Molecular Crossroads: Shared and Divergent Molecular Signatures in Alzheimer’s Disease and Dementia with Lewy Bodies

1. Introduction

On 3 November 1906, at a German psychiatrists meeting in Tübingen, Germany, Alois Alzheimer, a psychiatrist and neuropathologist gave a talk on “A peculiar severe disease process of the cerebral cortex”, describing a woman named Auguste Deter with symptoms of presenile dementia, and reported the post-mortem autopsy of her brain showing senile plaques and neurofibrillary tangles [1,2]. A few years later in 1910, Emil Kraepelin, a German psychiatrist, named the disease condition as “Alzheimer’s disease” [3,4]. Alzheimer’s disease (AD) is the most common progressive neurodegenerative disorder caused by neuronal cell death. AD currently affects around 50 million patients globally and this number is projected to reach approximately 152 million by 2050 [5,6]. Most AD cases are prevalent particularly among adults aged 65 years or older, with a life expectancy of about 3-10 years [7] . Depending on the stage, AD patients often exhibit a severe decline in cognitive and behavioral ability, initially with memory and spatial navigation problems, and other symptoms such as executive dysfunction (planning/organisation), impaired reasoning and judgement, attention and concentration problems, personality and behaviour changes, mood changes, social withdrawal, agitation and aggression, paranoia/delusions, and sleep disturbances [8,9,10,11] .

Early stage AD exhibits histological alterations and atrophy in the entorhinal cortex within the hippocampus of the medial temporal lobe, a brain region crucial for episodic memory formation and consolidation and spatial navigation [12,13,14]. Biochemists George Glenner and Caine Wong reported the accumulation of a 4-kDa peptide, which they termed “amyloid-β protein” (Aβ), the major constituent of extracellular amyloid plaques [15,16]. Similarly, Inge Grundke-Iqbal and Khalid Iqbal identified microtubule-associated protein tau (MAPT) labeling some neurofibrillary tangles and plaque neurites, in AD brain [17,18].

Dementia with Lewy bodies (DLB) is an age-associated neurodegenerative disorder that is less common than AD [19]. DLB is a member of group of neurodegenerative diseases referred to as synucleinopathies, which also include Parkinson’s disease (PD) and Multiple system atrophy (MSA). Synucleinopathies are a group of disorders characterized by pathological accumulation of the misfolded form of the small presynaptic protein alpha-synuclein (α-Syn), which is encoded by the gene SNCA. Major synucleinopathies are Lewy body disease (LBD) [20]. LBD refers to a group of disorders comprising Parkinson’s disease (PD), Parkinson’s disease dementia (PDD), and dementia with Lewy bodies (DLB), which shows the shared neuropathological hallmark of deposition of Lewy bodies within neurons and extensive axonal Lewy neurites [21,22]. However, among synucleinopathies, MSA is characterized by accumulation of inclusions known as glial cytoplasmic inclusions containing synuclein aggregates in oligodendrocytes, while Lewy body pathology is absent and, therefore, the LBD group does not include MSA [21,23,24,25].

DLB has an incidence of 0.5–1.6 per 1,000 person-years and affects an estimated 1.4 million people in the United States [26]. It accounts for ~7.5% of all dementia cases [27]. DLB typically begins in the late fifties and constantly increases with age [28,29], and it is more prevalent in males than females [30]. Historically, in 1912 Jakob Heinrich Lewy, who was studying Parkinson’s disease at Alois Alzheimer’s laboratory, observed eosinophilic intraneuronal inclusion bodies, which were named Lewy bodies by Nikolaevich Tretiakoff in 1919 [31]. DLB gained momentum in 1961, when Okazaki et al. reported that patients with dementia who died shortly thereafter with severe extrapyramidal rigidity [32]. Later, Japanese psychiatrist Kenji Kosaka and colleagues observed Lewy bodies in the brainstem and cerebral cortex in post-mortem autopsy samples from more than 20 patients with different cognitive impairment [33,34]. In 1995, the term “dementia with Lewy bodies” was first proposed at the First International Workshop (Newcastle upon Tyne, England), with diagnostic criteria focused on three core features: impairment in cognitive function, hallucinations, and Parkinsonian features [33,35,36]. In 1997, Spillantini and colleagues identified α-Syn as a major component of Lewy bodies in the brain tissue from patients with DLB and PD [37].

AD and DLB frequently converge clinically and pathologically, complicating diagnosis, prognostication, medication selection, and eligibility for clinical trials. In this review, we briefly compare clinical and neuropathological features of AD and DLB. We integrate evidence across fluid biomarkers, genetics, and imaging to delineate shared versus disease-specific biology. Finally, we translate the evidence from the findings into therapeutic implications and outline new future research directions with potential next steps for targeted treatments.

2. Clinical and Neuropathological Overlap Between AD and DLB

2.1. Clinical Features of AD

AD usually co-exists with ageing and presents with cognitive, behavioral, and psychological symptoms. In the early stage, individuals with AD show amnestic features such as memory loss, trouble learning new information, repeating questions, misplacing items, and difficulty with spatial navigation [11,38]. As the disease progresses, the moderate stage includes memory loss; disorientation of time and place; language difficulties; and impaired executive function such as poor reasoning and planning, with changes in mood (apathy, irritability, or depression), difficulty in solving complex daily tasks, and impaired judgement. In the late stages, AD patients show global cognitive failure, dependency for activities of daily living, gait impairment, dysphagia, and other complex neurological and psychiatric symptoms such as agitation, delusions or occasional hallucinations, and extrapyramidal symptoms such as rigidity or gait disturbance [11,39]. Initially, in 1984, the criteria were set by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS–ADRDA criteria for AD) as a suitable procedure for a clinical diagnosis of AD and have recently been updated [40,41,42]. The National Institute on Aging and Alzheimer’s Association (NIA–AA) updated the clinical approach guidelines, introduced the use of biomarkers, and proposed a new biological scheme: amyloid, tau, neurodegeneration (A/T(N)) for research [43]. In day-to-day practice, clinicians still diagnose the disease based on medical records, physical and neurological examination, neuropsychological evaluation, and neuroimaging [44].

2.2. Clinical Features of DLB

Individuals with DLB also show cognitive with characteristic non-amnestic features that often different from typical AD. The major symptoms of DLB are progressive cognitive or visuospatial impairment with fluctuating attention and alertness, parkinsonism (rigidity, slowness, tremor, shuffling gait, falls), recurrent visual hallucinations, REM sleep behaviour disorder (RBD), and antipsychotic sensitivity causing worse parkinsonism, confusion, and sleepiness [45,46]. In the later stages DLB patients commonly develop autonomic instability (orthostatic hypotension, urinary incontinence), repeated falls, and increased sensitivity to antipsychotics [30]. The 2017 DLB Consortium criteria diagnose probable DLB with two or more core clinical features, or one core feature plus an indicative biomarker: reduced striatal dopamine-transporter uptake (Positron Emission Tomography / Single-Photon Emission Computed Tomography, PET/SPECT), reduced myocardial iodine-123-metaiodobenzylguanidine (MIBG) uptake, or polysomnographic confirmation of rapid eye movement (REM) sleep without atonia [47,48,49]. Many symptoms of DLB overlap with PD, hence, the timing of the symptoms is diagnostically crucial. The “1-year rule”: dementia that begins before or within 1 year of the onset of parkinsonism helps separate DLB from PDD. DLB is favoured when dementia precedes or appears within 1 year of the onset of parkinsonism [35,48,50,51].

AD and DLB have some distinct early symptoms that help differentiate them. AD starts with prominent episodic memory loss and a steady decline, spatial navigation deficits, and features like hallucinations or parkinsonism usually appear late [44,52]. However, in DLB, early visual hallucinations, fluctuating cognitive features, RBD, and parkinsonism are hallmarks; autonomic failure and marked antipsychotic sensitivity increase the diagnostic confidence for DLB [53,54]. On bedside testing, DLB shows visuospatial and attentional deficits disproportionate to memory impairment, whereas AD is characterised by predominant amnestic deficits. Timing of motor signs is important: parkinsonism near or preceding cognitive symptoms supports DLB, while late parkinsonism in a prolonged amnestic course favors AD with superimposed extrapyramidal signs [55]. Recognition of these discriminators reduces misdiagnosis and avoids toxic antipsychotic exposure in DLB.

2.3. Neuropathological Features of AD

At the autopsy, AD brains often show moderate cerebral cortical atrophy primarily with narrowed gyri and widened sulci, early and significant atrophy of the medial temporal lobe (hippocampal/entorhinal atrophy), and symmetrical dilation of the lateral ventricles (hydrocephalus ex vacuo) compared with healthy age-matched controls [56,57,58]. Microscopically, AD brains show accumulation of extracellular Aβ (diffuse and neuritic) and intraneuronal neurofibrillary tangles composed of hyperphosphorylated tau [59,60,61]. In a landmark study by Braak H. and Braak E., tau pathology progresses in a predictable Braak sequence: transentorhinal (Stage I–Stage II) → limbic/hippocampal (Stage III–Stage IV) → widespread neocortical (Stage V–Stage VI), tracking clinical severity; Aβ deposition follows Thal phases from neocortex to deeper structures [62,63,64,65]. Thal phases (1–5) stage the spread of amyloid-β plaques in AD: starting in the neocortex (1), then hippocampal/allocortical regions (2), striatum and diencephalon (3), brainstem (4), and finally the cerebellum (5) [65]. Neuroinflammation such as activated microglia and astrocytosis, synaptic loss, and neuronal death accompany these lesions, and tangle burden correlates best with cognitive impairment [56]. Current NIA–AA neuropathologic guidelines report “AD neuropathologic change” on a continuum, integrating Braak neurofibrillary tangle stage, Thal Aβ phase, and neuritic-plaque scores [66].

2.4. Neuropathological Features of DLB

DLB brains also show cortical atrophy, but it is less severe than in AD [67]. Most prominently in early stages, for instance in the medial temporal lobe (hippocampus), atrophy is more profound in AD patients than in DLB. DLB patients show Lewy pathology, which is observed in the brainstem (mainly the substantia nigra), limbic, and neocortical regions [68]. DLB brains mainly exhibit depigmentation of the substantia nigra and locus coeruleus due to loss of the majority of pigmented neurons. In later stages, DLB individuals’ brains show moderate cortical atrophy [67,69,70,71]. Microscopically, DLB patients show two general types of Lewy bodies: classic brainstem Lewy bodies with a dense eosinophilic core and a pale halo (visible by haematoxylin and eosin stain), and cortical Lewy bodies, which are smaller, lack the halo, and are mainly observed by α-Syn immunohistochemistry. α-Syn is the primary component of Lewy bodies [22,72,73]. However, post-mortem studies of DLB brains suggest that microglial activation is not prominent in DLB compared to AD [74].

Comparatively, DLB post-mortem tissue shows widespread cortical and limbic α-Syn–containing inclusions, whereas AD shows medial temporal (hippocampal) atrophy with abundant neuritic Aβ plaques and high-stage tau tangles. Co-pathology is common, but lesion type, regional burden, and staging (Braak/Thal) usually distinguish the two [67,75,76,77]. An overview of shared versus disease-specific features is shown in Figure 1.

3. Shared and Divergent Molecular Pathways

3.1. From Etiology to Genetics: AD Versus DLB

The exact cause of AD remains a mystery to researchers and clinicians; however, ageing is the most predominant risk factor. Various genetic and epigenetic factors, together with environmental factors including lifestyle, diet, chronic stress, infections, hearing loss, hypertension, alcohol, air pollution, midlife cholesterol levels, vision loss, neuroendocrine changes, and other lifelong chemical exposures can disrupt immune homeostasis triggering neurodegenerative cascades [78,79,80,81].

Despite many genetic associations, pathogenic variants in APP (amyloid precursor protein), PSEN1 (presenilin-1), PSEN2 (presenilin-2), are rare and cause early-onset familial AD, and the ε4 and ε2 variants of APOE (apolipoprotein E) are AD susceptibility alleles, however, AD still lacks confirmed modifiable risk factors or preventive therapy [82]. Phenomic and genomic studies such as GWAS (genome-wide association studies), PheWAS (phenome-wide association studies), and Mendelian randomization analyses of the UK Biobank dataset reported 75 loci (42 novel) linked to pathways enriched for amyloid/tau, lipid metabolism, endocytosis, and immunity [83,84,85].

Multiple hypotheses have been proposed to describe complex etiology of AD. These include the amyloid cascade hypothesis, tau propagation (prion-like spread of tau), cholinergic deficit hypothesis, the mitochondrial cascade hypothesis, calcium homeostasis dysregulation, neurovascular dysfunction, chronic neuroinflammation, dysregulation of metal ion homeostasis, and impaired glymphatic clearance of protein in the brain [86,87,88,89]. Each model captures a different dimension of the pathology suggesting that AD likely results from a convergence of multiple factors rather than a single cause.

The exact cause of DLB also remains unclear. Multiple lines of evidence support a multifactorial cause which includes ageing, genetic vulnerability and environmental factors as in AD [90,91]. Reported associations include education and living environment, contact with chemical toxins and air pollution, smoking, alcohol use, diet and nutrition, infections, and stress [92]. These links are largely observational, correlational and inconsistent across studies, so causality has not yet been clearly established.

In 2021, Chia et al. identified five genome-wide significant risk loci: GBA, BIN1, TMEM175, SNCA and APOE, using whole-genome sequencing in large DLB cohorts and neurologically healthy controls. Polygenic risk analyses supported shared genetic profiles of DLB with AD and PD [93]. Functionally and etiologically, these findings place DLB at the intersection of PD and AD co-pathology [94]. GBA, SNCA and TMEM175 are well-established PD risk loci [95], whereas APOE and BIN1 are known AD risk loci [96]. Similarly, Guerreiro and colleagues further complemented these results with an unbiased GWAS of DLB that used a two-stage design and identified five loci in discovery: APOE, BCL7C/STX1B, SNCA, GBA and GABRB3. Replication confirmed APOE, SNCA and GBA at genome-wide significance, while GABRB3 did not retain significance in pathologically confirmed subsets. The study also provided evidence for a novel candidate locus, CNTN1 [97].

3.2. Pathogenic Frameworks: Aβ/Tau/Cholinergic Hypotheses in AD and α-Syn Pathology in DLB

John Hardy and David Allsop in 1992 put forth the well-known “Aβ Cascade Hypothesis” which posits that the accumulation of misfolded Aβ peptides and their insufficient clearance in the brain initiates AD [98]. Aβ deposition is thought to set off a toxic cascade leading to synaptic dysfunction, tau tangle formation, and downstream widespread neurodegeneration [98,99,100]. Under normal physiological conditions, first, APP is cleaved by the brain’s major β-secretase, β-site APP cleaving enzyme 1 (BACE1), releasing the large soluble sAPPβ fragment and leaving behind a 99-amino-acid, membrane-tethered C-terminal fragment (CTF99). Second, the multisubunit γ-secretase complex cleaves this CTF99 within the membrane, producing Aβ peptides of various lengths. The two most common Aβ species are Aβ₄₀ (40 amino acids) and Aβ₄₂ (42 amino acids). Aβ₄₀ is the more abundant form, but the slightly longer Aβ₄₂ is prone to misfolding and aggregation into toxic fibrils [88,100,101,102,103].

The cholinergic signaling deficit in AD suggests that early degeneration of basal forebrain cholinergic neurons and cortical acetylcholine deficit correlates with cognitive impairment, explaining symptomatic benefit of acetylcholinesterase inhibitors, however, this hypothesis fails to explain the major neurodegeneration in AD [104,105,106].

In healthy neurons, MAPT-encoded tau supports different cellular functioning and transport of axonal nutrients. Tau is mainly involved in promoting the assembly and stabilization of neuronal microtubules [107]. In AD, tau is misfolded, hyperphosphorylated and becomes prone to self-aggregation, forming insoluble straight filaments and paired helical filaments as neurofibrillary tangles [108,109]. These aggregated forms of tau disrupt neuronal function and contribute to neurodegeneration. Misfolded, hyperphosphorylated tau can seed and further spread in a prion-like manner [110,111]. Furthermore, numerous studies have demonstrated that tau pathology is also triggered by aggregated Aβ [112,113]. In a study by Bencze et al. on postmortem middle frontal gyrus and anterior hippocampus samples from AD patients examined the regional relationship between Lemur Tyrosine Kinase 2 (LMTK2) and phospho-tau. They found that LMTK2 expression was significantly reduced in neurofibrillary tangle–affected regions and showed a strong inverse correlation with phospho-tau levels, indicating that decreased LMTK2 is associated with tau pathology rather than a general feature of AD brains [68]. Additionally, TAR DNA-binding protein 43 (TDP-43), a major hallmark of amyotrophic lateral sclerosis, was detected in human post-mortem brains from patients with AD and DLB; phosphorylated TDP-43 was present in most AD cases and in a subset of DLB cases [114].

Intraneuronal accumulation of α-Syn inclusions is a prominent feature of DLB, with Lewy bodies in neuronal soma and Lewy neurites in processes [22,115]. Natively unfolded α-Syn misfolds into β-sheet–rich oligomers that build into protofibrils and finally insoluble fibrils, a stepwise process often accelerated at membranes. These pathogenic assemblies spread cell-to-cell when seeds are released from neighbouring or dying neurons (including via vesicles) and then taken up by neighboring cells, where they template further aggregation and toxicity [116]. Importantly, α-Syn can adopt distinct “strains,” each with its own biochemical features and seeding potential [117]. In DLB, strain properties favor robust neuronal seeding and Lewy-type pathology which explain the distinct pathological signature of cortical and subcortical involvement [118].

3.3. Mitochondrial Dysfunction and Oxidative Stress in AD Versus DLB

Mitochondrial dysfunction and oxidative stress play an essential part in the pathology of AD. Mitochondrial impairment, leading to faulty energy metabolism and enhanced oxidative damage, has been observed in AD brains. Antioxidant enzymes: catalase, glutathione peroxidase, and superoxide dismutase are reduced in AD pathogenesis, supporting the “mitochondrial cascade” hypothesis [88,119]. Aβ aggregates can trigger a vicious cycle of mitochondrial damage, which further promotes Aβ aggregation [120]. Tau pathology can both trigger and can also be driven by mitochondrial bioenergetic failure, impaired mitophagy, and redox imbalance that worsen synaptic health [121].

In DLB, intracellular deposition of the α-Syn inclusions results in mitochondrial dysfunction and oxidative stress [122,123,124]. Literature using Mendelian randomization to probe mitochondrial proteins in DLB has identified several candidates with putative causal links. Higher levels of AIF1 and the ES1 protein homologue were associated with greater DLB risk, whereas higher GLRX2, C1QBP and mitochondrial GC2 were linked to reduced risk [125]. Recent research Millot et al. has demonstrated that peripheral blood mononuclear cells from patients with AD and DLB show disrupted mitochondrial dynamics, with altered levels of key fusion (MFN2, OPA1) and fission (FIS1) proteins. These findings underscore that mitochondrial dysfunction, manifesting as an imbalance in the fusion–fission machinery, may be a common pathophysiological signature in DLB as well as in AD, however, further studies are needed to confirm these findings [126].

3.4. Neuroinflammation and Cerebrovascular/Endothelial Dysfunction in AD and DLB

Activation of various immune cells, including microglia and astrocytes in AD patients’ brains is observed as a robust neuroinflammatory response [127,128]. Aβ and pathological tau bind microglial receptors (TREM2 and CD33), transforming homeostatic glia into reactive states that cluster around plaques and tangles [129]. Activated microglia release pro-inflammatory cytokines, chemokines, and inflammasome products such as IL-1, IL-6, TNF-α, and IFN-γ, as well as reactive oxygen species, which exacerbate neurodegeneration [128,130,131]. Microglial cells tracks Aβ and p-tau; for example, release of cardiolipin from microglia promotes internalization of Aβ [132]. Other brain cells as well as peripheral cells like reactive astrocytes, oligodendrocytes, lymphocytes (adaptive immune system), and peripheral myeloid cells (neutrophils and monocytes) also contribute to pathogenesis and drive neuroinflammation in AD [130].

Cerebrovascular and endothelial dysfunction is another convergent pathological contributor in AD. Blood–brain barrier (BBB) breakdown and endothelial injury causing impaired Aβ clearance may occur at early stage of the diseases even before the appearance of cognitive symptoms [133,134,135]. BBB dysfunction leads to the chronic hypoperfusion and cerebral hypoxia impairing Aβ clearance further enhancing its accumulation and aggregation [136]. Deposition of Aβ in vessel walls (cerebral amyloid angiopathy) impairs vascular integrity and triggers local inflammation [134,137]. A dysfunctional BBB also permits peripheral immune cells to infiltrate the brain; infiltrating leukocytes interact with the neurovascular unit and intensify neuroinflammation and neuronal injury [138,139]. Additionally, vascular risk factors such as hypertension, diabetes, atherosclerosis favours AD pathology [140].

Evidence from numerous studies in DLB and PDD links neuroinflammation to region-specific changes in microglial state, including activation in the substantia nigra, putamen and several cortical regions in early DLB. Whether this response is overall protective or detrimental remains unclear [141].

Intriguingly, a post-mortem cohort study of DLB patients suggests that cortical microglial activation is not a dominant feature. Several inflammatory markers, such as Iba1, HLA-DR, CD68, CD64, CD32b, IL4R and CHI3L1, were largely unchanged, with lower CD32a, higher CD16, and no increase in neuropil degeneration [74]. Taken together, multiple studies suggest that neuroinflammation in DLB may be time- and region-dependent, whereas AD shows a pronounced, phagocytic microglial phenotype [142]. This evidence is also crucial in shaping how researchers design anti-inflammatory strategies across the two disorders.

Utilising the autopsy data from the Institute of Clinical Neurobiology, Vienna, Jellinger compared 96 DLB brains (subtyped as “pure” DLB and Lewy-body variant of AD (LBV)), 291 PD brains, and 390 age-matched controls, using routine histology and immunohistochemistry to grade cerebrovascular lesions and cerebral amyloid angiopathy (CAA). This research reported cerebrovascular lesion rates in DLB were similar to controls and PD, but severe lesions were far less frequent (~2% vs ~11% in PD and ~6% in controls) and no acute ischaemic or haemorrhagic strokes were present. CAA was common mainly in LBV (~78% vs ~28% in pure DLB). In DLB, cognitive impairment tracked neuritic AD co-pathology rather than vascular burden, suggesting a limited cerebrovascular endothelial contribution to DLB dementia [143].

3.5. Emerging Pathways from Multi-Omics of AD and DLB

Recent state of the art research such as large transcriptomic meta-analyses and single-cell RNA sequencing is revealing new molecular signatures of AD and DLB. Recently, Johnson et al. using TMT-MS to large post-mortem AD cohorts and built a proteomic co-expression network identified enrichment in MAPK signalling and metabolism showed strong correlations with AD neuropathology and longitudinal cognitive decline [144]. Using state-of-the-art single-nucleus RNA-seq of human AD prefrontal cortex, Mathys and colleagues identified disease-associated subpopulations and the myelination regulator LINGO1 consistently perturbed across neuronal and glial populations [145].

Similarly, mouse models revealed coordinated shifts across cortex, hippocampus, and plasma and highlighting altered glucose metabolism (elevated lactate and pyruvate, decreased plasma glucose-6-phosphate) and oxidative-stress signatures, including a significant cortical decrease in galactitol that may reflect its oxidation to an aldehyde and hydrogen peroxide [146]. Similarly, a complementary human AD Neuroimaging Initiative (ADNI) cohort, untargeted hydrophilic metabolomics and targeted lipidomics identified palmitoleamide, oleamide, diacylglycerols, and ether/plasmalogen lipids as significantly altered at baseline, with several of these (e.g., oleamide) associating with faster progression from mild cognitive impairment to AD [147].

Rajkumar et al. performed post-mortem RNA-seq of the anterior cingulate and dorsolateral prefrontal cortex identified and validated twelve genome-wide significant differentially expressed genes in DLB: RBM3, ALPI, OXTR, SELE, CSF3, GALNT6, SLC4A1, ABCA13, MPO, SST, RAB44 and CTSG [142]. Greally et al. performed proteomic profiling on pathologically confirmed DLB brains using sarkosyl-insoluble cortical tissue in two subgroups: Tau^positive and Tau^negative for tau pathology. They found that DLB with Tau^positive showed higher level of insoluble tau as well as enrichment in other pathways such as ubiquitin/p62 pathway, vesicle-mediated transport signatures, and Aβ. In contrast, DLB with Tau^negative showed increased in cytokine signaling and metabolic pathways and increased abundance of immunoproteasome components (e.g., PSMB8/10, PSME2) indicating molecularly distinct DLB subtypes [148].

Goralski and co-authors used GeoMx spatial transcriptomics on the human cingulate cortex in disease brains, as well as in a pre-formed fibril (PFF) mouse model, to profile NeuN⁺ neurons with versus without pSer-129–α-Syn (α-Syn phosphorylated at serine 129) positive inclusions. Inclusion⁺ neurons showed downregulation of synaptic, mitochondrial, proteasome, endolysosomal, and cytoskeletal genes, and upregulation of DNA-damage/repair and complement/cytokine pathways [149]. Using a high-throughput proteomic approach, label-free liquid chromatography–tandem mass spectrometry of cerebrospinal fluid (CSF) samples of DLB patients identified PCSK1N, NPTXR, VGF, PDYN, SCG2, and NPTX2. A three-marker panel (SCG2 + PDYN + VGF) distinguished DLB from non-DLB with high accuracy and high specificity, further supporting synaptic dysfunction as a core fluid biomarker signature [149].

In a targeted plasma metabolome profiling of patients with DLB and AD, Pan et al., 2024 reported shared reductions in serotonin/taurine and multiple glycerophospholipids/triglycerides, but DLB uniquely showed higher glutamine and lower hydroxylated to non-hydroxylated sphingomyelin ratios. A simple glutamine to lysophosphatidylcholine C24:0 (a common lipid with a 24-carbon chain) ratio distinguished DLB from AD with very high sensitivity and specificity demonstrating a potentially crucial biomarker in diagnostic practice and precision medicine differentiating these clinically overlapping dementias [150].

Comparative brain proteomics identified what AD and DLB share versus what’s distinct pathways. For instance, elevated levels of synaptic/mitochondrial proteins are associated with better cognition; however, increased extracellular matrix/complement and autophagy pathways are associated with greater cognitive impairment. More specifically, TRIM33 and SLC7A11 were significantly increased, and the autophagy protein p62/SQSTM1 was significantly reduced, in DLB compared with AD [151]. Multi-omics, single-cell RNA-seq, bulk proteomics and transcriptomics, single-nucleus transcriptomics, spatial transcriptomics and metabolomics, combined with machine learning and artificial intelligence to analyse complex datasets, are revealing dysregulated genes and pathways and cell-type programmes, and prioritising causal genes at GWAS risk loci, providing a foundation for integrative, data-driven biomarker discovery for AD and DLB [152,153,154].

4. Emerging Next-Generation Biomarkers

The current well-established tools for diagnosing AD are CSF biomarkers and neuroimaging. Imaging techniques include advanced PET, such as Aβ-PET to visualise Aβ deposition and tau-PET to assess tau pathology, together with ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET and structural magnetic resonance imaging (MRI). The latter two serve as indicators of neurodegeneration [155,156]. Major reliable CSF and advanced blood-based biomarkers for pathological changes in AD are Aβ42 (often expressed as the Aβ42/Aβ40 ratio), total tau, phosphorylated tau, neurofilament light chain (NfL), the microRNA subtypes miR-21-5p and miR-451a, and glial fibrillary acidic protein (GFAP) [157,158].

Beyond Aβ and tau, other non-traditional and utilising minimally invasive alternative biomarkers such as salivary (e.g., lactoferrin, acetylcholinesterase), urinary (AD7c neuronal thread protein, extracellular vesicle proteins), lipidomic shifts (ceramide/sphingomyelin), synaptic/axonal proteins (neurogranin), circulating microRNAs, and gut-/fecal-derived readouts that reflects to inflammation, synaptic injury, mitochondrial stress, and membrane lipid dysregulation enhance diagnostic accuracy [159].

Interestingly, detection of Spatial navigation ability measured with the Sea Hero Quest smartphone game is emerging as a sensitive, early “cognitive fingerprint” of AD risk. In large, real-world large demographic samples, navigation performance can flag preclinical risk and has been reported to outperform some standard memory tests, positioning such early digital biomarkers [38,160].

The well-known current clinical practices for diagnosing DLB are (i) ¹²³I-FP-CIT dopamine transporter SPECT (DAT-SPECT) for nigrostriatal uptake, (ii) polysomnography for REM sleep without atonia, and (iii) ¹²³I-MIBG cardiac scintigraphy showing reduced cardiac uptake. DAT SPECT offers high specificity; however, it has only moderate sensitivity in prodromal stages [161]. A new line of research has offered detection of α-Syn in biological samples as an additional avenue for biomarker development [162]. Development of α-Syn seed amplification assays (SAAs) such as RT-QuIC (Real-Time Quaking-Induced Conversion) and PMCA (Protein Misfolding Cyclic Amplification) show high sensitivity and specificity in CSF for PD and hold promise for stratifying synucleinopathies, including DLB, using seeded-aggregation kinetics [163]. No blood test is yet clinically validated for DLB, although plasma/exosomal α-Syn are under study and some non-traditional matrices including SAAs on peripheral tissues and body fluid olfactory mucosa and urine have shown poor agreement with CSF [164].

Sjaelland and colleagues, in a systematic review that screened 4,295 records and included 20 studies, reported that 17 different portable and wearable digital health technologies have been used to capture digital biomarkers of non-cognitive symptoms in DLB, but validation and feasibility reporting remain uncertain. Future digital biomarker–related studies with standardized evaluation are needed before adoption for clinical use [165].

A multiplex approach that integrates core clinical features with neuroimaging and fluid biomarkers is needed to improve the diagnostic accuracy of DLB [166,167]. Given the complex nature of AD and DLB pathobiology, no single marker can fully capture disease progression across all stages. Combined use of fluid biomarkers with imaging biomarkers and emerging new biomarkers can provide comprehensive biological insights. Using fluid and imaging measures together provides multilevel insight that better supports precision diagnosis, prognosis and treatment monitoring, thereby offering new opportunities for diagnosis, hence, effective disease management.

5. Conclusions and Therapeutic Future Directions

Increasing cases of AD and DLB impose a steep and rising societal and economic burden such as caregiver stress and health-system costs. Despite impressive progress in the field, disease-modifying options remain limited aside from recently approved anti-amyloid monoclonal antibodies for AD [168], so there is an urgent need for therapies that halt progression. AD and DLB frequently co-occur and share pathology, complicating diagnosis, prognosis, and medication choice. Hence, it is important to investigate the contribution of co-pathology in each patient. Determining whether a patient has pure AD, pure DLB, or a mixed phenotype is crucial for clinical trial selection and interpretation.

While established imaging and CSF tests are valuable, the current biomarker toolkit for distinguishing mixed versus pure disease is still incomplete. Larger, longitudinal, and standardized studies including α-Syn SAAs, A/T(N) panels, proteomic/metabolomic signatures, peripheral or blood-based signatures and rigorously validated digital biomarkers could help in clarifying these complex questions. Leveraging patient-derived induced pluripotent stem cells (iPSCs) and advanced brain organoid models that recapitulate α-Syn, tau, and Aβ pathologies will accelerate target discovery and mechanism-driven pharmacology. High-throughput CRISPR perturbation platforms such as unbiased screens [169] should be used to reveal druggable targets. Modern drug-design toolkits: computer-aided drug design, integrative drug designing approach [170], enhancing targeted protein degradation (e.g., PROTACs), molecular chaperone modulators, and peptide/binder design platforms such as BindCraft [171] can be deployed to design and identify selective modulators of α-Syn, tau, and upstream trafficking/clearance machinery. Future studies should be directed at investigating a combinatorial treatment approach, such as pairing an anti-amyloid agent with an anti-α-Syn therapy, with or without an anti-tau component and other additional strategies for synaptic repair, mitochondrial support, glial-immune modulation, and neurovascular protection.

In conclusion, deeper investigation of the molecular mechanisms, particularly the synergistic interplay of Aβ, α-Syn, and tau, is needed to uncover overlapping pathology and points of divergence. Clarifying these networks will open new avenues for earlier and more accurate diagnostics and for rational, mechanism-based therapeutic interventions.