The Most Overlooked Nootropic: A Body-Brain-Blood Investigation of Metformin’s Cognitive Enhancement Potential

Track 3: Cognition: Neuroplasticity, Memory, and Attention

Faye McKenna, PhD

Research Assistant Professor, Department of Radiology

Albert Einstein College of Medicine / Montefiore Medical Center

Director, The Neuroimaging Research Lab

1. Abstract

The nootropic and longevity communities have long sought compounds that enhance mitochondrial function, promote neuroplasticity, and support brain clearance, yet the most clinically validated AMPK activator on the planet has been almost entirely overlooked in this context. Metformin activates the same AMPK-mitochondrial pathway targeted by peptide mimetics such as MOTS-c and humanin^1,2, induces endogenous GLP-1 secretion through gut-brain axis signaling^3,4,penetrates the blood-brain barrier via organic cation transporters⁵,upregulates BDNF, and has demonstrated neuroplasticity-promoting effects in preclinical models^6,7.It has an exceptional safety profile with decades of real-world use at scale, but remarkably full body-to-brain imaging signatures of its CNS and metabolic effects remain entirely uncharacterized at population scale. Epidemiological data suggest that metformin users show slower cognitive decline and reduced dementia incidence^8,9, but the neurobiological mechanisms underlying these effects remain entirely uncharacterized. This proposal leverages the UK Biobank (UKB)¹⁰, the world's largest population neuroimaging dataset, to characterize the full metabolic-to-brain axis in metformin users versus propensity-matched non-users. Using advanced quantitative MRI metrics computed from brain MRI data (DTI-ALPS index, free water fraction, perivascular space volume), pre-derived abdominal MRI phenotypes (liver proton density fat fraction, visceral adipose tissue, muscle fat infiltration), and high-throughput NMR metabolomics (inflammatory, metabolic, and lipid biomarkers)¹¹, we will deliver the first rigorous, population-scale neuroimaging characterization of metformin as a cognitive enhancement compound, testing its relationship to glymphatic brain clearance, neuroinflammation, peripheral body composition, blood metabolomics, and cognitive performance simultaneously. The study is preregistered, fully open-access, and conducted by a specialist quantitative neuroimaging laboratory with established MRI pipelines validated in large aging cohorts. The award will support a graduate research analyst and cloud computing costs for data processing at scale.

2. Introduction

The search for compounds that reliably enhance human cognition, promote neuroplasticity, and protect against brain aging has accelerated dramatically in recent years. Interest has ranged from nootropic peptides and BDNF mimetics to psychoplastogens, mitochondrial enhancers such as urolithin A and the mitochondrial-derived peptides MOTS-c and humanin, and gut-brain axis modulators. The unifying principle across these approaches is mechanistic specificity: compounds with defined molecular targets, plausible CNS mechanisms, and translational potential in humans. Against this landscape, one compound stands out as dramatically undercharacterized from a cognitive enhancement perspective: metformin. Originally developed as an antidiabetic drug, metformin is increasingly recognized as a pleiotropic biology modulator whose CNS-relevant mechanisms map directly onto the targets being pursued by the enhancement field. Nevertheless, it has never been rigorously studied from this angle using advanced neuroimaging.

Metformin’s primary mechanism is activation of AMP-activated protein kinase (AMPK) which is the same pathway targeted by mitochondrial-derived peptide mimetics including MOTS-c and humanin, and by compounds such as urolithin A and NAD+ precursors that have attracted intense interest in the longevity and performance optimization communities¹².Beyond mitochondrial biology, metformin exerts a remarkably broad set of CNS-relevant actions it: 1) penetrates the blood-brain barrier via organic cation transporters (OCT1/OCT2)⁵; 2) modulates neurovascular coupling and cerebrovascular endothelial function, upregulates brain-derived neurotrophic factor (BDNF); 3) promotes hippocampal neurogenesis^6,7; 4) attenuates neuroinflammation through NF-kB suppression, 5) and induces endogenous GLP-1 secretion from enteroendocrine L-cells through gut-brain axis signaling^3,4 (partly recapitulating the CNS effects of exogenous GLP-1 agonists at a fraction of the cost and with a decades-long safety record⁸). This gut-brain-mitochondrial mechanistic profile is further supported by epidemiological outcome data, the Sydney Memory and Ageing Study found that metformin users showed significantly slower global cognition and executive function decline and substantially lower dementia incidence⁹. In the ADNI dataset, metformin use was associated with better cognitive performance and more favorable CSF and imaging biomarkers, particularly with early treatment initiation¹³.

What is entirely missing from this picture is rigorous human neuroimaging data, linked with body and peripheral marker data. No study has examined whether metformin’s detectable changes in brain neuroinflammation, glymphatic clearance, or synaptic plasticity-related imaging markers in humans. This is a striking gap: the enhancement field has invested heavily in compounds with far less mechanistic evidence and no human outcome data, while the most clinically validated AMPK/mitochondrial modulator available sits largely unstudied at the neuroimaging or body imaging level. The glymphatic and neuroinflammation system is particularly relevant here: a glial-dependent perivascular clearance network whose activity can be quantified non-invasively using diffusion MRI, and whose impairment is associated with neurotoxic protein accumulation, neuroinflammation, and cognitive decline¹⁴. Impaired glymphatic function is increasingly recognized as a core feature of neurodegeneration, and MRI-based proxies of glymphatic activity , including the DTI-ALPS index, free water fraction in white matter, and perivascular space volume, have been validated against CSF biomarkers and cognitive outcomes in multiple independent cohorts^15,16. Crucially, prior work has demonstrated that metabolic syndrome, hyperglycemia, and vascular risk factors are associated with reduced ALPS index and elevated free water fraction, suggesting that peripheral metabolic burden directly impairs glymphatic clearance¹⁷. A small study of 20 patients with type 2 diabetes confirmed a reduced ALPS index relative to controls, with impaired water diffusivity correlating with disease severity¹⁸. Whether pharmacological metabolic improvement through metformin reverses these imaging signatures has never been tested.

The UK Biobank provides the only dataset in the world where this question can be answered at scale¹⁰. With complete medication records derived from linked GP prescription data, diffusion MRI available for over 40,000 participants, pre-derived abdominal MRI phenotypes capturing liver fat, visceral adipose tissue, and muscle composition¹⁹, and high-throughput NMR metabolomics in over 118,000 participants the UKB enables a comprehensive end-to-end characterization of metformin as a cognitive enhancement compound from peripheral metabolic remodeling through gut-brain axis signaling, BBB modulation, and glymphatic brain clearance, to cognitive outcomes that is impossible in any other existing dataset. The proposed analyses are preregistered, fully open-access, and designed to generate the first rigorous neuroimaging evidence base for metformin as a next-generation cognitive enhancer.

Figure 1. Metformin metabolic-to-brain enhancement axis and UK Biobank measurement strategy.

Hypotheses

H1. Primary Hypothesis (Brain Changes)

Metformin users will show significantly higher DTI-ALPS, fractional anisotropy and lower free water fraction compared to propensity-matched non-users, reflecting improved glymphatic function, myelin integrity and reduced neuroinflammatory burden^15,17,18, after controlling for age, sex, BMI, HbA1c, diabetes duration, and vascular risk factors.

H2. Secondary Hypothesis (Peripheral Metabolic MRI)

Metformin users will show lower liver proton density fat fraction (PDFF), lower visceral adipose tissue volume, and lower muscle fat infiltration compared to matched controls, after controlling for age, sex, BMI, HbA1c, diabetes duration, and vascular risk factors, confirming peripheral metabolic effects of treatment in this sample¹⁹.

H3. Secondary Hypothesis (Cognition)

Metformin users will show better performance on UKB cognitive battery measures (fluid intelligence, reaction time, trail making, pairs matching), after controlling for age, sex, BMI, HbA1c, diabetes duration, and vascular risk factors, compared to matched controls.

H5. Mediation Hypothesis

The association between metformin use and cognitive performance will be partially mediated by brain health imaging metrics (ALPS index, free water fraction), and this brain mediation pathway will itself be partially mediated by peripheral metabolic MRI phenotypes (liver PDFF, VAT), consistent with a metabolic-to-brain-to-cognition cascade^8,9,15,17.

Figure 2. Three-level serial mediation model. Indirect effects estimated via 5,000-iteration bootstrapping.

3. Methods

3.1 Dataset

All analyses will use data from the UK Biobank (UKB), a prospective population cohort of approximately 500,000 UK adults aged 40-69 at recruitment (2006-2010), with ongoing longitudinal follow-up¹⁰. The imaging sub-cohort includes approximately 100,000 participants who have undergone multimodal MRI scanning at dedicated imaging centers. Brain MRI includes 3T diffusion MRI (100 directions, two shells at b=1000 and b=2000 s/mm2) and T1-weighted structural MRI for registration. Abdominal MRI provides pre-derived image-derived phenotypes (IDPs) for liver PDFF, visceral and subcutaneous adipose tissue, and muscle composition.[19] Medication data are available via linked GP prescription records. NMR metabolomics are available for approximately 118,000 participants¹¹.

3.2 Exposure Definition

Metformin exposure will be defined using UKB GP prescription records (data field 42039). Participants will be classified as metformin users if they have at least two metformin prescriptions on record prior to or within six months of their imaging visit. Non-users will be defined as participants with no metformin prescription records and no other antidiabetic medication, to ensure a clean comparison group. Dose and duration data will be extracted where available for sensitivity analyses.

3.3 Propensity Score Matching

To control for confounding by metabolic indication, metformin users will be propensity-matched 1:2 to non-users on the following covariates: age at imaging, sex, BMI, HbA1c, self-reported type 2 diabetes diagnosis, systolic blood pressure, smoking status, physical activity level, Townsend deprivation index, and APOE4 carrier status. Matching will use nearest-neighbor matching with a caliper of 0.1 standard deviations. Balance will be assessed using standardized mean differences, with a threshold of less than 0.1 indicating acceptable covariate balance. Sensitivity analyses will include inverse probability weighting as an alternative to matching.

3.4 Primary Brain Imaging Outcomes

The following quantitative neuroimaging MRI metrics will be computed from raw UKB diffusion MRI data using established pipelines adapted for the UKB acquisition protocol. All preprocessing will follow a standardized pipeline including eddy current and motion correction (FSL eddy/topup), brain extraction (FSL BET), and registration to MNI space. Quality control will include automated outlier detection and visual inspection of a random subset.

• DTI-ALPS Index: Diffusion tensor image analysis along the perivascular space20, computed as the ratio of diffusivity in the x-direction to mean diffusivity in y and z directions at the level of the lateral ventricles, as a non-invasive proxy for glymphatic activity. Computed bilaterally and averaged.

• Free Water Fraction (FW): Free water-corrected diffusion tensor imaging separating tissue signal from interstitial fluid. Elevated FW in white matter reflects neuroinflammatory burden and extracellular fluid accumulation²¹.

• Tissue Fractional Anisotropy (FAt): Free water-corrected fractional anisotropy reflecting true axonal and myelin integrity after removing the isotropic free water compartment. Reduced FA t indicates white matter microstructural damage independent of extracellular fluid contributions.

3.5 Abdominal MRI Outcomes

Pre-derived abdominal MRI IDPs will be extracted directly from UKB Showcase without additional processing¹⁹. These include liver proton density fat fraction (PDFF), a validated MRI-derived hepatic steatosis measure reproducible at scale; visceral adipose tissue (VAT) volume derived from automated segmentation; abdominal subcutaneous adipose tissue (ASAT) volume; muscle fat infiltration (MFI), a marker of metabolic muscle quality; and thigh muscle volume as a lean mass proxy.

3.6 Blood Biomarkers

UKB NMR metabolomics data (Nightingale Health platform, approximately 325 metabolites in over 118,000 participants)¹¹ will be used to characterize peripheral metabolic and inflammatory signatures associated with metformin exposure. Key metabolite classes of interest include glycoprotein acetyls (GlycA), a validated circulating marker of systemic inflammation; branched chain amino acids, markers of insulin resistance; VLDL and triglyceride subfractions reflecting hepatic lipid export; HDL subfractions and apolipoprotein A1; and glucose and HbA1c equivalent measures. Standard clinical blood biomarkers available in UKB (CRP, HbA1c, total cholesterol, LDL, HDL, triglycerides) will supplement the NMR panel.

3.7 Cognitive Outcomes

UKB cognitive battery measures will serve as behavioral endpoints, including fluid intelligence score, reaction time (mean of 4 trials), trail making test (numeric and alphanumeric path duration), pairs matching (number of errors), and digit span. A composite cognitive score will be derived from standardized z-scores across available measures to provide a summary outcome with improved statistical power.

3.8 Statistical Analysis Plan

All analysis code will be preregistered and openly shared prior to any outcome data access. The following planned analyses will be conducted:

• Primary group comparisons: Linear regression models comparing brain imaging metrics (ALPS, FW, FAt) between metformin users and matched controls, adjusting for residual covariate imbalance post-matching. Effect sizes reported as Cohen's d and standardized beta coefficients with 95% confidence intervals.

• Dose-response analysis: In metformin users with available prescription duration and dose data, linear regression of imaging metrics on cumulative exposure.

• Cognitive outcomes: Linear regression of cognitive composite and individual test scores on metformin exposure in matched sample.

• Mediation analysis: Bootstrap mediation models (5,000 iterations) testing whether glymphatic imaging metrics mediate the metformin-cognition association, and whether peripheral metabolic MRI phenotypes (liver PDFF, VAT) mediate the metformin-brain association, with bootstrapped indirect effects and 95% confidence intervals.

• NMR metabolomics: Metabolome-wide association analysis (MWAS) comparing 325 NMR metabolites between metformin users and controls, with FDR correction (q < 0.05). Metabolites surviving correction will be examined as potential mediators in the brain-cognition pathway.

• Multiple comparisons: FDR correction applied within imaging metric families and within cognitive outcome domains. Primary hypotheses H1 and H5 are specified a priori and will be evaluated at alpha = 0.05.

3.9 Preregistered Acceptance Criteria

Consistent with the registered report framework, the following criteria define a positive result:

1. At least one primary brain metric shows a statistically significant difference between metformin users and matched controls at FDR-corrected q < 0.05, in the expected direction.

2. The association between metformin exposure and at least one cognitive outcome survives FDR correction across the cognitive battery.

3. Bootstrap mediation analysis reveals a significant indirect effect (95% CI excluding zero) of at least one brain imaging metric in the metformin-to-cognition pathway.

Null results will be reported in full with effect size estimates and confidence intervals. A null result for primary H1 will be interpreted as evidence that metformin does not produce detectable glymphatic microstructural differences at the population level in this dataset, which is itself an informative and publishable finding.

3.10 Ethics and Open Science

This is a secondary analysis of de-identified, pre-existing data from the UK Biobank, conducted under an approved UKB access application. No new human subjects research is involved. All analysis code, processed data outputs, and statistical results will be shared openly on the Open Science Framework (OSF) and GitHub. A preprint will be submitted to bioRxiv concurrently with journal submission. The study will be preregistered on ResearchHub prior to any analysis of outcome data.

4. Timeline

Phase	Months	Activities
1. Setup and Preregistration	1-2	Finalize analysis plan; complete ResearchHub preregistration; extract UKB metformin cohort and matched controls; quality-check medication and covariate data
2. dMRI Processing	3-6	Download raw UKB dMRI for metformin subsample and matched controls; run preprocessing pipeline (eddy, topup, BET); compute ALPS, FW, and FAt metrics; QC and outlier removal
3. Statistical Analysis	7-9	Primary group comparisons; dose-response analyses; cognitive outcomes; NMR metabolomics MWAS; mediation models
4. Manuscript and Data Release	10-11	Draft manuscript; prepare open data release; submit preprint to bioRxiv; submit to peer-reviewed journal
5. Buffer and Revision	12	Respond to reviewer comments; finalize open data repository; submit progress report to ResearchHub

5. Budget

Item	Cost (USD)
Graduate student research analyst (part-time, 12 months)	$8,300
Cloud computing: raw dMRI processing at scale (UKB imaging cohort subsample)	$1,500
Open-access data repository and preregistration infrastructure (OSF, Zenodo)	$200
Total	$10,000

The graduate student will be responsible for UKB data extraction, dMRI pipeline execution, statistical analyses, and manuscript preparation under direct supervision of the PI. Cloud computing costs reflect parallelized processing of raw diffusion MRI data for the metformin-exposed subsample and matched controls, which exceeds available institutional cluster allocations. Pre-derived abdominal MRI IDPs and NMR metabolomics will be accessed directly from UKB Showcase at no additional cost.

6. Principal Investigator Background

Dr. Faye McKenna is a Research Assistant Professor in the Department of Radiology at Albert Einstein College of Medicine / Montefiore Medical Center, and Director of The Neuroimaging Research Lab, an independent nonprofit research entity. Her research program centers on advanced quantitative MRI biomarkers, including DTI-ALPS, free water imaging (FWI), IVIM, NODDI, and perivascular space segmentation -- applied to glymphatic biology, neuroinflammation, and cognitive aging across multiple populations. She has active expertise in all imaging metrics proposed here, with established analysis pipelines for ALPS, FWI, and FAt segmentation validated in large aging cohorts^15,16,20. Her collaborative network includes specialists in aging neuroscience, epidemiology, and clinical neuroimaging. This proposal directly extends her laboratory's existing quantitative glymphatic imaging program into the pharmacoepidemiology domain, leveraging a dataset ideally suited to the scientific question.

8. References

1. Lee C, Zeng J, Drew BG, et al. The mitochondrial-derived peptide MOTS-c promotes metabolic homeostasis and reduces obesity and insulin resistance. Cell Metabolism. 2015;21(3):443-454.

2. Zhan Y, Yin H, Yin J. MOTS-c: a mitochondria-derived peptide and new therapeutic target for ischemia-reperfusion injury. Frontiers in Pharmacology. 2023;14:1140815.

3. Duca FA, Cote CD, Rasmussen BA, et al. Metformin activates a duodenal AMPK-dependent pathway to lower hepatic glucose production in rats. Nature Medicine. 2015;21(5):506-511.

4. Forslund SK, Bhatt DL, Sundstrom J, et al. Metformin and the gut microbiome: a systematic review and meta-analysis. Gut Microbes. 2024;16(1):2309183.

5. Sharma S, Bhatt DK, Bhattacharya A, Bhalla A. Permeability of metformin across an in vitro blood-brain barrier model during normoxia and oxygen-glucose deprivation: role of organic cation transporters. Pharmaceutics. 2023;15(5):1357.

6. Wang J, Gallagher D, DeVito LM, et al. Metformin activates an atypical PKC-CBP pathway to promote neurogenesis and enhance spatial memory formation. Cell Stem Cell. 2012;11(1):23-35.

7. Jiang T, Xu S, Shen Y, Li B. Metformin alleviates neurocognitive impairment in aging via activation of AMPK/BDNF/PI3K pathway. Scientific Reports. 2022;12(1):17084.

8. Samaras K, Makkar SR, Crawford JD, et al. Metformin use is associated with slowed cognitive decline and reduced incident dementia in older adults with type 2 diabetes: The Sydney Memory and Ageing Study. Diabetes Care. 2020;43(11):2691-2701.

9. Pomilio C, Perez NG, Calandri I, et al. Diabetic patients treated with metformin during early stages of Alzheimer's disease show a better integral performance: data from the ADNI study. Geroscience. 2022;44(3):1791-1805.

10. Miller KL, Alfaro-Almagro F, Bangerter NK, et al. Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nature Neuroscience. 2016;19(11):1523-1536.

11. Sudlow C, Gallacher J, Allen N, et al. UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLOS Medicine. 2015;12(3):e1001779.

12. Ou Z, Kong X, Sun X, et al. Actions of metformin in the brain: a new perspective of metformin treatments in related neurological disorders. Pharmacological Research. 2022;182:106340.

13. Xiao M, Yan H, Zhang J, et al. Long-term use of metformin is associated with reduced risk of cognitive impairment with alleviation of cerebral small vessel disease burden in patients with type 2 diabetes. Frontiers in Aging Neuroscience. 2021;13:773797.

14. Jessen NA, Munk AS, Lundgaard I, Nedergaard M. The glymphatic system: a beginner's guide. Neurochemical Research. 2015;40(12):2583-2599.

15. Rajanayagam J, Soon CS, Rahimi J, et al. Association of MRI indices of glymphatic system with amyloid deposition and cognition in mild cognitive impairment and Alzheimer disease. Neurology. 2022;99(18):e2136-e2147.

16. Perrotta G, Bhatt H, Zhao L, et al. Perivascular space and white matter hyperintensities in Alzheimer's disease: associations with disease progression and cognitive function. Alzheimer's Research & Therapy. 2025;17:62.

17. Ismaiel AAK, Hari AR, Papandreou D. The role of metformin in neurogenesis and synaptic plasticity in Alzheimer’s disease: a systematic review. Current Neuropharmacology. 2023;21(5):1174-1183.

18. Yang G, Deng N, Liu Y, Gu Y, Yao X. Evaluation of glymphatic system using diffusion MR technique in T2DM cases. Frontiers in Human Neuroscience. 2020;14:300.

19. Linge J, Borga M, West J, et al. Body composition profiling in the UK Biobank imaging study. Obesity. 2018;26(11):1817-1826.

20. Taoka T, Masutani Y, Kawai H, et al. Evaluation of glymphatic system activity with the diffusion MR technique: diffusion tensor image analysis along the perivascular space (DTI-ALPS) in Alzheimer's disease cases. Japanese Journal of Radiology. 2017;35(4):172-178.

21. Pasternak O, Sochen N, Gur Y, Intrator N, Assaf Y. Free water elimination and mapping from diffusion MRI. Magnetic Resonance in Medicine. 2009;62(3):717-730.

The Most Overlooked Nootropic: A Body-Brain-Blood Investigation of Metformin’s Cognitive Enhancement Potential

The Most Overlooked Nootropic: A Body-Brain-Blood Investigation of Metformin’s Cognitive Enhancement Potential

Funding Progress

Nonprofit Recipient

Peer Reviews

Topics

DOI