During the past few years findings from two studies—the International Study of Asthma and Allergies in Childhood (ISAAC)1 and the European Community Respiratory Health Survey (ECRHS)2—seem to have dominated much of the thinking about the geographical variations in the prevalence of asthma. These studies suggest important variations by geographical site in different countries and, to a lesser extent, in study centres within countries. However, these studies only raise new questions as to how geographical factors contribute to the aetiology of asthma. For example, are the geographical variations in the prevalence of asthma caused by unique exposures to disease causing agents found only in local environments? Do differences in geographical location in some way alter host susceptibility? Do local environmental conditions, either physical or sociocultural, alter the host/agent interactions in ways that are critical to the expression of this condition?

In this issue of Thorax Duran-Tauleria and Rona3 further elucidate how social and physical environmental factors may explain some of the geographical expression of asthma prevalence. Their study focuses on the differences in asthma and respiratory symptoms among three populations—two nationally representative samples of English and Scottish children and a third sample of children living within the inner cities of England. The authors found a higher prevalence of children with persistent (as opposed to occasional) wheeze within the inner city population than in the other two sampled populations. They also found that the prevalence of asthma attacks was higher in the non-inner city sample of English children than in the other two groups. Social and physical environmental factors—mainly social deprivation (as measured by the Townsend score)—seemed to explain much of the geographical variation in children with persistent wheezing. This finding held even after adjusting the analysis for many of the well known risk factors that might otherwise have explained these differences such as age, sex, parental atopy, and maternal smoking.

The importance of social and physical environmental factors in disease expression is well known and has been studied extensively for many health conditions, including asthma. A number of studies suggest an association between higher asthma morbidity/mortality rates and geographical areas of lower socioeconomic standing. The literature on the effects of social environmental factors on the prevalence of asthma is less definitive, but there have been several studies that suggest findings similar to those of Duran-Tauleria and Rona. Yet this current study also provides new insights into the possible mechanisms by which poverty may influence the development of asthma based on its impact oncommunities independent of its effect onindividuals.

The results of this study suggest that individual measures of social and physical environmental risk (specifically, measures of socioeconomic status) may be less important to the prevalence of asthma than geographical ecological measures (specifically, the Townsend deprivation score of the area of residence). The finding that an ecological measure may explain risk better than individual measures is perhaps at first disconcerting. On further reflection, however, it would be expected that the relative deprivation index (Townsend score) may be acting as a surrogate for other sociocultural or physical environmental exposures common to parochial experience—for example, it is possible that high levels of outdoor pollutants such as ozone, small particulate matter, and sulphur dioxide might be disproportionately represented in areas of poverty. It is also possible that any one or more of these exposures may be more closely associated with geographical ecology than individual socioeconomic status.

Similarly, allergens, cigarette smoke, inadequate ventilation, dampness, and lack of air conditioning have all been cited as possible indoor environmental risk factors for the expression of asthma.4 Areas of increased poverty may be associated with risks of exposure that are less amenable to individual risk avoidance than population risk avoidance. For example, areas of poverty tend to have a higher density of population with more persons sharing residences—either more persons per household or more households per building. In shared living arrangements many exposures such as moulds, cockroaches, and cats are more likely to be community problems, requiring community based solutions, and are less under the control of any one individual, regardless of individual attitudes or behaviours towards changing these exposures.

Furthermore, these two examples of how poverty may impact asthma at the community level focus on asthma specific risk factors. Other literature on social and physical environmental deprivation suggests that more generic factors may be significant—for example, local community social and cultural practices may alter the general perception of disease within a community either through basic health beliefs, through susceptibility to disease, or by likelihood of contact with health care providers that would in some way influence awareness of clinical symptoms.5 ,6

While it is interesting to ponder the possible differences as to how geographical community factors (compared with individual factors) affect the risk of asthma, Duran-Tauleria and Rona caution us about some of the key limitations of their findings. Most noteworthy is the fact that there was variable reporting of socioeconomic measures for the populations with the father’s occupation more likely to be missing from the inner city population. The authors state that the lack of willingness to disclose the father’s occupation may be due to a higher frequency of single parent families headed by the mother. The effect of this loss of information on the analysis of socioeconomic status is not clear. In addition, one must question whether measures of individual socioeconomic status, based on constructs developed in the mid 1900s—for example, father’s occupation—are still valid and reliable in today’s society. Finally, the overall limitations of self-reported compared with observed health experience need to be considered.

In spite of these limitations, this study provides additional evidence that there is still much to be learned about the phenotypic expression of the syndrome called “asthma” through the study of small area variations. While studies such as the ISAAC and ECRHS are essential for understanding the global burden of asthma, studies of variations in the prevalence of asthma and morbidity in small areas within communities are more likely to elucidate some of the key interrelations between host, agent, and environment for this disease.