O'Reilly, D., Rosato, M., Maguire, A. & Wright, D. (2015) Journal of Epidemiology and Community Health, 69(Suppl 1), A29 [NILS]
Other information: Abstract:
Background There is a significant body of evidence associating caregiving with poor mental health and recent studies suggest that caregivers are more likely to contemplate suicide than non-caregivers. However, no study to date has explored the relationship between caregiving and actual suicide. This study uses a large population-based linkage database to (i) explore the association between caregiving and self-reported mental morbidity, and (ii) quantify the relationship between caregiving and suicide risk.
Methods A 2011 Census-based study of 1,037,475 people (including 183,811 caregivers) aged 25–74 with mortality follow-up for thirty-three months. Caregiving was categorised as either light (1–19 h/week) or intense (twenty or more). Suicide risk was assessed using Cox Proportional Hazards models with adjustment for and stratification by mental health status at census.
Methods The cohort experienced 390 suicides during follow-up (31 to caregivers). Although caregivers were more likely to suffer from poor mental health (ORadj = 1.17: 1.14, 1.20) for those providing twenty or more hours/week, the suicide risk for caregivers was lower – HRadj = 0.60 (0.37, 0.98) and HRadj = 0.58 (0.33, 1.00) respectively for those providing light and intense caregiving. However, this was modified by base-line mental health: for those not reporting poor mental health, caregiver suicide risk was about half that of non-caregivers, while for those reporting poor mental health there was no significant difference between caregivers and non-caregivers.
Conclusion Although this large and representative population study demonstrates that suicide risk is not raised for the majority of caregivers, it is evident that a large proportion of caregivers report poor mental health and there is a need to support caregivers to reduce potential physical and mental ill-effects associated with their role.