Abstract
The burdens of chronic obstructive airway diseases among the elderly in Europe, and worldwide, are increasing. Although asthma is common in all ages, the main airway disease affecting the elderly is chronic obstructive pulmonary disease (COPD). The aim of this paper is to review the prevalence and incidence of COPD on the basis of population studies. As the prevalence estimates of asthma are probably well known, only the incidence and remission of asthma will be discussed. The underdiagnosis of obstructive airway diseases is huge. A Dutch programme for early detection of obstructive airway disease among the elderly has, thus, been included in the presentation.
A prerequisite for fighting COPD is to acquire data on illnesses and death. COPD has only recently been defined by cut-off points of spirometric outcomes, which is why measures of the prevalence of COPD have been distorted by use of a large number of different diagnostic terms and lung function criteria. The prevalence of clinically-relevant COPD has been estimated in several community studies to 4–6% in adult population samples, with a considerable increase by age, particularly among smokers. The incidence of COPD not only increases heavily with age and smoking, but also occupational exposure to dust, gas and damp. Precise estimates of the incidence of COPD or spirometric airflow limitation are not available.
Demographic changes will result in a further substantial increase of chronic obstructive airway disorders, mainly chronic obstructive pulmonary disease, among the elderly. The increasing burden of chronic obstructive pulmonary disease has to come to the awareness of the public, governments, health authorities, and industry.
Chronic obstructive pulmonary disease (COPD) is a common, costly and preventable disease that has great implications for health 1–5. After the ages of 40–50 yrs, its prevalence increases considerably with increasing age, particularly among smokers; however, the underdiagnosis is great 6, 7. The slowly progressive nature of COPD 3 means that the disease usually remains undetected for many years, and most patients are first identified when they present with an exacerbation. By the time COPD is diagnosed, often ≤50% of lung function has already been lost and the need for healthcare utilisation is high 8. Though the incidence of asthma is large in children and teenagers, the prevalence of asthma remains relatively stable and may also increase slightly among the elderly, in whom it is around 5–10% in most European countries 9. Both asthma and, in particular, COPD are chronic and common diseases among the elderly, although asthma may remit 10. The burdens among the elderly from obstructive airway diseases are increasing.
In the last decades, a number of studies have shown an increasing prevalence of obstructive airway diseases worldwide 9, 11, 12. It is expected that in the near future demographic changes will result in a further substantial increase of chronic obstructive airway disorders. The main part among the elderly is expected to be due to COPD.
Dutch general practice data from the Continuous Morbidity Registration show that, over time, incidence has remained stable, whereas prevalence has more than doubled 13. As prevalence follows population demography, a clear increase can be expected for the period 1996–2020, levelling off in the period 2030–2050 14.
Within the perspective of obstructive airway disease among the elderly, a review of the current discussion of the criteria for diagnosing COPD, the burden of COPD in terms of prevalence, in part incidence, disability and mortality, has been made. Regarding asthma, mainly incidence has been reviewed. As the underdiagnosis of obstructive airway disease is huge, a Dutch programme for the detection of COPD in society is included.
Methodological considerations
Diagnostic criteria of chronic obstructive pulmonary disease
Because the term COPD is used in a variety of ways, data on prevalence, incidence, mortality and morbidity from COPD must be interpreted with caution. The diagnostic term COPD has not, in contrast to asthma, been widely used and distributed to the public by physicians and other health professionals. Asking a general population about a physician's diagnosis of COPD will generate few affirmative answers from those with irreversible airflow limitation. Most patients will answer that they have asthma, chronic bronchitis, emphysema, or the disease is unrecognised 6, 7. A prerequisite for the diagnosis is spirometry, which is still insufficiently used.
The CIBA Guest Symposium in the late 1950s defined generalised obstructive lung disease as a condition with widespread narrowing of the airways, at least on expiration, causing an increased resistance to airflow. They divided the obstructive lung disease into two subgroups: 1) intermittent or reversible obstructive lung disease, or asthma, and 2) irreversible or persistent obstructive lung disease, later labelled as COPD 15.
Several international and national guidelines have recommended diagnostic criteria for COPD. Operative criteria for the cut-off of the forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) or FEV1/vital capacity (VC) ratio may, however, differ between recommendations. The British Thoracic Society (BTS) 3 and the Global Initiative for Obstructive Lung Disease (GOLD) 4 have recommended a simple cut-off point with a FEV1/FVC ratio of <0.7, while the European Respiratory Society (ERS) 1 has recommended a FEV1/VC ratio of <88% predicted in males and <89% pred in females. The GOLD criteria are easily applicable, but they may cause bias by including too many cases with the elderly and too few with young adults.
The severity of COPD is defined by level of FEV1 in % pred. The grading of severity varies considerably among the recommendations. The grading of the BTS 3 seems clinically appropriate, with severe COPD as FEV1 levels <40% pred, moderate COPD as FEV1 ≥40<60% pred, and mild COPD as FEV1 ≥60<80% pred. No agreement exists on what prediction equations should be used when standardising for age, sex and body size. The recommended summary equations of the ERS give predicted FEV1 as 300–400 mL lower than that observed in healthy nonsmoking Europeans 16. Using these summary equations will, therefore, underestimate the prevalence of COPD in most European populations.
Measures of incidence and remission
When measuring incidence, two terms are used. Cumulative incidence is the proportion of a population that becomes affected by a disease during a defined time period. When studying diseases with low incidence and remission, the incidence rate for a time period may be approximated with the cumulative incidence for the period. The incidence is usually given as the number of new cases per 1,000 persons at risk per year. Persons at risk are the subjects not having the studied condition at the start of the observation period.
Regarding remission of asthma, with the exception of asthma in children, only a few articles have been published thus far, and only one focusing on the elderly 10. The definitions of remission and the methods used vary. COPD, with established lung function impairment defined by today's criteria, may, by definition, not remit. Chronic lung disease, formerly defined by symptom criteria (e.g. chronic nonspecific lung disease (CNSLD)), may remit 17.
Disability adjusted life years lost
Disability adjusted life years (DALYs) have been launched by the World Bank and supported by the World Health Organization (WHO) as a measure of burden of disease. Thus, DALYs are the sum of years lost because of premature mortality and years lived with disability, adjusted for the severity of the disability. The disability weight for untreated COPD was estimated to be 0.43, using a person trial off method 12. Death is given the disability weight of 1.0, and complete health the weight of 0.0 (fig. 1⇓). Estimated DALYs for the world are calculated to 29 million DALYs in 1990, of which 2.3 million are in countries that have an established market economy 12.
Disability adjusted life years lost (DALYs) and years of healthy life lived (QALYs) are complementary concepts. Both approaches multiply the years (x‐axis) by the quality of those years (y‐axis). DALYs use disability weights to reflect the burden of disease. The average disability weight of severe chronic obstructive pulmonary disease (COPD) is in this example estimated to be 0.43, the duration of disease to be 8 yrs, and the life expectancy to be 80 yrs.
Burden of chronic obstructive pulmonary disease
Prevalence of chronic obstructive pulmonary disease
The prevalence of COPD has been distorted by the use of different diagnostic terms and lung function criteria. Furthermore, a meaningful evaluation of estimates demands knowledge of age distribution and smoking habits in the examined populations.
An Italian general population showed very large differences in prevalence estimates of COPD, when using the 1986 American Thoracic Society (ATS) criteria, the 1995 ERS criteria or “clinical criteria” 18. For the last 40 yrs, the prevalence of COPD has been estimated in community surveys in several Nordic countries 7, 19–23. Studies from the last two decades indicate that 4–6% of the adult population suffer from clinically-relevant COPD. Prevalence increases heavily with age, and recent surveys have shown only small differences between males and females.
Prevalence studies in the UK have been reviewed by Strachan 24, who stated that only one national study of ventilatory function had been conducted. This study of 5,547 males and females aged 16–65 yrs showed that a FEV1 of ≥2 standard deviations (sd) below the age and height predicted value was present in 10% of the males and 11% of the females. A Norwegian spirometry survey of a general population sample aged 18–73 yrs found a prevalence of 6% of airflow limitation, defined as FEV1/FVC <0.7, 12% had FEV1 <80% pred using Norwegian reference values 19, and 4.5% had both the ratio and FEV1 <80% pred 21. Using stages advocated by the BTS 3 only 0.2% of the adult population could be defined as severe cases of airflow limitation having FEV1 <40% pred (fig. 2⇓). A spirometry survey in the Po River delta area in Italy 18 in a random sample aged 25–73 yrs yielded spirometric prevalence estimates that are similar to those obtained in the Nordic countries. Approximately 0.5% of the population could be defined as having severe COPD using the cut-off point of FEV1 <50% pred.
Percentage distribution of the prevalence of a) chronic obstructive pulmonary disease (forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) <0.7 and FEV1 >80% pred) and b) restrictive or other lung function impairment (FEV1/FVC ≥0.7 and FEV1 <80% pred) in a general population aged 18–73 yrs in Bergen, Norway. □: 18–44 yrs; └: 45–73 yrs.
In a study carried out in random population samples aged 40–69 yrs in seven different areas of Spain 25, COPD was defined as FEV1/FVC ratio <88% pred in males and <89% in females, with a <12% increase in FEV1 at bronchodilator test. An overall prevalence of 10.6% for COPD was observed. In a USA population survey of 16,695 subjects aged 17–89 yrs in 1988–1994, airflow limitation was defined as in the Norwegian study 21 and in the criteria of the BTS for COPD 3. The overall prevalence of a low lung function was 6.8%, while 8.5% reported previous or current diagnosis of obstructive lung disease 26. Data from the same study indicate that 50% of smokers may develop COPD 27, similar to recent Swedish data 23.
The third National Health and Nutrition Examination Survey (NHANES III) and other national healthcare surveys in the USA suggest that between 2.4–7.0 million people are actually diagnosed with COPD 27. However, the National Center for Health Statistics estimates that COPD affects >16 million US citizens 28. The model, which has been developed from the NHANES III to estimate prevalence of COPD from the known smoking status of the population, estimated the cases of COPD to be 1.5 million persons with COPD in Spain, 3.0 million persons in the UK, 2.7 million persons in Germany, 2.6 million persons in Italy, and 2.6 million in France 27.
The Global Burden of Disease Study has estimated the worldwide prevalence of COPD to 834 per 100,000, which yields ∼44 million cases of COPD 29. In the regions of countries with established market economies, the prevalence rate was estimated to be as low as 535 per 100,000. This estimate includes all ages. Since most cases of COPD usually become clinically apparent after the age of 45 yrs, the study grossly underestimates the prevalence of the disease in adults and the elderly. These estimates may be too low in countries with a life expectancy of ≥80 yrs and with >20% daily smokers.
Estimates of incidence of chronic obstructive pulmonary disease
Precise estimates of incidence of physician-diagnosed COPD or spirometric airflow limitation as defined by modern guidelines, are so far not available. Longitudinal studies of lung function 30–32 have shown that the development of airflow limitation is heavily dependent on smoking habits and dust exposure. In most cases, the decline seems to occur along a slowly accelerating curvilinear path.
A limited number of studies have performed incidence measures of COPD; however, in these measures the cut-off levels of the outcome variable differ largely. A Finnish study 30 of a population aged 40–64 yrs in 1961 was re-examined in 1971. The average annual incidence of a spirometric ratio of FEV1/FVC <0.6 was 0.2% for the whole population and 1.0% in smokers. In a 13-yr follow-up study, conducted in a population aged 19–70 yrs in Cracow, Poland 33 between 1968–1981, COPD was defined as FEV1 <65% pred. The average incidence per year was 0.5%. The incidence was twice as high in males than in females, and a considerable number of new cases of COPD was also seen in never-smokers.
In the Netherlands, the incidence of CNSLD (mainly defined by symptomatic criteria) was studied from 1965–1985 in the Zutphen community population of males aged 40–59 yrs. The average incidence was estimated at 1.5% per year 17.
Within the Global Burden of Disease Study 29, estimates were made of incidence of COPD in eight health regions of the world. The estimate for countries with market economies was calculated to be 84 per 100,000, which is less than one-fifth of that observed for Cracow, Poland 33.
Disability
Chronic obstructive airway disorders are also a growing cause of disability. In 1990, COPD was the 12th most burdensome condition in the world. According to the projection from the 1990 WHO Global Burden of Disease Study, COPD will rank 5th in 2020, being responsible for 4% of total DALYs lost 34. Only ischaemic heart disease, major depression, traffic accidents and cerebrovascular disease will cause a greater burden.
In 1990, the WHO Global Burden of Disease study showed COPD to be the sixth leading cause of death worldwide 34. By 2020, COPD is expected to rise to the third position, after ischemic heart disease and cardiovascular disease. In 1990, COPD caused 4% of deaths (>2 million) worldwide; in 2020, this figure is projected to rise to 7% (4–5 million).
Asthma
Incidence
Studies reporting the incidence of asthma among adults show a wide range of results, and vary from 0.4–11·1,000 persons−1 under risk per year, indicating that the results are largely dependent on the methods used, including the definitions of incident asthma and the population under risk as well 35. In contrast to adulthood and old age, the incidence of asthma in childhood is considerably higher and has been estimated at 7–10·1,000 persons−1·yr−1 from 7–10 yrs of age, and the incidence decreases slowly with increasing age 36.
Prospective studies of samples of the general population, although still few, have reported incidence rates of 2–4·1,000 persons−1·yr−1 37 and 2–8·1,000−1·yr−1 38. Lower incidence rates of ∼1·1,000−1 have been found in other studies 39–41. Register studies generally provide even lower incidence rates. In a Finnish study 42, the incidence was estimated prospectively during a 6‐yr period by two repeated questionnaires, and also from a medical register in the same sample. The population study-based measure of incidence was 1.5·1,000 persons−1·yr−1, while the incidence calculated from hospital admissions was one-half of that. Retrospective studies also provide lower incidence rates than prospective studies. Incidence of asthma (or wheeze) measured prospectively in a cohort from birth to age 33 yrs was 13·1,000−1·yr−1, but recalling asthma or wheeze when interviewed resulted in an incidence of 9·1,000−1·yr−1 43. A recent Swedish retrospective study reported the incidence to be lower in the past, which may reflect an increase of asthma, or an effect of forgetfulness, or both 44. Adults often forget their childhood asthma.
The incidence of asthma among the elderly has been estimated, not only in the USA, but also among the middle-aged and elderly, by longitudinal studies during 10 yrs within the Obstructive Lung Disease in Northern Sweden Studies (OLIN). After exclusion of all subjects with respiratory disease or symptoms common in asthma at the start of the observation period, the incidence of asthma was estimated at 2.3·1,000−1·yr−1. The incident cases were confirmed by clinical examinations including methacholine tests. Risk factors for incident asthma were a family history of asthma, current and exsmoking, and female sex. Smoking, as a risk factor, behaved differently compared with results from cross-sectional studies in the same cohorts. The incidence of wheezing and of new users of asthma medicines was both 11·1,000 persons−1·yr−1, indicating that the incidence of asthma may have been underestimated 35.
Remission
Approximately one-half of the asthmatic children may outgrow their asthma before adulthood; however, asthma may recur later in life. Only a few studies that allow conclusions about the general population have focused on remission of asthma in adults. The criteria for remission have varied. A study in Tucson, AZ, USA, found remission of asthma to be 22% during 9 yrs. The remission rate was highest (65%) in subjects aged 10–19 yrs 45. In a 25-yr follow-up study in the Netherlands, different outcomes of remission of asthma in asthmatics aged 13–44 yrs at baseline were measured. After 25 yrs, 21% did not have bronchial hyperresponsiveness (BHR), while 11% were in remission if normal lung function and freedom from symptoms also were required 46. The remission of asthma in a population-based asthma cohort of the middle-aged and elderly within the Swedish OLIN studies was found to be 6% during 10 yrs. Remission was defined as no symptoms without using asthma medicine 10.
Screening of obstructive airway disease: the Dutch DIMCA study
In 1990, it was hypothesised that the maximum benefit of inhaled corticosteroids (ICS) treatment, such as moderating the course of lung function decline, could be expected if treatment was instituted as early as possible in the development of the disease. This concept resulted in the establishment of the Dutch Detection, Intervention and Monitoring of COPD and Asthma in family practice (DIMCA) project in Nijmegen in 1991 (fig. 3⇓) 47. The study was designed to be representative for the Dutch general population. Because of the specific nature of the Dutch healthcare system, virtually all Dutch citizens are automatically registered at a general practice and receive ongoing care. In the DIMCA study, a random sample was selected from a screening of a 10% sample of the population and recruited for the trial if persistent signs and/or symptoms could be demonstrated over time. The subjects were taken from the practice lists of 10 general practices in the eastern area of the Netherlands. The sample was divided into an intervention and a control group.
The Dutch Detection, Intervention and Monitoring of COPD and Asthma in family practice (DIMCA) study design.
The study had two objectives: 1) to assess the efficacy of ICS treatment of early obstructive airway disease; and 2) to assess the effectiveness of early case-finding and early treatment.
Subjects in the general population aged 25–70 yrs with signs of COPD or asthma at an early stage were detected. This was done by a two-stage protocol involving screening and a subsequent 2-yr monitoring of all subjects with positive results in the screening. Subjects were excluded from screening in case of diagnosed asthma or COPD, other lung diseases, congestive heart failure, serious morbidity (life expectancy <4 yrs), a severe physical or mental handicap, or corticosteroid dependency (inhaled or oral). A total number of 1,749 subjects met the inclusion criteria, and 1,155 subjects (66%) agreed to participate in the screening. Screening consisted of lung function testing and assessment of current symptoms.
Measurement included assessment of symptoms, exacerbations, smoking behaviour, health-related quality of life, functional status, allergy, lung function, reversibility, BHR, work disability and healthcare cost. At the 10-yr follow-up in 2001, two independent clinicians allocated all subjects as: 1) evidently developed asthma or COPD; 2) at risk of developing such a disorder; or 3) not at risk for development of asthma or COPD.
The screening procedure is described in table 1⇓. The study covers the whole range, from detection of formerly undiagnosed, monitoring the natural course of the early stage of disease, and early treatment of subjects that are considered to be “at risk”, towards identification of prognostic factors for the development of COPD or asthma.
Screening procedure in the Dutch Detection, Intervention and Monitoring of COPD and Asthma in family practice (DIMCA) study
Symptoms or signs of obstructive airway disease during the screening (“positive”) were found in 604 subjects (52%). Of these, 384 subjects (64%) agreed to participate in a 2-yr monitoring programme, consisting of periodical lung function measurements and assessment of symptoms and exacerbations.
During this stage (table 2⇓), three mutually exclusive groups were distinguished. The first consisted of subjects with persistently reduced lung function or increased BHR. Once identified, these subjects were withdrawn from monitoring. Subjects with a rapid decline of lung function with signs of BHR were assigned to the second group. In these groups, ethical considerations prompted intervention. Subjects with a moderate increase in the decline in lung function or signs of BHR were assigned to the third group. Subjects who completed the two stages of the detection programme, but did not meet the aforementioned criteria were regarded as “healthy”.
Detection procedure in a 2‐yr monitoring programme in the Dutch Detection, Intervention and Monitoring of COPD and Asthma in family practice (DIMCA) study
During the monitoring, 252 subjects were detected with signs of obstructive airway disease at an early stage. Although early airway obstruction could be identified, a final diagnosis of either asthma or COPD was often not possible. As no evidence of biased recruitment or selection was found, the proportion of subjects with signs of airway obstruction at an early stage could, therefore, be extrapolated in the general population 47. It was concluded that 7.7% of the general population showed a persistently reduced lung function or increased BHR, another 12.5% showed a rapid decline in lung function (>80 mL·yr−1) in combination with BHR, and a further 19.4% of the general population showed mild objective signs of obstructive airway disease.
The two-stage detection programme 47 revealed that a large proportion of the general undiagnosed population showed symptoms and objective signs of obstructive airway disease. As found by others 19, ∼50% of the general population had respiratory symptoms or signs of obstructive airway disease at the screening. The monitoring stage of the programme further showed that in ∼20% of this general population, a persistently decreased lung function or a rapid decline in lung function was observed, and in these subjects, treatment should be considered 48, 49.
Subjects, although experiencing respiratory complaints, only consult their general practitioner when the quality of their everyday life becomes affected, or when they experience variability in lung function 50. The large number of undetected and untreated subjects raises the question of practice case-finding or screening. However, potential health gains of screening should be set against the costs inflicted and the potential negative effects, including medicalisation with overconsumption of medical facilities. Furthermore, the over-reliance on medical intervention is a potential adverse effect of screening.
Conclusions
The burdens among the elderly from obstructive airway diseases, particularly chronic obstructive pulmonary disease, are increasing. Considerable underdiagnosis exists. The public, governments, health authorities, and industry must become aware of the increased burden of chronic obstructive pulmonary disease, the only leading cause of death that is increasing in prevalence worldwide. Health professionals should implement strategies for early detection, effective prevention, diagnosis and treatment of the disabling and life-threatening obstructive airway diseases, like the Dutch Detection, Intervention and Monitoring of Chronic obstructive pulmonary disease and Asthma in family practice (DIMCA) study whose follow-up has been started in 2001.
- Received April 12, 2002.
- © ERS Journals Ltd
References
