Aging/Research reports/2026-10-18
Comparing wound closure: what counts as one outcome?
Guiding question: Are we measuring average wound closure — or closure of the total wounded area?
Today's decision: The counterevidence on p21-cell removal remains. Its magnitude is not recalculated today: the original linked source-data workbook was not delivered as an Excel file. AG-C06 therefore gains a more precise analysis criterion. Experimental unit, baseline area and aggregation of repeated measurements must be defined before comparing groups.
What was checked today
Gasek et al., Nature Aging, AG-D09, online 13 November 2024, January 2025 issue; retrieved today, 18 October 2026. The original PMC manuscript links “raw data” as an XLSX file. A regular request to the verified target returned HTML rather than spreadsheet contents. Methods describe two wounds per animal and normalization to baseline area. How the wounds were aggregated remained unverified. This is an access limitation, not evidence of missing data or erroneous study statistics.
Bibliographic information and the opposing animal finding are already documented in the 17 October report. No additional study was admitted and no new efficacy result obtained today.
Project mathematical check: two reasonable questions, two estimands
For one animal, let a₁ and a₂ be the positive baseline areas of its two wounds, and r₁ and r₂ their remaining area fractions at the same time. Each fraction refers to its own baseline area. Areas must be non-overlapping, additive components on the same scale; r is dimensionless. These are symbols, not measured values.
| Estimand | Definition and meaning |
|---|---|
| Average closure per wound | Q = (r₁ + r₂)/2; Cequal = 100 × (1 − Q). Both wounds receive equal weight. |
| Closure of total baseline area | S = (a₁r₁ + a₂r₂)/(a₁ + a₂); Carea = 100 × (1 − S). Larger baseline wounds receive more weight. |
| Animal-level comparison | Use one predefined summary per animal, then determine group means and uncertainty using independent experimental units. Multiple wounds and time points do not automatically increase the number of independent animals. |
Algebra gives exactly: S − Q = (a₁ − a₂)(r₁ − r₂) / [2(a₁ + a₂)]. Carea − Cequal is −100 × (S − Q) percentage points. If baseline areas are equal or remaining fractions are equal, the estimands coincide. Otherwise they may differ. Neither definition is universally wrong; they answer different questions. The same nominal wound size does not establish identical measured baseline areas.
Consequence for the research decision: A group comparison must not silently switch definitions. With verified animal mapping, a sensitivity analysis of both estimands could be useful. Normalized fractions alone, without baseline areas, cannot in general reconstruct the area-weighted quantity; their baseline-area ratio would suffice. A verified per-animal summary may suffice for a single prespecified time point; missing raw areas do not preclude every possible analysis. Between-animal weighting also needs definition: averaging per-animal values generally differs from pooling all areas. The experimental unit follows the design; additional dependence and missing values need separate checks. Which fields are actually available remains unknown here. This algebra is a project methods check, not a new biological discovery or a calculated treatment effect.
Countermodel, uncertainty and next executable step
An apparently different benefit could partly reflect weighting, baseline areas or dependent measurements. This is an alternative to assess, not an established explanation for the disagreement. If a direct animal-level comparison is stable across reasonable definitions, this measurement alternative weakens; cell selection and other design differences remain unresolved. A change in significance alone neither refutes nor confirms the effect.
Park the exact download route until an original workbook is actually accessible. Next executable source check: Inspect original legends and methods of the already registered AG-D08, online 28 May 2026 for the connection between surface-area closure and histological restoration: same animals, matched times, defined tissue endpoint? Only then decide which endpoints within this model are comparable. Do not pool with AG-D09. External confirmation would additionally require meaningful tissue function and durability; no experiments were performed here.
Evidence and project status
- Established knowledge: The algebraic relationship holds under the stated definitions. It is not evidence of biological efficacy.
- Experimental findings: No new measurements today. Published animal findings retain their previous limitations.
- Promising indication: More precise cell selection remains a testable approach from earlier literature; no additional confirmation today.
- Hypothesis / project inference: Measurement definitions could influence the comparison. Whether this affects the examined findings is unverified.
⚪ Level 1: Literature approach AG-A08, limited preclinical evidence. 🔵 Level 2: AG-C06 synthesis and AG-T01 target hypothesis unchanged. No Levels 3–5, independent reproduction or established reversal of human aging. The same assessment principle applies to axolotl and brain research: measurement, experimental unit and functional target must align; this does not establish a shared mechanism.
Search and teaching
Targeted original data-link, normalization and regular file-access check; publisher access also limited. No access restriction bypass, spreadsheet contents or numerical effects inspected. Yesterday's limited correction/retraction search was not repeated; no new notice-status finding today. No comprehensive recency search. Source delta 0: 31 DOI-distinct papers, 30 peer reviewed and 1 preprint; 12 approaches, 6 claims.
This methods case extends the existing expert colloquium without a duplicate wish. AG-P21 remains unpublished pending a separate review of the changed native package contract. No course generation, points booking or personal treatment plans.
Continue exploring
