Delirium is a serious neuropsychiatric syndrome presenting with inattention and global changes in cognition
[1–3]. Delirium arises as a consequence of a neurological or systemic illness, medications and psychological stress. It is well-recognized that there is a relationship between predisposing (ageing, cognitive impairment) and precipitating (illness severity) factors such that in the setting of multiple (or severe) predisposing factors, fewer (or less severe) precipitating factors are required
. Delirium is therefore a sensitive marker of acute illness in vulnerable older people. This association with acute illness has resulted in the vast majority of delirium studies being undertaken in hospital cohorts
. However, this introduces selection biases as not all persons with delirium may reach medical attention. In addition, comparisons to pre-morbid cognitive functions are difficult.
In hospital samples, a common finding is that delirium contributes to persistent cognitive deficits, independently of predisposing and precipitating factors
. This has also been reported for subsyndromal delirium, where individuals have one or more of the diagnostic features of delirium
. Indeed, any examination of the utility of a delirium definition should incorporate criterion validity tests for future dementia. In prospective community cohort studies, hospitalization predicts adverse cognitive outcomes
[8–10], though none has been able to specify if delirium is a key determinant. Delirium is also associated with increased mortality
, and this should be another criterion by which any definition of delirium should be validated.
The point-prevalence of delirium in the community is thought to be low (0.7%, 95% CI 0.5 to 1.0 in the population aged ≥60 years), though this understanding is based on a systematic review identifying only three prevalence estimates in population samples
. Furthermore, epidemiological studies may under-estimate acute illness and/or prevalent delirium because people who are unwell are less likely to be interviewed. However, the period-prevalence may be higher. The Gerontological Regional Database (GERDA) study reported that 27% of persons aged 85 and older in the general population with delirium in the previous month
. This suggests that whole population samples could potentially investigate delirium more efficiently if stratified subsamples at higher risk for cognitive impairment are more intensively studied.
Delirium is clinically defined by application of a psychiatric reference standard such as the Diagnostic and Statistical Manual (DSM), where the core features are inattention, altered consciousness, cognitive and/or perceptual disturbance, acute and fluctuating change, in relation to a general medical condition. Based on this, there is an opportunity to construct an algorithmic diagnosis for delirium in population-based cohort studies collecting psychiatric symptoms. Such an approach is well-established in dementia, but yet to be systematically applied in delirium, and particularly not in population studies. Accordingly, using data from the population-based Medical Research Council (MRC) Cognitive Function and Ageing Study (CFAS) the aims of this study were to: (1) construct an algorithm for the diagnosis of delirium in population-based studies using the Geriatric Mental State (GMS) examination based on clinical principles; (2) test the predictive criterion validity of this algorithm against mortality and dementia risk; (3) report the age-specific prevalence of delirium as determined by this algorithm.