Checking a set of hypotheses

Mechanism. Cholesterol transport follows ABCA1/ABCG1 → HDL → hepatic SCARB1 → ABCG5/ABCG8. Limited hepatic uptake or canalicular export can prevent increased cholesterol efflux from becoming increased final elimination.

Prediction. Connect cholesterol-labeled human macrophages, plasma with measured CETP activity and a polarized hepatic model with a separate canalicular outlet, comparing obicetrapib with control while standardizing ApoB and HDL particle number. With low hepatic reserve, standard efflux should increase while final flux increases by less than 10%; restoring the limiting component should increase flux by at least 20%. Separate reductions of SCARB1 and ABCG5/ABCG8 should locate the limitation. If final flux always follows efflux regardless of reserve, the hypothesis is rejected. These thresholds are proposed, not measured.

Mechanism. Enhanced protein degradation within neurons reduces the cargo transferred to microglia, relieving their clearance burden without suppressing them. This does not assume that every aggregate can be removed through LAMP2A.

Prediction. Restore LAMP2A only in neurons in aged neuron–microglia cultures, trace labeled neuronal proteins in both populations, and determine which are substrates of this pathway. Assess synaptic transmission after repeated loading as the primary outcome, and return an equal amount of isolated synaptic cargo to test whether unloading is causal. Increased neuronal degradation without changes in cargo transfer or functional reserve would reject the mechanism.

Mechanism. Colchicine binds tubulin and reduces pathological microtubule-dependent neutrophil activation.

Prediction. Pass neutrophils from older donors through an endothelial chip at clinical exposure, measuring adhesion, capillary patency and preservation of directed migration toward a nonpathogenic stimulus, and test the prediction in a separate series. The selected group is predicted to show at least 30% fewer obstructed microchannels while preserving at least 80% of protective migratory function; these are proposed criteria, not observed results.

Mechanism. An increased flow of damaged proteins into nonfunctional lysosomes makes enhancement of the SON-dependent clearance program harmful.

Prediction. In old neurons, compare correction of a measured lysosomal defect before SON intervention with the reverse order, using equal exposure. Correction first should yield better functional survival after repeated proteotoxic stress; measure flux rather than autophagosome number. The hypothesis is refuted if order does not affect function or SON enhancement is equally beneficial under a confirmed degradation block.

Mechanism. Stable remodeling of the thymic stromal niche by local FGF21. Partial dependence on epithelial KLB is to be tested separately, without assuming complete dependence.

Prediction. Compare brief local FGF21 elevation, continuous exposure and control in old mice, with total exposure matched in a separate group. The primary outcome is new functional naive T-cell output after withdrawal; also assess repertoire, tolerance and stromal architecture. Rapid loss of improvement after FGF21 disappears, or persistence of increased thymic volume alone, would reject sustained remodeling.

Mechanism. Suppression of Clu within old HSC locally corrects their mitochondrial state while retaining the extracellular pool and chaperone function of CLU.

Prediction. Compare Clu suppression in isolated old HSC with extracellular protein neutralization and a control, assessing long-term polyclonal production of lymphoid and myeloid cells after washout. Adding purified extracellular CLU will not abolish the benefit of cellular suppression, whereas restoring expression in HSC with a suppression-resistant construct will abolish it. If the outcome is determined only by protein concentration in the medium, the proposed compartment distinction is incorrect.

Mechanism. Blocking HCRTR1/HCRTR2 with suvorexant restores continuous episodes of deep sleep; daytime benefit depends on continuity rather than total sleep duration.

Prediction. In a randomized comparison of drug and non-drug controls with equal increases in total sleep duration, an increase in continuous sleep episodes will predict at least 15% fewer attention errors without worsening balance. Polysomnography, morning attention and balance, and residual drug concentration will be measured, with mediation analysis and the alternative explanation of daytime sedation specified in advance.

Mechanism. Partial inhibition of MYH7 with mavacamten reduces expenditure on unproductive cross-bridges. ATP expenditure per unit of external work defines the useful window of myosin inhibition.

Prediction. In a load test of human cardiac microtissues, measure ATP flux, external work and recovery after increased stimulation frequency. Compare profiles with the same baseline force but different energetic efficiency and independently alter energy supply. The inefficient group is predicted to show at least a 20% increase in work before fatigue without a decrease in maximum force of more than 10%; these are proposed criteria, not observed results.

Mechanism. Targeted restoration of critically short telomeres in normal HSC changes competition between normal and mutant clones.

Prediction. In a mixed model of normal and mutant HSC, compare targeted correction of the normal pool, general correction and a control, using normal polyclonal haematopoiesis after stress as the primary outcome. Measure the distribution of the shortest telomeres, damage and the absolute growth of each clone, and check that cell-cycle control is preserved. If targeted correction does not change competition or predominantly expands dangerous clones, the proposed window is not supported.

Mechanism. The beneficial phagocytic program can be separated from other functions of the upstream kinase, with the downstream node sufficient during natural aging.

Prediction. In naturally old animals, compare microglia-restricted restoration of Bhlhe41, Aurka and a control after the same demyelinating injury. Bhlhe41 should improve conduction along the affected tract and debris clearance with less change in the cell cycle. Assess myelin thickness, debris removal and the microglial cell cycle as secondary outcomes; suppress Bhlhe41 in the Aurka group to test epistasis and verify completed digestion rather than particle uptake alone. If the effect occurs only in an artificial Aurka-loss model and is absent during natural aging, its applicability to age-related recovery is rejected.

Mechanism. Mitochondrial uncoupling reduces short, extreme oxidation peaks.

Prediction. In adult Drosophila, compare moderate tissue-specific uncoupling regimens with the same average redox signal but different peak frequencies. Use a dynamic redox reporter and independently restore pulsed oxidation without changing the average. If restoring the peaks does not abolish lifespan extension, or if the effect is explained solely by temperature and energy consumption, the mechanism is rejected.

Mechanism. Vesicular transfer of miR-29 relieves epigenetic repression of KL, allowing part of the exercise signal to act remotely on cartilage.

Prediction. Compare vesicles before and after exercise and engineered miR-29 vesicles in chondrocyte cultures with intact or suppressed KL, measuring α-Klotho, matrix anabolism and viability; separately test cartilage from old mice. miR-29 enrichment should reproduce part of the effect, and KL suppression should weaken it.

Mechanism. Exercise accelerates resolution of the inflammatory response.

Prediction. Compare exercise and control groups in old mice, serially measuring the response to standardized sterile injury. Selectively prolong the inflammation-resolution phase in the exercise group while maintaining the amount of exercise, then test restoration of the effect. If prolonged resolution does not reduce the survival benefit of exercise, the proposed mediator is rejected.

Mechanism. A persistent stiff scaffold maintains mechanical coupling of paracardiac adipose tissue to the pericardium and myocardium after fat volume decreases.

Prediction. Relate changes in fat to regional deformation and cardiac filling in permitted paired patient data during tirzepatide treatment, accounting for pressure, fluid balance and weight loss. In a mechanically calibrated cardiac-chamber model with human paracardiac matrix, reduce lipid volume while preserving the scaffold or disconnect the scaffold at the same volume, measuring the pressure required for the same diastolic volume and excluding direct dependence on sample thickness. Removing fat from a stiff scaffold should leave at least 20% of excess pressure, and mechanical disconnection should eliminate that portion. An effect fully explained by mass, pressure or fluid and not reproduced in the causal model would reject the hypothesis.

Mechanism. When fatty acid oxidation is limited, excessive suppression of esterification redirects lipids from triglyceride storage into toxic intermediate pools.

Prediction. In primary liver microtissues, compare graded DGAT2 suppression with a nontargeting control while independently varying oxidative reserve and measuring isotope fluxes, diacylglycerols, ceramides, acyl-CoA, albumin, cell death, and lipid export. At clinically achievable exposure, partial suppression will preserve at least 25% better function than maximal suppression in the low-reserve group. Monotonic functional improvement without toxic redirection, or failure to reproduce the proposed window, would refute the hypothesis.

Mechanism. The peak of proteasome activity aligns with the period of minimal input of new proteins.

Prediction. In Drosophila, shift the feeding phase and the tissue-specific phase of proteasome activity while keeping calories and fasting duration equal; measure protein turnover, aggregate accumulation and lifespan. If restoring phase alignment does not restore longevity while sleep and food consumption remain unchanged, the mechanism is not supported.

Mechanism. Reducing expression of the PLB1 ortholog alters the distribution of lipid chain lengths, reducing the age-related shift toward long acyl chains.

Prediction. Compare suppression of the PLB1 ortholog with a control using lipidomics and survival curves, separately examining feeding, development and fecundity as possible explanations. The expected combination is reduced chain lengths and increased lifespan; confirmed intervention with reproducible chain shortening but no lifespan increase would refute this combination.

Mechanism. Continuous PAGln signaling prevents recovery between exposures.

Prediction. Expose aged human vascular microtissues to constant or pulsed PAGln with equal area under the curve. Constant exposure is expected to cause poorer barrier recovery after complete washout; account for free concentration and metabolite stability. The hypothesis is refuted if the difference disappears after matching free exposure or both regimens are equally reversible.

Mechanism. Partial suppression of the BMP hydrolase PLA2G15 allows BMP accumulation, improving lipid cargo processing primarily in aged macrophages with low lysosomal BMP.

Prediction. Stratify aged macrophages by BMP and lipid flux, then compare partial PLA2G15 suppression, a control, and restoration of enzymatically active protein. Partial suppression should improve sequential digestion of multiple portions of apoptotic cells predominantly in macrophages with low BMP. Measure cholesterol and viability separately and monitor excess PLA2G15 substrates using lipidomics. An effect independent of BMP, or BMP accumulation without completed digestion, would not support the proposed mechanism for identifying responders.

Mechanism. Restoration of lysosomal reserve between pulses of mTOR suppression mediates the lifespan benefit of the intermittent regimen.

Prediction. In adult C. elegans, compare continuous and pulsed TOR suppression at comparable integrated reductions in translation, measuring autophagic flux and lysosomal acidity. Selectively disrupt recovery of acidity only during the intervals. Disrupting lysosomal recovery should eliminate the advantage of the pulsed regimen; otherwise, the proposed mechanism is not supported. Assess lifespan extension through overall survival, with independent replication and controls for feeding, reproduction, activity and toxicity.

Mechanism. INSIG1 stabilization is proposed to bypass part of the age-related RORγ deficit through the RORγ–INSIG1/SREBP2 axis when excessive SREBP2 activity is the key defect.

Prediction. Compare RORγ restoration, INSIG1 stabilization, their combination and control in aged tubular organoids, keeping intervention levels within the physiological range. Assess transport function after lipid loading as the primary outcome, measure cholesterol synthesis, STING and AMPK, and use forced SREBP2 activity to test the bypass. INSIG1 should restore transport function; normalization of cholesterol without transport recovery would refute the proposed sufficiency of the downstream component.

Mechanism. Remaining extracellular HERVK-containing particles may sustain secondary senescence after removal of the primary senescent-cell source.

Prediction. In a human tissue coculture, separate donor and recipient cells and compare selective particle removal with control after donor secretome input has stopped. Particle removal should improve recipient-tissue regeneration after repeated challenge; adding back the purified fraction while controlling for accompanying cytokines would test causality. An effect fully explained by soluble cytokines, or no change in recipient function after particle removal, would refute the mechanism of prolonged propagation.

Mechanism. Erythrocyte deformability constrains microcirculation independently of hematocrit: poorly deformable erythrocytes increase resistance in narrow capillaries.

Prediction. Use paired blood samples collected before and after therapy prescribed independently of the study. Separately normalize hematocrit, plasma viscosity and oxygen capacity in a microvascular network, cross-exchange plasma and washed erythrocytes, and measure oxygen delivery at a fixed pressure difference and flow stoppages at constrictions. Ektacytometry should define the predictor before outcomes are revealed, and restoring the original erythrocyte fraction at the same hematocrit should serve as a causal control. Oxygen delivery should differ by at least 15% at equal hematocrit, with the difference reproducible through erythrocyte transfer. Differences that disappear after hematocrit normalization or transfer only with plasma would reject the deformability hypothesis.

Mechanism. IL-11 can retain epithelial cells in a KRT8-high transitional state. The proposed epithelial block remains an independent limitation after intervention in fibroblasts, and epithelial IL11RA1 suppression is proposed to release this block.

Prediction. After documented, comparable injury in old animals, compare control, fibroblast intervention, epithelial IL11RA1 suppression and their combination using validated cell-specific interventions. The combination should improve function beyond fibroblast intervention alone, with transitional cells maturing into AT1 rather than merely dying; measure gas exchange, lung compliance and fibrosis. Changes in cell markers without additional functional improvement, or loss of KRT8-positive cells through death without alveolar restoration, would not support the hypothesis.

Mechanism. Early suppression of the inflammatory program may interfere with clearance. A later signal should accelerate the transition to repair without directly inhibiting protein synthesis in muscle progenitors.

Prediction. Compare early and delayed courses at equal total exposure, placebo and macrophage-directed delivery, initiating the delayed course according to a predefined reduction in debris. The primary outcome is recovery of strength after repeated standardized injury; assess clearance, the satellite cell reserve and dependence on Gcn2 in macrophages. The delayed course should outperform the early course in repeated regeneration despite comparable reductions in inflammatory markers. The phase-dependent macrophage explanation is rejected if differences disappear after exposure is matched or persist when Gcn2 is disabled in macrophages.

Mechanism. The effects of miR-29 delivery reflect a balance between KL-dependent benefit and possible costs from changes to other matrix targets.

Prediction. Compare doses of miR-29 vesicles, empty vesicles, and direct KL restoration in explants from old donors, using mechanical properties after loading as the primary outcome. Measure new collagen, aggrecan, and matrix degradation; suppress KL to test the beneficial branch and use equal delivery of other cargoes to test specificity. If the functional response is entirely monotonic and no matrix cost occurs within the studied range, the proposed boundary is not supported.

Mechanism. Lifespan extension depends on eliminating the critically low NAD state in a small fraction of cells.

Prediction. In old mice, compare two NAD restoration regimens with similar mean levels but different distributions across cells. Use cellular sensors, functional tests, and survival; separately restore NAD in the most deficient population. If eliminating the lower tail of the distribution provides no additional benefit, the hypothesis is rejected.

Mechanism. Brief SIRT3 restoration in old HSC selectively resets pathological inflammatory memory while preserving useful capacity for renewed training.

Prediction. After treatment and washout, compare myeloid descendants of treated and control HSC for basal activity and responses to a new standardized stimulation. The primary outcome is effective recovery after sterile tissue injury without chronic inflammation, controlling clonal composition, cell numbers, and chromatin memory. The selective-reset hypothesis is rejected if both the pathological baseline and all subsequent functional responses decrease.

Mechanism. Local TTR synthesis in affected ligaments and tendons supports continued deposition after vutrisiran suppresses hepatic TTR.

Prediction. In ligament/tendon explants with ATTR, distinguish newly synthesized local TTR from external supply using isotope labeling. Compare reduced external TTR with additional local siRNA, monitoring oligonucleotide availability, mRNA, and secretion, and measure new deposits and tissue gliding. In TTR-synthesizing explants, local suppression will reduce new deposition by at least 30% beyond the reduction achieved by lowering external supply. This threshold is a proposed criterion, not an observed result.

Mechanism. RhoA is linked to nuclear mechanics and the age-related state of HSC; the proposed restoration produces a sustained change in intrinsic HSC function.

Prediction. Treat old HSC ex vivo, wash out the intervention, and compare competitive hematopoiesis after standard and intrabone administration, normalizing to the number of early clones that actually engrafted using cellular barcodes. Improved polyclonal lymphoid-myeloid reconstitution should persist during long-term follow-up and secondary transplantation. An advantage confined to early colonization, unchanged productivity and balance among equally engrafted clones, or expansion of one dominant clone would not support the hypothesis.

Mechanism. During PTH1R activation with abaloparatide, advanced glycation end-product crosslinks in old collagen limit the mechanical benefit of new matrix depending on where it forms.

Prediction. In human bone constructs with a known glycation map, compare abaloparatide and control with spatial labeling of new collagen and assessment of total mineral gain, cracks at the new–old matrix interface, and work to failure. The proportion of the most glycated regions replaced will predict at least a 25% greater gain in work to failure independently of total mineral mass.

Mechanism. Lysosomal export of oxidized lipid breakdown products prevents their accumulation from making enhanced autophagic flux a source of damage.

Prediction. Enhance autophagy in adult worms with normal or moderately reduced lysosomal lipid export; measure flux, lipid products, lysosomal integrity and survival, and test rescue by restoring the transporter. Preservation of the full effect of autophagy despite confirmed export delay refutes the necessity of this step.

Mechanism. Following sGC stimulation by vericiguat, signaling through PRKG1 must reach titin TTN to reduce stiffness.

Prediction. In aged myocardial slices with comparable total cGMP, measure local cGMP/PKG sensors, titin phosphorylation, and passive tension. Use PRKG1 attenuation and its targeted restoration as mechanistic controls. The local signal will explain at least 20% more variation in the mechanical response than total cGMP; restoring the connection will reduce passive tension by 15%. These thresholds are proposed criteria, not observed results.

Mechanism. Preserved innervation limits whether the respiratory improvement associated with cardiolipin restoration translates into improved muscle force.

Prediction. In old mice, compare local restoration of Crls1 with a control, stratifying muscles in advance by endplate occupancy. Measure the increase in force under nerve stimulation as the primary endpoint, use direct muscle stimulation to distinguish nerve from muscle defects, and verify cardiolipin restoration by lipidomics. Equal force recovery in both strata and no interaction between treatment and baseline innervation with adequate statistical power would refute the hypothesis.

Mechanism. The benefit of systemic translation suppression depends on preserving local synthesis of repair proteins in axons.

Prediction. In adult Drosophila, combine moderate systemic translation suppression with selective preservation or suppression of axonal synthesis, avoiding developmental interventions. Measure organelle transport, motor function and survival. If restoring local synthesis does not change the effect of systemic suppression, the hypothesis is not supported.

Mechanism. Mitochondrial citrate export through SLC25A1 supports acetyl-CoA-dependent acetylation of SASP loci. Restricting suppression to senescent fibroblasts reduces SASP while avoiding the limitation of regenerating epithelial cell metabolism caused by systemic suppression.

Prediction. In organotypic aged skin, compare fibroblast-restricted and general SLC25A1 suppression at equal SASP reduction. Fibroblast-restricted suppression is expected to produce faster restoration of impermeability after standardized injury; monitor epithelial viability and energy balance.

Mechanism. Inhibiting C5 cleavage with avacincaptad pegol preserves C3-dependent tagging of cellular debris while blocking terminal complement injury, compared with C3 blockade using pegcetacoplan.

Prediction. In a human model of pigment epithelium, photoreceptors and microglia, match suppression of C5b-9 formation between the two approaches and measure outer segment removal, debris accumulation and light response. With equal protection from lysis, C5 blockade will preserve at least 25% better clearance and no lower light response; C3 depletion should eliminate its advantage. The threshold is a proposed criterion, not an observed result.

Mechanism. Mavacamten inhibits the ATPase cycle of cardiac myosin MYH7. The contribution of active cross-bridges distinguishes drug-reversible stiffness from scar-related stiffness.

Prediction. Separate active tension from matrix stiffness in myocardial slices and compare responses to mavacamten across different baseline fractions, controlling calcium transients, systolic work and free exposure. Independently modify cross-bridges to test causality. A high active-tension fraction is predicted to yield at least a 20% reduction in diastolic tension while preserving at least 90% of systolic work; these are proposed criteria, not observed results.

Mechanism. Continuing exposure to PAA accelerates recurrent endothelial dysfunction through endothelial NOX4 after the restorative intervention has been withdrawn.

Prediction. In old vascular explants and then old mice, compare a standardized restorative intervention at low and high measured PAA exposures within the observed physiological range, with and without endothelial Nox4 suppression. PAA will accelerate loss of endothelium-dependent relaxation after withdrawal, and Nox4 suppression will specifically weaken the dependence on PAA; endothelium-independent relaxation serves as a control for smooth muscle changes. Changes in senescence markers without altered vascular function, or an effect independent of endothelial Nox4, would not support this explanation.

Mechanism. Tadalafil inhibits PDE5A and thereby inhibits cGMP breakdown; its functional benefit depends on the retained ability of NO–sGC to synthesize cGMP.

Prediction. With preserved synthesis, the neurovascular response will increase by at least 20%; with low synthesis, the increase will be less than half that, independently of baseline tone. Human arterioles and a neurovascular chip will be tested with tadalafil and a control at equal perfusion pressure, measuring cGMP and responses to neuronal stimulation; experimental reduction of cGMP synthesis will test causality.

Mechanism. RIPK3–MLKL activity may damage mitochondria and shift surviving HSC into a functionally impaired state without necroptotic death.

Prediction. In mouse HSC, perturb MLKL and simultaneously assess its mitochondrial localization, mitochondrial condition, self-renewal, lymphoid output, and cell death. The supporting outcome is functional restoration that cannot be reduced to an increase in the number of living cells. No mitochondrial or functional changes after accounting for viability, or an effect entirely explained by preventing necroptosis, would refute the hypothesis.

Mechanism. SON-dependent reorganization of nuclear speckles determines the functional response to pyrvinium.

Prediction. Compare pyrvinium, moderate SON restoration, and their combination in directly converted neurons from elderly donors. The primary endpoint is preservation of a reproducible electrical response under repeated stimulation, with proteostasis and speckle dynamics as additional measures. The hypothesis is refuted if the response depends only on the initial aggregate burden and SON does not alter it despite confirmed intervention activity.

Mechanism. Daprodustat inhibits HIF prolyl hydroxylases EGLN1/EGLN2/EGLN3. Iron export through SLC40A1/ferroportin limits the useful response to HIF-induced erythropoietic demand.

Prediction. In linked hepatic, macrophage and erythroid cultures from older donors, compare EPO signaling, isotopic iron transfer and hemoglobinization; block and restore ferroportin, and compare with EPO at the same erythropoietic stimulus. Iron flux per unit of EPO is predicted to forecast the yield of mature hemoglobinized cells at least 25% more accurately than baseline ferritin. The hypothesis is refuted if the measure does not work independently of overall iron availability or restoring flux does not improve cell maturity.

Mechanism. Neutralizing APRIL/TNFSF13 while preserving BAFF/TNFSF13B may affect protective plasma cell clones differently according to their dependence on these signals. A narrow protective repertoire can make antigen-specific secretion particularly dependent on APRIL-dependent clones.

Prediction. In plasma cell niche models using authorized material from older donors, compare clonal composition and antibody secretion against previously administered vaccine antigens under sibeprenlimab, control and simultaneous APRIL/BAFF blockade, and separately vary BAFF while holding APRIL constant. Donors with a narrow APRIL-dependent repertoire should lose at least 30% of secretion against an individual antigen while losing less than 15% of total IgG; BAFF rescue should explain differences between niches. The primary measure is retained antigen-specific secretion per initial cell number, with avidity, viability and Gd-IgA1 also assessed. If specific secretion follows total IgG regardless of clonality, the hypothesis is rejected. These thresholds are proposed, not measured.

Mechanism. Hemopexin binds extracellular heme while targeted delivery of transferrin-bound iron corrects iron availability in T cells, jointly addressing extracellular heme toxicity and intracellular iron limitation.

Prediction. Conduct a factorial experiment in old animals comparing hemopexin, restricted delivery of transferrin-bound iron, their combination, and control, after confirming both excess available heme in the environment and iron deficiency in T cells. The primary outcome is the functional antigen-specific memory response after vaccination. Labeled iron should establish where the substrate goes, with monitoring of accumulation in other cells and oxidative stress. The combination should improve intracellular iron availability without increasing tissue toxicity. No additional functional benefit from heme correction, or increased damage from iron, would reject the proposed combination.

Mechanism. Accelerated removal of defective ribosomes mediates lifespan extension under protein restriction.

Prediction. In adult worms, compare protein restriction under normal and partially impaired ribosome quality control. Measure translation errors, ribosome turnover, and survival; restore ribosome quality by an independent method. If the effect of restriction persists despite confirmed accumulation of defective ribosomes, their removal is not a necessary mediator.

Mechanism. CMA and macroautophagy address different limitations on aged muscle strength, with soluble damaged proteins constituting a limiting cargo for which restoring CMA provides an additional benefit.

Prediction. In aged muscle fibers, conduct a factorial comparison of CMA restoration and independently confirmed enhancement of macroautophagic flux. CMA restoration should provide additional improvement in specific force and its recovery after loading when the limiting cargo consists of soluble damaged proteins. Separate flux reporters, proteomics of soluble and aggregated fractions, and LAMP2A inactivation test mechanistic independence; no additional CMA effect after macroautophagy normalization, assessed with sufficient precision, would reject the proposed independent limitation.

Mechanism. Macrophage efferocytosis rapidly clears the remnants of removed senescent cells without inflammation.

Prediction. In old mice, factorially combine inducible removal of senescent cells with a time-limited alteration of macrophage efferocytosis. Match the number of cells removed and measure cell remnants, inflammation resolution, organ function and survival. If efficacy does not depend on remnant clearance when cell removal is equal, the hypothesis is not supported.

Mechanism. Ferroptosis blockade alone preserves some damaged MuSC; KMT5A restoration is proposed to preserve the genomic quality of the reserve.

Prediction. Compare brief protection against ferroptosis with KMT5A restoration in aged MuSC, matching viable cell numbers before transplantation. KMT5A restoration is expected to produce better functional recovery of recipient muscle after repeated injury; additional outcomes are DNA damage and clonal stability. The hypothesis is refuted if functional outcomes are identical at equal cell numbers or if KMT5A increases genomic abnormalities.

Mechanism. A persistent external source of signals rapidly returns renewed microglia to dysfunction through migrasome transfer from border macrophages.

Prediction. Compare microglial renewal, border macrophage-restricted Tspan4 suppression, their sequential combination and a control, tracing the origin of new cells and the arrival of labeled migrasomes. Prior suppression of the source should delay relapse, and adding back isolated migrasomes should eliminate the advantage. The primary outcome is memory preservation after a withdrawal period, controlling for motor activity; secondary outcomes are microglial function and renewed accumulation of the dysfunctional state. If Tspan4 affects only border macrophage survival, or renewal alone is equally durable, the proposed mechanism of effect duration is not supported.

Mechanism. Efruxifermin activates FGFR/β-Klotho complexes, but the response depends on hepatic sensitivity to adiponectin signaling through ADIPOR2.

Prediction. In linked primary adipose and hepatic microtissues, vary the donor of each compartment and measure adiponectin, lipid fluxes, albumin synthesis and injury; separately neutralize adiponectin and restore ADIPOR2 to test this pathway's contribution. The two-tissue test is predicted to reduce the error in forecasting functional recovery by at least 25% compared with a single measure of adipose response. The hypothesis is refuted if hepatic sensitivity does not modify the effect or the adiponectin branch does not explain the difference.

Mechanism. A limited reset of persistent AP-1-dependent epigenetic memory after inflammation separates its pathological effects on clonal advantage from its role in normal repair.

Prediction. In old post-colitis organoids, compare a brief course after inflammation subsides, a continuous course and a control, washing out the intervention before repeated sterile injury. Measure barrier restoration as the primary outcome, track clonal diversity and persistent accessibility of AP-1 chromatin sites, and control for selection through widespread cell death. If memory is erased at the cost of regeneration or the brief course does not change memory after withdrawal, the proposed separation of functions is rejected.

Mechanism. Ferritin buffering determines the response to iron restriction through reduced fluctuations in labile iron.

Prediction. In adult worms, compare interventions affecting iron intake and ferritin storage while matching total iron content. Measure labile pool dynamics, lipid oxidation, and survival; control for anemia-like energetic defects at the level of respiration. If the effect is determined by total iron independently of labile pool fluctuations, the hypothesis is refuted.

Mechanism. Activation of ADRB3 with mirabegron mobilizes fatty acids; the balance between lipolysis, re-esterification and oxidation determines whether the mobilized fatty acids are utilized or transfer harm to muscle.

Prediction. In a primary adipose–muscle model from older donors, high fatty acid utilization will produce at least 20% greater improvement in muscle glucose uptake without ceramide accumulation; low utilization will not provide this advantage. Isotope tracing, ADRB3 blockade and independent alteration of utilization will test the mechanism at measured achievable exposure.

Mechanism. Brief GRHL2–CDK19 restoration reversibly separates reduction of chronic p21 arrest from preservation of the acute protective p53 response to new damage.

Prediction. Compare brief and continuous GRHL2 restoration in old organoids with control, using secretory function after withdrawal as the primary outcome. An independent mild challenge should then demonstrate preserved p53-dependent arrest in response to new damage; track clonal composition, genome integrity, and preservation of epithelial identity. If functional restoration requires persistent suppression of the protective response or selection of damaged clones, reversible separation is rejected.

Mechanism. IL-1α maintains residual cartilage catabolism after IL-1β blockade. Canakinumab neutralizes IL1B, whereas anakinra blocks IL1R1, affecting IL-1α/β signaling.

Prediction. Compare canakinumab, anakinra and control in paired synovial-cartilage explants, with separate IL-1α neutralization to test the explanation. Compare completeness of IL-1β suppression and measure cartilage mechanics, necrosis and defect repair rate. With high residual IL-1α dependence, anakinra is predicted to preserve mechanical properties at least 20% better than canakinumab, and removal of IL-1α is predicted to eliminate the difference; these are proposed criteria, not observed results.

Mechanism. Insufficient NUP50 limits metabolic remodeling despite retained AMPK activation.

Prediction. In primary old human cells, conduct a factorial experiment combining an AMPK activator and NUP50 normalization while ensuring equal AMPK activation. NUP50 normalization should restore the metabolic response, with ATP recovery during substrate switching as the primary endpoint and fatty acid oxidation and nuclear transport as additional endpoints. Failure to restore function despite confirmed NUP50 expression, or an effect occurring only through additional AMPK activation, would refute the hypothesis.

Mechanism. Recovery intervals restore the FOXO1-dependent responsiveness of osteogenic cells to repeated mechanical stimulation.

Prediction. Compare continuous and interval-based local loading in aged animals with identical peak strains and cycle counts, selecting intervals from FOXO1 dynamics in an independent pilot and fixing them before the main experiment. Interval-based loading is predicted to improve resistance to cyclic mechanical failure, and osteogenic suppression of Foxo1 is predicted to eliminate the advantage. Monitor microdamage and inflammation. A difference explained by less damage under noncomparable loads, or an effect independent of FOXO1, would not support the proposed mechanism.

Mechanism. Elastin fragments released during matrix breakdown may initiate inflammatory feedback that undermines restorative remodeling.

Prediction. Compare identical limited remodeling of old tissue explants with early fragment binding, late binding, and a control, measuring functional recovery after withdrawal at comparable final mechanics. Labeled elastin and reintroduction of isolated fragments will test the signal's source. The feedback hypothesis is rejected if interception timing does not change the late outcome or the effect is determined solely by reduced matrix breakdown.

Mechanism. Prior reduction of incoming fatty acid flux prevents early toxic accumulation in intermediate pools when DGAT2-mediated packaging into triglycerides is subsequently suppressed.

Prediction. In linked adipose and liver tissues, compare both monotherapies, simultaneous initiation, and both treatment orders at equal total exposure, measuring flux with isotopes, toxic fractions, functional protein synthesis, and injury. The FGF21→DGAT2 order will reduce integrated injury by at least 25% relative to the reverse order and the better monotherapy. An advantage explained by unequal exposure or absent after controlling for baseline flux would refute the hypothesis.

Mechanism. The endocochlear potential limits the benefit of restoring SLC35F1 by providing an electrical supply that cellular correction cannot fully replace.

Prediction. Stratify aged animals by endocochlear potential, then compare targeted Slc35f1 restoration with control at comparable hair cell counts. Assess DPOAE and hearing threshold as primary outcomes, measuring potential and cell survival separately; hearing recovery should depend on potential when delivery and cell preservation are equal. Altering the electrical gradient in an explant will test the proposed interaction.

Mechanism. Iptacopan inhibits complement factor B, CFB, suppressing CFB-dependent alternative complement amplification.

Prediction. In a human glomerular model, compare sera with equal C3 deposition and different CFB dependence. Iptacopan, CFB removal and its restoration should produce concordant effects on albumin permeability, podocyte injury and recovery after washout. A high proportion of CFB-dependent amplification will predict a reduction in albumin leakage of at least 30%; a low proportion will produce less than half the effect. These thresholds are proposed criteria, not observed results.

Mechanism. Pulsed activation of PTH1R with abaloparatide, with the interval between pulses determined by recovery of receptor signaling after washout.

Prediction. In primary osteoblasts from older donors on a bone scaffold, compare fixed and adapted intervals at equal total exposure. The adapted interval will increase strength by at least 20% without an additional increase in resorption.

Mechanism. Sprifermin activates FGF18/FGFR3, while the timing of mechanical loading affects attachment of the new cartilage matrix to the old matrix.

Prediction. In paired cartilage explants, compare sprifermin/control × immediate/delayed loading with equal total loading, measuring attachment of new and old matrix, friction, cracks, and stability after full loading resumes. Delayed loading will preserve at least 30% greater attachment strength and reduce delamination without loss of thickness.

Mechanism. Lp(a)-associated ENPP2/autotaxin activity modifies the response to LPA reduction with lepodisiran.

Prediction. In a human renal vascular chip, compare Lp(a) fractions with different ENPP2 activity while matching particles and oxidized phospholipids; remove the fraction and add it back with active or catalytically inactive ENPP2, measuring permeability and capillary patency. Removing the highly active fraction will reduce leakage by at least 25%; active, but not inactive, ENPP2 will restore most of the harm.

Mechanism. Without prior restoration of the intestinal barrier, the strengthened immune system responds to a continuous influx of microbial products.

Prediction. In old mice, compare the same two interventions in opposite orders, simultaneously, and separately. Measure permeability, microbial products in blood, immune response, and survival; standardize the microbiota. If order does not affect inflammation and longevity despite confirmed barrier restoration, the hypothesis is not supported.

Mechanism. Suppression of SAP/PTX2 production in bone marrow adipocytes reduces local amyloid deposition, allowing bone mechanical function to recover while the circulating pool is preserved.

Prediction. Compare inducible Apcs suppression in bone marrow adipocytes, systemic SAP neutralization, and control, verifying targeting in bone marrow, liver, and peripheral adipose tissue. The primary outcome is resistance to mechanical failure of bone; secondary outcomes are local deposits and hematopoietic reserve. Local correction should improve function while preserving the circulating pool, and local SAP replacement at an unchanged systemic concentration should weaken the benefit. Reduced deposits without improved strength, or a requirement to reduce systemic SAP, would reject the sufficiency of the local mechanism. Without demonstrated targeting, the experiment cannot resolve the question.

Mechanism. Peroxisomal localization of catalase spatially separates suppression of the chronic oxidant background from preservation of the brief cytosolic redox signal induced by exertion.

Prediction. Compare constructs with equal enzymatic activity and different localization in aged muscle fibers, confirming peroxisomal delivery before the main experiment and restoring import beforehand if PEX5 is defective. Peroxisomal targeting should increase contractile reserve after repeated exertion while compartment-specific H2O2 sensors show suppression of the chronic background and preservation of the pulse. Benefit determined solely by total antioxidant activity regardless of localization, or suppression of the beneficial response to exertion, would not support the hypothesis.

Mechanism. Temporary inhibition of miMOMP limits BAX/BAK-dependent mtDNA leakage involved in SASP, with apoptotic competence preserved after withdrawal.

Prediction. Compare short and continuous interventions in aged tissue models, monitoring mtDNA and the number of living damaged cells. The proposed separation would be supported by recovery of neighboring microtissue function after withdrawal and retained apoptotic competence in a separate standardized test; independent correction of leakage without globally suppressing cell death would strengthen the causal conclusion. Benefit requiring continuous blockade of cell death, or persistent damaged clones after withdrawal, would refute the proposed therapeutic separation.

Mechanism. Persistent remodeling of peripheral immune cells into a less inflammatory state determines the duration of the cognitive effect of PF4.

Prediction. Measure immune states and memory in old mice after treatment and washout. Transfer washed peripheral immune cells from treated and control donors into comparable old recipients, excluding transfer of free PF4. Memory improvement should persist after PF4 disappears and transfer with the cells, with memory after PF4 disappearance as the primary outcome and controls for motor activity and platelets. Failure of the effect to persist or transfer despite confirmed immune remodeling would fail to support the cellular-memory hypothesis.

Mechanism. Epidermal PANX1-dependent ATP release from keratinocytes re-excites the nerve network after capsaicin treatment targeting TRPV1.

Prediction. In a skin–nerve model, selectively reduce PANX1 in keratinocytes after identical capsaicin exposure, measuring extracellular ATP, nerve density and spontaneous activity; restore ATP as a mechanistic control while maintaining viability. Reducing the epidermal signal will reduce the return of spontaneous activity by 30% without delaying protective reinnervation by more than 10%. Restoring ATP should restore the effect attributed to ATP. These thresholds are proposed criteria, not observed results.

Mechanism. Neuronal GPR158 availability limits the memory benefit of restoring osteocalcin.

Prediction. Use a factorial comparison of osteocalcin and moderate Gpr158 restoration in the relevant hippocampal population. Their combination should improve memory more than the hormone alone at the same hormone exposure, with independent controls for anxiety and movement and measurements of BDNF and synaptic transmission. Gpr158 suppression tests the necessity of the receptor component. If hormone response does not depend on receptor availability, or the combination does not outperform the hormone despite confirmed correction, the proposed limitation is not supported. Check for reduced receptor availability before the experiment rather than assuming it is universal with age.

Mechanism. The regenerative demand associated with restoring structure drives distant functional rejuvenation.

Prediction. Compare injuries with different regenerative demands but comparable early stress responses, controlling nutrition, size and reproduction. Measure recovery of distant function after complete healing, and temporarily restrict the regenerative programme with toxicity controls to test its necessity. If the effect is determined solely by the stress pulse and persists without regeneration, the hypothesis is rejected.

Mechanism. Changing phosphatidylcholine synthesis may trigger different compensatory responses depending on the background mitochondrial state.

Prediction. Under identical conditions and with intervention beginning at the same age, compare controls with suppression of each gene in all three genotypes, measuring survival, phosphatidylcholine, and mitochondrial network integrity. Lifespan should increase in wild-type C. elegans and decrease in clk-1(qm30) and isp-1(qm150). The same direction of survival change in all three genotypes despite confirmed gene suppression and reproducible conditions would refute the hypothesis.

Mechanism. Targeted QPCT/L correction releases macrophages from contact-dependent, CD47–QPCT/L-dependent inhibition of efferocytosis by neighbouring senescent cells.

Prediction. In an old-tissue coculture containing senescent cells, macrophages and separately labelled apoptotic cargo, compare targeted QPCT/L correction, treatment of the macrophage alone and a control, with contact disrupted in a separate group. The primary outcome is completed clearance of repeated portions of cargo and restoration of the tissue barrier. If the effect depends solely on reduced soluble cytokines or direct macrophage stimulation, the contact-release hypothesis is rejected.

Mechanism. Aprocitentan inhibits EDNRA/EDNRB, removing the renal EDNRB-dependent reserve for suppressing sodium transport.

Prediction. Connect a human vascular chip to a collecting duct model with ENaC transport, match vascular dilation, and separately disable EDNRB in the tubular compartment. Measure sodium and water balance. Baseline ETB reserve will predict an increase in sodium retention of at least 25%; disabling it beforehand will eliminate the drug's additional tubular effect. These thresholds are proposed criteria, not observed results.

Mechanism. Drug binding and slow distribution in fibrotic interstitium may limit local free acoramidis exposure, leaving TTR insufficiently stabilized in some tissue regions.

Prediction. Perfuse human cardiac or tendon matrices of different densities with added standardized TTR at the same plasma-like acoramidis concentration, matching cell viability and time to equilibrium. Measure free drug, TTR subunit exchange and new aggregate formation at different depths. A difference of at least 20% in the local aggregate formation rate must be explained by measured free exposure and disappear when free concentration is equalized through the tissue; saturation of binding without changing TTR provides an additional control. Sufficient concentrations everywhere, or residual aggregation independent of exposure, rejects this cause of nonresponse. The 20% threshold is proposed, not measured.

Mechanism. Restoration of particular isoforms through brief SRSF1 correction produces a lasting functional change.

Prediction. Compare short SRSF1 induction, continuous induction, and control in old donor cells, using tissue-specific function after withdrawal in a validated microtissue as the primary outcome. Measure isoforms, identity, clonal composition, and aging clocks; independently correct the leading isoform to test causal sufficiency. Functional improvement should persist after withdrawal; changes confined to clocks or loss of the functional result together with SRSF1 expression would not support the proposed durable change.

Mechanism. Transient CMA activation durably prepares senescent target cells for macrophage-mediated removal by changing their recognizability.

Prediction. In a factorial coculture, pretreat and wash only senescent fibroblasts, only macrophages, both populations, or neither. Pretreatment of the targets should reduce the number of living senescent cells after repeated contact cycles even after withdrawal, with barrier restoration in a tissue model assessed subsequently. Suppress LAMP2A in targets and control for spontaneous target death; benefit requiring continuous compound exposure or direct macrophage stimulation would reject durable target preparation.

Mechanism. IAA suppresses ISC turnover through AhR–Wnt signaling; the proposed benefit depends on applying this brake after repair rather than during the expansion phase.

Prediction. Compare early and delayed IAA at equal exposure in old intestinal organoids and then mice, defining the start of the late phase by barrier restoration. Assess barrier quality after repeated injury and preservation of clonal reserve, test the mechanism through epithelial AhR inactivation, and measure local rather than substitute plasma concentration. If timing does not affect function despite a confirmed AhR response, phase-dependent inhibition is rejected.

Mechanism. Improved proteostasis reduces the release of active protein seeds and thereby reduces damage transmitted to neighboring cells.

Prediction. Treat only donor old neurons in a microfluidic culture and transfer their medium to untreated recipients after removing the drug. Assess recipient electrical function as the primary endpoint and measure seeding activity at equal extracellular protein concentration. The hypothesis predicts reduced seeding activity and recipient damage; improvement in donors without changes in seeding activity or recipient damage after controlling for drug carryover refutes it.

Mechanism. Slow removal of already inserted C5b-9 complexes delays barrier recovery after inhibition of new C5 cleavage with avacincaptad pegol; membrane recycling determines removal of these complexes.

Prediction. In pigment epithelium from elderly donors with equal initial C5b-9 loads, track labeled old complexes separately from new ones after blockade begins. Alter membrane recycling with a reversible research intervention while monitoring viability and phagocytosis. The rate of old complex removal will predict the time to barrier recovery, and accelerating removal will shorten it by at least 25%. The threshold is a proposed criterion, not an observed result.

Mechanism. After TNFSF11/RANKL blockade by denosumab, zoledronate inhibits FDPS. Treatment timing accounts for renewed fusion of recycling osteoclast cells and uptake of mineral-bound zoledronate as resorption resumes.

Prediction. In human bone culture modeling declining denosumab, a regimen timed by a cellular test will reduce peak resorption by at least 30% without additional osteoblast loss. Compare cellular-test timing with fixed timing at equal exposure, tracking cell fusion, uptake of mineral-bound zoledronate, resorption lacunae, strength and osteoblast viability.

Mechanism. The proposed functional boundary involves dynamic neuronal inhibition and a possible trade-off between memory stability and flexibility following PLA2G7 correction.

Prediction. In old mice, compare degrees of NSC-restricted PLA2G7 suppression with a control. The primary outcome is switching a learned rule while preserving the original memory; simultaneously measure inhibitory responses and neurogenesis, controlling for anxiety, movement and the speed of learning the first task. Excessive correction may preserve old memories while impairing rule switching. If memory and flexibility change in the same way throughout the correction range without a relationship to inhibition dynamics, the proposed functional boundary is not supported.

Mechanism. Cellular adrenergic sensitivity determines the functional harm caused by the same PAGln dose.

Prediction. In endothelial cultures from older donors, measure the baseline adrenergic response and then expose the cultures to PAGln; test candidate receptors through genetic suppression and restoration of expression. Receptor sensitivity is expected to predict the effect, and genetic restoration is expected to restore sensitivity to PAGln. Barrier function is the primary endpoint; AMPK and senescent features are secondary.

Mechanism. BCKDK inhibition removes a metabolic limitation so that a transient anabolic signal is again accompanied by substrate utilization rather than substrate accumulation.

Prediction. In old mice, compare control, physiological protein supplementation, BCKDK suppression and their combination, with matched caloric intake and physical activity. Measure isolated muscle specific force after treatment and a withdrawal period as the primary outcome, with recovery after exertion and labeled leucine flux through oxidation as secondary outcomes; an independent muscle-specific intervention targeting BCKDK tests the mechanism. Supplementation will have a greater effect when BCAA flux is restored, especially in animals with a baseline catabolic defect. Increased p-S6 or muscle mass alone is insufficient; restored flux without an interaction with nutrition and improved specific force, with sufficient precision of estimation, would not support the proposed causal relationship.

Mechanism. Combining BST2 with stable cell-cycle arrest distinguishes persistent senescence from a transient interferon response, preserving recoverable cells.

Prediction. In old retinal cultures, compare BST2 selection and dual selection after a short interferon pulse and washout. For postmitotic cells, do not use arrest as a standalone criterion; a validated set of senescence features is required. The primary outcome is the retinal model's light response; separately track the fate of transiently BST2-positive cells. The proposed advantage is rejected if the additional selection does not reduce functional cell loss at equal removal of persistently senescent cells.

Mechanism. Heterogeneous delivery leaves regions with insufficient local HSD17B13 suppression, which remain unprotected despite substantial average suppression. Tissue uptake is ASGR1-dependent.

Prediction. In perfused human liver slices, spatially compare oligonucleotide penetration, ASGR1, HSD17B13 and cell death; compare equal average suppression with different delivery heterogeneity and experimentally equalize delivery within the tissue. Reducing heterogeneity is predicted to reduce focal cell death by at least 25% at the same average target suppression.

Mechanism. Excessively prolonged mitochondria–lysosome contacts limit energy reserve; TBC1D15-dependent disengagement normalizes these contacts.

Prediction. First verify increased contact duration in naturally aged neurons, then compare TBC1D15 restoration, an inactive GAP control, and increased lysosomal mass. Normalizing contacts should restore the ability to sustain dopamine release during repeated stimulation. Live imaging should distinguish contact formation frequency from contact duration, and maintaining RAB7 in its active state should test epistasis. Absence of the defect during natural aging, or failure of contact normalization to restore function, would reject the extension of this mechanism to aging.

Mechanism. The vascular matrix retains apoB and releases it slowly, sustaining plaque toxicity after incoming LDL has been reduced.

Prediction. Expose perfused human plaque fragments with fast or slow release of labeled apoB to the same reduction in incoming LDL, reproducing plasma after evolocumab therapy. Match baseline necrosis, independently restore the original lipid flux, and measure barrier function and necrotic expansion. At the same LDL, barrier improvement in the fast-release group will be at least 20% greater; restoring lipid flux will eliminate most of the benefit. Failure of retention kinetics to predict effects in independent donors, or complete explanation of improvement by baseline necrosis, would refute the hypothesis.

Mechanism. Bempedoic acid inhibits ACLY after activation by SLC27A2 and separately activates PPARA.

Prediction. Independently attenuate SLC27A2 and PPARA in primary hepatocytes from elderly donors; measure bempedoyl-CoA, cholesterol synthesis, β-oxidation, apoB export, and albumin synthesis, restore expression of each target, and externally validate the classifier. The two-component test will reduce functional-response prediction error by at least 20% compared with the ACLY level alone.

Prediction. In old fibroblasts, compare AP2A1 suppression, a control, and mechanical restriction of cell area on micropatterns, with the quality of organized matrix in a tissue model after withdrawal as the primary outcome. Measure receptor trafficking, migration, DNA damage, and clonal composition, and restore AP2A1 to test specificity. The proposed independent role is not supported if changing area without AP2A1 reproduces all effects or if the effects disappear after shape equalization.

Mechanism. Inflammation is associated with loss of KMT5A/H4K20me1 and ferroptosis in aged muscle stem cells; the proposed restoration depends on KMT5A catalytic activity.

Prediction. Begin intervention in aged mice after confirming the defect, restrict expression to MuSC and compare active with catalytically inactive KMT5A. Active KMT5A is expected to improve force recovery after a second injury at equal initial cell numbers; track clonal contribution. The hypothesis is refuted if survival increases without improvement in the second regenerative response or if the effect does not depend on catalytic activity.

Mechanism. Sotatercept reduces the availability of TGF-β-family ligands through the ACVR2A/ActRIIA-Fc ligand trap. The proposed vascular vulnerability depends on the residual functional reserve of BMP10–ACVRL1/ALK1–ENG signaling in the endothelium.

Prediction. Measure sotatercept concentration-response curves in paired models of pulmonary-artery endothelial and smooth-muscle cells, assessing BMP10-induced pSMAD1/5 and barrier function under flow before treatment, followed by pathological proliferation and albumin leakage. Use selective BMP10 neutralization and restoration of free BMP10 while measuring its actual availability; analyze BMP9 separately. Restoring BMP10 should reduce leakage by at least 20% while preserving the main antiproliferative benefit. Inseparable effects or failure of baseline reserve to predict leakage would reject the proposed therapy-selection marker.

Mechanism. Transient AMPK–NUP50 activation establishes metabolic memory that requires NUP50 during the formation phase.

Prediction. Expose old cells to a brief AMPK pulse with or without temporary NUP50 suppression, then repeat substrate stress after complete washout. The primary endpoint is the rate of energy charge recovery during the second challenge, with AMPK activity confirmed to have returned to baseline. Improved recovery should persist after washout and depend on NUP50 during memory formation; disappearance of the improvement after washout or equal improvement with NUP50 suppressed would refute the hypothesis.

Mechanism. Phosphatidylcholine availability determines whether reducing SAM synthesis in old age is beneficial or harmful; lipid replenishment separates membrane harm from other effects of altered methylation.

Prediction. In old adult C. elegans, compare partial sams-1 suppression with and without phosphatidylcholine replenishment, without intervening during development. Lipid replenishment should change the functional effect of sams-1 suppression when membrane correction is confirmed, without itself restoring SAM. Remaining lifespan is the primary endpoint, with mobility, lipidomics, and SAM measured additionally.

Mechanism. Disrupted V-ATPase assembly limits degradation, so increasing component abundance is insufficient without restoring the assembled proton pump.

Prediction. In replicatively old yeast, compare restoration of V-ATPase assembly, increased component abundance and a control. The primary outcome is remaining replicative capacity and recovery of growth after stress; separately measure assembly, vacuolar pH, ATP and cargo degradation. An assembly-defective control separates abundance from competence. If increasing components is equally effective without improved assembly, or if benefit is explained solely by selection of young cells, the hypothesis is not supported.

Mechanism. Restoration of HuR/ELAVL1 acts on surviving but functionally impaired sensory cells. Regeneration of lost hair cells is not proposed.

Prediction. Randomize old mice with measured hearing loss to local HuR restoration or control, stratifying by baseline function and first verifying cellular targeting of delivery. The primary outcome is the change in ABR hearing threshold; wave I amplitude and DPOAE distinguish neural and sensory components. Compare function with hair cell and synapse counts at the endpoint under blinded assessment. Late intervention should help when the cellular substrate is preserved and help substantially less when it is lost. Prevention of further cell death without restoration of already lost function would not support reversibility.

Mechanism. Increased ribosomal translation fidelity provides a benefit independently of developmental effects and general slowing of translation.

Prediction. Use inducible RPS23 replacement in mature Drosophila after confirming ribosome assembly and increased fidelity. Compare early induction, late induction and controls under identical nutrition, with age-related loss of motor function as the primary outcome, followed by survival, and measure protein synthesis rate and misreading separately. Late induction should improve functional resilience without slowing overall translation; no benefit despite increased fidelity, or benefit only with overall translation slowing, would fail to support the proposed independent role of fidelity.

Mechanism. After vutrisiran suppresses hepatic TTR synthesis, the removal of soluble TTR oligomers determines the early contractile response.

Prediction. Reproduce reduced TTR production in a connected human liver–heart model with macrophages, measuring monomer, oligomers, and fibrils separately. Add oligomers back while keeping total protein unchanged, and compare with a nontoxic TTR fraction. Oligomer removal will predict an improvement in contractile reserve of at least 20%; adding them back will eliminate at least half of the gain. These thresholds are proposed criteria, not observed results.

Mechanism. Limited capillary reserve constrains oxygen delivery to responding adipocytes as metabolic demand increases, producing local hypoxia and inflammation.

Prediction. In perfused human adipose microtissues with equal ADRB3 responses and equal tissue drug concentrations, restoring perfusion will turn an adverse response into an improvement in muscle glucose uptake of at least 20%, without changing ADRB3 signaling. Capillary flow will be varied while measuring oxygen, lactate, lipid output and the effect of the secretome on the muscle insulin response.

Mechanism. An age-related NPC1 defect retains cholesterol in the lysosome, limiting its intracellular availability for oligodendrocyte maturation.

Prediction. In old OPC, verify cholesterol retention, then compare NPC1 restoration, lipid addition, their combination and control, using labeled cholesterol and monitoring the available membrane fraction. Correcting cholesterol availability should restore conduction in a myelinating neuronal culture, with mature myelin and OPC chromatin state as secondary outcomes. Absence of a transport limitation, or equally effective lipid addition without correcting distribution, would refute the mechanism.

Mechanism. Preserved NAM maintain the supportive function of the local immune niche during anti-inflammatory treatment.

Prediction. Compare selective targeting of the age-associated inflammatory population with broad macrophage suppression, matching reductions in a predefined inflammatory panel while differing in NAM preservation and confirming cell location and identity. Adipose tissue function should be better with preserved NAM even at equal cytokine levels, with insulin sensitivity as the primary outcome at equal body weight and food intake. Additional selective NAM loss should eliminate the advantage of the sparing regimen. If the difference disappears after controlling for weight, exposure and the composition of other cells, or does not depend on NAM, the proposed explanation is incorrect.

Mechanism. SOST neutralization by romosozumab is combined with a controlled direction of mechanical loading to coordinate loading with matrix formation and improve collagen orientation.

Prediction. In paired human bone constructs, the combination will yield at least 25% more cycles to mechanical failure than the drug with nondirectional loading at comparable mineralization. Compare drug/control × directional/nondirectional loading of equal energy, matching initial damage and measuring collagen orientation, mineral and cycles to failure.

Mechanism. Spatial and temporal coupling of macrophage-derived glutamine delivery with restorative metabolic flux and lipogenesis in old MuSC.

Prediction. Compare local macrophage production, external delivery with the same tissue exposure integral, and a control. Trace labeled carbon into new MuSC lipids and vary delivery timing separately; the primary outcome is muscle strength after repeated regeneration. Suppress restorative metabolic flux in MuSC to test the mechanism. The hypothesis is rejected if, at matched local exposure, source and timing do not change function or the effect does not depend on MuSC lipogenesis.

Mechanism. Preserved perfusion allows the tissue to withstand cell removal and clear the resulting debris.

Prediction. Compare CSF1→senolysis, senolysis→CSF1, simultaneous treatment, monotherapies and a control at equal total exposure, using a senolytic dose with confirmed selectivity and measuring resident macrophage death. Early niche restoration should improve healed wound strength at comparable removal of senescent cells; perfusion and capillary leakage are secondary outcomes. Macrophage depletion through a separate mechanism tests whether the restored niche is necessary. An effect explained only by less senolysis or direct CSF1 action on other cells would not confirm the causal hypothesis.

Mechanism. Senescent cells create an environment that inhibits muscle regeneration, and CD36 contributes to their harmful secretory effects. Selective CD36 suppression is proposed to reduce these effects while preserving clearance functions performed by other cells.

Prediction. Compare control, broad CD36 suppression and suppression restricted to senescent cells in old injured muscle, first validating selectivity and confirming comparable CD36 suppression in the target senescent population. Selective suppression should preserve clearance of necrotic material and provide greater functional recovery, assessed through inflammatory secretion, specific muscle force and the stem cell reserve after repeated injury. Reduced inflammatory secretion without recovery of force would refute the hypothesis; failure to preserve tissue clearance or provide a functional advantage would fail to support the proposed value of selectivity.

Mechanism. PEP acts as an endogenous competitive inhibitor of cGAS; targeting its intracellular deficiency in cells that sustain inflammation accounts for the proposed functional advantage.

Prediction. In an old skin model, compare PEP delivery predominantly to fibroblasts with distributed delivery, confirming intracellular levels after extracellular PEP removal. Targeted delivery should produce greater restoration of barrier function in coculture at equal tissue exposure; assess cGAMP and dependence on cGAS. Absence of a targeting advantage, or its persistence without cGAS, refutes the proposed mechanism.

Mechanism. Restoration of ZDHHC11 and reduction of pathological O-GlcNAc signaling converge on a shared GATA4-dependent source of SASP.

Prediction. In cartilage explants from old donors, compare each intervention, their combination, and a control in a factorial experiment. The primary outcome is preservation of mechanical properties after cyclic loading; additionally measure GATA4, p65, and matrix degradation. Restore GATA4 activity to test epistasis and use a dose curve to exclude apparent nonadditivity caused by saturation. The hypothesis is rejected if the combined benefit remains independent of GATA4 or nonadditivity is explained by toxicity.

Mechanism. Oxidation of cardiolipin limits the functional benefit of its new synthesis by making the lipid pool created through increased CRLS1 vulnerable.

Prediction. In aged myotubes, conduct a factorial experiment comparing CRLS1, suppression of lipid peroxidation, their combination and a control. At equal total cardiolipin mass, measure oxidized cardiolipin forms and contractile fatigability, then confirm the interaction in aged muscle. The combination should exceed the expected independent effect; failure to do so, or a lack of association between force and the oxidized fraction at equal cardiolipin content, would refute the hypothesis.

Mechanism. Maintenance of nuclear pore selectivity mediates the lifespan-extending effect of dietary restriction.

Prediction. In adult Drosophila, compare a normal diet and dietary restriction with a moderate neuronal alteration of nucleoporin turnover; measure nuclear reporter leakage, transcriptional noise and survival, and restore expression of the target nucleoporin. If restoring the barrier does not restore the effect of dietary restriction in the absence of toxicity, its mediating role is rejected.

Mechanism. Defective PEX5-dependent peroxisomal import limits metabolic flexibility and the functional benefit of increased mitochondrial biogenesis.

Prediction. In aged muscle models, first measure peroxisomal reporter import, then compare PEX5 restoration, inducible biogenesis through PGC-1α, their combination, and a control. Restoring import before enhancing biogenesis should improve contractile work during switching between carbohydrate and lipid substrates more than enhancing biogenesis alone. Test an import-incompetent PEX5 variant and control organelle number. If the effect of PGC-1α is independent of baseline import and its restoration, the proposed causal order is incorrect.

Mechanism. Suppress APOC3 with olezarsen and add the FGF21 analogue efruxifermin, acting through FGFR/β-Klotho complexes.

Prediction. Compare four groups in linked adipose and liver microtissues from elderly donors, using isotopes to distinguish lipid uptake, oxidation, and export and measuring diacylglycerols, ceramides, albumin, and cell death. The combination will reduce lipotoxic cell death by at least 25% compared with olezarsen while retaining at least 90% of the triglyceride reduction. β-Klotho blockade will test the mechanism of the additional effect; β-Klotho independence will refute the proposed mechanism.

Mechanism. Locally retained CD4-Eomes cells remove senescent cells without requiring prolonged systemic 4-1BB signaling.

Prediction. Compare local cell delivery, systemic stimulation, and control at comparable target removal, using function of a fibrotic tissue model after withdrawal as the primary outcome. Also assess healthy cells, antigen specificity, and preservation of ordinary immune memory. If the benefit requires continuous systemic activation or causes comparable damage to healthy tissue, local sufficiency is rejected.

Mechanism. Preserved lymphatic clearance of accumulated macromolecules is required for the lifespan benefit of suppressing inflammatory-factor secretion by senescent cells.

Prediction. In old mice, factorially combine a senomorphic intervention with moderate enhancement of lymphatic transport. Track clearance of labeled interstitial proteins, edema, inflammation and survival, separately controlling factor production and removal. If the effect does not depend on clearance despite clearance actually being altered, the proposed constraint is refuted.

Mechanism. Residual transport of oxidized phospholipids on Lp(a) continues to impair endothelial recovery despite low LDL.

Prediction. On a human vascular chip, compare four lipoprotein profiles from a hepatic model with the two interventions, matching LDL-apoB and measuring Lp(a) separately. Add back the isolated Lp(a) fraction, using the same protein without excess oxidized cargo as a control. The combination will reduce endothelial leakage by 25% relative to PCSK9 blockade; adding back pathological Lp(a) will restore at least half of the leakage. Loss of the advantage after LDL matching, or failure to reproduce it through Lp(a) removal and add-back, would refute the hypothesis.

Mechanism. Incorporation of phosphatidylcholine into mitochondrial membranes, rather than its total cellular level, determines restoration of metabolic flexibility.

Prediction. In cells from several elderly donors, track labeled lipid at equal cellular uptake, compare delivery methods, and verify mitochondrial fraction purity. Recovery of ATP after switching the energy substrate is the primary endpoint; the fraction of label in mitochondria should explain differences in response rather than the total cellular pool.

Mechanism. Dysfunction in some old HSC is reversible through blockade of P-selectin interactions, allowing the surviving cells to recover function.

Prediction. Compare non-depleting SELP blockade, removal of SELP-positive HSC and a control, confirming no cell death during blockade and matching the number of viable transplanted HSC. Assess functional clonal diversity and balanced blood recovery after repeated hematopoietic stress, measuring early homing separately and using washout and an independent genetic intervention. Some previously SELP-positive labeled clones should produce fully functional progeny after blockade. If the benefit depends only on disappearance of these clones or improved engraftment, the reversible recovery hypothesis is rejected.

Mechanism. PEP is an endogenous competitive inhibitor of cGAS. The benefit of accumulating PEP by limiting its consumption for ATP depends on the cell's respiratory reserve.

Prediction. Compare direct PEP delivery with partial restriction of its conversion to pyruvate in old fibroblasts at equal intracellular PEP. The primary endpoint is the ability to maintain the epithelial barrier in coculture, with ATP and cGAMP as additional endpoints. The functional difference between the methods should depend on respiratory reserve and be related to energetics.

Mechanism. Reversible inhibition of EGLN1/2/3 by daprodustat allows temporal separation of the EPO-dependent erythropoietic response and the VEGFA-dependent vascular leakage response.

Prediction. Compare temporal exposure profiles in linked human EPO-producing, erythroid and endothelial models at the same yield of mature erythroid cells, measuring HIF, EPO, VEGFA and barrier function; neutralize VEGFA to test the source of leakage. One profile is predicted to produce at least 30% less leakage while retaining at least 90% of the erythroid effect. The hypothesis is refuted if separation is impossible at achievable exposure or leakage is unrelated to the proposed VEGFA branch.

Mechanism. Restoring cardiolipin before exercise prevents early energetic damage in aged muscle.

Prediction. Compare CRLS1 before, during and after identical exercise in naturally aged mice, with equal total exposure and workload. Prior restoration is expected to produce greater specific force after completion of the course and withdrawal of the intervention; additional outcomes are muscle fiber damage and respiratory reserve. The hypothesis is refuted if the advantage disappears after matching workload or is explained solely by greater muscle mass.

Mechanism. Normalization of phosphatidylcholine restores responsiveness to moderate dietary restriction in the setting of impaired mitochondrial plasticity.

Prediction. Conduct a factorial experiment only after the reproductive period, comparing phosphatidylcholine, dietary restriction, their combination, and a control, while controlling actual consumption and bacterial food. The interventions should show a positive interaction on remaining lifespan, the primary endpoint, with mobility and respiration as secondary endpoints; the benefit should not be entirely attributable to altered bacterial growth.

Mechanism. Binding to plasma proteins changes the PAGln dose available to cells at a fixed total PAGln level.

Prediction. In a vascular model from older donors, independently vary total and free PAGln in defined protein media, controlling for the proteins themselves. Measure barrier permeability after washout and analytically measure the free fraction. At a fixed total level, changing the free fraction is expected to change cellular accumulation and function.

Mechanism. Periods of restored translation preserve the renewal of short-lived repair proteins.

Prediction. In adult Drosophila, compare reversible regimens of translation modulation and measure repair protein turnover, proteostasis, and survival. Selectively suppress their restorative synthesis during the pauses. If suppressing restorative synthesis does not eliminate the advantage of pulsing, the explanation is rejected.

Mechanism. Continued exposure to the microbial ADP-heptose signal limits the persistence of the effect of ALPK1 blockade; restoring the intestinal barrier prolongs the effect by removing the signal source.

Prediction. In a model of mixed normal and Tet2-mutant haematopoiesis, compare ALPK1 correction, barrier restoration, their combination and a control. Measure metabolite influx and use the slope of mutant-fraction growth after withdrawal, with overall haematopoiesis preserved, as the primary outcome. Reintroducing the metabolite should eliminate the benefit of the barrier intervention; failure to do so, or clone dynamics independent of exposure, rejects the relapse mechanism.

Mechanism. Restoring GPX4 protects CD4+CD8+ thymocytes from ferroptosis without globally correcting m6A.

Prediction. Compare targeted restoration of Gpx4, restoration of Mettl3, a catalytically inactive GPX4 control and a delivery control. GPX4 is predicted to restore a substantial fraction of functional output while the age-related m6A profile persists, assessed by production of new functional naive T-cells with a diverse receptor repertoire. Check the response to model vaccination, positive and negative selection, signs of autoimmunity and lipid peroxidation. Increased survival without improved output, or at the cost of impaired tolerance, would mean the proposed bypass does not solve the problem.

Mechanism. Cardiolipin synthesis and remodeling constrain the benefit of enhanced mitophagy by determining whether functional membrane creation keeps pace with organelle removal.

Prediction. In adult worms, factorially combine moderate mitophagy induction with changes in the activity of cardiolipin-remodeling enzymes; measure mitochondrial removal flux, the lipidome, respiration and lifespan, and use genetic restoration of the target enzyme. The hypothesis is refuted if the effect of mitophagy is independent of cardiolipin restoration at comparable baseline viability.

Mechanism. The functional benefit of reducing pathological O-GlcNAc burden depends on preserving glycosylated matrix synthesis.

Prediction. Compare partial brief and continuous OGT suppression with a control, matching early SASP reduction. Trace UDP-GlcNAc into cytosolic modifications and secreted components. The primary outcome is cartilage resistance to repeated compression after withdrawal; monitor viability and new aggrecan. The proposed constraint is not supported if equal SASP suppression always produces equal mechanical outcomes regardless of matrix flux.

Mechanism. Suppress hepatic LPA with lepodisiran to reduce Lp(a) delivery to the valve before the calcium front becomes established.

Prediction. Perfuse valve microtissues with equal total calcium but different proportions of new deposits with reconstructed plasma before and after Lp(a) reduction; test purified Lp(a) add-back, the rate of new calcium deposition, and mechanical compliance. Suppression of new calcification will be at least 30% stronger in tissues with a larger active front, with compliance preserved.

Mechanism. The benefit of CTSC/DPP1 inhibition with brensocatib depends on preserved ciliated-cell regeneration through the GMNC–MCIDAS–FOXJ1 program in bronchial epithelial basal cells.

Prediction. Grow air–liquid interface cultures of basal cells from older donors with bronchiectasis and measure their ability to generate FOXJ1-positive cells before treatment. Compare secretions from neutrophils matured with and without DPP1 blockade, normalizing cell numbers, with particle transport speed across the epithelium as the primary measure. MCIDAS suppression should abolish recovery despite a preserved antiprotease effect; restoring MCIDAS in a separate exploratory arm should restore ciliated differentiation. Control mucus viscosity and volume. The proposed minimum functional gain is 25% in transport speed. The hypothesis is rejected if this gain is unrelated to the previously measured reserve or persists after MCIDAS suppression. A change in FOXJ1 without mucus movement is insufficient.

Mechanism. Residual and returning soluble Aβ retains the ability to initiate aggregation after deposits have been reduced. Lecanemab neutralizes Aβ protofibrils.

Prediction. In a human neuroglial model after equal clearance of deposits, compare repeat-exposure intervals determined by a seeding assay or total Aβ mass; remove and reintroduce the soluble fraction, conduct blinded analysis and separately measure residual antibody in the assay. The regimen guided by seeding activity is predicted to preserve at least a 20% better network response at the same total antibody exposure.

Mechanism. PDE5A inhibition by tadalafil produces local vasodilation, but a concurrent fall in systemic perfusion pressure can offset the microvascular benefit when autoregulation is impaired.

Prediction. Holding pressure constant will restore at least 25% of the lost oxygen benefit specifically in the group with weak autoregulation. A neurovascular chip from older donors will compare equal tissue exposure under fixed pressure and a pressure-decline profile observed clinically, measuring capillary transit heterogeneity, oxygen and neuronal response, with autoregulatory reserve assessed separately.

Mechanism. Local TRPV1 activation by capsaicin is followed by defunctionalization of sensory endings; Schwann cell support enables subsequent recovery of protective sensation.

Prediction. In a primary skin–nerve model from older donors, compare capsaicin and vehicle at different levels of functional Schwann cell reserve, tracking ending regrowth, responses to a controlled thermal stimulus, and spontaneous discharges; replenish supporting cells as a causal control. With high reserve, recovery of the protective response will be at least 25% better at the same reduction in spontaneous discharges.

Mechanism. Exercise creates a limited recovery phase in which mitophagy induction can return muscle stem cells to a competent state; after that phase, it cannot. The limiting factor is the timing of induction relative to recovery rather than total dose.

Prediction. In people aged 65-80 years, measure muscle-biopsy mitophagy markers at 0, 24, 48, 72 hours after standardized exercise to establish the window. Randomize treatment to uralicin A within the identified window, the same course outside it, chronic administration and placebo, matching total exposure by plasma metabolite level. The window regimen should produce greater specific contractile force at 72 hours and after repeated exercise than the outside-window and chronic regimens, with confirmed mitophagy induction in biopsy samples. A parallel preclinical muscle-specific ATG5/PINK1 test should establish dependence on mitophagy in stem cells. No phase differences in markers, no superiority of the window regimen, persistence of benefit after eliminating stem cell mitophagy, or an effect fully explained by microbiota metabotype or plasma metabolite level without a relationship to mitophagy markers would refute the corresponding claims.

Mechanism. Activation of GIPR/GLP1R by tirzepatide can temporarily slow gastric emptying, potentially prolonging amino acid availability to muscle MTORC1. This effect weakens with repeated exposure; persistent gastroparesis in all recipients and direct effects of tirzepatide on myocytes are not assumed.

Prediction. Reproduce amino acid profiles measured in older patients before and early in prescribed therapy in mature muscle microtissues, keeping total amino acids, insulin and energy equal. Compared with a restored short pulse, the prolonged profile should produce at least 15% less protein synthesis, measured by stable-label incorporation, with synthesis restored by changing pulse shape at the same area under the curve; subsequent force is also measured, and MTORC1 inhibition serves as a mechanistic control. No difference, or disappearance of the difference after matching insulin and energy, rejects the mechanism. The 15% threshold is proposed, not measured.

Mechanism. Consumption of available NAD precursor by the stromal niche limits resource availability to oocytes; stromal CD38 reduction is proposed to relieve this barrier.

Prediction. Compare an NAD precursor, local stromal suppression of Cd38, their combination and a control. Confirm targeting and trace labeled precursor in stromal cells, granulosa cells and oocytes before functional assessment. The combination is predicted to increase resource availability to oocytes and preserve function without accelerating reserve depletion, assessed by the duration of preserved fertility, viable offspring and the primordial follicle reserve after treatment, while monitoring accelerated follicle recruitment. A change only in total tissue NAD, or no effect of stromal suppression on labeled-substrate distribution, would invalidate the proposed mechanism.

Mechanism. Neutralizing IL6 with ziltivekimab lowers HAMP/hepcidin and mobilizes iron through SLC40A1/ferroportin, but muscle benefit depends on iron uptake through TFRC.

Prediction. Connect hepatocyte cultures, iron-loaded macrophages and mature muscle microtissues, with a parallel erythroid compartment, and compare control with ziltivekimab using plasma from older donors with inflammation and a stable iron isotope label. Measure label distribution, heme synthesis and muscle contraction force. Restoring hepcidin should abolish iron mobilization; reducing TFRC only in the muscle compartment should abolish muscle benefit while preserving iron release. The proposed minimum effect is a 15% increase in muscle specific force with corresponding label uptake. The mechanism is rejected if muscle force is independent of iron flux or remains improved after disruption of the HAMP/SLC40A1/TFRC chain. Direct toxicity and iron overload must be excluded; increased hemoglobin alone does not confirm the hypothesis.

Mechanism. Loss of nuclear barrier selectivity independently limits the effectiveness of NUP50 in restoring metabolic adaptation.

Prediction. Before intervention, classify old cells by passive nuclear permeability and active import, then normalize NUP50 within each stratum while matching strata for baseline viability. Assess survival and energy recovery after the same metabolic challenge as the primary endpoint. Improvement should be restricted to cells with preserved nuclear transport selectivity; a functional response independent of baseline permeability across a sufficient range of damage would refute the hypothesis.

Mechanism. Canakinumab neutralizes IL1B, targeting the component of synovial-fluid bioactivity that disappears upon IL-1β neutralization.

Prediction. Apply synovial fluid from different donors to paired cartilage explants with antibody or isotype control, match baseline severity, and restore IL-1β after antibody removal. Assess proteoglycan loss and resistance to cyclic loading. The high neutralizable-bioactivity group is predicted to show at least a 25% reduction in loss of mechanical strength, while the low group shows less than half the effect; these are proposed criteria, not observed results.

Mechanism. The benefit of PLA2G7 correction depends on maturation of the inhibitory circuit and functional integration of neural progenitor descendants into it.

Prediction. Compare targeted PLA2G7 correction with control while tracing NSC descendants, using memory as the primary outcome and measuring excitation–inhibition balance. Temporarily silencing the labeled descendants after maturation should eliminate the benefit; control for total neuron number and movement. If improvement persists during functional silencing and is unrelated to the inhibitory network, the proposed mechanism of the lasting effect is rejected.

Mechanism. FOXO1 is a necessary mediator of the beneficial range of mechanical stimulation: restored cellular tension increases chromatin accessibility at its locus and enhances its expression.

Prediction. Compare several levels of mechanical stimulation by measuring cellular tension, FOXO1 accessibility, damage markers, and osteogenesis, then repeat with FOXO1 suppression. Effective FOXO1 suppression should eliminate the beneficial response, and the benefit should reflect restored cellular function rather than selection of surviving cells.

Mechanism. Physical activity preserves capillary reactivity, enabling rapid redistribution of flow during exertion.

Prediction. Measure microcirculatory dynamics in old mice under the same exertion. Compare interventions that preserve vessel density but differ in pericyte reactivity, and test whether selectively disrupting reactivity abolishes the effect of exercise. If longevity is preserved despite lost reactivity while the other effects of exercise remain intact, causal necessity is rejected.

Mechanism. An epithelial P2Y2 pulse activates the beneficial repair signal in a naturally aged thymus, contingent on an epithelial reserve capable of a FOXN1-program response.

Prediction. Compare an epithelium-directed P2Y2 pulse with control in old uninjured mice, using a young post-injury repair model as a positive control. The primary outcome is new functional naive T-cells, with checks of epithelial P2ry2 dependence, negative selection and absence of increased thymocyte death. A response confined to young injured tissue, or enlargement of the old thymus without quality T-cell output, would reject transfer of the mechanism.

Mechanism. Selective restoration strengthens the competitive advantage of cells capable of maintaining fully functional tissue.

Prediction. In reconstructed epidermis from elderly donors, selective restoration of COL17A1 will improve barrier function more than uniform restoration at the same average protein level. Establish functional competence before intervention through the ability to maintain organized epidermis in independent sister cultures; compare selective restoration, uniform restoration and an untreated control, measuring clonal composition and diversity, permeability and recovery after repeated injury. Control initial cell proportions and confirm functional COL17A1 localization. Altered competition without better barrier function, or an advantage caused only by uncontrolled expansion of one clone, would not support the hypothesis.

Mechanism. HSD17B13 has a proposed structural contribution to harm on lipid droplets that is independent of its catalytic activity. Reducing the protein through ARO-HSD would remove this contribution, whereas retaining catalytically inactive HSD17B13 would preserve it.

Prediction. After suppressing endogenous mRNA in primary hepatocytes, restore siRNA-resistant wild-type or catalytically inactive HSD17B13 with identical localization and stability, and measure droplet morphology, lipid stress, cell death and secretome effects on stellate cells. The catalytically inactive protein is predicted to restore at least half of the pathological signal eliminated by RNA suppression.

Mechanism. LAMP2A-dependent CMA affects endocytic removal and lysosomal degradation of SIRPα, thereby affecting inhibition of phagocytosis. SIRPα is not proposed to be a direct soluble substrate of CMA.

Prediction. Compare macrophages with intact and impaired LAMP2A-dependent CMA for SIRPα turnover, engulfment of senescent cells and response to a CMA activator, controlling the viability of both cell types. Restoring CMA flux should improve phagocytosis, potentially partially; no improvement at comparable viability, or SIRPα turnover unrelated to CMA status, would refute the hypothesis.

Mechanism. NCOA7 restores stress granule clearance through granulophagy in granulosa cells, improving the quality of existing follicles without accelerating depletion of the reserve.

Prediction. Compare granulosa-cell-restricted restoration of Ncoa7, a delivery control, and a functionally impaired construct previously validated for granulophagy. The primary outcome is the duration of the ability to produce viable offspring under standardized observation; additional measurements include oocyte quality, stress granule fate, primordial follicle count, and recruitment rate. Functional improvement should accompany preservation of the reserve after treatment ends. An early benefit accompanied by accelerated depletion, or a benefit independent of restored stress granule clearance, would reject the proposed mechanism.

Mechanism. Proteolytic TTR fragments, including 49–127, may continue to form and retain seeding activity despite tetramer stabilization, sustaining residual amyloid formation and cardiomyocyte toxicity.

Prediction. Use wild-type TTR and tissue extracts from older donors with ATTR at measured free acoramidis exposure and physiologically justified fluid shear. Distinguish pre-existing from newly formed fragments by isotope labeling, independently confirm tetramer stabilization, and measure new fibrils and cardiomyocyte toxicity. Removing fragments or selectively limiting the plasmin contribution in vitro should reduce residual toxicity, and adding back isolated fragments should restore it. Disappearance of new fragment flux during confirmed tetramer stabilization, or failure of fragment removal to reduce residual damage by at least 20%, would reject the hypothesis; old fragments in a biopsy alone are insufficient.

Mechanism. CMPK2 supports mitochondrial DNA accumulation and its appearance in the cytosol, thereby enhancing the type I interferon response in senescent p21+TREM2+ macrophages.

Prediction. Compare senescent macrophages with suppressed CMPK2 against a genetic control for mitochondrial and cytosolic DNA, IFN-α/IFN-β and viability, confirming the cellular state with several independent features. CMPK2 suppression should reduce cytosolic DNA and IFN-α/IFN-β expression, with the reduction not fully explained by loss of living cells.

Mechanism. Trained immunity involves myelopoietic progenitors, and SIRT3 is associated with age-related HSC dysfunction; their combination predicts that the effect depends on the stage of cell differentiation.

Prediction. In mouse models, compare interventions in HSC and mature myeloid populations using identical training and restimulation. Intervention at the HSC stage is expected to reduce the excessive inflammatory response more strongly, assessed through IL-6/TNF in descendants and preservation of HSC function, with separate checks of expression level and intervention cell specificity. The hypothesis is refuted if comparable interventions show no advantage for early intervention or if cell death fully explains the anti-inflammatory effect.

Mechanism. Suvorexant blocks HCRTR1/HCRTR2. In the presence of nocturnal neuronal hyperactivity, the proposed effect is reduced Aβ production rather than increased Aβ removal.

Prediction. In a controlled mechanistic study, compare groups with different levels of nocturnal hyperactivity, measure Aβ production and removal kinetics using stable isotopes, and relate these to EEG, sleep and drug concentration at matched circadian phases. In the hyperactivity group, Aβ production is predicted to decrease at least 15% more, without a comparable change in the removal rate.

Mechanism. Functional benefit from FTL1 reduction depends on recovery of energy reserve, rather than normalization of iron measures alone.

Prediction. Use a factorial comparison of partial Ftl1 suppression and validated metabolic support, including verification that the support reaches the intended compartment. Assess synaptic transmission during prolonged stimulation as the primary outcome, followed by memory, while matching viability and iron load and independently limiting energy reserve. Functional benefit should depend on energetic recovery; independence from that recovery or complete explanation by death of cells with high FTL1 would refute the hypothesis.

Mechanism. SEMA3A–NRP1 regulates fenestrae. Suppressing this signal in liver endothelium is proposed to restore fenestrae and insulin transport, thereby improving insulin action in the liver.

Prediction. An intervention directed predominantly at liver endothelium should restore fenestrae and labeled insulin transport and improve suppression of hepatic glucose production at a controlled insulin level in old animals compared with controls, with young animals as a physiological reference. Measure insulin clearance and separately test isolated hepatocyte responses, controlling for liver perfusion, systemic hemodynamics, tissue injury and intervention specificity. Restored pores and insulin transport without improved liver metabolic function would refute the hypothesis; an effect explained solely by altered blood flow or direct action on hepatocytes would not support the proposed endothelial chain.

Mechanism. The capacity of cardiac macrophages to engulf and complete lysosomal digestion of exported exophers limits the benefit of increased cardiomyocyte export.

Prediction. First establish an intervention that increases exopher export from cardiomyocytes without increasing cell death, then compare it factorially with restoration of MERTK-dependent cardiac macrophage function. The combination is predicted to outperform increased export alone in the relaxation and contraction reserve of the aged heart under a standardized load. Distinguish exopher uptake from completed lysosomal digestion and quantify extracellular residue. Benefit from export independent of clearance, or changes in autophagy markers without functional improvement from the combination, would not support the hypothesis.

Mechanism. Inhibit ACLY with bempedoic acid instead of HMGCR with a statin at comparable reductions in hepatic apoB; hepatic activation and the absence of the active metabolite in muscle underlie the proposed preservation of specific force.

Prediction. In linked primary liver and muscle microtissues from older donors, match exposures for equal apoB reduction; measure both drugs and the active metabolite by compartment, specific force, and recovery after exercise, and separately alter muscle SLC27A2 to test causality. Specific force will be preserved at least 15% better; artificially increasing activation in muscle will diminish the advantage.

Mechanism. Neutralizing IL6 with ziltivekimab acts on endothelial signaling through the IL6–soluble IL6R–IL6ST/gp130 complex. The proposed determinant of barrier restoration is the contribution of IL-6 trans-signaling to plasma activity.

Prediction. Apply plasma from elderly patients with chronic kidney disease to mature endothelial microvessels under flow, comparing control, ziltivekimab and soluble gp130-Fc within each donor and confirming pSTAT3 suppression. Measure the trans-signaling fraction before unblinding; the primary outcome is albumin permeability and the secondary outcome is monocyte adhesion. The proposed supporting result is at least 20% lower permeability in samples with a high trans-signaling fraction, with the effect reproduced by gp130-Fc; selective removal and return of soluble IL6R at equal IL6 should change the response, with cytokines and osmolarity controlled. The hypothesis is rejected if this benefit is excluded despite confirmed pSTAT3 suppression or if IL6R removal and return do not change the effect; CRP correlation alone is not causal confirmation.

Mechanism. PDE3/PDE4 inhibition by ensifentrine produces a smooth-muscle cAMP response, while a low lysosomal acidification reserve in alveolar macrophages limits phagolysosomal processing.

Prediction. Compare ensifentrine, a selective PDE4 control and a PDE3 control at measured exposure in paired cultures of bronchial smooth muscle and alveolar macrophages from older donors. Measure acidification reserve before treatment and uptake and degradation of standard nonliving particles, lysosomal pH and cytokines afterward, using the fraction of cargo degraded per viable cell as the primary metric. Cargo processing should decrease by at least 20% only in donors with low reserve, and restoring acidity should reverse the defect without abolishing muscle relaxation. Failure to reproduce the association, or a defect explained solely by cell death, would reject the mechanism.

Mechanism. Changes in the state of surviving old cells contribute to functional recovery in distant tissues.

Prediction. Before injury, label stable cell populations in a distant tissue and separately track new cells. Compare regenerating old animals, uninjured animals and a control with comparable nutrition, using the function of the selected distant tissue as the primary outcome and analysing molecular indicators of age within previously labelled cells. If improvement occurs only in new cells while surviving cells retain their previous dysfunction, the hypothesis is rejected.

Mechanism. Precursor recruitment by the Postn-positive perivascular mesenchymal niche and restoration of the epithelial program through FOXN1 are proposed to address two sequential limitations in the same old thymus.

Prediction. In old animals, compare FOXN1, Postn-positive cells, their combination and control. With mesenchymal cell engraftment and epithelial program restoration confirmed, the combination should exceed either intervention alone in the absolute output of new functional T cells; track labeled precursor entry, receptor diversity and responses to a new antigen, while checking self-tolerance, autoantibodies and preservation of existing immune memory. Increased thymus size without superior functional T-cell output, or impaired self-tolerance, would refute the hypothesis. If precursor entry does not limit the response to FOXN1, the proposed sequence of limitations would not be supported.

Mechanism. Complement-dependent inflammatory signaling modifies the vascular barrier benefit of ZFYVE21 restoration, with NIK suppression proposed to remove the adverse interaction.

Prediction. First verify ZFYVE21 deficiency in the aged endothelium under study. In a factorial experiment, compare protein restoration under low and high controlled complement activation, and separately test NIK blockade. The primary microvascular outcome is albumin permeability, followed by confirmation of the direction of the effect on renal filtration in vivo; measure NO, inflammatory response, and viability. Complement should weaken or alter the benefit of restoration, while NIK suppression should remove the adverse interaction. Equal effects in both conditions despite confirmed effects on the pathways would reject the proposed condition for therapy selection.

Mechanism. Hepatic expression compensates for insufficient external signals without requiring prior correction of all intrinsic HSC defects. FLT3L supports thymic recovery, IL-7 supports T-cell homeostasis, and DLL1 provides a contact-dependent Notch signal.

Prediction. Compare temporary hepatic expression of the three factors with a control for thymopoiesis, T-cell composition, and response to a vaccine antigen; track changes after withdrawal and HSC composition. Expected support is increased thymopoiesis and improved functional immune response, with an effect that weakens after withdrawal and no required change in HSC composition. Confirmed expression without improved functional immune response, or only redistribution of existing T-cells without increased thymopoiesis, refutes the hypothesis.

Mechanism. The amount of eIF5A, rather than the supply of spermidine, limits the response of a subset of old B cells with a preserved proportion of hypusinated eIF5A.

Prediction. Stratify B cells from elderly donors by total and hypusinated eIF5A, then compare spermidine, eIF5A restoration and their combination. The primary outcome is the quality of the antigen-specific response upon repeat stimulation; measure TFEB and completed autophagic flux, and use a TFEB bypass to test the location of the limitation. Restoring eIF5A should overcome nonresponse better than spermidine alone. Failure of the baseline ratio of eIF5A forms to predict response, or failure of protein restoration to overcome nonresponse, would reject the selection hypothesis.

Mechanism. TIMP2 improves the selectivity of microglial clearance of damaged synaptic components relative to functioning synapses, with dependence on cargo recognition.

Prediction. Compare TIMP2 with control in aged neuron–microglia cultures and then aged mice, separately labeling damaged fragments and functionally active synapses and measuring completed digestion. Increased selectivity should be associated with preserved synaptic transmission, the primary outcome, followed by memory assessed with movement controlled for. Independently altering opsonization tests dependence on cargo recognition; increased CD68 alone does not count as success.

Mechanism. Sufficiently prolonged extracellular matrix softening resets persistent mechanical memory in fibroblasts.

Prediction. First determine in cultures of old fibroblasts the exposure duration after which activation does not return at the original stiffness. Then compare short and long reversible matrix-alteration regimens in old mice at equal total exposure, assessing fibrosis, organ function and survival. If duration does not affect the persistence of the cellular state and longevity, the hypothesis is rejected.

Mechanism. The benefit depends on completing differentiation and switching off the SOX5 program after matrix formation without fixing a foreign cellular identity.

Prediction. Compare brief and continuous SOX5 induction with a control, using the same initial number of cells. The primary outcome is the mechanical function of the formed cartilage after withdrawal. Track identity, clonal diversity, hypertrophy and mineralization; temporarily restore SOX5 to test reversibility. If benefit requires continuous expression or is accompanied by pathological mineralization, the hypothesis of completed restoration is rejected.