Return to search

Area effects on health inequalities: the impact of neighbourhood deprivation on mortality and morbidity

There is a long history of research showing that material deprivation indicators are important predictors of health inequalities. The gap between the richest and poorest members of society is more predictive' of ill-health than the deprivation per se in countries that are above a certain income threshold level. However, although . . . , demonstrated across nations and states, the exact nature of the association between the context of the local area and local health outcomes is unclear. The aim of this thesis is to examine area effects on health inequalities. It explores whether the socioeconomic conditions of neighbouring localities influence the mortality and self-reported morbidity of a target locality, analysing at different levels of small geographical units across the whole of England. It tests the hypothesis that areas which are geographically close, but divergent in terms of deprivation, have different health outcomes than those where deprivation is similar across neighbouring localities. The Index of Multiple Deprivation (IMD) was used to create a neighbouring deprivation index (Adjacent Locality Deprivation-ALD) using a gravity model. Principal component analysis was then applied to disaggregate the strong correlation between the deprivation of a target area (lMD) and that of its surrounding areas (ALD). At lower super output area (LSOA) level (n=32,482, population size approximately 1,500 persons), the first principal component (PC1) represents the common effect of IMD and ALD. The second principal component (PC2) represents the additional effect due to the difference between IMD and ALD, which is interpreted as relative deprivation. However, at middle super output area (MSOA) level (n=6,780, population 7,500), IMD and ALD are already orthogonal, with the PCl being the equivalent of IMD and PC2 being ALD. Linear regression models were applied to scrutinize the association between the two principal component scores (predictors) and Directly Standardised Rate (DSR) mortality, self-reported morbidity (outcome variables), identifying the strength of the relative effect of the two principal component scores on health in different segments of the population. At LSOA level, areas that were deprived compared to their immediate neighbours . , . , suffered higher rates of mortality (~=0.162, p<O.OOl), 'not good health' (~=0.306, p<O.OOl) and 'Limiting Long-Term Illness' (~=0.299, p<O.OOl), after controlling for the deprivation of the area itself (mortality-~=0.457, p<O.OOl, not good health-~=0.659, p<O.OOl; Limiting Long-Term IlIness-~=0.578, p<O.OOl). The effect of the deprivation differential relative to the effect of deprivation per se was strongest in the most affluent portions of the population. A stronger effect of relative deprivation was observed for self-reported morbidity compared with the effect on mortality. At MSOA level, there were no significant effects associated with relative deprivation. Since ill-health was greater in less affluent areas surrounded by more affluent areas, this thesis provides evidence to support the psychosocial theory that perception of inequality exacerbates ill-health. A neighbourhood size of around 1,500 persons may be the critical population size for making such comparisons. In conclusion, this study presents a novel methodology to measure relative deprivation at different geographical levels. The findings support Wilkinson's psychosocial theory (1992) that negative social comparisons could cause ill-health if less affluent areas are surrounded by more affluent areas. This psychosocial effect is greater still in more affluent segments of the population, further supporting the notion that psychosocial effects become more important when material (absolute) deprivation is less relevant. The global trend towards more of the population residing in cities, where experience of relative inequality may be more common, may increase the number of people experiencing negative health effects. 11 -- .

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:583375
Date January 2012
CreatorsZhang, Xin
PublisherUniversity of Liverpool
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation

Page generated in 0.0021 seconds