IJPDS International Journal of Population Data Science IJPDS 2399-4908 Swansea University 10.23889/ijpds.v7i3.2052 7:03:276 Bespoke automated linkage to enable analysis of covid deaths by ethnicity. Iveson Matthew 1 The University of Edinburgh 092022 2022 7 3 2052 This work is licenced under a Creative Commons Attribution 4.0 International License. This article is available from the IJPDS website at: https://ijpds.org/article/view/2052 Objectives

Suicide rates are high among older adults, with self-harm as an important risk factor. In middle-aged adults, self-harm and suicide risk appears to be predicted by early-life factors including cognitive ability. The present study examines whether associations between early-life factors and self-harm and suicide can be observed among older adults.

 Approach

We construct a large, representative cohort using participants of the Scottish Mental Survey 1947 – a nationwide assessment of cognitive ability and socioeconomic conditions administered to all 11-year-olds attending a Scottish school (N ~ 70,000). We link research data from childhood to later-life (age 50+) routinely-collected hospital admissions and deaths data.

 Results

Using survival analyses, we report the associations between early-life predictors – including childhood cognitive ability – and the risk of self-harm and suicide in later-life, further adjusting for proximal socioeconomic conditions and comorbidities.

 Conclusion

We demonstrate the importance of early-life factors for predicting self-harm and suicide among older adults, highlighting potential mechanisms, modifiable factors and markers. The implications of the results for research and policy are discussed.