The purpose of the study is to determine if pre-operative neuropsychological characteristics, apolipoprotein E4 status and/or cerebrospinal fluid (CSF) and serum biomarkers can predict those most at risk of post-operative delirium. Patients over the age of 65 years undergoing elective primary hip or knee arthroplasty in Musgrave Park Hospital are invited to take part. Participants undergo a pre-operative neuropsychological assessment which is repeated 6-12 weeks post-operatively at the time of surgical review appointment. At the time of surgery advantage is taken of the necessary venous cannulation and spinal anaesthetic to gain venous blood and CSF samples respectively. Venous blood is also sampled post-operatively. Patients are assessed daily for delirium, using the Confusion Assessment Method, until discharge. The association between delirium and subsequent cognitive decline is now well established in the literature. It is not yet clear whether this is solely because delirium following elective surgery indicates a previously unrecognised neurodegenerative process or whether the delirium itself exerts an additional pathophysiological insult. In order to investigate this, a follow-up study of these participants, approximately 8 years later is currently being conducted. Neuropsychological assessment and venous blood sampling will be repeated with consenting participants. Furthermore, the significance of subtle symptoms, not sufficient to fulfil the criteria for delirium and known as subsyndromal delirium, in relation to cognitive function years later will be analysed.