Blood Metabolite Test Detects Early Cognitive Decline

Timely identification of individuals at risk of dementia remains difficult because symptoms commonly appear only after significant neurodegeneration. Accessible screening tools that flag subtle cognitive changes before clinical diagnosis could improve monitoring and enable earlier intervention. Blood-based assays tied to upstream biology are a practical avenue for risk stratification in primary and specialty care. A new study shows that metabolite profiling of fasting blood, linked to the gut–brain axis, can distinguish early cognitive decline with notable accuracy. 

University of East Anglia (Norwich, UK) scientists developed an artificial intelligence (AI) machine-learning model built from six blood metabolites to classify cognitive status. The approach analyzes 33 gut and diet derived molecules measured from fasting blood alongside stool profiling to characterize gut bacterial communities. Investigators state that the metabolic signals reflect shifts in specific gut bacteria, supporting the role of the gut–brain axis in early memory changes.

The study evaluated 150 adults aged 50 years and older, spanning healthy participants, individuals with subjective memory concerns who tested normally, and people with mild cognitive impairment (MCI). All participants provided fasting blood samples for targeted metabolite assays and stool samples for microbiome mapping. Using advanced computer modeling and AI, researchers assessed whether combinations of these circulating molecules could separate the groups. 

A model based on six metabolites classified participants across the three groups with 79% accuracy and distinguished healthy adults from those with MCI at over 80% accuracy. The work, published in Gut Microbes on March 27, 2026, involved researchers from the University of East Anglia and Queen Mary University of London. Investigators suggest the approach could help identify people at higher risk of memory decline years before dementia is typically diagnosed, and highlight the microbiome as a potential target for protecting brain health. 

“This study is exciting because it links advanced data analysis with a clinically realistic goal of finding a simple and acceptable way to detect risk earlier. It could add to a new set of methods that help bridge the gap between discovery science and practical tools for prevention,” said Dr. Saber Sami, Alzheimer’s Research UK UEA lead and co-author. 

Related Links
University of East Anglia