Structural Conditions for Cellular Senescence Entry:
A Co-occurrence Hypothesis

1. Introduction

Cellular senescence is widely described as an irreversible growth-arrest state, yet the logic by which cells enter that state remains incompletely specified. The boundary between reversible arrest and irreversible senescence is still difficult to draw, and the same class of damage can lead some cells toward cell death while directing others toward senescence (Salama et al., 2014). This uncertainty matters not only conceptually but also practically, because it complicates the design of interventions that aim either to prevent senescence entry or to target senescent cells once they appear.

Current consensus already treats senescence as a multi-marker state. In practice, senescence is usually judged through the joint presence of several indicators, such as growth arrest, SA-beta-gal activity, p16 or p21 expression, and secretory changes, rather than through a single universal marker (Gorgoulis et al., 2019). However, this consensus does not yet provide a formal rule for entry. It does not specify which combinations are necessary, which are sufficient, or whether co-occurrence itself should be treated as an independent variable rather than as a descriptive feature of an already-established senescent state. To our knowledge, this level of co-occurrence has not been formalized as an explicit structural variable.

This paper approaches the problem from outside cell biology. It draws on D-Architecture, a structural framework derived from the question of what must exist if a selection system is to remain open to future-selectable continuation (Jung, 2026a; Jung, 2026c). The present argument does not use the full framework. Instead, it uses only the minimal maintenance invariant and the regulation layer needed to derive a pattern-based collapse-approach signal. The aim is not to claim that cells are literally identical with D-Architecture, but to ask whether a minimal fragment of that framework yields a testable organizational rule for senescence entry.

The resulting working hypothesis is straightforward. Senescence entry is proposed to depend not on the severity of any single marker, but on the sustained co-occurrence of at least two among three structurally distinct channels: restoration blockage, selection invalidation, and boundary rigidification. On this reading, the problem is not simply whether markers become abnormal, but whether multiple recovery-related paths are closing at the same time.

The paper makes four specific contributions. First, it introduces the minimal structural premises used for the argument and defines the two-of-three co-occurrence condition in a form that is not reducible to a numerical threshold. Second, it presents a role-level mapping between the three channels and candidate cellular phenomena. Third, it derives falsifiable predictions and explicit counterexample conditions, including conditions under which the model should be rejected. Fourth, it proposes an experimental design in which zero-, one-, two-, and three-channel conditions can be compared under fixed operational definitions. The claim is therefore neither that senescence has been deduced from first principles nor that cells are exhaustively captured by a single framework. The narrower claim is that co-occurrence can be treated as a testable structural entry condition.

2. Minimal Structural Premises

2.1 Minimal Premise

D-Architecture is a formal framework that asks what structures are necessary if a selection system is to persist without closing its own future-selectable continuation. The present paper does not use the full framework. It relies only on two premises: the minimal maintenance invariant, I_min, and the regulation layer derived from that invariant.

I_min is the weakest persistence condition used in the argument. At time t, the system must remain such that some future-selectable option is still available at a later time t'.

The point of this invariant is not optimization, success, or performance. What matters is only whether future-selectable openness remains non-empty. In that sense, I_min is not a goal function but a minimal non-collapse condition.

Regulation follows when repeated selection begins to consume the very openness on which future selection depends. A persistent selection system therefore requires structures that slow, buffer, postpone, or suspend its own narrowing tendencies. Regulation is not introduced here as an efficiency device. It is introduced as the minimal structural layer required to keep I_min from being undermined by the system's own repeated selections.

2.2 Signal of Collapse Approach: Theta_near

Within this regulation layer, the relevant signal is not a numerical threshold but a structural one. The present paper uses Theta_near to denote a pattern-based signal that justifies regulatory intervention when future-selectable openness is beginning to erode.

For the purposes of this paper, the expression is read through three operational categories.

• Restoration blockage: restoration candidates are no longer being regenerated.
• Selection invalidation: repeated selection no longer produces cumulative improvement.
• Boundary rigidification: suspension and regulatory holding no longer relax but remain persistently engaged.

The important point is not simply that the system has become worse. The important point is that recoverability is shrinking. Theta_near therefore should not be read as an absolute-severity threshold. It should be read as a structural signal that recoverable openness is being reduced by the sustained co-occurrence of distinct patterns.

2.3 Why Two of Three?

The two-of-three condition is not introduced here as an empirical tuning parameter. It follows from the operational logic of the framework itself. If only one pattern is present, transient fluctuation and structural approach remain difficult to distinguish. If all three patterns are required before intervention becomes justified, the signal arrives too late, at a point where recoverability may already be nearly exhausted.

For that reason, the threshold is placed at the smallest co-occurrence level that exceeds isolated fluctuation without waiting for full collapse approach. One sustained pattern is insufficient. Three sustained patterns may be diagnostically stronger but operationally delayed. Two sustained patterns therefore function as the minimal structural condition for judging that the system is approaching recoverability loss rather than merely exhibiting noise or temporary stress.

This is also why the rule is tied to pattern co-occurrence rather than to the amplitude of a single marker. Within the framework, single-indicator judgment is structurally disallowed, because state approximation is always partial and error-bearing. What matters is the simultaneous and sustained presence of distinct patterns, not the extremity of any one indicator taken in isolation.

2.4 Mapping Rule

The present paper treats cells as systems that move across multiple possible state-transitions, including continued cycling, reversible arrest, senescence entry, and cell death. In that restricted sense, cells can be read as state-dependent selection systems.

However, the paper does not claim that cells are literally identical with Theta_near systems. The role of the framework is narrower. It provides three structural categories against which known senescence-related phenomena can be organized and tested. The question is therefore not whether cells instantiate D-Architecture in full, but whether restoration blockage, selection invalidation, and boundary rigidification can serve as a useful structural partition for interpreting senescence entry.

The following section examines whether candidate cellular phenomena can be arranged under those three categories, and whether their sustained co-occurrence is plausibly related to entry into senescence.

3. Cellular Mapping of the Three Channels

3.1 Mapping Overview

The broader cellular comparison from which the present paper is drawn examined 33 D-Architecture structures against candidate functions in eukaryotic cells. On that comparison, 26 items were judged to show strong homology, 6 were treated as partial correspondences, and no explicit contradiction was identified. In this context, however, strong homology does not mean ontological identity. It means role-level concordance: the relevant structural function can be observed in cells and its direction of operation is consistent with the formal role assigned in D-Architecture.

The present paper does not reproduce the full mapping catalogue. It restricts attention to the three categories required for the senescence-entry hypothesis: restoration blockage, selection invalidation, and boundary rigidification. The question is whether known senescence-related phenomena can be arranged under these categories without forcing a literal one-to-one identity between cellular machinery and the formal framework.

3.2 Channel-by-Channel Mapping

Restoration blockage corresponds, at the cellular level, to conditions in which the cell's ability to process damage and reopen viable continuation paths is reduced. Candidate cellular proxies include stalled autophagic turnover and accumulated burden on DNA repair pathways. Bulk autophagy functions as a route for sequestration, degradation, and recycling of damaged components (Mizushima and Komatsu, 2011), while DNA repair operates through multiple partially parallel pathways (Jackson and Bartek, 2009). When these restoration-related processes are jointly impaired or overloaded, the cell can be read as losing room for recovery rather than simply accumulating damage. Persistent DNA-damage signaling provides one representative biological sign of such unresolved burden (Rodier et al., 2009).

Selection invalidation corresponds to situations in which repeated state-selection no longer yields meaningful recovery or re-entry. A candidate cellular pairing is sustained p21-associated arrest together with persistent gamma-H2AX foci. Persistent p21 expression can lock the cell into continued cell-cycle arrest (Abbas and Dutta, 2009), while persistent gamma-H2AX marks unresolved damage signaling rather than successful return to baseline (Rogakou et al., 1998). Read together, this pairing does not merely indicate stress. It suggests that repeated attempts at stabilization are no longer producing cumulative state improvement.

Boundary rigidification is the least certain of the three channels. The relevant structural meaning is that the system no longer reopens its internal flexibility once regulatory holding has been engaged. Candidate cellular proxies include senescence-associated heterochromatin foci and broader forms of proteostasis collapse. SAHF has often been discussed as a reorganization of chromatin into a more rigidized compartment, but it is neither universal across cell types nor present in every senescence program (Kosar et al., 2011). Chronic proteostasis disruption also points to reduced internal flexibility and impaired adaptive reorganization (Labbadia and Morimoto, 2015). For that reason, this third channel is treated not as an established biological equivalent, but as a plausible candidate class for operational testing.

3.3 Scope and Limits of the Mapping

Three cautions are necessary. First, the weakest part of the present mapping is not the existence of candidate phenomena but the lack of a settled integrated indicator for boundary rigidification. The problem is therefore not that the relevant biology is absent, but that a stable operational index for that structural category is not yet established.

Second, the three channels are not biologically independent in a strict sense. DNA damage, secretory burden, mitochondrial dysfunction, proteostasis stress, and chromatin reorganization can reinforce one another through multiple feedback routes (Dorr et al., 2013; Mathew et al., 2007; Correia-Melo et al., 2016; Goodarzi et al., 2008). As a result, observed co-occurrence may represent either genuinely distinct channel convergence or a cascading consequence initiated by one dominant disturbance. This limits causal interpretation even where the mapping remains descriptively useful.

Third, the present mapping is operational rather than identificatory. The claim is not that cells literally are Theta-near systems, nor that D-Architecture provides a complete ontology of senescence. The narrower claim is that known senescence-related phenomena can be organized under three structural categories in a way that yields explicit predictions about entry conditions.

4. Predictions and Falsification Criteria

4.1 Core Prediction

If the preceding mapping is broadly valid, a clear prediction follows for senescence entry. Persistence of any single channel on its own should not be sufficient for entry into senescence. Under such a condition, the cell may remain in a reversible arrest-like state rather than crossing into irreversible senescence. By contrast, sustained co-occurrence of at least two among the three channels should predict senescence entry.

The stronger form of the hypothesis is that the decisive variable is not the identity of one privileged combination, but co-occurrence itself. In other words, any two-channel combination should be able to support senescence entry if the model is correct. The proposal therefore differs from current multi-marker practice in an important way. Existing practice accepts that multiple markers are needed, but it does not formalize what kind of multi-marker relation matters. The present hypothesis proposes that what matters is not simply which markers appear, but whether markers drawn from structurally distinct categories persist together.

4.2 Falsification and Counterexamples

For the hypothesis to count as a scientific claim, its failure conditions must be explicit. The model is challenged if only one particular two-channel combination induces senescence while other two-channel combinations do not. Under that outcome, the explanatory variable would no longer be co-occurrence itself, but the specific biological content of one favored pairing. In that case, the structural claim of the present paper would fail.

A more direct challenge comes from reports that sustained p21 expression may induce irreversible arrest or senescence-like outcomes on its own (Macip et al., 2002). This observation stands as a potential counterexample to the claim that one-channel persistence is insufficient. The paper allows two possible readings. Either the model is wrong, or p21-driven arrest is in fact accompanied by an unmeasured second-channel change. However, the second reading is not granted automatically. It is acceptable only if at least one additional channel change can be independently confirmed by a separate readout under the operational definitions adopted in this paper. If p21 alone reproduces senescence entry without such accompanying change, the model should be rejected.

4.3 Implications

If the prediction is supported, two practical implications follow. First, co-occurrence may provide a basis for defining an early warning signal for senescence entry. A cell need not already display maximally abnormal markers for the model to become relevant. If two among restoration capacity, repair effectiveness, and internal flexibility are simultaneously trending downward, that pattern may indicate approach toward the entry boundary even before any single marker becomes extreme.

Second, the proposal may reorient intervention strategy. If entry depends on the simultaneous closure of at least two channels, then intervention need not begin by correcting every altered process at once. A more tractable strategy may be to stabilize whichever channel can most plausibly be kept open, thereby preventing multi-channel closure from consolidating into senescence entry.

These implications remain conditional on an important assumption already discussed in the mapping section: the three channels must be sufficiently separable for operational use. If positive feedback between channels proves too strong, maintaining only one channel may not be enough to prevent transition.

5. Experimental Design and Operationalization

5.1 Experimental Overview

The central experimental prediction of the present hypothesis is simple: any two-channel combination, under simultaneous induction, should be associated with increased senescence entry, whereas a single channel on its own should remain insufficient for entry. To test this claim, the paper proposes an intervention design in which each of the three channels can be operationally assigned and compared across a fixed set of condition groups.

After establishing conditions putatively assigned to each channel separately, senescence-entry rates can be compared across four condition groups, yielding eight total experimental arms.

Control (0-channel): no channel deliberately activated.
Single-channel (1-channel): restoration blockage, selection invalidation, and boundary rigidification induced separately, for three total arms.
Dual-channel (2-channel): all possible two-channel combinations induced, for three total arms.
Triple-channel (3-channel): all three channels induced together.

For the model to be supported, senescence-entry rates should remain low in the control and single-channel conditions, but rise in the two-channel and three-channel conditions. More specifically, every possible two-channel combination should show a significant increase in senescence entry. At the same time, each arm must also be checked for unintended co-activation of non-target channels. If senescence is induced only by one privileged pairing, the model fails, because the explanatory variable would then be the specific content of that pairing rather than co-occurrence itself.

5.2 Structural Conditions in D-Architecture

D-Architecture provides a three-pattern structural scheme with a two-of-three threshold for sustained co-occurrence, but it does not specify which molecules should be measured or how long persistence must last in a cell-based experiment. It does, however, impose several structural constraints that should guide the design.

First, "sustained" should be read in terms of trend rather than absolute value. The framework disallows judgment based on a single indicator taken at a single magnitude. Accordingly, experimental co-occurrence should be operationalized not as a snapshot threshold alone, but as a non-reversing or non-recovering pattern observed across time.

Second, the two-of-three condition is an early structural signal, not a declaration of completed outcome. In the formal framework, Theta_near justifies regulatory intervention; it does not certify that collapse has already occurred (Jung, 2026c). Translated into the present cellular setting, two-channel co-occurrence is read as a senescence-prone state, whereas three-channel co-occurrence may be read as a stronger near-collapse or senescence-prone signature.

Third, the core question is not whether markers have become worse, but whether recoverability is shrinking. Channel activation should therefore be judged less by the absolute severity of any one marker than by evidence that recovery is slowing, stalling, or no longer reopening.

Fourth, time-course measurement is more appropriate than single-point observation. Because Theta_near is defined by sustained co-occurrence rather than by a static threshold, the experimental design should favor repeated measurement across time over one-time readout whenever feasible.

The operational definitions proposed below are therefore best understood as candidate experimental translations of these structural conditions, not as direct molecular deductions from the formal framework.

5.3 Operational Definitions

To implement the structural conditions of Section 5.2 at the cellular level, each channel must be given an explicit activation rule. Those rules must be fixed in advance. If the definitions are adjusted after inspection of the results, the falsification conditions lose force. The following candidates are proposed on the basis of current literature.

The threshold for each readout must also be fixed before the experiment begins. Acceptable approaches include fold-change from baseline, percentile-based cutoffs, or a pilot-defined criterion of non-decrease across time. The exact threshold may vary by cell line, but the rule must be fixed before outcome evaluation.

The temporal criterion for "sustained co-occurrence" must likewise be prespecified. In practical terms, two stages should be separated. First, channel co-activation observation: at least two channel markers must be simultaneously active and show no recovery trend across a defined time window, for example 48 to 72 hours, using time-course measurement. Second, senescence-entry judgment: senescence entry itself should be assessed later, for example 7 to 14 days after channel induction. Under the structural reading adopted here, two-channel co-occurrence indicates approach toward senescence entry, whereas three-channel co-occurrence may indicate a stronger approach signal. A faster entry point under the three-channel condition can therefore be treated as an exploratory outcome. The exact time windows should be calibrated in advance by cell-line-specific pilot studies.

As already noted in Section 3.3, boundary rigidification remains the least settled operational category. SAHF may fail to form depending on cell type (Kosar et al., 2011), which is why the present paper proposes chromatin accessibility and UPR activation as a candidate combination. The adequacy of that combination is itself part of what future testing must determine.

Together with the channel definitions, the criterion for senescence entry itself must also be fixed in advance. A reasonable candidate definition, consistent with current consensus (Gorgoulis et al., 2019), is a combination of SA-beta-gal activity, irreversible growth arrest, and p16 expression. Because p16 may remain low in early entry windows, sustained p21 expression or lamin B1 loss (Freund et al., 2012) may serve as auxiliary markers. At the same time, viability assessment and apoptosis-exclusion readouts are required so that acute toxicity or cell death is not misread as senescence entry.

5.4 Expected Results and Technical Challenges

The major technical challenges are as follows.

First, feedback control across channels. As discussed in Section 3.3, induction of one channel may trigger change in another. If that occurs, a nominal one-channel condition may in practice become a hidden two-channel condition, weakening the logic of the design itself. After each intended induction, non-target channel readouts must therefore be checked explicitly.

Second, cell-type selection. Because SAHF formation and related chromatin behaviors depend on cell type (Kosar et al., 2011), results may vary substantially across cell lines. At minimum, the model should be tested in both a line that readily forms SAHF, such as IMR90, and a line in which SAHF is weak or absent, such as BJ, so that the generality of the result is not mistaken for a single-cell-line artifact.

Third, distinction between senescence and cell death. Strong two-channel or three-channel induction may produce acute toxicity or apoptosis rather than genuine senescence entry. Senescence readouts must therefore be interpreted together with viability and apoptosis-exclusion measurements so that arrested or dying cells are not conflated with cells that have entered senescence.

6. Limitations

6.1 Operational Mapping Is Not Proof of Identity

The present paper offers an operational mapping that asks whether known markers of cellular senescence can be organized under the three categories associated with Theta_near. Even if that mapping proves useful, it does not establish the ontological claim that cells are literally implementations of D-Architecture. Observing role-level structural concordance is not the same thing as demonstrating mechanistic identity between the two systems.

6.2 Operational Definitions of the Three Channels Remain Incomplete

The testability of the present hypothesis depends on whether the three channels can be measured as operationally separable units. That condition is not yet satisfied evenly across all three channels.

Boundary rigidification remains the most uncertain channel. In D-Architecture, its structural meaning is that regulatory holding, or the functional equivalent of persistent buffer or delay engagement, no longer relaxes once activated. No settled direct index of that condition currently exists in cell biology. SAHF is one candidate, but it is not formed in all cell types (Kosar et al., 2011), and proteostasis collapse is another candidate, but without a settled quantitative criterion. Section 5.3 therefore proposed a combination of chromatin accessibility and UPR activation as a provisional proxy, but whether that proxy actually captures the intended structural meaning is itself part of what future testing must determine.

Restoration blockage carries a similar incompleteness. The current operational definition is centered on autophagy-related impairment and does not yet include an independent marker that directly captures DNA-repair saturation as a distinct dimension of reduced recoverability. Persistent DDR signaling (Rodier et al., 2009) is an indirect sign that repair has not been successfully resolved, but that is not the same as directly measuring exhaustion of restoration capacity.

Selection invalidation is relatively more clearly operationalized, but the three channels are not all defined at the same level of maturity. In other words, even if the hypothesis is structurally sound, the tools currently used to measure its channels remain uneven. If the measurement layer is weak, the experimental result may become unstable even when the underlying structural proposal is correct.

6.3 Channel Independence Is an Analytical Convenience

For purposes of testing, the present hypothesis treats the three channels as operationally separable units. In real cells, however, one failing channel can pull the others with it through broad positive feedback routes (Dorr et al., 2013; Mathew et al., 2007; Correia-Melo et al., 2016; Goodarzi et al., 2008). Channel independence is therefore an analytical convenience, not a biological fact.

This feedback structure creates two levels of limitation. Experimentally, even a nominally single-channel induction may co-activate other channels, making a pure one-channel condition difficult to construct. Interpretively, even when two-channel co-activation is observed, it may remain unclear whether two channels were independently impaired or whether one dominant disturbance triggered the other. In the latter case, the observed co-occurrence cannot be read straightforwardly as independent two-channel convergence. Time-course measurement may partially address this problem by tracking which channel emerges first, but fast feedback may still prevent clear causal separation.

6.4 Cell Type, Context, and Scope

The present predictions are proposed as a working hypothesis across multiple eukaryotic cell contexts, but actual testing can only begin in particular cell lines. Channel behavior may differ by cell type. Just as SAHF formation depends strongly on cellular context (Kosar et al., 2011), the relative fragility of the three channels and the strength of feedback coupling may also vary from one cell line to another. A positive result in one line would therefore not, by itself, establish a universal law.

In addition, all three channels in the present paper are defined within a cell-autonomous frame. Yet senescence in vivo is also shaped by influences outside the cell itself. SASP can promote neighboring-cell senescence through secreted factors (Rodier et al., 2009), and immune clearance can selectively remove senescent cells. The present three-channel model is limited to internal entry conditions and does not attempt to model those higher-level tissue processes. Whether SASP should be treated as part of one of the three channels or as an external pathway remains a valid follow-up question, but it is outside the scope of this paper.

Finally, the interpretive boundary among senescence, quiescence, and apoptosis remains difficult to stabilize fully. Strong induction conditions may produce acute toxicity or cell death rather than senescence entry, while weak single-channel conditions may remain difficult to distinguish from reversible arrest. Section 5.3 proposed fixed entry criteria together with apoptosis-exclusion measurements, but that uncertainty cannot be eliminated completely.

6.5 Theory-Driven Mapping Bias and Alternative Explanations

The mapping presented here was carried out by the proposer of D-Architecture. That creates a real risk of confirmation bias, because the same person who developed the framework is also selecting and organizing the corresponding biological cases. Since D-Architecture is abstract, it may also create the impression that post hoc mappings could be constructed for many complex systems unless negative cases are made explicit.

The most direct response to that concern would be to provide negative controls: systems in which the mapping fails or in which the predicted co-occurrence rule does not hold. The present paper does not yet include such negative cases. A useful next step would be to examine systems in which the regulation layer is organized differently and ask whether the same three-channel structure still applies. A second step would be blind mapping by independent evaluators who do not already know the D-Architecture partition in advance.

Finally, even if the predictions proposed here are experimentally supported, that would not mean D-Architecture is the only framework capable of explaining the phenomenon. Multistability models, network-based approaches, and other systems frameworks may also be able to generate similar predictions. The claim of the present paper is therefore not one of uniqueness, but of proposing a testable organizational rule for senescence entry.

7. Discussion

7.1 Re-reading Existing Observations

The present paper proposes one testable answer to the gap stated at the outset: the absence of a formal account that treats marker co-occurrence itself as an independent variable in senescence entry. The proposed answer is that senescence entry is predicted when any two of the three channels are sustained together, and that this claim should stand only together with explicit rejection conditions.

Seen through that lens, a re-reading of existing observations becomes possible. The co-appearance of multiple markers in senescent cells is already widely reported. Replicative senescence and oncogene-induced senescence alike have been described through concurrent DDR, cell-cycle arrest, chromatin reorganization, SASP, and related markers (Salama et al., 2014; Gorgoulis et al., 2019), with persistent DDR serving as one representative case (Rodier et al., 2009). Previous work has usually described such co-occurrence as a result or characteristic of senescence. The present hypothesis proposes a different reading. Co-occurrence may be interpreted not only as an effect of an already established state, but also as a candidate transition condition for entry into that state. The observations themselves do not change; what changes is the organizational rule under which the same observations are arranged.

7.2 Comparison with Alternative Frameworks

D-Architecture is not the only attempt to formalize cellular senescence. Multistability-style models, for example, treat senescence-related fate decisions as transitions among stable states in regulatory systems (Mombach et al., 2014). General systems theory describes open systems and their maintenance at a broad level (von Bertalanffy, 1950), but it does not directly provide the specific kind of co-occurrence rule used here. Autopoietic approaches focus on self-maintaining boundaries and system identity (Varela et al., 1974), but they do not formally derive the particular kind of regulatory layer and co-occurrence condition proposed in the present paper.

The distinctive claim made here is narrower. It is that the prediction that any two-channel combination should predict senescence entry is not introduced as a fitted summary of existing marker data, but as a structural consequence of the framework itself. Within the range of materials consulted for the present study, we did not find an explicit proposal of this same rule in the form used here. At the same time, as discussed in Section 6.5, the possibility remains open that other frameworks may generate similar predictions. The point is therefore not uniqueness, but difference in derivation and organizational form.

7.3 Extension Possibilities

The structure of Theta_near may not be limited to cellular senescence alone. Whether the same framework can be applied, in a complementary way, to other phenomena such as cancer remains a separate hypothesis worth examining. For example, cancer might tentatively be read not as a case in which regulatory holding becomes jointly blocked, but as a case in which such holding is progressively weakened or released in sequence.

That possibility, however, belongs to a separate line of argument. The present paper does not pursue it in detail. It is mentioned here only to indicate that the proposed organizational rule may have a wider range of possible application than the senescence-entry case developed in the main body of this paper.

7.4 Closing Remark

What determines entry into cellular senescence may not be the magnitude of damage alone, but the structural pattern by which recoverability is being depleted. If that is correct, then the focus of early detection and intervention may shift away from the absolute level of any single marker and toward the question of whether distinct recovery-related paths are closing together at the same time.

8. References