The relationship between informal care-giving and mortality: an analysis using the ONS-LS (Phase I)

Dermot O'Reilly, Michael Rosato and Sheelah Connolly,Queen's University, Belfast

[Project number 30081]

Informal care is a fundamental component of care in the community which, given current demographic trends and increasing prevalence of debilitating chronic disease, is likely to assume even greater significance in future. Research indicates that caregivers are more likely than non-carers to report poor health, though this has usually been measured in terms of psychological or emotional health such as depression or 'caregiver strain'. While relatively little is known about the effects of caring on physical health (with much of the relevant analysis carried out in the United States using limited data sources - and small numbers of deaths), a recent analysis of the whole of the enumerated Northern Ireland 2001 Census population examined the extent care-giving and subsequent mortality in the following four years (see above). In that analysis, carer-givers were found to be a heterogeneous group, with those providing fewer hours of care being relatively more affluent than those providing care at greater intensities. Overall, those providing care had a lower mortality risk than non-carers - effects more pronounced for women, older people and those reporting poorer health at the start of the study period - lending support to the growing body of literature which suggests that the positive aspects of caring have been underreported. The aim of the proposed study is to corroborate these findings using the ONS-Longitudinal Study.

Using data from the 2001 Census this study will examine the relationship between informal care-giving as recorded at the census and subsequent mortality. Its purpose is (a) to replicate an analysis recently completed using the Northern Ireland Longitudinal Study (O'Reilly, Connolly, Rosato & Cardwell: Social Science and Medicine, 2008, submitted), and (b) to extend the analysis to examine for effects not possible with the data from Northern Ireland (for example - by standard region, or by ethnic minority status).