Delirium is a complex neuropsychiatric syndrome that impacts adversely upon patient outcomes and healthcare outcomes. Delirium occurs in approximately one in five hospitalised patients and is especially common in the elderly and patients who are highly morbid and/or have pre-existing cognitive impairment. However, efforts to improve management of delirium are hindered by gaps in our knowledge and issues that reflect a disparity between existing knowledge and real-world practice. This review focuses on evidence that can assist in prevention, earlier detection and more timely and effective pharmacological and non-pharmacological management of emergent cases and their aftermath. It points towards a new approach to delirium care, encompassing laboratory and clinical aspects and health services realignment supported by health managers prioritising delirium on the healthcare change agenda. Key areas for future research and service organisation are outlined in a plan for improved delirium care across the range of healthcare settings and patient populations in which it occurs.
Statistics from Altmetric.com
Delirium: a prime target for early intervention
Delirium is a serious neuropsychiatric syndrome that occurs in approximately one in five hospitalised patients, with rates of up to 90% reported among patients in palliative and intensive care settings.1 No other disorder has equivalent penetration across healthcare settings such that delirium has been likened to a ‘cognitive superbug’ that warrants the often effective preventative focus that infectious diseases have received in our healthcare environments.2 It is the commonest complication among hospitalised elderly and is associated with longer hospital stays, more complications, greater costs of care, reduced subsequent functional independence, and increased inhospital and subsequent mortality.1 ,3 Importantly, these adverse health and social outcomes are relatively independent of confounding factors such as morbidity level, baseline cognition, age and frailty. A quarter of older medical patients with delirium die within 1 month of its onset4 with a recent study estimating that the risk of mortality increases by 11% for every additional 48 h of delirium.5 In addition, delirium may be an accelerating and possibly causal factor in the development of dementia.6–9
Despite this impact upon healthcare outcomes, delirium has been understudied in comparison with other neuropsychiatric conditions2 and there is much evidence to indicate that its management in everyday practice is suboptimal with inconsistent detection and treatment practices. Typically two-thirds of cases are undetected or misdiagnosed.1 ,10 Many cases are not formally recognised until symptoms have been present for days or weeks, presenting an avoidable delay in providing effective treatment. This reflects the challenges of recognising the often complex and fluctuating clinical presentation of delirium, but also relates to a lack of robust and consistently applied procedures for detecting and monitoring of delirium, even in patients at high risk. Non-detection is associated with especially poor outcomes including mortality.11 Non-detection also means that especially distressing aspects of delirium (such as hallucinations and delusions) are inadequately managed, and the lack of a clear explanation from professionals often provokes bewilderment and distress in carers.12 The frequency of delirium, its impact upon healthcare outcomes and evidence that delirium care can be substantially improved, all point towards delirium as a key target for improved practices.
Bench: unravelling the psychobiology of delirium
The clinical presentation of delirium includes a broad range of features indicative of widespread disturbance of brain function. In addition, the heterogeneous aetiology of delirium emphasises the connection between brain and body wellness. The principal pathophysiological theory of delirium emphasises the combined effect of brain insults (eg, hypoxia) with aberrant stress responses (eg, excess cortisol and exaggerated central nervous system (CNS) inflammation) perhaps inducing neurochemical (eg, relative cholinergic deficiency and dopaminergic excess) abnormalities that are expressed as delirium.13 The substantial interactions among cholinergic function, cognition and inflammatory processes suggest common mechanisms but their respective roles as predisposing or precipitating factors or merely epiphenomena requires clarification.
Animal models of delirium can allow direct investigation of pathophysiology and testing of novel therapies. Trzepacz et al14 developed a hypocholinergic model based on atropine administration to rats that induces changes in EEG, behavioural and maze performance that mirror features of human delirium. More recently, a model based upon activation of CNS inflammatory processes via peripheral inflammatory stimuli in animals with pre-existing neurodegenerative pathology has been used to explore the interaction between delirium and dementia. This work suggests that prior neurodegenerative pathology is associated with greater susceptibility to delirium due to an exaggerated response to systemic and CNS inflammatory signals due to factors such as microglial priming and synaptic loss at key areas of the brain (such as the hippocampus) which in turn are associated with heightened and more prolonged transcription of inflammatory mediators.15
Biomarkers can illuminate our understanding of delirium pathophysiology but can also assist in determining delirium proneness, allow for more accurate detection of delirium and assessment of illness severity, as well as monitoring of illness course and treatment response. Recent studies have identified evidence for serum, cerebrospinal fluid and genetic factors that (a) relate to increased baseline risk of delirium (APOE4 allele, low IGF-1), (b) link to the emergence of delirium (elevated C-reactive protein, cortisol, interleukin (IL)-8), (c) relate to delirium presence and course (IL-6, IL-8, IGF-1, chemokines) and (d) indicate neurotoxicity and possibly delirium severity (S-100B, NSE).16 ,17 Importantly, these studies take account of possible confounding factors such as age, prior cognitive function including dementia and severity of morbidity. Their integration into predictive models that include known clinical risk factors may allow for enhanced delirium prediction as well as illness monitoring.
Other measures can also assist: the EEG shows generalised slowing in delirium and can help in distinguishing delirium from dementia, but this pattern lacks specificity18 and the practicalities of performing EEG on delirious patients limit widespread use. The few neuroimaging studies in delirium suggest that white matter disease may be an important predisposing factor for delirium.19 More recently, a small study of intensive care unit (ICU) patients found that duration of delirium correlated with degree of neurocognitive disturbance and white matter disruption at both 3- and 12-month follow-up.20 ,21 Other work has linked peripheral measures of anticholinergic activity (serum anticholinergic assay (SAA)) to delirium proneness but findings are inconsistent and SAA may not correlate closely with central cholinergic function.18 Melatonergic metabolism appears to be differentially affected in hypoactive versus hyperactive clinical subtypes of delirium22 highlighting the potential importance of circadian function as a key factor in delirium. Postoperative delirium is characterised by reduced serum levels of amino acids with low tryptophan (<40 µg/ml) suggested as a measure for delirium detection.23 However, altered amino acid levels were not found in elderly medical patients24 and may reflect the physiological response to surgery rather than a specific pathophysiological mechanism for delirium.
A variety of neuropsychological measures can also identify patients at risk of delirium for whom preventative measures are especially recommended. Pre-existing dementia is a well-recognised risk factor for delirium, but more subtle impairments of attention, vigilance, memory, visuospatial function, graphomotor speed and executive function25 are also linked to elevated delirium risk. A key challenge is to apply this evidence to identifying a neuropsychological profile that reliably predicts delirium risk and that can be applied along with other clinical predictors to routine screening in real-world settings and linked to remedial/preventative interventions especially in those who are at high risk or facing periods of elevated risk due to elective interventions.
Delirium at the bedside
The diagnosis and management of delirium in everyday clinical practice is challenging because (a) delirium is a complex neuropsychiatric syndrome with considerable heterogeneity in clinical presentation that overlaps with dementia and depression, both phenomenologically and as comorbidities; (b) delirium involves patients with complex multi-morbidity and is frequently trivialised as a symptom rather than a syndrome (with a legion of synonyms such as acute confusion and brain failure) and therefore not addressed as a condition in its own therapeutic right; (c) delirium occurs across all healthcare settings and thus falls outside the clear remit of any one particular medical specialty; and (d) issues of consent complicate the conducting of high quality clinical trials of delirium treatment and consensus regarding optimal management remains lacking.
Delirium is a highly predictable occurrence and thus amenable to primary prevention. Models of delirium pathogenesis emphasise ‘delirium readiness’26 where the interaction of a range of predisposing factors with acute precipitants results in the emergence of delirium. The most consistently studied model is that of Inouye and Charpentier27 who identified four predisposing factors (cognitive impairment, severe illness, visual impairment and dehydration) and five precipitating factors (polypharmacy, catheterisation, use of restraints, malnutrition and any iatrogenic event) that predicted a 17-fold variation in the relative risk of older medical inpatients developing delirium. Subsequent work has validated this model in older patients undergoing hip surgery.28 An important implication is that patients with high baseline vulnerability can develop delirium even in response to minor precipitants.
Prevention: non-pharmacological interventions
The incidence of delirium can be reduced by multicomponent interventions that minimise exposure to known risk factors for delirium. Moreover, incident delirium is less severe and of shorter duration when it occurs in the setting of active management of delirium proneness.29 ,30 Interventions include many common elements: assisting orientation, enhancing efficacy (eg, sensory), promoting healthy sleep–wake cycles, pain relief, optimising physiological parameters (electrolytes, hydration), mobilisation and active review by specialist nurses or geriatricians/geriatric psychiatrists.29–32 Success relates closely to degree of implementation; Marcantonio et al,33 for example, trebled detection in postacute nursing facilities with a nurse-led delirium detection and treatment programme but achieved only modest implementation rates for the actual intervention and did not reduce delirium persistence at 1-month follow-up.
The complexity of delirium is such that relatively simple interventions (eg, consensus guidelines, educational interventions) have limited impact.34 ,35 Novel programmes that are tailored to the needs of specific settings such as postacute facilities,33 community-based care homes36 and palliative care services34 reflect the demands of providing targeted and individualised care for a multifactorial condition. Moreover, existing evidence suggests that preventative measures are more effective in patients at high risk of delirium for reasons other than dementia.32 ,33 ,37–39 A key concern for future work will be to enhance alignment between specific interventions and the needs of individual patients.
Additional opportunities for earlier and more optimal intervention exist for patients who are entering periods of increased risk for delirium by virtue of elective medical and surgical interventions. ‘Prehabilitation’ involves risk factor reduction to optimise preparedness for non-emergency procedures or other interventions. Bjorkelund et al40 significantly reduced delirium incidence (from 34% to 22%) in patients undergoing hip fracture surgery with an intervention addressing hydration, oxygenation, analgesia, polypharmacy and optimising care environment routines that commenced during the prehospital period.
Pharmacological management of delirium proneness
Delirium risk is closely linked to pharmacological factors that include polypharmacy in general, sudden withdrawal of certain classes of drugs and administration of particular deliriogenic agent classes with anticholinergic agents, benzodiazepines and opioid medications consistently implicated across clinical populations.41 It is important to be aware that many commonly prescribed agents have clinically significant anticholinergic effects, including some that would not immediately be recognised by many clinicians.42 Avoiding non-essential medications is fundamental to minimising delirium predisposition. Where drug treatments are necessary, prudent use can be guided by a careful analysis of risk to benefit ratio. Of note, uncontrolled pain due to underuse of opioid analgesia is also linked to elevated delirium incidence,43 emphasising the need to balance symptom control and toxicity. Protocolised care can guide the optimal use of medications in populations with complex needs. For example, Skrobik et al44 reported reduced duration of mechanical ventilation, improved pain management and reduced delirium symptoms with protocolised management of pain and agitation in the ICU. However, the application of such systematic practices remains highly variable in everyday practice.45
In accordance with the prevailing acetylcholine–dopamine imbalance neurochemical theory of delirium, studies exploring pharmacological prevention have focused upon agents that diminish dopaminergic or enhance cholinergic function in high-risk populations. Overall, the evidence for antipsychotic agents is encouraging with placebo-controlled studies demonstrating prophylactic impact for typical and atypical agents in elderly patients undergoing surgical interventions,46–48 but there are uncertainties as to their role across settings, with limited work in high risk older medical inpatients and patients with comorbid dementia. In addition, optimal dosing and timing of treatment are uncertain since studies have focused on short term use at low doses. Studies of procholinergic strategies have had largely negative findings despite the theoretical basis for a possible effect.49 ,50 The interaction between pharmacological and non-pharmacological interventions remains poorly studied to date.
Improved recognition of delirium represents a major opportunity for improving patient outcomes because typically two-third of cases are diagnosed late or missed in everyday practice across settings.1 ,10 Early detection of delirium can allow for more timely and effective intervention while non-detection is associated with poorer outcomes that include elevated mortality.11 Poorer detection rates occur with hypoactive presentations, cases complicated by comorbid dementia, and in patients with a history of previous psychiatric problems, prominent pain or receiving care in surgical settings.25 Importantly, these features also characterise patients with the poorest outcomes.
Improving delirium detection requires better appreciation of its varied clinical presentation which can include a broad range of cognitive impairments and neuropsychiatric symptoms. Educational and other interventions need to raise awareness of less obvious presentations such as those that involve hypoactivity as they are easily mistaken for fatigue or frailty in the seriously unwell or postoperative patient. Although the onset of delirium is typically acute, with symptoms occurring over hours or days, full syndromal illness is frequently preceded by a prodromal phase during which early detection and prevention of transition into full syndromal illness can be achieved.51 However, efforts to characterise this prodromal phase have variously emphasised non-cognitive symptoms, deterioration in cognitive function, and a variety of non-specific complaints including reduced pain tolerance, anxiety, general malaise and being ‘not themselves’.52–55 Efforts to identify early indicators of delirium should therefore include additional symptoms beyond those that characterise full syndromal illness.
A range of tools can assist in delirium detection but choosing the optimal approach in any setting requires careful consideration of the characteristics of the patient population to be assessed, time available, the expertise of healthcare staff responsible for assessment and the availability of expert clinicians to clarify diagnosis. In general, delirium detection is a two stage process involving initial screening with a brief, simple and sensitive instrument followed by formal diagnosis using DSM-IV criteria and/or assisted by an instrument with high specificity for delirium. Initial screening can be done using simple tests of attention which have high sensitivity and relative specificity for delirium such as the digit or spatial span, serial sevens test, or months backwards test.56 More delirium specific tools include the Confusion Assessment Method,57 Nu-DESC58 and Delirium Observation Scale.59 The last two are observational scales that are preferable for use by staff who are less adept at cognitive assessment. Patients who screen positive can be more formally assessed for delirium by applying DSM-IV criteria. These can be supplemented by more detailed instruments that provide diagnostic cut-off scores for delirium such as the Memorial Delirium Assessment Scale,60 Cognitive Test for Delirium (CTD),61 Delirium Rating Scale (DRS)62 and Revised Delirium Rating Scale (DRS-R98).63 A key challenge in more complicated cases is to distinguish delirium from other neuropsychiatric conditions such as depression and dementia and the DRS versions as well as the CTD have demonstrated ability in this regard.
Other measures may be useful markers of functional changes that indicate emerging delirium. Matsushima et al64 found that background slowing on EEG and sleep disturbance preceded the emergence of overt delirium in coronary care unit patients, while Osse et al65 identified activity changes measured on an actigraphic device indicative of early delirium. Other recent work has highlighted changes in motor functional performance (measured by the Trunk Control Test and the Tinetti Scale) as a marker of emerging and resolving delirium and that distinguishes the delirious state from dementia without delirium.66 Cognitive tests that are sensitive to delirium can be readily incorporated into computerised devices that allow for greater consistency and ease of interpretation in cognitive assessment in routine practice and by a range of healthcare staff.67 Newer technologies (eg, Smartphone applications) are widely available and could facilitate improved delirium management by (a) increasing access to screening tests, (b) providing decision support to guide diagnosis and management in real-time and (c) allowing for accurate and convenient measurement of sleep, activity and other physiological functions. The minimal inconvenience and discomfort for patients make these approaches particularly attractive in the study of delirium.
Management of delirium episodes
Secondary prevention to reduce the impact of delirium requires more systematic management with judicious use of all available effective interventions, both pharmacological and otherwise. Delirium is a challenging condition to study due to issues with capacity and consent in cognitively impaired patients. In addition, concerns regarding the potential for adverse effects in highly morbid patients, many of whom have concomitant dementia, act as considerable deterrents to pharmacotherapy. As a consequence, there is much inconsistency both in recommended guidelines and actual clinical practice across disciplines with, for example, pharmacotherapy perceived more favourably by palliative and intensive care physicians compared with specialists in care for the elderly.68
Poor outcomes associated with delirium are linked to a variety of factors: the underlying causes of delirium, complications of active delirium, poor cooperation with medical care, direct neurotoxicity of the delirious state and adverse effects of treatment. Delirious patients are prone to serious complications that include malnourishment, dehydration, pneumonia, urinary incontinence, falls, intravenous line removal, pressure sores, uncooperative behaviour and problems with consent. These are closely related to particular symptoms of delirium (eg, hypoactivity and pressure sores, hyperactivity and falls) with poor outcomes predicted by the severity and duration of symptoms.5 ,69–71 More timely and effective intervention thus has considerable potential to reduce these problems but evidence for reduced delirium duration with pharmacotherapy is limited to (a) two small placebo-controlled studies of quetiapine which indicate more rapid recovery in elderly medical72 and ICU73 patients and (b) studies of prophylaxis where some evidence suggests that emergent delirium is of shorter duration and less severe in active treatment arms74 but this is not consistently reported.46 Specific studies exploring the relationship between timing of intervention and outcomes, and exploring the impact of earlier and proactive pharmacotherapy are needed.
An important implication of recent research is that the traditional concept of delirium as a highly reversible condition is not borne out in practice where many patients experience persistent illness and/or longer term cognitive impairment.7 The extent to which these difficulties reflect unresolved or recurring delirium, a separate neuropsychiatric state or emergent dementia remains uncertain, but studies suggest that delirium can accelerate the course of existing dementia and also seems to be associated with an increased risk of dementia in patients without evidence to indicate neurocognitive impairments prior to the delirium episode.6–8 ,75 Other work suggests that persistent neurocognitive impairments occurring after delirium may be different in character to those of dementias8 ,20 but further work is needed, including studies that investigate for the neuropsychological footprint associated with long term cognitive impairment. In practical terms, these observations further emphasise the need for sustained therapeutic effort until full symptom resolution. However, efforts are restricted by a lack of clarity as to what represents recovery in delirium and in real-world practice many delirious patients are discharged to postacute facilities and nursing homes before full resolution of symptoms even though these settings are often not well equipped to cater for the many challenges that delirious patients pose.76 Studies need to explore the therapeutic impact of interventions delivered beyond the period of acute hospitalisation, including cases that have settled to subsyndromal severity. Such evidence can guide optimal practices around continued care for patients who have experienced delirium.
Psychological management of sequelae
Delirium is a distressing experience for patients and their caregivers. Delirium-recovered patients may be uncomfortable discussing their episodes because they equate it with loss of mental or self-efficacy. Particularly distress is particularly linked to the experience of psychosis and psychomotor disturbance.77 The psychological aftermath of delirium in understudied, but around 50% of patients can recall the episode and in many cases are still distressed by their recollections 6 months later.78 Persistent psychological disturbances are a particular target because they can impact upon subsequent help-seeking behaviour. Follow-up visits can facilitate postdelirium adjustment by allowing for discussion of the meaning of delirium and planning of how to minimise future risk (eg, by addressing risk factors such as medication exposure and sensory impairments).
Better bedside care in practice
Improved delirium care in everyday practice can be achieved by identifying a set of brief and convenient screening procedures that allow for more consistent and efficient identification of cognitive and/or behavioural changes that indicate possible delirium. These must be linked to specific diagnostic and evidence-based therapeutic interventions which should be used as a matter of urgency. For identified cases of delirium, the course of illness should be monitored closely to assess treatment response and whether patients experience full recovery across all symptom and functional domains. These efforts can guide care planning and ongoing rehabilitative efforts (see figure 1).
Boardroom and beyond: penetrating the healthcare agenda
Delirium warrants special status in our increasingly aged and thus delirium-prone society because of its unrivalled penetration of healthcare settings, severe impact upon outcomes and high economic cost, coupled with its predictability and demonstrated preventability. It is a multifactorial condition with a highly heterogeneous course that provides multiple opportunities for cost effective intervention.79 Patients at high risk are readily identified and many factors that predict the likelihood of developing delirium are highly preventable—including many iatrogenic elements. Moreover, evidence exists for preventative and treatment interventions that are underutilised in real-world practice. Research needs to focus on identifying how this gap between knowledge and practice can be bridged. However, delirium has been overlooked by health services management and is not emphasised in routine activity data or as a key performance indicator.80 Ongoing service development provides ideal opportunities to incorporate delirium-friendly practices that minimise delirium risk and provide consistent detection aligned to effective evidence-based management.
The challenges of delirium are increasingly rooted between the sometimes conflicting needs of improved risk management within organisations versus the provision of individualised, patient-focused care that promotes autonomy and dignity. Complications are often documented more thoroughly than the actual clinical conditions that underlie poor outcomes.81 Concerns regarding the safety of delirious patients at risk of falls and wandering can drive restrictive care practices that inhibit reorientation, mobility and self-efficacy, all of which are recognised elements of good delirium care. A compromise can be achieved with creative solution-focused collaboration. For example, less restrictive care can be facilitated by electronic alarms and pressure mats to monitor patient behaviour and alert staff when vulnerable patients are at risk of wandering or falls.82
Costs of care and patient safety are performance indicators that are prioritised by healthcare managers and substantially impacted upon by delirium. However, the increased emphasis on rapid patient turnover is poorly conducive to addressing the detailed assessment and care planning that patients with delirium need. This philosophy has extended to proposals to withhold payments for costs generated by preventable complications like delirium.83 However, such developments risk further complicating accurate documentation of delirium and have the potential to exclude high risk patients from care. In essence, the recognition and documentation of delirium are unlikely to be improved by linking it to financial disincentives. Managers and clinicians must work in partnership towards solutions that consider all these elements. One obvious target is to develop systems that allow for better coordination of interdisciplinary efforts and monitoring of delirium-relevant outcomes. Rudolph et al84 propose a strategy that includes a more systematic approach to screening for delirium, targeted intervention for those at high risk and better education for providers and the public. However, there is also a need for health service managers to recognise that delirium requires change at a systems level. This encompasses clinical care pathways, appropriate training opportunities for staff, suitable ward facilities and an incentivised system to reduce the incidence of delirium.
The gathering evidence for delirium prevention by addressing common risk factors can provide opportunities for improved care in community-based settings and primary care. In addition to prehabilitation of patients who are facing periods of elevated delirium risk, attention to more general risk factors such as polypharmacy and use of anticholinergic and benzodiazepine agents in elderly patients can also reduce the burden of delirium within society. Public health and other health service campaigns have been successful in reducing the extent of these problems but could be extended to specific programmes that emphasise how vulnerability to delirium can be minimised in high risk populations.
The past decade has witnessed considerable advances in our understanding of how to prevent and manage delirium. The emergence of the European Delirium Association and the American Delirium Society can foster and support research while advocating improved delirium care. However, everyday practice is still characterised by inconsistent application of existing knowledge. The quality of delirium care is a marker of the dysfunctionality of our health systems, highlighting the need for greater cohesion in everyday clinical management, more emphasis on educational programming and research funding, as well as administrative structures to ensure that this extremely common and prognostically significant condition is addressed across healthcare services. A set of proposals to improve delirium care are listed in box 1. Optimum delirium care requires better appreciation of its importance in research and education, at the bedside interface with individual patients and by policy makers at national and international levels. Meaningful change is required across the spectrum from bench to bedside to boardroom.
A 12-point plan for improved delirium care.
Promote awareness of delirium by including its detection and management as a fundamental and assessed educational component of medical and nursing programmes at undergraduate and postgraduate level.
Ensure that delirium risk status monitoring becomes embedded into daily routines across all healthcare settings, including hospital and community-based settings.
Routine and systematic implementation of known preventative measures, particularly in high risk patients.
Conduct high quality studies to explore the respective roles of non-pharmacological and pharmacological interventions in delirium management across populations.
Promote earlier and more consistent detection of delirium through formal systematic screening in everyday practice across healthcare settings.
Explore the utility of computer-assisted and other technologies in facilitating efficient and accurate assessment of mental status as well as in decision support around diagnosis and management.
Link delirium identification to evidence-based action through protocolised management based upon a more coherent understanding of the role of interventions in managing specific clinical presentations of delirium.
Research key uncertainties regarding management, including the optimal timing and duration of treatment, risk-benefit ratio in highly morbid patients, barriers to implementing evidence-based practices and the relevance of different clinical presentations / comorbidities to treatment choices.
Promote active management of the post-delirium phase, including the risk for subsequent episodes and prevention of secondary psychological sequelae.
Recognition (and prioritisation) of delirium in routine healthcare activity data to allow for more robust monitoring of the frequency of delirium and its impact upon outcomes.
Establish public health campaigns to promote awareness of delirium in the community, including risk reduction through pharmacovigilance and other measures.
Prioritization of delirium-relevant research by funding agencies with particular emphasis upon improved knowledge of detection, pathobiology and knowledge translation into routine clinical practice.
This work was supported by grants from the Health Research Board (DM, CE, WC, NO'R, CD) and the Medical Research Council (AMJM). DM, CE, WC, NO'R and CD are members of the Cognitive Disorders Research Group at the 4i Institute at the University of Limerick, Ireland. AMJM is a member of The University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology, part of the cross council Lifelong Health and Wellbeing Initiative. Funding from the HRB and MRC is gratefully acknowledged.
Contributors All the authors were involved in: paper conception and design; paper drafting; and final approval.
Funding This work was supported by grants from the Health Research Board (DM, CE, WC, NO'R, CD) and the Medical Research Council (AMJM). DM, CE, WC, NO'R and CD are members of the Cognitive Disorders Research Group at the 4i Institute at the University of Limerick, Ireland. AMJM is a member of The University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology, part of the cross council Lifelong Health and Wellbeing Initiative. Funding from the HRB and MRC is gratefully acknowledged.
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.