Abstract

Serum neurofilament light chain (sNfL) is a biomarker of neuroaxonal damage with growing prognostic value in neurological and systemic diseases. However, its role in older adults, especially in relation to cognitive and kidney function, is not fully understood.

We analyzed data from 1,052 hospitalized patients aged ≥65 years (Report-AGE project, INRCA, Ancona) and 915 home-dwelling individuals from Bologna and Milan cohorts, who underwent geriatric assessment and blood tests, including sNfL. We investigated sNfL trajectories by age, sex, cognition, kidney function, and their association with 10-year mortality.

sNfL levels increased with age across all groups, with the steepest rise in hospitalized patients. Among them, non-survivors had significantly higher sNfL levels than survivors, and elevated sNfL was linked to comorbidities, frailty, and lower eGFR.

Hospitalized patients were stratified by cognitive status and sNfL levels (<40, 40–66, ≥67 pg/mL). Higher sNfL correlated with older age, cognitive impairment, reduced eGFR, and higher mortality. Those with sNfL ≥67 pg/mL had double the mortality risk compared to those with sNfL <40 pg/mL. This association remained significant after adjusting for demographic, clinical, and laboratory variables, and was strongest in patients without cognitive impairment. Spline models showed sNfL’s predictive value weakened in patients with severe renal dysfunction (eGFR <30 mL/min), but remained consistent in cognitively impaired individuals.

In conclusion, sNfL is a strong predictor of long-term mortality in older hospitalized adults. Its prognostic significance is influenced by kidney function and cognitive status, underlining the need for personalized interpretation in geriatric care.

Extended Abstract

Background:

Serum neurofilament light chain (sNfL), a marker of neuroaxonal injury, is increasingly recognized for its role in aging and neurological disorders. However, its prognostic significance in hospitalized older adults remains unclear, especially in relation to cognitive impairment and renal function. Given the complex interplay of aging, comorbidity, and cognitive decline in this population, sNfL could serve as a valuable biomarker for mortality risk stratification.

Methods:

We analyzed sNfL levels in 1,052 older adults (≥65 years, median age 84) admitted to geriatric wards and compared them to age-matched community-dwelling cohorts. Particularly, a cohort of 465 home-dwelling individuals (≥65 years, median age 87) from Bologna and 450 home-dwelling individuals (≥65 years, median age 80) from Milan.

sNfL levels were measured by automated enzyme linked immunoassays kits designed for the Simple Plex Ella microfluidic platform. Routine laboratory techniques were employed to assess laboratory biomarkers. Associations with age, sex, cognitive impairment, frailty, kidney function (eGFR), comorbidity, and long-term (10-year) mortality were evaluated.

Results:

sNfL increased with age across all cohorts but was markedly higher and more variable among hospitalized patients, compared to both home-dwelling patients, which showed similar trends.

In hospitalized patients, elevated sNfL level was linked to cognitive impairment, reduced kidney function, and higher frailty. Deceased individuals had significantly higher sNfL levels than survivors (81.1 pg/mL vs. 39.3 pg/mL). Then, we stratified patients by cognitive impairment and sNfL levels (<40, 40–66, ≥67 pg/mL. Patients with sNfL ≥67 pg/mL had a twofold increased mortality risk compared to those with sNfL <40 pg/mL.. Cox models confirmed a robust association between elevated sNfL and increased long-term mortality, independent of other clinical and laboratory variables.

The association was strongest in patients without cognitive impairment. Spline models revealed that the predictive power of sNfL diminished in patients with advanced kidney dysfunction (eGFR <30 mL/min), suggesting altered clearance, but remained stable across eGFR levels in cognitively impaired individuals.

Implications:

sNfL is a robust biomarker for long-term mortality in older hospitalized adults, capturing both neurological and systemic decline. Its prognostic value is modulated by cognitive status and renal function, highlighting the need for integrative biomarker interpretation in geriatric risk stratification.

Introduction: Plasma phospho-tau 217 (pTau217) is a biomarker for Alzheimer’s disease (AD) pathology, reflecting amyloid (Aβ) and tau burden, but its role in Parkinson disease (PD) and 4-repeat(4R)-tauopathies remains incompletely understood. We measured plasma pTau217 across the cognitive spectrum of Lewy body diseases (PD, Dementia with Lewy bodies [DLB]) and in 4R-tauopathies, comparing these groups to cognitively unimpaired (CU) and mild cognitive impairment (MCI) individuals.

Methods: Participants included 18 cognitively normal PD (PD-NC), 32 PD with MCI, and 7 PD with dementia (PDD), alongside 4 DLB patients, grouped as PDD/DLB. The 4R-tauopathy group included 28 Progressive Supranuclear Palsy (PSP) and 4 corticobasal syndrome (CBS) patients, compared to 51 CU and 26 MCI individuals. Ptau217 was measured using the fully automated Lumipulse platform, with values adjusted for creatinine levels. Further, the presence of AD-pathology was defined using a validated cut-off based on Aβ-PET.

Results: PTau217 levels were significantly lower in PD-NC and CU individuals compared to those with greater cognitive impairment (PD-MCI, PDD/DLB, and PSP/CBS), and MCI individuals. AD co-pathology was identified in 28% of PDD/DLB and PSP/CBS patients, 16% of PD-MCI, and none of PD-NC. MCI showed the highest pTau217 positivity (35%), while 8% of CU individuals were positive despite normal cognition. In PD, pTau217 negatively correlated with cognitive performance, as assessed by Montreal Cognitive Assessment (MoCA: r² = –0.38, p = 0.004) and Mini-Mental State Examination (MMSE: r² = –0.37, p = 0.006).

Discussion: Plasma pTau217 levels serve as a scalable, non-invasive marker of AD-pathology across Lewy body diseases, PSP/CBS, and MCI/CU populations. AD co-pathology independently contributes to cognitive deficits in PD, but not in PSP/CBS.

Background and Objective: Alzheimer’s disease (AD) is characterized by a long preclinical phase during which neuropathological changes accumulate in the absence of overt clinical symptoms. Identifying individuals at risk during this early phase offers a unique opportunity for intervention aimed at delaying symptom onset and preserving functional independence. In this context, the Brain Health Service (BHS) of San Raffaele Hospital in Milan was established as a specialized clinical and research initiative focused on early identification and risk stratification for preclinical AD. The present study aimed to comprehensively assess the clinical/neuropsychological, neurophysiological, neuroimaging, genetic, and fluid biomarker profiles of individuals accessing the BHS.

Materials: From February 2024 to April 2025, fifty-two individuals accessed the BHS due to a positive family history of dementia, self-perceived cognitive changes, and/or a proactive interest in maintaining cognitive health. All participants underwent a standardized protocol including clinical/neurological and neuropsychological assessments, magnetic resonance imaging (MRI), electroencephalography (EEG), optical coherence tomography (OCT), and genotyping for the apolipoprotein E (APOE) gene. Participants also consented to blood sampling for plasma biomarker analysis as part of an ongoing research project. Based on plasma phosphorylated tau-217 (p-tau217) levels (cutoff = 0.1325 pg/mL), individuals were classified as AD biomarker-positive (AD+, n=22) or AD biomarker-negative (AD–, n=30). Clinically, they were further categorized as cognitively normal (CN; CN+ n=6, CN– n=9), subjective cognitive decline (SCD; SCD+ n=3, SCD– n=15), or mild cognitive impairment (MCI; MCI+ n=13, MCI– n=6).

Methods: Clinical evaluation included medical history and assessment of lifestyle and vascular risk factors. Neuropsychological testing covered domains such as memory, language, attention, executive function, and visuospatial abilities. MRI scans were evaluated using standardized semiquantitative rating scales for gray matter (GM) atrophy and white matter (WM) lesion load. EEG analyses focused on spectral metrics, particularly power distribution across frequency bands, while OCT assessed average retinal layer thickness bilaterally. Statistical comparisons between groups were conducted using ranked ANCOVAs, Kruskal-Wallis tests, or chi-square tests, as appropriate.

Results: AD+ participants were older than their AD– counterparts, reported higher alcohol consumption, and engaged in lower levels of physical activity. No significant group differences emerged for sex, education level, or APOE genotype. Cognitively, the AD+ group performed significantly worse on tasks assessing verbal learning, fluency, visuospatial skills, attention, and executive functioning. Within clinical subgroups, MCI+ individuals showed slower performance on timed tasks compared to MCI–, and CN+ exhibited reduced semantic fluency compared to SCD–. MRI revealed that AD+ individuals had lower GM volume in the left medial temporal lobe, left fusiform gyrus, right precuneus, and right anterior temporal gyrus, along with greater WM lesion burden in both periventricular and deep regions. EEG findings in the AD+ group indicated a shift in power spectral density toward lower frequencies (delta band: 1–4 Hz), predominantly in temporal and frontal regions. OCT analysis showed increased global retinal nerve fiber layer thickness in AD+ participants.

Main Implications: This study supports the clinical feasibility and utility of incorporating plasma biomarkers, such as p-tau217, into early risk stratification protocols offered by a BHS. Biomarker-positive individuals demonstrated distinct cognitive, neuroimaging, electrophysiological, and retinal profiles consistent with early AD pathology. These findings underscore the importance of early detection for implementing timely lifestyle interventions that could potentially delay clinical progression and maintain autonomy. The BHS model represents a viable framework for identifying at-risk individuals and integrating personalized preventive strategies into routine care.

Age-related macular degeneration (AMD) is a leading cause of irreversible vision loss among the elderly, characterized by the degeneration of retinal pigment epithelial (RPE) cells. Apoptosis and cellular senescence are critical contributors to RPE dysfunction in AMD. Despite extensive studies, the molecular mechanisms underlying these processes remain poorly understood. Shp1, a protein tyrosine phosphatase involved in regulating cell survival, apoptosis and inflammation, has recently emerged as a potential modulator of these pathways.

To investigate the role of Shp1 in RPE cell death, we generated Shp1 knockout ARPE-19 cells using CRISPR-Cas9 technology. When exposed to genotoxic agents, Shp1-KO cells exhibited reduced apoptosis compared to wild-type cells, as indicated by decreased levels of cleaved caspase-3, phospho-Akt, and p53. These results suggest that Shp1 facilitates apoptotic signaling in RPE cells under stress conditions, possibly via modulation of the Akt/p53 pathway.

Given the relevance of cellular senescence in AMD progression, we further explored Shp1 involvement in this process. Senescence was induced in ARPE-19 cells using hydrogen peroxide, which led to changes in Shp1 phosphorylation, likely reflecting alterations in its catalytic activity. Additionally, we found that targeting Shp1 activity with its chemical inhibitor resulted in the modulation of specific senescence markers, such as cell cycle progression formation of nuclear γH2ax foci and the SASP phenotype.

Collectively, our findings highlight a dual role for Shp1 in promoting apoptosis and regulating cellular senescence in RPE cells, supporting its potential as a therapeutic target in age-related macular degeneration.