Asthma is a relatively common chronic disease affecting approximately 10% of the population (1). Literature on the epidemiology of asthma focuses primarily on prevalence, morbidity (usually measured by hospitalizations), and mortality. It appears from this research that these measures may be showing an increase throughout the world; however, there is debate over whether or not these changes are real. In order to better understand the nature of the disease, its burden on society, and what can be done to prevent the disease and its resultant morbidity, we must have a solid understanding of its epidemiology.
The information included here is intended to provide an overview of the epidemiology of asthma and the factors that may be contributing to its development, morbidity, and mortality. Application of this epidemiologic research in preventing avoidable morbidity and mortality will also be presented.
What is asthma? Although this may seem like a fairly simple question to answer, there has been much controversy surrounding the definition of asthma. A large part of this stems from the fact that although there are certain pathologic features that are found in many asthmatics, there are none that apply solely to all patients diagnosed with asthma. The National Asthma Education Expert Panel Report (2) recommends the following clinical definition of asthma:
"Asthma is a lung disease with the following characteristics: 1) airway obstruction that is reversible (but not completely so in some patients) either spontaneously or with treatment; 2) airway inflammation; and 3) increase airway responsiveness to a variety of stimuli."(2)
A clinical description of asthma is not the same as an epidemiologic definition of the condition. Epidemiologic research of asthma often relies on self-reporting of the asthma diagnosis and/or reporting of common symptoms such as wheezing; however, there are problems associated with relying on this diagnosis for such research. In general, asthma is underdiagnosed in the population (3,4). Asthma and chronic bronchitis are overlapping diagnoses among children. Wheezing tends to be the predominant symptom; however, if it is mild, infrequent, or happens only with infection, the physician may not make the diagnosis of asthma. Some asthmatics may only experience a cough and no wheezing. In addition to possibly not being diagnosed as asthmatic, such cases may also be excluded from studies focusing exclusively on the report of wheezing. Another factor to consider is that parents may underreport asthma or its symptoms in children (3,4).
Assessment of bronchial reactivity of the airways to exercise or pharmacologic agents is another commonly used method of identifying cases of asthma. However, this lacks specificity and sensitivity for asthma--up to 50% of those with bronchial reactivity have neither symptoms nor diagnosed asthma, and up to 50% of those with asthma do not have bronchial reactivity (3).
Incidence rates, measures of the onset of a disease, provide information about the probability of developing the disease. The incidence rate of asthma has been estimated to be between 2.65 to 4/1000 per year. In childhood the onset of asthma is more common among children less than five years of age and among boys versus girls. In this age group, incidence rates have been estimated to vary between 8.1 to 14/1000 per year for boys and between 4.3 to 9/1000 for girls. Sex differences tend to disappear in adulthood. The incidence of asthma in persons over 25 years of age has been estimated at 2.1/1000 per year (3)
When defining asthma in epidemiologic studies, the prevalence of asthma can be described in a number of ways. It is important to be aware of this when interpreting results of these studies. Point prevalence (asthma at the time of the study survey), although the easiest measurement of prevalence to obtain, will likely underestimate asthma's presence within the population. Lifetime prevalence (asthma at any time in a person's life) presents the problem of recall bias. In general, the most commonly accepted measurement used is period prevalence, usually over the past 12 months (5).
Given differing definitions and methods of identifying cases, it is difficult to interpret variations in asthma prevalence both over time and between geographic areas. In looking at the review of studies presented by Cookson (6), Gergen and Weiss (3) came to several conclusions about the international prevalence of asthma:
"The point prevalence of current (or active) asthma for children in 'industrialized' countries ranges between 0.7% in Tokyo, Japan to 8.5% in Tucson, Arizona, US and averages about 5% worldwide. Rates from 'nonindustrialized' countries vary from high levels in the Western Caroline Islands, 49%, to 0.6% in South Fore, Papua New Guinea. However, rates for 'nonindustrialized' countries tend to be lower."
The prevalence of asthma has been reported to be increasing worldwide, but this apparent trend continues to be debated. Changes in prevalence can be confounded by factors such as: differing levels of awareness of the disease by health care providers and/or families; changes in access to medical care; and changes in medical diagnosis. Within the same population the prevalence of asthma can vary to a significant degree based on the definition used and questions asked of the study population. The following table shows the variance in estimates of asthma prevalence from the National Health and Nutrition Examination Survey (NHANES) and the National Health Interview Survey (NHIS)--national surveys monitoring the health of the United States population (3). Notice that within the same population the highest estimate of prevalence is over three times that of the lowest depending on the question asked.
Estimates of asthma prevalence: results from two national population-based surveys.(From Evans et al(10))
|Survey and questionnaire||Rate per 100 population|
(persons 6 months-74 years, 1976-1980)
|Did a doctor ever tell you that you had asthma?||06.2|
|Do you still have asthma?||03.0|
|During the past 12 months, not counting colds
or the flu, have you frequently had trouble with wheezing?
|Has a doctor ever told you that you had asthma
|Do you still have asthma and/or wheezing?||07.7|
(all ages, 1979-1981)
|During the past 12 months, did you have asthma?||03.1|
The NHANES data can be used to examine trends in the reported cumulative prevalence (ever having asthma and/or frequent wheezing within the year) among six to 11-year-old children in the U.S. One should note, however, that survey questions have changed, and NHANES 2 included questions about allergy which may have resulted in an increased ascertainment of asthma cases. From this data, there appears to be a significant increase (p<0.05) in asthma prevalence in this age range from 4.8% in NHANES 1 (1971-1974) to 7.6% in NHANES 2 (1976-1980) (7)
In Finland, Sweden, Australia , new Zealand, and England asthma prevalence has been reported to be increasing. A retrospective analysis of the prevalence of asthma in cohorts of 19-year-old Finnish candidates for military conscription (representing 98% of the male Finnish population at this age) showed prevalence increasing from 0.29% in 1966 to 1.79% in 1989. Asthma here was defined as asthma recognized by a physician at call up examination, exemption from military service because of disabling asthma, and discharge during the year of required military service because of asthma (8). In 18-year-old Swedish military conscripts, the prevalence of asthma increased from 1.9% in 1971 to 2.8% in 1981 (9).
Using a questionnaire, Robertson et al. (10) determined the prevalence of reported asthma or wheezing in the past 12 months in Melbourne Australian children aged 7, 12, and 15 years and then compared this to the prevalence of a history of asthma as determined 26 years earlier in a 1964 survey of 7-year-old Melbourne schoolchildren. Their analyses showed a 141% increase in asthma prevalence among 7-year-olds, from 19.1% in 1964 to 46% in 1990. The authors noted that part of this increase may have been due to increased awareness of both asthma and the symptom of wheeze by health care providers and the public. Peat et al. (11) conducted serial cross-sectional studies in 1982 and 1992 of 8-10 year old children in two Australian towns (Belmont and Wagga Wagga). Using a self-administered questionnaire, they found significant increases in both populations in the prevalence of doctor diagnosed asthma, recent use of an asthma drug, and episode of wheeze. A histamine inhalation test was included as an objective measurement of airway responsiveness, and those children with both recent wheeze and airway hyperresponsiveness were classified as cases of asthma. The prevalence of airway hyperresponsiveness increased significantly in Belmont (10.7%, 95%CI 7.3-14.1, p<0.001) and in Wagga Wagga (6.4%, 95%CI 2.9-9.9),p<0.05). Likewise, the prevalence of current asthma increased in Belmont (7.5%, 95%CI 4.9-10.1, p<0.001) and in Wagga Wagga (2.8%, 95%CI 0.1-5.5,p<0.05).
Prevalence studies from England are not consistent in their results. In reviewing asthma prevalence studies conducted in the United Kingdom between 1964 and 1986, Anderson (4) concluded that there is "little evidence" of a trend in estimates of prevalence of wheezing (described as 'current' or 'recent'). Prevalence of diagnosed asthma did show a tendency to increase in recent years. In a longitudinal study of English children 4 to 12 years of age, Burney et al. (12) found significant increases over a 13-year period (1973-1986) in reported asthma (p<0.001) and persistant wheeze (p<0.001) in both boys and girls.
Geographic comparisons can provide clues about factors possibly contributing to the development and/or promotion of asthma. However, as was stated earlier, comparisons of studies can be limited by differences in research methodologies. A few studies have employed standardized methodologies in order to provide a comparative analysis of prevalence in different countries. The prevalence of asthma in children in New Zealand and Australia has been reported to be higher than that found in other countries (3). One study compared the prevalence of asthma in 12-year-olds in New Zealand and South Wales. Using the same questionnaire and an exercise provocation test in both countries, this study found New Zealand children to have a higher prevalence of reported history of asthma or wheeze and were more likely to have a positive exercise provocation test (12). Burr et al. (13) also used a similar questionnaire and exercise challenge test in their comparative survey of childhood asthma in New Zealand, Wales, South Africa, and Sweden. Their data showed that children in New Zealand had the highest prevalence of asthma ever, current asthma, and wheeze without a cold. Swedish children had the lowest prevalence of asthma and asthma-like symptoms (14).
A comparison of reported asthma mortality rates in 20 countries during the period of 1985-1987 showed a variation from over 9/100,000 in West Germany to less than 2/100,000 in the Netherlands, the U.S., and Hong Kong (14). Throughout the world asthma morbidity and mortality appear to be increasing (4,15,16), and as with prevalence, the reasons for these trends are not clear. Defining and identifying the disease again are obstacles in understanding trends. Various factors hypothesized to be contributing to increasing mortality trends include: 1) changes in 1979 in the ninth revision of the World Health Organization International Classification of Diseases that resulted in "asthma with bronchitis" being coded to "asthma" in ICD9 versus "bronchitis" in ICD8; 2) changes in physician diagnostic patterns, such as over-reporting asthma deaths over age 50 and under-reporting in younger ages; 3) increased diagnosis of asthma; 4) increased prevalence and/or severity of the disease; and 5) adverse drug effects (4)
With the change from ICD8 to ICD9 came an increase in asthma mortality rates in 1979. The effect of this change was evaluated in the UK where it was determined that the expected increase in coding of asthma deaths in those under 45 years would be only 6% (16). The effect was similarly evaluated in New Zealand 5-34 year olds, and an expected increase of just 2.4% in coded asthma deaths was calculated (17). Additionally, trends from 1980 to 1987, after the coding change, show a continued increase in many countries, such as Italy, Denmark, Israel, Australia, France, and the United States. New Zealand, Sweden, and Japan experienced decreased mortality during this time period (15).
In the United States an analysis by the Centers for Disease Control and Prevention (CDC) shows the overall annual age-adjusted death rate for asthma increased 40%, from 13.4/1 million in 1982 to 18.8/1 million in 1991 (mortality data was not available for 1992). In 5-34 year olds, this rate increased from 3.4 to 4.9 deaths per 1 million population (18).
Trends in hospitalization among children (0-17 years of age) in the U.S. from 1979 to 1987 were reported by Gergen and Weiss (19). Using the National Hospital Discharge Survey, they found a 4.5% per year increase in asthma hospitalizations (95%CI, 2% to 7.1%), with the largest increase being 5.0% per year among 0- to 4-year-olds (95%CI, 3.4% to 6.7%). African American children 0 to 4 years old had about 1.8 times greater increase than Caucasian children of the same age. Their analyses of data regarding admissions for lower respiratory tract disease and bronchitis showed decreases that suggested diagnostic transfer may have contributed in part to increased hospitalizations.
In the UK, hospitalizations due to asthma in children were increasing until the mid-1980s. A study using data from the Hospital In-Patient Enquiry (which provides hospital admission data for England and Wales), described trends in admission rates in England and Wales (1976 to 1985), the East Anglian region (1976 to 1991-2), and Wales (1980-1990). The rates for England and Wales combined increased through 1985. In East Anglia, rates rose until 1985 when they peaked and then appeared to decline. Rates in Wales increased until a peak in 1988 and a subsequent decline. A change in the information system used to collect hospital admission data may have confounded these trends, but the observation that some of the downwards trends were started before the system change suggests that this change may have only partly contributed to the observed trends. Other possible factors included changes in medical care delivery, treatment, hospital admission and readmission policies, and severity or prevalence of asthma. The researchers also noted that recent asthma mortality data supports a possible true change in asthma morbidity (20).
Of special note are the two "epidemics" of asthma mortality in New Zealand, the first in the 1960s and another in the 1970s. In Chapter 6 of the text "Asthma and Rhinitis", Neil Pearce et al. provide a review of these mortality epidemics. I will just mention a few key points. In the 1960s large increases in asthma mortality were also seen in England and Wales, Scotland, Ireland, Australia, and Norway; however, mortality rates remained relatively stable in the U.S. and Germany. Various possible reasons for these trends were investigated including changes in medical diagnosis, disease classification, death certification, asthma prevalence, and treatment methods. Ecologic data correlating asthma mortality with sales of beta-agonist bronchodilators appeared to show some support for the possible association between use of these asthma treatment drugs and increasing asthma mortality. However, there are limitations to this type of data, and formal case-control studies were not done (21).
There continues to be debate about the possible dangers of beta-agonist bronchodilators. It has been hypothesized that underuse of corticosteroids and overuse of beta-agonists have contributed to increases in asthma deaths. Fenoterol, a selective B2-agonist, was hypothesized to play a large role in the second asthma mortality epidemic, this time experienced only in New Zealand. Sales of fenoterol appeared to correlate with trends in asthma deaths. This along with additional observations as well as laboratory studies of its possible physiologic effects, prompted researchers to carry out several case-control studies. These studies supported the association between fenoterol use and increasing asthma mortality, even when controlling for asthma severity. Last of all, additional ecologic data showed that mortality rates dropped significantly after warnings about the use of fenoterol were announced and sales of the drug decreased (21).
Asthma prevalence appears to differ between certain races and by socioeconomic status. In NHANES 2 (1976-1980), the cumulative prevalence of asthma for all age groups was estimated to be 10.6% This prevalence was greater in males than in females (11.4% vs 9.7%, p<0.05). The prevalence in African Americans was greater than in Caucasians (12.2% and 10.4%, respectively, but this difference was not significant. Those living below poverty level appeared to experience significantly more asthma than those living at or above poverty level (13.1% vs 10.3%, p<0.05) (7).
In the U.S., asthma mortality is more common among nonwhites, those living in urban areas, and the poor (22). The CDC analysis, mentioned in the "Morbidity and Mortality" section, found that African Americans more so than Caucasians experienced a higher annual death rate and a higher age-adjusted hospital discharge rate for asthma as a primary diagnosis (18). Lang et al. (23) analyzed asthma mortality rates in Philadelphia, Pennsylvania (a large urban area from 1969-1991 in order to identify trends in rates and possible associations with changes in air population. Mortality rates were found to increase during this time and were highest in census tracts with the highest percentages of low income and minority residents, especially African Americans. Concentrations of major air pollutants declined during this time, suggesting that other factors in this urban area contributed to this trend.
Poverty is also associated with increases in asthma morbidity and mortality in New Zealand. Maoris and Pacific island Polynesians experience higher mortality rates and hospitalization rates than other New Zealanders (3)
Understanding the distribution of asthma over time and between populations can help to identify risk factors, develop effective interventions, and identify and target persons at risk so that we may decrease the unnecessary burden of asthma morbidity and mortality. Possible risk factors for the development of asthma as well as for morbidity and mortality resulting from asthma are addressed in "Risk Factors".
Many factors appear to play a role in both the etiology of asthma and its exacerbation. It is not clear which factors are causative and which are "promoters" of the disease. In part for this reason, etiologic factors will not be discussed in great detail here. Genetic factors, atopy, birth factors, tobacco smoking, indoor and outdoor air pollution, infections, treatment, and occupational exposure have all been proposed to contribute to the prevalence, morbidity and/or mortality from asthma.
At the host level, genetic factors seem to contribute to the development of asthma. Twin studies have shown symptom concordance of 19% for monozygotic twins versus 4.8% for dizygotic twins. The prevalence of asthma appears to be greater in first degree relatives of nonasthmatics--13% versus 4%. A study of college students found that asthmatic students were more likely than nonasthmatic students to have a family history of asthma or hay fever (3).
Several studies support that atopy is the most important single risk factor for asthma. The majority of asthmatics are atopic. Atopy is highly prevalent among children--it has been estimated that between 30% and 50% of children are atopic (16). A study comparing the prevalence of asthma and atopy in two areas of East and West Germany found both disorders to be more prevalent in West German children (p<0.0001). After stratification by atopic status, the difference in asthma prevalence was no longer significant (24).
Asthma has been classified as "intrinsic" (nonallergic) and "extrinsic" (allergic), but these classifications have been challenged. An often sited study by Burrows et al. (25) found that doctor diagnosed asthma rates after six years of age were strongly associated with serum immunoglobulin (IgE) levels, and none of the population with the lowest IgE concentrations had asthma. Additionally, those with normal allergy skin tests (intrinsic asthmatics) had higher IgE levels than nonasthmatic persons. Sears et al. (26) looked at the correlations between bronchial hyperresponsiveness, atopic status, and asthma. They found a strong correlation between atopy and BHR. They also found that children with diagnosed asthma and BHR were almost always atopic.
The allergens that appear most important in triggering asthma are allergens from dust mites, cats, cockroaches, and molds. Dust mites have also been identifies as playing a role in the etiology of asthma (3). In order to evaluate the relationship between house-dust mite (Der p) exposure and the development of allergic sensitization and asthma, Sporik et al. (27) carried out a prospective study of children at risk for atopy. They found that children exposed to high levels of Der p at one year of age were significantly more likely to have asthma at 11 years of age (RR 4.8, p=0.05). Data from another study show that children admitted to the hospital with asthma exacerbations had a significantly higher prevalence of house-dust mite exposure and house-dust mite sensitivity than control children (p<0.001) (28).
Various factors of birth and early childhood have been investigated as possible contributors to asthma. Premature and low birth weight infants appear to be at increased risk for developing asthma. Exposure to second-hand tobacco smoke in utero or in early childhood appear to be risk factors. It is difficult, though, to separate the effects of in utero exposure from exposure after birth. Tobacco smoke has also been shown to increase asthma's severity (3,29).
Infections have been hypothesized to influence the development of asthma, but their role is being debated. Some research has shown that bacterial infections may not be important precipitators of asthma, and it has even been suggested that if these infections occur in the first few weeks of life, they may actually be protective against the development of an allergic phenotype (29). Studies that have shown small family size to be associated with increased atopy (29) lend some support to this theory--i.e., smaller family sizes would provide decreased opportunity for infection. On the other hand, viral infections early in life have been associated with the development of asthma (3,29).
As with many of the other factors discussed, the role of pollution in the development and exacerbation of asthma is under debate. Investigation of possible associations is limited by the difficulty of measuring individuals' exposure to pollution. Trend data show that asthma prevalence has increased at the same time that the concentrations of major air pollutants have decreased (3,23,30). Indoor air pollution has also been considered as a possible risk factor. Modern homes with improved insulation and less air exchange could lead to increased concentrations of allergens and tobacco smoke (30).
Additional risk factors hypothesized include: changes in treatment, particularly as this relates to asthma drugs (this is discussed in Morbidity and Mortality); occupational exposure; and dietary factors. Flour dust, laboratory animal urine proteins (30), and wood dust (29,30) are among the occupational exposures investigated. Dietary intakes of sodium, antioxidants, and fatty acids are all nutrition factors being researched.
Identifying those populations at increased risk for developing asthma or incurring increased morbidity and mortality from the disease can provide important information needed to develop and implement effective interventions. As was already pointed out, it appears that throughout many countries asthma's morbidity and mortality is increasing. Certain populations are affected disproportionately--e.g., the poor and certain minority groups. A number of the risk factors are controllable. Other factors are related to broader issues such as housing and access to health care.
Clearly, asthma places a burden on the individuals affected and their families. A burden is also placed on society. Asthma is a major source of childhood disability, the leading cause of school absences, and a frequent reason for emergency department visits (31). When asthma is managed well, hospitalization is rare (32). In the U.S. in the 1980s children receiving Medicaid or with no insurance contributed to an increasing percentage of asthma hospitalizations (19).
What are the economic effects of this disease? In the U.S. health care spending for asthma medication alone is close to $1 billion per year. It was estimated that the total cost of asthma in the U.S. in 1985 was almost $4.5 billion ($6.2 billion when adjusted to 1990 dollars), with approximately $2.4 billion being direct costs and $2 billion indirect costs. Inpatient hospitalization accounted for the greatest portion of direct costs (32). The economic cost of asthma in New South Wales in 1989 was $209 million, of which $142 million could be attributed to direct health care costs, $19 million to direct non-health care costs, and $48 million to indirect costs (33).
Reducing avoidable morbidity and mortality through appropriate treatment and effective preventive strategies is necessary to limit the public health and economic burdens of asthma. Much focus has been placed on asthma self-management. This approach to asthma care involves education of and committment by the family to manage asthma at home. Research supports that educational program can improve families' ability to manage asthma at home. Hughes et al. (31) conducted a 2-year randomized controlled trial in order to evaluate the impact of a comprehensive asthma management program. They found that the experimental group compared to controls had significantly less school absenteeism (p=0.04), and spent significantly fewer days in the hospital (p=0.02). Another randomized trial evaluated the effects of a health education program to improve home asthma management in a low income population. Among those children who had been hospitalized during the previous year, the program was found to significantly decrease emergency room use and the mean number of hospitalizations among the experimental group versus the control group (p<0.05) (34).
Please send comments and suggestions to: email@example.com
Created by Lorraine Ettaro
Last updated April 16, 1997