Alpha memo — exercise

Headline: Exercise signals are endpoint-specific, not uniformly generalizable Alpha score: 100/100 (internal triage score; not a certainty claim) Confidence: evidence_backed_signal Memo surface: broad alpha memo Source scope: 16 A/B-bound receipts across 10 evidence sources; 12 total references in the source bundle. Two source-bundle entries are corpus-context references for breadth, not thesis proof; schema labels still use primary/review only.

One-sentence thesis

The exercise corpus in this run supports a broader alpha signal: exercise effects are endpoint- and context-specific rather than uniformly generalizable. The strongest bound receipts span inflammatory markers, exercise capacity and mortality, tumor-model response, angiogenesis biomarkers, muscle-protein synthesis, peripheral oxygen extraction, dietary inflammatory index, and NO/exercise muscle effects.

Why this is surprising

The common story treats exercise as a single beneficial geroscience intervention. This evidence map says the more useful signal is specificity: different populations, endpoints, and mechanisms move differently, so the publishable claim is not “exercise works,” but “exercise signals need endpoint- and context-specific interpretation.”

Evidence receipts

• fact_id=12572 (A_core) — Mixed-muscle protein fractional synthetic rate increased by 42% at 3 h postexercise and 69% at 24 h postexercise in CON. DOI 10.1152/ajpendo.00600.2013
• fact_id=12573 (A_core) — Mixed-muscle protein fractional synthetic rate increased by 42% at 3 h postexercise and 69% at 24 h postexercise in CON. DOI 10.1152/ajpendo.00600.2013
• fact_id=41245 (A_core) — tumor-bearing mice with access to running wheels showed reduced growth of MDA-MB-231 (-66%, P < 0.01) tumors DOI 10.1158/0008-5472.can-16-3125
• fact_id=12544 (A_core) — Mortality risk was 11% lower (hazard ratio, 0.89; 95% confidence interval, 0.86-0.93; P<0.001) for every 1-MET increase in exercise capacity. DOI 10.1161/hypertensionaha.114.03510
• fact_id=22119 (A_core) — exercise and metformin reduced sTNFαR2 (-13.1%; 95% CI, -22.9, -1.0) DOI 10.1158/1940-6207.capr-20-0188
• fact_id=22117 (A_core) — exercise alone reduced hs-CRP and IL6 (-30.9%; 95% CI, -47.3, -9.5) DOI 10.1158/1940-6207.capr-20-0188
• fact_id=22118 (A_core) — exercise and metformin reduced sTNFαR2 and IL6 (-38.7%; 95% CI, -52.3, -18.9) DOI 10.1158/1940-6207.capr-20-0188
• fact_id=22116 (A_core) — Compared with control, exercise alone reduced hs-CRP [-30.2%; 95% CI, -50.3, -1.0] DOI 10.1158/1940-6207.capr-20-0188
• fact_id=30392 (A_core) — IGF-I/IGFBP-3 ratio increased significantly in the diet + exercise group (+5.4%, P < 0.01) compared with control. DOI 10.1158/1055-9965.epi-13-0337
• fact_id=15480 (A_core) — patients with PAD had a greater reduction in SmO2 (-54 ± 10 vs. -12 ± 4%, P = 0.001) DOI 10.1152/japplphysiol.01110.2016
• fact_id=29372 (A_core) — Participants randomized to the diet and diet + exercise arms had statistically significantly greater reductions in PEDF (-9.90%, P < 0.0001) compared with controls. DOI 10.1158/0008-5472.can-16-0399
• fact_id=29371 (A_core) — Participants randomized to the diet and diet + exercise arms had statistically significantly greater reductions in VEGF (-9.98%, P < 0.0001) compared with controls. DOI 10.1158/0008-5472.can-16-0399
• fact_id=29370 (A_core) — Participants randomized to the diet + exercise arms had statistically significantly greater reductions in PAI-1 at 12 months compared with controls (-19.3% vs. +3.48%, respectively, P < 0.0001). DOI 10.1158/0008-5472.can-16-0399
• fact_id=19252 (A_core) — protein synthetic response to exercise was blunted in obese rats: mixed -7%, myofibrillar +6%, cytosolic +7% vs. lean +34%, +40%, +17%. DOI 10.1096/fj.12-224006
• fact_id=27808 (B_context) — Participants randomized to diet and diet+exercise arms had greater reductions in E-DII (-104.4% and -84.4%), versus controls (-34.8%, both P < 0.001). DOI 10.1158/1940-6207.capr-20-0181
• fact_id=34200 (A_core) — NO-donor treatment with isosorbide dinitrate for 6 wk, in combination with voluntary exercise for 3 wk, increased muscle mass by 25% DOI 10.1152/ajpcell.00305.2011

Source bundle map

1. Activation of mTORC1 signaling and protein synthesis in human muscle following blood flow restriction exercise is inhibited by rapamycin (2014). DOI 10.1152/ajpendo.00600.2013 — primary
2. Exercise-Induced Catecholamines Activate the Hippo Tumor Suppressor Pathway to Reduce Risks of Breast Cancer Development (2017). DOI 10.1158/0008-5472.can-16-3125 — primary
3. Exercise Capacity and All-Cause Mortality in Male Veterans With Hypertension Aged ≥70 Years (2014). DOI 10.1161/hypertensionaha.114.03510 — primary
4. Effect of Exercise or Metformin on Biomarkers of Inflammation in Breast and Colorectal Cancer: A Randomized Trial (2020). DOI 10.1158/1940-6207.capr-20-0188 — primary
5. Effects of Dietary Weight Loss and Exercise on Insulin-Like Growth Factor-I and Insulin-Like Growth Factor-Binding Protein-3 in Postmenopausal Women: A Randomized Controlled Trial (2013). DOI 10.1158/1055-9965.epi-13-0337 — primary
6. Blood pressure and calf muscle oxygen extraction during plantar flexion exercise in peripheral artery disease (2017). DOI 10.1152/japplphysiol.01110.2016 — primary
7. Dietary Weight Loss and Exercise Effects on Serum Biomarkers of Angiogenesis in Overweight Postmenopausal Women: A Randomized Controlled Trial (2016). DOI 10.1158/0008-5472.can-16-0399 — primary
8. Abnormal protein turnover and anabolic resistance to exercise in sarcopenic obesity (2013). DOI 10.1096/fj.12-224006 — primary
9. Changes in Dietary Inflammatory Index Patterns with Weight Loss in Women: A Randomized Controlled Trial (2020). DOI 10.1158/1940-6207.capr-20-0181 — primary
10. Nitric oxide and voluntary exercise together promote quadriceps hypertrophy and increase vascular density in female 18-mo-old mice (2012). DOI 10.1152/ajpcell.00305.2011 — primary
11. Exercise training undertaken by people within 12 months of lung resection for non-small cell lung cancer (2019). DOI 10.1002/14651858.cd009955.pub3 — primary
12. Aerobic Exercise Attenuates Frailty in Aging Male and Female C57Bl/6 Mice and Effects Systemic Cytokines Differentially by Sex (2021). DOI 10.1093/gerona/glab297 — primary

What this changes

Do not treat exercise as one generic anti-aging lever in review or trial design. Split claims by endpoint family before making a broad conclusion: inflammation, tumor biology, vascular oxygenation, muscle synthesis, mortality/exercise capacity, and diet-linked biomarker changes are different evidence lanes.

Limitations

• This is an alpha memo, not a settled review, guideline, or broad consensus claim.
• Only 10 references in the 12-source bundle have A/B-bound receipts in this run; two source-bundle entries are corpus-context references for breadth and should not be used as direct proof.
• The topic is heterogeneous, so the memo should guide subtopic extraction rather than support one pooled treatment-effect claim.
• Animal, mechanistic, and human endpoints are deliberately separated; cross-lane analogies remain hypotheses.

Strongest counter-evidence

• No A_core/B_context counter-evidence was classified in this run, but the heterogeneity of endpoints is itself a constraint on broad claims.
• If independent receipt expansion shows the same endpoint families behave uniformly across populations and models, this context-specific framing weakens.

Next extraction

• Split exercise into endpoint-specific child topics: inflammatory-axis response, cancer/tumor biology, vascular oxygenation, muscle synthesis/anabolic resistance, fracture/bone outcomes, and mortality/exercise capacity.
• For each child topic, require at least two independent A/B-bound source papers before publishing a broad claim.
• Re-run counter-evidence search against each child topic so “no counter-evidence found” is not confused with proof of absence.

Provenance / priority

• Topic: exercise
• Author: Dom Lynch
• Version: 1.0
• License: CC BY-NC 4.0
• Suggested citation: Dom Lynch. (2026). Exercise signals are endpoint-specific, not uniformly generalizable. ReseaRka Evidence Index. Version 1.0.
• Run: exercise-evidence-2026-05-16T18-12-38Z