Founder's role line (READ): "Co-inhibitory receptor pathways that restrain T-cell activation to maintain self-tolerance and prevent excessive immune responses." Key components (READ): PDCD1 (PD-1), CD274 (PD-L1), CTLA4, LAG3, TIGIT
STATED FIRST, BEFORE ANY EVIDENCE. This node is about self-tolerance and the restraint of T-cell activation, which is what the founder's own role line says. Nothing in this node is a cancer treatment claim, a statement that fasting treats any disease, or advice to anyone to fast. The one human trial that measured checkpoint receptors under a fasting-type diet reported no tumor response, progression or survival endpoint of any kind, and said so itself. The one animal study that combined restriction with a checkpoint drug reported the drug losing its effect. Those two facts are the top of this node, not a footnote at the bottom.
The two things kept rigorously separate in this node:
- (a) Human lymphocyte and immune-cell response to fasting. Real, measurable, and in the one controlled-follow-up study located, null for T-cell subsets.
- (b) Checkpoint-receptor signaling under fasting. One human single-arm trial measured checkpoint-receptor expression on cells. No human study located measured checkpoint signaling, and no human study located tested a checkpoint intervention under fasting. Everything about checkpoint pharmacology under restriction is mouse.
DRY status
DRY: none located. Queries run:
"dry fasting" DNA methylation epigenetic OR immune OR VEGF OR Myc OR EZH2 human study(English)Ramadan fasting lymphocyte count CD4 CD8 natural killer cells human immune changes(English)prolonged fasting human T cell PD-1 expression lymphocyte subsets flow cytometry fasting mimicking diet humans(English)сухое голодание метилирование ДНК иммунитет лимфоциты исследование Филонов(Russian)
Two human dawn-to-dusk dry-fasting studies were located and opened:
- Mindikoglu AL, Park J, Opekun AR, Abdulsada MM, Wilhelm ZR, Jalal PK, Devaraj S, Jung SY. Metabolism Open. 2022 Nov 1;16:100214. doi:10.1016/j.metop.2022.100214. PMID 36506940. PMCID PMC9731888. 14 adults with metabolic syndrome, dawn-to-dusk for more than 14 h daily for 29 days, PBMC proteomics, no non-fasting control group. READ, authors' own words for the protocol: "strict dry fasting without any liquid or food intake" and "without eating or drinking (nothing by mouth)." MEASURED: PD-1, PD-L1, CTLA-4, LAG3 and TIGIT were not measured. The study measured LSP1 up (log2 4.25, P=0.004), an immune-cell cytoskeletal protein, and nothing on this node's components. MEASURED: no FDR or multiple-testing correction is reported for the differential-expression comparisons; the 1 percent FDR in the Methods applies to peptide identification. With n = 14, no control arm and many comparisons, the P values are nominal.
- Mindikoglu AL, Eckel-Mahan K, Opekun AR, Alzubaidi MM, Crochet ZR, Jalal PK, Jung SY. Clin Proteomics. 2025 Jun 24;22:25. doi:10.1186/s12014-025-09547-3. PMID 40551123. PMCID PMC12186377. Pilot, author-described proof-of-concept. 4 subjects with metabolic syndrome and MASLD plus 4 healthy men, 4 weeks of dawn-to-dusk fasting over 14 h daily, serum proteomics on 3 versus 3 after outlier removal. MEASURED: PD-1, PD-L1, CTLA-4, LAG3 and TIGIT were not measured. B2M fell (log2 -0.977, P=0.033), CD93 fell (log2 -2.567, P=0.041), CD109 fell (log2 -1.321, P=0.039). None is on this node's component list. MEASURED: no FDR correction reported; several log2 decreases of 9 to 12 in healthy subjects on n = 3 are implausibly large and the extraction itself flags them as needing caution.
So: the dry-fasting human channel is productive as a literature and empty for this node, and by the brief's own definitions dawn-to-dusk with no food and no water is RAMADAN, not absolute DRY. Absolute DRY and checkpoint signaling: nothing located.
Disconfirming and null first
1. Ramadan fasting did NOT change T-cell or B-cell subpopulations in humans, over two years of follow-up. Siadat Z, Rastin M, Tabasi N, Rezaee A, Mahmoudi M. "Evaluation of the Effects of Ramadan Fasting on Lymphocyte Subpopulations in a Two-year Follow-up." J Fasting Health. 2014;2(1):31-36. 38 healthy Muslims (9 female, 29 male), aged 17 to 51, Ramadan in 1999 and 2000 (winter), roughly 12 to 12.5 h/day, at least 22 of 30 days observed, bloods two days before and one day before the end. MEASURED, all non-significant:
- CD3+ T cells 70.12 to 70.26 percent; absolute 1.61 to 1.5 x10^3/mm3. NS.
- CD4+ 53.46 to 52.8 percent; absolute 0.087 to 0.081 x10^3/mm3. NS.
- CD8+ 37.7 to 37.8 percent; absolute 0.6 to 0.54 x10^3/mm3. NS.
- CD19+ B cells 14.56 to 14.74 percent; absolute 0.35 to 0.3 x10^3/mm3. NS. READ, authors' own words: Ramadan fasting "does not affect these cells."
- HLA-DR+CD3+ activated T cells 11.14 to 10.54 percent; absolute 0.14 to 0.11 x10^3/mm3. NS.
- CD4/CD8 ratio 1.48 to 1.5. NS.
- Total lymphocytes 25.82 to 26.23 percent; absolute 2.32 to 2.19 x10^3/mm3, P = 0.06. MEASURED internal contradiction in the source, reported and not smoothed: the abstract calls the total lymphocyte difference significant while the table gives P = 0.06, which does not meet 0.05. Mean age is 35.4 in the abstract and 32.5 in the results. Table 3 reports the CD4 values differently from the abstract. The paper does not state clearly whether the same individuals were sampled in both years. MEASURED: NK cells (CD16/CD56) were not measured. PD-1, CTLA-4, LAG3 and TIGIT were not measured. This is the control on the whole node. The claim "fasting redistributes human lymphocytes" does not survive a 12-hour daily fast measured over two Ramadans.
2. Prolonged water-only fasting INCREASED systemic inflammation in humans. Commissati S, Lastra Cagigas M, Masedunskas A, Petrucci G, Tosti V, De Ciutiis I, Rajakumar G, Kirmess KM, Meyer MR, Goldhamer A, Kennedy BK, Hatem D, Rocca B, Fiorito G, Fontana L. "Prolonged fasting promotes systemic inflammation and platelet activation in humans: A medically supervised, water-only fasting and refeeding study." Mol Metab. 2025 Apr 21;96:102152. doi:10.1016/j.molmet.2025.102152. PMID 40268190. PMCID PMC12088818. 20 middle-aged volunteers (11 women, 9 men; mean age 52.2; mean BMI 28.8), single-arm, no control, medically supervised water-only fast averaging 9.8 +/- 3.1 days, then about 5.3 days of guided plant-based refeeding. Protocol deviations MEASURED: six participants switched to broth or juice on medical advice, and a few drank small amounts during the fast. MEASURED: hsCRP rose about 129 percent during fasting, Wilcoxon P = 0.0004, returning to baseline after refeeding in all but one participant. Hepcidin, ferritin, midkine, MMP-9, IL-8 and PAFAH (PLA2G7) all increased significantly on SOMAScan. Urinary 11-dehydro-TXB2 rose 21 percent during fasting and 36 percent after refeeding. vWF and soluble GP1Balpha mildly increased. Platelet counts did not change. Prothrombin did not change. MEASURED unit inconsistency in the source, reported: Table 1 gives CRP 1.7 to 3.9 mg/dL while the abstract gives 2.8 to 4.3 mg/L for the validation cohort. MEASURED: in a retrospective validation cohort of 1,422 people on modified fasting (Wilhelmi de Toledo et al., 2019), 66.6 percent had a significant CRP rise. The authors describe that cohort as retrospective. MEASURED: this study did not report lymphocyte counts or immune-cell subsets. Its immune findings are plasma proteins and inflammatory markers. This cuts directly against any reading of fasting as immune-quieting. The direction is the opposite of what a reader expects, it is human, and the n is 20 with no control arm. 3. In mice, caloric restriction made a PD-1 drug stop working. Dung NT, Susukida T, Ucche S, He K, Sasaki S, Hayashi R, Hayakawa Y. "Calorie Restriction Impairs Anti-Tumor Immune Responses in an Immunogenic Preclinical Cancer Model." Nutrients. 2023;15(16):3638. doi:10.3390/nu15163638. 40 percent CR from the day of tumor inoculation, C57BL/6J mice, subcutaneous B16-OVA melanoma, endpoints through day 16. MEASURED: CR significantly slowed tumor growth on its own. MEASURED: CR did not change CD4+ or CD8+ tumor infiltration, the CD8/CD4 ratio, or CD44-high effector CD8+ cells. MEASURED: CD8 depletion increased tumor growth in control-diet mice but made no difference in CR mice, which the authors read as CR blunting CD8-dependent control; the authors' own results section notes those depletion comparisons were not statistically significant. MEASURED: CD8+ cells from tumor-draining nodes of CR mice had lower oxygen consumption and lower extracellular acidification; anti-PD-1 shifted those cells toward glycolysis in control mice but not in CR mice. READ, authors' own words: "Anti-PD-1 Ab treatment lost its efficacy to suppress tumor growth." MEASURED: anti-PD-1 still increased tumor-infiltrating CD8+ T cells in CR mice. MEASURED limits: authors call it exploratory, n = 3 to 21 per group with large variation, mouse only; PD-L1 and CTLA-4 were not examined. The extraction notes the abstract's patient-level conclusions are extrapolations. This is the single most important row for keeping this node gotcha-proof. The restriction-plus-checkpoint literature contains a published failure, in the same direction of intervention, and it is listed before any positive. 4. The human positive has no efficacy endpoint, no randomization and no control arm. See 41.1 below. READ, authors' own words: "Phase II/III clinical trials are needed to investigate FMD antitumor activity/efficacy." MEASURED: the primary endpoint was safety. No tumor response, progression or survival endpoint was reported. 5. Checkpoint-pathway direction under restriction is contested in animals. Located and NOT opened, recorded as retrieval debts and named so both sides are visible: a Molecular Cell 2021 report that energy status dictates PD-L1 protein abundance; a Nature Communications 2023 report on FMD in mice and immunotherapy side effects; a Nutrients-indexed and a Nature-indexed report that dietary restriction reprograms CD8+ T cell fate to enhance immunotherapy response; a biorxiv 2026 report that calorie restriction upregulates islet PD-L1 in NOD mice; and the Dung 2023 impairment above. The animal literature points in both directions. No winner is picked here.
Graded rows
| # | Claim | Modality | Species / system | Tier | Grade | Source | Tag |
|---|---|---|---|---|---|---|---|
| 41.1 | A 5-day FMD repeated every 21 to 28 days (median 4 cycles, up to 8) changed checkpoint-receptor expression on human blood cells: PD-L1+ monocytes down, activated PD-1+CD69+ CD8+ T cells up, PD-1+LAG3+TIM3+ exhausted and hyperexhausted cells down, PD-1+ CD8+ cells lacking LAG3 and TIM3 up | FMD | human, single-arm, 101 enrolled, 38 immune-profiled, plus 22 in a window-of-opportunity trial | H3 | C | Vernieri C, Fucà G, Ligorio F, et al. Cancer Discov. 2022;12(1):90-107. doi:10.1158/2159-8290.CD-21-0030. PMID 34789537. PMCID PMC9762338 | MEASURED, direction; no effect sizes given in the text opened |
| 41.2 | In the same trial, total CD14+ monocytes, HLA-DR-negative monocytic MDSCs, CD15+ granulocytes, CD3+CD25+ cells and Tregs all fell, and CD3-CD16+CD56dim NK cells, granzyme B and IFN-gamma in CD8+ T cells rose | FMD | same | H3 | C | Vernieri 2022 | MEASURED, direction |
| 41.3 | In the window-of-opportunity sub-analysis, intratumoral CTLA4 and LGALS9 transcripts rose, and they were the only checkpoint-related genes that did | FMD | human tumor tissue, 22 patients, mostly biopsy versus surgical specimen, unplanned interim analysis, no control arm | H3 | C | Vernieri 2022 | MEASURED |
| 41.4 | That trial reported no efficacy endpoint. Primary endpoint was safety: grade 3/4 FMD-related adverse events in 12.9 percent (90 percent CI 7.8 to 19.7), below the prespecified 20 percent threshold | FMD | human | H3 | C | Vernieri 2022 | MEASURED |
| 41.5 | TIGIT and absolute lymphocyte counts were not reported in the text opened, so two of the founder's five components are unmeasured even in the one human checkpoint study | FMD | human | n/a | n/a | Vernieri 2022 | MEASURED as an absence of reporting |
| 41.6 | Ramadan fasting produced no significant change in CD3+, CD4+, CD8+, CD19+, HLA-DR+CD3+ or the CD4/CD8 ratio in 38 healthy adults across two Ramadans | RAMADAN | human, pre/post, two-year follow-up | H3 | C | Siadat 2014, J Fasting Health 2(1):31-36 | MEASURED, null |
| 41.7 | A 9.8-day water-only fast raised hsCRP about 129 percent (P=0.0004) and raised hepcidin, ferritin, midkine, MMP-9, IL-8 and PAFAH; platelet counts unchanged | PF | human single-arm, n=20, no control | H3 | C | Commissati 2025, Mol Metab 96:102152 | MEASURED |
| 41.8 | In a retrospective cohort of 1,422 people on modified fasting, 66.6 percent had a significant CRP rise | PF / FMD-adjacent, "modified fasting" | human, retrospective | H2 observational | C | Wilhelmi de Toledo et al. 2019, READ through Commissati 2025, NOT opened | READ, secondhand; retrieval debt |
| 41.9 | 40 percent CR abolished the tumor-growth effect of anti-PD-1 in mice while not changing CD4+ or CD8+ tumor infiltration | CR | mouse, B16-OVA melanoma | A1 | C | Dung 2023, Nutrients 15(16):3638 | MEASURED |
| 41.10 | Energy status dictates PD-L1 protein abundance | C1 or A1, not determined | not opened | C1 | C | Molecular Cell 2021, located, NOT opened | located only; retrieval debt |
| 41.11 | Calorie restriction upregulates islet PD-L1 and lowers autoimmune diabetes onset in NOD mice | CR | mouse | A1 | C | biorxiv 2026.02.15.705935, located, NOT opened, preprint | located only; retrieval debt |
| 41.12 | Fasting lowers insulin and IGF-1 and raises ketones, and those change T-cell metabolism, therefore checkpoint signaling changes. Hops: fasting to those fuels (other slices), fuels to T-cell metabolism (measured in mice by Dung 2023), T-cell metabolism to checkpoint signaling (not measured in any human source located). Inference, not a human fasting effect | n/a | hops shown | M | C | derivation stated | inference, shown and labeled |
Conflict left unresolved, named on both sides
- Does fasting change human T-cell populations? Side A: no, Ramadan fasting changed nothing measurable across two years (Siadat 2014, H3, explicit null on six readouts). Side B: yes, FMD shifted monocyte, MDSC, Treg, NK and PD-1+ CD8 compartments (Vernieri 2022, H3, single-arm, patients on concurrent chemotherapy). Unresolved. The modalities differ (12 h daily versus 5-day FMD), the populations differ (healthy adults versus patients with cancer on chemotherapy), and neither is randomized against the other. Both stay on the record.
- Does fasting quiet or provoke human inflammation? Side A: provokes. hsCRP up 129 percent in a 9.8-day water-only fast, and 66.6 percent of 1,422 people in a retrospective modified-fasting cohort had a CRP rise (Commissati 2025, H3). Side B: FMD lowered MDSCs and Tregs and raised effector markers (Vernieri 2022, H3). Unresolved. These are different markers and different modalities and they are not reconciled by anything located here.
- Does restriction help or hinder checkpoint pathways in animals? Side A: hinders. 40 percent CR abolished anti-PD-1 efficacy (Dung 2023, A1, opened). Side B: helps. Several located-but-unopened reports claim dietary restriction enhances CD8+ fate and immunotherapy response, and that CR raises islet PD-L1. Unresolved, and side B is unopened in this sweep. Named, not adjudicated.
- Whether PD-1 expression going up is good or bad. Vernieri 2022 reports PD-1+CD69+ activated CD8+ cells up and PD-1+LAG3+TIM3+ exhausted cells down, in the same samples. PD-1 is both an activation marker and an exhaustion marker. The source itself does not collapse this, and neither does this file.
Open questions
- TIGIT under any fasting modality in humans: nothing located.
- CTLA-4 in human blood under any fasting modality: nothing located. The one CTLA4 human row (41.3) is intratumoral transcript in an unplanned interim analysis of 22 patients.
- No randomized human trial of any fasting modality with a checkpoint molecule as a prespecified endpoint was located.
- Whether the Vernieri immune shifts are attributable to the FMD or to the concurrent chemotherapy is unresolved by the source itself. READ, authors' own position: adverse-event attribution is confounded by concurrent chemotherapy, the mechanistic explanations are largely speculative and draw on mouse studies, and future trials must determine which dietary components or metabolic changes drive the effects.
# CROSS-NODE NOTES FOR THIS SLICE
What the founder's component lists reveal
Of the 28 named key components across these 8 nodes, the number with any human fasting-intervention measurement located in this sweep is two: VEGFA (node 40, under CR only) and PDCD1/PD-1 plus CD274/PD-L1 plus CTLA4 plus LAG3 (node 41, under FMD only, single-arm). Counting generously that is 5 of 28. MEASURED as an absence of measurement. Absence is not evidence of no effect.
Modality distribution across all rows in this file
CR is the most-represented modality and CR is not fasting. Every human row in nodes 37 and 40 is CR. The only human rows under a true fasting modality in the whole slice are: PF in node 37 (Luo 2026, preprint, null at day 10), PF in node 41 (Commissati 2025, inflammation up), FMD in node 41 (Vernieri 2022, single-arm), and RAMADAN in node 41 (Siadat 2014, null). That is four human studies under true fasting modalities across eight nodes, three of which are null or point against the expected direction. Those four studies supply nine graded rows; the per-row count is in the receipt block.
Why the DRY column is empty and why that is the protective finding
Human dry-fasting research exists and was opened (two Mindikoglu studies, 14 and 8 subjects, both uncorrected for multiple testing, one with no control group and one a 3-versus-3 pilot). It is dawn-to-dusk, which the brief defines as RAMADAN. It measured none of the 28 components across these eight nodes. Writing any of these eight nodes as a dry-fasting mechanism would require an identifier that does not exist in the retrieved literature. The empty column is the correct column.
# RECEIPT BLOCK
Nodes
- Nodes done: 8 / 8 (34, 35, 36, 37, 38, 39, 40, 41)
- Nodes with ZERO human fasting evidence of any modality: 5 of 8 = 34 (Myc), 35 (PRC2/EZH2), 36 (SWI/SNF), 38 (EMT), 39 (Integrin/FAK)
- Nodes with human evidence: 3 of 8 = 37 (CR H1, PF H3), 40 (CR H1, CR H3), 41 (FMD H3, RAMADAN H3, PF H3)
- Nodes whose best row is C1 or M, and which therefore must not read as a human fasting effect: 36 and 39 (best row A1 for an adjacent readout, M for the node's own identity). Nodes 34, 35 and 38 have a genuine A1 for the node itself, animal only.
- Nodes with absolute DRY evidence: 0 of 8
Rows by tier
Counts computed off the graded-row tables in this file by enumerating every row id, not estimated. Total distinct graded rows: 53.
| Tier | Rows | Row ids |
|---|---|---|
| H1 | 7 | 37.1, 37.2, 37.3, 37.4, 40.1, 40.2, 40.3 |
| H2 | 3 | 40.5 (observational), 41.8 (observational), 39.5 (human crossover, but the intervention was carbohydrate, so it is not this node under fasting) |
| H3 | 10 | 37.5, 37.6, 37.7, 40.4, 41.1, 41.2, 41.3, 41.4, 41.6, 41.7 |
| A1 | 22 | 34.1, 34.2, 35.1, 35.2, 35.3, 35.4, 35.5, 36.1, 36.2, 36.3, 37.8, 37.9, 38.1, 38.2, 38.3, 38.4, 39.1, 39.2, 39.3, 40.7, 41.9, 41.11 |
| A2 | 0 | none located in this slice |
| C1 | 2 | 40.6, 41.10 |
| M | 7 | 34.3, 35.6, 36.4, 38.5, 39.4, 40.8, 41.12 |
| absence-of-measurement rows, no tier applicable | 2 | 37.10, 41.5 |
Human rows (H1 + H2 + H3) = 20 of 53, and 19 of those 20 sit in only three nodes: 37, 40 and 41. Human rows under a true fasting modality (PF, FMD, RAMADAN, TRE or IF; excluding CR, KD, observational rows and the one non-fasting row) = 9: 37.5, 37.6, 37.7 (PF), 41.1, 41.2, 41.3, 41.4 (FMD), 41.6 (RAMADAN), 41.7 (PF). Of those nine, four are explicit nulls or point against the expected direction (37.5, 37.6, 41.6, 41.7), and the remaining five are all from one single-arm unrandomized trial with no efficacy endpoint. Stated exactly: of the 20 human rows, 7 are CR (37.1 to 37.4, 40.1 to 40.3), 1 is CR single-arm (40.4), 3 are PF (37.5 to 37.7), 1 is PF (41.7), 4 are FMD (41.1 to 41.4), 1 is RAMADAN (41.6), 2 are observational non-intervention (40.5, 41.8), and 1 is not a fasting intervention at all (39.5).
Grades by node
- 34: best C (A1, single species, confounded by hypothermia, no second stream)
- 35: best C (A1, mouse liver, one study, writer unchanged)
- 36: best C (A1 for accessibility, M for the node's identity)
- 37: best B+ (H1, one randomized trial, one clock of five, EWAS null)
- 38: best C (A1, mouse tumor, direction against, intended dose not delivered)
- 39: best C (A1 for the ligand side, M for the receptor and kinase side)
- 40: best B+ (H1, one randomized trial, effect confined to >=5 percent weight loss, CR not fasting)
- 41: best C (H3 only; no randomized human trial; animal literature conflicting)
No node in this slice reaches grade A. No node has multiple independent human trials converging.
Language channels
| Language | Query run | Result |
|---|---|---|
| English | many, including CALERIE caloric restriction trial epigenetic clock DunedinPACE null Horvath PhenoAge GrimAge results, "dry fasting" DNA methylation epigenetic OR immune OR VEGF OR Myc OR EZH2 human study, fasting EZH2 H3K27me3 nutrient deprivation starvation Polycomb PRC2 mouse liver, fasting SWI/SNF BRG1 SMARCA4 ARID1A chromatin remodeling nutrient starvation liver, fasting epithelial mesenchymal transition EMT Snail ZEB1 E-cadherin nutrient starvation mouse in vivo, "caloric restriction" OR fasting integrin FAK phosphorylation reduced adipose tissue extracellular matrix in vivo, fasting VEGF human serum plasma time-restricted eating randomized trial angiogenesis biomarker, Ramadan intermittent fasting serum VEGF human levels study, fasting PD-1 CTLA-4 immune checkpoint expression T cells mouse study caloric restriction, Ramadan fasting lymphocyte count CD4 CD8 natural killer cells human immune changes, prolonged fasting human T cell PD-1 expression lymphocyte subsets flow cytometry fasting mimicking diet humans, fasting MYC expression liver mouse starvation c-Myc downregulation nutrient deprivation, intermittent fasting epigenetic age DNA methylation human randomized trial Horvath clock, time-restricted eating DNA methylation human trial epigenome-wide results methylome, Ramadan fasting DNA methylation epigenetic age human study | PRODUCTIVE. Every source in this file came through this channel. |
| Russian | сухое голодание метилирование ДНК иммунитет лимфоциты исследование Филонов | RUN, EMPTY for these eight nodes. Returned Filonov's dry-fasting books (books.google.com, vesbook.ru, filonov.net, e-reading.club, ruslit.net) and unrelated indexed papers. No primary study measuring any of the 28 components was located. The Filonov tradition is present as books and practitioner essays, not as indexed primary measurement, and nothing in it was opened or cited here. |
| Turkish | not run as a Turkish-language query. A Turkish journal source was reached through the English channel: Işık T, IGUSABDER 2025;26:722-737, a Turkish-language review with a Turkish title, opened and cited at node 37. | PARTIAL. One Turkish-journal source opened; no Turkish-language query was issued. |
| Chinese | not run | BLOCKED. See channel block below. Intended query: 禁食 DNA甲基化 表观遗传时钟 人体 研究 plus 禁食 EZH2 H3K27me3 小鼠 肝脏 |
| Japanese | issued and refused mid-run: 短期断食 DNAメチル化 ヒト 研究 エピジェネティック 時間制限摂食 | BLOCKED. The query was submitted and the search tool returned a budget refusal, not a result set. Not an empty channel. |
| German | not run | BLOCKED. Intended query: Heilfasten DNA-Methylierung epigenetische Uhr Mensch Studie Buchinger |
| French | not run | BLOCKED. Intended query: jeûne prolongé méthylation ADN horloge épigénétique humain étude |
| Arabic | not run | BLOCKED. Intended query: الصيام مثيلة الحمض النووي الخلايا اللمفاوية دراسة بشرية |
| Persian | not run | BLOCKED. Intended query: روزه داری متیلاسیون DNA لنفوسیت مطالعه انسانی |
| Spanish | not run | BLOCKED. Intended query: ayuno prolongado metilación del ADN reloj epigenético humanos ensayo |
| Korean | not run | BLOCKED. Intended query: 단식 DNA 메틸화 후성유전 시계 인간 연구 |
Languages done / total: 2 of 11 run (English productive, Russian empty); 1 of 11 partially served through a foreign-language journal reached in English (Turkish); 8 of 11 blocked by the shared WebSearch budget, not by absence of literature. This is a channel block and a debt, not a null. A worldwide all-language sweep for these eight nodes has not been completed and this file does not claim one.
Channels blocked, with reason, never retried after the block
| Channel | Reason | Action taken |
|---|---|---|
api.crossref.org | Agent proxy refused, connect_rejected, HTTP 403. Host not allowed for this session. | Abandoned. Citations sourced from publisher pages instead. |
www.nature.com | client_error on fetch, nothing retrieved. | Routed to PMC and link.springer.com. |
pubmed.ncbi.nlm.nih.gov | Returned a shell page with no record content. | Routed to publisher hosts. |
pmc.ncbi.nlm.nih.gov | Google reCAPTCHA "Checking your browser" interstitial on repeated articles (PMC10458233, PMC5033683, PMC9799188, PMC5712070, PMC8091980). Productive earlier in the run (PMC10148951, PMC12088818, PMC7010817, PMC5741368, PMC9731888, PMC12186377, PMC9762338, PMC7602406, PMC5033683/table/T4) then captcha-walled. | Stopped using after the captcha pattern established. Routed to academic.oup.com, mdpi.com, journals.plos.org, link.springer.com, biorxiv, medrxiv, dergipark.org.tr, intechopen.com, sochob.cl, applications.emro.who.int, eurekalert.org. |
aacrjournals.org (and the article-pdf path) | client_error on fetch, nothing retrieved. | Volume, issue and first page read from the publisher's article URL path only. Authors taken from the AACR press release via eurekalert.org. Full citation remains a debt. |
www.cell.com | client_error on fetch, nothing retrieved. | Routed to a publisher-identical PDF on sochob.cl for the Harney 2021 paper. The iScience 2025 paper could not be routed. |
onlinelibrary.wiley.com | client_error on fetch, nothing retrieved. | Zhao et al. Aging Cell left as a debt. |
journals.plos.org | One RATE_LIMIT_EXCEEDED from the target, cleared on a single re-fetch as the error text instructed. | Fetched successfully on the second, single attempt. |
WebSearch tool | Turn budget exhausted at 200 calls, shared across every agent in the turn. | Switched to WebFetch-only for the remainder of the run. No workaround attempted through curl, a browser, a reader, a proxy or an archive service. |
Retrieval debts: located and NOT opened
1. Cancer Research 2016;76(14):4226, "Dietary Weight Loss and Exercise Effects on Serum Biomarkers of Angiogenesis in Overweight Postmenopausal Women: A Randomized Controlled Trial", PMCID PMC5033683. Table T4 was opened; the article body was not. Full author list, DOI and PMID not retrieved. Authors Duggan and senior author McTiernan taken from the AACR press release, not from the paper. This is the H1 source for node 40. 2. Waziry et al., medRxiv 2021.09.21.21263912, Table 3. Holds the confidence intervals for the Horvath, Hannum, PhenoAge, GrimAge and DunedinPACE estimates. Not retrievable from the full-text page. This is why the preprint-versus-published DunedinPACE sign conflict at node 37 item 7 is unresolved. 3. Waziry 2023 Supplementary Information, Supplementary Figure 3a and 3b, Supplementary Data 1 and 2. Hold the per-clock Horvath and Hannum results for the published version. Not opened. 4. Zhao et al., Aging Cell, doi:10.1111/acel.12947, SLC1A5 glutamine transporter as a MYC target in long-lived Myc hypomorphic mice. Wiley refused. Node 34 row 34.3 rests on title and indexing only. 5. Hamad Bin Khalifa University thesis, epigenome-wide DNA methylation analysis of Ramadan fasting effects on pediatric MASLD. Located at elmi.hbku.edu.qa. Not opened. Relevant to node 37 and to the RAMADAN modality. 6. Nutrients 2024;16(19):3260, EZH2-mediated H3K27 trimethylation in mouse liver as an early event induced by high-fat diet. Located at mdpi.com. Not opened. Relevant to node 35 as a feeding comparator. 7. Dalmas et al. 2011, surgery-induced weight loss and VEGF in humans. Located through Korybalska 2017. Not opened. Node 40 row 3 of the disconfirming section. 8. Halberg et al. 2009, adipocyte HIF-1alpha overexpression with no VEGF induction. Located through Korybalska 2017. Not opened. Node 40 row 4 of the disconfirming section. 9. Lolmede et al. 2003, hypoxia and VEGF in 3T3-F442A adipocytes. Located through Korybalska 2017. Not opened. Node 40 row 40.6. 10. Csiszar et al., 10 years of CR in rhesus macaques and VEGF. Located through Korybalska 2017. Not opened, and the direction is ambiguous in the secondary source. Node 40 row 40.7. 11. Miyazawa-Hoshimoto et al., serum VEGF and anthropometry in obese humans. Located through Korybalska 2017. Not opened. Node 40 row 40.5. 12. Silha et al. 2005, angiogenic factors in overweight and obese humans, VEGF direction not specified in the secondary source. Located through Korybalska 2017. Not opened. 13. Wilhelmi de Toledo et al. 2019, 1,422-person modified-fasting cohort with the 66.6 percent CRP rise. Located through Commissati 2025. Not opened. Node 41 row 41.8. 14. Molecular Cell 2021, "Energy status dictates PD-L1 protein abundance and anti-tumor immunity to enable checkpoint blockade." cell.com refused. Node 41 row 41.10. 15. Nature Communications 2023, FMD in mice, cancer growth and immunotherapy-associated side effects. nature.com refused. Node 41 item 5. 16. "Dietary restriction reprograms CD8+ T cell fate to enhance anti-tumour immunity and immunotherapy responses." Located at PMC12727518 and experts.umn.edu. PMC captcha-walled. This is a named side of the unresolved animal conflict at node 41 and it is unopened. 17. biorxiv 2026.02.15.705935, calorie restriction upregulates islet PD-L1 in NOD mice. Preprint. Located, not opened. Node 41 row 41.11. 18. "Caloric Restriction Impairs Regulatory T cells Within the Tumor Microenvironment After Radiation and Primes Effector T cells." Located at sciencedirect.com. Not opened. Node 41. 19. PMC5712070, "Ramadan Fasting Exerts Immunomodulatory Effects: Insights from a Systematic Review." PMC captcha-walled. This is the systematic review for the RAMADAN immune channel and it is unopened. Node 41's RAMADAN evidence therefore rests on one primary study (Siadat 2014) and not on the review. 20. iScience 2025, S2589-0042(25)01133-2, "Altered immunometabolic response to fasting in humans living with obesity." cell.com refused. Human fasting immunometabolism. Node 41. 21. Nutrients 18(15):2514, doi:10.3390/nu18152514, "Exploring the Immunomodulatory Effects of Fasting-Mimicking Diet in Healthy Individuals: A Pilot Study." Located, not opened. Node 41. 22. Işık 2025's primary sources for rows 37.8 and 37.9 (the mouse offspring DNMT study and the 16:8 vascular-dementia methylation study). Read secondhand through the review only. Not opened. 23. PMC6953068 and PMC8091980, hepatic and multi-tissue fasting and refeeding transcriptional response in mice, for a MYC check at node 34. PMC captcha-walled. 24. Vernieri 2022 beyond character 100,000, including TIGIT and absolute lymphocyte counts. Not read. Node 41 row 41.5. 25. Korenfeld 2025 beyond character 100,000, including the Discussion, where a remodeler could still be named. Not read. Node 36. 26. Byun 2020 beyond character 100,000, where EED and SUZ12 could still be mentioned. Not read. Node 35.
Compliance statement
- Every modality named with one of the nine brief labels. One deviation declared in full at node 38 (a once-weekly 24-hour fast, for which the brief has no label; recorded as IF-ADF family with the deviation stated in the row).
- DRY status stated explicitly for all 8 nodes with the queries run. All 8 are "none located."
- Every row tagged with a system tier. The two nodes whose own identity is supported only at M (36, 39) are written as mechanism and say so in the row text.
- Disconfirming evidence, nulls and failed replications placed first in all 8 nodes.
- Conflicts reported unresolved with both sides named, at nodes 35, 37, 39, 40 and 41. No winner picked.
- No PMID, DOI, author, year, journal or number appears here that was not retrieved from an opened source or, in the single declared case of Cancer Research 2016;76(14):4226, read from the publisher's own article URL path with that provenance stated in the row.
- No cancer treatment claim, no cure language, no advice to any reader to fast. No em dashes. American spelling.
← 40. Angiogenesis (VEGF axis) All 41 nodes
dryfasting.com has been published since 2014. This evidence record was compiled on 22 November 2026. Every row names the fasting modality the study actually ran, the system it was measured in and the grade. Where no source carries a DOI or PMID the row says so rather than supplying one. Nothing is sold on this site.