Cortisol and inflammation in delirium and long-term cognitive decline after hip fracture

Abstract

Delirium, or “acute confusion” is a common and serious acute neuropsychiatric syndrome mainly affecting older people. It is associated with multiple adverse outcomes, including an increased risk of developing dementia and increased mortality. The underlying mechanisms of delirium are poorly understood, and there are currently no specific treatments. This thesis investigated the roles of the hypothalamic-pituitary adrenal axis and inflammation in the pathophysiology of delirium, persistent delirium and cognitive decline following delirium. It investigated whether levels of cortisol in blood and cerebrospinal fluid (CSF) are elevated in delirium, with elevated pro-inflammatory and reduced anti-inflammatory cytokines. It also investigated whether there is loss of cortisol diurnal rhythm (in saliva) with elevated afternoon cortisol levels. The thesis investigated whether any hypercortisolaemia was sustained during the year after delirium, and whether this was associated with deterioration in cognition during the year after hip fracture. Finally, it also tested whether there are high levels of a marker of central nervous system damage (S100B) and of a dementia marker (tau) in CSF in delirium. A prospective observational cohort study was conducted in N=108 patients aged over 60 who had sustained a hip fracture, in whom 40% developed delirium. Participants gave informed consent or if they lacked capacity to give informed consent, this was given by their next of kin. Participants were assessed regularly for delirium, according to DSM IV criteria, during the two weeks after hip fracture. A sample of CSF was collected during the spinal anaesthetic performed for the operation to repair their fracture. Samples of blood and saliva were collected during the two weeks after the hip fracture operation. Participants were visited three, six and twelve months after their hip fracture for further delirium assessment, and a cognitive test battery was completed. Further samples of blood and saliva were collected at these visits. The study found evidence of high levels of cortisol and of S100B in CSF in those with active delirium, but there were no differences in levels of tau or cytokines in CSF. Those with delirium had elevated serum cortisol during the perioperative period, and elevated afternoon salivary cortisol, suggesting flattening of cortisol diurnal rhythm with failure to reach the normal diurnal nadir. After adjusting for confounders in a multivariate logistic regression analysis, serum cortisol was still predictive of delirium, but salivary cortisol AM:PM ratio had a trend towards significance. Those who had persistent delirium features in the months after hip fracture had significantly higher serum cortisol three months after hip fracture. There was a change in serum inflammatory profile in those with delirium, with a shift towards a pro-inflammatory state. Testing the study hypotheses surrounding cognition after delirium was very challenging, due to patient attrition and other factors. Some participants showed a trajectory of cognitive improvement, which was probably due to resolution of delirium during the year after hip fracture. Those with resolved delirium had deficits in verbal and visual memory. This study has improved understanding of the mechanisms of delirium, suggested further avenues for research and identified possible new therapeutic targets.