European Respiratory Society
Respiratory Epidemiology

Over the last decade, the volume of research into the pathophysiology and genetics of pulmonary diseases has increased greatly. This has led to the development of new treatments and therapies for many diseases, including lung cancer, asthma and cystic fibrosis. This issue of the ERS Monograph comprehensively demonstrates the developments in respiratory medicine in recent years. It outlines the importance of epidemiology in respiratory medicine, and will prove a methodological tool that will help disease management. It should also be used as an advocacy tool for the sake of public health.

  • ERS Monograph
  1. Page ix
  2. Page xi
  3. Page xiv
  4. Page 1
    Correspondence: Bo Lundbäck, Krefting Research Centre, Institute of Medicine, University of Gothenburg, P.O. Box 424, Gothenburg, SE-40530, Sweden. E-mail:

    COPD is today the third leading cause of death worldwide and its prevalence has steadily increased. Prevalence in Europe seems to be levelling, and in western and northern Europe, recent data even indicate a decrease. Beyond tobacco, other major risk factors have been identified, while objective possibilities for prevention exist. New medicines and treatment strategies can slow down disease progression. COPD heterogeneity is huge and restricts treatment options, and epidemiology can contribute to identifying clinically relevant phenotypes of COPD.

  5. Page 18
    Correspondence: Deborah Jarvis, Dept of Respiratory Epidemiology and Public Health, Imperial College London, Emmanuel Kaye Building, Manresa Road, London, SW3 6LR, UK. E-mail:

    Asthma and rhinitis are common chronic diseases in children and adults. Both conditions show marked geographical variation in disease prevalence, and the prevalence of each has increased over the past 60 years in many, but not all, parts of the world. Children who have regular contact with other children early in life either within their family or by attending day care have less rhinitis, and this “protection” appears to last into adult life. However, these exposures do not appear to be as beneficial for asthma. There has been intense and extensive research into associations of lifestyle (smoking, diet and obesity) and environmental factors (infections, allergen exposure, farming environments and pollution) with both asthma and rhinitis, but the underlying causes for the geographical variation and temporal trends remain unclear.

  6. Page 37
    Correspondence: Giovanni Battista Migliori, WHO Collaborating Centre for TB and Lung Diseases, Fondazione S. Maugeri, Care and Research Institute, via Roncaccio 16, 21049, Tradate, Italy. E-mail:

    TB is an infectious disease caused by strains of Mycobacterium tuberculosis. It is one of the most important diseases worldwide, together with malaria and HIV/AIDS. The last World Health Organization report highlighted an estimated global incidence of 8.6 million cases (i.e. 122 cases per 100 000 population) in 2012; the highest figures were estimated in India, China, South Africa, Indonesia and Pakistan. In the majority of the cases, high-income countries show an estimated incidence <10 patients per 100 000 population. HIV/AIDS plays an important role in the development of TB disease; consequently, geographical areas characterised by an high HIV/AIDS prevalence show a high TB incidence. Disorders which impair the immune system (e.g. diabetes mellitus or exposure to immunosuppressive drugs) favour the occurrence of pulmonary and/or extrapulmonary forms of TB. A new World Health Organization public health strategy has been recently launched to reduce the global incidence to less than one TB case per 100 000 population by 2050.

  7. Page 48
    Correspondence: Giuseppe Di Maria, Pneumology Unit, A.O.U. Policlinico-Vittorio Emanuele, Dept of Clinical and Molecular Bio-Medicine, Bldg 4, Via Santa Sofia 78, 95123 Catania, Italy. E-mail:

    Overall lung cancer incidence increased steadily during the second half of the 20th century, reaching the highest rank among all cancers for both the whole population and age-standardised population worldwide. Today, lung cancer is the leading cause of cancer mortality in many regions. Its major causative agents have been clearly identified over the last six decades, with more than 80% being attributable to cigarette smoking, which makes lung cancer one of the most preventable tumours. In males, the effect of smoking still determines the high risk of dying from lung cancer in industrialised countries, although incidence and mortality have shown a trend to decline among younger generations. In contrast, females in developed and developing countries have similar morbidity and mortality levels; however, given the increasing number of females who smoke, these are likely to grow worldwide.

    Pleural mesothelioma is a highly lethal tumour that is much less frequent, except among subjects exposed to occupational or environmental asbestos and asbestos-like fibres. The global burden of mesothelioma is unclear, but its age-standardised mortality rate has doubled during the last two decades.

  8. Page 61
    Correspondence: Francesco Blasi, Fondazione Ospedale Maggiore Cà Granda IRCCS, Dipartimento Fisiopatologia Medico-Chirurgica e dei Trapianti, Università degli Studi di Milano, Padiglione L. Sacco, U.O. Broncopneumologia, 20122 Milan, Italy. E-mail:

    This chapter describes the epidemiology of CAP and influenza in Europe. CAP is still a substantial cause of hospitalisations and death across Europe. Hospitalisation rates differ widely across Europe. Aetiological studies across the continent consistently report Streptococcus pneumoniae as the leading pathogen, irrespective of the site of patient management. Costs associated with pneumonia are largely associated with hospitalisation, with diagnostics and antibiotic treatment covering a marginal quota of expenses. In addition to direct mortality, patients with pneumonia are more prone to long-term mortality compared with matched populations.

    Seasonal epidemics of influenza typically spread through Europe during the winter months carrying a considerable socioeconomic burden. Influenza pandemics occur at 10–40-year intervals. The last pandemic, caused by an H1N1 virus, was of moderate intensity and was associated with excess deaths primarily among children, often without underlying diseases.

    Data on influenza vaccination coverage in Europe indicate that, on average, coverage levels are below those set by World Health Organization and European Union standards.

  9. Page 79
    Correspondence: Dominique Valeyre, Hôpital Avicenne, 125 rue de Stalingrad, 93009, Bobigny, France. E-mail:

    The interpretation of epidemiological studies of interstitial lung diseases (ILD) must take into account methodological issues that may impact results: changes in the classification of ILD and of diagnosis criteria, particularly for idiopathic pulmonary fibrosis; the broad or narrow definition of diseases; and sources. The prevalence and incidence of ILD are around 60–80 cases per 100 000 population and 20–30 cases per 100 000 person-years, respectively. There are four main groups of ILD: sarcoidosis (23–38% of ILD cases); ILD secondary to identified causes (30% of ILD cases); certain rare diseases; and idiopathic interstitial pneumonias (26% of ILD cases), with idiopathic pulmonary fibrosis the most frequent of this last group (19% of ILD cases). The epidemiology of various ILD depends on age, sex, race and smoking habits. For some of them, genetic factors and occupational or environmental factors play a significant role. Epidemiological studies are particularly useful for public health knowledge, optimising diagnosis and better understanding pathogenesis.

  10. Page 88
    Correspondence: Thorarinn Gislason, Dept of Respiratory Medicine and Sleep (E7), Landspitali – University Hospital, 108 Reykjavik, Iceland. E-mail:

    Over the past 30 years OSA has evolved from being a very rare disorder, almost solely characterised by loud snoring and daytime sleepiness, to a common systemic disorder with multiple adverse consequences, which affect a substantial percentage of the middle-aged population. Obesity is the strongest risk factor, but not the only pathogenic mechanism. Other important mechanisms include differences in craniofacial dimensions and even lifestyle factors such as smoking, alcohol consumption and sleeping position. Given the systemic effects, there is developing evidence for the role of OSA as an independent risk factor for: excessive sleepiness with increased risk of traffic accidents, hypertension, cardiovascular disease (including atherosclerosis), myocardial infarction, stroke, impaired glucose metabolism, neurodegeneration, and both cancer mortality and incident cancer. Subjects with untreated OSA have increased mortality and are at higher risk when undergoing surgical procedures. The major clinical challenge today is how simplified diagnostic procedures can confirm or reject the possibility of OSA in various situations.

  11. Page 106
    Correspondence: Alain Vergnenègre, Unité d'Oncologie Thoracique et Cutanée, CHU Limoges, 2 avenue Martin Luther King, 87042 Limoges cedex, France. E-mail:

    Cost analyses have become an increasingly important consideration in healthcare decisions over the last decade. This is especially the case in the field of respiratory diseases, the most serious of which have a major impact on public health. This review focuses on the cost of lung cancer (overall costs, costs at different disease stages, and the use of chemotherapy and targeted therapies) and oxygen therapy. The aim is to provide the clinician with data complementary to clinical outcomes, and to expand their knowledge of the economic impact of these major respiratory diseases and conditions.

  12. Page 116
    Correspondence: Rune Grønseth, Lungeavdelingen, Haukeland Universitetssykehus, N-5021 Bergen, Norway. E-mail:

    The economic burden of COPD and asthma in Europe is substantial. According to the European Lung White Book, the annual cost per COPD case was approximately €2100 in 2011, whereas asthma costs were almost €3500 per case. Indirect costs accounted for 52% and 43% of COPD- and asthma-related costs, respectively. Recent studies were in line with these findings, but did indicate that previous estimates might be too conservative. For pneumonia, TB, interstitial lung diseases and pulmonary embolism there was scant information regarding population-based costs.

  13. Page 125
    Correspondence: Elisabeth Zemp, Dept of Epidemiology and Public Health, Unit Society, Gender and Health, Swiss Tropical and Public Health Institute, Socinstrasse 57, PO Box 4002 Basel, Switzerland. E-mail:

    Research on the relationships between sex and gender and respiratory health suggests an impact on incidence, susceptibility and life-course pattern for respiratory diseases. This chapter reviews evidence of the impact of sex and gender on smoking, lung cancer, COPD and asthma, and outlines the possible mechanisms underlying gender differences.

    Worldwide, reductions in age-standardised daily smoking rates have been reported since 1980 from 41.2% to 31.1% for men, and from 10.6% to 6.2% for women. Gender differences in smoking affect the rates of lung cancer and COPD, which are on the increase worldwide, varying considerably across regions and countries. Sex ratios of asthma rates are more homogenous (around 1.2 to 1.5), but exhibit a characteristic lifetime pattern.

    Considerable gender differences are reported for smoking, lung cancer, COPD and asthma. To improve the quality and effectiveness of healthcare by gender sensitive approaches, we need unbiased comparisons of men and women and insights into the mechanisms involved in producing sex and gender differences.

  14. Page 139
    Correspondence: H. Marike Boezen, Dept of Epidemiology, University of Groningen, University Medical Center Groningen, E3.22, Hanzeplein 1, 9700 RD, Groningen, The Netherlands. E-mail:

    In the past few decades, identification of genes predisposing to development of asthma or COPD has been the focus of many candidate-gene studies and genome-wide association studies. Acknowledging the role of environmental factors like cigarette smoking, air pollution and job-related exposures, current studies focus on identification of genetically susceptible groups for these specific exposures. Gene-by-environment interaction studies (candidate gene-by-exposure interaction studies, genome-wide interaction studies and epigenetic studies) aim to identify genetic loci that predispose to asthma and COPD onset after exposure to such environmental risk factors. This chapter gives an overview of the current knowledge of genetic risk factors and gene–environment interactions underlying asthma and COPD.

  15. Page 152
    Correspondence: Anne Greenough, NICU, 4th Floor Golden Jubilee Wing, King’s College Hospital, Denmark Hill, London, SE5 9RS, UK. E-mail:

    Many children suffer chronic respiratory morbidity with troublesome symptoms and lung function abnormalities persisting into adulthood. There are many early life events that increase the risk of this adverse outcome. These include intrauterine growth retardation, antenatal smoking and premature delivery, particularly if the infant then develops bronchopulmonary dysplasia. Respiratory viral infections in infancy are associated with increased wheeze and asthma at follow-up, although some children may have a genetic and/or a functional predisposition to developing a symptomatic LRTI. Further longitudinal studies are required to better understand how early life events may regulate transgenerational epigenetic effects and lead to asthma.

  16. Page 165
    Correspondence: Laura Carrozzi, Pulmonary Unit, Cardio-Thoracic and Vascular Dept, University Hospital of Pisa, Via Paradisa 2, 56124 Pisa, Italy. E-mail:

    Despite reduction in prevalence of active smoking since 1980, the absolute number of smokers has increased worldwide from 718 to 967 million in 2012. In Europe, the prevalence of smoking in 2012 ranged from 13% to 40%, with higher values in south and east. Trends in smoking-related mortality and morbidity follow changes in prevalence; control of smoking prevents premature death and smoking cessation is associated with longer life. Research on the relationship between smoking and health is still in progress: 10 new diseases causally linked to active smoking were introduced in the 2014 US Surgeon General’s Report, along with enhanced causality for TB and idiopathic pulmonary fibrosis. There is no safe level of exposure to passive smoking. Authoritative health organisations agree that passive smoking exposure leads to serious and fatal diseases, including cardiovascular and respiratory diseases, and cancers. Children, including fetuses and newborns, can develop health effects from exposure. About 35% of nonsmokers are exposed worldwide to passive smoking at home, at work and in public places.

  17. Page 179
    Correspondence: Annunziata Faustini, Dept of Epidemiology, Regional Health Service, Lazio Region, via di Santa Costanza 53, 00198 Rome, Italy. E-mail:

    The effects of air pollution on health became evident after episodes of extraordinarily high levels of airborne pollutants in the 1930s. Thereafter, research addressed the health effects of current levels of outdoor air pollution, which made it possible to implement monitoring of air pollutants and policies to reduce the impact on health. Worldwide, studies have assessed an increase in mortality, mainly due to respiratory and cardiovascular diseases, associated with long- and short-term exposure to air pollutants (especially particulate matter). Evidence of the effects on morbidity is increasing. The findings clearly indicate the need to reduce population exposure to air pollutants in order to protect human health. Some scientific issues related to exposure assessment and biological mechanisms of damage remain unaddressed. We provide a brief synthesis of the current knowledge of the effects of particulate matter, with specific attention given to the most innovative contributions.

  18. Page 198
    Correspondence: Marzia Simoni, Pulmonary Environmental Epidemiology Unit, CNR Institute of Clinical Physiology, Via Trieste 41, 56126 Pisa, Italy. E-mail:

    Daily activity requires individuals to spend as much as 90% of their time indoors. Some pollutants present in larger quantities indoors than outdoors, and with long-term exposure, even low concentrations of pollutants may have a significant biological impact. Indoor air pollution has been related to a higher risk of upper and lower respiratory tract irritation, respiratory and allergic symptoms and diseases, lung function reduction, BHR, and lung cancer.

    Nearly half of the world’s population uses biomass fuels for cooking, heating and/or lighting. Solid-fuel smoke causes huge health problems in developing countries but wood-burning populations in developed countries may also be at risk. According to the World Health Organization, 4.3 million people died from exposure to household air pollution (HAP) in 2012 (7.7% of the total annual global burden of disease). Globally, HAP ranks third among risk factors for DALYs. Air quality is particularly important for frail subpopulations (i.e. children, elderly people, and subjects with cardiorespiratory diseases or socioeconomic deprivation).

  19. Page 211
    Correspondence: Dick Heederik, University Utrecht, IRAS EEPI Division, PO Box 80178, Utrecht, Utrecht, 3583 VH, The Netherlands. E-mail:

    Occupational respiratory diseases contribute considerably to the total burden of respiratory diseases. The occurrence of classical respiratory occupational diseases, such as pneumoniconioses, which are related to mining and mineral processing industries, has declined in most countries because of exposure control and automation. However, these diseases are still being observed in the construction industry due to the use of mechanical hand tools. In general, the role of asthma and COPD has become more dominant over the past decades. Lung cancer and mesothelioma generally occur after an active working life, but occupational exposure clearly contributes to their occurrence. The burden of these diseases related to occupational exposures, and expressed in DALYs, are amongst the highest worldwide. The need to control exposure by setting standards for levels of pollutants in the work environment should continue to receive attention and priority.

  20. Page 224
    Correspondence: Foteini Malli, University Hospital of Larissa, Respiratory Medicine Dept, Biopolis (Mezourlo), Larissa, 41110, Greece. E-mail:

    Several lung diseases may be related to dietary intake changes. Dietetic hypotheses have primarily related antioxidants, vitamin D, polyunsaturated fatty acids and dietary patterns (mainly the Mediterranean diet) with the pathophysiological events related to various lung diseases. Asthma has been related to reduced antioxidant intake and decreased omega-3 to omega-6 ratio; however, the beneficial effect of nutrient supplementation in asthma remains to be proven. Vitamin C may be related to better lung function, while β-carotene intake has been associated with increased lung cancer incidence in smokers and should be used with caution. Active TB risk is increased in subjects with vitamin D deficiency, but the role of vitamin D manipulation in TB prevention and treatment remains to be established. A Mediterranean diet is unequivocally associated with reduced lung cancer incidence. Further high quality investigations are needed in order to substantiate the nature and the importance of various nutrients as risk factors for lung health and disease.

  21. Page 241
    Correspondence: Valeska Padovese, National Institute for Health, Migration and Poverty, Via San Gallicano 25/a, 00153 Rome, Italy. E-mail:

    Poverty deprives individuals of the freedom to satisfy basic needs and rights. The poor are exposed to greater personal and environmental health risks, are less well nourished, are less informed about health, and are less able to access healthcare; they thus have a higher risk of illness and disability.

    We reviewed scientific literature and disclosed unpublished data concerning health conditions of migrants, the homeless and Roma people in Europe with particular regard to respiratory conditions, i.e. the re-emergence of TB related to the influx of migrants from highly endemic countries; the prevalence of respiratory allergies and acute respiratory infections in migrants, residents and Roma people due to unhealthy environments; and COPD in homeless people related to high nicotine consumption. Furthermore, we discuss public health policies addressing these fragile segments of the population and propose inclusion models.

  22. Page 249
    Correspondence: Riccardo Pistelli, Dept of Respiratory Physiology, Complesso Integrato Columbus, Università Cattolica del Sacro Cuore, Via Moscati 31, 00168 Rome, Italy. E-mail:

    Epidemiological research is carried out by both experimental and non-experimental study designs. Experimental studies are rarely used in epidemiology. These studies are performed under the control of the investigator, who can assign treatment (exposure) to patients via randomisation in order to achieve quite similar experimental subgroups and avoid bias, due to the effect of unbalanced factors. However, an experimental study often cannot be carried out in all areas of research, epidemiology being one, for technical or ethical reasons.

    In non-experimental studies, the investigator purely acts as an observer. They do not control the exposures, only the selection of included subjects and the proper treatment of confounding factors. There are several non-experimental study designs, the most popular of which will be described in this chapter, along with their specific targets, strengths and weaknesses. For each design, an example from the current literature will be used to analyse some methodological issues.

  23. Page 257
    Correspondence: Francesco Pistelli, CardioThoracic and Vascular Dept, University Hospital of Pisa, via Paradisa 2 - Cisanello, Pisa 56124, Italy. E-mail:

    Questionnaires and lung function testing are important instruments of measurement in respiratory epidemiology.

    Questionnaires allow the investigation of large population samples through the collection of subjective data. Their basic methodological requirements are validity, reliability and bias. The first standardised respiratory questionnaire was developed by the British Medical Research Council in the 1950s, and was the basis of subsequent questionnaires developed in Europe and the USA. More recently, ad hoc questionnaires for specific studies have been produced, even though existing validated questionnaires still appear to be largely used.

    Lung function testing has several applications in respiratory epidemiology. Standardised procedures have been developed to minimise variability of lung function testing measurements, and updated reference equations have been derived to improve interpretation of results. Spirometry is the most frequently used test in population-based studies but other tests, including diffusing capacity or methacholine challenge test, have been shown to be applicable in respiratory epidemiology.

  24. Page 273
    Correspondence: Christer Janson, Dept of Medical Sciences, Respiratory Medicine and Allergology, Uppsala University, 75185 Uppsala, Sweden. E-mail:

    BHR is one of the hallmarks of asthma. In asthmatic patients, BHR is often regarded as a consequence of airway inflammation, but BHR is also found in other disorders such as COPD, where the association with airway inflammation is less clear. The aim of this chapter is to investigate how allergy and inflammation, together with nonallergic exposures such as smoking, are related to BHR.

    The median prevalence of BHR was 13% in the European Community Respiratory Health Survey, but the study also showed an eight-fold variation in BHR between the centres. Urban living, female sex, higher age and obesity have been associated with a higher prevalence of BHR. Allergic sensitisation to perennial allergens, such as cats and mites, and smoking increase the likeliness of having BHR. BHR is related to increased levels of eosinophilic inflammatory markers in blood and sputum and to increased FeNO. The association between FeNO and BHR is, however, only found in nonsmokers and in subjects who are atopic. BHR is related to neutrophil inflammation in COPD.

    We conclude that allergy and smoking are the main determinants of BHR, but the type of inflammation in BHR associated with smoking is different from the type of inflammation in BHR associated with allergy. BHR that is associated with allergy is related to eosinophilic inflammation and high FeNO, while neutrophilic inflammation is important in smoking-associated BHR.