Reduction of glycation stress as a geroscience intervention: protocol for a pilot RCT in postmenopausal women

Abstract

Advanced glycation end products (AGEs) drive metabolic dysfunction, inflammation, and age-related disease, and their accumulation accelerates after menopause. Preclinical studies show that GLYLO, a five-compound glycation-lowering formulation (alpha-lipoic acid, nicotinamide, pyridoxine, benfotiamine, and piperine), reduces AGE burden and improves metabolic health, but its translational relevance in humans is unknown. The Glycation Reduction and Aging: a Clinical Evaluation (GRACE) trial is a randomized, double-blind, placebo-controlled pilot study testing whether six months of GLYLO supplementation lowers circulating AGEs and methylglyoxal (MGO) in postmenopausal women (45–65 years) with elevated adiposity (BMI ≥ 25 kg/m² or waist circumference ≥88 cm) and elevated HbA1c (5.5–6.4%). Secondary and exploratory endpoints include HOMA-IR, body composition, and reproductive hormones, cognitive and physical function, retinal aging, and systemic inflammation. GRACE is designed to address a critical gap in geroscience by evaluating whether targeting glycation can mitigate early metabolic and functional decline in postmenopausal women. ClinicalTrials.gov identifier NCT06813261.

Introduction

Geroscience, or the study of aging biology, postulates that aging is driven by conserved biological processes, referred to as the “hallmarks of aging” that operate across multiple systems. These hallmarks include genomic instability, loss of proteostasis, mitochondrial dysfunction, deregulated nutrient sensing, cellular senescence, and altered intercellular communication. Interventions that target one or several of these hallmarks are therefore hypothesized to simultaneously delay or ameliorate multiple age-related diseases. In women, the onset of menopause accelerates the accumulation of multiple “hallmarks” of aging, such as metabolic and inflammatory changes, coinciding with increased vulnerability to age-related disease. Postmenopausal women represent a growing population at heightened risk for metabolic dysfunction, hormonal dysregulation, and accelerated physiological aging. By the year 2030, over 1.2 billion women worldwide will be aged 50 years or older¹, with a substantial proportion transitioning through menopause into postmenopause. This demographic shift is not confined to high-income countries but reflects a global trend across continents, contributing to mounting healthcare demands and substantial socio-economic pressure worldwide. Most interventions used in post menopause target a single symptom, such as hot flashes or mood changes (e.g., selective serotonin reuptake inhibitors, neurokinin antagonists), rather than addressing underlying aging biology. Multimodal geroscience interventions that target fundamental mechanisms of aging, therefore, hold promise to slow menopause-induced biological aging at a stage that could be highly responsive to intervention and reduce downstream disease risk.

The transition to menopause is characterized by a marked decline in endogenous ovarian hormone production and concomitant increase in anterior-pituitary release of follicle-stimulating hormone (FSH), both of which are strongly associated with increased adiposity in women^2–4 and long-term health issues^5,6. This hormonal shift contributes to the accumulation of visceral fat and loss of lean mass, key features of the postmenopausal phenotype^7,8. Abdominal fat accumulation is both a hallmark and contributing factor of metabolic syndrome⁹, closely linked to insulin resistance¹⁰, dyslipidemia^11,12, chronic inflammation¹³, and increased cardiometabolic risk¹⁴. These metabolic alterations are compounded by the age- and menopause-related decline in lean muscle mass⁸, further elevating the risk of age-related diseases such as type 2 diabetes¹⁵, cardiovascular disease¹⁶, sarcopenia¹⁷, and neurodegenerative conditions^18,19, highlighting the population’s vulnerability to aging-related dysfunction. Despite the growing size of this vulnerable population and the rising prevalence of central obesity across global settings²⁰, effective and well-tolerated interventions specifically tailored to the biology of postmenopausal women remain limited.

Hormone replacement therapy (HRT) remains the primary intervention for postmenopausal women and effectively alleviates vasomotor symptoms^21,22. However, it is not universally accepted or tolerated and may carry potential risks such as coronary heart disease, stroke, and breast cancer²³, depending on therapy type, timing, and duration. Moreover, it does not broadly address the underlying biological mechanisms of aging. Lifestyle and dietary approaches are often recommended, but few have been rigorously studied in this population. This further highlights the need for safe, accessible, and geroscience-grounded strategies that address the multifactorial drivers of accelerated aging in women.

One intervenable geroscience mechanism is glycation stress, characterized by the accumulation of advanced glycation end-products (AGEs). The formation of AGEs is accelerated under hyperglycemia²⁴ and during aging²⁵, enhancing the risk of diseases such as diabetes²⁶ and Alzheimer’s disease²⁷. Elevated AGEs contribute to adipogenesis²⁸, dyslipidemia^29,30, inflammation^31,32, vascular dysfunction³³, impaired insulin signaling³⁴, elevated blood pressure³⁵, and arterial stiffness³⁶, processes highly relevant to the postmenopausal metabolic phenotype. Importantly, AGEs accumulate in estrogen-deprived states^37,38, suggesting a mechanistic link between menopause and glycation-associated tissue damage that drives various age-related diseases^39,40. Despite this, no clinical interventions have been specifically designed to target glycation in postmenopausal women. In broader populations, only a small number of clinical trials have evaluated glycation-modulating compounds, including those testing benfotiamine and a combination of related AGE inhibitors^41,42. This represents a substantial therapeutic gap and underscores the need to test novel glycation-lowering strategies in a population particularly vulnerable to AGE accumulation, such as postmenopausal women.

To address this therapeutic opportunity, we developed GLYLO, a glycation-lowering dietary supplement, through rigorous screening of thousands of natural compounds⁴³. GLYLO is formulated with five ingredients with FDA-designated GRAS (Generally Recognized As Safe) status: alpha-lipoic acid, nicotinamide, pyridoxine, benfotiamine, and piperine. Together, these compounds act through additive and complementary mechanisms to enhance endogenous detoxification of methylglyoxal (MGO), a reactive α-dicarbonyl produced from glycolytic intermediates and a key precursor to AGEs, thereby reducing downstream glycation stress. Based on our recent mechanistic work, individual components provided partial protection against MGO-induced cellular toxicity, while the combined formulation conferred greater protection than any single component alone, even when each component was present at one-fifth of the total concentration. Consistent with these findings, the formulation upregulates cellular defenses against MGO and minimizes MGO-induced damage, likely by supporting glyoxalase-related detoxification pathways rather than directly neutralizing MGO. Further, we have demonstrated that GLYLO supplementation significantly reduced MGO levels, improved insulin sensitivity, reduced insulin resistance, decreased body weight and central adiposity, enhanced neuromuscular performance, and extended lifespan in mice⁴⁴. However, it is unknown whether these effects translate to humans. As postmenopausal women face increased metabolic risk, they are more likely than younger adults to experience measurable benefits from GLYLO on glycation stress and related metabolic outcomes, making them an ideal population in which to test its translational potential.

To directly address this gap, we designed the GRACE study (Glycation Reduction and Aging, a Clinical Evaluation), a randomized, double-blind, placebo-controlled pilot clinical trial in postmenopausal women with overweight or obesity. The primary objective is to determine whether 6 months of GLYLO supplementation reduces circulating AGEs and MGO levels. Secondary and exploratory outcomes include changes in metabolic and hormonal parameters, functional measures related to aging, and novel, non-invasive biomarkers of aging, such as a novel retinal imaging-based aging clock (eyeAge)⁴⁵. Findings from this trial will help guide the design of future larger-scale studies and may support the development of non-pharmacological strategies to promote healthy aging and extend healthspan in women. Figure 1 illustrates the conceptual framework underlying the GRACE study based on promising preclinical findings demonstrating the beneficial effects of GLYLO. These effects are proposed to act through mechanisms that alleviate glycation stress and its physiological consequences. The GRACE study is designed to evaluate whether GLYLO effectively lowers MGO and AGEs in humans and seeks a signal of improvement in healthspan and functional outcomes in postmenopausal women.

Fig. 1. Conceptual framework of the GRACE study.

This framework illustrates the cellular and hormonal changes associated with the postmenopausal phase, contributing to metabolic dysfunction and accelerated aging. Based on preclinical findings, the glycation-lowering intervention, GLYLO, is hypothesized to counteract these age-related effects through multiple pathways, ultimately aiming to improve healthspan and functional outcomes in postmenopausal women.

Results

Trial status

The GRACE study is an ongoing randomized, double-blind, placebo-controlled pilot trial. Recruitment began on April 1, 2025. As of December 2025, enrollment and follow-up assessments are in progress. No interim efficacy analyses have been conducted. Final analyses will be performed after completion of the intervention phase.

Discussion

The GRACE (Glycation Reduction and Aging: a Clinical Evaluation) trial addresses a critical and underexplored question in geroscience: whether targeting glycation stress, a fundamental and modifiable mechanism of aging, can improve metabolic, hormonal, and functional health in postmenopausal women. Menopause is accompanied by profound hormonal changes⁴⁶ that accelerate multiple biological aging processes, particularly metabolic dysfunction¹⁴, central adiposity⁵³, and loss of lean mass⁸. Despite the high prevalence of these risk factors, few interventions in this population are explicitly grounded in the biology of aging, and even fewer adopt a multimodal approach. GRACE is designed to close this gap by translating strong preclinical evidence for GLYLO⁴⁴, a combination of five GRAS compounds with synergistic glycation-lowering activity, into a well-controlled human trial.

A key methodological strength of GRACE is its comprehensive, multimodal outcome framework, which pairs direct, proximal biomarkers of glycation stress (serum MGO, advanced glycation end products (AGEs), skin autofluorescence) with secondary, more distal measures of metabolic health, body composition, and hormonal status. The inclusion of diverse exploratory endpoints, spanning cognitive and physical performance, quality of life (QoL), and non-invasive aging biomarkers such as retinal age gap⁴⁵, enables an integrated assessment of whether changes in a targeted molecular mechanism translate into broader physiological and functional benefits. This design reflects the core tenet of the geroscience hypothesis: that intervening in fundamental aging biology may yield multi-system gains.

Several design choices enhance the translational value of the study. The selected cohort, postmenopausal women aged 45–65 years with elevated central adiposity or HbA1c, represents a life stage and risk profile in which metabolic decline is emerging but potentially reversible. By focusing on a well-defined, at-risk group, GRACE directly addresses the vulnerability of midlife women to insulin resistance, adiposity, and functional decline, which are accelerated by hormonal shifts. The intervention uses a commercially available, well-tolerated supplement, allowing for a relatively rapid path from proof-of-concept to wider dissemination should efficacy be demonstrated. Integration of wearable activity monitors, standardized at-home dietary recalls, and biospecimen banking further supports the generation of ecologically valid data and mechanistic insight for future precision-aging studies.

The trial also has limitations inherent to early-phase pilot work. The modest sample size and short intervention period preclude definitive conclusions about long-term clinical outcomes. Findings may not be generalizable to older adults, men, or individuals with advanced comorbidities. As with many behavioral and nutritional interventions, adherence to supplementation and device wear may vary; however, these risks are mitigated by a structured adherence-monitoring strategy combining pill counts, paper logs, and biweekly check-ins. Objective tracking of wearable data provides an additional safeguard against nonadherence. Finally, while the biomarker panel is mechanistically grounded, the study is not powered to detect changes across all exploratory endpoints.

Despite these constraints, GRACE is positioned to make several important contributions. First, it will generate the clinical data on the combined glycation-lowering effects of GLYLO in humans, bridging a translational gap between promising preclinical findings and potential clinical application. Second, it will provide effect size estimates for a broad range of outcomes, informing the design and powering of future multi-site efficacy trials. Third, by focusing on an accessible, low-cost, non-pharmacological intervention, GRACE aligns with public health goals to expand healthy aging strategies that are both scalable and acceptable to midlife women.

In summary, GRACE is a pilot trial that combines molecular, physiological, and functional measures to evaluate a glycation-lowering strategy in postmenopausal women at elevated metabolic risk. Grounded in the principles of geroscience, the study examines whether targeting a fundamental aging mechanism, glycation stress, can influence both proximal biochemical markers and broader health-related outcomes. Findings from this pilot will help determine the feasibility, safety, and potential impact of this approach, and will inform the development of multimodal strategies and the design of larger clinical trials to test glycation-lowering interventions for healthy aging in more diverse populations.

Methods

Study design overview

The GRACE study is a randomized, double-blind, placebo-controlled, parallel arm, pilot study designed to assess the impact of a glycation-lowering non-pharmacological intervention (GLYLO) for 24 weeks compared to placebo on glycation biomarkers, metabolic function, hormonal health, and aging-related outcomes in postmenopausal women (Fig. 2). Following informed consent and screening (Visit 1), eligible participants are enrolled in the study. Baseline measures are collected at Visit 2 (week 0), after which daily administration of GLYLO or placebo begins (day 1). An interim assessment is conducted at Visit 3 (week 12), and final post-intervention measures are collected at Visit 4 (week 24). Tolerability and adherence are monitored through biweekly phone calls. The study protocol described here has been approved by the Advarra Institutional Review Board (current version 3.0 approved on 24th November 2025). Any further amendments will be reported to the sponsor and approved by the IRB. The study is being conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and applicable regulatory requirements⁴⁸. The study is registered with ClinicalTrials.gov (NCT06813261) on 2/3/2025, and all items from the WHO Trial Registration Data Set are provided in Supplementary 1. This manuscript outlines the study protocol in alignment with the SPIRIT 2025 guidelines, with the corresponding SPIRIT checklist available in Supplementary 2.

Fig. 2.

Schedule of GRACE study illustrating screening, assessments, and data endpoints SPIRIT Figure.

Study setting

The GRACE Study is conducted by investigators and clinical research staff at the Buck Institute for Research on Aging, an academic institution located in Novato, California, USA. All clinical visits and study-related activities take place within the Institute’s Clinical Research Unit. Most participants are anticipated to reside in the surrounding Northern California region.

Study population and eligibility criteria

The GRACE study aims to enroll 30 community-dwelling postmenopausal women aged 45–65 years who are in generally stable health but exhibit early signs of metabolic risk. No restrictions were placed on race or ethnicity. Recruitment was open to participants of all racial and ethnic backgrounds. Race and ethnicity are self-reported and will be summarized descriptively. This population was intentionally selected to reflect a life stage where hormonal and metabolic shifts, such as increased adiposity, insulin resistance, and glycation stress, are emerging but may still be modifiable through targeted intervention. Given the high prevalence of metabolic dysfunction and elevated glycation stress in women during the postmenopausal period^37,38, this group is likely to benefit from glycation-lowering strategies. Promising preclinical⁴⁴ and early clinical data^41,42 support the rationale for studying these interventions in this demographic.

GRACE study directly assesses whether glycation stress in humans can be reduced through GLYLO supplementation, and eligibility criteria were developed to optimize the likelihood of detecting such a signal. To enrich the cohort for individuals at increased risk, participants must exhibit elevated glycemic markers and/or central adiposity but not meet criteria for overt diabetes. Additionally, the study excludes conditions that could obscure metabolic endpoints or pose safety concerns, such as uncontrolled hypertension, significant hepatic abnormalities, active malignancy, infection, or gastrointestinal bleeding, and other investigator-determined safety risks. Inclusion and exclusion criteria were kept as streamlined as possible to facilitate recruitment, ensure safety, and preserve the interpretability of biological endpoints. Participants must be able to attend four in-person visits over 6 months and be willing to comply with all study procedures. Full eligibility criteria are provided in Table 1.

Table 1.

Inclusion/exclusion criteria

Inclusion criteria 1. Adults identified as female at birth with ovaries present (self-report) 2. Post menopause >1 year since last menses (self-report) 3. Aged 45–65 years 4. Anthropometric criteria (either of the following must be met):  • BMI ≥ 25 kg/m², based on self-reported weight and height  • OR Waist circumference ≥88 cm, based on self-measured values. Participants may provide average home weight measurements over two consecutive days if their BMI at the screening visit is slightly below 25 kg/m². 5. Glycemic criteria: HbA1c 5.5–6.4% inclusive (screening measurement).

Exclusion criteria 1. Bilateral oophorectomy 2. Receiving systematic hormone replacement therapy (HRT). Use of local vaginal estrogen therapy (e.g., estrogen creams, vaginal tablets, or estrogen rings such as Estring) is permitted. (self-report). 3. Currently prescribed or received weight loss medications within the past 6 months or currently enrolled in a defined weight loss program. Weight must be stable (>4%) within the last 3 months. 4. Regular use of high-dose GLYLO ingredients within the last 3 months. 5. Diabetes, T1DM or T2DM (self-report and screening tests): Treatment with any hypoglycemic agents (self-report), fasting glucose >125 mg/dL (screening test; may reassess once), current use of hypoglycemic drugs for non-diabetic reasons (self-report). 6. Elevated blood pressure readings (screening test): Resting Systolic Blood Pressure (SBP) ≥180 mmHg or resting Diastolic Blood Pressure (DBP) ≥100 mmHg. If a participant’s blood pressure is elevated at the screening visit but not consistent with this threshold, they may provide home blood pressure readings (twice daily for two consecutive days) for the study team to evaluate eligibility. 7. Psychotropic and/or other medications known to significantly impact weight unless on a stable dose for ≥6 months (self-report). 8. Liver enzyme tests (alanine transaminase, aspartate transaminase) (screening test): >2 times the laboratory upper limit of normal. Reassessment during screening may be allowed under some conditions (e.g., recent use of acetaminophen). 9. Immunosuppressive disorders, taking immunosuppressive medications (including oral prednisone >10 mg/day and biological immunosuppressants), or receiving chemotherapy. 10. Active gastrointestinal bleeding, or active bleeding diathesis (or resolved within 6 months prior to randomization) (self-report) 11. Active peptic ulcer disease (or resolved within 6 months prior to randomization) (self-report) 12. Active malignancy (or resolved within 6 months prior to randomization), except non-melanoma skin cancer not undergoing treatment (self-report). 13. Active infection (or resolved within 1 month prior to randomization) (self-report) 14. Allergy or hypersensitivity to any component of the supplement (self-report) 15. History of hyperthyroidism or thyroid cancer(self-report), current abnormal thyroid function (blood test at screening). 16. Cognitive status: Unable to provide informed consent to participate in and safely complete the protocol, as based on the judgment of the investigators (screening visit) 17. Psychiatric status: any condition that might affect the ability to comply with the protocol in the opinion of the Clinical Investigator or Medical Officer (screening visit) 18. Active eating disorders (self-report). 19. Active diagnosis of Gout (self-report) 20. Any change to prescription medications within 3 months prior to randomization that are judged by the study physician to impact the results of the study (self-report) 21. No overnight hospitalization within 1 month prior to randomization (self-report) 22. The presence of a condition or abnormality that in the opinion of the Investigator or Medical Officer, would compromise the safety of the patient or the quality of the data

Study intervention and placebo

The formulation of GLYLO (Juvify Ltd, Kensington, CA) consists of five ingredients classified as GRAS by the U.S. FDA: benfotiamine (100 mg), nicotinamide (200 mg), vitamin B6 (pyridoxine, 50 mg), alpha-lipoic acid (150 mg), and piperine (10 mg). The placebo used in the study is a capsule containing microcrystalline cellulose (Progressive Placebo Inc, Brigham City, UT). Placebo capsules are visually identical to the active product in size, shape, color, and packaging (60 capsules per bottle). Both active and placebo capsules are bottled, labeled, and dispensed in a blinded manner to maintain allocation concealment throughout the study.

Based on our previous experience, incrementally introducing the supplement and splitting the dosage across the day improves tolerability. Therefore, participants take one capsule daily after their first meal during the first week, and from week 2 onward, they take two capsules daily, one after their first and last meal, for the remainder of the 24-week intervention period.

Compliance is assessed by daily completion of a paper Study Product Log reported by study participants. The log is compared to the number of leftover capsules returned at week 12 and week 24 visits. Participants who consume between 80 and 120% of their allocated study product are considered “per protocol”. Participants are contacted by phone by the study team every 2 weeks to review their study product use and any other comments, which is expected to improve adherence and retention.

GRACE outcomes

Primary outcome

The primary outcome of the GRACE study is the reduction in serum levels of glycation stress markers, MGO and key MGO-derived and glycation-related AGEs (primarily Nε-(1-carboxyethyl)-L-lysine [CEL], MGO-derived hydroimidazolone [MG-H1], and Nε-(carboxymethyl)lysine [CML]), from baseline to week 24. These biomarkers reflect the biochemical burden of glycation, a key driver of metabolic dysfunction and aging-related decline⁴⁴. They will be quantified using a validated targeted LC–MS assay based on the QuarkMod workflow, enabling structurally specific, isotope-dilution quantification of MGO- and glycation-derived protein post-translational modifications, as established in our preclinical work⁴⁴. As the most proximate mechanistic targets of the intervention, reductions in MGO and its downstream AGEs will provide critical translational evidence linking clinical efficacy in humans to mechanisms identified in model systems. To our knowledge, this is the first clinical study to evaluate these specific glycation biomarkers in response to this specific glycation-lowering combination (GLYLO) in postmenopausal women.

Secondary outcomes

Secondary outcomes of the GRACE study focus on metabolic health markers that are closely linked to glycation stress and age-related metabolic dysfunction. Based on the known role of AGEs in impairing insulin signaling and promoting adiposity, we hypothesize that GLYLO supplementation may improve key metabolic parameters.

Fasting glucose, insulin, and HbA1c are measured at a CLIA-certified Quest Diagnostics laboratory, with insulin resistance calculated using the homeostatic model assessment (HOMA-IR). These indices provide an integrated view of glycemic control and insulin sensitivity, central features of metabolic risk. Body composition is assessed via Bioelectrical Impedance Analysis (BIA; RJL Systems Quantum V Segmental System, Clinton Township, MI). In postmenopausal women, age-related shifts in fat distribution and reductions in skeletal muscle mass are strongly associated with increased insulin resistance^46,49 and functional decline^50,51. Improvements in body composition could therefore reflect favorable metabolic adaptation. Anthropometric measurements, including weight, height, body mass index (BMI), and waist and hip circumference, further characterize central adiposity, a key driver of cardiometabolic risk. Together, these secondary outcomes offer a comprehensive assessment of metabolic status and are expected to respond to glycation-lowering interventions.

A final secondary outcome is circulating FSH, analyzed by Quest Diagnostics. In women, FSH levels rise significantly during the menopausal transition due to ovarian follicle depletion and reduced estrogen feedback^52,54. Elevated FSH is not only a biomarker of menopausal status but has also been implicated in the pathophysiology of central adiposity, bone loss, and other aging-related phenotypes^3,6,55,56. Monitoring FSH in this study provides insight into whether glycation-lowering therapy may influence hormonal aging trajectories in postmenopausal women.

Exploratory outcomes

Aging is a multidimensional process, and interventions targeting fundamental mechanisms, such as glycation stress, may influence a wide range of downstream physiological systems. To capture this broader impact, the GRACE study includes exploratory outcomes assessing physical function, cognitive performance, QoL, and biological aging markers. These measures reflect domains of whole-person function that are clinically meaningful in the context of aging and resilience, particularly in midlife women.

Given the pilot nature of this study and its modest sample size, these exploratory endpoints are not powered for definitive conclusions. Instead, they are intended to generate preliminary data and assess potential signals of efficacy that could inform larger, follow-on trials. Because these outcomes are more distal to the direct mechanism of action of the intervention, it remains uncertain whether they will be affected. However, their inclusion allows us to evaluate whether improvements in glycation stress translate into functional and quality-of-life benefits, critical endpoints for future geroscience-guided interventions.

Physical function is assessed using a combination of validated tools to capture both performance-based and real-world measures of mobility and activity. These include the Short Physical Performance Battery⁵⁷; hand grip strength via hydraulic handheld dynamometer (Jamar®, Model 563213, Performance Health, Downers Grove, IL)⁵⁸, 6-min walk test, and actigraphy via FitBit Inspire 3 (FitBit, San Francisco, CA), equipped with a triaxial accelerometer and photoplethysmography sensors. Each measure provides complementary insights. Grip strength is a robust, low-cost indicator of musculoskeletal health and has been consistently linked to morbidity, disability, and all-cause mortality in large population-based studies. The 6MWT is a practical, well-validated assessment of cardiorespiratory fitness (CRF) and functional endurance, particularly relevant in aging populations where CRF is a strong predictor of future disability and cardiovascular risk. The Fitbit Inspire 3 enables passive, continuous measurement of physical activity and sleep, providing ecologically valid data on movement patterns that may influence key secondary outcomes such as body composition and insulin resistance. Collectively, these measures allow us to evaluate how glycation-lowering intervention may impact physical function and real-world activity over the course of the study.

Cognitive function is assessed using three validated tools: the Montreal Cognitive Assessment (MoCA), the Digit Symbol Substitution Test (DSST), and the Trail Making Test (TMT). MoCA and Trails A/B are components of the National Alzheimer’s Coordinating Center Uniform Data Set Version 3^91,92. The MoCA is a 30-point screening instrument that evaluates multiple cognitive domains, including memory, attention, and executive function; higher scores indicate better performance⁵⁹. The DSST, a subtest of the WAIS-III, measures processing speed, attention, and working memory⁶⁰. Participants match digits to symbols in the 90 s; the total correct responses serve as the score. The TMT⁶⁰ assesses visual attention, psychomotor speed, and cognitive flexibility through two timed parts: Part A evaluates sequencing and motor speed, while Part B requires set-shifting between numbers and letters, a task that is particularly sensitive to executive dysfunction and frontal lobe impairments commonly observed in aging. In the TMT, slower times indicate worse performance. Together, these tools provide a multidimensional view of cognitive function. As exploratory outcomes, they allow us to assess whether improvements in glycation stress may translate into cognitive benefits, particularly in domains known to decline with age and linked to functional independence in older adults.

QoL is assessed using three validated instruments selected to capture general health status, menopause-related symptoms, and sleep quality domains known to influence well-being and functional independence in midlife women: (1) the EuroQol 5-Dimension 5-Level (EQ-5D-5L), (2) the Greene Climacteric Scale, and (3) the Pittsburgh Sleep Quality Index (PSQI). The EQ-5D-5L provides a global measure of health-related QoL across five key dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression^61,62. It includes a visual analog scale for self-rated health (higher scores are better) and is widely used in aging and clinical research due to its simplicity, strong psychometric properties, and broad applicability across disease states^63–66. The EQ-5D-5L has demonstrated sensitivity to change following interventions in middle-aged women and adult populations, supporting its use to capture improvements in general health and functional status^67–69. The Greene Climacteric Scale was selected to assess symptom burden specifically related to menopause (higher scores relate to higher symptom burden), including psychological, vasomotor, somatic, and sexual symptoms⁷⁰. As postmenopausal women are the target population for this study, this tool provides nuanced insight into domains that may be indirectly affected by the intervention, including mood, hot flashes, and sleep disturbances, many of which are also linked to metabolic and hormonal dysregulation. It has been shown to detect reductions in menopausal symptom burden following lifestyle and non-pharmacologic interventions in perimenopausal and postmenopausal women^71,72. The PSQI captures subjective sleep quality^73,74, a critical component of overall health and a known modifiable factor linked to metabolic function, mood, and cognitive performance. Poor sleep is common in midlife women⁷⁵ and may interact with increased glycation stress and worsening health outcomes. Higher PSQI scores indicate poor sleep quality. Prior studies have shown improvements in PSQI scores following lifestyle and mind-body interventions in perimenopausal and postmenopausal women^76–78. Together, these QoL measures enable a multidimensional understanding of how the intervention may impact daily functioning, symptom burden, and perceived well-being, key outcomes for translating geroscience into meaningful clinical benefit.

In addition to functional and quality-of-life outcomes, the GRACE study includes a suite of molecular and systems-level exploratory measures that reflect biological aging processes and metabolic health. These outcomes, ranging from non-invasive imaging to dietary intake and biobanked biomarker analysis, were selected to provide mechanistic and translational insight into how glycation-lowering therapy may influence systemic physiology. These measures offer important downstream readouts of glycation stress and have potential value in identifying early signals of efficacy that could inform endpoints for future, larger trials.

Retinal imaging (TopCon NW500 Retinal Imaging System, Oakland, NJ) offers a non-invasive window into biological aging and systemic metabolic health. Prior studies have linked retinal features to metabolic syndrome⁷⁹, obesity⁸⁰, and mortality risk⁸¹, supporting their relevance as aging biomarkers. Leveraging this, our lab developed eyeAge⁴⁵, a deep learning model trained on longitudinal fundus images to predict biological age. The resulting retinal age gap, the difference between predicted retinal age and chronological age, serves as an index of accelerated or decelerated aging. Given its high accuracy and stability over time, it is well-suited for use in this 6-month intervention study and may offer insight into whether glycation-lowering therapy influences biological aging trajectories.

Skin autofluorescence provides a complementary, tissue-level measure of glycation burden. Using an AGE Reader device (Diagnoptics Technologies, Groningen, The Netherlands), we assess the accumulation of AGEs in the skin non-invasively. Elevated skin AGE levels have been associated with a range of adverse outcomes, including cardiovascular mortality⁸², kidney disease^83,84, and chronic inflammatory conditions^85,86. As AGEs are a direct mechanistic target of intervention, this measure may serve as an integrative readout of cumulative glycation stress over the study period. In addition, we will explore the relationship between skin autofluorescence and circulating serum AGEs to evaluate whether tissue-level glycation reflects systemic glycation, providing complementary insight into intervention effects.

To contextualize these biological measures, we assess habitual dietary intake using the Automated Self-Administered 24-h Recall (ASA24), a web-based tool developed by the National Cancer Institute⁸⁷. Dietary patterns, particularly those high in exogenous AGEs, can influence glycation stress and metabolic outcomes. Participants complete the ASA24 in week 1 and again in week 24, allowing us to track potential changes in intake that could affect or interact with the intervention’s effects.

Finally, banked plasma, serum, and urine samples will be analyzed for a range of exploratory biomarkers to support mechanistic investigation and future hypothesis generation. These include circulating ovarian hormones, estradiol (E2), luteinizing hormone (LH), progesterone, and testosterone, which may provide insight into how glycation stress interacts with the hypothalamic–pituitary–gonadal axis in postmenopausal women. We will also quantify markers of systemic inflammation, including interleukin-6 (IL-6), interleukin-10 (IL-10), tumor necrosis factor alpha (TNF-α), and C-reactive protein (CRP), given their established links to both metabolic health and AGE accumulation. Notably, several of these inflammatory biomarkers, including IL-6 and CRP, have been recommended by the TAME (Targeting Aging with Metformin) Biomarkers Working Group⁸⁸ as part of a core panel for assessing biological aging in geroscience-guided clinical trials. That group proposed criteria for biomarker selection that include biological plausibility, robust association with health risk, feasibility in clinical trials, and responsiveness to intervention, principles that align with the goals of GRACE.

In addition, exploratory analyses will include assessment of aging-relevant endocrine markers within the Klotho–FGF23–phosphate axis. Serum Klotho, fibroblast growth factor-23 (FGF23), and phosphate levels will be measured in banked serum samples at baseline and post-intervention to provide complementary insight into endocrine and metabolic aspects of biological aging. Plasma markers of phenotypic age will be assessed to explore whether short-term reductions in glycation stress are accompanied by changes in composite measures of biological aging. While GRACE is not powered to definitively assess these outcomes, they offer an opportunity to explore whether modulating a specific molecular mechanism, glycation, may produce broader systemic effects in line with the Geroscience Hypothesis, which posits that targeting hallmarks of aging can simultaneously impact multiple age-related conditions. Urine samples will be analyzed for renal function, oxidative stress, and AGE-related metabolic byproducts, further enriching our ability to characterize physiological responses to the intervention. Together, this suite of exploratory measures enables a multidimensional investigation of glycation-lowering as a geroscience intervention, even within the constraints of a short-term pilot trial.

Recruitment procedure

Participants are recruited from Novato, California, and the broader northern San Francisco Bay Area using a staged, multimodal outreach strategy. Recruitment materials, including flyers and social media ads, are distributed digitally and physically through community postings, emails, mailing lists, and brief presentations at local events. Targeted geolocation and demographic filters are applied to social media ads to enhance outreach. Interested individuals complete a brief telephone pre-screening with study staff, during which they receive an overview of the study and are asked about key eligibility criteria such as menopausal status, anthropometrics (e.g., self-reported height, weight, waist circumference), major health conditions, and the absence of recent hospitalizations. Individuals who meet preliminary criteria are invited for an in-person screening visit, where written informed consent is obtained by study staff and full eligibility is assessed.

Participant timeline

Pre-visit guidelines for all in-person study visits include fasting for at least 10 h and refraining from alcohol, vigorous physical activity, and psychoactive cannabis use during the same period. Use of new medications or supplements, as well as any recent changes in their health status, are queried at each visit. A detailed overview of visit-specific assessments and procedures is provided in Fig. 2.

Screening visit (Visit 1)

Participants attend an in-person clinic visit. Participants provide written informed consent before undergoing any study-specific procedures. A fasting blood sample is collected for clinical lab tests, and basic demographic data, vital signs, anthropometrics, menstrual/menopause history, and current medication and supplement use are recorded. Eligibility is assessed by the study Medical Officer prior to enrollment.

Baseline visit (Visit 2, week 0)

During this in-person visit, fasted blood and urine samples are collected for clinical labs and biomarker analysis. Vital signs, anthropometrics, and body composition are assessed. Participants complete QoL questionnaires, followed by physical and cognitive function tests. Non-invasive assessments include skin and retinal imaging. Participants receive a 12-week supply of study product, a wearable physical activity monitor, and instructions for using and completing the ASA24 dietary recall tool. Participants are also given the Study Product Log for at-home assessment of adherence and product tolerability.

At-home procedures

At home, participants are instructed to maintain consistent diet and activity levels throughout the study duration. Participants consume one capsule per day after their first meal for the first week, followed by two capsules per day from Week 2 onward, one after each of their first and last meals. They complete a daily Study Product Log to indicate the time of product consumption. For the first 14 days, the Study Product Log also includes a structured daily product tolerability questionnaire (PTQ) to record pre-specified side effects (gas/flatulence, abdominal cramping, stomach rumbling, burping, and reflux/heartburn); after 14 days, there is space for free-form recording of side effects.

While paper logs are simple and low-cost, adherence data from self-reports can be affected by recall bias and incomplete recording over time⁸⁹. To improve accuracy, the GRACE study uses a multi-pronged approach that combines paper logs with pill counts at Visits 3 and 4 and bi-weekly phone calls to review entries and reinforce compliance⁴⁷. The Study Product Log is reviewed for completeness and PTQ responses during these calls (week 2, 4, 6, 8, 10, 14, 18, and 22) and in person at Visit 3 and Visit 4. Potential adverse events (AEs) are assessed during each visit and follow-up call. Participants complete ASA24 recalls at Weeks 1 and 23 and sync physical activity data from their Fitbit throughout the study.

Interim visit (Visit 3, week 12)

Approximately 90 days post-Baseline, participants complete interim assessments at an in-person visit. They return the unused product and the Study Product Log for monitoring of adherence. Biospecimen collection and all assessments from Visit 2 are repeated. Participants receive a second 12-week supply of study product and a new Study Product Log. This mid-point assessment is important for detecting the timing and trajectory of any changes in biomarkers or functional measures, providing insight into whether effects of the intervention emerge early, progress gradually, or require the full 24 weeks to become apparent.

Final visit (Visit 4, week 24)

Approximately 180 days post-Baseline, participants return for the final in-person visit. Biospecimen collection and outcome measure assessments performed at Visit 3 are repeated. They return the unused product and Study Product Log Part II for monitoring adherence. To evaluate blinding, participants are asked to guess their group assignment.

Sample size

Evaluating serum AGEs in response to a dietary supplement in postmenopausal women is a novel area with limited prior data. For this reason, sample size estimation for the GRACE trial was based on the most relevant available randomized study by Kahleova et al.⁹⁰. Although that study assessed dietary rather than serum AGEs, it is the only trial in postmenopausal women reporting AGEs-related variability. Because dietary AGE intake is strongly associated with circulating AGE levels, the variability estimates from this study provided the most appropriate proxy for planning this pilot trial.

In their study, Kahleova et. al.⁹⁰, observed large between-group differences at 12 weeks (intervention mean 2037 ± 1055 ku/day; control mean 6839 ± 3417 ku/day). Standard deviations (SDs) were derived from the reported confidence intervals and sample sizes, yielding within-group SDs of approximately 1055 and 3417 ku/day for the intervention and control arms, respectively. Using these SDs with a two-sided α of 0.05 and 80% power, the minimum required sample size was approximately 4 participants per group. To ensure stable effect-size estimation and accommodate expected variability in this pilot study, we will enroll 15 participants per group (30 total). Assuming a 20% post-randomization dropout rate, we plan to randomize 38 participants to retain 30 completers. This anticipated attrition reflects the demands of the study, i.e., 6-month duration of the study, multiple in-person visits, fasting blood draws, daily supplement adherence, and completing the study product log. Given that up to 50% of screened individuals may be ineligible or withdraw before randomization due to stringent eligibility criteria, up to 76 participants may be consented and screened (Fig. 3). This sample size is appropriate for a pilot trial and will generate effect-size estimates needed for designing future efficacy studies.

Fig. 3.

Outline of the planned enrollment, randomization, intervention, follow-up, and primary end-point assessment across the GRACE study timeline.

Randomization and blinding

Eligible participants at the screening visit were randomized in a 1:1 ratio to receive either GLYLO or a visually matched placebo for 24 weeks. The block randomization sequence (block size of four) is generated by external personnel using a computer-based program. Allocation occurred at Visit 2, with assignments implemented sequentially according to participant enrollment. Study staff involved in recruitment, data collection, and outcome assessment remain blinded to product identity but not to coded group allocation. The study product and placebo were packaged identically by personnel not involved in data handling or analysis. Emergency unblinding procedures are in place.

Participant retention

To promote continued engagement throughout the study, retention strategies include establishing a strong rapport with participants during initial visits and maintaining regular biweekly communication through phone calls and email reminders about upcoming appointments. Participants receive a stipend following each in-person visit, with the amount increasing at each subsequent visit to encourage sustained participation and recognize their time and commitment.

Data storage, management, and availability

During each study visit, data is recorded by trained research staff into REDCap (Research Electronic Data Capture), a secure, web-based electronic data management system hosted by the Buck Institute for Research on Aging. Paper source documents are maintained in locked cabinets in restricted-access areas, and REDCap data are stored on secure institutional servers with access limited to authorized personnel. All research records and biospecimens are retained for a minimum of 10 years following study closure, in accordance with institutional and regulatory requirements. Every effort will be made to safeguard participant confidentiality throughout the study and in all data handling processes.

This study adheres to current NIH guidelines on data sharing and transparency. Any data prepared for sharing will be fully de-identified, with no personal identifiers or variables that could allow re-identification. Participant-level data, free of direct and indirect identifiers, is securely stored at The Buck Institute for Research on Aging under the supervision of the Principal Investigator. Upon reasonable written request, a de-identified dataset may be shared with investigators. Study findings will be disseminated through peer-reviewed articles, scientific conferences, and other academic forums. Authorship decisions will follow the guidelines established by the International Committee of Medical Journal Editors (ICMJE).

Statistical analysis

The primary outcome, MGO and AGEs (CEL, MG-H1, CML) from baseline to 6 months, will be analyzed using analysis of covariance (ANCOVA). The 6-month biomarker value will serve as the dependent variable, with treatment group (GLYLO vs. placebo) as the independent variable and baseline biomarker levels included as covariates. This approach accounts for baseline variability and increases statistical power. Effect sizes will be reported as adjusted mean differences with 95% confidence intervals, and statistical significance will be evaluated using a two-sided alpha of 0.05. Sensitivity analyses will explore the influence of baseline metabolic status and other covariates, with age, BMI, and HbA1c included where appropriate, noting that these variables are restricted by study eligibility criteria.

Secondary outcomes, including metabolic, hormonal, and functional measures, will be assessed at 6 months post-randomization. Exploratory analyses will use data collected at both three and 6 months. Baseline characteristics and outcome variables will be summarized using descriptive statistics (means, medians, standard deviations, interquartile ranges, or frequencies, as appropriate). For group comparisons, parametric methods will be applied as the primary approach. Box-Cox transformations will be applied to normalize skewed variables, with nonparametric or robust alternatives considered if model assumptions are not met. Between-group comparisons of continuous outcomes will be performed using independent two-sample t-tests, while within-group changes will be assessed with paired t-tests. Missing data will not be imputed. Treatment effect heterogeneity will be explored using subgroup analyses or interaction terms, as appropriate. No interim efficacy analysis is planned. All analyses will be conducted using validated statistical software such as R, GraphPad Prism, or SAS.

Study oversight and participant safety

Participant safety and data integrity are closely monitored throughout the duration of the GRACE study. Oversight includes both remote and in-person review of study conduct. These efforts prioritize informed consent documentation, participant eligibility confirmation, adverse event tracking, protocol compliance, endpoint data accuracy, and accountability for investigational product use. Medical oversight is provided by an independent physician (their role in the study is limited to that of a medical officer) with clinical experience in geriatrics. The medical officer is responsible for confirming participant enrollment, assessing eligibility, and making determinations about early discontinuation if necessary. In addition, the medical officer evaluates all serious adverse events and consults on any notable clinical findings during the study. Reports of adverse events are collected during scheduled site visits and telephone follow-ups and are documented in the eCRFs. Any serious adverse events (SAEs) are promptly reported to the overseeing medical officers, the institutional officer, and the reviewing IRB. A subset of adverse events, referred to as “events of note” includes those that result in study discontinuation, require clinical treatment, or prompt diagnostic testing.

Author contributions

V.T. conceptualized the study, developed the study protocol, will contribute to the investigation, perform the primary data analysis, and drafted the original manuscript and prepared figures. P.K. contributed to the study conceptualization, secured funding for the study, and provided critical revisions to the manuscript. B.S. assisted in protocol development, provided key methodological insights and supervision, supported project administration, contributed to manuscript writing and editing and will contribute to data analysis. J.N. provided clinical oversight and contributed to manuscript review and editing. All authors reviewed and approved the final version of the manuscript and are responsible for the decision to submit it for publication.

Data availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Competing interests

Dr. Pankaj Kapahi is the founder of Juvify Ltd. and holds an equity interest in the company. Dr. Kapahi is an inventor on patents related to the study product, which are assigned to the Buck Institute and licensed to Juvify Ltd. The Buck Institute also holds an equity interest in Juvify Ltd. Individual (for Dr. Kapahi) and institutional (for the Buck Institute) conflict management plans were developed and approved by the Buck Institute and the Institutional Review Board. Dr. Kapahi contributed to study design; however, neither Dr. Kapahi nor Juvify Ltd. has any role in data collection, data analysis, or interpretation of the results. Study staff, including the principal investigator, will remain blinded through study completion unless unblinding is required for participant safety. All other authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41514-026-00373-x.