| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Social and Preventive Medicine (Drs. Schünemann, Dorn, and Trevisan), Pulmonary and Critical Care Division (Dr. Grant), Department of Medicine, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, Buffalo, NY; and the Division of Population Biology and Epidemiology (Dr. Winkelstein), School of Public Health, University of California at Berkeley, Berkeley, CA.
Correspondence to: Holger J. Schünemann, MD, PhD, Departments of Medicine and Social and Preventive Medicine, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, 270 Farber Hall, 3435 Main St, Buffalo, NY 14214-3000; e-mail: HJS{at}Buffalo.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Design: Prospective study with 29-year follow-up of the Buffalo Health Study cohort.
Participants: Randomly selected sample of 554 men and 641 women, aged 20 to 89 years, from all listed households of the city of Buffalo, NY.
Measurements and results: Baseline measurements were performed in 1960 to 1961. Pulmonary function was assessed based on FEV1 expressed as the normal percent predicted (FEV1%pred). FEV1%pred adjusted by age, body mass index, systolic BP, education, and smoking status was inversely related to all-cause mortality in both men and women (p < 0.01). A sequential survival analysis in participants who had a survival time of at least 5, 10, 15, 20, and 25 years after enrollment in the study was also performed. Except for men who survived for > 25 years, we observed a statistically significant negative association between FEV1%pred and all-cause mortality. FEV1%pred was also inversely related to ischemic heart disease (IHD) mortality. When participants were divided into quintiles of FEV1%pred, participants in the lowest quintile of FEV1%pred experienced significantly higher all-cause mortality compared with participants in the highest quintile of FEV1%pred. For the entire follow-up period, the adjusted hazard ratios for all-cause mortality were 2.24 (95% confidence interval [CI], 1.60 to 3.13) for men and 1.81 (95% CI, 1.24 to 2.63) for women, respectively. Hazard ratios for death from IHD in the lowest quintile of FEV1%pred were 2.11 (95% CI, 1.20 to 3.71) and 1.96 (95% CI, 0.99 to 3.88) for men and women, respectively.
Conclusions: These results suggest that pulmonary function is a long-term predictor for overall survival rates in both genders and could be used as a tool in general health assessment.
Key Words: cohort study • FEV1 • ischemic heart disease • lung function • mortality
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Despite numerous reports, little is known about whether or not there is a causal relationship between impaired pulmonary function and increased mortality. A recognized problem with the interpretation of this association is that study participants at baseline may have undiagnosed or asymptomatic disease that is associated with both lower levels of pulmonary function and short-term mortality risk.1 8 In fact, it has been suggested that pulmonary function is a proxy for other disease conditions that lead to mortality and, therefore, that pulmonary function is not causally related to mortality. In an effort to overcome this concern, Hole et al3 stated that the exclusion of participants who died during the first 5 or 10 years after follow-up did not significantly alter results, but periods of exclusion of > 10 years have not been examined until now. Persistent elevated risk levels after the exclusion of participants who die earlier during follow-up would suggest a direct effect.
The purpose of this study was to investigate the association between pulmonary function and mortality for periods of > 25 years of follow-up and, in particular, to examine this relationship for follow-up of > 15 years in women. We also sought to determine the time span over which pulmonary function remains a significant predictor of mortality by sequential survival analysis in participants who had minimum survival times of 5, 10, 15, 20, and 25 years after enrollment in the study.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
A total of 369 participants (16.2%) were African American or other minorities and are not included in this analysis because of incomplete follow-up for this group and limitations in establishing reference equations for pulmonary function in race-gender subgroups from a relatively small sample. Of the remaining 1,904 participants, 156 (8.2%) had insufficient data available to initiate follow-up. We excluded another 254 participants (13.3%) because of their refusal to perform pulmonary function tests, an inability to perform the forced expiration maneuver, or a lack of spirometry curves to confirm the pulmonary function test data retrieved from magnetic tapes. Vital status was determined beginning in 1990; participants were followed up until their death or until the end of the study. The primary outcome of this analysis was vital status and cause of death, and no longitudinal measurements of covariates were performed.
Pulmonary Function Testing
Pulmonary function measurements were obtained with a spirometer
(Vitalor; McKesson; Toledo, OH) adapted for field use. FVC and
FEV1 were recorded for at least three FVC
maneuvers, as described.13
Each of the test records was
evaluated, and the FEV1 was calculated by a
physician with experience in pulmonary function testing using back
extrapolation. The best FEV1 test result was used
for analysis according to the current American Thoracic Society
standardized protocol.16
Of the 1,494 participants with
complete pulmonary function test data, 21 had unsatisfactory tracings
of the spirometry curves, which exhibited insufficient expiratory times
to calculate the FEV1 (1.4%). We excluded 145
participants (9.7%) who were < 20 years of age at baseline, because
pulmonary function in this age group shows variation owing to
differences among participants in growth and because of the possibility
that participants < 20 years would not have reached their adult
height for body mass index (BMI) calculation.17
Baseline
data (eg, weight, height, and smoking status) were
incomplete for 44 participants (2.9%), and vital status at any time
during follow-up could not be confirmed in 69 women (4.6%) and 20 men
(1.3%). The remaining 1,195 subjects, 641 women and 554 men, were
included in the current analysis. Raw FEV1 values
were multiplied by 1.10 to correct for body temperature pressure
saturation (BTPS) for better comparison with other published data, as
previously described by Sorlie et al.18
A constant
correction factor for BTPS was used because no measurements of internal
spirometer or room temperature were available. The use of this
adjustment, is likely to introduce random error in the determination of
FEV1, but it is unlikely to cause
bias.18
FEV1 relative to the
predicted value was used to estimate pulmonary function impairment.
Predicted values of FEV1 were calculated from
linear regression on age and height. The following equations were
obtained,
 |
 |
Interview
Information on anthropometric, demographic, and lifestyle
variables was obtained at baseline by an interviewer-administered
questionnaire, and systolic BP (SBP) and diastolic BP were measured
three times in a standardized manner, as previously described in
detail.15
BMI was calculated from self-reported weight
divided by height squared.
Follow-up and Outcome Measures
Participants were enrolled from June 1960 through December 1961
and were followed up for an average length of 29 years until the end of
the follow-up on December 31, 1989. Baseline data and measurements were
obtained at enrollment, and participants were followed up until their
death or the end of the study, respectively. Details of the follow-up
and ascertainment of vital status have been described by Dorn et
al.15
Briefly, vital status was determined with
computer-based searches of the New York State Department Health Vital
Records Death Registry, the Cancer Tumor Registry, the Department of
Motor Vehicle records of drivers’ licenses and automobile
registrations, the US Social Security Administration Death Master
Files, and manual searches of the telephone and Polk directories of the
city of Buffalo, including its suburbs. We also attempted to contact
the participant’s last employer, neighbor, church, or other contact
address if recorded in the original questionnaire. Living participants
were followed up either through direct telephone contact or mail
correspondence or through contact with relatives or other immediate
contact persons (eg, nursing home personnel). Rigid matching
criteria (exact spelling of the name and exact date of birth) were used
and were manually reviewed to determine the death of participants from
death certificates. The primary aim of this study was to assess
mortality from all causes. The date and underlying cause of death were
obtained from the death certificate. A trained nosologist upgraded
Ninth International Classification of Diseases codes, as
previously described.15
All recorded deaths during the
follow-up period from the day of the interview until the end of 1989
were taken into account for the current analysis. Ninth
International Classification of Diseases codes 410 through 414
were classified as IHD (acute myocardial infarction, 410; other acute
and subacute forms of IHD, 411 [eg, postmyocardial
infarction syndrome, 411.0; intermediate coronary syndrome, 411.1];
old myocardial infarction, 412; angina pectoris, 413; other forms of
chronic IHD, 414), codes 390 through 459 were classified as
cardiovascular, codes 140 through 209 were classified as cancer, and
codes 460 through 520 were classified as respiratory deaths.
Statistical Analysis
Student’s t test and 
2
analysis were utilized to compare the included participants to those
persons excluded or unavailable for follow-up. We used the Cox
proportional hazards model20
to calculate the hazard
ratios associated with pulmonary function status
(FEV1%pred or QFEV1%pred)
and mortality rate, with adjustment for known risk factors. Age,
education, smoking status (never, former, or current smoker), SBP, and
BMI were utilized as covariates because other studies showed these to
be predictors of mortality. In the models exploring the subgroups with
longer survival times, these covariates were included independently of
the level of statistical significance. No statistically significant
interactions were observed. Never-smokers were defined as those who
smoked < 100 cigarettes in their lifetime. Hazard ratios were
calculated utilizing FEV1%pred as continuous
variable or in quintiles (QFEV1%pred) with the
highest QFEV1%pred serving as the reference
category. Survival analysis was repeated after the exclusion of
participants who survived < 5, 10, 15, 20, and 25 years. Computer
software (SPSS; SPSS Inc; Chicago, IL)21
was used for the
analyses. Statistical significance was considered for p values < 0.05
(two-tailed), and 95% confidence intervals (CI) were computed around
the risk-point estimates.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Baseline characteristics of study participants included in the analysis are shown separately for men and women in Table 1 . The age ranges were 20 to 89 years in men and 20 to 84 years in women. On average, men were slightly older and taller than women. BMI and SBP were higher in men. Overall smoking prevalence was high, and men were more likely to be current or ex-smokers than women. The number of years of education was similar in both sexes. While absolute FEV1 was higher in men, FEV1%pred was lower in men than in women.
|
Table 2 shows the vital status and underlying cause of death for all participants included in the present study. During the entire follow-up period, 54.5% of the men and 43.4% of the women died. Cardiovascular disease and IHD, in particular, represented the predominant causes of death in both genders. Respiratory disease was found to be the underlying cause of death in 9.6% of the deceased men but in only 3.6% of the deceased women.
|
|
25 years after
enrollment. In women, FEV1%pred was a
statistically significant predictor in all but one survival group
(survival for > 20 years; p = 0.053). The analysis of
QFEV1%pred indicated that for the 29-year
follow-up period as a whole compared to the highest quintile, the
relative hazard ratio for those in the lowest quintile was 2.24 for men
and 1.81 for women. Increased risks also were observed for participants
with moderately lower levels of FEV1 (in the
second quintile). Similar results were observed for men when
participants who survived for < 5 or 10 years were excluded, but
hazard ratios declined and did not reach statistical significance after
the exclusion of participants with survival times of < 15 years. In
women, the risk for all-cause mortality for the first quintile of
FEV1%pred was significantly elevated in all
survival groups except after the exclusion of participants who survived
for > 20 years.
|
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
The results of this study confirm previous reports that pulmonary function is an independent risk factor for overall all-cause mortality and IHD mortality and suggest that this risk is evident for longer periods than those studied to date.1 3 5 8 9 18 22 23 Previous studies with long follow-up periods are limited by the characteristics of the study population. Both the study by Beaty et al,8 which examined a follow-up period of 24 years in volunteers, and the study by Weiss et al,4 which examined a follow-up period averaging 25 years, are restricted to men. Other studies that included women had significantly shorter follow-up periods.2 3 6 7 9 18 No previous study has reported an association of lower levels of pulmonary function with all-cause mortality for a follow-up period of 29 years. Our risk estimates for FEV1%pred and all-cause mortality for both men and women are similar to those in other studies.3 4 9 10 As a continuous variable, FEV1%pred was a significant predictor of IHD mortality in both men and women. The risk estimates for death from IHD in the lowest QFEV1%pred of 2.11 for men and 1.96 in women are in agreement with the study by Hole et al,3 which examined participants in terms of QFEV1. It is important to note that elevated hazard ratios were observed also for participants with lower levels of lung function in the second quintile. This observation suggests that the increased risk is not confined to a small fraction of the population with severely reduced pulmonary function.
There has been speculation about the underlying mechanism that could explain the predictive properties of pulmonary function with regard to mortality. The initial concerns that FEV1%pred may be a proxy for smoking status have been solved by reports showing that this association is independent of smoking status and that it is present in never-smokers.3 6 7 The number of never-smoking men (n = 57) in our study was too small for separate analysis. In the group of never-smoking women as a whole (n = 288), FEV1%pred as a continuous variable was inversely related to all-cause mortality with a hazard ratio of 0.993 (95% CI, 0.985 to 1.000; p = 0.053) after adjustment for other risk factors. Compared to the highest quintile in this subgroup, the all-cause mortality risk in the lowest QFEV1%pred was 1.58 (95% CI, 0.94 to 2.67; results not shown), which is similar to the results of Hole et al.3
It is not clear whether the observed association reflects a cause-effect relationship of reduced pulmonary function with mortality. The lung is a primary defense organ against environmental toxins, and impaired pulmonary function could lead to decreased tolerance against these environmental toxins. This hypothesis has been described in detail by Cohen,24 who speculated that impaired pulmonary function contributes to a variety of disease processes that ultimately lead to disease and death. On the other hand, factors may be involved that affect both pulmonary function and mortality. Weiss et al4 speculated that FEV1 is an indicator of general health influenced by environmental toxic exposure and is, therefore, related to survival. FEV1 levels could affect physical activity, which may prolong survival times through its influences on metabolism with decreased IHD mortality. These authors also hypothesized that oxidants, which influence FEV1 and health status, might be responsible for the observed relationship.4 In fact, oxidants have been shown to play a role in the etiology of various diseases, including IHD.25 In 1997, we showed in a cross-sectional study26 that indicators of oxidative stress are negatively correlated with FEV1%pred. Several other authors have reported positive correlations between antioxidant vitamins and respiratory function,27 28 29 30 31 32 but reduced pulmonary function also could be the underlying factor responsible for increased oxidative stress.26 Further research is needed to investigate the hypothesis that oxidative stress is related to both pulmonary function and mortality.
In female participants in this study, pulmonary function was a predictor of all-cause mortality for a period of > 25 years, while in male participants, pulmonary function lost its predictive value after 20 years from baseline. The underlying reason for the observed difference between men and women needs to be explained. Men and women compared well with respect to baseline characteristics in the different survival groups except for smoking status. A change in lifestyle factors (eg, smoking cessation) after baseline measures were taken may have led to a shift between quintiles of FEV1%pred. This shift, in turn, could have influenced the relative hazard associated with FEV1%pred. We may have observed a decline in risk in men that was influenced by smoking but was not controlled for adequately in the analysis by utilizing smoking status alone. The latter is less likely, because when we included pack-years as a summary measure of smoking exposure in the analysis, the results did not change significantly. Finally, men die earlier than women, and the lack of statistical power in men after > 25 years of survival may be the most likely explanation for the observed difference, since only 44 men were classified into the quintile with the lowest FEV1%pred.
In our study, as in other studies,3 we included all participants independent of disease status at baseline. The presence of disease at baseline could have been related to both pulmonary function and mortality. However, we controlled for this bias by excluding those persons who died early in the sequential analysis. An advantage of this approach is that it also decreases the bias of including participants with undiagnosed disease, which could be responsible for an observed relationship in other studies in which only those persons with diagnosed disease at baseline were excluded.
For the proper analysis of pulmonary function data, the use of an adequate prediction equation and adjustment for known covariates is important. Neas and Schwartz10 reported that the choice of an internal vs an external prediction equation for FEV1 did not alter the observed results. We performed our analysis with an internal linear equation using age and height as predictors, because models that included nonlinear terms did not improve prediction. Given the relatively wide age range of participants and the association of reduced pulmonary function with mortality, older participants in this study could represent healthy survivors selected for superior pulmonary function. This selection could explain the observed linear relationship of FEV1 with age. Additionally, the equation in men was derived from only 57 never-smokers. For these reasons, we repeated the analysis using external linear and nonlinear prediction equations. Our results did not change significantly when we used external linear or nonlinear equations, such as the ones used by Morris et al33 or Dockery et al,34 to predict FEV1. It seems, therefore, that the observed results were not influenced by the choice of an internal prediction equation.
The present study has a number of limitations. Participant characteristics, except for vital status and underlying cause of death, were obtained at baseline only. Several variables (eg, smoking) could have changed during the follow-up period, as described above. Since we do not have this information, we are unable to determine how changes in the independent variables would affect survival. Our analysis was adjusted for significant risk factors, such as BMI, SBP, gender, education, and smoking status; however, no information was available on another important IHD risk factor, serum cholesterol level. Cholesterol levels at baseline have been shown to be a significant risk factor for IHD mortality in prospective studies.35 36 Our criteria for the determination of vital status were rigid, and overall follow-up for vital status was high. A general problem in cohort studies is the exclusion of participants and the unavailability for follow-up. While the lack of baseline data, including missing spirometry curves, is less likely to be associated with pulmonary function status, we excluded a small percentage of participants because they were unable to perform or refused to perform spirometry. Although these participants did not differ significantly in baseline characteristics from those with available pulmonary function data, it has been shown that the refusal or inability to perform spirometry is a stronger risk factor of mortality than low FEV1.9 However, subjects who are unable to perform or refuse to perform spirometry would also be likely to have lower FEV1 levels and would fall in the lower percentiles of the distribution. An exclusion of these subjects from the lower quintiles, despite their elevated mortality risk, could lead to an underestimation of the overall risk associated with poor pulmonary function. We were unable to follow-up on 69 women and 20 men who had sufficient data to be included at baseline. These participants did not significantly differ in baseline characteristics from those who were included. If the majority of these subjects died shortly after the baseline information was obtained and if their early death was the reason for the unavailability for follow-up, then our risk estimates would represent overestimates of the true risk because they were similar to included participants in baseline characteristics. Overall, the number of subjects with pulmonary function test information who were unavailable for follow-up is small. Therefore, the results are unlikely to be biased severely by the loss to follow-up. There is also the potential for inaccurate coding on death certificates. Such misclassification probably affected the mortality attributed to IHD, but a relationship of coding inaccuracy to baseline lung function is highly unlikely. Therefore, the finding that reduced pulmonary function is related to IHD mortality is less likely to be affected by coding errors on death certificates.
|
Conclusion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
|
Acknowledgements |
|---|
|
Footnotes |
|---|
This research was supported by the National Heart, Lung, and Blood Institute (grant No. HL5487402) and Deutsche Forschungsgemeinschaft (grant No. SCHU 1056–1/1).
Received for publication September 14, 1999. Accepted for publication March 13, 2000.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. K. Han, V. V. McLaughlin, G. J. Criner, and F. J. Martinez Pulmonary Diseases and the Heart Circulation, December 18, 2007; 116(25): 2992 - 3005. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Alexeeff, A. A. Litonjua, H. Suh, D. Sparrow, P. S. Vokonas, and J. Schwartz Ozone Exposure and Lung Function: Effect Modified by Obesity and Airways Hyperresponsiveness in the VA Normative Aging Study Chest, December 1, 2007; 132(6): 1890 - 1897. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Aronson, I. Roterman, M. Yigla, A. Kerner, O. Avizohar, R. Sella, P. Bartha, Y. Levy, and W. Markiewicz Inverse Association between Pulmonary Function and C-Reactive Protein in Apparently Healthy Subjects Am. J. Respir. Crit. Care Med., September 15, 2006; 174(6): 626 - 632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Hollenberg, J. Yang, T. J. Haight, and I. B. Tager Longitudinal changes in aerobic capacity: implications for concepts of aging. J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2006; 61(8): 851 - 858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G.B. J. Mancini, M. Etminan, B. Zhang, L. E. Levesque, J. M. FitzGerald, and J. M. Brophy Reduction of Morbidity and Mortality by Statins, Angiotensin-Converting Enzyme Inhibitors, and Angiotensin Receptor Blockers in Patients With Chronic Obstructive Pulmonary Disease J. Am. Coll. Cardiol., June 20, 2006; 47(12): 2554 - 2560. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M. Ochs-Balcom, B. J.B. Grant, P. Muti, C. T. Sempos, J. L. Freudenheim, M. Trevisan, P. A. Cassano, L. Iacoviello, and H. J. Schunemann Pulmonary function and abdominal adiposity in the general population. Chest, April 1, 2006; 129(4): 853 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lindberg, B. Eriksson, L.-G. Larsson, E. Ronmark, T. Sandstrom, and B. Lundback Seven-Year Cumulative Incidence of COPD in an Age-Stratified General Population Sample. Chest, April 1, 2006; 129(4): 879 - 885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Guenegou, B Leynaert, I Pin, G Le Moel, M Zureik, and F Neukirch Serum carotenoids, vitamins A and E, and 8 year lung function decline in a general population. Thorax, April 1, 2006; 61(4): 320 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V M Pinto-Plata, H Mullerova, J F Toso, M Feudjo-Tepie, J B Soriano, R S Vessey, and B R Celli C-reactive protein in patients with COPD, control smokers and non-smokers Thorax, January 1, 2006; 61(1): 23 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Shaukat, J. L. Freudenheim, B. J.B. Grant, P. Muti, H. M. Ochs-Balcom, S. E. McCann, M. Trevisan, L. Iacoviello, and H. J. Schunemann Is Being Breastfed as an Infant Associated with Adult Pulmonary Function? J. Am. Coll. Nutr., October 1, 2005; 24(5): 327 - 333. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D A Lawlor, S Ebrahim, and G Davey Smith Association of birth weight with adult lung function: findings from the British Women's Heart and Health Study and a meta-analysis Thorax, October 1, 2005; 60(10): 851 - 858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. S. von Ehrenstein, D. N. G. Mazumder, Y. Yuan, S. Samanta, J. Balmes, A. Sil, N. Ghosh, M. Hira-Smith, R. Haque, R. Purushothamam, et al. Decrements in Lung Function Related to Arsenic in Drinking Water in West Bengal, India Am. J. Epidemiol., September 15, 2005; 162(6): 533 - 541. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. K. Myint, R. N. Luben, P. G. Surtees, N. W. J. Wainwright, A. A. Welch, S. A. Bingham, N. J. Wareham, N. E. Day, and K-T. Khaw Respiratory function and self-reported functional health: EPIC-Norfolk population study Eur. Respir. J., September 1, 2005; 26(3): 494 - 502. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hoogendoorn, M. P. M. H. Rutten-van Molken, R. T. Hoogenveen, M. L. L. van Genugten, A. S. Buist, E. F. M. Wouters, and T. L. Feenstra A dynamic population model of disease progression in COPD Eur. Respir. J., August 1, 2005; 26(2): 223 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J G Schanen, C Iribarren, E Shahar, N M Punjabi, S S Rich, P D Sorlie, and A R Folsom Asthma and incident cardiovascular disease: the Atherosclerosis Risk in Communities Study Thorax, August 1, 2005; 60(8): 633 - 638. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Stavem, E. Aaser, L. Sandvik, J. V. Bjornholt, G. Erikssen, E. Thaulow, and J. Erikssen Lung function, smoking and mortality in a 26-year follow-up of healthy middle-aged males Eur. Respir. J., April 1, 2005; 25(4): 618 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Sin and S. F. P. Man Chronic Obstructive Pulmonary Disease as a Risk Factor for Cardiovascular Morbidity and Mortality Proceedings of the ATS, April 1, 2005; 2(1): 8 - 11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Q. Gan, S.F. P. Man, and D. D. Sin The Interactions Between Cigarette Smoking and Reduced Lung Function on Systemic Inflammation Chest, February 1, 2005; 127(2): 558 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. Gauderman, E. Avol, F. Gilliland, H. Vora, D. Thomas, K. Berhane, R. McConnell, N. Kuenzli, F. Lurmann, E. Rappaport, et al. The Effect of Air Pollution on Lung Development from 10 to 18 Years of Age N. Engl. J. Med., September 9, 2004; 351(11): 1057 - 1067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Shohaimi, A. Welch, S. Bingham, R. Luben, N. Day, N. Wareham, and K-T. Khaw Area deprivation predicts lung function independently of education and social class Eur. Respir. J., July 1, 2004; 24(1): 157 - 161. [Abstract] [Full Text] |
