Population evidence generally points toward lower dementia incidence after cancer and lower cancer incidence after dementia, although results differ by cancer type, dementia subtype, population and analytical design. In a cohort of 3,021,508 adults aged 60 or older, a previous cancer diagnosis was associated with a 25% lower subsequent dementia risk (HR 0.75, 95% CI 0.74–0.76). The association remained after adjustment for competing mortality (SHR 0.56, 95% CI 0.55–0.56), while Mendelian-randomization analysis yielded a smaller inverse association between genetically predicted cancer susceptibility and Alzheimer’s disease (OR 0.97, 95% CI 0.95–0.99). A bidirectional meta-analysis found 8% to 14% lower risks among cancer survivors for all-cause dementia, Alzheimer’s disease and vascular dementia, while dementia was associated with approximately 25% lower subsequent cancer risk.

These findings are sensitive to ascertainment and analysis. In UK Biobank, neither prevalent nor incident cancer was significantly associated with Alzheimer’s disease. All-cause dementia diagnoses rose during the first year after an incident cancer diagnosis (HR 1.83, 95% CI 1.42–2.36) before attenuating toward the null, a pattern compatible with detection bias caused by increased healthcare contact, treatment-related cognitive symptoms or diagnostic investigation. Rotterdam Study dementia estimates ranged from RR 0.71 to 1.10, depending on how cancer exposure, confounding, selective survival and death were addressed. Reduced screening after dementia, delayed diagnosis and long preclinical phases further complicate causal interpretation.

Evidence beyond diagnostic records nevertheless suggests a possible biological component. In a community autopsy cohort of 785 people, including 190 with a history of cancer, cancer was associated with lower odds of advanced Braak neurofibrillary-tangle stages (OR 0.38, 95% CI 0.26–0.55), neuritic plaques (OR 0.52, 95% CI 0.36–0.75) and cerebral amyloid angiopathy, but not with TDP-43, α-synuclein or cerebrovascular pathology. Genetic-correlation and transcriptome-wide association analyses also identified opposing associations across Alzheimer’s disease and several cancers involving cell-cycle regulation, apoptosis, DNA-damage responses and metabolism. The pattern is not uniform across neurodegenerative disorders, including Parkinson’s disease and melanoma, so it does not support a universal cancer–neurodegeneration antagonism.

Breast and prostate cancers illustrate why tumor effects must be separated from endocrine treatment. Estrogen- and androgen-receptor signaling supports tumor growth while also regulating neuronal metabolism, mitochondrial function, synaptic activity, vascular homeostasis and inflammation. Among 70,701 Korean breast cancer survivors, Alzheimer’s disease risk was 8% lower than among 180,265 controls (SHR 0.92, 95% CI 0.86–0.98), but the association did not persist in landmark analyses restricted to more than five years of survival. A Swedish study of 26,741 five-year survivors found no overall association with dementia, Alzheimer’s disease or vascular dementia, although diagnosis after age 65 was associated with a 35% higher Alzheimer’s disease risk (SHR 1.35, 95% CI 1.05–1.75).

Treatment findings are also heterogeneous. In a US claims cohort of 57,843 women with breast cancer, hormone-modulating therapy was associated with lower Alzheimer’s disease incidence (RR 0.82, 95% CI 0.75–0.90) and dementia incidence (RR 0.88, 95% CI 0.83–0.93). A SEER–Medicare analysis of 18,808 women aged 65 or older found a smaller reduction in Alzheimer’s disease and related dementias (HR 0.93, 95% CI 0.88–0.98), varying by age, race and treatment class. For prostate cancer, a meta-analysis of 28 studies with approximately 2.54 million participants associated androgen-deprivation therapy with 20% higher all-cause dementia risk (HR 1.20, 95% CI 1.11–1.29) and 26% higher Alzheimer’s disease risk (HR 1.26, 95% CI 1.10–1.43). Individual cohorts have not consistently reproduced these associations, so the pooled findings do not establish causality.

The review describes overlapping insulin/IGF–PI3K–AKT, mTOR, AMPK, mitochondrial, cell-cycle, chaperone and autophagy–lysosomal pathways with context-dependent consequences. Tumors use them for metabolic adaptation, proliferation, apoptosis resistance and proteotoxic-stress tolerance, whereas Alzheimer’s disease involves energetic failure, impaired proteostasis and neuronal degeneration. Evidence comes mainly from human tissue, cancer-cell systems, organoids and animal models rather than direct studies of people with both diseases, making these biologically plausible contributors rather than demonstrated explanations of inverse comorbidity.

Protein self-assembly provides another possible bridge. In Alzheimer’s-model mice, lung, prostate and colon tumors secreted cystatin C, which interacted with amyloid-beta and activated TREM2-dependent microglial clearance. Disrupting TREM2 or mutating cystatin C eliminated the effects, establishing causality in that experimental system. Other preclinical studies found that amyloid-beta, human islet amyloid polypeptide and calcitonin could alter cancer-cell fitness, but uncertain physiological concentrations, aggregation states and selectivity over nonmalignant cells limit interpretation. Human plasma findings do not support a straightforward systemic anticancer effect, and melanoma-secreted amyloid-beta has promoted brain metastasis experimentally.

The bounded conclusion is that inverse comorbidity may reflect interacting epidemiological, endocrine, metabolic, proteostatic and protein-assembly mechanisms rather than one pathway. The authors recommend treatment-aware longitudinal studies that distinguish cancer subtype and stage from treatment class, dose and duration; incorporate amyloid and tau biomarkers, neuroimaging, endocrine markers and metabolic profiles; and test molecular interactions at physiologically relevant concentrations and aggregation states in patient-derived organoids, immune-competent cocultures and animal models. Whether the tumor-derived cystatin C–TREM2 pathway or other preclinical mechanisms contribute materially to inverse comorbidity in humans remains unknown.

Subject of Research: Epidemiological and experimental evidence on inverse comorbidity between cancer and Alzheimer’s disease

Article Title: Cancer–Alzheimer inverse comorbidity from epidemiology to protein self-assembly

Article References: Zhang, Y., Wang, L., Wei, J., Shuzan, N., & Zheng, J. (2026). Cancer–Alzheimer inverse comorbidity from epidemiology to protein self-assembly. Frontiers in Endocrinology, 17, Article 1992522. https://doi.org/10.3389/fendo.2026.1992522

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1992522

Keywords: cancer, Alzheimer’s disease, inverse comorbidity, epidemiology, endocrine therapy, amyloid-beta, protein self-assembly, proteostasis, dementia risk, Alzheimer, inverse, comorbidity

The central idea is that spontaneous speech carries Alzheimer’s-related information in both acoustic and linguistic channels. Acoustic features can include pauses, pitch, rhythm and other properties of delivery, while transcripts reveal vocabulary, syntax and semantic organization. Earlier multimodal systems often trained separate models and combined their predictions through majority voting or simple early- or late-stage fusion. Other systems converted audio into log-Mel spectrogram images so that pretrained vision models such as ViT or DeiT could analyze it. The new approach processes audio directly with Whisper-small, a speech-recognition model, and text directly with bert-base-uncased, avoiding an intermediate image conversion that the researchers argue can discard fine-grained acoustic information.

Training the two encoders as part of one architecture creates a technical obstacle because speech and transcripts do not naturally arrive in matching forms. Whisper produces a variable-length sequence of frame-level acoustic embeddings, with longer recordings divided into 30-second segments, while BERT returns token-level features padded or truncated to a maximum of 512 tokens. The authors solve this mismatch with Optimal Transport Kernel Embedding, or OTKE, which maps the variable-length Whisper sequence onto a compact, fixed-size representation. The method uses a kernel approximation in a Reproducing Kernel Hilbert Space, a trainable reference set and an optimal-transport plan computed with Sinkhorn’s algorithm. This step allows audio features to retain a structured temporal representation while being aligned with the text sequence for joint processing.

After alignment, the model applies two cross-attention layers. In the text-to-audio branch, the audio sequence supplies queries and the text sequence supplies keys and values, enabling each acoustic position to incorporate information from the transcript. The reverse branch allows text positions to attend to the audio sequence, making the interaction bidirectional rather than one-way. Each output retains the sequence length of its query modality, and the two results are reduced by global average pooling, concatenated and passed to a classification head containing batch normalization, a ReLU activation, dropout and a final sigmoid. In matrix terms, each branch scales query-key compatibility by the square root of the key dimension, applies a softmax and multiplies the resulting weights by the other modality’s values. The design therefore models dependencies within each representation and relationships across the two modalities.

The researchers trained the architecture in PyTorch with the AdamW optimizer, a learning rate of 2×10⁻⁵, batches of four and dropout of 0.1. For ADReSS, the official training set of 108 participants was divided 65% into training and 35% into validation, while the separate 48-participant challenge test set was reserved for final evaluation. The model was trained five times with different random seeds, with early stopping after 20 consecutive steps without a reduction in validation loss and a maximum of 50 epochs. The reported test metrics were means with standard deviations across the five runs, and typical convergence occurred between epochs 35 and 45. Each run took about 1.5 hours on a single NVIDIA RTX A6000 GPU, according to the paper.

p>An acoustic-only version of the system, called Whisper-OTKE, reached 79.17% accuracy and 78.23% F1 on ADReSS, exceeding the challenge’s reported 62.00% acoustic baseline. The ablation analysis showed that text carried more predictive information when used alone: BERT achieved an F1 of 85.67%, compared with 73.89% for Whisper alone. Combining BERT and Whisper through simple concatenation produced an F1 of 85.17%, slightly below the text-only result, suggesting that the audio was not being fully exploited. Adding cross-modal attention raised F1 to 87.98%, and combining that mechanism with OTKE produced the best reported score of 89.92%. Adding a position-wise feed-forward layer caused a slight decline, which the authors attribute to increased complexity and possible overfitting in a small clinical dataset.

Generalization was tested on the VAS dataset, which contains recordings of people issuing predefined commands to an Amazon Alexa device. The cohort included 36 healthy controls, 35 people with mild cognitive impairment and 29 people with dementia in the analysis; one person with advanced dementia was unable to complete the task. Participants produced 34 utterance types, including 30 commands commonly used by older adults and four response-type commands such as “yes” and “pause.” Because much of the command text was shared across participants, the transcript contained little information for distinguishing diagnostic groups, making this a stringent test of whether the acoustic channel contributed meaningful signal. Using stratified ten-fold cross-validation repeated three times, Cross-Attention-OTKE achieved 72.3% accuracy and 71.2% weighted F1 in the three-class task of distinguishing healthy control, mild cognitive impairment and dementia.

Probing all 12 layers of Whisper-small provided further evidence that the location of the acoustic representation mattered. The researchers aggregated variable-length hidden states from each encoder block with OTKE and classified ADReSS samples using speech alone. Earlier layers, which primarily encode lower-level voice quality and spectral characteristics, produced modest results. Classification metrics rose from around the sixth block onward as the representations became more abstract and temporally organized. The final layer performed best, reaching 79.17% accuracy, 78.23% F1 and a 79.06% G-mean. The authors interpret this pattern as evidence that higher-level Whisper representations contain the diagnostic information needed for the task, although the layer analysis alone does not establish which particular acoustic or linguistic features drive individual predictions.

The paper also compares the final architecture with a general-purpose large multimodal model evaluated zero-shot as a diagnostic assistant on audio and transcript inputs. The reported multimodal baseline achieved 70.83% accuracy, 72.73% precision and 66.67% recall, misclassifying eight of the 24 dementia cases in the test set. The specialized Cross-Attention-OTKE system was more accurate overall, but the comparison does not establish clinical superiority because the two systems were evaluated within a limited benchmark study. The paper’s authors describe the result as evidence that a purpose-built architecture can capture complementary acoustic and linguistic information more effectively than broad heuristic reasoning. They also acknowledge that hyperparameters were selected empirically rather than through a systematic grid search, and the model has no built-in method for explaining individual classifications.

p>For now, the findings are best understood as a promising computational result rather than a ready-made diagnostic test. The ADReSS cohort is small, and the paper does not report prospective validation across independent hospitals or diverse populations. Speech tasks also differ across studies: ADReSS participants described the Cookie Theft picture, whereas VAS participants interacted with an Alexa device, so the two datasets test related but non-identical behavior. The proposed future work includes systematic hyperparameter tuning, interpretability methods such as SHAP value analysis and visualizations of cross-attention matrices, and testing with automatically generated speech-recognition transcripts to better resemble workflows without manually prepared text. If those steps confirm the performance and clarify how the system reaches its decisions, the architecture could become a component of scalable speech-based cognitive assessment, but clinical adoption would require validation that the reported benchmark accuracy does not overstate performance in broader use.

Subject of Research: Experimental evaluation of a cross-modal speech and transcript model for detecting Alzheimer’s disease in speech datasets

Article Title: Alzheimer’s multimodal diagnosis: enhancing cross-attention between speech and text

Article References: Ding, K., Noori Hoshyar, A., Bhattacharya, T., & Chetty, M. (2026). Alzheimer’s multimodal diagnosis: enhancing cross-attention between speech and text. Neural Computing and Applications, 38(20), Article 824. https://doi.org/10.1007/s00521-026-12482-5

Image Credits: AI Generated

DOI: 10.1007/s00521-026-12482-5

Keywords: Alzheimer’s disease, speech analysis, multimodal AI, cross-modal attention, Whisper, BERT, machine learning, clinical benchmarks, Alzheimer, multimodal, diagnosis, enhancing

The research team, led by Rebecca Z. Rousset and Charlotte E. Teunissen of Amsterdam UMC, set out to answer a question that has practical consequences for the growing field of blood-based Alzheimer’s diagnostics: how do genetic risk factors, age, and sex interact to influence the concentrations of plasma biomarkers that clinicians increasingly rely on? Plasma biomarkers are attractive because they are far less invasive and less costly than positron emission tomography scans or cerebrospinal fluid analysis, yet their interpretation depends on understanding how factors such as age and genetics shift baseline values across the population. Without that context, a single biomarker reading could be misleading.

The cohort included 64.1 percent women and 29.8 percent APOE ε4 carriers, a proportion consistent with the general population. The APOE gene comes in three common variants — ε2, ε3, and ε4 — and carrying one or two copies of ε4 substantially raises lifetime risk of Alzheimer’s disease while lowering the age of onset. Beyond APOE, the researchers also computed a polygenic risk score for Alzheimer’s disease that excluded the APOE region, combining hundreds of smaller genetic effects scattered across the genome into a single numerical estimate of inherited risk. This allowed the team to separate the outsized influence of APOE ε4 from the cumulative, subtler contributions of the rest of the genome.

The results were striking in their specificity. Among ε4 carriers, the ratio of amyloid-beta 42 to amyloid-beta 40 — a marker that falls as amyloid plaques accumulate in the brain — began declining more steeply after age 50 compared with non-carriers. Concentrations of p-tau217, a phosphorylated form of the tau protein that is currently among the most accurate blood markers of Alzheimer’s pathology, started rising faster after age 55 in carriers. GFAP, a protein released by astrocytes when they respond to amyloid-driven injury, accelerated upward after age 60 in the carrier group. Each biomarker had its own age threshold, painting a picture of a genetically determined timeline in which amyloid changes precede tau and glial activation changes.

Perhaps the most consequential negative finding involved the polygenic risk score. Unlike APOE ε4, the PRS for Alzheimer’s disease excluding APOE showed no association with biomarker concentrations across age. In other words, hundreds of small genetic variants that collectively raise Alzheimer’s risk did not measurably shift the trajectory of plasma biomarkers in this population-based sample. This dissociation implies that APOE ε4 occupies a unique mechanistic position: its effect on biomarker aging is large enough to detect in a general population, while the polygenic background appears to influence risk through pathways that do not leave the same measurable fingerprint in peripheral blood.

Sex, by contrast, did enter the picture as a factor worth accounting for. With nearly two-thirds of the cohort being female, the dataset had adequate power to evaluate sex differences in biomarker trajectories, and the authors treated sex as an interacting variable in their models. The study’s design — a population-based cohort rather than a clinic-based sample — matters here, because it captures the full spectrum of brain health from cognitively healthy young adults to older individuals with undetected pathology, avoiding the referral biases that complicate interpretation in patient cohorts.

To verify that the cross-sectional age trends they observed actually reflected progressive disease biology rather than cohort artifacts, the team turned to a separate longitudinal resource: the EMIF-AD cohort. In this independent dataset with repeated measurements over time, rising p-tau217 concentrations in ε4 carriers were confirmed to reflect genuine Alzheimer’s pathology rather than nonspecific aging. This confirmation is critical, because p-tau217 has emerged as the leading candidate for blood-based Alzheimer’s screening, and its trajectory in genetically at-risk individuals defines the window in which intervention could begin before symptoms appear.

The technical underpinnings of the study reflect the state of the art in plasma neurochemistry. p-tau217 is measured using ultrasensitive immunoassays capable of detecting femtomolar concentrations, and its specificity for Alzheimer’s-type tau pathology has been validated against amyloid and tau PET imaging in multiple prior studies. The amyloid-beta 42/40 ratio captures the shift in amyloid peptide production that accompanies plaque deposition, while NfL serves as a nonspecific marker of axonal damage and GFAP indexes reactive astrogliosis — the brain’s inflammatory response that often begins years before clinical symptoms. Together, these four analytes form a layered view of Alzheimer’s biology, from amyloid accumulation through glial activation to neurodegeneration.

The age thresholds identified — 50 for amyloid ratio decline, 55 for p-tau217 acceleration, and 60 for GFAP rise in ε4 carriers — carry direct implications for screening strategy. If blood tests for Alzheimer’s are to be deployed preventively, the results suggest that ε4 carriers may benefit from earlier and more frequent testing, with p-tau217 offering the clearest signal of preclinical pathology. For non-carriers, the same biomarkers change more gradually with age, meaning that a single threshold value applied uniformly across the population could miss early disease in some individuals while overdiagnosing it in others. Genetically informed reference ranges could therefore sharpen the accuracy of population screening.

The study also demonstrates the value of twin register resources for disentangling genetic and environmental contributions to brain aging. The Netherlands Twin Register, with its decades-long collection of biological samples and health data, provided the statistical power to model interactions between genotype, age, and sex across a wide age range in a single harmonized dataset. As plasma biomarkers move from research laboratories toward clinical deployment, studies of this scale — grounded in population-based sampling, ultrasensitive assays, and longitudinal validation — will define how these tests are interpreted, who gets tested, and when the molecular countdown of Alzheimer’s disease first becomes visible in a vial of blood.

Subject of Research: APOE ε4 carriership status enhances age-related changes in Alzheimer’s disease plasma biomarkers concentrations

Article Title: APOE ε4 carriership status enhances age-related changes in Alzheimer’s disease plasma biomarkers concentrations

Article References: Rousset, R. Z., den Braber, A., Bongers, B., Boonkamp, L., Wilson, D., Ligthart, L., Pool, R., Visser, P. J., van der Landen, S. M., de Geus, E., & Teunissen, C. E. (2026). APOE ε4 carriership status enhances age-related changes in Alzheimer’s disease plasma biomarkers concentrations. Nature Communications. https://doi.org/10.1038/s41467-026-78723-2

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78723-2

Keywords: APOE, carriership, status, enhances, age-related, changes, Alzheimer, disease, plasma, biomarkers, concentrations, scientific research

The research team, led by Yang Jing and colleagues at Nanchong Psychosomatic Hospital in China, focused on miR-615-3p because it has repeatedly surfaced in Alzheimer’s disease datasets. MicroRNAs are tiny non-coding RNAs, roughly 22 nucleotides long, that fine-tune gene expression by binding to the 3′-untranslated region of target messenger RNAs, either blocking translation or promoting mRNA degradation. Previous work had shown that miR-615-3p is abnormally elevated in the brains and biofluids of Alzheimer’s patients, and that it appears among a small set of microRNA markers with detection accuracy exceeding 84 percent. It is also one of the few microRNAs found dysregulated in both Alzheimer’s and Parkinson’s disease. Yet its precise contribution to mitochondrial dysfunction remained unclear until now.

To dissect the mechanism, the researchers used HT22 cells, a mouse hippocampal neuronal cell line, and treated them with amyloid-β1-42, the peptide that forms the hallmark plaques of Alzheimer’s disease. This created an in vitro model that recapitulates key features of the disease, including amyloid accumulation, oxidative stress, and mitochondrial damage. The team then manipulated miR-615-3p levels using synthetic mimics to overexpress it and inhibitors to knock it down, while simultaneously measuring cell viability, apoptosis, ATP production, mitochondrial membrane potential, reactive oxygen species, lactate dehydrogenase release, and inflammatory cytokines such as IL-1β and IL-18.

The results were striking. When the researchers overexpressed miR-615-3p in amyloid-stressed neurons, cell viability and ATP levels dropped significantly, while apoptosis, amyloid-β accumulation, lactate dehydrogenase release, and reactive oxygen species production all increased. Mitochondrial membrane potential deteriorated, and inflammatory cytokine levels climbed. Crucially, the team observed elevated levels of mitophagy markers, including the LC3-II to LC3-I ratio, Parkin, PINK1, and p62. At first glance, this might suggest that mitophagy was being activated. But a closer look revealed something more insidious.

Using bafilomycin A1, a lysosomal inhibitor that blocks autophagic degradation, the researchers performed an autophagic flux assay. In control and model cells, bafilomycin A1 treatment caused LC3-II and p62 levels to accumulate further, indicating that the autophagic machinery was still functional and that the lysosomal blockage was being detected. But in miR-615-3p-overexpressing cells, bafilomycin A1 failed to further increase LC3-II and p62 levels. This meant that autophagic flux was already attenuated or completely blocked in these cells. The accumulation of mitophagy-related proteins was not a sign of increased autophagosome formation but rather of impaired autophagosome-lysosome degradation, a state the researchers describe as autophagy stress.

Having established that miR-615-3p disrupts mitophagy, the team turned to bioinformatics to identify its molecular target. Using the TargetScan and miRDB databases, they predicted that miR-615-3p binds to a conserved site within the 3′-untranslated region of SESN2 messenger RNA. SESN2, also known as Sestrin2, is a stress-inducible protein that regulates cellular metabolism, oxidative stress responses, and autophagy. The researchers confirmed this interaction using a dual-luciferase reporter assay, in which co-transfection of miR-615-3p mimics with a wild-type SESN2 3′-untranslated region construct significantly reduced luciferase activity, while a mutant construct lacking the binding site showed no such reduction. This proved that miR-615-3p directly targets SESN2.

The mechanistic picture became clearer when the team examined the AMPK/ULK1 signaling axis. SESN2 is known to activate AMP-activated protein kinase, or AMPK, which in turn phosphorylates ULK1 at the Ser555 site, triggering the autophagic machinery. The researchers found that miR-615-3p overexpression significantly reduced phosphorylation of both AMPK at Thr172 and ULK1 at Ser555, indicating genuine suppression of the cascade rather than an inhibitory phosphorylation event. Co-transfection with SESN2 siRNA produced similar reductions, while SESN2 overexpression or miR-615-3p inhibition restored phosphorylation levels. This established a clear regulatory chain: miR-615-3p downregulates SESN2, which suppresses AMPK/ULK1 phosphorylation, which blocks the initiation of mitophagy.

The therapeutic implications became apparent when the researchers overexpressed SESN2 in the amyloid-stressed neurons. SESN2 overexpression reversed nearly all the damage caused by amyloid-β1-42 treatment, restoring mitochondrial membrane potential and ATP levels, reducing reactive oxygen species and lactate dehydrogenase release, lowering LC3-II, Parkin, PINK1, and p62 levels, and suppressing inflammatory cytokines. Cell viability improved and apoptosis decreased. Conversely, knocking down SESN2 with siRNA exacerbated the damage, confirming that SESN2 is functionally protective in this context. The findings align with previous reports that SESN2 promotes mitophagy through AMPK activation in hypoxic nerve injury models, and that SESN2 deletion weakens mitophagy signaling and aggravates neuronal damage.

The study also highlights a paradox in mitochondrial quality control that has been gaining attention in Alzheimer’s research. Simply upregulating mitophagy-related proteins without restoring the complete autophagic flow may lead to autophagy accumulation and increased cytotoxicity. The activation of Parkin, for instance, has been shown to reduce amyloid load and improve synaptic plasticity in mouse models, and exercise intervention can enhance mitophagy by upregulating Parkin. But the new findings suggest that the real therapeutic goal is not merely to increase the expression of mitophagy markers but to restore the full autophagic flux from autophagosome formation through lysosomal degradation. This distinction matters because lysosomal dysfunction is increasingly recognized as a primary trigger of autophagic build-up and neuronal death in Alzheimer’s disease.

The researchers acknowledge several limitations. The study was conducted only in the HT22 cell line, which does not fully recapitulate the molecular features of primary or human neurons. Validation in primary neurons and in vivo Alzheimer’s models, such as APP/PS1 transgenic mice, will be necessary to confirm the pathological relevance of the miR-615-3p/SESN2/AMPK/ULK1 axis. The team also focused exclusively on the PINK1/Parkin-mediated mitophagy pathway and did not examine whether receptor-mediated mitophagy participates in the same regulatory network. Additionally, miR-615-3p may have multiple target genes beyond SESN2, and future studies using RNA sequencing or proteomics will be needed to identify other potential targets. Despite these caveats, the study provides a compelling mechanistic framework and suggests that developing miR-615-3p antagonists or SESN2 agonists could represent a promising therapeutic strategy for Alzheimer’s disease, one that targets the mitochondrial quality-control failure at the root of neurodegeneration rather than its downstream consequences.

Subject of Research: miR-615-3p targets SESN2 to inhibit neuronal mitophagy in an in vitro Alzheimer’s disease model by regulating the AMPK/ULK1 axis

Article Title: miR-615-3p targets SESN2 to inhibit neuronal mitophagy in an in vitro Alzheimer’s disease model by regulating the AMPK/ULK1 axis

Article References: Jing, Y., Li, D., Chen, L., Li, S., & Yuan, X. (2026). miR-615-3p targets SESN2 to inhibit neuronal mitophagy in an in vitro Alzheimer’s disease model by regulating the AMPK/ULK1 axis. Frontiers in Molecular Neuroscience, 19, Article 1873619. https://doi.org/10.3389/fnmol.2026.1873619

Image Credits: AI Generated

DOI: 10.3389/fnmol.2026.1873619

Keywords: miR-615-3p, targets, SESN2, inhibit, neuronal, mitophagy, vitro, Alzheimer, disease, model, regulating, AMPK

Amyloid-beta is not simply a waste product. The peptide is produced continuously in the brain, circulates in blood and cerebrospinal fluid, and is constantly exported, degraded and recycled by organs far from the site of plaque formation. The review’s central contribution is to trace this full trafficking cascade in one integrated picture, from intracerebral drainage routes to hepatic and renal elimination. The authors emphasize that cerebral amyloid-beta transport and peripheral amyloid-beta clearance have typically been investigated as independent processes, and that the coordinated interplay between the two remains insufficiently characterized. It is precisely this interplay, they argue, that holds the key to understanding why amyloid accumulates in some aging brains and not others.

Inside the cranium, the review identifies four major clearance pathways. The first is the blood-brain barrier, the specialized endothelial lining of cerebral capillaries that expresses transporters capable of shuttling amyloid-beta out of brain interstitial fluid and into the bloodstream. Receptors such as LRP1 mediate efflux of the peptide across this barrier, while receptors like RAGE can transport it in the opposite direction, and the balance between these opposing fluxes shifts with age and disease. When barrier transport falters, amyloid-beta that should leave the brain lingers and aggregates instead.

The second route runs through the blood-cerebrospinal fluid barrier, at the choroid plexus and arachnoid membranes, where amyloid-beta can be cleared into the cerebrospinal fluid that bathes and cushions the brain. The third pathway is venous: arachnoid granulations and associated outflow structures allow cerebrospinal fluid and its dissolved cargo to drain into the venous sinuses of the dura mater, carrying amyloid-beta back toward the systemic circulation. The fourth and most actively researched system combines the meningeal lymphatic vessels with the glymphatic system, the network of perivascular channels through which cerebrospinal fluid flushes interstitial solutes from brain tissue, particularly during sleep. Disruption of these lymphatic and glymphatic routes, whether through aging, poor sleep or impaired meningeal drainage, is increasingly linked to amyloid accumulation in animal models and human imaging studies.

Once amyloid-beta exits the brain, the peripheral organs take over. The review systematically illustrates how circulating amyloid-beta is handled by the blood, the liver, the kidneys and the gastrointestinal tract. The liver emerges as the dominant peripheral eliminator: hepatocytes and Kupffer cells take up the peptide from the portal and systemic circulation and degrade it, and the organ’s capacity to sequester circulating amyloid-beta is thought to substantially influence how much of the peptide recirculates to the brain. The kidneys filter and catabolize amyloid-beta as well, and the gastrointestinal tract contributes both through direct degradation and through modulation of the gut microbiome, which in turn affects systemic inflammation and amyloid metabolism.

Comparative analysis of central and peripheral clearance, as summarized in the review, indicates that the liver appears to dominate peripheral elimination of amyloid-beta. But the authors’ most consequential conclusion concerns synergy rather than supremacy. The real driver of cerebral amyloid deposition, they argue, is not the failure of any single route, but the loss of coordination among multiple pathways. A modest decline in blood-brain barrier efflux can be compensated by robust hepatic uptake; a sluggish glymphatic system can be offset by efficient renal clearance. Disease emerges when several routes degrade simultaneously, tipping the whole-body balance toward deposition. This mechanistic framework explains a persistent puzzle in Alzheimer’s drug development: why therapies aimed solely at the brain have struggled to produce durable benefits.

The therapeutic implications are substantial. The review surveys a broad spectrum of emerging strategies that target amyloid-beta transport and clearance rather than plaque formation alone. Specific pathway modulators aim to enhance transporter-mediated efflux at the blood-brain barrier or stimulate meningeal lymphatic drainage. Microglia regulators seek to restore the brain’s resident immune cells to a state where they efficiently phagocytose amyloid rather than fueling inflammatory cycles. Amyloid aggregation inhibitors attack the chemistry of plaque nucleation, preventing soluble monomers from assembling into the oligomers and fibrils that damage synapses.

Beyond these targeted approaches, the review highlights natural multi-target compounds, many derived from traditional medicinal plants, that simultaneously modulate inflammation, oxidative stress and amyloid clearance across several organs. Ultrasound-mediated blood-brain barrier opening, demonstrated in both preclinical models and early clinical trials, transiently increases barrier permeability so that amyloid-beta can be drained more readily and therapeutics can reach brain tissue. Exosome-based interventions, using engineered extracellular vesicles to deliver amyloid-degrading enzymes or modulatory cargo across biological barriers, represent another frontier. Each of these approaches, the authors note, could be paired with strategies that bolster peripheral organ function, such as supporting hepatic amyloid uptake or protecting renal clearance capacity.

This systemic perspective reframes the search for disease-modifying therapies. If Alzheimer’s reflects a breakdown of whole-body amyloid-beta homeostasis, then purely brain-centric therapeutics are unlikely to correct the underlying imbalance, no matter how effectively they target a single molecular pathway. The review argues that future translational efforts should give higher priority to combination regimens that simultaneously facilitate intracerebral drainage and bolster peripheral organ clearance. Such regimens might pair a glymphatic enhancer with a hepatic clearance booster, or combine barrier-opening ultrasound with systemic amyloid sequestration, addressing the trafficking cascade at multiple points rather than one.

The authors also confront current hurdles in systemic Alzheimer’s drug discovery. Measuring amyloid-beta kinetics across the brain-blood-organs axis in living patients remains technically difficult, and clinical trials designed around peripheral clearance endpoints are still scarce. Yet the opportunities are growing: advances in neuroimaging, liquid biopsy and organ-on-chip models are making whole-body amyloid dynamics increasingly tractable to study. By offering a unified conceptual framework, one that treats the brain and periphery as a single homeostatic system, the review aims to guide the exploration of novel therapeutic paradigms for a disease that has resisted decades of single-target approaches. For a field long focused on what happens inside the skull, the message is clear: the answer to Alzheimer’s may lie partly in the liver, the kidneys and the lymphatic vessels that connect them.

Subject of Research: Systemic amyloid-beta transport and clearance pathways in Alzheimer's disease

Article Title: New insight into Alzheimer’s disease: small molecule-Aβ in the brain transport and peripheral organ clearance

Article References: Ma, T., Li, Y., Ding, Y., Liu, S., Gao, C., Li, X., Tong, Z., & Lv, Y. (2026). New insight into Alzheimer’s disease: small molecule-Aβ in the brain transport and peripheral organ clearance. Frontiers in Aging, 7, Article 1921485. https://doi.org/10.3389/fragi.2026.1921485

Image Credits: AI Generated

DOI: 10.3389/fragi.2026.1921485

Keywords: Alzheimer's disease, amyloid-beta, blood-brain barrier, glymphatic system, liver clearance, meningeal lymphatics, peripheral clearance, drug development, insight, Alzheimer, disease, small

The hippocampus is the brain’s primary memory-encoding structure, and its CA1 and CA3 regions are critical for both storing new information and retrieving stored patterns. Long-term potentiation (LTP), the strengthening of synaptic connections that underlies memory formation, depends heavily on NMDA receptors—ion channels that open in response to glutamate binding and postsynaptic depolarization, allowing calcium to enter the cell and trigger the molecular cascades that consolidate synaptic change. In early Alzheimer’s disease, pathological levels of amyloid-β peptide (Aβ) bind to NMDA receptors and drive excessive activation, ultimately disrupting the delicate balance between potentiation and depression that memory storage requires.

Previous experimental work had shown that pharmacologically blocking GluN2B-containing NMDA receptors with the antagonist ifenprodil could rescue impaired LTP in CA1 pyramidal neurons exposed to Aβ. Computational models at the single-neuron level had further elucidated the biophysical mechanisms underlying this rescue. But whether these cellular-level mechanisms remain effective when synaptic impairments are distributed across a population of neurons—and how they ultimately affect memory storage and recall at a network level—could not be readily addressed experimentally. The new study by Fabio Librizzi, Saana Seppälä, Marja-Leena Linne, Michele Migliore, and Ausra Saudargiene set out to answer exactly that question.

The researchers built their model on a well-established CA1 network architecture comprising 100 hippocampal CA1 pyramidal neurons with four types of inhibitory interneurons: oriens-lacunosum moleculare, basket, bistratified, and axo-axonic cells. Each CA1 neuron was modeled with 15 compartments reflecting the soma, axon, and dendritic regions. The network receives excitatory inputs from the entorhinal cortex (EC) and CA3 Schaffer collaterals, activated by gamma bursts (30–80 Hz) coupled with a theta rhythm (4–8 Hz), mirroring the oscillatory patterns that govern memory encoding in the living brain. Each CA3 input creates two synaptic connections on the proximal dendrites of each CA1 neuron—one AMPA-mediated and one NMDA-mediated—while EC inputs arrive 9 milliseconds earlier and form AMPA-mediated connections on distal dendrites.

To simulate early Alzheimer’s pathology, the team modeled the effects of accumulated Aβ by reducing synaptic density to 80 percent of its original value, increasing free glutamate concentration to 120 percent, and modifying the shape of the GluN2B-dependent LTP function so that potentiation was abolished even when the subunit was activated. These adjustments, grounded in prior electrophysiological observations, produced a strong inhibition of LTP and a complete lack of synaptic weight increase during the storage protocol. The researchers then tested whether partially blocking GluN2B function could restore the impaired plasticity.

The results were striking. Reducing GluN2B conductivity to 40 percent of its original value produced a recovery of roughly 90 percent of the original synaptic weight increase during the 25-second storage protocol. The recovery fraction depended strongly on the degree of GluN2B attenuation, with 0.4 microsiemens providing the maximum rescue for the network. Importantly, control simulations in which AMPA receptor conductance was reduced across a range of values produced no recovery whatsoever, confirming that the rescue is specific to the GluN2B-dependent mechanism rather than reflecting a nonspecific reduction of excitatory drive. This specificity aligns with the known role of GluN2B–NMDAR-dependent calcium/calmodulin kinase II (CaMKII) signaling in Aβ-induced LTP disruption.

The team then evaluated network performance in storing and retrieving multiple patterns, both orthogonal (non-overlapping) and non-orthogonal (overlapping). With 100 CA1 cells and 20-input patterns, a maximum of five orthogonal patterns could be presented. Recall quality was nearly perfect for orthogonal patterns, as expected, since they activate entirely separate sets of synapses. For non-orthogonal patterns, quality decreased even with only five stored patterns, mainly due to spurious spiking of cells not involved in the recalled pattern. When 50 non-orthogonal patterns were stored, recall quality approached the threshold corresponding to the spiking of all 100 CA1 neurons—effectively rendering the output uninformative about which memory was retrieved.

When the researchers introduced different fractions of AD-affected synapses—0, 20, 50, 80, and 100 percent—they found that recall quality decreased significantly as the fraction of impaired cells increased. However, reducing GluN2B activity to 40 percent of its original value strongly counteracted this effect, to the point that in rescue conditions the quality factor almost approached its maximum value even when 100 percent of synapses were AD-affected. For non-orthogonal patterns, the partial GluN2B inhibition appeared to reduce spurious spiking of cells outside the recalled pattern, improving recall specificity, although it simultaneously reduced the average number of spikes, weakening quality in some cases. The interplay between these two effects produced a nuanced network-level outcome that could not have been predicted from single-synapse mechanisms alone.

The authors emphasize that the degree of recovery depends on population-level variables such as the fraction of affected CA1 neurons and the structure of the stored patterns. The model’s robust behavior across different levels of synaptic impairment and pattern separation supports the presented mechanisms for synaptic recovery in Alzheimer’s-like conditions. The results indicate that accumulated Aβ disrupts GluN2B function in hippocampal microcircuits, impairing memory storage and recall especially by affecting theta-gamma coupling, which can be partly restored by suppressing GluN2B conductivity.

While no specific GluN2B antagonist has yet been approved as a therapeutic agent for Alzheimer’s disease, unspecific NMDA receptor antagonists such as memantine have received FDA approval for their modest symptomatic effects. The new findings provide a proof-of-concept that selective GluN2B blockade could counteract early AD impairments not only at the single-neuron level but also at the network level in memory storage and recall. The authors stress that the 40 percent reduction identified here should not be considered a universal optimum, but rather that an intermediate level of GluN2B activity exists that optimally restores LTP under the conditions studied. Future work incorporating astrocytic dynamics and other molecular mechanisms into large-scale network models could further elucidate how early synaptic dysfunction propagates to network-level deficits and how targeted interventions can restore function in Alzheimer’s disease.

Subject of Research: Network-level restoration of impaired memory function in early Alzheimer’s disease via GluN2B modulation: insights from a hippocampal CA1–CA3 model

Article Title: Network-level restoration of impaired memory function in early Alzheimer’s disease via GluN2B modulation: insights from a hippocampal CA1–CA3 model

Article References: Librizzi, F., Seppälä, S., Linne, M.-L., Migliore, M., & Saudargiene, A. (2026). Network-level restoration of impaired memory function in early Alzheimer’s disease via GluN2B modulation: insights from a hippocampal CA1–CA3 model. Frontiers in Computational Neuroscience, 20, Article 1953352. https://doi.org/10.3389/fncom.2026.1953352

Image Credits: AI Generated

DOI: 10.3389/fncom.2026.1953352

Keywords: Network-level, restoration, impaired, memory, function, early, Alzheimer, disease, GluN2B, modulation, insights, hippocampal

The research team analyzed 188 posts shared between March 2014 and May 2025 on Facebook, Instagram, and TikTok that discussed Alzheimer’s blood tests. These posts originated from accounts with a combined audience exceeding 114 million followers, illustrating the enormous reach that health-related content can achieve on social platforms. The study employed a cross-sectional content analysis methodology, allowing researchers to categorize posts according to their claims, the presence or absence of supporting evidence, and whether they aligned with established clinical guidance regarding appropriate use of the tests.

Among the most concerning findings was that more than a third of the analyzed posts suggested the blood tests could be used to screen healthy individuals who show no symptoms of cognitive decline. This recommendation directly conflicts with current clinical guidelines, which advise against routine testing of asymptomatic people. The blood tests are designed to support diagnosis of Alzheimer’s disease specifically in individuals already experiencing cognitive symptoms, not as general screening tools for the broader population. Despite this, social media content frequently blurred the line between diagnostic use and widespread screening.

Dr. Jenna Smith, the lead author of the study from the University of Sydney’s School of Public Health and the Sydney Health Literacy Lab, emphasized the gap between social media enthusiasm and scientific evidence. She noted that social media is presenting exciting claims about what these new blood tests might be able to do, but that excitement can outpace the evidence. According to Dr. Smith, people increasingly turn to social media for health information, and when content omits important context about who these tests are for, their limitations, and potential downsides, there is a real risk that individuals may develop unrealistic expectations or seek testing that is not clinically recommended for them.

The analysis revealed that 59 percent of posts promoted the blood tests as useful for earlier or improved diagnosis of Alzheimer’s disease. While this claim has some basis in clinical practice, many posts failed to mention that a comprehensive clinical assessment remains essential to any diagnosis, or that the tests are only recommended for people experiencing symptoms of cognitive decline. This omission is significant because blood biomarkers for Alzheimer’s disease, while increasingly accurate, are intended as one component of a broader diagnostic process that includes cognitive evaluation, clinical history, and in some cases imaging or cerebrospinal fluid analysis.

Perhaps most striking was the finding regarding evidence quality and financial incentives. Only 36 percent of the posts included any evidence supporting their claims, and fewer than one quarter of those cited a specific study. Even more concerning, accounts with explicit financial interests were more than three times as likely to promote testing for people without symptoms compared to accounts without such interests. This correlation suggests that commercial motivations may be driving some of the most misleading content, potentially encouraging healthy individuals to pursue testing that offers no proven benefit and carries real risks of harm.

The researchers highlighted several potential harms associated with testing asymptomatic individuals that were rarely mentioned in the social media posts discussing this use case. These harms include overdiagnosis, false positive results, and the broader psychological and social impacts that an Alzheimer’s diagnosis can have on individuals and their families. Dr. Smith explained that dementia is one of the most feared health conditions among adults, and understandably, people want answers and may be drawn to new tests that appear to offer certainty or early detection. However, for people without symptoms, there is currently no evidence that screening provides any benefit.

A positive test result in an asymptomatic person may create significant anxiety and uncertainty, particularly given that there is no cure for Alzheimer’s disease and that guidance on how these new tests should be used is still evolving. Furthermore, people with no symptoms could potentially be diagnosed with Alzheimer’s disease based on biomarker evidence but never go on to develop dementia that actually impacts their daily life. This phenomenon of biomarker-positive but clinically asymptomatic individuals underscores the complexity of interpreting these tests outside of a proper clinical context and highlights why current recommendations advise against routine screening of healthy populations.

The study’s authors are calling for several responses to address these issues. They emphasize the need for clear, evidence-based information to help the public understand who may benefit from testing and what the results can and cannot tell them. Dr. Smith noted that these tests may play an important role in supporting diagnosis for some people experiencing cognitive symptoms, but they should be interpreted with appropriate clinical oversight. She stressed the importance of consumers receiving balanced information rather than relying on social media alone, and recommended that posts should always include advice to speak with a health professional before undertaking any Alzheimer’s blood testing.

Looking forward, the researchers say clinicians should also be aware of how social media may shape public perceptions of Alzheimer’s blood tests, as patients increasingly arrive at appointments with preconceived notions formed by online content. They are calling for improved regulatory enforcement to address misleading health claims on social platforms, alongside the development of resources to support meaningful conversations between patients and healthcare providers about the benefits, limitations, and uncertainties surrounding Alzheimer’s blood testing. As these diagnostic tools become more widely available and public interest continues to grow, ensuring that accurate, balanced information reaches consumers will be essential to preventing harm while preserving legitimate access to valuable diagnostic advances for those who truly need them.

Subject of Research: Social media misinformation about Alzheimer's disease blood tests and their appropriate clinical use

Article Title: Social media spreading misleading claims about Alzheimer’s blood tests, study finds

Article References: Social media spreading misleading claims about Alzheimer’s blood tests, study finds. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: Alzheimer's disease, blood tests, social media, health misinformation, dementia, clinical recommendations, Social, media, spreading, misleading, claims, Alzheimer

Alzheimer’s disease is characterised by the accumulation of extracellular amyloid-β plaques and intracellular Tau neurofibrillary tangles, leading to progressive memory impairment and cognitive dysfunction. Parkinson’s disease, by contrast, involves the degeneration of dopaminergic neurons in the substantia nigra and the aggregation of α-synuclein, producing motor impairments such as tremors, rigidity, and bradykinesia. Both conditions involve progressive neuronal loss, misfolded protein accumulation, and irreversible decline in cognitive or motor function. Yet the review suggests that neither disease is solely brain-centric. Alterations in peripheral systems, especially the gastrointestinal tract, play a major role in disease pathogenesis.

Gut dysbiosis in both Alzheimer’s and Parkinson’s disease commonly involves reduced microbial diversity, altered taxonomic composition, depletion of beneficial short-chain fatty acid-producing bacteria, and enrichment of pro-inflammatory taxa. These changes compromise the integrity of the intestinal barrier, leading to increased intestinal permeability—often referred to as leaky gut. Microbial products such as lipopolysaccharides translocate into the systemic circulation, triggering immune activation and chronic inflammation. This, in turn, disrupts the blood-brain barrier, activates microglia, and promotes neuroinflammatory cascades that contribute to neuronal damage and disease progression.

Short-chain fatty acids, primarily acetate, propionate, and butyrate, are produced when gut bacteria ferment dietary fibres. These metabolites maintain intestinal integrity, regulate immune responses, and modulate neuronal activity. Butyrate acts as a histone deacetylase inhibitor, influencing gene expression associated with neuroprotective and anti-inflammatory pathways. In both diseases, a reduction in SCFA-producing bacteria such as Faecalibacterium, Roseburia, and Eubacterium has been reported, diminishing production of these beneficial metabolites and contributing to increased intestinal permeability, systemic inflammation, and altered microglial activation within the brain.

The review places particular emphasis on inflammatory mediators including tumour necrosis factor alpha, interleukin 1 beta, interleukin 6, the toll-like receptor 4/nuclear factor kappa B pathway, and the NLRP3 inflammasome. Translocation of bacterial endotoxins such as lipopolysaccharide activates toll-like receptor 4, triggering NF-κB-mediated inflammatory cascades, driving microglial activation, and elevating cytokines. Gut barrier dysfunction and increased exposure to microbial products provide inflammatory priming through TLR4/NF-κB signalling, followed by NLRP3 inflammasome activation and caspase-1-mediated maturation of IL-1β and IL-18. In Alzheimer’s disease, NLRP3 activation has been associated with amyloid-β-induced microglial activation and impaired inflammatory regulation. In Parkinson’s disease, α-synuclein aggregates can activate NLRP3-dependent inflammatory responses in microglia, contributing to neuronal injury.

One of the most striking comparative findings concerns the vagus nerve. In Parkinson’s disease, pathological propagation of α-synuclein from the enteric nervous system to the central nervous system via vagal pathways has been proposed, supported by the observation that gastrointestinal symptoms such as constipation and altered motility often precede motor signs by years. Germ-free α-synuclein-overexpressing mice exhibit attenuated motor deficits that reappear after colonisation with Parkinson’s patient microbiota, implicating a causal role for the microbiota. In Alzheimer’s disease, however, current evidence does not support the generalisation that pathology originates in the gastrointestinal tract. Instead, Alzheimer’s is closely linked to metabolic dysfunction, insulin resistance, and systemic inflammatory priming, with microbial metabolites contributing to mitochondrial dysfunction, oxidative stress, and a shift in tryptophan metabolism towards the kynurenine pathway, yielding neurotoxic metabolites such as quinolinic acid.

Microbial amyloids represent another shared mechanism. The gut microbiota produces amyloid-like proteins that structurally resemble pathological proteins found in neurodegenerative diseases. These microbial amyloids can stimulate immune responses and potentially promote cross-seeding of host proteins, reinforcing amyloid-β accumulation in Alzheimer’s and α-synuclein aggregation in Parkinson’s. Certain bacteria, such as Escherichia coli, produce functional amyloids that cross-react with host proteins, promoting misfolding and aggregation. The review also highlights the role of neurotransmitters and neuromodulators, including serotonin, dopamine precursors, GABA, and acetylcholine, which the gut microbiota synthesises and which influence central neurotransmission via vagal and immune pathways.

The review then turns to recent advances in artificial intelligence and machine learning for integrating microbiome, clinical, neuroimaging, and multi-omics datasets. Traditional diagnostic approaches depend on clinical symptoms, neuropsychological testing, and imaging findings that typically reflect late-stage pathology. AI-driven models facilitate early diagnosis, disease prediction, and patient stratification by identifying subtle patterns across multimodal data, including neuroimaging, biofluid biomarkers, genetic profiles, and digital health metrics. Convolutional neural networks can detect subtle hippocampal atrophy and cortical degradation in early Alzheimer’s, while PET-based machine learning classifiers differentiate amyloid-positive from amyloid-negative individuals with high sensitivity and specificity. In Parkinson’s disease, machine learning algorithms applied to dopamine transporter imaging help identify nigrostriatal degeneration and differentiate idiopathic Parkinson’s from atypical Parkinsonian syndromes.

Microbiome analysis platforms such as QIIME 2 provide reproducible workflows for processing sequencing data, including quality assessment, feature generation, taxonomic classification, and microbial diversity analysis. Multi-omics integration tools including mixOmics and MOFA+ support multivariate integration and latent-factor modelling across heterogeneous omics datasets, facilitating the identification of coordinated molecular signatures associated with either disease. Machine learning algorithms such as random forests, support vector machines, and XGBoost can integrate clinical characteristics, genetic variants, inflammatory biomarkers, microbiome-derived features, and metabolomic profiles for disease classification, risk prediction, and patient stratification. XGBoost is particularly applicable to structured biomedical datasets because it can model non-linear relationships and interactions amongst multiple predictors whilst incorporating regularisation to reduce overfitting. Deep neural networks can analyse high-dimensional datasets derived from genome-wide transcriptomics, epigenomics, and metabolomics, and integrating genetic risk factors such as APOE with machine learning models might improve predictive accuracy for disease progression.

Digital biomarkers are also redefining early detection. AI-based wearable sensors and smartphone-based motion tracking systems generate continuous data on tremor amplitude, bradykinesia, gait instability, and motor fluctuation in Parkinson’s disease. Recurrent neural networks and long short-term memory models detect longitudinal motor patterns and identify subtle abnormalities early, enabling remote monitoring. In Alzheimer’s disease, natural language processing techniques analyse speech patterns, semantic errors, pause frequency, and linguistic diversity to detect early cognitive decline, providing a non-invasive, scalable screening tool with potential for community-level implementation. The review concludes that integrating gut-brain axis biology with AI-driven analytics provides promising opportunities for early diagnosis, personalised therapeutic strategies, and precision medicine. However, the authors caution that high predictive performance within a development cohort should not be interpreted as evidence of clinical validity without robust external and prospective validation, and that microbiome-based models present additional challenges because microbial abundance data are compositional and strongly influenced by diet, medication, geography, and sample-processing procedures.

Subject of Research: Health & Medicine

Article Title: The gut-brain axis in Alzheimer’s and Parkinson’s diseases: a comparative perspective

Article References: Upadhyay, R., C., C., Kar, S., Sathyamoorthy, S., Pitchaimuthu, X., & Sevanan, M. (2026). The gut-brain axis in Alzheimer’s and Parkinson’s diseases: a comparative perspective. Frontiers in Microbiomes, 5, Article 1942228. https://doi.org/10.3389/frmbi.2026.1942228

Image Credits: AI Generated

DOI: 10.3389/frmbi.2026.1942228

Keywords: gut-brain axis, Alzheimer's disease, Parkinson's disease, gut microbiota, neuroinflammation, artificial intelligence, precision medicine, neurodegenerative disease, gut-brain, axis, Alzheimer, Parkinson

Alzheimer’s disease remains the most prevalent cause of dementia worldwide, characterized clinically by progressive memory and executive deficits and pathologically by amyloid-β plaques, tau neurofibrillary tangles, synaptic loss, and chronic neuroinflammation. Current pharmacological interventions confer at best modest symptomatic benefit and have limited impact on long-term disease progression. Mounting evidence indicates that circadian rhythm disruption is both highly prevalent in patients and may actively exacerbate amyloid deposition, tauopathy, and neuroimmune dysfunction. Core clock components regulate sleep-wake cycles, energy metabolism, redox homeostasis, and immune signaling, and disturbances in clock-controlled neuroimmune circuits in microglia and astrocytes are increasingly recognized as central contributors to sustained neuroinflammation and synaptic vulnerability in the aging and diseased brain.

REV-ERBα (NR1D1) and REV-ERBβ (NR1D2) are atypical members of the nuclear receptor superfamily. Structurally, they diverge from classical ligand-activated receptors by lacking the carboxy-terminal activation function-2 helix within their ligand-binding domain, a motif normally required for coactivator recruitment and transcriptional activation. As a consequence, they behave predominantly as constitutive transcriptional repressors, often competing with the transcriptionally activating ROR nuclear receptors at shared RORE sites. Ligand engagement within the hydrophobic pocket of the ligand-binding domain, most notably by endogenous heme or synthetic agonists, stabilizes a conformation that enhances high-affinity binding to the nuclear receptor corepressor NCoR1/2 and enables assembly of a multi-protein repression complex containing histone deacetylase 3. Recruitment of NCoR-HDAC3 results in targeted deacetylation of histone tails, chromatin compaction, and durable silencing of REV-ERB target enhancers and promoters. Within the core clockwork, REV-ERBs oppose the activity of the BMAL1-CLOCK complex, repressing Bmal1 transcription and shaping 24-hour oscillations in gene expression.

Beyond their canonical clock roles, REV-ERB complexes regulate extensive transcriptional networks controlling inflammatory signaling, intermediary metabolism, and the core clock machinery. In immune cells, REV-ERBα directly binds regulatory elements near Il-1β, Il-6, and Tnfα, limiting their inducible expression and constraining NF-κB-driven inflammatory programs. In peripheral macrophages, NR1D1 directly binds the Nlrp3 promoter to constrain NLRP3 inflammasome activity and the maturation of interleukin-1β and interleukin-18, positioning the REV-ERBs as gatekeepers of innate immune activation. In metabolic tissues, REV-ERBs occupy enhancers of genes such as Pgc1α and Srebp1, rhythmically recruiting HDAC3 to coordinate mitochondrial biogenesis, fatty acid oxidation, and lipogenesis with circadian nutrient availability. The review frames this convergence as molecular integration of immune-metabolic timing, translating oscillations in heme and redox state into coordinated changes in chromatin accessibility across diverse cell types.

Transcriptomic and proteomic analyses reveal significantly reduced levels of REV-ERBα in Alzheimer’s disease brains, particularly within vulnerable regions such as the frontal cortex, hippocampus, and suprachiasmatic nucleus. In both human postmortem tissue and AD mouse models, these reductions coincide with disrupted clock gene rhythmicity and impaired coupling between the SCN and downstream oscillators. Single-cell and single-nucleus RNA-seq datasets from human AD cortex and hippocampus now provide cellular resolution, demonstrating a consistent downregulation of NR1D1 transcripts in disease-associated microglia and subsets of reactive astrocytes relative to homeostatic counterparts in control brains. Clusters with reduced NR1D1 show enrichment of pro-inflammatory gene modules and complement and chemokine signatures, implicating loss of REV-ERBα repression as a key step in the transition from physiological to pathological glial states. The authors caution, however, that direct measurements of NR1D1 or NR1D2 proteins in CSF or blood remain limited, and most evidence linking REV-ERB pathways to human disease derives from transcriptomic datasets and postmortem tissue analyses rather than validated fluid biomarkers.

The microglial story is the most mechanistically developed. In global Nr1d1 knockout mice, loss of this repression leads to spontaneous hippocampal microgliosis and secondary astrogliosis, accompanied by enhanced NF-κB p65 nuclear accumulation and upregulation of pro-inflammatory cytokines and complement factors including Il-1β, Tnfα, and C1q. Beyond cytokines, Nr1d1 deletion disrupts the circadian control of complement gene expression, producing exaggerated diurnal swings and an overall elevation of chemokines such as Ccl2 and Cxcl10, complement components C3 and C4b, and inflammasome-associated transcripts like Nlrp3 and Casp1. These changes collectively favor excessive synaptic tagging and microglial phagocytosis. Strikingly, depletion of microglia in these knockout mice normalizes many of the behavioral and cognitive abnormalities and attenuates neuroinflammatory signatures, demonstrating that hyperactivated microglia are the principal effectors of the neuropathology that emerges when REV-ERBα signaling is lost.

Tauopathy models provide further evidence that microglial REV-ERBα links inflammatory tone, lipid metabolism, and protein aggregate clearance. In PS19 mice expressing mutant human P301S tau, microglia-specific Nr1d1 deletion exacerbates tau aggregation and neurofibrillary tangle burden. Notably, these effects are sex-specific, occurring only in male mice, while female mice show no exacerbation of tau pathology. In affected males, loss of REV-ERBα increases inflammatory cytokine expression and drives accumulation of lipid droplets within microglia. This clogged metabolic state impairs their phagocytic capacity for pathological tau species, and pharmacologic or genetic blockade of lipid droplet formation partially restores tau uptake. The review also distinguishes experimentally validated functions from extrapolated models: direct evidence supports regulation of NF-κB signaling, complement expression, microglial synaptic phagocytosis, lipid droplet accumulation, and tau pathology, whereas links to broader inflammasome networks, oxidative stress pathways, and certain aspects of amyloid pathology remain largely mechanistic inferences.

Astrocytic REV-ERBα exerts context-dependent effects that diverge sharply from the microglial picture. In the astrocyte compartment, absence of this receptor induces an increase in bulk brain and astrocytic NAD+ levels via an NFIL3-CD38 signaling axis, where REV-ERBα deficiency leads to derepression of NFIL3, which subsequently limits CD38-dependent NAD+ consumption and preserves intracellular NAD+ pools. In P301S tauopathy models, this astrocytic REV-ERBα deficiency attenuates tau aggregation, neuroinflammation, and neuronal loss, indicating that reduced astrocytic REV-ERBα activity can paradoxically confer neuroprotection by enhancing NAD+-dependent stress resilience. Yet in primary astrocyte cultures and human iPSC-derived astrocytes, reduction of NR1D1 drives a robust pro-inflammatory shift characterized by upregulation of IL-6, NOS2, and NF-κB reporter activity alongside a neurotoxic phenotype that compromises cocultured neuron survival. Astrocytic REV-ERBα thus functions as a dual-purpose regulator, simultaneously restraining inflammatory gene expression and tuning redox and NAD+ metabolism, and the net outcome appears to depend on disease stage, injury context, and the balance between immune repression and metabolic control.

Pharmacological activation confers broad neuroprotective effects across multiple experimental paradigms. The first-generation synthetic agonist GSK4112 established proof-of-concept that small molecules can ligand-dependently enhance REV-ERB-NCoR interactions and repress canonical target genes such as Bmal1, but its low potency and poor metabolic stability limited its use largely to in vitro studies. Second-generation benzimidazole derivatives SR9009 and SR9011 exhibited greater bioavailability, CNS penetration, and in vivo efficacy. In APP/PS1 mice, systemic administration of these agonists reduces microgliosis and astrogliosis, attenuates NF-κB-dependent inflammatory gene programs, and restores performance in hippocampal-dependent memory tasks. In microglia and macrophages, these agonists attenuate pro-inflammatory cytokine production, suppress NLRP3 inflammasome activation, and shift cellular metabolism toward a less oxidative, lower-ATP state. Nonetheless, SR9009 and SR9011 display relatively short plasma half-lives, off-target effects at higher concentrations, and modest receptor affinity, which together constrain their translational potential. The competitive antagonist SR8278 tells a bidirectional story: short-term treatment enhances Aβ clearance by upregulating phagocytic receptors including TREM2 and P2Y12 and boosting microglial chemotaxis, but chronic or high-dose exposure has been associated with exacerbated astrogliosis and increased pro-inflammatory gene expression. More recently, the high-affinity pan-agonist STL1267 has emerged as a next-generation tool compound with submicromolar binding to REV-ERBα, enhanced metabolic stability, and superior in vivo exposure. Structural studies reveal that STL1267 occupies a deep hydrophobic cavity within the REV-ERBα ligand-binding domain, and structure-activity relationship analyses have identified a cluster of residues, particularly phenylalanine 484, 488, and 497 together with leucine 498, as critical determinants of agonist binding, forming a hydrophobic microenvironment that cradles the ligand and stabilizes the agonist-bound conformation favoring productive NCoR/HDAC3 recruitment.

The rhythmic nature of REV-ERB signaling adds another layer of complexity and opportunity. Circadian disruption happens early in Alzheimer’s disease, and core clock components like REV-ERBα directly affect neuroinflammation and neuronal vulnerability. Preclinical studies already provide proof-of-principle for chronotherapeutic REV-ERB modulation: in rodent models, restricting dosing of REV-ERB ligands or other anti-inflammatory agents to specific circadian phases improves restoration of clock gene rhythms in hippocampus and cortex, enhances diurnal variation in glial activation markers, and more effectively reduces cytokine levels and complement activation than phase-mismatched schedules. Time-restricted administration also appears to better preserve sleep-wake architecture and cognitive performance. Long-term systemic activation of REV-ERBs may produce adverse consequences because these receptors regulate lipid metabolism, glucose homeostasis, skeletal muscle energetics, immune responses, and cardiovascular physiology, a challenge particularly relevant for elderly AD patients who frequently exhibit multimorbidity and altered sleep-wake rhythms. The review argues that therapeutic success will rely on cell-type-selective modulation, stage-specific strategies, and precise calibration rather than a simple push for maximal receptor activation. Future research should prioritize cell-type and time-resolved mapping of REV-ERB chromatin occupancy in human and experimental AD, development of brain-penetrant and isoform-selective next-generation ligands, integration of REV-ERB-targeted treatments with existing AD therapies such as anti-amyloid antibodies or NLRP3 inhibitors, and tracking of circadian biomarkers like sleep-wake metrics and melatonin rhythms alongside standard neurodegeneration markers in early clinical studies. Emerging ATX(N)-based biomarker frameworks increasingly recognize inflammatory, vascular, metabolic, and synaptic pathways as clinically relevant components of AD biology beyond amyloid and tau alone, and future studies integrating NR1D1/NR1D2 expression, circadian biomarkers, and blood-based inflammatory markers may help determine whether REV-ERB pathway dysfunction can serve as a clinically informative biomarker of disease progression. The evidence positions REV-ERBs as modular, homeostatic sensors at the intersection of circadian timing, metabolic flux, and neuroimmune function, active determinants of whether glial responses remain protective or shift into chronic, neurotoxic inflammation in the Alzheimer’s brain.

Subject of Research: REV-ERBs as regulators of circadian rhythm, neuroinflammation, and glial lipid homeostasis in Alzheimer’s disease. A narrative review

Article Title: REV-ERBs as regulators of circadian rhythm, neuroinflammation, and glial lipid homeostasis in Alzheimer’s disease. A narrative review

Article References: Shabnam, F., Singh, M., Amin, R., & Griffett, K. (2026). REV-ERBs as regulators of circadian rhythm, neuroinflammation, and glial lipid homeostasis in Alzheimer’s disease. A narrative review. British Journal of Biomedical Science, 83, Article 16121. https://doi.org/10.3389/bjbs.2026.16121

Image Credits: AI Generated

DOI: 10.3389/bjbs.2026.16121

Keywords: REV-ERBs, regulators, circadian, rhythm, neuroinflammation, glial, lipid, homeostasis, Alzheimer, disease, narrative, review

The team started with the publicly available sequencing dataset GSE203206, downloaded from the Gene Expression Omnibus repository, which contains transcriptomic data from 39 Alzheimer’s disease samples and 8 normal control samples. Using t-tests implemented in the limma R Bioconductor package, they identified 958 differentially expressed genes, with a strikingly symmetric result of 479 genes upregulated and 479 genes downregulated in the diseased brains. That balance alone is noteworthy, because it suggests widespread, bidirectional remodeling of the transcriptome rather than a simple overshoot of one or two processes, and it gives researchers a rich catalogue of candidate players to investigate.

To make biological sense of that gene list, the researchers ran Gene Ontology and pathway enrichment analyses. The upregulated genes clustered mainly around response to stimulus, cytoplasmic functions, small molecule binding and signal transduction, painting a picture of a brain under persistent stress, actively reorganizing its signaling machinery. The downregulated genes, by contrast, were enriched for multicellular organism development, cell junction biology, ion binding and cardiac conduction, hinting that fundamental structural and developmental programs are being quietly dismantled as the disease progresses. These divergent functional signatures reinforce the idea that Alzheimer’s is not one pathway gone wrong but an entire coordinated system drifting out of tune.

The next step was network analysis. By mapping the differentially expressed genes onto a protein-protein interaction network, the team built a graph containing 4,886 nodes and 10,342 edges, an enormous molecular web from which they extracted the most connected hubs. Ten genes rose to the top: HSP90AA1, FN1, KIT, YAP1, LSM2, SKP1, EIF5A2, TAF9, DDX39B and CDK7. Several of these names will be familiar to neuroscientists. HSP90AA1 encodes a heat shock protein chaperone already flagged by proteomic studies in the entorhinal cortex of Alzheimer’s patients, while FN1, the fibronectin gene, has recently been linked through rare genetic variants to protection against the notorious APOE ε4 risk factor.

Beyond static expression patterns, the study probed the regulatory layers that might control these hub genes. Constructing a microRNA-hub gene regulatory network, the authors identified hsa-mir-545-3p and hsa-miR-548f-5p as microRNAs that could help orchestrate Alzheimer’s pathology by fine-tuning multiple hub genes simultaneously. Similarly, a transcription factor-hub gene network implicated PLAG1 and MEF2A as master regulators that may be involved in disease development. This kind of multi-level regulation matters because it suggests that upstream control points, rather than individual downstream genes, could offer the most efficient targets for intervention, and because non-coding RNAs and transcription factors are increasingly seen as versatile biomarkers in neurodegeneration.

Perhaps the most clinically intriguing part of the work is the drug-hub gene interaction analysis, which predicted four existing drug molecules as candidates for Alzheimer’s treatment: Sulindac, Infliximab, Norfloxacin and Gemcitabine. The logic of repurposing is simple and appealing. These compounds already have established safety profiles, known pharmacokinetics and, in some cases, mechanisms that touch inflammation or cell survival, processes central to Alzheimer’s biology. Sulindac is a nonsteroidal anti-inflammatory drug, Infliximab is an antibody targeting tumor necrosis factor, and both speak to the chronic inflammatory component that has shadowed Alzheimer’s research for decades. Gemcitabine and Norfloxacin add further molecular diversity to the candidate pool.

To test these predictions computationally, the researchers performed molecular docking between hub gene products and corresponding active molecules. The docking analysis revealed that Isocryptomerin and Macrophylloside D, natural product-derived compounds, showed strong binding activities to HSP90AA1 and FN1 respectively. Docking scores cannot substitute for wet-lab validation, but favorable computed binding poses provide a rational starting point for medicinal chemists and suggest that these molecules are worth experimental follow-up as potential modulators of the disease’s most central protein hubs.

The study also evaluated the clinical diagnostic potential of the hub genes using receiver operating characteristic curve analysis, a standard method for estimating how well a biomarker separates diseased from healthy samples. Strong diagnostic performance would mean that measuring these genes, or the proteins and microRNAs they encode, could one day help identify Alzheimer’s earlier and more reliably than current cognitive assessments and imaging alone. With disease-modifying therapies finally emerging, the window for meaningful intervention depends critically on early detection, making reliable molecular biomarkers one of the field’s most valuable goals.

As with any computational study, caveats apply. The findings are hypotheses generated from correlation in a modest sample of 39 disease and 8 control brains, and hub genes identified by network centrality do not automatically equal causal drivers. The predicted drug interactions and docking results require validation in cell and animal models before any clinical translation. Nevertheless, the pipeline used here mirrors successful approaches in other diseases and provides a transparent, reproducible map of where future experimental effort should be directed.

Taken together, the work offers new insights into Alzheimer’s pathogenesis by nominating ten hub genes, two microRNAs, two transcription factors and several repurposable compounds as candidate diagnostic and therapeutic markers. It exemplifies a broader trend in biomedical research: as public sequencing archives grow, computational biology can extract biomedical value from data that already exists, at a fraction of the cost of new clinical trials. For a disease that affects tens of millions of people worldwide and still defies a cure, every new molecular lead, however preliminary, matters. The hub genes, microRNAs and transcription factors identified here now join the growing arsenal of targets that researchers worldwide will test in the years ahead.

The dataset at the heart of this study, GSE203206, is one of a growing number of publicly deposited transcriptomic datasets in the Gene Expression Omnibus, a repository maintained by the National Center for Biotechnology Information. Because such archives are openly accessible, any laboratory with computational resources can reanalyze raw sequencing reads, apply its own statistical thresholds, and cross-check published conclusions. This transparency has become a cornerstone of modern genomics, allowing independent teams to validate biomarker candidates across independent cohorts, a step that will be essential before any of the ten hub genes identified here can be considered a robust diagnostic marker.

Several of the hub genes carry biological stories that extend well beyond Alzheimer’s research. YAP1, for instance, is the key effector of the Hippo signaling pathway, a conserved cascade best known for controlling organ size and cell proliferation, and it has been increasingly implicated in neural regeneration and glial responses to injury. CDK7 is a cyclin-dependent kinase that functions as part of the transcriptional machinery, phosphorylating the RNA polymerase II tail and thereby regulating the expression of broad gene programs, which makes its dysregulation potentially consequential for many downstream pathways at once. LSM2 and SKP1 participate in RNA processing and ubiquitin-mediated protein degradation respectively, both processes that intersect with the protein homeostasis failures characteristic of neurodegenerative disease.

The involvement of HSP90AA1 is particularly interesting from a therapeutic standpoint. Heat shock protein 90 acts as a molecular chaperone that stabilizes numerous client proteins, many of which are involved in signaling cascades, and chaperone overload has been proposed as a mechanism by which misfolded and aggregated proteins, such as tau and amyloid-beta species, persist in the aging brain. Inhibitors of this chaperone have been explored in oncology for years, meaning that a substantial body of pharmacological knowledge and chemical tool compounds already exists and could be adapted for neurodegeneration research.

The microRNA findings also fit into a broader literature. MicroRNAs are short non-coding RNAs that each typically regulate dozens to hundreds of messenger RNAs, so a single microRNA shift can ripple across entire pathways. Previous work has documented widespread microRNA alterations in Alzheimer’s brain tissue and even in circulating blood, fueling interest in these molecules as minimally invasive biomarkers detectable in plasma or cerebrospinal fluid. The specific candidates reported here, hsa-mir-545-3p and hsa-miR-548f-5p, now join that expanding catalogue and can be tested for reproducibility in independent sample sets.

Molecular docking itself deserves a note of context. The method models how a small molecule fits into the three-dimensional structure of a target protein and estimates binding strength computationally, often within hours and at negligible cost compared with laboratory screening. Its predictions, however, are only as good as the protein structures and scoring functions used, and docking affinities frequently fail to translate into cellular activity. Natural products such as Isocryptomerin, derived from coniferous plants, and Macrophylloside D offer chemical diversity that synthetic libraries sometimes lack, which is why they attract attention as starting scaffolds. The sensible next steps are biochemical binding assays, neuronal cell models, and ultimately animal studies to determine whether any of these computational leads survives contact with biological reality.

Subject of Research: Identification of Candidate Biomarkers and Signaling Pathways Associated with Alzheimer’s Disease Using Bioinformatics Analysis of Next Generation Sequencing Data and Molecular Docking Studies

Article Title: Identification of Candidate Biomarkers and Signaling Pathways Associated with Alzheimer’s Disease Using Bioinformatics Analysis of Next Generation Sequencing Data and Molecular Docking Studies

Article References: Vastrad, B., Pattanashetti, S., & Vastrad, C. (2026). Identification of Candidate Biomarkers and Signaling Pathways Associated with Alzheimer’s Disease Using Bioinformatics Analysis of Next Generation Sequencing Data and Molecular Docking Studies. Ageing International, 51(3), 34. https://doi.org/10.1007/s12126-026-09665-9

Image Credits: AI Generated

DOI: 10.1007/s12126-026-09665-9

Keywords: Identification, Candidate, Biomarkers, Signaling, Pathways, Associated, Alzheimer, Disease, Bioinformatics, Analysis, Next, Generation