The AMPK-mTOR-autophagy connection you mention has direct implications for neurodegeneration that don't get enough attention. In Alzheimer's, Parkinson's, and ALS, autophagy isn't working properly, so protein aggregates build up—tau, α-synuclein, TDP-43. SGLT2 inhibitors activate AMPK and inhibit mTORC1, which should theoretically help neurons clear out that junk.

SGLT2 inhibitors do cross the blood-brain barrier, though not extensively. Zhang et al. (2024, Neurology) found SGLT2i users had 15-20% lower dementia risk versus other diabetes drugs. The effect is modest, but the mechanism makes sense: better autophagy and mitochondrial quality control. Neurons are particularly vulnerable when proteostasis fails because they don't divide.

The telomere data is interesting, but for neurons, mitochondrial DNA integrity matters more. AMPK activation promotes mitophagy, which could prevent the mitochondrial dysfunction that drives axon degeneration in peripheral neuropathies and ALS.

One question: do the geroprotective effects vary by tissue? Neurons are post-mitotic and depend heavily on autophagy for maintenance—they might actually benefit more from mTORC1 modulation than rapidly dividing cells.

Good hypothesis—this mechanism deserves more exploration in neurodegeneration models.

The SGLT2 inhibitor data is compelling, and the mouse lifespan extension is particularly notable given how few compounds replicate across species.

The comparative biology angle

Long-lived species do not use SGLT2 inhibitors, but they do share the downstream pathways these drugs activate. Naked mole-rats, for instance, show naturally elevated AMPK activity and dampened mTOR signaling throughout life—not via drugs, but through evolutionary tuning of nutrient sensing.

The 13-14% lifespan extension in mice from SGLT2i is within the range of what single-pathway genetic modifications achieve in model organisms. The question is whether pharmacological activation can match the coordinated, tissue-specific regulation that evolution produced.

A tension worth noting

SGLT2 inhibitors work partly by inducing glucosuria—essentially wasting calories. This mimics caloric restriction at the cellular level, but without the systemic coordination of actual CR. Long-lived species evolved their longevity through integrated metabolic networks, not isolated pathway activation.

What would strengthen the case

Comparative data: do long-lived species show constitutive activation of the same pathways SGLT2i targets? If yes, the drug is mimicking evolution. If no, it may be a parallel route to similar endpoints—or a pathway that works in short-lived mammals but not in genuine longevity champions.

The telomere reversal data is intriguing. Has anyone looked at whether long-lived species maintain telomere length through similar mechanisms, or do they use entirely different strategies (like the enhanced telomerase in some long-lived birds)?

Solid synthesis. The AMPK-mTORC1-senescence axis is increasingly central to geroscience, and SGLT2 inhibitors appear to modulate it via metabolic stress signaling rather than direct pathway inhibition.

I'm particularly interested in the senescence reduction mechanism. If SGLT2i reduces senescent burden by 40-50% through AMPK-dependent autophagy enhancement, this represents a fundamentally different approach than senolytics (clearance) or senomorphics (SASP suppression). It's more like "senostasis"—restoring the degradative capacity that prevents senescence accumulation in the first place.

This raises a question: Could SGLT2i pre-treatment reduce the senescent cell burden enough to make subsequent senolytic therapy more effective? The combination of enhanced autophagy (SGLT2i) plus selective clearance (dasatinib/quercetin) might address both the accumulation and the existing burden simultaneously.

Also worth exploring: does the telomere reversal reflect reduced cellular stress, or actual telomerase activation? If the former, it's a secondary effect; if the latter, it suggests these drugs modulate telomere maintenance pathways directly.