International Journal of Epidemiology O International Epidemiologlcal Association 1996 Vol. 25, No. 2 Printed in Great Britain Severity of Childhood Asthma by Socioeconomic Status A MIELCK,' P REITMEIR* AND M WJST+ Mielck A, (Institut fur Medizinische Informatik und Systemforschung (MEDIS), GSF—Forschungszentrum fur Umwelt und Gesundheit, Neuherberg, Germany), Reitmelr P, and Wjst M. Severity of childhood asthma by socioeconomic status. International Journal of Epidemiology 1996; 25: 388-393. Background. A review of studies on the association between childhood asthma and socioeconomic status (SES) in industrialized countries leads to the conclusion that there does not seem to be a clear association. A study from Aberdeen published 25 years ago, however, shows that among children with asthma, severe asthma is most prevalent in the lower social class, but this distinction between grades of asthma severity has been largely ignored since. Methods. We screened all fourth grade schoolchildren of German nationality in Munich (4434 children, response rate 87%), distinguishing three severity grades in the same way as the study in Aberdeen. Results. Prevalences of childhood asthma are reported by severity grade and SES. Prevalence of severe asthma was found to be significantly higher in the low as compared with the high socioeconomic group (Odds ratio = 2.37; 95% confidence interval: 1.28-4.41). This association could not be explained by established risk factors. Conclusions. More attention should be paid to the association between severe asthma and SES, with measures such as targeting early diagnosis and treatment towards low socioeconomic groups. Keywords: asthma, children, socioeconomic status, health inequalities Asthma is one of the most prevalent chronic illnesses among children in industrialized countries, 1 - 2 and the prevalence seems to be increasing. 3 - 4 Empirical infor- mation on the association between childhood asthma and socioeconomic status (SES) can be found in 24 studies published since the late 1960s. The overview presented in Table 1 includes studies published in English only and is not intended to provide more than a brief summary concerning the major features of the studies. Also, it may suffice to express the information on the observed association between childhood asthma and SES in a simplified form in order to show that, overall, the association is less than clear: negative associations (i.e. increasing prevalence with decreasing SES) are about as numerous as positive ones, and most studies show no association at all. This confusing pic- ture could be due, for example, to different definitions of asthma, different methods for assessing asthma (by questionnaire, medical record or medical examination) and the different age groups studied. Similar problems of comparability are well known in asthma research, 5 but from the studies listed in Table 1 no clear picture emerges concerning the association between childhood * Inst. filr Medizinische Informatik und Systemforschung (MEDIS) and f Inst. fDr Epidemiologie (EPI), GSF - Forschungszentnim fUr Umwelt und Gesundheit, Neuherberg, Germany. asthma and SES even if studies are compared which use similar designs and methods. A potential explanation for this puzzling result is provided by a study published by Dawson et al. 6 about 25 years ago. The study does not present prevalences but instead it gives the distribution of severity grades among children who were diagnosed with asthma. Of the 31 children with high SES, 65% were classified as having mild asthma and 6% as having severe asthma; the corresponding figures for the 53 children with mid- dle SES were 32% and 15% respectively, and for the 37 children with low SES, 35% and 27% respectively. Thus, the association between childhood asthma and SES is strongly modified by the degree of asthma sev- erity, indicating that the prevalence of severe asthma increases with decreasing SES. This result seems to have been largely ignored, as only two studies were subsequently published that pursued the idea of disting- uishing degrees of asthma severity. One of these studies reports a similar result, 2 but the second 7 unfortunately only includes a brief statement that no significant differences by SES were found. In order to clarify whether the distinction between degrees of asthma severity offers a basis for under- standing the association between childhood asthma and SES, we used the approach from the Aberdeen study 6 for analysing recent data from Germany. There were 388 by guest on August 29, 2016 http://ije.oxfordjournals.org/ Downloaded from