# SGLT2 inhibitors

Rating: Early — Mostly animal or cell studies; few or uncontrolled human data
Type: Compound
Updated: 2026-10-14
Canonical: https://heretic.bio/articles/sglt2-inhibitors-as-senotherapeutics-evidence-taxonomy.html

> SGLT2 inhibitors are approved diabetes drugs that unexpectedly engage aging biology: AMPK and SIRT1 activation, SASP suppression, and immune-mediated clearance of senescent cells shown in cell and mouse studies. The human evidence is large but indirect: EMPA-REG showed a 32% reduction in all-cause mortality in diabetics with cardiovascular disease, and one small RCT found leukocyte telomere lengthening at 26 weeks. No trial has tested senescence or healthspan endpoints in healthy people, and canagliflozin shortened female mouse lifespan in the ITP dataset.

> Sodium-glucose co-transporter 2 (SGLT2) inhibitors — a drug class already proven to reduce cardiovascular mortality and kidney failure in large RCTs — unexpectedly converge on core aging biology: they clear senescent cells, suppress the SASP, activate AMPK/SIRT1, and in one small RCT lengthened leukocyte telomeres in humans within 26 weeks. A 2025 _npj Aging_ review positions them as "a novel senotherapeutic approach." The honest caveat: the senescence-specific evidence is preclinical; the human RCT evidence is in diabetic/diseased populations; and one lifespan study found the drug _shortened_ female mouse lifespan while extending males'.

**Where this sits in this library:** this is the drug-class companion to [Senolytics](senolytics.html) and [Metformin and Rapamycin](metformin-and-rapamycin.html) — a cheap, generic, already-approved geroprotective _candidate_ whose anti-senescence biology overlaps the Dasatinib+Quercetin/fisetin agenda but with vastly more human outcome data (in disease populations).

## Mechanism Primer

### Primary Action (What They Were Built For)

  * The class mimics aspects of caloric restriction: mild glycosuria, shifted substrate metabolism (ketogenesis, fat oxidation), lower insulin/IGF-1 signaling — see [Fasting Protocols](fasting-approaches.html) and [mTOR and AMPK Pathways](mtor-and-ampk-pathways.html) for the pathway logic



### The Senotherapeutic Layer (Why They're Now Studied as Anti-Aging Drugs)

Pathway | Effect | Evidence Base  
---|---|---  
**AMPK activation** | Reduces ROS, improves mitochondrial function, anti-senescence signal | Cell + animal  
**SIRT1 upregulation** | Anti-senescence effect is _blunted by SIRT1 inhibitors_ (causality shown) | Cell (HUVEC, HK-2) + mice  
**Senescent cell reduction** | ↓ SA-β-gal, p16, p21, p53 in endothelium, heart, aorta, kidney, adipose | Mice (multiple models)  
**SASP suppression** | ↓ IL-1β, IL-6, IL-8, TNF-α, TGF-β1 in high-glucose-exposed cells | Human primary cells (RPTEC)  
**NO restoration** | ↑ eNOS phosphorylation, improved vasodilation | db/db mice, human coronary endothelial cells  
**Autophagy modulation** | Corrects hyperactivated autophagy via AMPK/GSK3β; links to [Autophagy](autophagy.html) | Diabetic cardiomyopathy mice  
  
**Key in-vitro detail:** high glucose _induces_ senescence in human renal tubular cells (↑γ-H2AX DNA damage, ↑SASP); dapagliflozin normalizes all of it. SGLT2 expression itself rises with inflammation (IL-1β/IL-6/TNF-α from COVID-19 patient plasma upregulates endothelial SGLT2), suggesting a feed-forward loop the drug breaks.

## Evidence Taxonomy (Human vs Animal vs Cell)

### Human RCT Evidence — Indirect but Massive

Trial | Population | Key Result  
---|---|---  
**EMPA-REG OUTCOME** (NEJM 2015; PMID [26378978](https://pubmed.ncbi.nlm.nih.gov/26378978/)) | T2D + established CVD | 38% ↓ CV death, 35% ↓ HF hospitalization, 32% ↓ all-cause mortality — the result that shocked cardiology  
**EMPEROR-Reduced / -Preserved** (NEJM 2020; PMID [32865377](https://pubmed.ncbi.nlm.nih.gov/32865377/)) | HFrEF + HFpEF, incl. **non-diabetics** | Benefit in non-diabetics ⇒ effect is not glucose-mediated  
**DAPA-CKD** (NEJM 2020; DOI [10.1056/NEJMoa2024816](https://doi.org/10.1056/NEJMoa2024816)) | CKD ± diabetes | 44% ↓ composite renal death/decline; benefits independent of glycemia  
**Zhang et al. 2025** (_Cell Reports Medicine_ ; [PMC12490245](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12490245/)) | T2D, henagliflozin 10 mg × 26 wk | **First RCT of any drug showing leukocyte telomere lengthening in humans** (90.5% of treatment arm lengthened vs 65.6% placebo; +0.06, p=0.011) — full analysis in [Telomere Science](telomere-science.html)  
  
**Honest reading:** outcome RCTs are in _disease_ populations; the telomere RCT is small, single-agent (henagliflozin, approved mainly in China), in diabetics. No SGLT2i trial with lifespan/healthspan primary endpoint in healthy non-diabetics exists (unlike TAME for metformin).

### Animal Evidence

  * **Lifespan:** canagliflozin and empagliflozin extend male mouse lifespan; canagliflozin works even when started at 16 months of age
  * ⚠️ **Sex-specific toxicity:** the same group later showed canagliflozin **significantly reduced female mouse lifespan** — mechanism unclear; a standing warning against extrapolating male rodent data to humans of both sexes
  * **Nature Aging 2024** (DOI [10.1038/s43587-024-00642-y](https://doi.org/10.1038/s43587-024-00642-y)): canagliflozin "eliminates senescent cells and alleviates pathological aging" in high-fat-diet mice — ↓SA-β-gal, ↓cell-cycle regulators, ↓SASP factors in adipose/liver
  * **Atherosclerosis:** canagliflozin reduced senescent cell burden and plaque in ApoE-knockout mice
  * **Progeria model:** canagliflozin delayed premature aging in Zmpste24-knockout mice
  * **Neuro:** dapagliflozin improved learning/memory in D-gal-aged mice (↑SOD, ↑SIRT1, ↑PGC-1α in brain)



### Cell / In-Vitro Evidence

  * Human proximal tubule cells (HK-2), HUVECs, human coronary arterial endothelial cells: consistent anti-senescence across dapagliflozin, empagliflozin, canagliflozin — appears to be a **class effect** via SIRT1/AMPK
  * Dapagliflozin anti-senescence effect disappears with SIRT1 inhibition (EX-527) — cleanest causality data in the field



## Risk Ledger (Why This Isn't a Free Lunch)

Risk | Magnitude | Notes  
---|---|---  
**Euglycemic DKA** | Rare, serious | Ketone production + low insulin ⇒ risk spikes when combined with **keto diet, fasting, or heavy alcohol** — directly relevant to biohackers stacking this with [Fasting Protocols](fasting-approaches.html)  
Genital mycotic infections / UTI | Common (5–10%) | Glycosuria feeds yeast; women affected more  
Volume depletion / hypotension | Common early | Additive with diuretics  
Fournier's gangrene | Very rare | Class warning  
Fractures / ↓ bone density | Signal (canagliflozin) | Consider with [Magnesium Types and Protocols](magnesium-types.html) / bone health context  
**Female lifespan reduction** | Mouse data | Unknown human relevance; do not assume class is sex-neutral  
Off-label longevity use | Unproven | No non-diabetic healthy-human trial data at all  
  
## Open Questions

  1. Does the senescence effect occur at doses used in humans, or only in rodent dosing equivalents?
  2. Is henagliflozin's telomere effect shared by Western-market agents (empagliflozin/dapagliflozin), and does it persist past 26 weeks?
  3. Why did canagliflozin shorten female mouse lifespan — and does it translate? (See 2026 update below — partially answered.)
  4. Does SGLT2i + senolytic (D+Q/fisetin) stack additively or redundantly? (See 2026 update below — Navitoclax+SGLT2i preclinical only.)
  5. Do the class's CR-mimetic effects interact with [Rapamycin Dosing Protocols](rapamycin-mtor-inhibition-pearl-trial-evidence-and-clinical-translation.html) (both push the same metabolic direction)?



## 2026 Update (since last revision)

### A. Katsuumi & Minamino — JCEM follow-up review (Aug 2026)

**Katsuumi G, Minamino T. (2026).** _Dissecting out the unexpected effects of SGLT2 inhibitors on human aging._ J Clin Endocrinol Metab 111:2154–2165. PMID [42132232](https://pubmed.ncbi.nlm.nih.gov/42132232/). \- The same Juntendo group that published the foundational Nature Aging 2024 mechanistic study (PD-L1/AMPK/immune-surveillance → senescent cell clearance) returns with a clinical-translation-focused review. Headline framing: SGLT2i are now positioned as **"indirect senolytics"** because they don't kill senescent cells directly (Extended Data Fig. 5a of the 2024 paper confirmed no direct cytotoxicity even at high concentrations) — instead they rewire PD-L1 on senescent cells so endogenous NK/T-cell immunosurveillance clears them. \- The review formalizes a **two-step clearance model** : (1) AMPK activation (via AICAR elevation) → PD-L1 downregulation on senescent cells → (2) T-cell re-engagement → senescent cell apoptosis. CD3-depleting antibody experiments blocked the effect in vivo. \- Implication: SGLT2i senolytic action **requires intact immune function**. This is a previously unrecognized constraint and matters for the biohacker demographic — immunosenescence may blunt the benefit in the very old (≥80), who ironically have the most senescent cell burden.

### Uses

**Manchanayake DH et al. (2026).** _Dose and sex-specificity in Canagliflozin-mediated lifespan extension._ bioRxiv 2026.07.02.734998. DOI [10.64898/2026.07.02.734998v1](https://doi.org/10.64898/2026.07.02.734998v1). \- Re-analysis of the ITP UM-HET3 heterogeneous mouse stock confirms: **canagliflozin extends male median lifespan by ~14%; no statistically significant effect in females** (the earlier JCI Insight 2020 PMID [33048855](https://pubmed.ncbi.nlm.nih.gov/33048855/) finding holds across cohorts). \- Leading mechanistic explanation (correlative): **male UM-HET3 mice have higher baseline fasting glucose and insulin** than females at the age of treatment onset — i.e., SGLT2i has more metabolic dysregulation to correct in males. Hormonal modulation of SGLT2 expression is a contributing factor. \- Practical takeaway: do **not** assume the class is sex-neutral. Until human data on sex-stratified outcomes is in (no current trial reports it), female-specific longevity claims should be treated as unproven.

### C. Frontiers Endocrinology March 2026 — SGLT2i in senolytic combination strategies

**Syed AR, Rangraze I et al. (2026).** _Cellular senescence and metabolic aging in type 2 diabetes: mechanistic insights and translational implications._ Front Endocrinol 17:1799261. DOI [10.3389/fendo.2026.1799261](https://doi.org/10.3389/fendo.2026.1799261). \- The most current comprehensive review on T2D-as-metabolic-aging. Notable: lists **Navitoclax + empagliflozin/dapagliflozin** as one of three "experimental combinations" for senescent renal/adipose cell clearance (preclinical; apoptosis induction + glucosuria). Translational progress blocked by **Navitoclax thrombocytopenia/neutropenia** (BCL-xL on platelets). \- Review classifies SGLT2i as **"senomorphic"** (SASP suppression) more than direct senolytic — the distinction matters for trial design. \- Authors explicitly note: "no human clinical trials have yet evaluated" SGLT2i + senolytic combination in T2DM. Regulatory agencies (FDA/EMA) classify all such combinations as **investigational**.

### D. Simpson 2026 — frailty-stratified outcome review

**Simpson M et al. (2026).** _SGLT2 inhibitors in frailty: a review of current evidence._ [PMC13410979](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13410979/). \- Re-analysis of DAPA-HF (32-item FI) and EMPEROR-Preserved (44-item FI) outcome RCTs across frailty tertiles: **cardiovascular benefit is preserved across all frailty strata** (HR for worsening HF/CVD 0.71–0.77 in most-frail tertile of DAPA-HF; OR 1.12 for being in a lower frailty category in EMPEROR-Preserved). \- 2025 PLOS Medicine IPD meta-analysis (34 trials) confirms: only modest attenuation in HbA1c efficacy with rising FI (0.08%-point smaller reduction per 0.1-point FI increase), well below clinical significance — i.e., SGLT2i **work in frail people** , not just robust ones. \- Cognitive function sub-analyses in frail HFpEF/diabetic populations: empagliflozin improved 5 m gait speed and cognitive scores vs control over 3 months — small studies, hypothesis-generating.

### E. [NCT05975528](https://clinicaltrials.gov/study/NCT05975528) — first dedicated "SGLT2i for senescence" trial in T2D

  * Active, recruiting (per ClinicalTrials.gov). Designed specifically to test whether dapagliflozin reduces senescence biomarkers (p16^INK4a, SASP panel) in older T2D patients with frailty. **No results yet** — but this is the first trial designed with senescence as the _primary_ endpoint rather than glucose/CV outcome. Watch for read-out in 2026-2027.



## Practical Notes (Non-Prescriptive)

  * This file does **not** constitute dosing guidance; SGLT2 inhibitors are prescription drugs with a real adverse-effect profile
  * The class is cheap and generic in most markets; legitimate access requires a physician — typically only for T2D, HF, or CKD indications
  * Anyone (especially low-carb/fasting) using an SGLT2i should understand euglycemic DKA symptoms: nausea, vomiting, abdominal pain, malaise despite normal glucose
  * Biomarker-oriented physicians track: HbA1c, ketones, eGFR, BP, and — in research contexts — p16^INK4a, SASP panels, leukocyte telomere length (see [Telomere Science](telomere-science.html) testing table)

## Sources

1. Zinman B et al. (2015). EMPA-REG OUTCOME. NEJM 373:2117–28. PMID 26378978 . (PubMed 26378978)
2. Packer M et al. (2020). EMPEROR-Reduced. NEJM 383:1413–24. PMID 32865377 . (PubMed 32865377)
3. Heerspink HJL et al. (2020). DAPA-CKD. NEJM . DOI 10.1056/NEJMoa2024816 . (DOI 10.1056/NEJMoa2024816)
4. Zhang et al. (2025). Henagliflozin telomere RCT. Cell Reports Medicine . PMC12490245 . (PMC12490245)
5. SGLT2 inhibition eliminates senescent cells and alleviates pathological aging. Nature Aging (2024). DOI 10.1038/s43587-024-00642-y . PMID 38816549 . (DOI 10.1038/s43587-024-00642-y, PubMed 38816549)
6. Katsuumi G, Minamino T. (2026). Dissecting out the unexpected effects of SGLT2 inhibitors on human aging. J Clin Endocrinol Metab 111:2154–2165. PMID 42132232 . — JCEM clinical-translation follow-up to Nature Aging 2024. (PubMed 42132232)
7. Manchanayake DH et al. (2026). Dose and sex-specificity in Canagliflozin-mediated lifespan extension. bioRxiv 2026.07.02.734998. DOI 10.64898/2026.07.02.734998v1 . — confirms male-specific ITP lifespan effect. (DOI 10.64898/2026.07.02.734998v1)
8. Miller RA, Harrison DE, Astle CM et al. (2020). Canagliflozin extends life span in genetically heterogeneous male but not female mice. JCI Insight 5(21):e140019. PMID 33048855 . — original ITP finding, sex-specific. (PubMed 33048855)
9. Syed AR, Rangraze I et al. (2026). Cellular senescence and metabolic aging in type 2 diabetes. Front Endocrinol 17:1799261. DOI 10.3389/fendo.2026.1799261 . — T2D-as-metabolic-aging review; Navitoclax+SGLT2i combination. (DOI 10.3389/fendo.2026.1799261)
10. Simpson M et al. (2026). SGLT2 inhibitors in frailty: a review of current evidence. PMC13410979 . (PMC13410979)
11. NCT05975528 . Effect of SGLT2 inhibition on cellular senescence in older T2D patients with frailty. ClinicalTrials.gov (recruiting). (NCT05975528)
12. Yesilyurt-Dirican ZE, Qi C, Wang Y-C, et al. (2025). "SGLT2 inhibitors as a novel senotherapeutic approach." npj Aging 11:35. DOI 10.1038/s41514-025-00227-y . PMC12065912 . — the review this file was built on (full PDF archived in Sources/Papers/ ) (DOI 10.1038/s41514-025-00227-y, PMC12065912)