Introduction

Apathy is defined as a decrease in “goal-directed” behavior and impaired motivation not determined by a diminished level of consciousness, cognitive impairment, emotional distress, and depression [1, 2]. Apathy can be a symptom of neurological or psychiatric conditions, although it was recently detected also in older adults with healthy cognitive functions [3], and can display three different phenotypes. “Emotional-affective” apathy is defined as the failure to create the required relationship between emotional-affective impulses and current or future conduct; “cognitive” apathy refers to difficulties in elaborating the plan of actions required for the ongoing or future behavior; “auto-activation” apathy refers to the inability to self-activate thoughts or self-initiate actions, as opposed to a relatively spared ability to generate externally driven behavior [4]. Apathy can be consistently found in several neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease dementia (PDD), frontotemporal dementia (FTD), or vascular dementia (VaD) [5]. Its prevalence increases during disease progression [6], and it is linked to faster cognitive deterioration and increased risk of institutionalization [7]. Overall, existing evidence shows that apathy strongly impacts on the quality of life of both patients and caregivers [8]. Nevertheless, data on apathy in different forms of dementia are sparse [9,10,11]: there is no agreement on a recommended tool for apathy assessment [12, 13], it is rarely investigated as a primary outcome in literature [14,15,16], its neurobiological correlates are unclear, and evidence on possible treatments is limited [17]. For the above reasons, apathy in dementia is often misdiagnosed and undertreated [18, 19].

The aim of this paper is to provide an overview on prevalence, assessment tools, neurophysiological and neuroimaging correlates, and therapeutic options for apathy in different forms of dementia including AD, PDD, FTD, and VaD. We will identify gaps in the literature and future perspectives and provide information for a proper management of apathy, to improve the quality of life of both patients and caregivers.

Methods

The following databases were used for this narrative review: WebOfScience, Scopus, Pubmed, and PsycINFO. The script used for the search is (apathy) AND (dementia). General keywords focusing on the role of apathy in dementia were included.

Inclusion and exclusion criteria

We included randomized (RCT) and non-randomized clinical trials, controlled before–after studies, and interrupted time series addressing diagnosis, assessment, or treatment of apathy in dementia. Meta-analysis, systematic reviews, and animal models were excluded. Abstracts and full texts from the original search were reviewed individually by different authors. Disagreement at the screening stage was resolved by including a third author at the full-text stage, if required. Studies on subjects with non-neurodegenerative diseases, head trauma, stroke, brain tumors, and psychiatric diseases, as well as those not addressing apathy were excluded. Inclusion criteria entailed studies on adults or older adults (aged over 65 years old) and the presence of cognitive impairment. The research included only articles in English language published after 2017.

Data extraction

Data from individual studies were extracted in accordance with the PICOS approach [20]. The following information was extracted from each study: (1) author/s and year of publication, type of study, (2) characteristics of the participants, (3) cognitive and/or psychological domains investigated, (4) apathy evaluation, (5) type of treatment or intervention, and (6) primary outcome of the study. Continuous variables are expressed as value ± standard deviation, while categorical variables as absolute number and/or percentage.

Results

The research produced 3570 articles. A total of 836 records were found in Web of Science, 1266 records were identified on Scopus, 533 records were identified through Pubmed, and 935 records were identified on PsycInfo. Identified records were entered into the Mendeley software for their management and elimination of duplicates. A total of 2223 records were screened according to title and abstract. After 1821 records were excluded, 402 articles were evaluated for full-text. Therefore, 102 articles were selected according to eligibility criteria (records excluded 300). Seventy-five articles were included in the qualitative analysis (Fig. 1). We described the main results, according to each major clinical disease. For each clinical condition, after a preliminary description, we reported prevalence, available treatments, tools used for apathy assessment, and results from neuroimaging and functional studies.

Fig. 1
figure 1

PRISMA flowchart of article selection

Full size image

Alzheimer’s disease

AD is a neurodegenerative disease characterized by the accumulation of neuropathological abnormalities such as β-amyloid plaques and neurofibrillary tangles. The prodromal phase of this disease involves neural loss, typically affecting the hippocampus, leading to a progressive atrophy in large-scale of networks widely undermining cognitive and neuropsychiatric functions. AD negatively affects activities of daily living with loss of motivation and interest in several aspects of life. A summary of studies addressing apathy in AD included in the review is shown in Table 1.

Table 1 Summary of studies addressing apathy in Alzheimer’s disease and vascular dementia included in the review
Full size table

AD—prevalence

Thirty-nine articles met the inclusion criteria for apathy in AD. Five of them investigated AD together with bvFTD [12, 21,22,23,24], primary progressive aphasia (PPA) [23], and subcortical ischemic vascular disease (SIVD) [13]. The prevalence of apathy varied from 26 to 82%, and it increases, along with its severity, with the progression of cognitive and functional decline and brain atrophy.

AD—apathy assessment

Seven apathy-specific scales were used: Apathy Evaluation Scale (AES) [13, 23, 25,26,27,28,29,30,31], Dementia Apathy Interview and Rating (DAIR) [32, 33], Dimensional Apathy Scale (DAS) [22, 34, 35], Lille Apathy Rating Scale (LARS) [12], Apathy Rating Scale (ARS) [36], Apathy in Dementia-Nursing Home Version Scale (AES-NH) [37]. Seven apathy subscales were included in more global questionnaires, such as Neuropsychiatric Inventory (NPI) [12, 21, 24, 32, 33, 38,39,40,41,42,43,44,45,46,47,48,49,50] or Frontal Behavioral Inventory (FBI-a) [12]. The most utilized tools were AES and NPI. Using AES, the highest score was reported in a sample of 32 patients (mean age 84.5 ± 9.5 years old) and was 56.5 ± 13.0 points [31], while the lowest score of 34.4 ± 13.40 points was found in a sample of 34 patients (with a lower mean age of 77.8 ± 6.51 years old) [13]. The highest NPI score was 7.8 ± 2.4, and it was observed in a sample of 200 patients aged 76 ± 5 years old [45], and the lowest NPI score was 0.6 ± 0.8, and it was reported in a sample of 95 subjects aged 75.6 ± 7.4 years [39]. In a large sample of 1925 patients suffering from AD, apathy was diagnosed in more than two third of the sample [47].

AD—apathy treatment

Nine studies addressed different therapeutic options, and methylphenidate was the most investigated treatment.

Methylphenidate

Efficacy of methylphenidate was investigated by two RCTs. Padala and colleagues demonstrated an increase on the clinician/researcher version of the AES after 12 weeks of treatment with 10 mg/day of methylphenidate. A greater improvement in apathy in multiple domains (behavioral, cognitive, emotional, and motivation), global cognition, functional status, caregiver burden, and depression started after 4 weeks [51]. Mintzer et al. in a 6-month RCT demonstrated significant beneficial effects of 20 mg/day methylphenidate on apathy. The largest decrease in the NPI apathy score was observed during the first 100 days. Conversely, no further improvements in cognitive measures and quality of life was reported after 6 months [45].

Sertaline

An RCT investigated the efficacy of sertraline, escitalopram, and nicergoline to treat apathy and depression in AD. Regarding apathy, the authors observed a significant improvement with sertraline only [52].

Ninjin’yoeito (NYT, TJ-108)

An open-label pilot study reported significant positive effects on apathy after using Ninjin’yoeito (NYT, TJ-108) in 20 AD subjects. NYT, TJ-108 is a traditional Japanese multicomponent herbal medicine. The authors reported an increase on NPI apathy and anorexia subscales and global cognitive functions [38].

BrainUp-10® (BU-10)

An RCT by Guzman-Martinez and colleagues investigated the effect of BrainUp-10® (BU-10) in 74 AD subjects. After 24 weeks of treatment, the authors obtained a statistically significant improvement of apathy. In contrast, no differences on cognitive functions were observed [53].

Oil diluted cannabis extract (Bedrocan)

Palmieri and Vadalà demonstrated benefits on apathy-subdomain score of NPI and cognitive status assessed by mini-mental state examination (MMSE) (in 45% of the sample), as well as on the caregiver’s quality of life, 3 months after the administration of Bedrocan [49].

Stimulation techniques

Two studies investigated the effect of repetitive transcranial magnetic stimulation (rTMS), showing significant effects on apathy [30, 40]. Nguyen and colleagues examined the efficacy of 5 weeks of rTMS combined with cognitive training in a sample of 10 AD patients. rTMS was delivered over 6 brain areas, namely, right and left dorsolateral pre-frontal cortex (DLPFC), right and left posterior parietal cortex associative areas, and Broca and Wernicke language areas, and it was combined with cognitive tasks stimulating the three different cortical regions (spatial attention tasks for the parietal cortex, naming of actions and objects, word recall and spatial memory tasks for the prefrontal cortex, and syntax and grammar tasks for language areas). For each region, 40 s of cognitive training were performed between each train of 10 Hz-rTMS stimulation. A series of 20 trains per session was administered, and the patients received 3 sessions. At the end of the treatment, a significant improvement of apathy, disability, and cognitive functions was observed [40]. A study on 20 AD patients by Padala and colleagues reported significantly greater improvement of apathy and general cognitive functions after 4 weeks of rTMS (10 Hz on DLPFC). Positive effects started after 4 weeks of rTMS treatment and were still durable at 12 weeks of follow-up [30].

Other treatments

Inel Manav and colleagues evaluated the effect of reminiscence therapy on apathy and cognitive performance in a cohort of 72 mild AD subjects. The protocol involved internet-based videos for 60 min once a week for 3 months and resulted in improving both apathy levels and cognitive functioning [36]. A pilot study showed the positive effects of 10 weeks of horticultural therapy on apathy, cognitive, and functional abilities in 32 AD subjects. Over the course of 10 weeks of activities, a statistically significant reduction in apathy was observed in the experimental group [31].

AD—neurological correlates of apathy

Four magnetic resonance imaging (MRI) studies identified the brain regions associated with apathy in subjects with AD. In patients with moderate levels of apathy, Huey and colleagues found a greater degree of atrophy in ventromedial and ventrolateral prefrontal cortex (PFC), posterior cingulate cortex (PCC), and the superior temporal sulcus. These regions entailed brain networks responsible for arousal, threat response, and reward processing [24]. Differently, a study in patients with moderate to severe AD and with moderate apathy, performed using tensor imaging (DTI), showed bilateral damage associated with the severity of apathy in the corpus callosum and internal capsule [37]. Wei and colleagues found specific grey matter (GM) atrophy for each apathy subdomain. Emotional apathy involved cerebellum, ventral PFC, and the amygdala; executive apathy entailed left orbito-frontal cortex (OFC), bilateral frontal pole, lateral temporal regions including the right middle temporal gyrus and temporal pole, as well as the left supramarginal and angular gyri; initiation apathy involved right medial PFC, left frontal pole, OFC, the right paracingulate, and anterior cingulate cortex (ACC) [22]. Kumfor and colleagues identified three different apathy domains through NPI, Cambridge Behavioral Inventory (CBI-R), and Disability and Dementia scale (DAD). For affective apathy, GM atrophy was detected in the left temporal poles, extending into the bilateral OFC, subcallosal cortex, and bilateral insula; behavioral apathy in the frontal cortex and subcortical areas, including the left caudate, extending into the nucleus accumbens, the right precentral gyrus, and the cerebellum; cognitive apathy in left orbitofrontal and subcallosal regions, extending dorsally into the medial PFC, anterior cingulate and superior frontal gyrus, inferior temporal gyrus, and posterior cingulate cortex [21].

Three studies used resting state functional MRI (rs-fMRI) to evaluate cerebral functional connectivity (FC) underlying apathy in AD. FC is a measure of how different brain regions interact with each other [54]. Buyukgok and colleagues compared 10 apathy-early-stage-AD subjects to 10 non-apathy-AD subjects and 10 cognitively normal groups and addressed a significant hypo-functioning at a trend-level in the anterior component of default mode network (DMN), specifically in the preangular ACC in apathy-AD subjects [44]. In contrast, another rs-fMRI study on subjects with mild to severe AD did not detect any decrease in the activity of DMN between subjects with apathy (n = 35) and subject without apathy (n = 35). Downregulation of DMN, which is a typical feature of AD, was observed only between AD and the healthy control group. Moreover, reduced connectivity between the left insula and the right superior parietal cortex and increased connectivity between DLPFC and the right superior parietal cortex emerged [41]. The different findings of the two studies may be due to several reasons. First, different characteristics of the sample and small sample size; second, different methods used to detect the networks; third, taking into account that the down-regulation is seen only at a trend-level, it may be possible that DMN does not mediate apathy in AD as hypothesized by Buyukgok and colleagues [44]. A recent study of Altunkaya and colleagues described the right inferior frontal gyrus (FG), left middle FG, and left anterior insula as critical hubs for apathy in AD [13].

Yeh and colleagues used proton magnetic resonance spectroscopy to investigate apathy-related neurochemical alterations. They explored only the frontal brain regions and discovered that each subdomain of apathy was associated with neurochemical variations in the ACC, without alterations in the OFC [26].

Jeong et al. assessed the association between regional cellular blood flow (rCBF) and neuropsychological symptoms (NPS) in early AD using single-photon emission computed tomography (SPECT) in a sample of 59 patients. Results showed apathy was associated with a decrease rCBF in prefrontal, pre/postcentral, and midbrain areas [48]. Apathy-related rCBF reduction in the midbrain may be one of the novel findings in patients with early AD. This finding supports previous evidence from preclinical and clinical studies suggesting that alterations in the midbrain structure and the dysfunction of the dopaminergic system may result in the expression of apathy.

Three studies using positron emission tomography (PET) in apathy-AD subjects showed dysfunction of ACC [12, 28], DLPFC [28], and OFC [55]. More specifically, Fernández-Matarrubia and colleagues compared the features of apathy in patients with behavioral variant of FTD (bvFTD) and AD [12]. Their study demonstrated that patients with bvFTD displayed lower metabolism in the left lateral prefrontal cortex, medial frontal/anterior cingulate, and OFC and anterior insular cortices, while patients with AD were characterized by dysfunction of medial/anterior cingulate circuit without an involvement of OFC and DLPFC [12]. According to the authors, those differences may be accountable for the greater impairment of the emotional dimension observed in apathy in bvFTD compared with AD patients [12]. On the other hand, Marshall and colleagues found an association between apathy and small tau clusters within the right ACC and DLPFC, which were more pronounced in individuals with greater amyloid burden [28]. Finally, according to Kitamura and colleagues, tau tangle formation in OFC was found in patients with apathy in AD. For this reason, the authors suggest a possible therapeutic role of novel antitau drugs on apathy; however, further neuroimaging studies are needed to monitor structural and functional changes in neuronal regions involved in apathy with and without therapeutic interventions [55].

Vascular and mixed dementia

VaD is the second most common dementia worldwide after AD, accounting for 15–20% of all cases in North America and Europe [56, 57]. Recent studies identified a continuum between VaD and AD, so that the coexistence of the two is defined as mixed dementia (MxD). Although pure AD and pure VaD can be diagnosed with good accuracy, identifying mixed forms can be challenging [58, 59]. In this review, we included 10 studies about VaD and MxD. Among them, only two analyzed the MxD subgroup [10, 60] (Table 1).

VaD—apathy prevalence

Akyol et al. estimated a prevalence of 55.2% [61]. Similarly, Santos and colleagues addressed a prevalence of 56.6% [60]. A large study by Schwertner and colleagues on 10,405 patients with dementia estimated a prevalence of 30.5% among individuals affected by VaD (n = 1.708) and 29.7% in MxD (n = 1.621) [10]. Two studies reported apathy prevalence according to VaD subtype: Bhat et al. compared large vessel disease (LVD) versus small vessel disease (SVD) (28.6%vs 54.2%) [62], while according to Manso-Calderon et al., apathy was more frequent in subcortical VaD (sVaD) than in cortical VaD (cVaD) [63]. Kazui et al. stratified the prevalence of apathy according to dementia severity, assessed through the Clinical Dementia Rating scale (CDR). For VaD, they reported 52.5% for CDR 0.5, 87.7% for CDR 1, 90.3% for CDR 2, and 91.7% for CDR 3 [64].

VaD—apathy assessment

The most used tool for apathy evaluation was NPI (n = 8) [10, 13, 60,61,62,63,64,65]. Three studies used the AES [13, 66, 67]. One study used both NPI and AES [13]. Tu and colleagues stratified NPI scores according to CDR in SIVD, reporting higher scores for both the symptom and the domain of apathy in more severe dementia [65]. According to Schwertner and colleagues, of the 1708 patients affected by VaD, 11.3% suffered from mild apathy (NPI 1–3), while for 19.7% apathy was clinically significant (NPI > 3). Among 1621 individuals with MxD, 12.9% suffered from mild apathy, while for patients 17.3% apathy was clinically significant [10]. Altunkaya and colleagues analyzed the prevalence of apathy subdomains, obtained by grouping the 18 items composing AES into “initiation” (factors involved in the act of beginning an action), “motivation” (the reason for acting), and “socially” (social involvement). They reported a mean total score of 45.2 ± 14.69 and a score of 0.63 ± 1.29 for “initiation,” 0.32 ± 0.97 for “motivation,” and 0.09 ± 1.05 for “socially” [13]. To analyze the behavioral pattern in apathy and compare it to depression, Saleh et al. designed an effort-based decision-making task via psych-toolbox (psychtoolbox.org) [66].

VaD—apathy treatments

No studies explored apathy treatments as primary outcomes in VaD.

VaD—neurological correlates of apathy

Seven studies used MRI to evaluate ischemic brain damage and structural and functional variations. Tu and colleagues and Saleh and colleagues analyzed DTI parameters like fractional anisotropy (FA) and mean diffusivity (MD) [65, 66]. Altunkaya et al. performed brain mapping using rs-fMRI to evaluate FC between four different resting state networks (RSNs), which are areas of the brain showing synchronous activity at rs-fMRI, in patients with SIVD and AD [13]. Akyol et al. investigated the relation between apathy and basic and instrumental activities of daily living. According to their study, cognitive and functional decline were risk factors associated with apathy, while no statistically significant relationship was found between age, gender, and apathy [61]. According to Altunkaya et al., the volume of white matter (WM) hyperintensities (WMH), dementia staging, and Beck Depression Inventory (BDI) were significant clinical predictors of apathy [13]. Saleh et al. found that in patients with apathy, the main criterion driving decision-making was reward magnitude, as they were less responsive to low rewards and high efforts. In apathy, there was a reduction of drift rate to the decision parameter (the rate of evidence accumulation), and this change positively correlated with WM alterations. This means that patients with apathy spent less time accumulating evidence before taking a decision, and action was rejected in a shorter time, as the reward was less appealing [66]. Tu et al. found that disruption of the right superior longitudinal fasciculus predicted apathy [65]. Tay et al. showed an impaired connectivity in premotor and cingulate regions [67].

Parkinson’s disease

Parkinson’s disease (PD) is caused by degeneration of dopaminergic neurons of the substantia nigra and striatum and manifests with cardinal motor symptoms (bradykinesia, rigidity, tremor, gait/postural instability), as well as non-motor symptoms, including NPSs and autonomic dysfunction [68]. Apathy, together with depression, is one of the most common NPSs both in PD and PDD [15, 16]. We found ten studies addressing apathy in PDD. All of them addressed apathy and common NPSs in PDD, and none of them exclusively investigated apathy. A summary of studies addressing apathy in PDD included in the review is shown in Table 2.

Table 2 Summary of studies addressing apathy in Parkinson’s disease dementia included in the review
Full size table

PD and PDD—apathy prevalence

According to Modreanu and colleagues, the prevalence of moderate-to-severe apathy was 28–40% in PD and 29–52% in PDD [11]. Two studies reported a percentage of 77% [16] and 97.5% [15] in PDD.

PD and PDD—apathy assessment

Three studies used NPI [16, 69, 70]. Nine studies used apathy-specific screening tools, and among them, the Starkstein’s Apathy Scale (AS) was the most frequently used [11, 68, 71, 72]. Two studies used the AES [15, 73]. One study used the Frontal Systems Behavior Scale—Apathy subscale [70]. Studies using NPI showed lower prevalence than those using apathy-specific scales. On the other hand, Moretti and colleagues used both AES clinician and caregiver versions and addressed the highest rates of apathy [16]. Campbell et al. used a general screening tool for psychiatric functions and an apathy-specific tool; however, the authors did not report the scores [70]. Bugalho and colleagues reported a significant worsening of apathy during a follow-up period of 4 years [72], while Horne et al. did not observe significant differences during the same interval [69].

According to three studies, in patients with PDD, apathy was more severe than in subjects with PD-mild cognitive impairment (PD-MCI) and cognitively normal PD patients (PD-CN) [11, 68, 73]. In contrast, Barbosa and colleagues found similar scores among PD-CN, PD-MCI, and PDD [71], and Camargo and colleagues did not observe differences in severity in PD subjects with and without dementia [15]. Interestingly, Barbosa and colleagues found different scores among participants with and without subjective cognitive impairment (SCI). Additionally, both PD and PD-MCI subjects with SCI reported higher apathy scores than same-category participants not addressing SCI [71]. No association was found between baseline apathy measures and future dementia risk [15, 69], as well as between demographic characteristics (such as age and gender) and levels of apathy. Patients progressing to dementia had significantly greater proportions of moderate-to-severe depression, visual hallucinations, memory complaints, and non-motor predominance at baseline compared to those remaining dementia free [11].

PD and PDD—apathy treatments

Moretti and colleagues conducted an RCT on the effect of rivastigmine on apathy and addressed apathy as the most constant NPS in PDD. Patients first received a 9.5 mg/24-h dose of transdermal rivastigmine for 3 months. Therefore, a maintenance dose of rivastigmine was administered for 12 months. The first evaluation reported that 77% of patients experienced apathy, with a serious impact on daily life and with severe relevance to caregivers. After 6 months, apathy decreased in prevalence (62%), but still severely affected daily living. After 12 months, the levels of apathy increased up to 69%, while other NPS (irritability, disinhibition, and euphoria) improved [16]. Camargo et al. did not evidence a correlation between disease duration or duration of levodopa therapy and depression or apathy. However, they indicated a correlation between motor changes and NPS. More specifically, depression scores were greater in patients with more severe motor impairment, but apathy scores did not correlate with worsening motor symptoms [15].

PD and PDD—Neurological correlates of apathy

No studies exploring neural correlates of apathy in PD or PDD as a primary outcome were found.

Frontotemporal dementia

FTD is a group of neurodegenerative diseases characterized by four clinical variants distinguished by early and predominant symptoms: bvFTD, PPA, semantic variant PPA (svPPA), and nonfluent variant PPA (nfvPPA) [74]. bvFTD is the most common subtype and presents with a heterogeneous combination of socio-affective symptoms and executive deficits [75, 76]. In 2011, the consensus diagnostic criteria for bvFTD proposed six core features of the disease, namely, apathy, disinhibition, loss of empathy, change in eating behavior, stereotypical behavior, and executive dysfunction [75]. A summary of studies addressing apathy in FTD included in the review is shown in Table 3.

Table 3 Summary of studies addressing apathy in frontotemporal dementia included in the review
Full size table

FTD—apathy prevalence

The prevalence of apathy in FTD was 54.8–88.0%. Apathy showed the highest reported frequency in FTD including bvFTD in ten studies [12, 21, 23, 76,77,78,79,80,81,82], while only two studies addressed PPA [76, 78]. In studies comparing FTD with AD, FTD patients had a higher prevalence of apathy [12, 21] and higher apathy scores [83, 84] than AD patients. Only one study highlighted a higher prevalence of apathy in patients with bvFTD in its early stage [82].

FTD—apathy assessment

Eight studies used NPI [21, 76, 78, 80, 82, 84,85,86], and five studies used AES [14, 23, 78, 81]. One study used LARS [12], three used CBI-R [21, 76, 77], two used the Motivation and Energy Inventory (MEI) [76, 78], three used the Snaith-Hamilton Pleasure Scale (SHAPS) [76, 78, 79], one used DAD [21], two used the Social and Emotional Assessment (Mini-SEA) [82, 87], one used AS [87], DAS was used in four studies [23, 79, 83, 88], one used the goal-directed behaviors (GDB) [81], and one used FBI [89]. Interestingly, one of the studies investigating bvFTD phenotypes related to WM changes found discordant scores on several apathy administration tests [78].

FTD—apathy treatments

None of the selected articles proposed pharmacological treatments for apathy. Godefroy et al. used a remote monitoring system named ECOCAPTURE@HOME to assess the evolution of apathy in individuals with dementia and its associated impact on their caregivers. They recruited 60 dyads of patients and caregivers (20 bvFTD, 20 AD, and 20 healthy control). Each dyad was monitored through a multisensory wearable bracelet and questionnaires on a smartphone application for 28 consecutive days. Building on the knowledge acquired through this first ECOCAPTURE@HOME study, they expected that the following phase of such a process will be to test a machine learning system, which could automatically estimate the behavioral markers of apathy and the associated caregiver’s perception of the dyad’s status using solely passive data from sensors [81].

FTD—neurological correlates of apathy

Included studies revealed that different clinical phenotypes of FTD have different profiles of functional decline and distinct patterns of associated cortical changes. Seven studies used MRI [14, 21, 77,78,79, 84, 85, 90]. One study used PET [12], and another one used SPECT [80]. In one study, 88 patients with bvFTD were included in a cluster analysis focusing on levels of disinhibition and apathy. Four phenotypic subgroups were identified in this study: primary severe apathy (n = 26), severe apathy and disinhibition (n = 26), mild apathy and disinhibition (n = 27), and primary severe disinhibition (n = 9). Results showed that apathy scores were associated with increased atrophy in the insula, inferior frontal, and anterior temporal cortex. Patients with severe apathy phenotypes were more functionally impaired with more extensive brain atrophy than those with mild apathy or primary severe disinhibition [77]. In another study, WM correlates of apathy and impulsivity were identified in the major syndromes associated with frontotemporal lobar degeneration, using diffusion-weighted imaging. The results highlighted three components associated with significant WM tract abnormalities: (1) care-rated change in daily skills, (2) changes in self-care, and (3) motivation correlated with widespread changes in dorsal frontoparietal and corticospinal tracts. In the neuropsychological tests of cognitive control, reflection impulsivity, and reward responsiveness were associated with focal changes in the right frontal lobe and supplementary motor area [78]. Another study showed that bvFTD subjects had greater impairment of emotional apathy and self-awareness than the AD sample. Such differences in the qualitative aspects of apathy appear to be associated with different functional and neuroanatomical substrates, as shown by FDG-PET imaging analysis of left lateral prefrontal, medial frontal/anterior cingulate, lateral insular cortex, and lateral OFC in bvFTD and right ACC [12]. Furthermore, neuroimaging results showed that among dementia syndromes (bFTD and AD), greater affective apathy was associated with reduced GM intensity in the ventral PFC and cognitive apathy with dorsal PFC. The presence of behavioral apathy was associated with the degradation of subcortical regions, including the caudate region of the basal ganglia, as well as the premotor cortex and cerebellum [21]. Moreover, they found that apathy progressed significantly in pre-symptomatic carriers of FTD-associated mutations and those individual differences in apathy at baseline predicted the severity of progressive deterioration in digit symbol test performance over time. In pre-symptomatic carriers, progression of apathy over 2 years was associated with baseline frontal lobe and cingulate gyrus atrophy. In contrast, subclinical cognitive impairments did not predict the worsening of apathy [90]. Another study investigated the topographic feature of the rCBF network in bvFTD patients with and without apathy. They found that both bvFTD groups retained the global function and characteristics typical of small mundanity but showed the loss of hubs distributed mainly in the PFC. Compared with the bvFTD no-apathy group, the bvFTD apathy group showed further loss of hub in the frontostriatal circuit, but recruited hubs in the angular gyrus, precuneus, and PCC. Overall, their results support previous findings from other neuroimaging modalities, in which apathy was related to frontostriatal circuit disruption in bvFTD [80]. Concerning cognitive apathy, an MRI study demonstrated that the presence of a deficit in planning ability is a predictor for a worsening in cognitive decline in both bvFTD and AD and it shares similar pathways with apathy. More precisely, alterations in the lateral and medial prefrontal and lateral temporal cortices as well as subcortical regions including the hippocampus, striatum, and thalamus were detected in both patients with apathy affected by AD and bvFTD [88]. Finally, O’Connor et al. evaluated the importance of considering anhedonia as a primary presenting feature of bvFTD and semantic dementia, with neural drivers distinct from those of apathy or depression. Overall, the results underscore the importance of apathy in functional impairment, highlighting the role of the right temporal region in disinhibition [77]. Lansdall and colleagues also suggested that apathy and impulsivity are correlated with WM tract abnormalities in the major syndromes associated with frontotemporal lobar degeneration, while daily skills, self-care, and motivation were probably related to diffuse changes in the frontoparietal and dorsal corticospinal tracts [78].

Discussion

Prevalence of apathy in different forms of dementia

Apathy was shown to have broad prevalence intervals in all types of dementia (26–82% for AD, 28.6–91.7 for VaD, 29–97.5% for PDD, and 54.8–88.0 for FTD). One possible explanation for the ambiguous prevalence is the use of multiple tools to quantify apathy and dementia severity. Most studies assessed the severity of dementia using MMSE, which may be not sufficiently accurate in moderate-to-severe dementia [91]. Nonetheless, while apathy appears to be more often connected with later stages of dementia, previous studies demonstrated that it could be prodromal for executive functions deterioration and also associated to an increased risk of developing cognitive decline in community-dwelling people [92]. Moreover, although apathy underlies similar symptoms to other disorders, such as depression [93], agreement on a standardized diagnostic system has not been established yet. Finally, except for a few research [14,15,16], all the included studies did not have apathy as their primary outcome. No studies investigating predictive risk factors for dementia primarily investigated the role of apathy, as opposed to depression or to other behavioral disorders. This has made it even more complex to outline a clear profile of apathy in relation to patients with different forms of dementia, although apathy is extremely related to these conditions.

Apathy assessment in different forms of dementia

The use of multiple diagnostic techniques for assessing apathy may explain the differences in outcomes. Several studies employed generic scales, such as the NPI, which might lead to inaccurate identification. Apathy is frequently underestimated and misdiagnosed as depression [4]. Due to the necessity for specialist training, apathy-specific measures like the AES are seldom used in clinical practice. Apathy has been discovered as a predictive biomarker for the likelihood of acquiring major diseases such as PD, cognitive decline, and severe dementia [94, 95]. An early diagnosis of apathy could help clinicians to detect dementia in earlier stages, as well as prevent frailty [96] or a misdiagnosis of depression. For these reasons, a more specific test appears to be important in assessing therapy response across multiple domains [97]. Given the prevalence of several phenotypes of apathy, the NPI appears to be a deficient scale in clinical practice for diagnosing apathy, whereas AES [98], DAS, or LARS appeared more appropriate to investigate the domains of apathy. More specifically, a systematic review reported that DAIR and AES clinician versions for AD, and LARS for PD, have the highest methodological quality and psychometric properties among the other validated apathy scales [99]. Correct diagnosis of apathy has substantial consequences for caregiver quality of life, [16] and it is likely that caregiver distress has important implications on the perception of the apathetic behavior of their assisted. Thus, employing a unique neuropsychological instrument and distinguishing scales administered by caregivers from those administered by clinicians may be relevant in future research to show possible variations in apathy identification and quantification due to the type of clinical scale.

Apathy treatments in different forms of dementia

Despite a recent increased research interest in apathy in dementia, the number of RCT for pharmacological treatments is poor. Methylphenidate produced the most meaningful findings for apathy in AD, as it improved different apathy domains. However, included studies reported discordant results on cognitive performance and caregiver distress [45, 51]. However, in these circumstances, the inclusion of non-apathy-specific testing may have a detrimental impact on the study’s results. rTMS paired with cognitive training was found to significantly enhance apathy as well as cognitive functioning among non-pharmacological therapies in AD [40]. However, as noted by a systematic review by Theleritis and colleagues, apathetic patients with dementia may benefit from personalized treatments combining both pharmacological and non-pharmacological interventions, including lifestyle changes and environmental adjustments [100].

Neurological correlates of apathy in different forms of dementia

The neurological basis of apathy is thought to be a disruption in networks controlling goal-directed behavior. Although the PFC is one of the key regions engaged in these tasks [4], changes in other brain regions might result in distinct phenotypes of apathy [101]. Recently, a new categorization of apathy was developed based on clinical findings of individuals with damage to PFC and basal ganglia [102]. Emotional apathy seems to be due to lesions in the connection between ventral-striatum fibers and orbitomedial PFC; cognitive apathy seems to involve neural connections between lateral PFC and dorsal caudate nuclei; auto-activation apathy seems to be due caused by bilateral lesions in limbic and cognitive territories in basal ganglia (internal globus pallidus and bilateral paramedian thalamus) and dorso-medial part of PFC. The finding of the neurological mechanisms underlying these illnesses is intimately connected to the function of apathy as a risk factor for neurodegenerative diseases. Neuroimaging studies in AD identified brain areas primarily related with apathy; nevertheless, even in this situation, the multiple neuropsychological tools employed to quantify apathy degree resulted in diverse and sometimes contradictory results. Indeed, Huey et al. using NPI found a greater degree of atrophy in ventromedial and ventrolateral PFC, posterior cingulate cortex, and the superior temporal sulcus [24]. Instead, Aguera-Ortiz L. et al., using the Apathy in Dementia Nursing Home Version Scale, addressed bilateral damage of the corpus callosum and internal capsule [37]. Similarly, to evaluate brain atrophy related to different apathy domains, Wei et al. applied the DSA [22], while Kumfor et al. used NPI, CBI-R, and DAD [9]. In the first study, they found GM atrophy along the brain regions implied in emotional apathy, executive apathy, and initiation apathy [22]. Conversely, in the second study, GM atrophy was detected in affective apathy, behavioral apathy, and cognitive apathy. These results demonstrated that the use of different neuropsychological tests may lead to heterogeneous and misleading data; therefore, an apathy-specific tool is recommended. Studies investigating the neuroanatomical substrate of apathy in AD and VaD supported the disruption of pathways connecting brain regions implicated in effort-based decision-making (disconnection theory) as the neurological substrate of apathy [13, 65,66,67]. In comparison to AD patients, FTD patients had a more widespread emotional deficiency, which was related with decreased self-awareness. According to an FDG-PET investigation, changes in qualitative characteristics of apathy appear to be related with several functional and neuroanatomical substrates [12]. Another neuroimaging study found that in both AD and bvFTD, affective apathy was associated with reduced GM intensity in the ventral PFC, cognitive apathy with dorsal PFC, and behavioral apathy with degradation of subcortical regions, premotor cortex, and cerebellum WM correlates of apathy and impulsivity were related with frontotemporal lobar degeneration in FTD using DTI. These WM anomalies, which were linked to alterations in daily skills, self-care, and motivation, were shown to be associated to changes in the dorsal frontoparietal and corticospinal circuits. The neuropsychological tests, on the other hand, yielded different results, linking these behavioral changes to alterations in the right frontal lobe and pre-supplementary motor area [78]. In conclusion, in both AD and VaD, apathy increases as dementia and WM atrophy progress [64, 103]. In FTD, the role of the right temporal region in disinhibition has been consistently highlighted, together with the correlation of apathy with WM tract abnormalities [77]. No data on neuroimaging correlates have been reported for PDD.

Conclusions

Despite its high frequency in neurodegenerative disorders, apathy is a neurobehavioral condition that is too often misdiagnosed. This review highlights that the appropriate diagnostic methods are still not employed in both clinical practice and research, as well as a paucity of studies that examine apathy as a primary outcome. Inadequate assessment leads to a lack of evidence on appropriate treatments for this disease. Possible therapeutic strategies to modulate this condition could be neurostimulation or cognitive training. However, agreement on the various assessment measures should be reached in order to test and reduce discrepancies between clinical studies. As far as we know, this is the first review that reports the literature on apathy based on the most common dementias. As the main limitation, the present study reflects the findings of a narrative review.