61.3 82.3 90.8 72.5
Spirometry
(9.4) (20.0~1
(17.0)t (10.1~1
20.8 39.6 39.6
15.4 48.5 36.1
Pack-years Current smokers 48.4 32.1 Exsmokers 9.3 Chronic cough, % 25.7 Usual phlegm, 9& <:!:1 positive allergy 21.7 skin test,% Log,. [lgE] 1.50 Geometric mean 31.6 [lgE], IU/ml 3.76 Log,. WBC Geometric mean WBC, 5,750 cellslmm• Log.. eosinophil counlt 2.25 Geometric mean EOS, 168 cellslmm"§ 0.609 Log.. EOS as % ofWBCi Geometric mean EOS 3.06 as%ofWBC§
(29.4) (27.5)
47.4 30.7 24.5•• 47.2** 30.6
(28.3) (22.1)
(0.66)
1.93 85.1
(0.65)11
(0.12)
3.77 5,890
(0.30)
2.43 259
(0.221)
o. 738
(0.11)
(0.33XI
(0.251XI
4.47
*WBC =blood total leukocyte count; EOS =blood eosinophil count; NS =not significant. Values in parentheses are SD. tNumber is smaller fur some characteristics because of missing data. iConstant added before logarithmic transformation to eliminate values of zero. §Constant added before logarithmic transformation has been subtracted. IIP<0.001 compared with subjects without asthma by Student's t test. 'p
Subjects perfOrmed forced vital capacity maneuvers in the standing position without noseclips. An 8-L water-filled spirometer with microprocessor (Eagle II, Warren E. Collins Inc, Braintree, Mass) was used to measure FVC and FEV., which were corrected to body temperature pressure saturated. A minimum of three acceptable maneuvers were perfOrmed, unless this could not be accomplished after eight attempts. Additional maneuvers were perfOrmed, up to a muimum of eight, until 5 percent reproducibility of the highest two values ofboth FVC and FEV1 were obtained if possible. The highest values of FVC and FEV, were used for analysis. The FEV.;FVC ratio from the spirogram with the highest sum of FVC plus FEV, was used fur analysis. Predicted values for FEV., FVC, and FEV.;FVC ratio were derived from regressions of these measurements on age and height among 250 asymptomatic never smokers in the Normative Aging Study cohort.
CeUCountB Total blood leukocyte was measured using an automated cell counter (either a Hemalog 8, Technicon, Inc, 18rrytown, NY, or a Coulter Model S-plus VI, Coulter Electronics, Hialeah, Fla). Eosinophil counts were perfOrmed by a trained technician using a hemacytometer after staining blood with a reagent system (Unopette, Becton Dickenson and Co, RutherfOrd, Nn.
Serum 'lbtollmmunoglobulln E (IgE) Concentration Serum totallgE concentration was determined by paper radioimmunosorbent test (Pharmacia Diagnostics, Piscataway, NJ). The mean of two determinations on each serum specimen was used for analysis.
Allergy Skin 'lllsfing Subjects were tested with fOur common aeroallergens (ragweed, house dust, mixed trees, and mixed grasses) along with a glycerin control using the prick method. Reactions were measured at 20 min. Wheal size was recorded as the longest diameter of the wheal plus the perpendicular diameter, and the mean of these measurements was used fur analysis. For the current analysis, a positive skin test was defined as a mean wheal diameter <:!:5 mm after subtracting the size of any wheal reaction to the glycerin control.
DataAnolym
T-tests and )(- tests (18ble 1), analyses of covariance (18bles 2 and 3 and Figs 1 and 2), and stepwise multiple linear regression (18ble 4) were perfOrmed using a computer software system (SAS, SAS Inc, Carey, NC). Values of FEV, presented in 18bles 2 and 3 and Figures 1 and 2 are adjusted for age, height, pack-years, and (fur the total sample) smoking status by analysis of covariance. Residual analysis was perfOrmed to assess the fit of models used in analyses
Table 2-&latioruhip ofFEV, (UtMs)* to Sldn 1Nt ~
Skin tests negative C!: 1
positive skin test
p value by analysis of covariance
Total
Never Smokers
FOrmer Smokers
Current Smokers
3.14 (0.02) n= 865 3.10 (0.03) n=247 NS
3.25 (0.03) n=323 3.17 (0.05) n=89 NS
3.15 (0.03) n=412 3.12 (0.05) n=125 NS
2.83 (0.05) n=130 2.91 (0.11) n=33 NS
*Values in ( ) are mean (SE). Adjusted by analysis of covariance for age, height, pack-years, and (total sample) smoking status. Number in each group is given below mean value. NS=not significant. CHEST I 103 I 2 I FEBRUARY, 1883
377
No cough or phlegm Chronic cough without usual phlegm Usual phlegm without chronic cough
Both cough and phlegm p value by analysis of covariance
1btal
Never Smokers
FOnner Smokers
Current Smokers
3.18 (0.02) n=900 3.16 (0.09) n=35 3.05 (0.03) n=242 2.98 (0.06) n=93 <0.001
3.26 (0.02) n=364 3.27 (0.14) n=10 3.17 (0.~ n=75 2.91 (0.12) n=14 <0.05
3.18 (0.02) n=462 3.00 (0.16) n=10 3.08 (0.05) n=ll7 3.01 (0.10) n=26 NS
3.01 (0.07) n=74 3.03 (0.15) n=15 2.77 (0.08) n=SO 2.76 (0.08) n=53 <0.05
*Values in ( ) are mean (SE). Adjusted by analysis of covariance for age, height,
pack-years, and (total sample) smoking status. Number in each
group is given below mean value. NS =not significant.
ofcovariance and of models derived by stepwise regression. Because of their approximately log-normal distributions, IgE concentration and cell counts were logarithmically transfonned (base 10) prior to statisticalllll8lysis.
Subjects reporting a history of asthma displayed signi&cantly lower levels of pulmonary function, higher levels of serum total lgE, higher blood eosinophil counts, and a greater prevalence ofchronic cough and phlegm compared with subjects denying asthma. Subjects reporting a history of asthma did not differ signi&cantly in terms of smoking history from those denying asthma. A weak inverse relationship between FEV1 and blood eosinophil count was not statistically signi&cant (Fig 1). FEV1 was unrelated to blood eosinophil count expressed as a percentage of total leukocytes (Fig 1). A weak inverse relationship between FEV1 and serum total lgE concentration was not statistically significant (Fig 1). In analyses restricted to subjects in a particular
REsuLTS Characteristics of the 1,298 subjects with complete questionnaire data concerning asthma history are given in 'lllble 1 (three additional subjects included in later analyses failed to complete the question about asthma). Fifty-three of 1,298 subjects (4.1 percent) reported ever having asthma, although only 16 subjects (1.2 percent) reported that they still have asthma. Of the 53 subjects reporting asthma, all but 2 reported that this diagnosis had been made by a physician. 3.6 3.!1
3.4
...
-
"jj)3.3 Cl)
5 3.2
~
3.S
"0 Cl)
1ii 3.0 ::l
:c < 2.9 2.1
2.7 (331)
(211)
(311)
(214)
2
3
•
Quartile of eosinophils/mm3
(310)
(311)
(311)
(310)
2
3
•
Quartile of eosinphils/WBC
(217)
(217)
(211)
(217)
2
3
•
Quartile of serum lgE
FIGURE 1. Adjusted FEV, (mean± SE) in liters according to quartile of blood eosinophil count, blood eosinophil count as a percentage of total leukocytes (eosinophilsiWBC), and serum total IgE concentration. FEV, bas been adjusted for age, height, pack-years, and smoking status by analysis of covariance. Adjusted FEV, was not significantly related to quartile of eosinophil count, eosinophiiiWBC, or IgE concentration. The number of subjects in each quartile is given in parentheses.
378
Flak Facklnl for Venlllalory lmpalrmant (O'Connor et el)
3.&
Total
Never smokers
Former smokers
Current smokers
p<.001
p<.05
p<.01
Not significant
(317) (323) (323) (324)
(144) (124) (1 13) ( 8!1)
(1!18) (172) (152) (131)
( 15) ( 27) ( !18) (tOO)
3.!1 3.4
-
Uj'3.3
....Q)
5 3.2
~
3.1
"Q Q)
1ii 3.0 :I
:0 < 2.9 2.8 2.7
23
..
2
3
..
2
3
23
..
..
Quartile of WBC F'IcuRE 2. Adjusted FEV, (mean±SE) in liten according to quartile of total leukocyte count (WBC) derived from the total sample. FEV, has been adjusted for age, height, pack-yean, and (total sample) smoking status by analysis of covariance. The following values (cellslmm") define the quartiles of WBC: ::S4,700; 4,800 to 5,700; 5,800 to 6,900; '?::.7 ,000. The number of subjects in each quartile ofWBC is given in parentheses.
smoking status group (ie, never, former, and current smokers), FEV1 was not significantly related to eosinophil count, eosinophil count as a percentage of total leukocytes, or lgE concentration in any smoking status group. FEV1 was not significantly related to cutaneous immediate hypersensitivity to one or more common aeroallergens (Thble 2). There was a significant inverse relationship between FEV1 and blood total leukocyte count that appeared to be limited to never and former smokers (Fig 2). For the total sample, the FEV1 of subjects in the highest quartile oftotalleukocyte count was on average 0.17
L lower than that of subjects in the lowest quartile of total leukocyte count. The report of usual phlegm production, with or without chronic cough, was associated with lower FEV1 even after adjusting for smoking status and packyears (Thble 3). This relationship was observed in all three smoking groups, although it was not statistically significant among former smokers. Chronic cough without phlegm was relatively unusual (2. 7 percent of total sample) and not significantly related to FEV1 • In a stepwise multiple regression analysis (Thble 4), significant predictors ofFEV1 included current smok-
Table 4-&latiotuhip ofFEV, to Predictor l&riabla in Stepcoiae Mullipk .Regn.ion Antllpa: .Regn.ion Coej}icienta (SE) in Final Modela* All Variable Intercept Age, yrt Height, incbest Pack-yearst Current smoking (0,1) Phlegm (0,1) Log,. WBC, Log,. Cells, mm"t
Subjects (N=1,274) 3.21 -0.032 0.078 -0.0039 -0.240 -0.134 -0.463
(0.002)* (0.006)* (0.0005)* (0.045)* (0.033)* (0.121)*
Never Smoken (N=462)
FOrmer Smoken (N=613)
Current Smoken (N=199)
3.32 -0.035 (0.003)* 0.084 (0.008)*
3.20 -0.032 (0.003)* 0.078 (0.008)* -0.0046 (0.0008)*
2.93 -0.030 (0.006)* 0.072 (0.016)* -0.00087 (0.0014)
0.122 (0.052)§ -0.552 (0.181~1
-0.082 -0.613
-0.260 0.072
(0.051) (0.183)*
(0.079~1
(0.324)
"The following variables failed to enter the model for all subjects or within any smoking status category: former smoking (0,1); chronic cough (0,1); log,. eosinophil count; log,. eosinophil percentage; log,. lgE concentration; skin test positivity (0,1). WBC=totalleukocyte count; NS =failed to enter model as significant predictor of FEV,. tMean-centered.
*P
378
ing status, total pack-years, usual phlegm production, and total leukocyte count. Former smoking as a dichotomous indicator variable, chronic cough, absolute eosinophil count, eosinophil count as a percentage of total leukocytes, serum total IgE concentration, and allergy skin test positivity failed to enter the model for all subjects or within any smoking category. Predictor variables that were significantly related to FEV1 in the model for all subjects were retained in the models for specific smoking status groups to facilitate comparison between groups ('Thble 4). Repeating the model for all subjects with a term for the interaction between current smoking and total leukocyte count revealed that the relationship between leukocyte count and FEV1 was significantly weaker among current smokers than among former and never smokers (p=O.Ol). A term for the interaction between current smoking and phlegm was not significant, indicating that the relationship of reported phlegm to FEV1 did not vary significantly with smoking status. Stepwise regression analyses were repeated after excluding subjects who reported a recent respiratory infection from which they had not been symptom free for at least 3 weeks. Exclusion of these 270 subjects did not reduce the relationship of FEV1 to total leukocyte count observed among never and former smokers. The only difference resulting from this exclusion was the entry of chronic cough into the model (regression coefficient= -0.214, p=0.04) among never smokers. The analyses described in 'Thbles 2 through 4 and Figures 1 through 4 were repeated after excluding the 53 subjects with a history of asthma and the 3 subjects who failed to complete the questionnaire item about asthma. Exclusion of these 56 subjects weakened the relationship between usual phlegm production and FEV1 among never smokers but did not appreciably alter any of the other relationships described. DISCUSSION
In this sample of middle-aged and elderly men who were free of chronic diseases, including asthma, when enrolled in the Normative Aging Study 20 to 25 years ago, smoking history was the most important determinant of FEV1 • After adjusting for smoking status, markers ofallergy (eosinophil count, serum total [IgE], and skin test reactivity to common aeroallergens) were not significantly related to FEV1 in this sample. Eosinophil count displayed a weak inverse relationship to FEV1 in the total sample but was not related to FEV1 after adjustment for total leukocyte count. There was a significant inverse relationship between blood total leukocyte count and FEV1 • The report of usual phlegm production was associated with significantly lower FEV1 • Restricting analysis to subjects who denied any history of asthma did not appreciably alter
-
the observed relationships. Previous reports have provided conflicting data concerning the relationship of eosinophil count to ventilatory impairment. In their examination of subjects between 40 and 74 years of age, Burrows and coworkersS observed that the presence of 5 percent or more eosinophils on the differential leukocyte count was associated with significantly lower FEV1 only among nonsmokers; however, this relationship was not significant after excluding subjects with a history of asthma. In contrast, Kauffinann et aJ8 studied 912 working men between the ages of 22 and 55 years and observed an inverse relationship between FEV1 and eosinophil count (expressed either as cells per cubic millimeter or as a percent of total leukocyte count) among nonsmokers that was present even after excluding asthmatics. Mensinga et al10 have reported an inverse relationship between FEV1 and absolute eosinophil count (total leukocyte count was not measured) in a population-based sample with an age range of 15 to 64 years, even after excluding asthmatics. Modification of this relationship by smoking status was not addressed. In a longitudinal study of nonasthmatic men initially between the ages of 25 and 61 years, 'Thylor and coworkers7 observed a direct relationship between eosinophil count, measured at the end of follow-up, and rate ofFEV1 decline among nonsmokers but not among current or former smokers. The current data reveal no significant relationship between FEV1 and eosinophil count as a percent of total leukocyte count, a finding that did not change when subjects reporting asthma were excluded. This observation is in agreement with that of Burrows et al, 5 whose sample had a similar age range, but it is in conflict with the above-cited studies of younger samples. 8.7·10 'Ihken together, these data suggest that the relationship of eosinophilia to ventilatory impairment may diminish with age. It remains possible that eosinophilia in earlier life may contribute to the development of irreversible ventilatory impairment even though a cross-sectional association is no longer apparent after eosinophilia has diminished with advancing age. The absence of a significant relationship between total serum IgE and FEV1 among subjects without a history of asthma is in agreement with most previous reports. Burrows and coworkersS observed an inverse relationship between serum IgE and FEV1 among all subjects between the ages of 40 and 74 years but found no significant relationship after excluding subjects with a history of asthma. Vollmer et al11 observed an inverse relationship between IgE and FEV1 among subjects attending an emphysema screening clinic, but this relationship was not seen in a younger random sample of municipal employees. Three longitudinal studies7 •11 ·1! have indicated that serum total IgE is not Rlllk FB1Ds lor Ventilatory lmplllr111111d (O'Connor et el)
related to rate of decline of FEV1. Thus, serum total IgE may be elevated in association with reduced FEV1 in the setting of asthma or asthmatic bronchitis, but IgE does not appear to be an important determmant of FEV1 in individuals denying asthma. The lack of association between atopy, as assessed by cutaneous immediate hypersensitivity to common aeroallergens, and ventilatory impairment is also consistent with prior reports. In a population sample with an age range from 40 to 74 years, Burrows and coworkers& observed only a weak inverse relationship between skin test reactivity and FEV" and no such relationship was present after excluding subjects reporting a history of asthma. Longitudinal studies have indicated that decline of FEV1over time is not related to atopic status assessed by skin testing. 7·12•13 An inverse relationship between total leukocyte count and FEV1 has been reported previously in this population14 and by other groups. 15•16 This finding is consistent with the hypothesis that inflammation of the small airways and/or lung parenchyma results in the chronic obstructive bronchiolitis and/or pulmonary emphysema that is responsible for chronic airHow obstruction in adults with COPD. Longitudinal studies have provided contradictory data concerning a possible relationship between total leukocyte count and the rate ofFEV1 decline over time. 7·14 The absence of a relationship between FEV1 and total leukocyte count among current smokers in the present sample may reHect selection factors. The proportion of current smokers in the cohort has declined from 37 to 16 percent during the approximately 20 years since the Normative Aging Study began. Subjects who continue to smoke may represent a select group of"survivors" that is less susceptible to the adverse effects of inflammation on pulmonary function compared to never smokers and former smokers in the current cohort. The finding of reduced FEV1 in association with reported sputum production is in agreement with cross-sectional data from other community-based studies.5·17 Chronic bronchial mucous hypersecretion may simply be a more accuate indicator of actual tobacco smoke exposure than estimated pack-years. Alternatively, mucous hypersecretion may be a marker of increased susceptibility to chronic obstructive bronchiolitis and pulmonary emphysema. Mucous hypersecretion may be associated with bronchiolar inflammation and goblet cell hypersecretion that increase resistance in the peripheral airways. ConHicting reports concerning chronic productive cough and the rate of FEV1 decline over time18·19 contribute to uncertainty about the explanation for this cross-sectional association. A potential limitation of the current analysis relates to its generalizability in light of the particular charac-
teristics of the Normative Aging Study cohort. Initial screening produced a cohort that was healthier than the average US population; however, such a screening effect tends to dissipate with the passage of time as a cohort is followed.• The health of the current sample may also be enhanced by a healthy survivor effect, possibly biasing observed relationships toward the null. A previous report from the Normative Aging Study indicated that markers of atopy were associated with greater methacholine airway responsiveness whereas blood total leukocyte count was unrelated to responsiveness.z1 The present report reveals a contrasting pattern of relationships for FEV" which is significantly related to total leukocyte count but not independently related to markers ofallergy. In this population, allergic airway inHammation may contribute to increased airway responsiveness while having little impact on baseline level of pulmonary function; in contrast, nonallergic inflammation manifested by elevation of the total leukocyte count may be more important in the development of persistent ventilatory impairment. In summary, factors that were independently related to FEV1 in this sample of mostly nonasthmatic middle-aged and older men included current cigarette smoking, total pack-years of cigarette smoking, usual phlegm production, and blood total leukocyte count. Markers of allergy including skin test reactivity, serum total IgE concentration, and eosinophil count did not display significant independent relationships to FEV1. Physiologic factors associated with ventilatory impairment may be different in younger populations and in persons with lifelong asthma. REFERENCES 1 Yan K, Salome CM, Woolcock AJ. Prevalence and nature of bronchial hyperresponsiveness in subjects with chronic obstructive pulmonary disease. Am Rev Respir Dis 1985; 132:25-9 2 Burrows B, Bloom Jw. Traver GA, Cline MG. The course and prognosis of different forms of chronic airways obstruction in a sample from the general population. N Eng! J Med 1987; 317:1309-14 3 Burrows B, Knudson RJ, Cline MG, Lebowitz MD. Quantitative relationships between cigarette smoking and ventilatory function. Am Rev Respir Dis 1977; 115:195-205 4 Beck GJ, Doyle CA, Schachter EN. Smoking and lung function. Am Rev Respir Dis 1981; 123:149-55 5 Burows B, Knudson RJ, Cline MG, Lebowitz MD. A reexamination of risk factors for ventilatory impairment. Am Rev Respir Dis 1988; 138:829-36 6 Kauffmann F, Neukirch F, Korobaell" M, Marne MJ, Claude JR, Lellouch J. Eosinophils, smoking, and lung function. Am Rev Respir Dis 1986; 134:1172-75 7 'lilylor RG, Joyce H, Gross E, Holland F, Pride NB. Bronchial reactivity to inhaled histamine and annual rate of decline in and ex-smokers. Thorax 1985; 40:9-16 FEV, in male ~~mokers 8 Bell B, Rose Cl, Damon H. The Normative Aging Study: an interdisciplinary and longitudinal study of health and aging. Aging Hum Devel1972; 3:5-17 9 Ferris BG Jr. Epidemiology standardization project. Am Rev CHEST 1103 I 2 I FEBRUARY. 1993
381
Hespir Dis 1978; 118(6, pt2):1-88 10 Mensinga TI, Schouten JP, Rijcken B, Weiss ST, Speizer FE, Van der Lende R. The relationship of eosinophilia and positive skin test reactivity to respiratory symptom prevalence in a community-based population study. J Allergy Clio Immunol 1990; 86:99-107 11 Vollmer WM, Buist AS, Johnson LR, McCamant LE, Halonen M. Relationship between serum IgE and cross-sectional and longitudinal FEV, in two cohort studies. Chest 1986; 90:416-23 12 Parker DR, O'Connor GT, Sparrow D, Segal MR, Weiss ST. The relationship of nonspecific airway responsiveness and atopy to the rate of decline of lung function. Am Rev Hespir Dis 1990; 141:589-94 13 Annesi I, Neukirch F, Orvoen-Frija E, Oryszczyn MP, Korobaeff M, Dore MF, et al. The relevance of hyperresponsiveness but not of atopy to FEV, decline: preliminary results in a working population. Bull Eur Physiopathol Hespir 1987; 23:397-400 14 Sparrow D, Glynn RJ, Cohen M, Weiss ST. The relationship of the peripheral leukocyte count and cigarette smoking to pulmonary function among adult men. Chest 1984; 86:383-86 15 Yeung MC, Buncio AD. Leukocyte count, smoking, and lung
function. Am J Med 1984; 76:31-7 16 Carel RS, 1bckman MS, Baser M. Smoking, leukocyte count, and ventilatory lung function in working men. Chest 1988; 93:1137-43 17 Fletcher CM, Tinker CM. Chronic bronchitis: a further study of simple diagnostic methods in a working population. BMJ 1961; 1:1491-98 18 Fletcher CM, Peto R, Tinker C, Speizer FE. The natural history of chronic bronchitis and emphysema: an 8 year study of early chronic obstructive lung disease in working men in London. New York: Oxford University Press, 1976 19 Higgins MW, Keller JB, Becker M, Howatt W, Landis JR, Rotman H, et al. An index of risk lOr obstructive airways disease. Am Rev Respir Dis 1982; 2:144-51 20 Doll R. Retrospective and prospective studies. In: Witts LJ, ed. Medical surveys and clinical trials. London: Oxford University Press, 1959; 64-90 21 O'Connor GT, Sparrow D, Segal MR, Weiss ST. Smoking, atopy, and methacholine airway responsiveness among middle-aged and elderly men. Am Rev Hespir Dis 1989; 140:1520-26
20th Annual Harvard Intensive Care Medicine Course This course will be held at the Marriott Hotel, Cambridge, Massachusetts, April 16-18, under the direction of Dr. Warren M. Zapol. For information, contact Dr. Zapol, Department of Anesthesia, Massachusetts General Hospital, Boston 02114 (617:726-3030).
382
Flllk Factors tor VenlllatDry Impairment (O'Connor 81111)