What we know about the possible link between delirium and dementia with Lewy bodies, and why we need to learn more.

Delirium is associated with an increased risk of long-term cognitive decline and dementia, and in particular dementia with Lewy bodies (DLB).1 Richardson et al. in this issue highlight the challenge of distinguishing delirium from Lewy body disease (LBD), which includes DLB and dementia in Parkinson's disease, as they share many key features. The authors review the published literature on diagnosis and management of delirium in LBD and report on three main concepts: (1) the lack of validated methods or tools to diagnose incident delirium in LBD; (2) scant evidence on predictors of delirium onset in populations with diagnosed LBD, although poorer olfactory function and REM sleep behavioural disorder appear to predict delirium in Parkinson's disease; and (3) further lack of clarity on which factors associated with delirium should increase the index of suspicion for prodromal DLB. As a result, effective and safe management of delirium in LBD is challenging, also because there is increased risk of adverse outcomes associated with the pharmacological options frequently recommended for the management of delirium (e.g., haloperidol).2 In a related paper in this issue, Cullinan et al. describe an investigation of the accuracy of delirium diagnoses in patients admitted to hospital with Parkinson's disease. Prevalence of delirium in this patient group was as high as 57% and even higher in the older patients. Hypoactive delirium was significantly less likely to have been identified and formally diagnosed and only 12% of discharge summaries included any diagnosis of delirium. The discrepancy between the high prevalence of symptoms of delirium during hospitalisation and the low percentage of discharge summaries with recording or indication of delirium underlines the need for increased education about delirium symptomatology and diagnosis in Parkinson's disease. The mechanisms underlying the increased risk of dementia and DLB after delirium are unknown, but likely multifactorial. Possible explanations include neurotoxic effects of delirium, and common risk factors and etiologic mechanisms. The central cholinergic deficit typical for dementia and more pronounced in DLB3 is one of the leading hypotheses for delirium co-occurrence.4 Specifically, delirium may result from altered external information processing because of a derangement in the neural subcortical ascending activating systems that use acetylcholine as a neuromodulator. Particularly implicated here are effects on thalamocortical neurons modulating the excitability of widespread cortical areas. Derangement of cholinergic neurotransmission affects a significant group of these neurons and is considered one of the major drivers for thalamocortical dysrhythmia.5 Thalamocortical dysrhythmia, in turn, takes consciousness to non-ordinary states, including sleeping or dreaming, visual hallucinations, cognitive fluctuations, and psychotic or dissociative states.6 All these conditions are characterised by altered states of consciousness, particularly when deprived of real external stimuli, as typically occurs during delirium. Interestingly, these symptoms are also core features of DLB. The progressive appearance of a pre-alpha rhythm, characteristic of thalamocortical dysrhythmia, on the electroencephalogram (EEG) is considered a supportive biomarker for DLB diagnosis, and abnormalities typical of DLB and characteristics of thalamocortical dysrhythmia have also been found in hospitalised cognitively intact patients who experienced delirium.7 These changes tended to disappear, suggesting that abnormal EEG patterns might represent a marker of the acute delirium state. During the course of the disorders that precipitate delirium, the thalamocortical system may therefore become deranged, but this can recover after the acute condition resolves if a neurodegenerative condition is not yet in progress. Research criteria for different types of prodromal DLB have been published,8 and these include delirium. Considering that a disturbance of consciousness is an essential prerequisite for diagnosing delirium and the role of thalamocortical system in consciousness, attention, and awareness, evidence of thalamocortical dysrhythmia on the EEG could be a potential indicator of the delirium type of prodromal DLB and thus worthy of further investigation. More longitudinal studies pursuing potential and established biomarkers are required to examine the trajectory and interaction of delirium and LBD. In this respect, a longitudinal study conducted on older adults admitted to acute hospitals in Norway will examine the frequency, risk factors, and prognosis of hospital delirium. Patients included will undergo clinical examination with questionnaires and cognitive testing, and biomarker collection including cerebrospinal fluid sampling, MRI, and EEG. The study will rely on interdisciplinary collaboration involving several departments, including emergency and geriatric medicine, anaesthesiology, and orthopaedics. From this cohort, 400 patients will be recruited for a longitudinal study, to address whether development of DLB is more common in people with a history of hospitalised delirium, compared with hospitalised older adults without delirium. The study will also test whether DLB-specific CSF and/or EEG findings are more common in older people with delirium compared with those without. In conclusion, frustratingly little is known about links between delirium and DLB despite their impact and important clinical, and potential neurochemical and electrophysiological overlap. Longitudinal studies following patients who have suffered from delirium are needed to find how often delirium may be a presentation of prodromal DLB. The potential for future research is exciting, with relevance to patients and healthcare personnel, both in acute and psychiatric wards. The study is financed by a grant from the Western Norwegian Regional Health Authority. The authors declare no conflict of interest. The peer review history for this article is available at https://publons.com/publon/10.1111/acps.13526. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.