Notably, clinically and pathologically, there is significant overlap between AD and PD, suggesting that they may have common underlying mechanisms. Both diseases contain similar pathological features, particularly protein aggregations: amyloid-beta plaques as well as neurofibrillary tangles in AD, as previously outlined, and Lewy bodies composed of alpha-synuclein protein in PD. Post-mortem analyses have revealed that these protein aggregations can co-occur in both diseases as well (Kalaitzakis et al., 2008; Burack et al., 2010; Irwin et al., 2012; Twohig and Nielsen, 2019; Visanji et al., 2019; Fernández-Calle et al., 2022). Approximately 50% of AD patients exhibit alpha-synuclein pathology, while up to 50% of PD patients with dementia have amyloid-beta proteins and neurofibrillary tangles that are sufficient enough for secondary AD diagnosis (Irwin et al., 2013; Lim et al., 2019). Moreover, both AD and PD also exhibit overlapping clinical symptoms, such as changes in cognitive function and extrapyramidal features (Morra and Donovick, 2014; Collins and Williams-Gray, 2016; Foguem and Manckoundia, 2018; Pierzchlińska et al., 2018; Wakasugi and Hanakawa, 2021). Considering these commonalities, the association between APOE4 with PD and PD-related cognitive impairment has been debated over the years, with discussion that APOE4 may potentially influence AD-type changes that contribute to the development of cognitive changes in PD subjects (Fernández-Calle et al., 2022; Haikal et al., 2024).

In the context of Parkinsonisms, the APOE4 allele is prominently found in individuals with PD-MCI, PDD, and DLB, when compared to cognitively unimpaired healthy controls (HCs) and/or PD patients with normal cognition (PD-NC) (Parsian et al., 2002; Lashley et al., 2008; Tsuang et al., 2013; Outeiro et al., 2019). In PD, APOE4 carriers are significantly more prevalent among those with PDD, with a 20–35% higher prevalence than in HCs and PD-NC subjects (Parsian et al., 2002; Williams-Gray et al., 2009; Tsuang et al., 2013). This association between PDD and APOE4 is further supported by multiple studies demonstrating that those with PD and at least one APOE4 allele have a greater likelihood in developing PDD when compared to those without the allele (Pankratz et al., 2006; Monsell et al., 2014; Huertas et al., 2017; Chung et al., 2021). Monsell et al. (2014) specifically reported a five-times higher risk for dementia within 5 years of diagnosis among PD APOE4 carriers (Monsell et al., 2014). Interestingly, PD patients with APOE4 also progress to PDD much earlier in comparison to non-carriers (Huertas et al., 2017; Tunold et al., 2021; Szwedo et al., 2022). Hence, APOE4 may be a significant risk factor for dementia in PD and serve as a potential clinical biomarker for PDD. To note, individuals with PD who carry APOE4 are diagnosed at a younger age compared to non-carriers as well, indicating that APOE4 may also be a significant predictor for early onset PD (Pankratz et al., 2006; Szwedo et al., 2022).

In terms of PD-related cognitive impairment, PD patients who are APOE4 carriers show greater cognitive impairment (Pavlova et al., 2014; Paul et al., 2016; Chung et al., 2021). Tropea et al. (2018) report a 3.6 fold increased risk of cognitive impairment in PD APOE4 carriers when compared to non-carriers (Tropea et al., 2018). Additionally, Szwedo et al. (2022) observed that PD APOE4 carriers demonstrate an accelerated rate of cognitive impairment over a 10-year-follow up (Szwedo et al., 2022). Thus, this indicates that APOE4 may play an essential role in facilitating an enhanced likelihood and faster progression of cognitive impairment in those with PD. Further, PD patients with APOE4 also exhibit a decline in specific cognitive processes. This includes executive and visuospatial function, impaired short-term memory, language, and attention, contrasting to patients without the polymorphism and/or HCs (Mata et al., 2014; Pavlova et al., 2014; Paul et al., 2016; Samat et al., 2017; Myers et al., 2022). Studies utilizing the Mattis Dementia Rating Scale-2 (DRS) for neuropsychological evaluation also demonstrate that PD APOE4 carriers have a greater decline in distinct cognitive abilities, including conceptualization, construction, initiation, and overall memory performance when compared to those without the allele (Morley et al., 2012; Tropea et al., 2018). Therefore, this may also suggest that APOE4 adversely impacts certain cognitive processes in individuals with PD.

Despite this evidence, some studies have failed to find a link between APOE4 and cognitive impairment in PD (Ezquerra et al., 2008; Kurz et al., 2009; Williams-Gray et al., 2009; Federoff et al., 2012). For instance, Federoff et al. (2012) found no relationship between memory loss at PD diagnosis and APOE4 amongst a large PD cohort (Federoff et al., 2012). Similarly, an observational cohort study by Mengel et al. (2016) found that APOE4 was not associated with the diagnosis of PD-MCI and PDD, nor with declines in specific neuropsychological domains in PD subjects (Mengel et al., 2016). Several studies also reported no association between APOE4 carrier status and PDD (Ezquerra et al., 2008; Kurz et al., 2009; Williams-Gray et al., 2009). Some findings have shown that possessing the polymorphism does not affect the risk, development, and progression to dementia, as well as the overall rate of cognitive impairment in PD instead (Ezquerra et al., 2008; Kurz et al., 2009; Williams-Gray et al., 2009). Discrepancies in the literature have been attributed to several factors. For example, small sample sizes resulting in low statistical power have been noted to weaken findings on the association between APOE4 and cognitive impairment in PD (Kurz et al., 2009; Federoff et al., 2012). In addition, results from cross-sectional studies have been discussed to potentially be misleading, as the effect of APOE4 may be time-dependent in PD by emerging later in disease progression like in AD (Collins and Williams-Gray, 2016). Studies that did not find an association between APOE4 and cognitive impairment in their cohorts have emphasized the need for longitudinal studies to further evaluate the effect of APOE4 and cognitive function in PD, which are currently limited (Ezquerra et al., 2008; Collins and Williams-Gray, 2016; Mengel et al., 2016; Zhu et al., 2024). It should be noted however that these studies are outdated, and over the past 5 years, an increasing number of research demonstrate that APOE4 in PD patients is indeed linked to a greater risk in cognitive impairment, a decline in specific cognitive processes, and serves as a predictor for PDD, with an earlier and more rapid onset (Rawle et al., 2018; D’Souza and Rajkumar, 2020; Tan et al., 2021; Tipton et al., 2021; Tunold et al., 2021; Myers et al., 2022; Okubadejo et al., 2022; Szwedo et al., 2022; Umeh et al., 2022; Chen et al., 2023; Liu et al., 2023; Real et al., 2023; Zenuni et al., 2023). Thus, understanding the role of APOE4 toward PD-related cognitive impairment may be worthwhile due to growing evidence showing its association with such non-motor symptom.

In the context of DLB, it is also rationalized that APOE4 may play a role in its pathogenesis, and a strong risk factor of the disease (Tsuang et al., 2013; Rongve et al., 2019; Kaivola et al., 2021). Genome-wide association studies (GWAS) examined APOE4 to be a significant loci linked to DLB (Bras et al., 2014; Guerreiro et al., 2018; Rongve et al., 2019; Chia et al., 2021; Kaivola et al., 2021; Bousiges et al., 2023). A GWAS study of 828 clinically diagnosed DLB subjects showed that APOE4 alongside the GBA gene was significantly associated with higher DLB risk (Rongve et al., 2019). Another GWAS study including 495 DLB cases showed that APOE4 was significantly associated in DLB subjects with AD-co-pathology when compared to those with no-co-pathology, potentially highlighting the relevance of APOE4 and its relationship to AD-like changes in DLB (Kaivola et al., 2021). Similar to PDD and AD, neuropathological studies show that DLB subjects also exhibit amyloid-beta and tau pathology co-occurring with alpha-synuclein pathology, where approximately 40–70% harbor amyloid-beta plaques across studies (Vrillon et al., 2024). DLB subjects also show decreases in Cerebrospinal (CSF) biofluid markers that measure amyloid-beta and tau proteins, indicating that amyloid and tau metabolism may be disrupted in DLB as well (van Steenoven et al., 2019; Bolsewig et al., 2024; Vrillon et al., 2024). Thus, this may suggest that APOE4 is also critical in DLB pathogenesis, especially in contributing to AD-like pathology to exacerbate cognitive impairment. Furthermore, more pronounced cognitive impairment is observed in DLB APOE4 carriers when compared to non-carriers, including in attention, learning, memory, and executive function (Ballard et al., 2001; Mirza et al., 2019; Nasri et al., 2022). Therefore, further reinforcing APOE4’s potential role in cognitive impairment in those with DLB.

Overall, in the current literature, mechanisms involving APOE4 that may be associated with cognitive impairment in PD and DLB is less understood. However, a possible mechanism may include enhancing the accumulation of alpha-synuclein aggregations in Lewy bodies (McKeith et al., 2005; Davis et al., 2020; Zhao et al., 2020; Attems et al., 2021; Haikal et al., 2024). This is supported with preclinical studies using alpha-synuclein mice models, which examined that those expressing the allele developed the most burden of alpha-synuclein pathology, as well as greater and faster cognitive impairment, including memory dysfunction, when compared to those with APOE2 and APOE3 (Davis et al., 2020; Zhao et al., 2020). It has been proposed that this relation between APOE4 and enhanced alpha-synuclein pathology may be due to various factors. APOE4’s role in the disruption of cholesterol homeostasis has been discussed, considering that one of the preclinical studies by Zhao et al. (2020) observed in their APOE4 cerebral organoids to have both increases in alpha-synuclein and lipid droplet accumulation (Zhao et al., 2020). Another alternative explanation may involve alpha-synuclein directly interacting and co-aggregating with APOE, as observed in post-mortem studies, suggested to be accelerated at the presence of APOE4 (Rohn and Mack, 2018; Paslawski et al., 2019; Davis et al., 2020; Zhao et al., 2020). However, further research is needed to characterize the cellular and molecular processes involving APOE4 that may exacerbate alpha-synuclein pathology (Davis et al., 2020). Other mechanisms may include APOE4’s role in enhancing amyloid-beta pathology, as established in AD, due to the significant prevalence in amyloid-beta in the brains of both DLB and PDD subjects. However, a study by Tsuang et al. (2013) examined a high proportion of individuals with APOE4 among those with PDD and pure DLB, a type of DLB characterized by the presence of Lewy bodies without AD-like pathology, but they exhibited low amyloid-beta plaques. This is consistent with other studies analyzing no association between APOE4 and pure DLB (Dickson et al., 2018; Kaivola et al., 2021). Therefore, APOE4 may also play a role in amyloid-independent mechanisms in PD and DLB to induce neurodegeneration and cognitive changes, just like in AD.

Abnormal accumulations of amyloid-beta proteins have been observed in the brains of subjects with PD-related cognitive impairment (Gomperts et al., 2013; O’Callaghan and Lewis, 2017). Several studies have reported higher levels of amyloid-beta proteins in individuals with PD-MCI and PDD when compared to HCs and PD-NC subjects (Irwin et al., 2012; Gomperts et al., 2013; Hepp et al., 2016). These increases in amyloid-beta in PD subjects with cognitive impairment is also observed in specific brain areas, including the temporal, parietal, and frontal regions (Maetzler et al., 2008; Hepp et al., 2016; Mihaescu et al., 2022).

Pre-clinical and clinical research have shown that APOE4 carriers have higher amyloid-beta protein accumulation, as observed in AD subjects and transgenic mice (Liu et al., 2013; Kanekiyo et al., 2014). Similarly, this association between APOE4 and enhanced amyloid-beta accumulation is also observed in PD. Various Positron Emission Tomography (PET) studies using radiotracers that measure these proteins in the brain, including [11C]-Pittsburgh Compound B (PiB), [18F]-Florbetaben (FBB), or [18F]-florbetapir ([18F]-AV-45), revealed significant increased retention and standardized uptake value ratios (SUVR) in PD APOE4 carriers than non-carriers (See Table 1 for more details) (Maetzler et al., 2009; Gomperts et al., 2012; Vijayaraghavan et al., 2014; Akhtar et al., 2017; Jung et al., 2021). Notably, these studies showed that higher retention and SUVR values for amyloid-beta proteins in specific brain regions was associated with APOE4 in PD subjects with mild cognitive impairment and/or dementia, highlighting APOE4’s potential role in influencing enhanced regional amyloid-beta pathology. For instance, Gomperts et al. (2012) discovered that the number of APOE4 alleles (0, 1, 2) was correlated with amyloid-beta protein within the precuneus (via [11C]-PiB PET) across their PD-MCI, PDD, DLB, and HC groups (Gomperts et al., 2012). This brain region has been associated with memory deficits when containing higher amyloid-beta protein depositions (Perrotin et al., 2012; Farrell et al., 2018; Fang et al., 2020). However, the authors did not find a significant correlation between precuneus amyloid-beta and cognitive function, as measured by the Mini-Mental State Examination (MMSE), in their PD-MCI (mean age = 69.4) and PDD groups (mean age = 71.7) (Gomperts et al., 2012). It was discussed that this may be due to the possibility that other pathological mechanisms could underlie the progression of PDD, rather than amyloid-beta burden (Gomperts et al., 2012). Such rationale may align with findings that amyloid-beta deposition is not well-correlated with cognitive impairment in older adults, and that other abnormalities known to be more strongly correlated with cognitive changes may be the reason for cognitive impairment instead, such as tau pathologies (Villemagne et al., 2013; Jack et al., 2014; Jansen et al., 2015; Bejanin et al., 2017; Brookmeyer and Abdalla, 2018; Ramanan et al., 2021). Low-levels of amyloid-beta burden in their PDD group was also discussed to potentially affect their ability to robustly detect an association between amyloid levels and cognitive function (Gomperts et al., 2012). Correspondingly, Maetzler et al. (2009) identified that all their subjects positive for amyloid-beta (via [11C]-PiB PET), measured in their frontal brain regions, posterior cingulate, as well as in superior parietal, lateral temporal and occipital lobe, had dementia (PDD and DLB) and a higher prevalence of APOE4 when compared to non-demented and demented subjects negative for [11C]-PiB (Maetzler et al., 2009). This may support APOE4’s association with dementia in PD and DLB, as previously discussed, which may be explained by enhanced amyloid-beta pathology in specific brain regions. Similarly, Akhtar et al. (2017) also found that higher amyloid-beta (via [18F]-AV-45 PET SUVR) in memory-related brain areas, including the anterior and posterior cingulate, precuneus, as well as temporal, frontal and parietal cortices, was associated with APOE4 in their PD-MCI and PD-NC groups when controlling for age (Akhtar et al., 2017). This may indicate that APOE4 enhances amyloid-beta burden in regions essential for cognitive domains affected in those with PD. Moreover, Vijayaraghavan et al. (2014) observed in their combined sample of APOE4 carriers (PD-NC, PD-MCI, PDD, and DLB) to have higher amyloid-beta (assessed via [11C]-PiB retention) in their striatal and cortical regions, as well as a lower MMSE score when compared to non-carriers (Vijayaraghavan et al., 2014). Jung et al. (2021) also found using PET with [18F]-FBB that typical LBD was linked to increased occipital beta-amyloid burden though its interaction with APOE4, as analyzed in their combined group including subjects with DLB with amyloid and pure LBD (PD and DLB). However, it should be noted that the heterogeneity of the samples in these two studies since they combined subjects with distinct Parkinsonisms and different stages of cognitive impairment may complicate interpretations. Combining PD subjects with varying cognitive levels may make it challenging to interpret the relationships between APOE4, cognition, and amyloid-beta accumulation in PD, since a previous systematic review of 11 PiB studies analyzed that each cognitive stage differs in amyloid-beta burden (Petrou et al., 2015). Including DLB subjects alongside PD subjects may further introduce complexity, given the distinct cognitive profiles and trajectories of DLB and PD (Aarsland et al., 2021). Overall, all these studies suggests that APOE4 may influence cognitive impairment in those with PD by contributing to region-specific amyloid-beta protein burden (Maetzler et al., 2008; Hepp et al., 2016; Mihaescu et al., 2022). Though the brain regions differ across studies, it is apparent that higher retention and SUVR values are found in areas known to play a role in cognitive processes impaired in PD subjects.

In contrast, several studies have reported no association between APOE4 and amyloid-beta protein burden instead. A study by Campbell et al. (2013) using principal component analysis provided evidence that APOE4 status did not affect amyloid-beta burden (assessed via mean [11C]-PiB binding) in 13 PD subjects experiencing cognitive impairment (Campbell et al., 2013). Mashima et al. (2017) also found that PD APOE4 carriers and non-carriers were both negative for amyloid-beta protein pathology using PET with [18F]-FBB (Mashima et al., 2017). However, the analysis of only 3 carriers and 1 non-carrier in a total of 33 subjects may lack statistical power, in contrast to larger studies showing a higher prevalence of amyloid-beta protein pathology in PD individuals with APOE4 as previously discussed (Maetzler et al., 2008; Gomperts et al., 2012; Vijayaraghavan et al., 2014; Akhtar et al., 2017; Mashima et al., 2017). Similarly, Gomperts et al. (2013) study claimed that APOE4 was not strongly associated with amyloid-beta burden (via [11C]-PiB) in their PD group, including those with PD-MCI, but proposed that such observations may be due to their sample sizes (n = 46). Observed discrepancies amongst PiB studies may be attributed to the types of groups analyzed and their levels of cognitive severity. Some studies identifying a significant association between APOE4 and amyloid-beta burden included cohorts with more severe cognitive impairment (i.e., classified to have DLB and PDD), whereas all the studies finding no association included participants with PD-MCI or PD without dementia only (Maetzler et al., 2009; Gomperts et al., 2012, 2013; Campbell et al., 2013; Vijayaraghavan et al., 2014; Mashima et al., 2017). This indicates that the severity of cognitive impairment may potentially play a critical role in the relationship between APOE4 status and amyloid-beta accumulation in PD. Discrepancies may also be attributed to the presence of co-morbidities of AD and PD protein pathology, as well as their co-localization with neuroinflammation, which may complicate the accurate measurement of amyloid-beta proteins. Individuals with advanced PD are more likely to exhibit overlapping protein pathologies, as both AD and PD pathologies commonly occur at later stages of the disease (Visanji et al., 2019). Similarly, enhanced neuroinflammation, measured through reactive astrocytes, has been shown to increase with age in DLB and PDD subjects, highlighting the prevalence of neuroinflammation in later stages of the disease also (van den Berge et al., 2012). Thus, since some significant studies included individuals with advance stages of the disease (PDD and DLB) while studies finding no significant association included none, it is also possible that prevalence of heightened co-pathology and neuroinflammation may potentially influence radiotracer uptake. This is further supported with evidence showing that PiB can bind to Lewy bodies and neurofibrillary tangles in vivo (Fodero-Tavoletti et al., 2007; Lockhart et al., 2007).

Furthermore, higher amyloid-beta protein pathology is potentially associated with greater impaired cognition in DLB subjects (Nelson et al., 2009; Gomperts et al., 2012). However, several studies reported that DLB APOE4 carriers had higher amyloid burden (via [11C]-PiB) than non-carriers, while others did not (See Table 1) (Rowe et al., 2007; Maetzler et al., 2009; Gomperts et al., 2012; Vijayaraghavan et al., 2014; Nedelska et al., 2019; Ferreira et al., 2020; Jung et al., 2021). Thus, further research is warranted to mitigate these inconsistencies to make robust conclusions on how APOE4 influences DLB-related cognitive impairment via increasing amyloid-beta proteins also.

Numerous studies, summarized in Table 2, have investigated the relationship between APOE4 and structural brain measures, including those that quantify gray and white matter, as well as ventricular size, in those with PD and DLB.

Measurements known to help assess gray matter changes have been explored to be associated with PD-related cognitive impairment. This includes cortical thickness, where its reduction has been examined in PD-MCI and PDD subjects using Magnetic Resonance Imaging (MRI) (Segura et al., 2014; Koshimori et al., 2015; O’Callaghan and Lewis, 2017; Gasca-Salas et al., 2019), and is proposed to be a potential marker for the conversion from PD-MCI to PDD (Segura et al., 2014; Chung et al., 2019). Two studies identified a relationship between APOE4 and decreases in cortical thickness of specific regions of PD patients, especially within posterior cortical areas, including in the temporal and parietal regions (Rane et al., 2018; Sampedro et al., 2019). Rane et al. (2018) showed that 28 APOE4 carriers (24 AD and 4 PD) who had amnestic MCI (aMCI) showed significant decreases in the cortical thickness of the middle temporal gyrus when compared to non-carriers, a brain region essential for language, visual perception, and episodic memory processing (Onitsuka et al., 2004; Rane et al., 2018). This is consistent with previous literature showing that the middle temporal gyrus is atrophied and thinner in those with PD-MCI when compared to PD subjects without cognitive impairments or HCs (Zhang et al., 2015; Hanganu and Monchi, 2016). Rane et al. (2018) also examined cortical thinning in the parietal and precuneus regions in their APOE4 carriers, areas previously found to be affected in PD individuals experiencing cognitive impairment and are also essential for memory function (Perrotin et al., 2012; O’Callaghan and Lewis, 2017; Farrell et al., 2018; Rane et al., 2018; Jawabri and Sharma, 2023). The results of Rane et al.’s (2018) study may be challenging to generalize however, as only a small sample size of PD subjects within their APOE4 group was observed (Rane et al., 2018). Furthermore, Sampedro et al. (2019) reported that an interaction between APOE4 and a specific genotype for the Brain Derived Neurotrophic Factor (BDNF) was associated with greater cortical thinning in posterior cortical regions in both their PD-NC and HC groups (Sampedro et al., 2019). These results may also suggest that a combination of genetic factors, including APOE4, may contribute to reduced cortical thickness of specific regions in PD that may potentially predict cognitive impairment. This is supported by evidence examining that posterior-cortical degeneration in PD is associated with more severe cognitive impairment in PD and is strongly associated with PDD (Pagonabarraga and Kulisevsky, 2012; Uribe et al., 2018). Collectively, we infer that both studies indicate that APOE4 influences reduced cortical thickness in regions essential for cognitive function, potentially contributing to future cognitive changes in those with PD. However, findings from Apostolova et al. (2012) may contradict this, as the thickness of the hippocampus, a region well-known for memory function, was not significantly affected by APOE4 status in PD patients instead, despite strong positive correlations with MMSE score and cognitive status (PD vs. PD-MCI) (Apostolova et al., 2012). Yet, methodological differences in thickness analysis, such as the use of radial distance measured by Apostolova et al. (2012) versus the FreeSurfer pipeline in the two studies, may have impacted sensitivity in examining APOE4-related thickness changes to contribute to these discrepancies (Apostolova et al., 2012; Rane et al., 2018; Sampedro et al., 2019).

Decreases in gray matter volume (GMV) have also been a widely examined gray matter change to be associated with PD-related cognitive impairment (Xia et al., 2013; Noh et al., 2014; O’Callaghan and Lewis, 2017; Hou and Shang, 2022). Reduced GMV was detected in PD patients with cognitive deficits, and associated with worsened performance in the Montreal Cognitive Assessment (MoCA) and MMSE (Xia et al., 2013; Noh et al., 2014; O’Callaghan and Lewis, 2017; Li et al., 2022). A meta-analysis by Xu et al. (2016) also found that individuals with PD-MCI and PDD had greater decreases in GMV when compared to PD patients without cognitive impairments, including in the hippocampus, insula, left superior and inferior temporal lobe, and left superior frontal lobe (Xu et al., 2016). There is limited data however with regards to the association of APOE4 and GMV alterations in PD. One study by Nicoletti et al. (2016) reported no significant differences in whole brain GMV when comparing PD patients with and without APOE4, regardless of dementia status (Nicoletti et al., 2016). They also found no significant differences in the shape of deep gray matter structures between PD APOE4 carriers and non-carriers, including the hippocampus, thalamus, globus pallidus, and caudate nucleus (Nicoletti et al., 2016). This may indicate that APOE4 may not influence GMV changes relevant in PD individuals with cognitive impairment. However, caution is necessary as these conclusions are derived from a single study. Since there are existing studies that do observe a relationship between APOE4 and GMV alterations in regions relevant for cognitive function, it may be worthwhile to conduct more research in PD to further evaluate APOE4’s association with GMV (Spampinato et al., 2011; Haller et al., 2017; Cacciaglia et al., 2018). In terms of DLB, there is conflicting evidence with regards to the association of APOE4 and hippocampal GMV, a brain region moderately atrophied in those with the disease (Chow et al., 2012). Saeed et al. (2018) studied in their combined group of APOE4 carriers with DLB and AD, as well as after stratifying the two groups that APOE4 was linked to hippocampal volume loss, with two alleles causing greater atrophy than one or none, and that this reduction in volume was associated with impaired learning performance (Saeed et al., 2018). However, Barber et al. (2000) found no differences in volumetric measurements, including the hippocampus, of DLB patients when comparing those with and without APOE4 (Barber et al., 2000). Variations in analyzing volume measurements may influence the observed differences between studies, particularly given that each employed different imaging software (Barber et al., 2000; Saeed et al., 2018). Saeed et al. (2018) also offered a thorough characterization of their sample, detailing the characteristics of both APOE4 carriers and non-carriers, which may further strengthen their findings (Saeed et al., 2018). Contrastingly, Barber et al. (2000) lacked demographic information for their APOE4 carriers and non-carriers, such as the number of subjects in each group, which may limit the interpretability of their results and lead to weaker conclusions (Barber et al., 2000).

White matter changes are also found in subjects with PD-related cognitive impairment and DLB (Joki et al., 2018; Yang et al., 2023). Commonly seen are White Matter Hyperintensities (WMH), which are white matter lesions analyzed as bright regions in T2-weighted, fluid attenuated inversion recovery (FLAIR) and proton density-weighted MRI images, while they appear darker in T1-weighted MRI scans (Wardlaw et al., 2015; Yang et al., 2023). A meta-analysis identified that enhanced WMH volumes are more prevalent in PDD subjects when compared to those with PD-MCI and PD-NC, indicating it may be a structural marker for PDD (Liu et al., 2021). Previous studies also observed that more WMH volumes in PD and DLB patients predict greater cognitive impairments, especially regarding memory function and processing speed (Dadar et al., 2018; Linortner et al., 2020; Ferreira et al., 2021). APOE4 has also been associated with increased WMH volumes in PD and DLB subjects in some studies, suggesting it may also contribute to white matter abnormalities known to be linked to cognitive impairment in these conditions (Kandel et al., 2016; Saeed et al., 2018). For example, one study found that PD APOE4 carriers showed significantly higher WMH volumes when compared to AD APOE4 carriers (Kandel et al., 2016). This may indicate that PD APOE4 carriers show pronounced white matter changes, which could potentially contribute to an increase risk of cognitive impairment. However, interpretations must be made with caution, as it is based from a single study. For DLB, Mirza et al. (2019) found their pooled cohort of AD and DLB subjects with APOE4 to have higher levels of WMH volumes that were linked to worsened learning, language processing, memory, and executive functioning (Mirza et al., 2019). Yet, due to sample imbalances in the cohort (239 AD vs. 50 DLB subjects), interpreting if these observed relationships occurs solely in DLB subjects may be challenging (Mirza et al., 2019). In addition, other studies also did not find an association between APOE4 and WMH volumes in DLB instead, further complicating conclusions. Ferreira et al. (2023) observed higher WMH volumes in DLB patients without APOE4, indicating a lack of association between APOE4 and increased WMH volumes in DLB subjects (Ferreira et al., 2023). However, since the study focused on those with mild to moderate DLB, the influence of APOE4 on enhanced WMH volumes could be more apparent in individuals with severe DLB who were ineligible for their investigations (Ferreira et al., 2023). Barber et al. (1999) also found no association between the presence of APOE4 and white matter lesions including WMH in their APOE4 cohort that had DLB subjects (Barber et al., 1999). But, the allele frequency of APOE4 in their DLB cohort was small, specifically 0.30 amongst 22 patients, and should be considered (Barber et al., 1999). Further research is needed to robustly understand the effect of APOE4 on WMH that may be associated with cognitive impairment in PD and DLB.

Lastly, there is some evidence that changes in ventricular size may be associated with PD-related cognitive impairment. PD-MCI patients showed significantly larger ventricles than those with PD-NC, and that this brain change is associated with cognitive changes including a decline in memory performance (Dalaker et al., 2011). However, two studies found no significant association between enlarged ventricles in their pooled sample of APOE4 carriers that included those with PD-NC and PD-MCI (Apostolova et al., 2012; Beyer et al., 2013). Therefore, we theorize that the effect of APOE4 toward PD-related cognitive impairment may not involve ventricle enlargement.

The relationship between APOE4 and functional brain measures have been examined in the brains of PD and DLB subjects with APOE4 in various studies, as detailed in Table 3. However, most reviewed studies involved PD subjects, where there are limited studies in DLB subjects exploring these relationships.

Functional abnormalities as measured using functional MRI (fMRI), a non-invasive technique that quantifies neural activity by assessing blood oxygen levels, have been found to be relevant in subjects with PD-related cognitive impairment (Sasikumar and Strafella, 2020; Yen et al., 2023). This includes changes in functional activity in specific brain regions, such as reductions in the prefrontal cortex, caudate nucleus, and fronto-striatal regions when doing cognitive tasks (Sasikumar and Strafella, 2020). It also includes altered functional connectivity as measured using resting fMRI, which quantifies the communication between brain regions through distinct patterns of brain activity amongst them (van den Heuvel and Hulshoff Pol, 2010; Baggio and Junqué, 2019). For example, van Eimeren et al. (2009) determined disruptions in the functional connectivity of the Default Mode Network when investigating PD patients, a network crucial for executive function and greatly defective in those experiencing PD-related cognitive impairment (van Eimeren et al., 2009). Limited fMRI evidence is available on the relation between APOE4 and functional activity as well as connectivity in the brains of those with PD. Of the studies reviewed, reductions in functional activity and connectivity were overall found within specific subcortical regions, including the caudate and hippocampus, as well as posterior cortical regions, which was then associated with cognitive changes. In particular, Nombela et al. (2014) reported significant reductions of activity during a memory encoding task in 10 subjects with early PD and APOE4 when compared to APOE2 and APOE3 carriers (Nombela et al., 2014). These reductions were observed in their right and left hippocampus, and posterior cortical areas, specifically the right inferior frontal gyri pars triangularis, left inferior frontal gyrus, left occipital, and temporo-parieto-occipital areas (Nombela et al., 2014). However, Nombela et al. (2014) acknowledged that the sample size for their APOE4 carriers is small, which is inferred to potentially lead to biased findings (Nombela et al., 2014). To correspond, Shang et al. (2020) examined in 28 Chinese individuals with PD-MCI and APOE4 that significantly lower functional connectivity values between the bilateral caudate nucleus and posterior cortical regions, including the right superior occipital and middle temporal gyrus, was significantly associated with reduced semantic fluency scores when compared to PD-MCI APOE2 and APOE3 carriers (Shang et al., 2020). Notably, the cognitive domains analyzed in both studies have been previous analyzed to be associated with those with PD-related cognitive impairment. Memory encoding has been evaluated to be adversely impacted in PD-MCI patients when compared to those with normal cognition (Weintraub et al., 2004; Gratwicke et al., 2015; Devignes et al., 2021) Worsened semantic fluency predicts progression to PDD, and is more disrupted in those with PD-MCI than those with PD-NC (Bohnen et al., 2011; Ye et al., 2017; Yang et al., 2022). Thus, both studies suggest that APOE4 may play a role in PD-related cognitive impairment by affecting the activity of and communication between specific brain regions essential for cognitive processes deficient in PD, especially posterior cortical areas.

Changes in glucose metabolism as measured by [18F]-Fluodeoxyglucose ([18F]-FDG) is also prevalent in PD subjects with cognitive impairment. Both PDD and PD-MCI subjects exhibit distinct brain patterns of glucose hypometabolism associated with changes in memory, attention, executive function, and visuospatial processing (Garcia-Garcia et al., 2012; Meyer et al., 2017). However, only one study exhibited the relationship between APOE4, cognitive status, and altered glucose metabolism in those with PD and DLB (Vijayaraghavan et al., 2014). Vijayaraghavan et al. (2014) observed a trend of reduced glucose metabolism in the posterior cingulate and lateral temporal cortex in their combined cohort of APOE4 carriers (PD-NC, PDD, and DLB subjects) when compared to non-carriers (Vijayaraghavan et al., 2014). Interestingly, CSF APOE protein levels were also found to be significantly higher in their DLB and PDD APOE4 carriers compared to non-carriers, and that these levels were negatively correlated with glucose metabolism in the lateral temporal cortex and posterior cingulate, as well as with cognitive scores on the Boston Naming Test and MMSE. These results are consistent with previous literature that examined the prevalence of reduced glucose metabolism in these regions in PD patients who converted to dementia, and in PD-MCI and PDD patients when compared to cognitively unimpaired individuals (Vander Borght et al., 1997; Bohnen et al., 2011). In addition, the posterior cingulate is important for attention, and the lateral temporal cortex is essential for language comprehension, facial recognition, and visual processing, all of which are adversely affected in PD individuals with cognitive impairment (Vander Borght et al., 1997; Watson and Leverenz, 2010; Goldstein et al., 2017; Vogt, 2019; Fang et al., 2020; Vijayaraghavan et al., 2014). Thus, we infer that reduced glucose metabolism in specific brain areas important for cognitive function in PD and DLB is linked to APOE4, leading to adverse cognitive changes. However, such conclusions are based on limited evidence, and further research is warranted to explore these associations more comprehensively.

Lastly, reduced dopamine neurotransmission has been a key characteristic of PD, occurring from the degeneration of dopaminergic neurons in the substantia nigra and midbrain (Triarhou, 2013). Though its effects are associated with its motor features, some studies report that it may also be linked to cognitive impairment in PD patients (Triarhou, 2013; Fang et al., 2020). Dopamine-Transporter (DAT)-Single Photon Emission Tomography (SPECT) imaging measures dopamine neurotransmission by visualizing the quantity of DAT, a protein found in the endings of presynaptic dopaminergic neurons that facilitate dopamine reuptake from synapses (Akdemir et al., 2021). Specific to APOE4 and PD, two studies found no significant differences in DAT binding between patients with and without polymorphism in the caudate nucleus and putamen using DaT-SPECT (Huertas et al., 2017; Kim et al., 2021b). Similarly, Kim et al. (2021a) observed that individuals with APOE4 experience greater cognitive impairment than those without it. Yet, this relationship was not linked with significant reductions in DAT uptake in the caudate and putamen as well between baseline and follow-up (Kim et al., 2021a). Thus, we infer that APOE4’s impact toward PD-related cognitive impairment may not include dopaminergic processes.