🧬 Aging Wiki

interventions-by-modality

Properties1
aliasesintervention modality map, interventions by mode, modality MOC, interventions by delivery

9 min read

Interventions by modality — synthesis MOC

This page answers: “What interventions exist in modality X, and how do modalities compare for the same target?”

Companion to interventions-by-hallmark (axis: which hallmark) and intervention-classes (axis: mechanism). This page slices on the delivery modality (cell vs. drug vs. dietary pattern vs. gene-therapy vector vs. blood product). The same atomic intervention page may appear once here and once on each sister MOC; truth lives on the atomic pages, not here.

Mode-A discipline: this page is pure aggregation. No quantitative claims originate here. Each Dataview block queries interventions/ and molecules/compounds/ for the corresponding mode: value.

Round seeded: R23e (2026-05-06). Replaces the per-folder _overview.md design considered in R23a kickoff; matches the established R16 framework-page convention.

Modality taxonomy

The wiki recognises six canonical mode: values on type: intervention frontmatter (per CLAUDE.md):

ModalityDelivery substrateReversibilityRegulatory archetypeExample anchor
lifestylebehavioural — exercise, sleep, social, environmentalhigh (stop the behaviour)non-medical (consumer)exercise
dietarynutrition — fasting protocols, macronutrient patternshigh (resume normal eating)non-medical (consumer); some medical-food pathwayscaloric-restriction
pharmacologicalsmall-molecule + biologic drugsmedium (drug-PK-dependent)FDA / EMA / PMDA NDA-or-BLA pathwayrapamycin
gene-therapyviral-vector + DNA-modifying interventionslow (durable; integration or chronic expression)FDA RMAT / advanced-therapy regulationaav-tert
stem-cell-therapylive-cell or cell-derivative graftslow (engraftment durable; paracrine transient)FDA RMAT / cellular-therapy pathwayhematopoietic-stem-cell-transplantation
blood-productapheresis-based plasma manipulationmedium (per-session)FDA / blood-bank regulationplasma-exchange

Boundaries are not always sharp:

  • in-vivo-partial-reprogramming-therapy uses gene-therapy vectors but acts via a stem-cell-biology mechanism (cell-identity manipulation) — classified stem-cell-therapy per R23a decision; the AAV-vector view is on aav-osk (gene-therapy).
  • plasma-exchange involves no transplanted cells but uses heterochronic-parabiosis biology — classified blood-product after R21 schema escalation.
  • iPSC-derived β-cell grafts are stem-cell-therapy even though the cellular product is heavily engineered ex vivo; the modality marker tracks the therapeutic agent, not the manufacturing process.

When in doubt, classify by what physically enters the patient (drug / cells / DNA / plasma / nothing — i.e., behavioural).

Lifestyle

Behavioural interventions: nothing administered, but a sustained change in activity, sleep, or environmental exposure that the body responds to. Highest reversibility, lowest regulatory friction, but adherence is the central problem.

TABLE WITHOUT ID file.link AS Intervention, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", translation-gap AS "Gap"
FROM "interventions"
WHERE mode = "lifestyle"
SORT human-evidence-level DESC, file.name ASC

Cross-modality observation: lifestyle interventions overlap mechanistically with dietary (the meal-timing protocols straddle both); we classify pure-behaviour as lifestyle (exercise, sleep, heat-exposure) and food-composition-or-timing as dietary. Lifestyle interventions have the strongest population-evidence base (decades of cohort data) but the weakest mechanistic-RCT base.

Dietary

Nutrition-pattern interventions: fasting protocols, macronutrient compositions, specific-amino-acid restriction. Cohort + small-RCT evidence is strong for cardiometabolic markers; longevity-RCT evidence is uniformly absent.

TABLE WITHOUT ID file.link AS Intervention, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", translation-gap AS "Gap"
FROM "interventions"
WHERE mode = "dietary"
SORT human-evidence-level DESC, file.name ASC

Cross-modality observation: all dietary interventions converge on a few molecular axes — mtor inhibition, ampk activation, autophagy induction, sirtuin activation, fgf21-mediated stress signaling — with caloric-restriction as the canonical reference. The pharmacological-class axis (mtor-inhibitors, ampk-activators, nad-precursors) reproduces these signals via small molecules. Choosing dietary vs. pharmacological is a question of adherence, side-effect profile, and durability rather than mechanism.

Pharmacological

Small-molecule and biologic drugs. Largest historical R&D infrastructure; clearest regulatory pathway; mechanism-class is captured in detail on intervention-classes.

TABLE WITHOUT ID file.link AS Intervention, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", clinical-stage AS Stage, translation-gap AS "Gap"
FROM "interventions"
WHERE mode = "pharmacological"
SORT clinical-stage DESC, file.name ASC
TABLE WITHOUT ID file.link AS Compound, mechanisms AS Mechanisms, hallmarks AS Hallmarks, human-evidence-level AS "Evidence", clinical-stage AS Stage
FROM "molecules/compounds"
WHERE clinical-stage
SORT clinical-stage DESC, file.name ASC

Cross-modality observation: the pharmacological space subdivides cleanly by mechanism (covered on intervention-classes — senolytics, senomorphics, mTOR-inhibitors, NAD-precursors, AMPK-activators, sirtuin-activators, GLP-1 agonists, etc.). The class pages anchor specific compound pages; the mechanism MOC is the right entry point for “what hits hallmark X via pharmacology”. This block is here for modality-symmetry but is functionally redundant with the class MOC.

Gene therapy

Viral-vector + DNA-modifying interventions. Highest theoretical durability (single-administration permanent edit) but highest cancer + immunogenicity risk and the steepest regulatory bar. Modality is in early ascent: VERVE-101 / VERVE-102 (PCSK9 base-editing) is the first in-human in-vivo somatic-base-editing program with durable cardiovascular geroprotection as the implicit goal.

TABLE WITHOUT ID file.link AS Intervention, targets AS Targets, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", clinical-stage AS Stage
FROM "interventions"
WHERE mode = "gene-therapy"
SORT clinical-stage DESC, file.name ASC

Cross-modality observation: gene-therapy modality fragments by vector + edit-type:

VectorEdit typeDurabilityAging-relevant programs
AAV (serotypes 2/9/PHP.eB)Episomal expression cassetteMonths–years (non-integrating); persistent in non-dividing cellsaav-tert (telomerase), aav-klotho (Klotho), aav-follistatin (myostatin antagonism), aav-osk (Yamanaka-factor pulsing)
LNP-Cas9-base-editorPermanent single-base substitutionLifetime (DNA-level edit)crispr-base-editing-pcsk9 (VERVE)
mRNA-LNPTransient protein expression (hours–days)Single-dose; repeat as neededin-vivo-partial-reprogramming-therapy mRNA-LNP-OSK arm; multiple preclinical

The AAV-vector deep dive is on aav-tert (precedent) and aav-osk (most R23b detail). Cancer risk is the load-bearing translation barrier across the modality.

Stem-cell therapy

Live-cell + cell-derivative grafts. Decades of clinical experience for hematologic disease (hematopoietic-stem-cell-transplantation); aging-specific use is overwhelmingly preclinical or early-phase.

TABLE WITHOUT ID file.link AS Intervention, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", clinical-stage AS Stage, translation-gap AS "Gap"
FROM "interventions"
WHERE mode = "stem-cell-therapy"
SORT clinical-stage DESC, file.name ASC

Cross-modality observation — three sub-modalities that should not be conflated:

  1. Engraftment-based replacement (hematopoietic-stem-cell-transplantation, some ipsc-derived-cell-therapy applications like iPSC-RPE / iPSC-DA / iPSC-β) — the transplanted cells durably take up residence and provide function. Highest mechanistic confidence; highest cancer/teratoma surveillance burden.
  2. Paracrine / immunomodulatory (mesenchymal-stem-cell-therapy) — transplanted cells largely don’t engraft long-term; they secrete signaling molecules, modulate immune state, and are cleared. Easier safety profile; mechanism is contested (“Medicinal Signaling Cell” reframe per Caplan 2017).
  3. Cell-identity manipulation in situ (in-vivo-partial-reprogramming-therapy) — no cells are transplanted; instead, host cells are temporarily reprogrammed via vector-delivered Yamanaka factors. Modality-conventional classification puts this under stem-cell-therapy because the biological state change is cell-identity reprogramming, but the delivery is gene-therapy-typed. The narrower aav-osk page covers the AAV-specific implementation.

The three sub-modalities target different hallmarks (replacement → stem-cell-exhaustion; paracrine → chronic-inflammation + altered-intercellular-communication; reprogramming → epigenetic-alterations).

Blood product

Apheresis-based interventions: therapeutic plasma exchange, neutral blood exchange, young-plasma transfer, plasma-derived fractions. Conceptually anchored in heterochronic-parabiosis biology (R18 backlog); clinically anchored in plasma-exchange (verified) including the AMBAR mild-AD signal.

TABLE WITHOUT ID file.link AS Intervention, mechanisms AS Mechanisms, target-hallmarks AS Hallmarks, human-evidence-level AS "Evidence", clinical-stage AS Stage, translation-gap AS "Gap"
FROM "interventions"
WHERE mode = "blood-product"
SORT clinical-stage DESC, file.name ASC

Cross-modality observation: blood-product is the smallest modality bucket (one page as of R23 close) but is operationally simple — TPE infrastructure is FDA-approved for ~100 indications already. The distinction from stem-cell-therapy is that no cells are transplanted; only acellular plasma fractions are exchanged, replaced, or diluted. Mechanism debate (young-factor introduction vs. old-factor removal vs. dilution-only) overlaps directly with the heterochronic-parabiosis literature on satellite-cells (Conboy 2005) and hematopoietic-stem-cells (multiple groups).

Cross-modality decision tree

When the clinical target is a single hallmark, several modalities often compete. Some practical heuristics:

The full hallmark-by-modality matrix is the union of interventions-by-hallmark (per-hallmark blocks) and the modality blocks above. Both query the same atomic pages; this MOC’s contribution is the modality-axis grouping and the cross-modality-decision context.

Watchdog blocks

Watchdog 1 — Modality count (flag empty buckets)

TABLE WITHOUT ID rows.file.link AS Interventions, length(rows) AS Count
FROM "interventions"
GROUP BY mode
SORT length(rows) DESC

Every canonical mode: value should return ≥1 row. As of R26 close (2026-05-06, post mode: audit): lifestyle (3 — exercise, sleep, heat-exposure), dietary (5 — caloric-restriction, intermittent-fasting, time-restricted-eating, methionine-restriction, ketogenic-diet), pharmacological (7), gene-therapy (5), stem-cell-therapy (4), blood-product (1). If a bucket drops to zero, it should be flagged as a coverage gap (see planned-coverage).

Watchdog 2 — Missing modality

LIST FROM "interventions"
WHERE !mode

Every type: intervention page must have a mode: field. Should return empty.

Watchdog 3 — Stale clinical-stage

TABLE WITHOUT ID file.link AS Intervention, mode AS Modality, clinical-stage AS Stage, verified-date AS "Verified"
FROM "interventions"
WHERE contains(["phase-1","phase-2","phase-3","phase-4"], string(clinical-stage))
  AND (date(verified-date) < date(today) - dur(180 days) OR !verified-date)
SORT verified-date ASC

Clinical-stage entries decay every 6 months per sops/integrating-clinical-trials.md. Re-run ClinicalTrials.gov v2 query for any row returning here.

Cross-references

Limitations and gaps

  • The dietary vs. lifestyle distinction was reconciled in R26 (2026-05-06): caloric-restriction flipped from mode: lifestyle to mode: dietary. R23c outputs (intermittent-fasting, TRE, methionine-restriction, ketogenic-diet) were already correct. Pages physically located in interventions/lifestyle/ may carry either lifestyle or dietary mode tags — the folder name is not a constraint on the mode value (lifestyle/ is the catchall directory for non-pharmacological non-cellular interventions per CLAUDE.md). Lint pass should verify mode/folder agreement is not enforced.
  • Some modalities are under-represented: blood-product has a single anchor page; gene-therapy covers viral-vector AAV well but lacks dedicated lentiviral / mRNA-LNP-monotherapy pages. R24+ should consider pages for aav-foxn1 (thymic regeneration), crispr-lpa (Lp(a) editing), and mrna-lnp-vaccines-aging (analogous to COVID-vaccine immunosenescence reversal trials).
  • The decision tree above is based on current 2026-05-06 evidence; should be re-derived if any major cross-modality RCT (e.g., TAME results, PEARL-2, Verve heart-3) reports.
  • Interventions that aren’t yet seeded on the wiki but are clinically prominent: gdf15-blockade (anti-cachexia anti-aging), bimagrumab (sarcopenia, FDA fast-tracked 2024), rejuventate-thymic-aging (Fahy TRIIM-X). These should appear in the appropriate modality block when seeded.

Graph View

Backlinks