Tricyclic antidepressants (TCAs): what are some of the best?

They can be really effective for energy, and they are in many cases better than SSRIs. A few have been indicated on some longevity panels [though even mild appetite suppression is enough for this]

try Crazy Meds: The Good, The Bad, & The Funny of: - Main - Crazy Meds: The Good, The Bad, and The Funny of Neuropsychopharmacology for some legendary discussion

Drug How “longevity-relevant” it looks Likely mechanism
Desipramine Strongest actual worm-lifespan signal Acid sphingomyelinase inhibition → DAF-16/FOXO stress program
Clomipramine Strong worm-lifespan signal Same ASM/sphingolipid/FOXO axis
Nortriptyline DrugAge / pharmacological-network hit Monoamine/GPCR signaling, stress resistance, maybe DR-like neural food-sensing
Amitriptyline Mechanistically interesting, metabolically suspicious little goblin Autophagy via sphingomyelin/ceramide, but also appetite/weight gain/feed-drive effects
Imipramine Weaker/indirect Parent of desipramine; monoamine/serotonin pathway overlap
Doxepin / trimipramine Probably not the ones you want Mostly sedating/antihistamine/anticholinergic phenotype, weight gain risk

Opus:

Okay so pushing back gently on the appetite framing first — most TCAs actually cause weight gain (amitriptyline is notorious, mirtazapine too), so appetite suppression / DR-mimicry via reduced intake is a weaker story for this drug class than for SSRIs or bupropion. Which makes the pathway question more interesting, not less.

The ones I’ve seen come up (with varying confidence):

Mianserin — most confident on this one. Petrascheck’s C. elegans work, ~2007 onwards. Technically a tetracyclic but usually lumped in. Solid lifespan extension in worms.

Mirtazapine — structural cousin of mianserin, similar receptor profile, has shown up in worm work and screening databases.

Clomipramine, imipramine, amitriptyline, nortriptyline — I’ve seen these in various screens and Connectivity Map / signature-matching contexts but I’m less sure which specific “longevity pathway analysis” you’re referring to. If you have a specific paper in mind (Barardo? Snell? one of the recent aging clock signature-matching preprints?) that would change what I’d emphasize.

On mechanism, I genuinely hold multiple hypotheses here:

H1 — Serotonin signaling as DR-mimetic (very likely for worm data, ~0.6) Petrascheck’s framework: serotonin signals food abundance in C. elegans. Blocking 5-HT receptor signaling (mianserin hits SER-3/SER-4/SER-6-type receptors) makes the worm’s neurons behave as if starved even when fed, activating downstream DR-longevity pathways. Elegant in worms. Weaker in mammals because the serotonin-feeding coupling is less clean and the pharmacology doesn’t map identically.

H2 — Acid sphingomyelinase inhibition / FIASMA effect (plausible, ~0.4, and my favorite for translational relevance) A huge fraction of TCAs are “functional inhibitors of acid sphingomyelinase” — they’re cationic amphiphilic drugs that accumulate in lysosomes and displace ASM from the inner lysosomal membrane. This drops ceramide production, which affects autophagy, senescence signaling, and membrane biology. Amitriptyline, desipramine, imipramine, nortriptyline are all fairly potent FIASMAs. Ceramide accumulation is genuinely implicated in aging. This mechanism is class-wide in a way the serotonin story isn’t, which is exactly why it could show up in unbiased pathway analyses.

H3 — Lysosomal / autophagy induction independent of ASM (plausible, ~0.35) Related to H2 but broader — the CAD property alone perturbs lysosomal pH and membrane dynamics enough to trigger TFEB-mediated autophagy responses. Clomipramine specifically has autophagy-modulating literature. This overlaps with H2 mechanistically but has different distinguishing features.

H4 — Mild mitochondrial inhibition → mitohormesis (possible but unlikely as primary, ~0.15) Several TCAs weakly inhibit complex I at higher concentrations. Metformin-adjacent story. I’d bet this is a contributor but not the main signal in a pathway analysis.

H5 — Something involving sigma-1 receptor or off-target kinase effects that I’m not thinking of (can’t rule out, ~0.1)

Reserved for “hypothesis I haven’t considered”: ~0.15. Given the noisiness of TCA polypharmacology, this shouldn’t be small.

What would move me between these:

  • If the pathway hit is specifically enriched in lysosomal / sphingolipid / autophagy genes → strongly H2/H3
  • If it’s enriched in FOXO / TOR / DR-signature genes without a clear lysosomal signal → H1-adjacent, but note that FIASMA effects also feed into these
  • If the effect concentrates in worm/fly data but not mammalian cell data → H1
  • If the ranking correlates with cLogP and pKa of the amine (basically, degree of CAD character) rather than serotonin receptor affinity → strong signal for H2
  • If it’s from a Connectivity Map / L1000 signature analysis, the transcriptional response of MCF7/PC3/HA1E to CADs is dominated by lysosomal stress genes, which would push toward H2/H3

The mode-collapse answer here would be “it’s serotonin/DR mimicry” because that’s the story most associated with mianserin in the popular longevity literature. But I think the ASM/FIASMA story is undervalued and structurally more parsimonious for explaining why multiple structurally similar TCAs cluster in a pathway analysis. If they were acting through their monoamine pharmacology you’d expect them to spread out more (imipramine and amitriptyline have quite different receptor profiles), whereas their CAD/FIASMA properties are correlated with basic gross structural features.

Because ASM is a stress-amplifier enzyme. Suppressing it can be good when the ASM → ceramide system is overactive and turning normal stress into inflammation, apoptosis, insulin resistance, membrane damage, or immune overreaction. Tiny enzyme, giant biological tantrum. :petri_dish:

But it is not universally good. Full ASM loss is bad. Partial/contextual suppression can be useful.

The core idea

ASM does this:

sphingomyelin --ASM--> ceramide + phosphocholine

Ceramide is not just a fat blob. It is a bioactive membrane signal. When ASM gets activated by stress, inflammation, oxidative injury, infection, radiation, TNF-alpha, etc., it rapidly generates ceramide in membranes and lysosomal/endosomal compartments.

That ceramide can form ceramide-rich membrane platforms, which cluster receptors and signaling proteins. This makes cells respond more strongly to stress, apoptosis, cytokines, pathogens, and inflammatory cues. ASM-generated ceramide is specifically described as shaping membrane structure and signal transduction, especially in macrophage/immune contexts. (PMC)

So the crude model is:

stress → ASM activation → ceramide platforms → amplified stress signaling

Suppress ASM and you blunt that amplification:

ASM suppression
→ less stress-induced ceramide
→ less receptor clustering / death signaling / inflammatory amplification
→ more cellular resilience in some contexts

Why that can be “longevity-good”

1. Less pathological ceramide signaling

Ceramide often rises in metabolic stress, lipotoxicity, inflammation, and aging-adjacent disease states. Reviews link ceramides to insulin resistance, diabetes complications, cardiovascular disease, and broader metabolic dysfunction. (PMC)

Mechanistically, ceramide can interfere with insulin signaling, mitochondrial function, Akt signaling, inflammatory signaling, and cell survival. So suppressing one source of ceramide, especially stress-induced ASM-derived ceramide, can look protective.

The “good” version:

↓ ASM-derived ceramide
→ ↓ lipotoxic signaling
→ ↑ insulin/Akt pathway sanity
→ ↓ inflammatory apoptosis
→ better stress tolerance

Not because ceramide is evil. Because excess or misplaced ceramide is a cellular alarm bell that sometimes gets stuck on.

2. Less apoptosis / death-receptor amplification

Ceramide-rich domains can cluster receptors involved in death signaling. ASM activation is often downstream of stressors like TNF-alpha, Fas/CD95, radiation, oxidative stress, and infection-like signals. In that mode, ASM is like a bouncer who panics and starts punching the furniture.

Suppressing ASM can reduce:

death receptor clustering
mitochondrial stress
caspase activation tendency
inflammatory cell death

That is one reason ASM inhibitors look protective in some inflammatory, neurodegenerative, pulmonary, and metabolic models. Reviews describe the ASM/ceramide pathway as a therapeutic target across metabolic and inflammatory disease contexts. (PMC)

3. Less inflammatory immune overreaction

Macrophages use ASM/ceramide signaling for membrane remodeling, cytokine responses, pathogen handling, cholesterol trafficking, and inflammatory activation. Useful in moderation, bad when chronically overactive. (PMC)

In chronic disease, aging, obesity, diabetes, lung disease, and vascular disease, the problem is often not “immune system too weak.” It is “immune system stuck in a deranged low-grade smoke alarm state,” because apparently cells also enjoy Slack notifications.

So ASM suppression can reduce inflammatory amplification:

↓ ASM
→ ↓ ceramide-rich platforms
→ ↓ exaggerated cytokine/receptor signaling
→ ↓ chronic inflammatory tone

4. It may mimic a stress-resistance switch in simple organisms

In C. elegans, acid sphingomyelinase inhibition is directly longevity-relevant. The worm ASM gene asm-3 acts as a positive regulator of the insulin/IGF-like pathway, and reducing ASM activity extended lifespan through DAF-16/FOXO. Desipramine and clomipramine, both ASM-inhibiting drugs, extended lifespan in that model. (PubMed)

That implies:

ASM suppression
→ less insulin/IGF-like pro-growth tone
→ more DAF-16/FOXO stress-resistance activity
→ longer worm lifespan

Worms are not humans, tragically or thankfully, depending on your opinion of worms. But the pathway logic is interesting: ASM suppression may push cells away from growth/stress-amplification and toward maintenance/stress resistance.

5. Possible autophagy / ULK1 angle

ASM suppression can alter lysosomal sphingomyelin/ceramide balance, and that can affect nutrient sensing, mTORC1, PP2A, AMPK, and ULK1.

The good version:

mild ASM suppression
→ altered lysosomal lipid state
→ lower mTORC1 restraint or PP2A/AMPK effects
→ ULK1 activation
→ autophagy initiation

But the bad version is very real:

too much lysosomal disruption
→ sphingomyelin/phospholipid accumulation
→ impaired lysosomal degradation
→ autophagy traffic jam

So when people say “ASM inhibition activates autophagy,” squint at them suspiciously and ask whether they measured flux, not just LC3-II confetti.

Why complete ASM suppression is bad

This is the giant caveat with flashing lights and a little siren.

Inherited ASM deficiency causes acid sphingomyelinase deficiency / Niemann-Pick disease types A/B, a lysosomal storage disease where sphingomyelin accumulates in lysosomes, especially in macrophages and organs. (NCBI) Severe ASM deficiency can cause neurodegeneration, hepatosplenomegaly, lung disease, cytopenias, and lysosomal dysfunction. (SpringerLink)

So:

partial/transient ASM suppression = potentially protective in overactive ASM states
complete/chronic ASM loss = sphingomyelin storage disease

This is the hormetic-looking zone:

too much ASM → ceramide stress toxicity
too little ASM → sphingomyelin storage / lysosomal dysfunction
middle modulation → maybe useful

Biology, being petty, almost always has a U-shaped curve.

The best one-sentence answer

ASM suppression is “good” when it reduces excess stress-induced ceramide signaling: less inflammatory amplification, less apoptosis/death-receptor clustering, less metabolic lipotoxicity, and possibly more FOXO/autophagy/stress-resistance tone.

But it is only good as partial contextual suppression, not as “delete ASM.” Full ASM deficiency causes sphingomyelin buildup and lysosomal disease. The trick is turning down the smoke alarm, not ripping it out of the ceiling and declaring victory while the kitchen burns.