OBJECTIVES: To analyse whether the association between social inequality and health among the aged is mediated through social relationships (differential exposure hypothesis). Moreover, it is examined whether the association between social relationships and health is stronger in lower socio-economic groups (differential vulnerability hypothesis). METHODS: The analyses are based on a telephone survey of 682 older people (> or = 60 years) living in a private household in Germany. Education, income and occupational status are utilised as indicators of the socio-economic status. Quantitative (social contacts) as well as qualitative indicators (emotional support) of social relationships are used. Respondents were asked how often they meet their friends and relatives. Emotional support was measured by an instrument assessing the availability of a confidant and the perceived adequacy of the emotional support received. Self-rated health, depression (CES-D) and functional limitations are introduced as health indicators. RESULTS: Multiple logistic regression analyses show that the mediating effect of social relationships on the association between socio-economic status and health (self-rated health, depression and functional limitations) among the aged is weak. Results concerning the differential health impact of social relationships in different socio-economic groups are inconsistent. DISCUSSION: Results do not support the differential exposure hypothesis saying that social inequalities in health can be explained by social relationships. Results do not support the differential vulnerability hypothesis either as there is no evidence for a stronger association between social relationships and health in lower socio-economic groups.
