1. Scope

Covered:

– Somatic gene-transfer strategies proposed for ageing/healthspan (AAV-delivered TERT, KLOTHO, FGF21; senolytic and reprogramming gene therapies).[1][2][3][4]

– Epigenetic modifying interventions: senolytics/senomorphics, DNA methyltransferase (DNMT) and histone deacetylase (HDAC) inhibitors, partial epigenetic reprogramming, and repurposed agents with epigenetic effects (metformin, NAD⁺ precursors, semaglutide, pitavastatin).[5][6][7][8][9]

– Epigenetic measurement: DNA-methylation clocks as biomarkers/surrogate endpoints.[6][10][11]

Not covered: germline editing (prohibited in UK clinical practice); licensed disease-specific gene therapies (e.g. inherited retinal disease, SMA); general lifestyle geroprotection except where it modifies epigenetic clocks.

Positioning: All content is adjunctive to, and never a substitute for, conventional primary/secondary care. Nothing here is guideline-endorsed for healthy-ageing indications.

2. Background and pathophysiology

Biological rationale. Ageing is driven by interlinked hallmarks — genomic instability, telomere attrition, epigenetic alteration, loss of proteostasis, mitochondrial dysfunction, cellular senescence and chronic inflammation (“inflammaging”).[1][12][9] The geroscience hypothesis holds that targeting these shared mechanisms could delay multiple age-related diseases simultaneously.[5]

Key mechanisms targeted:

– Cellular senescence / SASP — senescent cells accumulate with age and secrete pro-inflammatory, tissue-damaging factors; clearance (senolysis) or SASP suppression (senomorphics) is the target.[13][14][12]

– Epigenetic drift — age-associated changes in DNA methylation, histone modification and chromatin architecture; theoretically reversible via reprogramming or epigenetic drugs.[7][8][9]

– Metabolic/NAD⁺ and mTOR/AMPK signalling — modulated by metformin, NAD⁺ precursors, rapamycin.[5][15][14]

– Growth/regenerative gene programmes — TERT (telomere maintenance), KLOTHO, FGF21, GDF11 delivered by AAV vectors.[1][4]

Robust human mechanistic evidence is limited to: (i) demonstration that senolytics reduce senescent-cell markers in some human tissues; and (ii) that certain interventions measurably shift DNA-methylation clocks.[13][14][6][10] Causal linkage from these mechanistic/biomarker shifts to improved human healthspan or survival is not established.[16][10]

Preclinical only (clearly separated — do not extrapolate to patients):

– AAV-TERT, AAV-KLOTHO, AAV-FGF21 extend healthspan/lifespan in aged mice; FGF21 gene therapy improved multi-organ function in geriatric mice.[4]

– Partial reprogramming (OSK/OSKM factors) resets methylation age and restores function in murine and in-vitro models, with oncogenic/dedifferentiation risk.[3][7]

– DNMT and HDAC inhibitors extend lifespan or alleviate ageing phenotypes in model organisms.[8][9]

3. Evidence base and grading

Types of evidence available. For hard clinical endpoints in ageing: essentially none from dedicated RCTs (the flagship TAME metformin RCT and VA-IMPACT are ongoing/not yet reported for geroscience endpoints).[14] For biomarker/surrogate endpoints: multiple small-to-moderate RCTs and systematic reviews of DNA-methylation clock responsiveness.[6][10] For gene therapy: preclinical plus early-phase/disease-specific clinical work; no anti-ageing RCTs.[1][2][3]

GRADE-style certainty by outcome:

Outcome	Intervention class	Evidence & certainty	Key limitations	Strength of recommendation
All-cause mortality / lifespan	Any gene or epigenetic therapy	Very low — no RCTs; preclinical/observational only	Indirectness (animal), imprecision, risk of bias	Only in research [1], [2], [3]
Healthspan / age-related multimorbidity	Metformin, senolytics	Very low–Low — observational (metformin) + ongoing RCTs; no completed geroscience RCT	Confounding, no completed trial	Conditional / research [5], [14]
Reduction in senescent-cell burden	Senolytics (D+Q, fisetin)	Low — small human trials show target engagement in some tissues	Small n, surrogate, heterogeneity	Conditional / research [13], [14]
Reduction in epigenetic age (next-gen clocks)	Lifestyle, semaglutide, pitavastatin, caloric restriction, omega-3	Low–Moderate (surrogate only) — systematic reviews of ≈41 human studies	Surrogate endpoint; clocks not validated as causal surrogates; inconsistency	Conditional for biomarker tracking; do not infer clinical benefit [6], [10]
Epigenetic clock change from senolytics / NR / rapamycin	Senolytics, nicotinamide riboside, rapamycin	Low — no detectable clock effect reported	Null/inconsistent findings	Recommend against using clocks to justify these [6], [16]
Multi-organ rejuvenation	AAV gene therapy (TERT/KLOTHO/FGF21)	Very low — preclinical only	Indirectness (animal)	Only in research [3], [4]

GRADE domain notes: Risk of bias — many biomarker studies are unblinded, industry-adjacent or single-arm. Inconsistency — clock responsiveness varies markedly by clock, population and study duration.[10] Indirectness — much data is murine or uses surrogate endpoints rather than clinical outcomes. Imprecision — small samples, wide CIs. Publication bias — plausible in a commercially active longevity field.[17]

A critical caveat: DNA-methylation signatures of senescence are not reversed by senolytics, and several clocks stayed unchanged or accelerated after geroscience interventions — undermining the assumption that clocks fall whenever an intervention “works”.[16]

4. Patient selection and indications

Who might reasonably be considered (adjunctive, informed-consent, off-label):

– Middle-aged adults (≈40–65) with elevated cardiometabolic risk where a repurposed agent already has a conventional evidence base for that risk (e.g. metformin in dysglycaemia; a statin/semaglutide for lipid/weight indications).[6][14]

– Adults seeking structured biological-age monitoring who understand clocks are investigational biomarkers.[11]

Inclusion considerations: capacity for informed consent; realistic expectations; stable comorbidities; willingness to remain within conventional care.

Exclusion / high-risk groups:

– Active or prior malignancy — caution with any pro-regenerative/telomerase or reprogramming strategy (oncogenic risk).[1][3][7]

– Pregnancy/breastfeeding, planned conception.

– Significant frailty, immunocompromise, active infection (relevant to senolytics/gene vectors).

– Renal or hepatic impairment (drug-specific — e.g. metformin and eGFR).

– Prior AAV exposure/neutralising antibodies (for any gene-transfer product).[2]

Regulatory/ethical status (UK):

– No gene or epigenetic anti-ageing product is MHRA/EMA/FDA licensed for a longevity indication. No geroscience-specific regulatory framework exists.[17]

– Somatic gene therapy and epigenetic reprogramming should be restricted to approved clinical trials / research frameworks.[1][2][3][17]

– Repurposed licensed drugs (metformin, statins, semaglutide) used for ageing are off-label and require explicit off-label informed consent, GMC-consistent prescribing, and ideally co-ordination with the patient’s GP.

– Germline modification is not permissible.

5. Assessment and baseline work-up

History & examination: full medical/medication/family history; malignancy and cardiovascular risk assessment; frailty screen (e.g. frailty index, gait speed, hand-grip strength); capacity and expectations.[11]

Baseline investigations (tailor to the specific agent):

– FBC, U&E/eGFR, LFTs, HbA1c, fasting lipids, hsCRP.

– Ageing-relevant biomarkers where monitoring is planned: hsCRP, IL-6, IGF-1, GDF-15, plus functional measures (grip strength, gait speed, Timed-Up-and-Go); the domains reaching expert consensus for intervention studies.[11]

– Optional DNA-methylation clock panel — document that it is an investigational biomarker, not a diagnostic test; prefer next-generation/mortality- or pace-of-ageing clocks (e.g. DunedinPACE, GrimAge, PhenoAge) which are more responsive and more strongly mortality-associated than first-generation clocks.[6][10][14]

– For gene therapy trials: anti-AAV antibody status, immunological screen (trial-specified).[2]

Risk stratification: categorise separately for (a) likelihood of benefit (usually low/uncertain) and (b) risk of harm (drug- and comorbidity-specific). High-harm groups: malignancy history, frailty, organ impairment, pregnancy.

Baseline documentation: off-label consent record; baseline biomarker/functional panel with assay and clock version; concomitant medications; agreed monitoring schedule and stopping rules.

The figure below benchmarks common epigenetic clocks and illustrates why chronological-age accuracy alone does not equal clinical/mortality prediction — relevant when selecting a clock for monitoring.

Figure 2 Systematic evaluation of epigenetic aging clocks across diverse datasets using Biolearn. a, Benchmarking of aging clocks across eight human datasets derived from whole blood (combined n > 3,000 samples), with mean R values indicating the average performance of each clock calculated using Pearson correlation. The dots colored based on the dataset represent Pearson’s R measured in each individual dataset (two-sided test). b, Code snippets demonstrating the streamlined curated data loading and implementation of aging clocks using the Biolearn library, highlighting its user-friendly interface and efficient data handling. c, Similar to a, but with dots representing the root mean square error (RMSE) for each clock. d, Scatter plots depicting the relationship between predicted epigenetic age and chronological age for six representative clocks across different datasets represented by different colors. The plots showcase the performance of each clock, with Pearson’s R and RMSE provided. The datasets shown include the Human Aging Rates Study (GSE40279), United Kingdom Ovarian Cancer Population Study (UKOPS) (GSE19711), Dutch schizophrenia case-control cohort (GSE41169), obesity genetics study (GSE73103), Young Finns Study (GSE69270) and newborns and nonagenarians study (GSE30870). Various age groups, ethnicities and populations are covered in the datasets. *Clocks using the sample’s chronological age as a predictor and are not directly comparable to other clocks. Illustration in a created in BioRender.com.

6. Dosing regimens and practical implementation

Important: Because no gene/epigenetic anti-ageing product is licensed, “dosing” below refers to repurposed agents or trial protocols. Anything beyond licensed use is off-label or research-only.

Regimens with the most human data (still off-label for ageing):

– Metformin — standard licensed dosing (e.g. 500 mg titrated to 1–2 g/day with meals); geroscience efficacy unproven pending TAME/VA-IMPACT. Avoid/adjust by eGFR.[14]

– Senolytics (research context): dasatinib + quercetin (“D+Q”) given as intermittent (“hit-and-run”) dosing in trials; fisetin under study — doses, schedules and safety are trial-specific and not established for routine use; treat as research-only.[13][14]

– NAD⁺ precursors (nicotinamide riboside/mononucleotide): raise blood NAD⁺ but human clinical-outcome benefit is unproven and mixed; no clock effect detected.[5][6][15]

Interventions extrapolated from early-phase/preclinical data — use only in research, caution:

– AAV gene therapy (TERT/KLOTHO/FGF21) — dose, vector, route entirely investigational; preclinical only.[1][4]

– Partial epigenetic reprogramming — no established human dosing; oncogenic risk.[3][7]

– DNMT/HDAC inhibitors for ageing — model-organism dosing only.[8][9]

Non-pharmacological interventions with human epigenetic-clock evidence (lower risk, reasonable adjuncts): structured exercise, plant-rich diet, caloric restriction, omega-3 supplementation, and a multivitamin-multimineral have each been associated with reductions in next-generation clocks.[6][10] These are supported for general health independently; the clock change should not be over-interpreted as proven life extension.

7. Monitoring, safety and follow-up

Monitoring plan (agent-dependent):

– Clinical: symptom review, weight, BP, functional measures (grip, gait).[11]

– Laboratory: FBC, U&E/eGFR, LFTs, HbA1c, lipids, hsCRP at baseline and periodically.

– Biomarker/clock re-testing: no earlier than ~3–6 months given assay noise; interpret cautiously — clocks may not move even with genuinely active interventions, and can paradoxically accelerate.[16][10]

Suggested intervals: short-term 4–12 weeks (tolerability, bloods); medium-term 3–6 months (biomarkers, function); long-term 6–12 monthly review of ongoing justification.

Adverse effects & safety:

– Repurposed drugs: per licensed safety profiles (e.g. metformin — GI upset, rare lactic acidosis, B12 deficiency; dasatinib — cytopenias, effusions, cardiotoxicity, bleeding risk — a major reason senolytics stay research-only).[12]

– Gene therapy classes: immunogenicity/anti-vector responses, off-target and insertional effects, theoretical oncogenesis (especially telomerase/reprogramming).[1][2][3][7]

– Senolytics: transient effects; long-term human safety unknown.[13][12]

Actions for abnormal findings: stop or reduce for significant lab/clinical toxicity; refer to the relevant specialty for new malignancy signal, cytopenias, cardiac or hepatic/renal derangement.

Interactions: dasatinib — CYP3A4 interactions, QT-prolonging drugs, anticoagulants/antiplatelets; metformin — iodinated contrast, other agents affecting renal function. Review polypharmacy in older adults.

Special populations: avoid in pregnancy/breastfeeding; dose-adjust or avoid in renal/hepatic impairment; heightened caution in frailty and extremes of age (limited evidence, higher harm potential).[5]

8. Contraindications and cautions

Absolute:

– Germline application (prohibited).

– Pregnancy/breastfeeding for cytotoxic senolytics and any gene-transfer product.

– Active malignancy for telomerase-activating, reprogramming or pro-regenerative strategies.[1][3]

– Use of unapproved gene therapies outside a regulated trial.

Relative / specialist input required:

– Prior malignancy; significant immunocompromise; active infection.

– Renal or hepatic impairment (drug-specific).

– Significant frailty or multimorbidity.

– Concurrent QT-prolonging or interacting medications (senolytics).

– Prior AAV exposure (gene therapy).[2]

Harm likely to outweigh benefit with current evidence: routine clinical (non-trial) use of AAV gene therapy, partial reprogramming, or DNMT/HDAC inhibitors for healthy ageing; use of senolytics outside a trial; and any decision to start/stop conventional therapy on the basis of an epigenetic-clock reading alone.[16][10][17]

9. Practical management scenarios (CKS-style)

Scenario A — Middle-aged adult with multiple cardiometabolic risk factors.

Recommendation: Offer conventional risk-factor management; consider off-label repurposed agents only where they independently meet conventional indications (e.g. metformin for dysglycaemia). Gene/reprogramming therapies: research only. (Conditional recommendation, low certainty.)

– Assessment: full CVD risk, HbA1c, lipids, renal/hepatic function, baseline function/biomarkers.

– Consent: explain off-label status and that clock/biomarker change is not proven to reduce disease or death.

– Initiation: optimise diet, exercise, sleep first (human clock evidence, low risk); add licensed agents per their own indications.[6]

– Monitoring: bloods at 6–12 weeks; function/biomarkers at 3–6 months.

– Escalate/stop: toxicity, no tolerability, or unrealistic expectations — revert to standard preventive care.

Scenario B — Older, frail patient with multimorbidity.

Recommendation: Avoid experimental gene/epigenetic therapies; senolytics research only. (Conditional-to-strong against, given harm potential and absent benefit data.)

– Prioritise deprescribing, exercise, nutrition, falls and frailty management.

– Any gerotherapeutic only within a trial with geriatric oversight.

– Stop threshold is low: any new cytopenia, functional decline, or intercurrent illness.

Scenario C — Patient already under specialist care seeking adjunctive therapy.

Recommendation: Restrict to research / co-managed off-label with the treating specialist’s agreement. (Conditional.)

– Assessment: confirm no interaction with disease-modifying therapy (e.g. oncology, transplant, cardiology).

– Consent & shared decision-making: document specialist liaison; avoid anything that could compromise established treatment.

– Monitoring: aligned with specialist schedule.

– Escalate/refer: any deterioration in the primary condition → defer to specialist; discontinue the adjunct.

10. Research gaps and future directions

– No completed RCTs link gene/epigenetic interventions to hard human outcomes (mortality, multimorbidity, disability). Await TAME and VA-IMPACT for metformin.[14]

– Surrogate validity of epigenetic clocks is unresolved — clocks respond inconsistently, may not reflect senescence, and can move in unexpected directions; they are not yet validated causal surrogate endpoints. Priority: prospective linkage of clock change to clinical outcomes.[16][10]

– Optimal dosing, schedule and long-term safety of senolytics in humans are unknown.[13][12]

– Gene therapy for ageing needs first-in-human safety, delivery/immunogenicity solutions, and oncogenicity data before any non-trial use.[1][2][3]

– Regulatory framework for gerotherapeutics does not exist and must be developed.[17]

– Recommended stance: partial reprogramming, AAV gene therapies, DNMT/HDAC inhibitors and senolytics should be confined to well-designed trials and registries; clinics should contribute standardised outcome data rather than deliver these as routine care.

References

1. Mini Review: Gene Therapy Targets for Aging-Associated Diseases. Elza S, Ilya E, Ilya M, Alexey M. Biochemical and Biophysical Research Communications. 2025;785:152708. doi:10.1016/j.bbrc.2025.152708.
2. Gene Therapy for Aging: Current Evidence and Future Directions. Zhang Y, Gan L, Wang S, Zhang J. Ageing Research Reviews. 2026;118:103144. doi:10.1016/j.arr.2026.103144.
3. Gene Therapy for Aging and Longevity. Everts SPA, Florea M, de Magalhães JP. Trends in Molecular Medicine. 2026;:S1471-4914(26)00116-4. doi:10.1016/j.molmed.2026.05.007.
4. AAV-mediated FGF21 Gene Therapy Promotes Health Span Extension by Whole-Body Tissue-Specific Adaptations. Jimenez V, Sacristan V, Garcia M, et al. Molecular Therapy : The Journal of the American Society of Gene Therapy. 2026;34(8):4546-4568. doi:10.1016/j.ymthe.2026.05.025.
5. Drugs Targeting Mechanisms of Aging to Delay Age-Related Disease and Promote Healthspan: Proceedings of a National Institute on Aging Workshop. Espinoza SE, Khosla S, Baur JA, de Cabo R, Musi N. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences. 2023;78(Suppl 1):53-60. doi:10.1093/gerona/glad034.
6. Turning Back Time: A Comprehensive List of Interventions That Decrease Next-Generation Epigenetic Aging Clocks in Humans. Johnson AA, Sinclair DA. Frontiers in Genetics. 2026;17:1836446. doi:10.3389/fgene.2026.1836446.
7. Epigenetic Reprogramming as a Key to Reverse Ageing and Increase Longevity. Pereira B, Correia FP, Alves IA, et al. Ageing Research Reviews. 2024;95:102204. doi:10.1016/j.arr.2024.102204.
8. Epigenetic Enzymes: A Role in Aging and Prospects for Pharmacological Targeting. Pasyukova EG, Symonenko AV, Rybina OY, Vaiserman AM. Ageing Research Reviews. 2021;67:101312. doi:10.1016/j.arr.2021.101312.
9. Epigenetic Regulation of Aging: Implications for Interventions of Aging and Diseases. Wang K, Liu H, Hu Q, et al. Signal Transduction and Targeted Therapy. 2022;7(1):374. doi:10.1038/s41392-022-01211-8.
10. Responsiveness of Epigenetic Aging Biomarkers to Longevity Interventions in Humans. Sehgal R, Borrus D, Armstrong JF, et al. Nature Medicine. 2026;:10.1038/s41591-026-04562-9. doi:10.1038/s41591-026-04562-9.
11. An Expert Consensus Statement on Biomarkers of Aging for Use in Intervention Studies. Perri G, French C, Agostinis-Sobrinho C, et al. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences. 2025;80(5):glae297. doi:10.1093/gerona/glae297.
12. Senolytics and Senomorphics: Natural and Synthetic Therapeutics in the Treatment of Aging and Chronic Diseases. Lagoumtzi SM, Chondrogianni N. Free Radical Biology & Medicine. 2021;171:169-190. doi:10.1016/j.freeradbiomed.2021.05.003.
13. Cellular Senescence and Senolytics: The Path to the Clinic. Chaib S, Tchkonia T, Kirkland JL. Nature Medicine. 2022;28(8):1556-1568. doi:10.1038/s41591-022-01923-y.
14. Impact of Geroscience on Therapeutic Strategies for Older Adults With Cardiovascular Disease: JACC Scientific Statement. Forman DE, Kuchel GA, Newman JC, et al. Journal of the American College of Cardiology. 2023;82(7):631-647. doi:10.1016/j.jacc.2023.05.038.
15. Nicotinamide Adenine Dinucleotide in Aging Biology: Potential Applications and Many Unknowns. Bhasin S, Seals D, Migaud M, Musi N, Baur JA. Endocrine Reviews. 2023;44(6):1047-1073. doi:10.1210/endrev/bnad019.
16. DNA Methylation Signatures of Cellular Senescence Are Not Reversed by Senolytic Treatment. Kasamoto J, González J, Markov Y, et al. Aging Cell. 2026;25(3):e70430. doi:10.1111/acel.70430.
17. Advancing Geroscience Research – A Scoping Review of Regulatory Environments for Gerotherapeutics. Muscedere J, Shorey CL, Duque G, et al. The Journal of Nutrition, Health & Aging. 2025;29(9):100637. doi:10.1016/j.jnha.2025.100637.
18. A unified framework for systematic curation and evaluation of aging biomarkers. Ying K, Paulson S, Eames A, et al. Nature Aging. 2025;5(11):2323-2339. doi:10.1038/s43587-025-00987-y.