Dementia epidemiology and public health significance

The World Alzheimer Report 2016 estimated that 47 million people are living with dementia worldwide (Alzheimer's Disease International, 2016). In the inaugural World Health Organization Ministerial Conference on Global Action against Dementia, six of the top ten research priorities were focused on prevention, identification, and reduction of dementia risk, and on delivery and quality of care for people with dementia and their carers (Shah et al., Reference Shah2016). While the Lancet Neurology Commission has suggested that even minor advances to delay progression or ameliorate symptoms might have substantial financial and societal benefits (Winblad et al., Reference Winblad2016), advances have been slow.

Clinical trial crisis

Despite the articulation in the USA of a national plan to effectively treat and prevent Alzheimer's by 2025, clinical trial enrollment and success hangs in the balance (Fargo et al., Reference Fargo, Carrillo, Weiner, Potter and Khachaturian2016), and requires aggressive exploration of all avenues for early detection. The Food and Drug Administration has not approved a dementia drug since 2003 and clinical trials have shifted to disease prevention, with trials conducted in at-risk populations (Graham et al., Reference Graham, Bonito-Oliva and Sakmar2017) resulting in an ever widening age gap between clinical trial participants and patients who could benefit from these drugs (Banzi et al., Reference Banzi, Camaioni, Tettamanti and Lucca2016). The vast majority of clinical trials in dementia have failed due to: (1) gaps of knowledge in the biological mechanisms of the commonest forms of dementias, particularly the role of comorbidities, both vascular and neurodegenerative, which are prevalent and relevant to disease progression; (2) low signal-to-noise ratio for sensitive, specific, valid, and reliable measurement of cognitive impairment in the face of confounding factors, as well as for validated biomarkers as entry and/or end-point criteria; and (3) poor recruitment and retention, particularly in the asymptomatic and early disease stages (Gauthier et al., Reference Gauthier2016). Identifying prodromal disease is challenging and at times speculative, given the difficulties in detecting what is traditionally thought of as a cognitive disorder, prior to the development of cognitive features. Alternative markers are required to indicate risk and susceptibility to neurodegenerative disease, in order to capture and potentially treat earlier for incident cognitive decline and dementia.

Non-cognitive markers of neurodegenerative disease

Non-cognitive signs and symptoms are increasingly cited as possible early indicators of neurodegenerative disease, and are important for clinical screening (Ismail and Mortby, Reference Ismail, Mortby, Chiu and Shulman2017). Changes in hearing (Lin et al., Reference Lin, Metter, O'Brien, Resnick, Zonderman and Ferrucci2011), eye tracking (Hellmuth et al., Reference Hellmuth2012; Ladas et al., Reference Ladas, Frantzidis, Bamidis and Vivas2014), olfaction (Stanciu et al., Reference Stanciu, Larsson, Nordin, Adolfsson, Nilsson and Olofsson2014), and gait (Del Campo et al., Reference Del Campo2016) can all potentially signal neurodegenerative disease. However, the inclusion of these markers has not fully penetrated into clinical care, and they are not always simple or accessible for general clinicians. If early dementia detection and clinical trial recruitment are to come from settings outside specialty clinics, simple and easily obtainable at-risk markers are still required, as neurological examinations and laboratory tests or imaging may not be available or feasible.

Mild behavioral impairment: a potential link between healthy aging and neurodegenerative disease

Later life acquired neuropsychiatric symptoms (NPS) are recognized as a hallmark of the dementia process. Increasingly, however, NPS are considered an intrinsic aspect of prodromal stages of neurocognitive impairment and an early marker of cognitive decline (Mortby and Anstey, Reference Mortby, Anstey and Pachana2015). NPS are associated with greater functional impairment in clinical populations referred for cognitive impairment and neurobehavioral symptoms, irrespective of cognitive status (Peters et al., Reference Peters2006). The NPS Professional Interest Area (PIA) of the International Society to Advance Alzheimer's Research and Treatment (ISTAART), a subgroup of the Alzheimer's Association (AA), formalized the assessment of later life onset NPS as an at-risk state for cognitive decline and dementia with the development of proposed research diagnostic criteria for mild behavioral impairment (MBI) (Ismail et al., Reference Ismail2016). MBI is a neurobehavioral syndrome characterized by acquired, sustained, and meaningful later life NPS of any severity that identify individuals at an increased risk of developing dementia, but who do not necessarily exhibit overt cognitive symptoms. Notably, MBI can occur in concert with or in advance of mild cognitive impairment, thus potentially identifying earlier stage disease (Ismail et al., Reference Ismail2016). While the ISTAART-AA MBI criteria stipulate age of onset of emergent NPS at ≥ 50 years, after which the vast majority of dementia cases occur, it is reasonable to consider that even with very early onset dementias NPS can precede overt cognitive impairment. With the development of proposed research diagnostic criteria for MBI (Ismail et al., Reference Ismail2016), and a rating scale based on these criteria, the MBI checklist (MBI-C) (Ismail et al., Reference Ismail2017a), NPS can be studied systematically for their prognostic utility for cognitive decline and dementia, in pre-dementia populations.

Despite the increasing recognition of the importance of NPS as an early marker of dementia risk, NPS are still commonly overlooked in clinical care. When patients with NPS who do not exhibit obvious cognitive impairment are identified, either nothing is done or they receive psychiatric diagnoses and the possibility of neurodegenerative disease is overlooked (Ting et al., Reference Ting2010; Jalal et al., Reference Jalal, Ganesh, Lau, Lysack and Ismail2014; Ismail et al., Reference Ismail2017a). Given the lack of robust treatment options for NPS in pre-clinical states and dementia, the MBI construct may provide a new avenue for clinical trials to target NPS and ultimately reduce the risk of dementia.

Does mild behavioral impairment provide a way to improve recruitment of prodromal illness?

As noted, MBI characterizes a population at higher risk group for cognitive decline and dementia. Broad recognition among clinicians, researchers, and clinical trialists of the concept of MBI as a possible prodromal stage of dementia will have significant implications for improving early detection and prevention and treatment of dementia, and will help provide a better understanding of the very early consequences of neurodegenerative disease (Ismail and Mortby, Reference Ismail, Mortby, Chiu and Shulman2017). For the most part, recruitment in early phase illness comes from participants identified with higher dementia risk based on family history, or based on presentation with early cognitive symptoms in clinical settings. Sometimes illness has advanced significantly by the time of detection or formal assessment. Including MBI screening in clinical care, and in trial recruitment settings, can potentially increase the yield for early phase illness in a population who may have otherwise been overlooked.

Evidence supports biomarker correlates of the MBI domains of apathy (Lanctôt et al., Reference Lanctôt2016; Sherman et al., Reference Sherman, Liu, Herrmann and Lanctôt2017), emotional dysregulation (Babulal et al., Reference Babulal2016; Ismail et al., Reference Ismail2017b), agitation/aggression (Nagata et al., Reference Nagata, Kobayashi, Shinagawa, Yamada, Kondo and Nakayama2014; Ng et al., Reference Ng2017), social disinhibition (Fernández-Matarrubia et al., Reference Fernández-Matarrubia, Matias-Guiu, Moreno-Ramos and Matias-Guiu2015; Desmarais et al., Reference Desmarais, Lanctôt, Masellis, Black and Herrmann2017), and psychotic symptoms (Donovan et al., Reference Donovan2014; Kwan-Chun Ting et al., Reference Kwan-Chun Ting, Fischer, Millikin, Ismail, Chow and Schweizer2015; Fischer and Agüera-Ortiz, Reference Fischer and Agüera-Ortiz2017). However, research is still required using MBI operationalized criteria to systematically capture later life sustained NPS. Recruitment into trials for MBI as possible prodromal disease is based on the assumption that the same neuropathology and biomarkers that account for progressive cognitive impairment also result in later life acquired NPS. This assumption needs to be tested, as study of MBI may bring to light novel biomarkers for treatment targets, monitoring progress, or treatment response.

Further research is required to validate MBI as a target population and an outcome for clinical trials. First, we need to determine the prevalence of neurodegenerative pathologies in patients with and without MBI, using in vivo biomarkers and autopsy-based neuropathology. It is critical to determine whether or not selecting MBI enriches a trial population for these pathologies. Second, prospective studies are required to determine the risk of dementia and rate of cognitive decline in a population diagnosed with MBI. Current studies where MBI has been studied retrospectively have likely been affected by selection bias, where “psychiatric” patients are systematically excluded. Prospective studies with the inclusion of these patients, with a rating scale designed specifically for the purpose of detecting MBI as defined, will inform the calculation of risk for incident cognitive decline and dementia. Third, while MBI is proposed as an additional measure to allow earlier detection of neurodegenerative disease, the MBI symptoms themselves are not benign, and may be treatment targets themselves, as they are clinically significant and associated with greater caregiver burden (Sheikh et al., Reference Sheikh2017), functional impairment (Peters et al., Reference Peters2006), and faster cognitive decline (Ismail et al., Reference Ismail2016).

How do we capture MBI and operationalize its incorporation into dementia prevention clinical trials (pharmacological and non-pharmacological)?

With the larger goal of increasing prodromal disease signal detection, further education is required with respect to MBI and later life NPS as possible prodromal dementia markers. Historically, meaningful NPS have been exclusion criteria for clinical trials. However, there has been little discrimination between chronic psychiatric illness (e.g. schizophrenia and recurrent major depression) and new onset neuropsychiatric illness (i.e. MBI). A confound in MBI diagnosis is the emergence in later life of psychiatric symptoms as part of a psychiatric diagnosis, in contrast to a neurocognitive one. A good clinical assessment is required to explore these symptoms, and longitudinal follow up is required. However, recent evidence has even muddied these diagnostic waters. Using depression an exemplar NPS, which can be seen in the context of late life major depression or in the context of an emerging neurocognitive disorder, three large longitudinal studies have described how this boundary is blurred. The first study (Singh-Manoux et al., Reference Singh-Manoux2017) describes the emergence of sustained depressive symptoms in the decade before dementia increasing risk of incident dementia in this 28-year follow up of a 10,189 person cohort. The second study (Almeida et al., Reference Almeida, Hankey, Yeap, Golledge and Flicker2017) describes the link between depressive symptoms and incident dementia in a 14-year longitudinal study of 4,922 cognitively intact men, demonstrating that “depression is more likely to be a marker of incipient dementia than a truly modifiable risk factor” due to a time-dependent link between onset of depression and incidence of dementia. The association between depression and dementia was only apparent during the initial five years of follow-up, and was not observed if depression emerged more than five years before the onset of dementia. The third study (Tapiainen et al., Reference Tapiainen, Hartikainen, Taipale, Tiihonen and Tolppanen2017) is a nationwide nested case–control study of 27,948 pairs, which established that the importance of the associations between mental and behavioral disorders and AD were modest and dependent on the time window. Hospital-treated behavioral and mental disorders, including depression and other mood disorders, were associated with a higher risk of AD in the five-year but not the ten-year time window. The authors concluded that “some of the disorders may represent misdiagnosed prodromal symptoms of AD, which underlines the importance of proper differential diagnostics among older persons.” Thus, the time window is an important feature of neuroepidemiological research. While these studies do not state that all later life onset depression reflects a dementia prodrome, they do raise the issue of how to truly diagnostically frame later life onset psychiatric symptomatology in the absence of biomarkers for psychiatric symptoms. In the case of depression, the duration requirement of major depressive disorder is only two weeks to meet criteria, whereas it is six months to meet MBI criteria. Overall, further research is required into the link between NPS in older adults and risk for incident cognitive decline and dementia, and the construct of MBI allows further research to include foundational structural elements of symptom duration and natural history of later life onset to help address some of the questions raised.

With these diagnostic issues in mind, primary care clinicians, and those involved in clinical trial recruitment, may consider more closely the natural history and onset of neuropsychiatric symptomatology, and consider those with late-onset symptoms for further assessment and evaluation, rather than exclusion, which is the current standard. This may increase the yield of potential clinical trial participants, earlier on in disease course.