Prevalence, risk factors and outcome of chronic obstructive pulmonary disease (COPD)

Thesis

Reference

Prevalence, risk factors and outcome of chronic obstructive pulmonary disease (COPD)

BRIDEVAUX, Pierre-Olivier

Abstract

This thesis summarizes some aspects of COPD epidemiology in 2011. Prevalence of COPD is increasing worldwide. Even mild COPD is associated with worse outcomes when present in smokers or symptomatic subjects. In Switzerland, prevalence of moderate or severe COPD is lower than elsewhere. Beyond smoking, which is decreasing in developed countries, novel risk factors have been documented. In developing countries, biomass exposure probably constitutes the most important risk factor for COPD in women. In developed countries, the role of environmental risk factors such as air pollution or occupational exposure is to be considered. But, air pollution slowly decreased and heavy industries and mining activities are outsourced in developing countries. Risk factors such as asthma or bronchial hyper-responsivness do play a role. Both may be considered as the results of interaction between genes and environment.

BRIDEVAUX, Pierre-Olivier. Prevalence, risk factors and outcome of chronic obstructive pulmonary disease (COPD). Thèse de privat-docent : Univ. Genève, 2011

DOI : 10.13097/archive-ouverte/unige:17511

P REVALENCE OF C HRONIC O BSTRUCTIVE P ULMONARY

D ISEASE (COPD) IN S WITZERLAND AND IN E UROPE

Privat Docent Thesis University of Geneva

2011

presented by

Dr Pierre-Olivier Bridevaux, MD, MSc

“Our knowledge can only be finite, while our ignorance must necessarily be infinite”

T ABLE OF CONTENTS

INTRODUCTION ... 4

HISTORICAL BACKGROUND... 6

THE CIBA GUEST SYMPOSIUM (1959) ... 6

THE NATURAL HISTORY OF CHRONIC BRONCHITIS AND EMPHYSEMA (1976) ... 7

The American Thoracic Society (ATS) Statements for the diagnosis and care of patients with COPD (1995) ...10

THE GLOBAL INITIATIVE FOR OBSTRUCTIVE LUNG DISEASE (GOLD)(2001) ...11

A SPIROMETRIC DEFINITION OF COPD.THE GOLD CRITERION ...11

IMPACT OF THE GOLD GUIDELINE ...13

THE GOLDCONTROVERSY ...14

THE OPEN LETTER TO THE MEMBERS OF THE GOLD COMMITTEE (AUGUST 26,2010) ...15

GLOBAL EPIDEMIOLOGY OF COPD ...17

COPD PREVALENCE IN 2004(META-ANALYSIS BY HALBERT,2006) ...18

COPD PREVALENCE IN 2007(THE BOLD STUDY) ...21

PREVALENCE OF COPD IN SWITZERLAND ...25

PREVALENCE OF COPD IN SWITZERLAND,METHODOLOGY ...25

RESULTS OF THE SAPALDIA STUDY ON PREVALENCE OF COPD IN SWITZERLAND ...31

PREVALENCE OF COPD/ AIRFLOW OBSTRUCTION ...31

PHYSICIAN DIAGNOSED ASTHMA PREVALENCE ...33

RISK FACTORS FOR COPD ...35

COPD IN NEVER SMOKERS ...42

COMMENTS ON THE PREVALENCE OF COPD IN SWITZERLAND ...47

INTERNATIONAL COMPARISONS ...47

COPD IN NEVER SMOKERS IN SWITZERLAND...50

GLOBAL TRENDS IN SMOKING PREVALENCE & COPD IN NEVER-SMOKERS ...53

IMPORTANCE OF NON-SMOKING RELATED COPD ...54

NOVEL RISK FACTORS OF COPD ...58

ENVIRONMENTAL TOBACCO SMOKE ...58

ASTHMA ...60

BRONCHIAL HYPER-RESPONSIVENESS ...62

AIR POLLUTION ...62

GENETIC FACTORS ...68

INFECTIONS ...69

OCCUPATIONAL EXPOSURE ...71

OBESITY, INFLAMMATION AND COPD ...74

THE ROLE OF HIGH SENSITIVITY C-REACTIVE PROTEIN (HS-CRP) ...75

COPD AS A CARDIOVASCULAR RISK FACTOR ...75

GENDER DIFFERENCES AND COPD ...77

SEX-SPECIFIC EFFECT OF BODY WEIGHT GAIN ON SYSTEMIC INFLAMMATION IN SUBJECTS WITH COPD ...78

COMMENTS REGARDING THE EFFECT OF WEIGHT CHANGE &COPD ON SYSTEMIC INFLAMMATION ...88

SIGNIFICANCE OF MILD FORM OF COPD ...91

LONG TERM IMPACT IN LUNG FUNCTION DECLINE, UTILISATION OF CARE AND QUALITY OF LIFE IN MILD COPD ...91

COMMENTS ... 101

FURTHER COMMENTS: FIXED RATIO VERSUS LLN ... 104

CONCLUSIONS ... 109

INTRODUCTION

HRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD) has been characterized by the slow deterioration of the lung function with little or no reversibility to

bronchodilators and little response to long-term pharmacological therapy [1, 2].

COPD is a major cause of morbidity and mortality and the only chronic condition for which figures show an upward trend. COPD is projected to become the third leading cause of death by 2020[3]. (Figure on page 4)

Figure: Percent change in age-adjusted death rate (USA) 1965-1998 (source: NHLBI, NIH)[1]

Legend : Compared to 1965, death rate for all conditions decreased except for COPD. COPD related death rate increased by 163%.

The classical but caricatural description of COPD as a chronic, debilitating disease with few if any therapeutic options to offer led generations of physicians to have a nihilistic approach toward COPD.

Up to now, the only therapies reducing mortality are smoking cessation and supplemental oxygen therapy in selected cases. Lung transplantation, the most aggressive therapy for end-

C

of lung transplantation in COPD. For instance, we showed that more than half of the patients with end-stage COPD lived longer than predicted by the BODE index (Body Mass Index, Obsruction, Dyspnea, Exercise capacity). However, this overall survival advantage was hampered by short-term post-operative mortality. (see appendix for full text)[4, 5].

Nevertheless, during the last 15 years, not only have researchers provided a better

understanding of the factors leading the COPD but therapeutic advances are now available for most patients with COPD. This is reflected by the regularly growing number of COPD

citations in Pubmed-medline over the last 40 years. In 1970, there were 4 articles citing

―COPD‖ compared to 2248 in 2009, outpacing the citation growth for ―asthma‖.(Figure on page 5).

Figure: Citations for “COPD” and “asthma” 1970-2009

In this thesis, recent research published in the field of COPD epidemiology will be presented.

Special attention will paid for COPD in never smokers and the novel risk factors possibly causing the conditions.

H

ISTORICAL BACKGROUND

T

HE

CIBA

GUEST SYMPOSIUM

(1959)

N THE LATE

50

^TH,THE

CIBA

GUEST SYMPOSIUM published definition of

emphysema, chronic bronchitis and asthma[6]. It was recognized that the terms were interchanged and freely used to describe the same clinical condition of the same patient.

Physicians in USA often used the term ―asthma‖ or ―emphysema‖ whereas those in England would call the same condition as ―chronic bronchitis‖.

The symposium was held in order to encourage people to use defined terms in making pathological, clinical and functional assessments, to investigate the reproducibility of

assessments in the hands of the same and different observers. The symposium was chaired by Dr C.M. Fletcher, who later published landmark article about the natural history of COPD[2].

The terms were defined as follows:

1) Emphysema: An (anatomical) condition of the lung characterized by increase beyond the normal in the size of air spaces distal to the terminal bronchiole either from dilatation or from destruction of their walls.

2) Chronic non-specific lung disease: The term was coined not to be used in clinical practice but to encompass the three different form of obstructive lung disease.

a. Chronic or recurrent excessive secretion of bronchial mucus b. Intermittent obstruction to

bronchial airflow

c. Persistent obstruction to bronchial airflow

These three different forms correlate with the conditions currently labelled as

―chronic bronchitis‖ ―asthma‖ and

―COPD‖. The symposium published a table to summarize its view.

The term ―generalized obstructive lung disease‖

I

lump subjects with ―asthma‖ (with intermittent or changing obstruction) and those

―irreversible or persistent obstructive lung disease‖ in the same clinical category.

The symposium report deserves comments. The term ―generalized obstructive lung disease‖

was not accepted or used in USA, where the term ―Chronic Obstructive Pulmonary Disease‖

was preferred. Greater influence of North-American journals and research contributed to popularize the term ―COPD‖. The term ―generalized obstructive lung disease‖ was soon forgotten. Retrospectively, this odd term can be view as an attempt from the symposium to lump both chronic bronchitis, asthma and emphysema in a same clinical category. This unifying theory explaining a final common pathway for diverse chronic respiratory condition (asthma, emphysema and chronic bronchitis) blossomed in the late eighties when the so- called ―Dutch hypothesis‖ was proposed to provide a common inflammatory pathogenesis for asthma and COPD [7, 8].

T

HE

N

ATURAL

H

ISTORY OF

C

HRONIC

B

RONCHITIS AND

E

MPHYSEMA

(1976)

Published by Charles Fletcher, Richard Peto et al, this book concluded that although other factors such as air pollution, social status or infections during infancy play a role, smoking was the most important factor contributing to the development of emphysema and chronic bronchitis. The book contained a famous figure (page 8) summarizing the natural decline of lung function in male never smokers and smokers.

Charles Fletcher Born, 1911, died 1996, professor of clinical epidemiology, London

Richard Peto, professor of epidemiology, Oxford. Richard Peto was knighted in 1999

Figure: The natural history of airflow obstruction revisited. Source AJRCCM 2009: [9]

Mean FEV1 values (expressed as percent of its value at the age of 25) by age in smokers who quit smoking before the age of 30 (Q , 30), between 30 and 40 years of age (Q30–40) and after the age of 40 (Q401). Curves from healthy never smokers (NS) and continuous smokers (CS) are also included for comparison. (A) Results in males. (B) Results in females. The mean FEV1

decline value (and 95% confidence intervals) for each curve of quitters for males was 15.5 ml (11.3–19.8) in quitters before the age of 30, 24.0 ml (20.0–28.1) in quitters between 30 and 40 years of age, and 28.9ml (26.4–31.1) in quitters after the age of 40; for females, 10.4 ml (6.3–14.5) in quitters before the age of 30, 16.5 ml (14.0–19.0) in quitters between 30 and 40 years of age, and 21.0 ml (18.8–23.2) in quitters after the age of 40. *P , 0.05 versus healthy never-smokers (NS). #P , 0.05 versus CS.

Source: Kohensal at al [9]

susceptible to tobacco smoke [2, 10]. Interestingly, the figure was built for teaching purpose without raw data. It was only in 2009, that the figure was recreated, both for men and women, using the Framingham offspring cohort [9] (Figure on page 9). The Framingham offspring cohort showed that lung function changes from adolescence to old age differ between sexes but smoking has a similar deleterious effect. Stopping early was better, thus validating

at this time: ―until [this] self-inflicted disease can be charged to patients, the National Health Service will be responsible for the treatment of chronic bronchitis and emphysema‖. The sentence suggests that smoking is a personal decision which is not influenced by factors such as nicotine dependence or social determinants. In 2010, clinical research and medical

sociology showed the importance of these factors. As a result, COPD cares are covered by health insurances. Nevertheless, nicotine dependence costs are still charged to patients.

The American Thoracic Society (ATS) Statements for the diagnosis and care of patients with COPD (1995)

In 1995, the ATS statement acknowledged the widespread use of the term COPD, which was adopted over other terms such as ―generalized obstructive lung disease‖ or ―chronic obstructive lung disease (COLD)‖[11]. This consensus can be considered as a landmark for researchers and clinicians in the field. The term COPD was defined as ―a disease characterized by the presence of airflow obstruction due to chronic bronchitis or emphysema;

the airflow obstruction is generally progressive, may be accompanied by airway hyperreactivity, and may be partially reversible‖. Since 1995, the COPD definition evolved marginally. Mainly, the GOLD consensus later added the terms ―avoidable‖ and ―treatable‖.

The ATS also presented for the first time the Venn diagram depicting the relations of asthma, emphysema and chronic bronchitis (Figure on page 11).

Figure: The Venn Diagram (source: ATS documents 1995 [11] )

Legend: The shaded area corresponds to COPD definition, acknowledging the different respiratory conditions leading to COPD.

The Venn diagram contributed not only to the widespread opinion that asthma causes COPD (the causal association is scientifically sound), but also allowed the pharmaceutical groups to heavily promote long-term use of inhaled steroids as an efficient therapeutic option to prevent lung function decline or even COPD related mortality. Unfortunately, in large randomized trials of patients with COPD, neither FEV1 decline nor mortality were reduced by long-term use of inhaled steroids.[12, 13].

T

HE

G

LOBAL

I

NITIATIVE FOR

O

BSTRUCTIVE

L

UNG

D

ISEASE

(GOLD) (2001)

In 2001, the first GOLD guideline was issued from a panel of experts in respiratory medicine, epidemiology, socio-economics and health education[14]. Based on available published evidence and sponsored by the pharmaceutical industry, the US National Heart, Lung and Blood Institute (NHLBI) and the World health Organization (WHO), the guideline provided definition of COPD for clinicians and researchers [14].

A SPIROMETRIC DEFINITION OF COPD.THE GOLD CRITERION

GOLD defined COPD as a disease state characterized by airflow limitations that is not fully

Figure: A 70 year old male with a history of smoking. According the 2010 GOLD Classification, this patient belong s to the GOLD IV category (GOLD 2001 would classify him as GOLD III)

Legend : Typical post bronchodilator spirometry of a patient with very severe COPD. The Forced Vital Capacity (FVC = [CVF, capacité vitale forcée]) is relatively normal (74%

predicted) and the FEV1 (Forced Expiratory Volume in 1 sec) = [VEMS, volume expiré maximal en 1 sec] is abnormal (23% predicted). The FEV1/FVC ratio is severely decreased (33 % predicted)

Figure: A man performing spirometry

Table: The first GOLD classification. Source: GOLD summary 2001 [14]

In 2001, The GOLD guideline also described the stage 0 ―at risk‖ category to encompass subjects with chronic respiratory symptoms (mostly smokers) and normal spirometry. The ―at risk‖ category was postulated as an early form of COPD which will progress to more severe forms (Stage 1+). It should be noted that the ―at risk‖ category dropped from subsequent updates of the GOLD consensus, because of doubt regarding the long term outcomes of patient with normal spirometry. In our own SAPALDIA study, subjects with respiratory symptoms and normal baseline spirometry exhibited larger lung function decline compared to those without respiratory symptoms. This suggests that the ―at risk‖ category is indeed

associated with worse long-term outcome.[15]

IMPACT OF THE GOLD ^GUIDELINE

The GOLD guideline states that respiratory symptoms or the mere exposure to noxious environmental factors were requested to screen for COPD using spirometry with the fixed ratio (FEV1/FVC <0.7). Given the high prevalence of respiratory symptoms (chronic cough

reached. The so-called fixed ratio was advocated because of simplicity of use. The purpose of this decision was to encourage spirometry performance by general practitioners or

pharmacists, acknowledging that COPD is largely underdiagnosed in the community.

The choice of the fixed ―clinical‖ FEV1/FVC ratio instead of the lower limit of normal value of FEV1/FVC caused inflamed controversy [16-18].

T

HE

GOLD C

ONTROVERSY

GOLD arbitrarily defined COPD on spirometric criteria that have been argued to be not based on sound scientific evidence. GOLD states that a FEV1/FVC <0.7 after inhalation of

bronchodilators confirms the presence of airflow limitation that is not fully reversible independently of age. This simplistic definition of airflow obstruction is not statistically sound. Because FEV1/FVC ratio naturally declines with increasing age, percentage of older subjects falsely classified as ―COPD‖ increases in parallel. The figure on page 15, based on 1649 spirometries performed in Switzerland for the Study on Air Pollution and Lung Disease in Adults (SAPALDIA) show a plot of FEV1/FVC ratio in non-smokers without respiratory symptoms versus age [19]. It shows that about half of subjects aged 60 or more have

―abnormal‖ lung function according to the faulty fixed FEV1/FVC ratio.

Figure: FEV1/FVC ratio of 1649 healthy, non smokers from the SAPALDIA study Source: SAPALDIA dataset, personal analysis

.5 .55 .6 .65 .7 .75 .8 .85 .9 .95 1

FEV1/FVC ratio

30 40 50 60 70

Age

Legend: The fixed FEV1/FVC ratio potentially misclassifies as having COPD subjects below the red line who are never smokers and free of symptoms. Risk of misclassification (over diagnosis) is greatest for older subjects. Data from 1649 subjects from the SAPALDIA cohort (2001-2003).

THE OPEN LETTER TO THE MEMBERS OF THE GOLD COMMITTEE (AUGUST 26,2010)

In 2010, before the 9th revision of the GOLD guideline, the Pulmonaria group and individuals drafted a letter to the GOLD Committee to welcome the effort to stimulate interest and

awareness of the high prevalence of COPD but also to criticize the fixed GOLD criterion defining obstruction. The researchers and clinicians (including the author of the present thesis) who signed the letter pleads for revision arguing that overdiagnosis of obstruction will lead to the inclusion of subjects who do not have COPD in clinical trials and epidemiological studies, thus adding noise to any signals and as a consequence limiting our ability to find new causes of COPD and finding new treatments.

Thus, when using the fixed ratio, the risk of misclassification of COPD is greatest for older subjects with normal or near normal FEV1, commonly labeled as ―GOLD stage 1 COPD‖ in the current literature. Because many population studies on the prevalence of COPD rely solely on airflow obstruction as measured by spirometry without requirement for reporting

symptom, exposure to noxious particles or fumes, we decided to study the long term impact of stage 1 COPD. We postulated that the long term impact of COPD diagnosed at some point in time, would vary according to the presence of respiratory symptoms. The study ―Long term impact in lung function decline, utilisation of care and quality of life in GOLD stage 1

COPD‖ was designed to test this hypothesis. [15] Methods and results of this study are detailed below (page 98)

Pathological image of emphysematous lung

GLOBAL EPIDEMIOLOGY OF COPD

NTIL RECENTLY

^,prevalence of COPD was unknown. This contrasts with other chronic conditions such as diabetes or hypertension which were estimated during the sixties and seventies for different populations. For example, the famous Framingham in New England cohort was established in 1948.

International comparisons are limited by under-diagnosis and report variability. For example EUROSTAT published mortality data (Standardised Death Rate) per 100,000 persons due to chronic lower respiratory diseases[20]. Large between-countries variations were observed which were not concordant with smoking prevalence in the same country. For example, in Greece, despite high smoking prevalence (men 62.4%; women 32.8%), death rates were low.

By contrast, in Denmark smoking prevalence (men 35.8%, women 29.4%) is lower and mortality related COPD ranked highest in Europe (see table on page 18). Between countries variations are best explained by the consequences of biased report from physicians or health authorities. This limitation highlights the need of population-based studies using spirometry, which is not dependent on physician’s decisions, to better estimate the true burden of COPD.

U

Table: Mortality (Standardized Death Rate, SDR) per 100,000 due to chronic lower respiratory diseases (ICD-10 code J40-J47) for men, women and total population in 2000; asthma (ICD-10 code J45-J46) subtracted in a number of countries (source:

Eurostat, 2006)

Men (SDR)

Women (SDR)

Total (SDR)

Denmark 53.9 42.5 45.7

Ireland 53 27.9 37.7

Netherlands 53.3 19.4 31.2

Belgium (1997) 57.9 14.4 30.6

United Kingdom 41.3 22.9 30.0

Hungary 46.9 17.0 28.2

Spain 49.3 9.0 24.6

Slovenia 49.4 11.3 24.3

Lithuania 46.3 10.2 22.6

Norway 30.6 17.9 22.5

Malta 38.9 7.9 20.8

Luxembourg 38.5 8.6 20.3

Germany 30.4 9.9 17.2

Italy 31.0 8.7 16.9

Austria 26.8 10.2 16.4

Portugal 27.7 8.5 16.2

Poland 29.1 6.8 15.1

Switzerland 26.2 8.3 15.1

Sweden 19.9 12.2 15.0

Finland 31.0 6.0 14.8

Slovakia 26.1 7.6 14.6

Czech Republic 23.2 7.9 13.8

Estonia 22.2 3.4 9.5

Greece 13.6 4.9 8.8

France 15.1 4.3 8.4

Latvia 16.5 3.3 7.8

COPD

PREVALENCE IN

2004 (M

ETA

-

ANALYSIS BY

H

ALBERT

, 2006)

Studying the period 1990–2004, Halbert et al found a pooled prevalence of spirometry- defined COPD of 9.2% whereas the prevalence of physician-diagnosed or self-reported COPD was almost twice lower (5.2%).[21] The meta-analysis included more than 100’000 cases from a population of 4 millions subjects living on 4 WHO regions (Africa and Eastern Mediteranean regions did not provided data). The table adapted from Halbert et al displays the prevalence estimates based on spirometry or various diagnostic methods such as self report, physician diagnosis, physical exam or radiologic testing. Overall, Halbert found that

spirometry, the COPD prevalences vary from 2.1% to 26.4% between countries. Spirometry defined pooled COPD prevalence was measured at 9.2% for subjects aged 40 or older.

Table: Pooled COPD prevalence estimates based on studies published before 2003.

Source: adapted from Halbert, ERJ 2006[21]

Studies (n)

Prevalence %, (between countries

range)

Pooled prevalence % (CI 95%)

Spirometry 26 10.1 (2.1–26.4) 9.2 (7.7–11.0)

Patient-reported diagnosis 7 3.7 (3.0–10.5) 4.9 (2.8–8.3)

Physician diagnosis 4 4.1 (2.3–18.2) 5.2 (3.3–7.9)

Physical/radiography 1 13.7 (12.9–14.5)

In the same study, Halbert reported mean prevalences with ranges and pooled prevalence with confidence interval for different age categories, smoking status, or WHO region (table on page 20). As expected, prevalences increased with age and smoking. Some large WHO areas such as whole Africa and Eastern Mediterranean were not studied at all. The meta-analyse reports for subjects aged 40 or more a prevalence of 10% with the following distribution (Halbert, personal communication):

Stage I: 6.6% (4.2% 10.3%) Stage II: 4.3% (3.7% 5.0%) Stage III/IV: 1.2% (0.8% 1.8%)

Table: Pooled COPD prevalence estimates based on studies published before 2003, by age categories, smoking status, sex and WHO regions. Source: adapted from Halbert, ERJ 2006[21]

Studies (n)

Cases (n)

Total (n)

Prevalence %, (between countries range)

Pooled prevalence

% (CI 95%) Overall 37 111261 4123646 8.9 (2.1–26.4) 7.6 (6.0–9.5) Age <40 yrs 9 1074 25362 2.7 (0.8–10.6) 3.1 (1.8–5.0)

40–64 yrs 23 2793 30942 7.6 (1.8–28.7) 8.2 (6.5–10.3)

≥65 yrs 11 2140 15153 15.0 (4.8–29.7) 14.2 (11.0–18.0)

Smoking status

Smoker 17 3133 24122 15.2 (5.1–39.7) 15.4 (11.2–20.7)

Ex-smoker 16 1240 14521 12.7 (2.8–27.7) 10.7 (8.1–14.0) Never-smoker 16 1235 32542 3.9 (0.7–14.6) 4.3 (3.2–5.7) Sex

Male 27 16480 327293 11.0 (2.5–28.0) 9.8 (8.0–12.1)

Female 27 12024 356398 5.0 (1.8–25.2) 5.6 (4.4–7.0)

WHO region

Africa 0 0 0

Americas 3 2666 27599 4.5 (3.2–14.0) 4.6 (2.8–7.6)

Eastern Mediterranean 0 0 0

Europe 28 104773 4015455 8.3 (2.1–26.4) 7.4 (5.9–9.3)

South-East Asia 2 747 6044 12.5 (7.1–17.9) 11.4 (4.4–26.4) Western Pacific 4 3075 74548 10.6 (3.0–18.2) 9.0 (3.0–24.1)

However, studies performed during the period 1990-2004 are hampered by several limitations.

1) Lack of spirometry with broncho-dilation.

The GOLD guideline recommend that post broncho-dilation spirometry only should be interpret to diagnose COPD. This is supported by the fact that subjects with asthma, which is reversible, may be misclassified as having COPD. Studies evaluated the misclassification risk to 30% [22, 23].

2) Unstandardized spirometric criteria

a. Use of the fixed FEV1/FVC ratio naturally leads to overdiagnosis of COPD in older subjects. Risk of overdiagnosis is largest for FEV1 predicted value above

induced by the fixed ratio as a sole criteria, as well as the partial correction when FEV1 <80% is also applied. At age 70 more than half of the subjects GOLD classified as COPD have ―mild COPD‖.

Figure: Comparison of spirometry criteria for the diagnosis of COPD: results from the BOLD study in never smokers. Source ERJ 2009 [24]

Legend : adapted from Vollmer et al[24].

b. Heterogeneity of spirometers may also lead to systematic bias limiting between studies comparison.

COPD

PREVALENCE IN

2007 (T

HE

BOLD

STUDY

)

he Burden of Obstructive Lung Disease (BOLD) was designed to address some of the above described limitations. The investigators decided for clear COPD criterion (FEV1/FVC <0.7 and FEV1 <80% predicted after broncho-dilation and used a single type of spirometer

(EasyOne spirometer, NDD). Different ethnic groups from the five continents were sampled in order to provide reliable estimates.

T

The EasyOne spirometer, NDD. Flow is measured in an open tube using ultrasonic transit time.

BOLD - COPD stage II

The BOLD study reported the prevalence of GOLD stage II COPD at 10.1% for adult men and 8.5% for adult women. Geographic and gender variations in COPD prevalence were reported between 3.4% in Chinese women and 14.2% in South African Men. Ageing and smoking were the main risk factors for COPD. However, geographical variations were not totally explained by country specific smoking prevalence or demographical pattern.

Compared to the meta-analyse by Halbert, the BOLD study reports higher COPD prevalences (BOLD stage II overall 9% vs Halbert stage II 4.3%). Explanations for the apparently

growing prevalence of COPD are multiple:

1) True increase in COPD prevalence due to global population ageing,

2) Methodological issues such as sampling method. In that case, it is likely that the sampling method of BOLD better reflect the true demographic structure of the world adult population.

This suggests an underestimation of the true burden of COPD before.

3) Changing causes of COPD (environmental exposure not related to smoking). Given the slowly decreasing prevalence of smoking at least in Europe and North-America, COPD prevalence was expected to be lower than observed. Higher than expected prevalence suggest that COPD is caused by other exposures not related to smoking.

BOLD - COPD stage I

The BOLD initiative also reported a prevalence of GOLD stage 1 COPD varying between 1.4% (Philippines) and 15.5% (Austria).[25]

The prevalence of post bronchodilator GOLD stage I was uniformly higher in men and but varied between sites. Interestingly, in contrast with previous studies, subjects with stage I COPD were outnumbered by subjects in stage II/IV disease in 6 (out of 12) study sites (Cape Town SA, Guangzhou China, Adana Turkey, Lexington USA, Manila Philippines, Sydney Australia). European and Canadian sites exhibited the expected higher prevalence of stage I COPD compared to stage II. (for example Salzburg Austria stage I 15.5%, stage II 9.2%. This finding was not observed before to our knowledge. It may reflect the negative selection of older subjects in the BOLD study or more likely the impact of broncho-dilation. The BOLD investigators discussed whether mild COPD should be regarded as a clinically important disease. They advocated the need for longitudinal studies comparing subjects in good health with ―mild COPD‖ and other with significant disease.

Nevertheless, in population studies, GOLD stage 1 COPD is frequently the most prevalent stage of the disease.

An example comes from the PLATINO study which was perfomed in South America. In five Latin American cities the prevalence of GOLD stage 1 COPD ranged from 5.2% to 12.5%

whereas stage 2 or higher was uniformly lower (2.6–7.1%).[26]

From the above reports, the clinical and epidemiological importance of mild COPD was debated:

1) Some advocated that, at a population level, mild COPD was associated with higher mortality[27],

2) Some underlined the risk of missing important clinical health consequences of COPD as defined by more stringent criteria because of the

misclassification induced by the fixed FEV1/FVC ratio which dramatically increases prevalence of ―mild COPD for older people.

3) Others pointed out the risk of overtreatment of subjects wrongly labelled as

On page 95, we will discuss this issue and present our personal findings on the long-term outcome of subjects with mild COPD.

PREVALENCE OF COPD IN SWITZERLAND

REVALENCE OF

COPD

^IN

S

WITZERLAND stayed unreported until recently. Largely incomplete data were published on number of admission or mortality due to COPD, solely based on hospital records. Because the disease is underdiagnosed, and the death certificates provide data of low quality, true death rate attributable to COPD could not be reliably inferred from Swiss statistics. Evidence of

underdiagnosis was brought by a study showing that spirometry, which is considered the gold standard regarding the diagnosis of COPD is underused by physicians in Switzerland. The diagnosis was frequently solely based on a probabilistic approach integrating respiratory symptoms such as chronic cough or phlegm and exposure to tobacco [28-30].

PREVALENCE OF COPD IN SWITZERLAND,METHODOLOGY

To address the limitations of previous population studies published elsewhere, we opted for the followings:

a) Our spirometry was analyzed using the lower limit of normal cut off (5^th percentile of normal value) for FEV1/ FVC ratio instead of the fixed ratio, in order to avoid

overdiagnosis among older subjects.

b) Only subjects in stage II COPD were considered (FEV1 <80% predicted). This point allows more direct comparisons between recent studies.

c) Subjects reporting respiratory symptoms were analyzed separetely, because those subjects experience worse long-term outcomes.

d) Subjects with a history of asthma were analyzed together and separately with subjects not reporting asthma.

e) Health-related quality of life and utilization of care were measured and compared between healthy subjects and those with COPD, in order to anchor spirometric categories into patient-centered outcomes. We thought that HRQoL and health care utilization better reflect the true burden of COPD even in population studies.

Finally, because COPD in never smokers received little attention so far, but appears to

P

STUDY DESIGN AND PARTICIPANTS

For this study, we included 6126 subjects from the SAPALDIA (Swiss Study on Air Pollution and Lung Diseases in Adults) cohort which has been described in detail elsewhere. [19, 31].

In brief, SAPALDIA is a representative sample of the Swiss adult population chosen in 1991 to cover geographical and cultural diversity of Switzerland. Eight study areas (Basel, Aarau, Geneva, Montana, Davos, Wald, Payerne and Lugano) participated in the study (table on page 26). From 9651 adult individuals aged 18 to 60 at cohort inception (SAPALDIA 1 [1991]) 8047 were included in the follow-up survey in 2002 (83.4%). Of those 6126 could be re-assessed with pulmonary function tests and respiratory symptoms questionnaires in 2002.

Table: SAPALDIA subjects by centers in 1991 and 2002. Source (SAPALDIA dataset, personal analysis

Study center

1991 N 2002 N (% of 1991

sample)

1191991)

Basel 1491 1192 (83.2%)

Wald 1518 1399 (94.0)

Davos 745 637 (87.4)

Lugano 1310 1140 (89.4)

Montana 794 694 (90.2)

Payerne 1495 1168 (81.1)

Aarau 1299 1080 (85.7)

Geneva 999 736 (75.7)

Total 9651 8047 (85.9)

The flow chart of SAPALDIA subjects for the present analyses is displayed below. Because SAPALDIA did not measure post broncho-dilation spirometry, we chose to report about airflow obstruction and not on COPD in the results of this study.

Flow chart of SAPALDIA subjects included in the prevalence of AO study (1991-2002)

AO: airflow obstruction; PFTs : Pulmonary functions tests; Obstruction: FEV1/FVC <Lower limit of normal value; Stage 1: FEV1 ≥0.8 predicted; Stage2-4: FEV1 <0.8 predicted

Because airflow obstruction (AO) develops after long lasting exposure to noxious agents, we based our estimates on data from the follow up survey SAPALDIA 2 [2002] where

participants had a median age of 53 years, [range 30-73]) instead of 42 at SAPALDIA 1.

However, in most follow-up population studies, subjects non-participation is not due to chance alone. Characteristics predictive of participation at SAPALDIA 2 were analysed, in order to evaluate potential bias due to non-participation. Subjects who never smoked, aged 40 or more, Swiss citizens with high education level, normal or low body mass index and normal baseline FEV1 were more likely to participate at SAPALDIA 2. Factors predicting

participation were comparable with other population studies. See table and figure on page 27 and 28.

Figure: participation rate at SAPALDIA 2 according to birth cohort at SAPALDIA 1

Table: Odds ratio* of participation to the pulmonary function test (PFTs†) at SAPALDIA 2 (2002),

SAPALDIA 1 characteristics (1991) OR* of participation at follow up (CI 95%) Participants / Non-participants / Total

n=6126/ 3524/ 9651 Year of birth

1963 and after 1

1953-1962 1.44 (1.24 1.67)

1943-1952 1.53 (1.32 1.77)

1933-1942 1.39 (1.20 1.62)

1932 and before 1.19 (0.97 1.46)

Gender

Male 1

Female 1.08 (0.98 1.46)

Nationality

Non-Swiss 1

Swiss 1.93 (1.71 2.18)

Education level

Lowest 1

Highest 1.49 (1.26 1.76)

Smoking status

Never smoker 1

Former smoker 0.82 (0.72 0.92)

Current smoker 0.64 (0.58 0.71)

Body mass index

<20 kg/m² 1

21 to 25 kg/m² 1.08 (0.93 1.24)

26 to 30 kg/m² 0.92 (0.78 1.08)

≥31 kg/m² 0.58 (0.47 0.72)

PFTs

FEV1 ≥80% predicted 1

FEV₁ <80% predicted 0.68 (0.56 0.84)

FVC ≥80% predicted 1

FVC<80% predicted 0.92 (0.71 1.19)

*Multivariate logistic regression model controlling for the above-mentioned variable and the study area (random effect); †: pre-bronchodilator pulmonary function tests.

DEFINITION OF AIRFLOW OBSTRUCTION

Pulmonary function tests (PFTs) were performed without bronchodilators by trained technicians according to the American Thoracic Society standards. We defined airflow obstruction according to the lower limit of normal FEV1/FVC derived from population specific prediction equations. [32] In accordance with the recently published studies on the prevalence of COPD, we report airflow obstruction in presence of FEV1/FVC <LLN and FEV1<0.8 predicted (modified stage 2-4 AO)[25]. To facilitate international comparisons we also report the prevalence of AO as defined by the fixed GOLD criterion (FEV1/FVC <0.7).

Because respiratory symptoms are important predictors of FEV1 decline and respiratory care use, we report prevalence of symptomatic airflow obstruction. [15, 33]

Chronic cough or chronic phlegm or chronic shortness of breath by walking were used to define respiratory symptoms. The underlying questions have been described in detail previously and are displayed below:[15]

Respiratory questionnaires at SAPALDIA

Subjects were considered symptomatic if they answered yes to one of the following questions:

Chronic cough:

―Do you usually cough first thing in the morning?‖ or ―Do you usually cough during the day, or at night?‖ or

―Do you cough like this on most days for as much as 3 months each year?‖

Chronic phlegm:

―Do you usually bring up phlegm from your chest first thing in the morning?‖ or

―Do you bring up phlegm like this on most days for as much as 3 months each year?‖ or

―Do you usually bring up any phlegm from your chest during the day, or at night?‖

Chronic dyspnea:

―Are you troubled by shortness of breath when hurrying on level ground or walking up a slight hill?‖ or

―Do you get short of breath walking with other people of your own age on level ground?‖ or

―Do you have to stop for breath when walking at your own pace on level ground?

METHACHOLINE BRONCHIAL CHALLENGE TESTS

Bronchial challenge tests were performed at SAPALDIA 1 with administration of

methacholine chloride in subjects who had no contraindication.[31] The test was considered positive if FEV₁ dropped by 20% or more from the pre-test level.

COVARIATES

Subjects who answered yes to both questions ―have you ever had asthma?‖ and, if yes, ―was this confirmed by a doctor?‖ were classified as having ―physician diagnosed asthma‖.

Education level, nationality, comorbid conditions, smoking status, lifetime smoking (packs of cigarettes/day * smoking duration [years]), environmental tobacco smoke exposure and level of physical activity were derived from the questionnaires. Detailed methods regarding the definition of physical activity have been published before.[34] Short Form 36-item (SF-36) was administered to assess health-related quality of life. Respiratory care utilization was considered when inhaler use or emergency room visit or hospitalisation or ambulatory visit (all for respiratory problems) was reported during the year preceding SAPALDIA 2.

STATISTICAL ANALYSIS

Prevalence was measured for different population categories such as age groups or sex. To evaluate potential biases related to differential non participation at SAPALDIA 2 which could have affected airflow obstruction prevalence at SAPALDIA 2, the probability of participation was estimated using the following predictor variables recorded at SAPALDIA 1: age, age squared, sex, nationality, smoking status, education level, categories of BMI, percent predicted FEV₁, percent predicted FVC and study area (table on page 28). Analyses were then repeated by assigning each observation a specific weight corresponding to the inverse of the respective subject’s probability of participation (inverse probability weighting).

Airflow obstruction risk factors were studied with multivariate analysis. Multivariate analysis involved mixed logistic regression models, systematically controlling for categories of age and smoking with the study area as a random effect variable. These variables were chosen a priori based on published literature. Covariates potentially associated with obstruction were tested one by one in models controlling for the above mentioned core variables.

Statistical analyses were carried out with Stata version 10 (StataCorp, 4905 Lakeway Drive, College Station, Texas 77845 USA).

R

ESULTS OF THE

SAPALDIA

STUDY ON PREVALENCE OF

COPD

AND ASTHMA IN

S

WITZERLAND

Prevalence of COPD / airflow obstruction

Table on page 32 compares the prevalence of AO stage 2 or higher as defined by the lower limit of normal of FEV1/FVC or the GOLD criterion. Compared to the LLN, the fixed FEV1/FVC ratio led to higher AO prevalence in older age categories (15.2%GOLD CI95%

(11.1 20.3) vs 8.0%_LLN CI95% (5.3 11.9)). Overall, stage 2 or higher AO was found in 5.1%

CI95% (4.3 5.9) according to the LLN and 7.0% CI95% (6.0 8.3) according to the GOLD criterion.

The figure below compares the AO prevalence using two different definitions. For teaching purpose, a graph showing similar trends from the NHANES study is shown.

Figure: comparisons of AO stage 2-4 prevalences using two different definitions: Lower limit of normal for FEV1/FVC ratio & the “fixed ratio”.

Table: Airflow obstruction prevalence in SAPALDIA 2 (2002) by age group and sex

Characteristics at

SAPALDIA 2 (2002) FEV1/FVC<LLN,

stage 2-4

FEV1/FVC<0.7, stage 2-4 Men

n=180

%, [CI95%]

Women n=127

%, [CI95%]

Men n=275

%, [CI95%]

Women n=152

%, [CI95%]

Overall, n=6126 6.1 [5.3 7.1] 4.0 [3.3 4.7] 9.4 [8.4 10.5] 4.8 [4.1 5.6]

Age 30-39, n=1109 3.2 [2.0 5.0] 1.9 [0.9 3.4] 3.4 [2.1 5.2] 0.9 [0.4 2.2]

Age 40-49, n=1525 3.2 [3.0 6.0] 4.0 [2.9 5.6] 5.8 [4.3 7.8] 4.0 [2.9 5.6]

Age 50-59, n=1811 6.1 [4.7 7.9] 4.7 [3.5 6.2] 9.2 [7.5 11.3] 5.4 [4.1 7.0]

Age 60-69, n=1378 8.9 [7.0 11.4] 5.0 [3.6 6.9] 15.2 [12.6 18.1] 7.4 [5.7 9.5]

Age 70+, n=303 15.0 [9.9 22.1] 2.4 [0.8 6.1] 26.3 [19.5 34.4] 5.9 [3.2 10.6]

<0.001 0.029 <0.001 <0.001

* P value from χ² test

LLN: lower limit of normal of FEV1/FVC ratio.

Symptomatic only* stage 1-4 obstructive lung disease

All stage 2-4 obstructive lung disease†

Symptomatic only* stage 2-4 obstructive lung disease

All SAPALDIA

subjects Men Women

1932 and before

1933-1942

1943-1952 1953-1962

1963 and after 0%

5%

10%

15%

20%

25%

30%

1991survey 2002 survey

1932 and before

1933-1942

1943-1952

1963 and after 1953-1962

0%

5%

10%

15%

20%

25%

30%

1991survey 2002 survey

1932 and before

1933-1942

1943-1952

1953-1962

0%

5%

10%

15%

20%

25%

30%

1991survey 2002 survey

1963 and after

1932 and before

1933-1942

1943-1952 1953-1962 1963 and after 0%

5%

10%

15%

20%

1991survey 2002 survey

1932 and before

1933-1942

1943-1952

1953-1962 1963 and after

0%

5%

10%

15%

20%

1991survey 2002 survey 1991sur vey 2002 sur vey

1933-1942 1943-1952 1953-1962

1963 and after 1932 and before

0%

5%

10%

15%

20%

1932 and before 1933-1942

1943-1952 1953-1962

1963 and after

0%

5%

10%

15%

20%

1991survey 2002 survey

1932 and before

1933-1942

1943-1952 1953-1962 1963 and after 0%

5%

10%

15%

20%

1991survey 2002 survey

1932 and before

1933-1942 1943-1952 1953-1962 1963 and after 0%

5%

10%

15%

20%

1991survey 2002 survey

Overall, 10.0% (CI95% 8.5% - 11.8%) of the adult population qualified for LLN defined AO.

More than half of subjects with stage 1 AO (n total=310) were free of respiratory symptoms (n=166, 53.6%).

Figure on page 33 provides the prevalence of AO (any stage) and 2-4 AO at SAPALDIA 2 (2002) with percentage reporting respiratory symptoms.

For subjects with FEV1/FVC below LLN, (all stage) and those with stage 2-4 AO, prevalence increased steadily with age and men were more frequently affected than women. Most

subjects with stage 2-4 AO reported one or more chronic respiratory symptoms (lower panel).

Figure: Prevalence of airflow obstruction at SAPALDIA 2*

*pulmonary function tests were performed without broncho-dilation. Respiratory symptoms:

chronic cough or phlegm or shortness of breath by walking.

PHYSICIAN DIAGNOSED ASTHMA PREVALENCE

Physician diagnosed asthma prevalence is detailed for categories of age in figure on page 34.

Asthma was less frequently reported in older age categories (5.7% [CI95% 3.5%-9.1%] for

age: up to 2.1% (1.4%-3.0%) for those aged 60 to 69. However in the oldest age group (70 or more, n=303), both conditions were found in 4 subjects only (1.3% [0.5%-3.5%]).

Figure: Physician diagnosed asthma, stage 2-4 airflow obstruction and concomitant asthma+AO by age group.

RISK FACTORS FOR COPD

Tables on page 36 & 37 compare the characteristics of normal subjects (normal spirometry and no report of respiratory symptoms) with a) subjects with respiratory symptoms but no AO, b) stage 1 AO or c) stage 2-4 AO.

Table on page 36 shows that AO prevalence increases with smoking, environmental tobacco smoke exposure, low education, non-Swiss citizenship and physical inactivity. Ever smokers (men 65.4%; women 49.8%) reported 26.2 pack years (median 20.4, iqr 27.4) for men and 17.0 (median 11.5, iqr 21.0) for women. Environmental tobacco smoke during childhood, professional exposure to dust smoke or fumes or outdoor fine particulate matter exposure were not associated with AO.

Subjects with stage 2-4 AO also reported higher rate of comorbid conditions (Table on page 37).

When examining risk factors for stage 2-4 AO as defined using the fixed FEV₁/FVC ratio instead of the LLN ratio, we found that those risk factors were associated in a very similar manner with AO.

Table: Risk factors prevalence by severity of airflow obstruction in SAPALDIA 2 (2002) Characteristics at SAPALDIA 2

(2002) FEV1/FVC≥LLN FEV₁/FVC<LLN,

stage 1

FEV₁/FVC<LLN, stage 2-4 n=6126

no symptoms (reference)

n=3342

with symptoms

n=2161

All n=310

With symptoms

n=144

All n=307

With symptoms

n=216 Tobacco smoking exposure

Ever smoker (2002) n (%) 1786 (53.5) 1265 (58.6) 191 (61.6) 99 (68.8) 217 (70.7) 161 (74.5)

P value <0.001 <0.001 <0.001 <0.001 <0.001

Pack year (ever smokers), median

(iqr) 10.0 (20.9) 16.0 (28.9) 16.2 (28.7) 20 (27.0) 31.2 (41.2) 35.0 (42.0)

P value <0.001† <0.001† <0.001† <0.001† <0.001†

Environmental tobacco smoke exposure

Not exposed, n (%) 2633 (78.8) 1536 (71.1) 229 (73.9) 98 (68.1) 212 (69.3) 147 (68.4)

≤ 3 hours/week 443 (13.3) 373 (17.3) 55 (17.4) 29 (20.1) 54 (17.7) 40 (18.6)

> 3 hours/week 265 (7.9) 250 (11.6) 26 (8.4) 17 (11.8) 40 (13.1) 28 (13.0)

P value <0.001 0.077 0.009 <0.001 0.001

Socio-educational level

Low education, n (%) missing n=6 129 (3.9) 190 (8.8) 22 (7.1) 15 (10.4) 26 (8.5) 21 (9.7)

P value <0.001 <0.001 <0.001 <0.001 <0.001

Non Swiss nationals, n (%) 390 (11.7) 341 (15.8) 30 (9.7) 16 (11.1) 50 (16.5) 35 (16.4)

P value <0.001 0.290 0.834 0.015 0.038

Physically active, n (%) missing n=64 1053 (31.8) 518 (24.2) 105 (34.4) 47 (33.1) 75 (24.8) 45 (21.2)

P value <0.001 0.347 0.746 0.011 0.001

All statistical comparisons [χ² test or †Wilcoxon rank sum test] made with ―FEV /FVC≥LLN & no respiratory symptoms‖ (reference group).

Table: Prevalence of self-reported comorbid conditions by severity of airflow obstruction in SAPALDIA 2 (2002) Characteristics at SAPALDIA 2

(2002) n=6126

FEV1/FVC≥LLN FEV1/FVC<LLN FEV₁/FVC<LLN, stage 2-4 no symptoms

(reference) n=3342

with symptoms

n=2161

All n=310

With symptoms

n=144

All n=307

With symptoms

n=216 Diabetes, n (%) missing n=6 71 (2.1) 84 (3.9) 4 (1.3) 2 (1.4) 16 (5.2) 10 (4.6)

P value 0.001 0.406 0.360 0.001 0.017

Hypertension, n (%) missing n=9 429 (12.8) 471 (21.8) 45 (14.5) 28 (19.4) 62 (20.2) 49 (22.7)

P value <0.001 0.402 0.022 <0.001 <0.001

Cardiac disease, n (%) missing n=5 152 (4.6) 220 (10.2) 13 (4.2) 6 (4.2) 26 (8.5) 22 (10.2)

P value <0.001 <0.775 0.831 0.002 <0.001

BMI <21 kg/m², n, (%) 223 (6.7) 127 (5.9) 20 (6.5) 8 (5.6) 16 (5.2) 12 (5.6) BMI 21 – 24.9 kg/m² n, (%) 1520 (45.5) 740 (34.2) 146 (47.1) 66 (45.8) 96 (31.4) 61 (28.4) BMI 25 – 29.9 kg/m² n, (%) 1233 (36.9) 813 (37.6) 108 (34.8) 46 (31.9) 124 (40.5) 88 (40.9) BMI 30+ kg/m² n, (%) missing n=19 366 (11.0) 481 (22.3) 36 (11.6) 24 (16.7) 70 (22.9) 54 (25.1)

P value <0.001 <0.895 0.159 <0.001 <0.001

Atopy with rhinitis‡, n (%) missing

n=601 487 (14.6) 288 (13.5) 37 (12.1) 14 (9.8) 50 (16.5) 33 (15.6)

P value 0.226 0.202 0.103 0.394 0.726

Physician diagnosed asthma, n (%)

missing n=4 132 (4.0) 215 (10.0) 39 (12.6) 30 (20.8) 88 (28.7) 72 (33.3)

P value <0.001 <0.001 <0.001 <0.001 <0.001

All statistical comparisons [χ² test or †Wilcoxon rank sum test] made with ―FEV1/FVC≥0.7 & no respiratory symptoms‖ (reference group).

‡:presence of atopy (positive phadiatop and rhinitis [2002]) BMI: body mass index

Table on page 39 details the adjusted odds ratios (OR) of stage 2-4 and stage 1 AO at

SAPALDIA 2 for different exposures. Smoking was the strongest risk factors for all stages of AO and age played a role for stage 2-4 AO. However, odds ratio of stage 2-4 AO in

association with smoking were higher than for stage 1(OR stage 2-4 1.25 [CI95%1.19-1.30] vs OR stage 1 1.12[CI95%1.05-1.30] for each ten unit pack-year increase). Obesity or physical inactivity were not associated with AO. Physician diagnosed asthma was associated with AO for stage 1 and stage 2-4.

In a sensitivity analysis, when examining these exposures for stage 2-4 AO as defined using the fixed ratio of FEV1/FVC instead of the LLN definition, we found a strong association between ageing and AO for stage 1 and 2-4.

Table: Adjusted odds ratio of stage 2-4 airflow obstruction*

Characteristics at SAPALDIA 2 (2002) n=6126

FEV1/FVC<LLN, stage 1

FEV1/FVC<LLN, stage 2-4 n=310/5819‡ n=307/6126

Age 30-39 ref ref

Age 40-49 0.95 (0.68 1.32) 1.37 (0.87 2.17)

Age 50-59 0.69 (0.48 0.98) 1.62 (1.05 2.51)

Age 60-69 0.70 (0.48 1.02) 2.09 (1.34 3.25)

Age 70+ 0.82 (0.46 1.47) 2.76 (1.55 4.91)

Women (vs men) 1.67 (1.31 2.13) 0.82 (0.64 1.05)

Tobacco smoke exposure

Ever smoker (2002) (vs never) 1.42 (1.11 1.80) 1.76 (1.36 2.28) Pack year (per 10 unit increase)∫ 1.12 (1.05 1.19) 1.25 (1.19 1.30 Environmental tobacco smoke exposure

Not exposed ref ref

≤ 3 hours/week 1.08 (0.79 1.49) 1.09 (0.79 1.51)

> 3 hours/week 0.79 (0.52 1.21) 1.25 (0.86 1.80)

ETS† during childhood (maternal

exposure) 1.44 (1.04 1.99) 0.82 (0.55 1.21)

Socio-economical status

Low education (vs high) 1.11 (0.66 1.85) 1.28 (0.78 2.11) Non Swiss nationals (vs Swiss) 0.80 (0.54 1.18) 1.25 (0.89 1.73) Professional exposure to dust, smoke or

fumes, 1.01 (0.76 1.36) 0.91 (0.68 1.22)

Physical activity 1.29 (1.00 1.67) 0.88 (0.67 1.17)

Associated conditions

Atopy with rhinitis‡‡ 0.90 (0.63 1.28) 1.67 (1.21 2.31) Physician diagnosed asthma 2.10 (1.47 3.00) 6.70 (5.04 8.91)

BMI <21 kg/m² ref ref

BMI 21 – 24.9 kg/m² 1.26 (0.77 2.05) 0.80 (0.46 1.39) BMI 25 – 29.9 kg/m² 1.12 (0.67 1.87) 0.91 (0.52 1.60)

BMI 30+ kg/m² 0.85 (0.48 1.52) 1.14 (0.64 2.05)

*adjusted for age, sex, smoking exposure and study area †ETS: environmental tobacco smoke. ∫: Lifetime smoking for ever smokers (per 10 pack-year increase) BMI: body mass index. ‡ subjects with stage 2-4 were excluded. ‡‡ at SAPALDIA 1

QUALITY OF LIFE AND RESPIRATORY CARE UTILIZATION FOR SUBJECTS WITH COPD Table on page 41 details the quality of life scores of normal subjects and subjects with obstruction. Out of 6126 subjects, 5278 (86.2%) filled the SF-36. AO and symptomatic AO were systematically associated with lower health-related quality of life and more so for those with stage 2-4 AO and symptoms. Respiratory care utilization increased with severity of AO and symptoms.

Table: Quality of life scores and respiratory health care utilisation in normal subjects, subjects with airflow obstruction and/or respiratory symptoms in SAPALDIA 2 (2002)

Characteristics at SAPALDIA 2

(2002) FEV₁/FVC≥LLN FEV1/FVC<LLN &

Stage 1

FEV1/FVC<LLN &

Stage 2-4 no symptoms

(reference) n=2897

With symptoms

n=1842

All n=279

With symptoms

n=129

All n=258

With symptoms

n=182 SF-36 scores, mean (SD)

Physical Functioning 93.7 (12.9) 84.2 (19.4)* 89.2 (15.7)* 84.5 (16.6)* 80.6 (20.2)* 78.3 (19.2)*

Role Physical 92.3 (21.8) 82.6 (31.4)* 88.4 (27.5)* 83.2 (33.1)* 81.1 (33.1)* 80.3 (32.5)*

Bodily pain 85.4 (21.1) 73.9 (24.3)* 80.0 (23.6)* 74.5 (25.1)* 76.8 (25.2)* 75.0 (25.1)*

General Health 65.0 (11.4) 61.1 (13.1)* 63.5 (12.9)† 61.7 (13.0)* 59.8 (13.5)* 58.2 (13.5)*

Vitality 67.5 (15.0) 59.0 (17.1)* 63.9 (17.3)* 59.6 (18.8)* 59.5 (18.4)* 58.2 (18.1)*

Social Functionning 90.3 (16.1) 82.5 (20.4)* 85.8 (20.9)* 79.5 (24.4)* 83.8 (19.7)* 82.8 (19.2)*

Role Emotional 92.1 (21.7) 82.6 (31.2)* 86.8 (29.7)* 78.3 (36.4)* 82.9 (32.2)* 81.6 (32.7)*

Mental Health 77.6 (13.7) 70.9 (16.6)* 74.6 (16.6)* 70.6 (18.8)* 72.8 (15.4)* 71.9 (15.7)*

Physical component summary 53.3 (6.5) 50.1 (9.4)* 52.0 (8.0)* 50.6 (8.8)* 48.6 (9.6)* 48.0 (9.1)*

Mental component summary 51.7 (7.5) 48.5 (9.3)* 50.1 (9.5)* 47.5 (11.4)* 49.8 (8.8)* 49.4 (9.1)*

Respiratory care utilisation# 206/3344 (6.2%)

405/2161 (18.7%)*

50/310 (16.1%)*

36/144 (25.0%)*

92/307 (30.0%)*

81/216 (37.5%)*

All statistical comparisons [χ² test or t-test for unequal variances] made with reference group ―FEV₁/FVC≥LLN & no respiratory symptoms‖. *:

P<0.001 †:P=0.06 #Report of emergency room visit, hospitalisation, ambulatory visit (all for respiratory problems) or report of inhaler use during the year preceding SAPALDIA 2 (2002).

COPD IN NEVER SMOKERS

Of 307 subjects with stage 2-4 AO at SAPALDIA 2, 90 (29.3%) were never smokers.

Examined by smoking status, prevalence of stage 2-4 AO was 6.3% [CI95% 5.3% – 7.6%] for ever smokers and 3.4% [CI95% 2.7% – 4.3%] for never smokers.

The figure on page 42 details the proportion of never smokers by age categories. COPD in never-smokers was more prevalent in younger age group.

Figure: Proportion of never smokers in subjects with COPD stage II/IV

Table on page 44 reports the distribution of various risk factors and covariates for stage 2-4 AO in subjects without or with smoking history. Never smokers with stage 2-4 AO were younger (p=0.003) and more often women (52.2% vs 36.9%, p=0.013). A third of never smokers with stage 2-4 AO reported asthma at SAPALDIA 1 (34.8% vs 18.0% for smokers with similar AO severity, p=0.001). Atopy was more frequent in never-smoker with AO.

Chronic cough and phlegm were similarly distributed and health-related quality of life equally impaired in smokers and never smokers with AO. Respiratory care utilization tended to be more frequent in never smokers with AO (36.7% vs 27.2%, P=0.099) despite higher rate of reported shortness of breath in smokers.

Table on page 45 shows the adjusted OR of stage 2-4 AO associated with various risk factors

0 2 4 6 8 10

39- 40-49 50-59 60-69 70+

%

Smokers

Never smokers

54%

41%

19%

27%

21%

However, for never smokers, male sex and asthma at SAPALDIA 1 were stronger risk factors of AO than for smokers. In contrast with never smokers, smokers were older and more

exposed to ETS. Interestingly in smokers, asthma was not associated with development of AO at SAPALDIA 2 after adjustment for covariates.

We found a significant interaction between smoking status and asthma (p=0.044).