RESUMO
OBJECTIVES: We supplement existing findings on a link between social support and cognitive function in later life by considering the role of personality as an antecedent to both, and of social support as a mediator of the link between personality and cognitive function. STUDY DESIGN: An observational cohort study. METHODS: We evaluated social support using the Lubben Social Network Scale, across 624 adults aged over 60 years, and investigated this measure as a mediator of the relationships between extraversion and neuroticism at baseline 2007-2009, and cognitive function at follow-up, 2 years later. A half-longitudinal mediation design, within a structural equation modelling framework, was used. RESULTS: There was a direct effect of extraversion, such that lower levels were related to higher scores of cognitive function. There was no significant direct effect of neuroticism on cognitive function at follow-up. Social support partially mediated the paths between both extraversion and neuroticism and cognitive function at follow-up. Decomposing the mediation effects by using social support subscales (measuring support from friends, relatives and neighbours) showed meaningful indirect effects for both predictors. CONCLUSION: Results suggest that social support may offer a target for interventions for cognitively at-risk older adults and add to the existing empirical evidence describing the link between personality and cognitive function.