Philibert M D, Pampalon R, Hamel D, Daniel M
Institut national de santé publique du Québec, 190, boulevard Crémazie, Montréal (Québec), H2P 1E2, Canada; Département de médecine sociale et préventive, université de Montréal, Montréal, Canada.
Rev Epidemiol Sante Publique. 2013 Oct;61(5):463-74. doi: 10.1016/j.respe.2013.05.020. Epub 2013 Sep 17.
Disability is understood to arise from person-environment interactions. Hence, heterogeneity in local-area characteristics should be associated with local-area variation in disability prevalence. This study evaluated the associations of disability prevalence with local-area socioeconomic status and contextual features.
Disability prevalence was obtained from the Canada census of 2001 for the entire province of Québec at the level of dissemination areas (617 individuals on average) based on responses from 20% of the population. Data on local-area characteristics were urban-rural denomination, social and material deprivation, active and collective commuting, residential stability, and housing quality. Associations between local-area characteristics and disability prevalence were assessed using multilevel logistic regressions.
Disability was associated with local-area socioeconomic status and contextual characteristics, and heterogeneity in these factors accounted for urban-rural differences in disability prevalence. Associations between contextual features and disability prevalence were confounded by local-area socioeconomic status. Some associations between local-area socioeconomic status and disability prevalence were moderated by contextual characteristics. The importance of this effect modification is greater when expressed in terms of the absolute magnitude of disability than in the relative likelihood of disability.
Explanation of rural-urban differences by the contribution of other local-area characteristics is consistent with the conceptualization of urban-rural categories as the reflection of spatially varying ensembles of compositional and contextual factors. Although local-area socioeconomic status explains most variability in disability prevalence, this study shows that contextual characteristics are relevant to analyses of the spatial patterning of disability as they predict spatial variations of disability, sometimes in interaction with socioeconomic status. This study demonstrates that absolute and relative perspectives on effect modification may lead to differing conclusions.
残疾被认为是人与环境相互作用产生的。因此，当地特征的异质性应与残疾患病率的地区差异相关。本研究评估了残疾患病率与当地社会经济地位及背景特征之间的关联。
基于20%人口的回答，从2001年加拿大魁北克省普查中获取传播区域（平均617人）层面的残疾患病率数据。关于当地特征的数据包括城乡划分、社会和物质匮乏、主动和集体通勤、居住稳定性以及住房质量。使用多水平逻辑回归评估当地特征与残疾患病率之间的关联。
残疾与当地社会经济地位及背景特征相关，这些因素的异质性解释了残疾患病率的城乡差异。背景特征与残疾患病率之间的关联受到当地社会经济地位的混淆。当地社会经济地位与残疾患病率之间的一些关联受到背景特征的调节。当以残疾的绝对程度而非残疾的相对可能性来表示时，这种效应修正的重要性更大。
用其他当地特征的贡献来解释城乡差异，与将城乡类别概念化为构成和背景因素空间变化集合的反映是一致的。虽然当地社会经济地位解释了残疾患病率的大部分变异性，但本研究表明，背景特征与残疾空间模式分析相关，因为它们预测了残疾的空间变化，有时与社会经济地位相互作用。本研究表明，对效应修正的绝对和相对视角可能导致不同的结论。
