Neuropsychiatric Symptoms, Endophenotypes, and Syndromes in Late-Onset Alzheimer's Disease: Focus on APOE Gene

Neuropsychiatric symptoms, previously denominated as behavioural and psychological symptoms of dementia, are common features of Alzheimer's disease (AD) and are one of the major risk factors for institutionalization. At present, the role of the apolipoprotein E (APOE) gene in the development of neuropsychiatric symptoms in AD patients is unclear. In this paper, we summarized the findings of the studies of neuropsychiatric symptoms and neuropsychiatric syndromes/endophenotypes in AD in relation to APOE genotypes, with special attention to the possible underlying mechanisms. While some studies failed to find a significant association between APOE and neuropsychiatric symptoms in late-onset AD, other studies reported a significant association between the APOE ε4 allele and an increase in agitation/aggression, hallucinations, delusions, and late-life depression or anxiety. Furthermore, some negative studies that focused on the distribution of APOE genotypes between AD patients with or without neuropsychiatric symptoms further emphasized the importance of subgrouping neuropsychiatric symptoms in distinct neuropsychiatric syndromes. Explanations for the variable findings in the existing studies included differences in patient populations, differences in the assessment of neuropsychiatric symptomatology, and possible lack of statistical power to detect associations in the negative studies.


Introduction
Dementia and age-related cognitive disorders are reaching epidemic proportions, given the significant increase in the aging population. Alzheimer's disease (AD), the most prevalent form of dementia [1], is a complex brain disorder that has effects on multiple cerebral systems, and characterized by relatively slow but progressive neurodegeneration and impairment in cognition, behavior, and functionality. Accurate AD epidemiological data have been recently released for the USA. The 2010 figures suggested that 5.3 million Americans have AD [2], with >26 million patients with AD worldwide, and an expected increase to more than 106 million by 2050 [3]. Neuropsychiatric symptoms, previously denominated as behavioral and psychological symptoms of dementia, are common features of AD [4,5] and are one of the major risk factors for institutionalization [6] as well as for increasing costs both in USA and Western countries [7,8]. Neuropsychiatric symptoms in AD include psychosis (delusions and hallucinations) as well as affective and behavioral changes such as depressive mood, anxiety, irritability/lability, apathy, euphoria, disinhibition, agitation/aggression, aberrant motor activities, sleep disturbance, and eating disorder [9]. Neuropsychiatric symptoms associated with AD tend to follow a trajectory of increasing severity over time, a feature they have in common with cognitive and functional decline. Neuropsychiatric symptoms may be associated to AD irrespective of cognitive impairment severity and may be the presenting complaint or may emerge in the course of the disease being important cause of a more rapid cognitive decline [9]. It has been estimated that up to 80% of patients with AD showed neuropsychiatric symptoms in the history of the disease [9,10]. Recently, the Behavioural Subgroup of the European Alzheimer's Disease Consortium has performed a factor analysis of the neuropsychiatric inventory (NPI) [11] in a homogeneous sample of patients with AD, analyzing the largest AD population ever studied for this purpose [5]. The Behavioural Subgroup of the European Alzheimer's Disease Consortium identified 4 separate neuropsychiatric syndromes: hyperactive, psychotic, affective, and apathetic [5], providing also evidence of the relative consistency of neuropsychiatric syndromes across dementia subtypes, age and gender [12], and stressing the importance of thinking about neuropsychiatric syndromes instead of separate symptoms in AD patients.
The majority of AD cases are sporadic (i.e., without an apparent familial patterns of inheritance) compared with fewer than 5% of cases that are caused by autosomal dominant inheritance of mutations in presenilin 1, presenilin 2, or amyloid precursor protein [13]. Several genes have been identified as possible risk factors for the development of sporadic late-onset AD, particularly the gene dose of apolipoprotein E (APOE) ε4 alleles [13]. In fact, the APOE gene is the only globally valid genetic determinant of sporadic AD to have been unambiguously identified in 15 years of intensive research [14]. However, as with other multifactorial diseases, this systematic inability to detect new genetic determinants has prompted more comprehensive investigations using genome-wide association studies. Three large genome-wide association studies in this field were performed and reported that the clusterin (CLU), phosphatidylinositol binding clathrin assembly protein (PICALM), complement component (3b/4b) receptor 1 (CR1), bridging integrator 1 (BIN1), and exocyst complex component 3like 2 (EXOC3L2) loci were associated with AD [15][16][17], in addition to the established APOE relationship. The genetic origin of the three common variants of the human APOE, known as E2, E3, and E4, was understood in 1981 [18], and since the mid of the 80s these are probably the most studied protein variants in human races. A further study in 1982 proposed a nomenclature for these protein isoforms, to be identified as E2, E3, and E4 [19]. These important findings have led several researchers to the identification of the APOE encoding gene, that was recognized in 1982 [20], localized on chromosome 19 at locus q13.31 [21], and definitively sequenced [22]. Three years later, also the DNA encoding these isoforms was definitively sequenced [23] and named as ε2, ε3, and ε4. The important implications of this study are the demonstration that these common gene variants were generated by the two single-nucleotide polymorphisms rs429358 and rs7412 in exon 4 of the APOE gene (AF261279) and that these three allelic forms of the APOE gene are indeed different haplotypes of the APOE gene, generated by the combination of the allele of these two single-nucleotide polymorphisms at the APOE locus [13]. Therefore, risk for late-onset AD is known to be associated with polymorphisms of the APOE gene; people with an ε4 allele have an increased risk of both familial and sporadic forms, accounting for 20-50% of the attributable risk [13]. Nonetheless, APOE is neither necessary nor sufficient to cause AD, and this is the main reason why APOE is classified as a risk factor for AD and not a causative one. Although this is currently a nonmodifiable risk factor, it has potential for modifying the impact of other factors, in particular vascular and lifestyle-related factors [24], implying that some interventions may perhaps best be restricted to people at genetic risk. At present, the role of the APOE polymorphism in the development of neuropsychiatric symptoms [25][26][27] or neuropsychiatric syndromes/endophenotypes in AD patients is unclear [28][29][30]. It has been suggested that neuropsychiatric symptoms in AD may be influenced by the APOE polymorphism in various case-control studies [25,26,28,29], but a recent longitudinal report showed a substantial lack of association between the APOE ε4 allele and neuropsychiatric symptoms in AD after correction for multiple testing [27].
In this paper, we summarized the findings of the studies of neuropsychiatric symptoms and neuropsychiatric syndromes/endophenotypes in AD in relation to APOE genotypes from the English literature published before September 2010. We reviewed clinical and epidemiological studies from the international literature, including both cross-sectional and longitudinal studies that involved subjects aged 65 years and older. We searched through the PubMed database of NCBI (available at http://www.ncbi.nlm.nih.gov/) by author and the following keywords: neuropsychiatric symptoms, neuropsychiatric syndromes, neuropsychiatric endophenotypes, apolipoprotein E, APOE, dementia, Alzheimer's disease, late-life depression, late-life anxiety, affective syndromes, apathy, apathetic syndrome in dementia, agitation, aggressiveness, hyperactive syndrome in dementia, psychosis, psychotic syndrome in dementia, and behavioral and psychological symptoms of dementia. Special attention was paid to the possible mechanisms linked to the development of neuropsychiatric symptoms and neuropsychiatric syndromes/endophenotypes in AD in relation to APOE genotypes.

APOE Genotypes and Affective
Syndromes/Endophenotypes in AD:

Late-Life Depression and Anxiety
In adults older than 65 years, late-life depression refers to depressive syndromes encompassing both late-onset cases as well as early-onset cases that recur or continue into later years of life [31,32]. Late-life depressive syndromes often arise in the context of medical and neurological disorders [33]. The prevalence of major depressive disorder is about 17% in patients with AD [34] and is even higher in those with subcortical dementias [31,32]. Furthermore, emerging research implicates also a consistent reciprocal relationship between late life anxiety and cognition [35,36]. In fact, anxiety disorders are the most common psychiatric diagnoses in International Journal of Alzheimer's Disease 3 late life with an estimated lifetime prevalence of 15.3% in older adults, surpassing population estimates for depressive disorders [37], and severe cognitive impairment [38]. There is evidence of more prevalent anxiety in cognitively impaired older adults, elevated anxiety related to poorer cognitive performance, and more severe anxiety symptoms predicting future cognitive decline [35]. Early studies examining the relationship between APOE genotypes and depression showed that the APOE ε4 allele may be a risk factor for depressive symptoms in patients with AD and dementia [39][40][41], and other investigators have demonstrated combined risks of developing AD among those with late-life depression and APOE ε4 allele in nondemented geriatric populations [42][43][44][45] (Table 1). These findings were confirmed in two very recent longitudinal population-based studies [46,47]. A recent study suggested that depression and APOE ε4 genotype may be higher in women with AD but not in men [48]. Consistent with these previous studies demonstrating a relationship between APOE genotype and depression in AD [29][30][31][32][33][34][35][36][37][38][39][40][41]48], a recent report found that depressed AD patients had a significantly higher frequency of APOE ε4 allele than nondepressed AD patients [49]. Also in this study the association between APOE genotype and depression in AD was primarily seen in women and not men, so conforming the results of a study showing an increased rate of the APOE ε4 genotype in late-onset depressed women with AD relative to men [43]. Finally, in a very recent study, when AD patients were subdivided according to the European Alzheimer's Disease Consortium classification of neuropsychiatric syndromes, in APO ε4-carriers there was an increased risk of affective syndrome (i.e., AD patients with the NPI symptoms depression and anxiety) [30]. These results on increased risk of an affective syndrome in AD patients APOE ε4-carriers confirmed recent findings indicating that neuropsychiatric symptoms in AD are not purely an epiphenomenon of cognitive impairment but could be attributed to specific biological brain dysfunction, suggesting the presence of specific AD endophenotypes [28,29,[50][51][52].
Despite these findings supporting an association between APOE and depression, several other published studies have not supported the notion that APOE ε4 genotype influences depression in AD [25,58,[66][67][68][69]71], also when specific AD endophenotypes were considered [28,29] (Table 1). Other reports suggested that in AD the APOE ε4 allele may be protective against depression [80] or that APOE ε2-carriers may be at higher risk of depressive symptoms [53][54][55]. In addition, others have shown trends toward association between APOE genotype and depression but have rendered conclusions of no relationship [60,77]. One possible explanation for these contrasting findings may be the different sample size, with several studies including few depressed AD patients [60,67,69], thus resulting in a different power to detect group differences. Also varying schemes for diagnosing depression in AD may be a source of variability [25,68,71]. Another factor may be that the use of limited screening measures for determining depression could be leading to poorer diagnostic accuracy and thus different prevalence rates of depression in AD (range 11-50%) [25,49,60]. Thus, the bulk of published studies reporting a lack of association between APOE ε4 allele and depression in AD have suffered from small sample sizes, low reported prevalence rates for depression, very brief inventories of depressive symptoms, and/or low diagnostic rigor in the determination of the presence or absence of a Diagnostic and Statistical Manual of Mental Disorders-based diagnosis of depression.
Anxiety is most common among older subjects with mild cognitive impairment [81] and AD patients with a younger age at onset (under age 65) [82]. APOE e4 allele is a risk factor for developing AD at an earlier age [83] and might contribute to this effect [84]. Interestingly, APOE-null mice carrying the human ε4 transgene had higher anxiety ratings than those expressing either wild-type murine APOE or the human ε3 [73]. Very recently, these findings were also confirmed in female APOE TR mice expressing APOE under control of the mouse APOE promoter. APOE4 mice showed decreased activity levels and higher measures of anxiety than mice expressing APOE2 or APOE3 across all ages [85]. Notwithstanding the higher prevalence and symptom expression of anxiety disorders in late life, not all measures of neuropsychiatric symptoms in dementia reported in the literature include an assessment of anxiety [82]. Two studies using longitudinal assessments with the NPI found no association with anxiety [25,65], and no association was found when assessed by the Behavioural Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD) in a cross-sectional study design [69] (Table 1). Consistent with the mouse studies [73,85], APOE also has isoform-dependent effects on measures of anxiety in probable AD patients [73]. In particular, male ε4/ε4 subjects had higher anxiety scores than gender-matched ε3/ε3 subjects. In males, but not in females, subjects with ε4/ε4 had also higher anxiety scores than those with ε3/ε4 [73], suggesting that APOE3 can antagonize the effects of APOE4 on measures of anxiety in males but not in females. The anxiety scores did not correlate with the Mini-Mental State Examination scores [73]. Pritchard and colleagues did not support these findings using the same methods of analysis [27]. However, they found an association with the APOE ε3/ε4 genotype, but not with the APOE ε4/ε4 genotype or the e4 allele [27], so confirming a previous study in which anxiety appeared more frequently in APOE ε3/ε4 demented patients, although this difference was not statistically significant [56]. Finally, when AD patients were subdivided according to the European Alzheimer's Disease Consortium classification of neuropsychiatric syndromes, for the affective syndrome (i.e., AD patients with the NPI symptoms depression and anxiety), these previous findings in AD and demented patients [27,73,74] were confirmed suggesting that the presence of an APOE ε4 allele may be a risk factor also for anxiety symptoms [56].

APOE and Apathy in AD
Apathy has been increasingly recognized as a distinct psychiatric syndrome, and defined as a lack of motivation, evidenced by diminished goal-directed overt behavior, diminished goal-directed cognition, and diminished emotional  International Journal of Alzheimer's Disease  concomitants of goal-directed behavior [86,87]. To date, there is no clear consensus on the definition of apathy, and it is not included in the glossary of the Diagnostic and Statistical Manual of Mental Disorders-lV [88] and mentioned merely as a nonspecific symptom of several disorders. Anhedonia or loss of interest or pleasure can be used as a principal symptom to diagnose major depressive disorder instead of or along with depressed mood. Because other criteria for Diagnostic and Statistical Manual of Mental Disorders-IV diagnosis of major depressive disorder [88] such as fatigue, hypersomnia or insomnia, loss of appetite, weight loss, and diminished ability to concentrate are prevalent among demented patients, a demented patient with apathy may be misdiagnosed as having major depressive disorder even in the absence of dysphoria [89]. This diagnostic challenge stems from the apparent overlap between apathy and depression [90]. Apathy has been reported to be common in AD outpatients, and the reported prevalence for apathy using the NPI was between 32.1% and 93.2% [91,92]. Also the relationship between the APOE e4 allele and the apathetic syndrome and/or the apathy symptom was studied [28-30, 36-38, 70, 74, 75, 93], although it has not been as thoroughly investigated as for depression or psychosis. In a sample of 197 subjects with probable AD, the presence of apathy was significantly associated with the APOE e4 allele independently from age, education, sex, duration of disease, Mini-Mental State Examination score, and other neuropsychiatric symptoms [74]. Furthermore, the recent results on an increased risk of the European Alzheimer's Disease Consortium apathetic syndrome (i.e., NPI symptoms apathy and eating abnormalities) in AD patients APOE ε4carriers [30] confirmed these previous findings suggesting a relationship between the APOE ε4 allele and apathy in patients with AD [74], although other studies did not show any correlation between apathy endophenotypes [28,29] or NPI apathy symptom [75] and APOE genotypes (Table 1).

APOE and Psychotic Symptoms and Syndromes/Endophenotypes in AD
Among neuropsychiatric symptoms in AD, major attention has been dedicated to psychotic symptoms, that typically consist of delusions and/or hallucinations [94]. The prevalence of psychosis is quite substantial, with estimates for delusions in AD ranging from 9.3% to 63% (median 36%) and estimates for hallucinations ranging from 4% to 41% (median 18%) [95]. However, psychosis proved to be a coherent grouping of psychiatric symptoms in AD in studies using cluster and factor analysis [94,96]. Therefore, it has been reported that AD patients with psychosis (AD + psychosis, AD + P) may be clinically different from AD [93,94,97]. Indeed, an increased familial risk for the AD + P endophenotype has been observed [70], suggesting heritability [88,98] and thus a possible genetic component of this clinical feature. In fact, genes that have been implicated in psychosis of AD include those for dopamine-3 and serotonin-2A receptors [98]. Major attention has been dedicated to psychotic episodes in relation to APOE genotypes, and a number of investigations have examined whether the APOE ε4 allele might contribute to the AD + P endophenotype [98]. Among initial studies conducted in unrelated AD + P and AD without psychosis subjects, one large [66] and two smaller studies [39,80] have found an association of APOE ε4 allele and the AD + P endophenotype. However, most successive reports have found no association of the APOE e4 allele with AD + P [25,27,40,56,61,99,100] ( Table 1), suggesting that it may be necessary to operate a distinction among different psychotic symptoms in AD in relation to APOE polymorphism. However, also association studies with hallucinations have been contradictory, with two copies of ε4 being reported as protective [68], ε4 being associated with increased risk [72], and hallucinations appeared to increase with more ε4 alleles [101] or also, in a small longitudinal study, to decrease 1 year after diagnosis in AD patients with one ε4 allele [78]. The association of hallucinations in AD with the e3/ε4 genotype [27] also added further inconsistencies to reported findings. Hallucinations are sometimes incorporated into a measure of psychosis, but also these studies have again been inconsistent, with significant associations being reported for the ε4 allele [66] and for ε3/ε4 versus ε4/ε4 carriers [39]. Furthermore, when the outcome was the presence of delusions in AD patients, in a longitudinal study, one ε4 allele carried 2.5-fold risk whereas the presence of two ε4 alleles carried 5.6-fold risk for development of delusions during a 9.3-year followup [68]. These findings were confirmed in other two longitudinal studies in which the presence of the APOE ε4 allele carried a 3.4-fold risk for developing delusions [72] and the presence of one e4 allele, 1 year after diagnosis, showed a moderate, but significant, increase in delusions [78]. Also in crosssectional studies, APOE ε4 allele possession was associated with increased levels of delusions within the last month from the first visit and with the presence of categorical delusions at the early stage until the first visit [59], delusions appeared to increase with more ε4 alleles [101] and ε4/ε4 genotype [75], and ε4 was preferentially associated with delusions also in a large sample of AD patients from the USA [76]. However, when AD patients were subdivided according to the European Alzheimer's Disease Consortium classification of neuropsychiatric syndromes, for psychotic syndrome, including AD patients with delusions, hallucinations, and night-time disturbances, no association was found between APOE polymorphism and this neuropsychiatric syndrome [30]. These findings were consistent with those of a larger number of studies in which no association was found between APOE genotypes and the psychotic endophenotype [28,29] or also single psychotic symptoms or measures incorporating these symptoms, that is, hallucinations [26,57,[62][63][64][65]69] and delusions [62,65] (Table 1).

APOE, Agitation/Aggressiveness, and Hyperactive Syndrome/Endophenotype in AD
Agitation/aggressiveness differs from psychosis and depression of AD in that it may be conceptualized as a single symptom or a symptom complex [94]. The prevalence of agitation in dementia ranges from 20% to 60%, depending on diagnostic definitions used and the population studied [102]. A formalized definition of agitation was proposed as "inappropriate verbal, vocal, or motor activity that is not judged by an outside observer to be an obvious outcome of the needs or confusion of the individual" [103]. Agitation/ aggressiveness in dementia often co-occurs with psychosis and depression. There is substantial evidence that verbal agitation is associated with depression, and there may be some relationship to delusions [94,104]. Psychosis, particularly delusions, and depression occur with increased frequency in aggressive patients and may be a causative factor [105]. Also the symptom complex agitation/aggressiveness has been linked to APOE genotypes. Agitation and disorientation were more common in demented patients ε4/ε4 carriers, whereas anxiety and sleep disorders were more common in ε3/ε4 carriers, although these differences were not statistically significant [40,56]. In another report, it was noted that the presence of the ε4 allele was associated with increased combativeness, agitation, wandering, and confusion [41]. These findings were confirmed in two other recent studies in which the APOE ε4 allele has been associated with agitation/aggressiveness using the Neuropsychiatric Inventory Caregiver Distress Scale (NPI-D) [26,75], and in a recent report in which the presence of the APOE e4+ genotype increased the risk for agitated behaviour, including restlessness and vocalizations, in nursing home residents with dementia [79] (Table 1). However, when AD patients were subdivided according to the European Alzheimer's Disease Consortium classification of neuropsychiatric syndromes, for hyperactive syndrome, including AD patients with agitation, euphoria, disinhibition, irritability, and aberrant motor behaviour, no association was found between APOE genotypes and this neuropsychiatric syndrome, as also was found for psychotic syndrome [30]. These findings on the European Alzheimer's Disease Consortium hyperactive syndrome were similar to previous negative results in endophenotype [29], single symptom approach [25,65], and in a study evaluating risk for incident neuropsychiatric symptoms in a prospectively followed cohort [68] (Table 1).

Mechanisms Linking Neuropsychiatric Symptoms and Syndromes/Endophenotypes with APOE Genotypes in AD
The present reviewed evidence on a possible role of the APOE polymorphism on the development of neuropsychiatric symptoms in AD patients confirmed that genetic factors may account for some of the neuropsychiatric heterogeneity associated with AD [106,107]. The role of APOE in AD-related depression is still controversial, and additional research is needed [108]. However, recent evidence has been accumulating to suggest that APOE may be linked to vascular risk factors in late life and, in turn, may be associated with depression. Indeed, it has been posited that the ε4 allele may be a predisposing genetic marker for ischemic cerebrovascular disease [109] given its association with hyperlipidemia [110], atherosclerosis [111], myocardial infarction [112], and subcortical white matter lesion pathology [113]. Some latelife depressive syndromes might predispose, precipitate, or perpetuate by cerebrovascular disease, the so-called "vascular depression" hypothesis [114,115]. In fact, a history of stroke was associated with a 3-4-fold increased risk of apathy and depression in AD [116], and cerebrovascular disease has previously been linked to depression in several studies of individuals with dementia [117]. In addition, the APOE ε4 allele may increase the odds that older depressed patients develop mild cognitive impairment [118]. Finally, it has been proposed that, for those carrying a copy of the ε4 allele, the destructive effect of subtle, underlying vascular risk factors, may be enhanced [113]. Furthermore, the APOE ε4 allele appears to be less efficient than other isoforms in inducing cholesterol transport [119], which may have an important role in maintaining the integrity of membranes, and in synaptic plasticity. Thus, it appears that the APOE ε4 allele is associated with impaired response to cerebral damage and diminished capacity for neuronal repair [120,121] and that this poorer neuronal reparative capacity may be implicated in the development of cognitive decline and depression in older adults with the APOE ε4 genotype. Finally, for affective syndromes and/or symptoms in AD, also the mechanisms underlying the suggested association between late-life anxiety and APOE in AD are, at present, not completely understood [73,84,85]. Interestingly, higher measures of anxiety were reported in male mice lacking murine APOE (APOE −/− ) than in sex-matched wild-type controls [73,84]. There were isoform-dependent effects of APOE on measures of anxiety in male APOE −/− and APOE3 and APOE4 mice expressing human APOE in neurons under control of the neuron-specific enolase (NSE) promoter or in astrocytes under control of the glial fibrillary acidic protein (GFAP) promoter [73,84]. APOE −/− and APOE4 male mice showed increased measures of anxiety, whereas APOE3 mice behaved like wild-type controls [73,84]. Very recently, these findings were also confirmed in female APOE TR mice expressing APOE under control of the mouse APOE promoter [85]. APOE4 mice showed decreased activity levels and higher measures of anxiety than mice expressing APOE2 or APOE3 across all ages [85]. The association of cerebrovascular disease with apathy may reflect stroke-related damage to areas of the prefrontal cortex or related neural pathways, involved in the planning and execution of goal-directed behavior [122]. In fact, apathy is associated with frontal and subcortical pathology [123,124], and more severe apathy has been related with a severe impairment of frontal executive functions [125]. Functional imaging studies revealed that apathy in AD is related to dysfunction of the right temporoparietal and anterior cingulate cortices [122,123], regions involved in frontal-subcortical networks. A recent neuropathological study reported a significant relationship between chronic apathy and anterior cingulated cortex tangle pathology [124]. Apathetic patients with AD treated with cholinesterase inhibitors showed a significant reduction in apathy [126]. This is probably related to loss of nucleus basalis of Meynert cholinergic input to prefrontal and subcortical regions [122]. Interestingly, the APOE ε4 allele has been associated with a severe loss of cholinergic activity in the frontal cortex [127] and with enhanced β-amyloid deposition in the frontal region of nondemented presenile subjects [128].
Although the ε4 allele has been shown to promote the neuropathological features of AD, including β-amyloid deposition [129,130] and neurofibrillary tangle formation [129,130], attempts to relate these features to psychotic symptoms in AD have yielded conflicting results [131,132]. Therefore, neuropsychiatric symptomatology could be associated with more severe neuropathological changes, although there is no clear consensus on this. Furthermore, the ε4 allele has been associated with more profound deficits in cholinergic neurons, in particular in the frontal cortex [127] and medial temporal lobe [133], whereas the development of neuropsychiatric symptoms in AD appears to be related to specific neurotransmitter imbalances, notably acetylcholine [134]. Psychotic manifestations in AD have been associated with pathology in the temporal lobe and hippocampus [131,135]. One single photon emission computed tomography study has suggested that delusions in AD may be associated with hypoperfusion in the temporal lobes [136], and some [137] but not all [138] functional imaging studies have shown that AD patients who carry the ε4 allele have reduced temporal lobe function. The structural magnetic resonance imaging literature is more unified in showing greater medial temporal lobe atrophy in association with the ε4 allele in AD [139][140][141]. It is conceivable that the detected association between APOE genotype and psychotic symptom, particularly delusions, might reflect neuropathology more heavily concentrated in the temporal lobe. Finally, also the mechanisms of the possible association between APOE and aggressive/agitated behaviour in AD are unclear. APOE is associated with more rapid progression as measured by single photon emission computed tomography and higher tangle burden in the brain [130,138,142,143]. The accepted spread of neuropathological damage seen in AD, from the hippocampus to frontal-temporalparietal regions, may encourage the development of those behavioural symptoms that not only localize regionally within the brain but are dependent on progressive neuronal loss and amyloid deposition away from the mesial temporal lobe. Frontal involvement is the best neuroanatomical correlate for aggression and agitation with secondary disruption of the serotonergic and dopaminergic systems [144][145][146][147]. High agitation scores correlate with bilateral orbitofrontal and left anterior cingulated tangle burden [144] and with left fronto-temporal hypoperfusion on single photon emission computed tomography scanning [145].

Conclusions
Environmental factors that have been associated with lateonset AD include depressive syndromes, various vascular risk factors, level of education, head trauma, and dietary factors. This complexity may help explain their high prevalence from an evolutionary perspective, but the etiologic complexity makes identification of disease-related genes much more difficult [148]. The "endophenotype" approach is an alternative method for measuring phenotypic variation that may facilitate the identification of susceptibility genes for complexly inherited traits [148]. The association of the APOE genotypes with neuropsychiatric symptoms or neuropsychiatric syndromes and endophenotypes in AD appeared to be still unclear. Neuropsychiatric symptoms in different times could coexist in a single patient showing a very complex psychological profile. The discrepancy in dementia syndromes between the occurrence of neuropsychiatric symptoms from rather linear cognitive decline implies independent pathophysiological pathways between these symptoms. In particular, contrasting findings existed on the possible association between affective symptoms or syndromes in AD and APOE genotypes, with studies with measures of late-life depressive syndromes and symptoms more frequent than studies that focused on late-life anxiety. At present, there is only limited evidence of an increased risk of apathy or agitated behaviour linked to the presence of APOE ε4 allele in AD patients. Furthermore, some studies have found an association between APOE ε4 allele and the AD plus psychosis endophenotype, with several studies suggesting a possible role of this polymorphism when the outcome was the presence of delusions in AD patients rather than of hallucinations. The fact that some studies assessing neuropsychiatric symptoms in relation to APOE status often analyzed the dichotomized APOE ε4 status rather than the number of APOE e4 alleles may partially account for these contrasting findings, suggesting a methodological limit for the correct evaluation of predisposing factors and the pathogenetic basis of neuropsychiatric symptoms in AD. Discrepant findings may be due also to other factors including small sample sizes, differences in sample compositions (e.g., general "dementia" groups versus strictly diagnosed patients with "probable AD," with some of the studies including patients with Lewy body disease), the use of very brief, wideranging, or unstructured psychopathology inventories, and possible lack of statistical power to detect associations in the negative studies. Furthermore, some negative studies that focused on the distribution of APOE genotypes between AD patients with or without neuropsychiatric symptoms further emphasized the importance of subgrouping these symptoms in distinct neuropsychiatric syndromes, suggesting also genetic basis for individual neuropsychiatric symptoms. In addition, many of the above studies did not control for potential confounding variables. Most important, many reviewed studies were cross-sectional, whereas it would be of paramount importance to evaluate the risk for incident neuropsychiatric symptoms in relation to the APOE genotypes in prospectively followed cohorts of AD patients. In fact, crosssectional studies have limitations in interpreting the causal relationship between APOE and neuropsychiatric symptoms or neuropsychiatric syndromes/endophenotypes. Further longitudinal studies on larger sample of patients are needed to clarify the possible role of this gene in neuropsychiatric symptoms or neuropsychiatric syndromes/endophenotypes in AD.