Housing, environment and cardio-respiratory health: the relative influence of the past and the present
Walker, Jeremy Joseph
The existence of socially-patterned health (with poorer health generally being experienced by those in more disadvantaged circumstances) is widely recognised. Social differentials have been observed for (inter alia) respiratory disorders, and for cardiovascular disease. One possible explanation for social inequality in these areas of health posits a mediating effect of housing conditions: disadvantaged individuals may face greater exposure to residential hazards (such as dampness), which may in turn adversely influence cardiorespiratory health. However, few studies have examined a complete posited causal chain linking socioeconomic position (SEP) with health via housing. Using pre-existing data, this study constructed detailed representations of the social and residential experiences over adult life (15 to 60 years) of a sample of elderly British people. Both measures of accumulated exposure (to disadvantage, and to housing hazards), and explicit trajectories of social and residential experience, were derived. Construction of trajectories required the development of methods for condensing individuals’ diverse experiences into higher-level groups, in the interests of analytical tractability. Relationships between the derived measures of lifetime exposure and a range of outcomes expressing aspects of cardio-respiratory health in old age were assessed. No persuasive evidence was observed to support the hypothesis that lifetime residential exposures may mediate the relationship between SEP and the health outcomes examined. In addition to testing this specific conceptual model, the study examined how exposure to social disadvantage and to residential risks varied over adult life, identifying distinctive features of the exposure experience which could not readily be captured by the infrequent sampling of SEP commonly featured in health inequality research. The respective merits of such ‘sparse’ sampling and the more intensive sampling used in the study were compared. It was concluded that fully exploiting the additional information captured by intensive sampling requires confronting a number of methodological challenges. Because of this, it is argued that the collection of detailed information on exposures over time does not automatically confer genuine advantages over the hitherto dominant approach of sampling at only a small number of time points. Future development of lifecourse epidemiology will require further debate over how lifetime exposure (to both social and environmental risk factors) can most effectively be represented in quantitative analysis.