Preprints.org

Functional-vs-Cognitive Decline Heterogeneity in Alzheimer’s Disease: LLM-Curation of Decades of Clinical Notes and GLP1RA Trial Design

At a glance

No. Among patients with both confirmed declines, cognition came first in 51.8%, function came first in 34.2%, and both were confirmed on the same date in 14.1%.

The research question

Do cognitive and functional decline unfold in the same order for everyone with Alzheimer’s disease?

Overview

LLM-curated clinical notes reconstructed decades of cognitive and functional assessments, revealing distinct decline sequences that conventional trial averages can obscure. Trials built around an average trajectory may miss clinically different disease phases. Recovering standardized assessments from notes creates a path toward trajectory-aware enrollment and endpoints. Trials built around an average trajectory may miss clinically different disease phases. Recovering standardized assessments from notes creates a path toward trajectory-aware enrollment and endpoints.

Study design & methods

The team identified 131,824 patients with structured Alzheimer’s diagnoses across a federated EHR network. LLM workflows extracted instruments including MMSE, MoCA, FAQ and ADCS activities-of-daily-living scores. Decline direction was instrument-specific: higher FAQ is worse, while lower MMSE, MoCA and ADCS activities-of-daily-living scores are worse.

Cognitive-first and functional-first groups both appeared

Of 2,169 patients with both confirmed declines, 1,123 had cognitive decline first, 741 had functional decline first and 305 had both confirmed on the same date. APOE ε3/ε3 was more frequent in the group whose functional decline preceded cognitive decline by at least twelve months, an exploratory association that needs replication.

Limitations & interpretation

The results suggest that functional reserve and the order of decline may help define clinically distinct Alzheimer’s populations for future research. The authors frame the study as a stratification hypothesis rather than evidence that semaglutide treats Alzheimer’s disease. Preprint; not peer reviewed.

Publication type and venue follow the original nference listing. Read the original paper for full methods, analyses, and interpretation. View source listing.

Keep exploring.

All publications
NeurosciencePeer-Reviewed Journal Publication

Somatosensory and peripheral neuropathic outcomes after semaglutide or tirzepatide initiation compared with other anti-diabetic medications

In this observational study we found no evidence of a population-level excess of documented somatosensory or peripheral neuropathic phenotypes after semaglutide or tirzepatide initiation; the modestly lower incidence was within the range of systematic error indicated by the negative controls.

Biology Methods and Protocols
November 4, 2026
NeurosciencePeer-Reviewed Journal Publication

GLP-1–based incretin therapy is associated with lower incident Alzheimer’s disease and cardiorenal events in adults with documented neuropsychiatric, cognitive, or sensory risk

In adults aged 50 or older with documented Alzheimer’s risk factors, a matched target-trial emulation found lower recorded Alzheimer’s incidence after GLP-1–based incretin initiation than after non-GLP-1 antidiabetic treatment. The observational design cannot establish prevention or causality.

Biology Methods and Protocols
August 3, 2026
Your next question could change everything.

What will you
discover next?

Let’s find out
Ideas are connected. Your next discovery starts here.