Delirium and long-term cognitive impairment after stroke: the role of the hypothalamic-pituitary-adrenal axis
Barugh, Amanda Jayne
Delirium is a severe neuropsychiatric syndrome, characterised by the acute onset of inattention, altered level of arousal, and other mental status abnormalities. Delirium is extremely common in acute stroke, affecting at least 1 in 5 such patients admitted to hospital. It is a serious complication of stroke, being associated with higher mortality, longer length of hospital stay and higher dependency at discharge. The pathophysiology of delirium is not completely understood, and there are no specific treatments. This thesis investigated the role of cortisol in the development of delirium after stroke and also investigated the role of delirium and of cortisol in the development of cognitive impairment in the 12 months after stroke. The thesis specifically investigated whether levels of cortisol in saliva are elevated in delirium and also whether there is a loss of the normal diurnal rhythm in delirium, evidenced by elevated afternoon salivary cortisol levels and reduced morning level to afternoon level ratio. The thesis also investigated whether cortisol levels are persistently elevated in the year after stroke in those who developed delirium and whether cortisol levels are associated with cognitive decline. Finally it investigated whether acute and/or chronic changes seen on Computed Tomography (CT) brain scans taken around the time of stroke onset are associated with the development of delirium after stroke A longitudinal cohort study was conducted in 95 participants aged 60 years or over, who were admitted to hospital with a clinically confirmed stroke. Participants gave informed consent, or proxy consent was obtained if they lacked capacity to consent. At baseline participants underwent brief cognitive testing and were then assessed for the presence of delirium, using DSM IV criteria, at regular intervals during the first two weeks after stroke. At each assessment a saliva sample was collected in the morning and in the afternoon, to measure cortisol. Participants were then visited at 1 month, 4 months and 12 months after stroke onset, at which point they were assessed for the presence of delirium, further saliva samples were taken and a cognitive test battery was completed. 26 (27%) participants developed delirium during the course of the study period. The study found elevated salivary cortisol levels in those with delirium at up to 4 months after stroke, but at 12 months there was no difference between the delirium and no delirium group. A loss of the diurnal rhythm was seen in those who developed delirium at 5 days after stroke, but the diurnal variation had returned to a normal pattern at follow-up. However, in a multivariate analysis, controlling for age, sex, stroke severity (NIHSS), current illness burden (APACHE II), chronic illness burden (CCI) and prior cognitive impairment (IQCODE), neither median salivary cortisol levels in the first two weeks after stroke, nor the ratio of morning to afternoon cortisol levels were independent predictors of delirium diagnosis, although median 9am cortisol approached significance (OR=0.95, 95% confidence interval (CI) 0.89-1.01, p=0.08). In a random effects logistic regression analysis, the probability of developing delirium decreased over time from stroke onset and increased per unit increase in salivary cortisol (nmol/L), however this effect was not statistically significant (OR 1.02, CI 0.84-1.19 P=0.70 for morning cortisol and OR 1.05, CI 0.82-1.25 p=0.46 for afternoon cortisol). Global cognition, measured by the MoCA, was significantly poorer in the delirium group at each time point throughout the 12 months after stroke. However, there was a trend towards improvement in MoCA scores in the whole cohort throughout the 12 month follow-up, with the exception of those who developed the most severe delirium. The presence of delirium at any point during the 12 month follow-up did not affect the rate of change of the MoCA scores over the 12 months after stroke. The presence of brain atrophy identified on admission CT brain scans was independently associated with delirium (OR 3.7, CI 1.15-11.88, p=0.02), however the presence of a visible acute or chronic stroke lesion and the presence of white matter lesions were not. Finally, those who developed delirium had a worse functional outcome, longer length of hospital stay and were more likely to require institutional care or a package of care at home, compared with those who did not develop delirium. This thesis has contributed to our understanding of the mechanisms and phenomenology of delirium after stroke, and has also highlighted areas for further research which will be required to unpick the complex pathophysiology of delirium.