- Top of page
- Material and methods
Background: It is disputed whether increases in self-reported respiratory allergy represent a true increase or merely increased recognition. We aimed to investigate whether the prevalence of skin-prick-test (SPT)-positive allergic rhinitis had increased in an adult general population in Copenhagen, Denmark.
Methods: Two cross-sectional surveys were carried out in 1990 and 1998. A screening questionnaire on respiratory symptoms in random samples of 15–41-year-olds preceded both surveys. Among the responders, random samples were invited to a health examination including SPT. Totals of 312 (participation rate 74.6%) and 482 (participation rate 53.4%) subjects were examined in 1990 and 1998, respectively. Diagnoses of SPT-positive allergic rhinitis were based on a history of nasal symptoms on exposure to allergens and SPT positivity to allergens.
Results: The prevalence of a diagnosis of SPT-positive allergic rhinitis increased from 12.9% to 22.5% (adjusted odds ratio 1.94, 95% CI 1.30–2.90), whereas the prevalence of a positive SPT (allergen histamine wheal ratio ≥0.5) to one or more of 10 allergens increased from 27.7% to 33.9% (adjusted odds ratio 1.47, 95% CI 1.05–2.05).
Conclusions: The prevalence of SPT-positive allergic rhinitis has increased significantly. Our findings indicate that a true increase in respiratory allergy has occurred.
There is considerable evidence that the prevalence of self-reported respiratory allergy has increased over recent decades. It has been discussed whether this increase represents a true increase or is due to an increased recognition of allergic disease. The inclusion of objective measures in studies of time trends is therefore important ( 1). Skin-prick-test (SPT) reactivity to aeroallergens is a generally accepted marker of respiratory allergy for use in epidemiologic studies ( 2, 3).
Most studies of time trends of respiratory allergy have been performed in children, and little is known about similar trends in adults ( 4). However, the reported increases in disease prevalence in children should now be apparent in increased disease prevalence in adults. Recently, we have reported an increasing prevalence of self-reported allergic rhinitis symptoms in an adult Danish population ( 5). This paper reports the prevalence of allergic rhinitis, using a definition based on the presence of both symptoms and SPT positivity, in an adult population in two surveys 8 years apart.
- Top of page
- Material and methods
Table 1 shows that the prevalence of the three specific diagnoses of SPT-positive allergic rhinitis and the two self-reported diagnoses of hay fever increased significantly. Accordingly, the prevalence of one or more of the three specific diagnoses of SPT-positive allergic rhinitis increased significantly. These increases were independent of sex and age group. In addition, the increases in SPT-positive allergic rhinitis were independent of the grade (1+ to 4+) that was used for defining SPT positivity in the definitions of SPT-positive allergic rhinitis (data not shown).
Table 1. Prevalence of diagnoses of allergic rhinitis
|Diagnoses of skin-prick-test (SPT)-positive allergic rhinitis||Prevalence 1990 (%) ||Prevalence 1998 (%) ||Odds ratio † (95% CI) |
|Allergic rhinitis due to pollen allergy (symptom a. or b. and SPT positivity to birch, grass, or mugwort)||10.0 (31/311)||16.8 (81/481)||1.81 (1.15–2.83)*|
|Allergic rhinitis due to dander allergy (symptom c. and SPT positivity to horse, cat, or dog dander)||2.6 (8/311)||7.7 (37/481)||3.27 (1.49–7.19)*|
|Allergic rhinitis due to mite allergy (symptom d. and SPT positivity to D. pteronyssinus or D. farinae) ||2.9 (9/311)||7.1 (34/481)||2.24 (1.05–4.79)*|
|Allergic rhinitis due to pollen, dander, or mite allergy (one or more of three above diagnoses)||12.9 (40/311)||22.5 (108/481)||1.94 (1.30–2.90)*|
|Self-reported hay fever||16.7 (52/311)||23.7 (114/482)||1.52 (1.05–1.36)*|
|Self-reported, doctor-diagnosed hay fever||8.0 (25/312)||15.6 (75/482)||2.14 (1.32–3.46)*|
Table 2 shows that the increased prevalence of SPT positivity to one or more allergens was due to a significantly increased prevalence of SPT positivity to pollen and animal dander, whereas SPT positivity to mite was unchanged. Furthermore, while the prevalence of subjects SPT-positive to two or more allergens increased significantly, the prevalence of subjects SPT-positive to only one allergen was almost unchanged.
Table 2. Prevalence of skin-prick-test positivity (allergen histamine wheal ratio ≥0.5)
|Allergens||Prevalence 1990 (%) ||Prevalence 1998 (%) ||Odds ratio † (95% CI) |
|Pollen (birch, grass, or mugwort)||19.3 (60/311)||24.1 (116/481)||1.52 (1.04–2.21)*|
|Dander (horse, dog, or cat dander)||6.1 (19/311)||13.7 (66/481)||2.72 (1.56–4.76)*|
|Mite (Dermatophagoidespteronyssinus or D. farinae) ||12.2 (38/311)||14.1 (68/481)||1.03 (0.65–1.62)|
|One or more of 10 allergens||27.7 (86/311)||33.9 (183/481)||1.47 (1.05–2.05)*|
|Only one allergen||10.9 (34/311)||11.2 (54/481)||1.13 (0.70–1.82)|
|Two or more allergens||16.7 (52/311)||22.7 (109/481)||1.55 (1.04–2.29)*|
The prevalence of nasal allergy symptoms, as reported in the screening questionnaire, increased significantly from 1990 to 1998 ( Table 3). This increase was slightly higher among the subjects participating in the health examination than among the subjects invited to the health examination (including both participants and nonparticipants).
- Top of page
- Material and methods
The results showed a significant increase in the prevalence of SPT-positive allergic rhinitis and a corresponding, but lower, significant increase in the prevalence of SPT positivity.
The prevalence of SPT-positive allergic rhinitis as assessed in the 1990 survey appears remarkably similar to that assessed in a Swiss general population of 18–60-year-olds in 1991. Thus, in Switzerland, the prevalences of atopic rhinitis (self-reported allergic rhinitis and at least one positive SPT or Phadiatop test) and hay fever (self-reported hay fever and a positive SPT to pollen) were 13.5% and 10.3%, respectively ( 8). In Finland the prevalence of SPT-positive hay fever among 5–17-year-olds was 10.9% ( 9), while in the UK the estimated prevalence of SPT-positive allergic rhinitis among 15–65-year-olds was 16% ( 10).
A few repeated cross-sectional studies have employed SPTs at two points in time. von Mutius et al. reported an increased prevalence of SPT positivity and self-reported hay fever from 1991–2 to 1995–6 among schoolchildren in the former East Germany during the years following the reunification with the former West Germany ( 11). In contrast, Peat et al. reported an unchanged prevalence of SPT positivity despite an increase in self-reported hay fever from 1982 to 1992 among Australian schoolchildren ( 12). In this Australian study, the prevalence of SPT positivity was already high in the initial survey, and we speculate that the prevalence of SPT positivity might have risen prior to the initial survey. One serologic study has reported an increased prevalence of specific IgE positivity to common allergens among 13–14-year-old Japanese schoolgirls between 1978 and 1991 ( 13).
Some methodological issues should be addressed in the interpretation of our results. First, the participation rates differed in the two surveys. The invitational procedures were kept the same, and it is therefore unlikely that the difference in participation rate was due to methodological differences between the surveys. However, the effect of possible selection bias may be enhanced by the lower participation rate in the second survey. Information on nonparticipants was obtained in the screening questionnaire, a fact which allowed us to analyze the effect of selection bias on the estimated odds ratios. Thus, the overall effect of nonparticipation was examined by comparing the increase in the prevalence of nasal allergy symptoms among the subjects invited to the health examination (including both those participating and those not participating in the health examination) to that among the subjects participating in the health examination ( Table 3). This increase was only slightly higher among the subjects participating in the health examination than among the subjects invited to the health examination. Thus, the odds ratios may be slightly overestimated when taking into account the effect of nonparticipation. However, this does not seem to explain all of the observed increase in SPT-positive allergic rhinitis.
Second, the SPT methods applied in the two surveys differed. A comparison in 1992 between the Phazet and the Soluprick methods showed that they were similar, although the Phazet tended to elicit a higher overall number of positive reactions to allergens in patients than the Soluprick method ( 14). According to this, a possible increase in the prevalence of SPT positivity would tend to be underestimated. Variations of SPT results between fieldworkers may bias epidemiologic studies ( 15), and it is therefore notable that, in the present study, the same technician performed all SPTs in both surveys.
Data on pollen counts during the pollen seasons in Copenhagen have been available since 1977 ( 16). There does not seem to be any consistent time trend in the cumulated pollen counts for each season that could account for the increase in sensitization to pollens. Furthermore, a comparison of the 1990 and 1998 pollen seasons showed no major differences ( 16); therefore, different levels of exposure to pollens during the two surveys were unlikely to have biased our results.
Marked urban/rural differences in the prevalence of respiratory allergy have been found in most countries ( 4, 17), including Denmark ( 18), suggesting that air pollution is a risk factor for the development of respiratory allergy. It would therefore be of interest to relate the observed increase in prevalence of pollen allergy to trends of air pollution measures in the study area. The Danish Air Quality Monitoring Program has been monitoring different air pollution measures in Copenhagen since 1982 ( 19). The concentrations of nitrogen oxide, sulfur dioxide, and total suspended particulate matter have been decreasing since the late 1980s, while the concentrations of nitrogen dioxide and ozone are almost unchanged. This apparent lack of a macroecologic relationship between pollen allergy and air pollution measures does not necessarily exclude air pollution as a potential risk factor for two reasons. Firstly, the critical components may not have been included in the monitoring program. For example, ultrafine particles from diesel combustion are not measured specifically. Secondly, the decrease in air pollution may not yet have had an effect on prevalence figures. Further research on the relationship between respiratory allergy and air pollution is needed to determine this issue. At present, it is not clear which factors are responsible for the recent increases in respiratory allergy. More knowledge about risk factors associated with lifestyle and environment may be provided by large-scale population-based prospective studies, some of which are already underway ( 20, 21).
There is no generally accepted definition of allergic rhinitis for use in epidemiologic studies ( 22, 23). The definition used in this study was based on both symptoms and objective measurements, as in the clinical assessment. Most studies of time trends of the prevalence of allergic rhinitis are based on self-reported diagnoses, such as self-reported hay fever. Table 1 makes it possible to compare the prevalence of SPT-positive hay fever (“allergic rhinitis due to pollen allergy”) to the prevalence of self-reported hay fever and self-reported doctor-diagnosed hay fever in the subjects studied. It appears that the prevalence of doctor-diagnosed hay fever was reasonably similar to, although slightly lower than, the prevalence of SPT-positive hay fever. This is consistent with previous reports that not all patients suffering from hay fever consult a physician ( 10). The sensitivity (76.0% and 69.3% in 1990 and 1998, respectively) and the specificity (95.8% and 92.9% in 1990 and 1998, respectively) of SPT-positive hay fever with self-reported, doctor-diagnosed hay fever as the reference indicated good agreement between these two definitions.
It is disputed to what extent the reported worldwide increases in respiratory allergy are real or due to changes in recognition or labeling of disease ( 1). The present study confirms that the previously reported increases in self-reported symptoms ( 5) and self-reported diagnoses ( 18) of respiratory allergy among Danish adults are real. However, the increase in SPT positivity to allergens appeared to be lower than the increase in SPT-positive allergic rhinitis. This may represent either increased morbidity or increased recognition of symptoms among SPT-positive subjects. In this way, our findings are not inconsistent with the hypothesis that the recognition of allergic disease has increased, although this does not fully explain the reported increases in symptoms and diagnoses. A possible increased morbidity, because SPT-positive subjects report more symptoms, might be explained by the increased prevalence of subjects sensitized to more than one allergen ( Table 2).
In the present study, the estimated prevalence of SPT-positive allergic rhinitis was high, and it increased over an 8-year period. Although some cases of allergic rhinitis are symptomatically mild, others are associated with considerable morbidity, and some subjects with allergic rhinitis have other atopic diseases, such as asthma. Diseases with substantial increases in prevalence over short periods of time are likely to be caused by environmental or lifestyle factors, and may therefore be prevented. Thus, allergic rhinitis should be considered a major public health problem, and more research into its causes is needed.