Does the context affect health rating? Area of residence and household influences on physical and mental perceived health in Italy

Patrizia Giannantoni, Università di Roma "La Sapienza"
Viviana Egidi, Università di Roma "La Sapienza"

Health represents one of the main dimensions of personal well-being, sensitive to both individual determinants and contextual factors. This research investigates the influence of context on physical and mental perceived health in Italy, defining context on two levels: (i) geographical, i.e. Aggregated Local Health Units (ALHU), and (ii) relational, i.e. households. The study is cross-sectional with data from the Italian Health Survey (2005). Outcome variables were: Physical Component Summary – PCS and Mental Component Summary – MCS, quantitative assessments of physical and mental health conditions as perceived by the respondent through a standardized questionnaire (SF-12). PCS and MCS were studied adjusted for objective health conditions, in order to isolate their perceived component. Data presented a hierarchical structure: individuals (level 1) living in different households (level 2), located in different “ALHU” (level 3). We adopted a multilevel approach in order to evaluate the proportion of variance on each level and to gain unbiased estimations of covariates’ effects. We documented a very limited, although significant, impact of ALHU on physical and mental perceived health (proportion of variance: 0.3% for PCS, 0.6% for MCS, adjusted for individual covariates). By contrast, the relevance of household was fairly substantive (proportion of variance: 15% for PCS and 33% for MCS). However, household covariates explained only a very limited part of this variability. This means that people in the same household tend to exhibit similar levels of health. We hypothesized this could depend on mutual influences of perceived health itself within households. Analyzing health homogeneity by family structure we found that similarity in perceived physical and mental health was indeed higher in those households where the links between members were supposedly tighter (2 components, couples, mono-nucleus families). These results parallel findings from social psychology, consistently documenting a similarity in mental illness, depressive symptom and distress between spouses.