Beyond Memory Loss: A Metabolic View of Cognitive Decline and Dementia

Published on
November 15, 2026

¹Department of Neurology, Institute of Medical Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh, India

²Department of Advanced Spectroscopy and Imaging, Centre of Biomedical Research (CBMR), Lucknow, Uttar Pradesh, India

Areas of Expertise
Clinical neurology, Dementia, Cognitive decline, Clinical biomarkers

Memory loss is one of the best-known signs of dementia, but it is only the visible part of a much more complex process. Cognitive decline can involve changes in attention, language, planning and everyday function, and different people may reach similar symptoms through different biological pathways. Alzheimer’s disease and vascular dementia are not identical conditions, yet both can be accompanied by disturbances in energy use, amino-acid balance, inflammation and lipid metabolism. Understanding these changes may help researchers develop accessible measures that complement neurological examination, cognitive testing and brain imaging.

The brain has a high and continuous demand for energy. Neurons depend on carefully regulated supplies of mainly glucose and other fuels, functioning mitochondria, and a balanced environment of amino acids and lipids. Inflammation and vascular or metabolic comorbidities can influence these systems. Metabolomics—the systematic study of small molecules in biological samples—offers a way to examine the combined biochemical consequences of genes, lifestyle, ageing and disease. Blood is particularly attractive because collection is relatively straightforward and can be repeated over time. However, a blood profile reflects the whole body; it should not be treated as a direct readout of brain chemistry.

Our work using proton nuclear magnetic resonance (¹H NMR) spectroscopy has examined circulating metabolic features across different stages of cognitive impairment. In a 2021 study of a North Indian population, plasma from a subset of participants with mild cognitive impairment and healthy controls was profiled using high-field ¹H NMR. Lysine, glycine and glutamine emerged as a candidate three-metabolite panel. The study also considered diffusion tensor imaging and cognitive assessment, placing metabolic findings alongside clinical and brain-imaging information. These results supported further investigation of blood-based metabolic markers, but the sample analyzed by NMR was modest and the findings require validation in independent populations.

In our more recent study published in Biophysical Chemistry, serum samples from 40 participants with cognitive decline and 54 cognitively normal controls were analysed using 800 MHz ¹H NMR spectroscopy. Thirty-three metabolites and selected metabolic ratios were quantified. The cognitive-decline group showed higher glucose, pyruvate, acetate, creatine and mannose, alongside lower levels of several features including lactate, leucine, valine and histidine. The glutamine-to-glucose ratio (QGR) and alanine-to-glucose ratio (AGR) were also lower. AGR was positively associated with Mini-Mental State Examination scores, while AGR and QGR declined with increasing severity in the study cohort. QGR, AGR and the histidine-to-tyrosine ratio showed promising discrimination within that cohort.

Small metabolites are only one part of the circulating biochemical picture. In a separate study currently under review at Clinica Chimica Acta, our team has explored complementary NMR experiments designed to capture both low-molecular-weight metabolites and broader glycoprotein- and lipoprotein-associated signals. The submitted analysis reports higher N-acetyl-glycoprotein signals and supramolecular phospholipid-composite signals in the cognitive-decline group, together with changes in glucose, mannose, amino acids and metabolic ratios. These findings are preliminary until peer review and independent validation are complete. Nonetheless, they illustrate the value of asking whether several biochemical layers—rather than a single metabolite—can provide a more informative picture of systemic changes accompanying cognitive decline.

Taken together, the studies support a view of cognitive decline as a condition that may be accompanied by interconnected metabolic changes, rather than a process understood only through memory symptoms. Ratios such as QGR and AGR may be useful because they express relationships between metabolites, not simply the level of one molecule. Glycoprotein-associated signals may add information about systemic inflammatory activity, while lipoprotein-related signals may reflect changes in lipid transport and composition. These are biological interpretations, not proof that a specific pathway caused cognitive impairment. Cross-sectional associations cannot establish whether metabolic changes precede cognitive decline, result from it, or arise partly from other conditions.

Several hurdles must be cleared before these observations can become clinical tests. First, findings need replication in larger, independent and diverse cohorts, including participants with different educational backgrounds, diets, vascular risks and comorbidities. Second, longitudinal studies should determine whether metabolic profiles change as cognition changes and whether they can predict conversion from mild cognitive impairment to dementia. Third, studies need careful control of factors such as age, sex, fasting status, medication, diabetes, kidney and liver function, sample handling and storage. Dementia is heterogeneous, and a marker that performs well in one group may not generalize to another.

Analytical and statistical standards matter equally. NMR offers reproducible profiling and can measure multiple metabolites in one experiment, but some spectral signals overlap and assignments may remain putative unless confirmed with additional methods. Multivariate models can appear more accurate than they truly are if data are over-fitted or validation is incomplete. Candidate markers should therefore be evaluated with transparent preprocessing, correction for multiple testing, suitable internal validation and external testing. Ultimately, their value must be compared with established clinical assessments and available blood, cerebrospinal-fluid or imaging biomarkers—not merely with healthy controls.

The next step is not to promote a metabolic panel as a stand-alone diagnostic test, but to test whether it adds useful information to existing care. Prospective multicenter studies could combine standardized clinical phenotyping, cognitive scores, brain imaging and established biomarkers with serum or plasma NMR profiles. Where feasible, metabolomics could be integrated with lipidomics, proteomics and other data layers, using machine-learning models that are independently validated and interpretable. Researchers should also examine whether metabolic patterns differ by dementia subtype and whether they reflect modifiable contributors such as vascular risk or nutritional status. The conceptual framework is presented in Figure1.

A diagnosis of dementia must remain grounded in a careful clinical evaluation. A blood test based on metabolomics is not currently a substitute for that process. Yet the ability to measure many metabolic features from a small sample offers a promising research route to understanding the biology that accompanies cognitive decline. Our studies—from mild cognitive impairment to broader serum metabolite profiles and exploratory glycoprotein/lipoprotein signals—add to this developing evidence base. The most important task now is rigorous validation and collaboration between neurologists, analytical scientists, statisticians and clinical researchers. Looking beyond memory loss does not mean looking away from patients; it means seeking a fuller biological understanding that may, in time, support more precise assessment and monitoring.

References

Kumar U, Kumar A, Singh S, Arya P, Singh SK, Chaurasia RN, Singh A, Kumar D. An elaborative NMR based plasma metabolomics study revealed metabolic derangements in patients with mild cognitive impairment: a study on north Indian population. Metabolic brain disease. 2021 Jun;36(5):957-68.

Article DOI

Trivedi R, Ahmad S, Chaurasia RN, Kumar D. NMR-based serum metabolomics of dementia patients and search for metabolic biomarkers associated with cognitive decline. Biophysical Chemistry. 2026 Sep 23:107715.

Article DOI

Science Factors.

Inside a Legume Root Nodule: Building the Network for Nitrogen Fixation and Sustainable Agriculture

Dr. Senjuti Sinharoy
0
What is the importance of nitrogen fixation? Nitrogen is one of the essential nutrients required for plant growth. Although nearly 78% of the Earth's atmosphere...

Beyond Potency: What Defines the Next Generation Selective HDAC8 Inhibitor?

Dr. Kondapalli Venkata Gowri. Chandra Sekhar
0
Histone deacetylases (HDACs) regulate the acetylation state of histone and non-histone proteins and therefore affect transcription, chromatin organization, cell cycle control and various signalling...

Overcoming Chemotherapy Resistance by Targeting Cancer’s Cellular Defences

Dr. N. Rajendra Prasad
0
Why does chemotherapy stop working? One of the greatest challenges in cancer treatment is not simply finding drugs that kill cancer cells, but understanding...

Understanding Antibiotic Resistance Through the Structure and Function of Bacterial Enzymes

Dr. Immanuel Dhanasingh
0
When life-saving medicines stop working Imagine having an infection that was once easily treated with antibiotics, but the same medicines no longer work. This is...

Strengthening Infection Prevention and Responsible Antibiotic Use through Communication and Teamwork

Dr Anup R Warrier
0
Consider a patient recovering after abdominal surgery. On the second postoperative day, a nurse notes that the urinary catheter no longer has a clear...

From engineered tissues to organ-on-a-chip: How bioengineering is transforming biomedical research and regenerative medicine

Prof. Sanjeev Kumar Mahto
0
Biomedical engineering is increasingly at the interface of living and engineered systems. Many biological questions are still difficult to answer using standard two-dimensional (2D)...

Beneath the Surface: Understanding Rhizosphere Interactions and Plant Development for Sustainable Agriculture

Dr. Mukesh Jain
0
Beneath every plant lies the rhizosphere, the narrow zone of soil surrounding the roots where plants and microbes constantly exchange chemical signals. Plants release...

The Cell’s Nuclear RNA Quality Control: Finding Faulty Messages and Fine-Tuning Gene Activity

Dr. Biswadip Das
0
Error-Prone mRNA Biogenesis In all eukaryotes, mRNA biogenesis consists of the transcription of the protein-coding genes by RNA polymerase II, capping of the primary transcript...