Founder's role line (READ): "Drives the growth of new blood vessels during development, exercise, and tissue repair." Key components (READ): VEGFA, VEGFR1-3 (FLT1, KDR, FLT4), FGF2, ANGPT1/2-TIE2, THBS1
The founder's own role line names exercise. Exercise and fasting are kept separate below, as instructed.
DRY status
DRY: none located. Queries run:
"dry fasting" DNA methylation epigenetic OR immune OR VEGF OR Myc OR EZH2 human study(English)Ramadan intermittent fasting serum VEGF human levels study(English, RAMADAN-targeted)сухое голодание метилирование ДНК иммунитет лимфоциты исследование Филонов(Russian)
MEASURED: neither opened human dawn-to-dusk dry-fasting study measured VEGF. Mindikoglu 2025 reports CD248 up by log2 8.124, P = 0.001 in the metabolic-syndrome group and discusses CD248 in relation to physiological angiogenesis in skeletal muscle, and LYVE1 up (log2 1.054, P = 0.029). MEASURED: angiogenesis itself was not measured, VEGF was not measured, n = 3 versus 3 after outlier removal, and no FDR correction is reported. These are not VEGF-axis rows and are not counted as DRY evidence.
Disconfirming and null first
1. Human VEGF does not fall unless weight falls, and exercise alone does not move it. Cancer Research 2016;76(14):4226 (volume, issue and first page READ from the publisher's article path aacrjournals.org/cancerres/article/76/14/4226/608215), "Dietary Weight Loss and Exercise Effects on Serum Biomarkers of Angiogenesis in Overweight Postmenopausal Women: A Randomized Controlled Trial", PMCID PMC5033683. Authors named from the AACR press release: Catherine Duggan (Fred Hutchinson Cancer Research Center) and senior author Anne McTiernan. Full author list, DOI and PMID are a retrieval debt; the publisher host refused and PMC captcha-walled. MEASURED, serum VEGF geometric means, baseline to 12 months, absolute change (percent), P:
| Arm and weight-change stratum | n | Baseline | 12 months | Change (percent) | P |
|---|---|---|---|---|---|
| Control | 87 | 325.0 | 321.0 | -3.92 (-1.2) | reference |
| Diet, no change or gained | 23 | 288.9 | 292.8 | +3.97 (+1.4) | 0.13 |
| Diet, lost <5 percent | 19 | 296.4 | 284.2 | -12.2 (-4.1) | 0.36 |
| Diet, lost >=5 percent | 76 | 295.6 | 260.2 | -35.3 (-12.0) | <0.0001 |
| Exercise, no change or gained | 41 | 265.9 | 262.2 | -3.69 (-1.4) | 0.89 |
| Exercise, lost <5 percent | 46 | 302.6 | 292.0 | -10.7 (-3.5) | 0.39 |
| Exercise, lost >=5 percent | 30 | 384.3 | 365.5 | -18.8 (-4.9) | 0.25 |
| Diet+exercise, no change or gained | 12 | 381.8 | 377.6 | -4.21 (-1.1) | 0.86 |
| Diet+exercise, lost <5 percent | 14 | 369.4 | 337.4 | -32.0 (-8.7) | 0.07 |
| Diet+exercise, lost >=5 percent | 90 | 294.5 | 261.2 | -33.3 (-11.3) | <0.0001 |
MEASURED: trend P was <0.0001 in the diet arm, <0.0001 in diet+exercise, and 0.20 in the exercise-only arm. MEASURED: seven of the nine intervention strata were null. The effect sits only where weight loss reached 5 percent, and the exercise-only arm has no significant trend at all. READ, press release wording: "those in the diet arm and the diet+exercise arm had significantly lower levels of the angiogenesis-related proteins, but such effects were not apparent in those in the exercise-only arm." The release does not break the result out by protein, so the arm-level sentence is not by itself a VEGF-specific statement; the per-protein numbers above are. This is a weight-loss-mediated result under CR, not a fasting result. The units of VEGF are not stated in the table that was opened. 2. Human CR reduced VEGF in one subgroup and not in another, inside one study that contradicts itself. Korybalska K. "Angiogenesis in Adipose Tissue: How can Moderate Caloric Restriction Affects Obesity-Related Endothelial Dysfunction?" In: Lenasi H, ed. Endothelial Dysfunction, Old Concepts and New Challenges. IntechOpen; 2017. doi:10.5772/intechopen.72624. Mixed review plus the author's own 50 obese adults on 8 weeks of moderate CR. MEASURED: the Results list VEGF among factors significantly reduced, at -11 +/- 6 percent. The Discussion states CR "did not reduce the level of VEGF" in the high-BMI glucose-intolerant group and that the decline occurred "only in the normoglycemic obese." MEASURED further internal problems in that same chapter, reported and not smoothed: endostatin is reported as "-126 +/- 5 percent," a reduction greater than 100 percent, which cannot be correct; "angiostatin-1" is written where angiopoietin-1 is meant; endothelial proliferation is reported as increased in the Results and decreased in the heavier weight-losers in the Discussion; and the glucose-intolerant participants were also taking metformin, which the chapter itself credits with anti-angiogenic effects. The conclusion that moderate CR improves endothelial function rests on 2 of 13 angiogenic factors changing in the glucose-intolerant group. This chapter is used here for its conflict, not as a clean positive. 3. Weight loss does not always reduce VEGF. READ from the Korybalska chapter: Dalmas et al. 2011, surgery-induced weight loss in humans, cited for the statement that weight loss does not always reduce VEGF. Primary study located through the chapter and NOT opened. Retrieval debt. 4. A mouse model built to raise VEGF did not raise it. READ from the same chapter: Halberg et al. 2009, transgenic mice overexpressing HIF-1alpha in adipocytes under hypoxia, reported no induction of classic HIF-1alpha targets including VEGF, with fibrosis instead. Primary study located and NOT opened. Retrieval debt. Stated because it breaks the hypoxia-to-VEGF hop that a fasting narrative would want. 5. Exercise, kept separate as instructed. The founder's role line names exercise. The one randomized human trial opened here found the exercise-only arm had no significant VEGF trend (P = 0.20) over 12 months. No acute-exercise VEGF measurement was opened in this sweep, so this file makes no claim about acute exercise and VEGF in either direction. Retrieval debt. 6. Tumor angiogenesis, kept separate as instructed. No tumor-angiogenesis study is used anywhere in this node. Every row above is from non-tumor human or animal physiology, except where a cited source's own framing is named. No cancer claim is made or implied. 7. No fasting modality has a VEGF trial. Nothing located for TRE, IF-ADF, IF-5:2, PF, FMD, RAMADAN or DRY with VEGFA, FLT1, KDR, FLT4, FGF2, ANGPT1, ANGPT2, TIE2 or THBS1 measured in humans.
Graded rows
| # | Claim | Modality | Species / system | Tier | Grade | Source | Tag |
|---|---|---|---|---|---|---|---|
| 40.1 | 12 months of dietary caloric restriction reducing weight by >=5 percent lowered serum VEGF by 12.0 percent, P<0.0001, n=76; diet+exercise with >=5 percent loss lowered it 11.3 percent, P<0.0001, n=90 | CR | human RCT, postmenopausal women with overweight | H1 | B+ | Cancer Research 2016;76(14):4226, PMCID PMC5033683 | MEASURED |
| 40.2 | In the same trial VEGF was null in all strata below 5 percent weight loss, and the exercise-only arm trend was null, P=0.20 | CR | human RCT | H1 | B+ | same | MEASURED |
| 40.3 | In the same trial PEDF, an anti-angiogenic factor, also fell with >=5 percent weight loss, by 13.7 percent in diet (P<0.0001), 8.7 percent in exercise (P=0.001) and 12.1 percent in diet+exercise (P<0.0001) | CR | human RCT | H1 | B+ | same | MEASURED; note the pro- and anti-angiogenic factors moved the same direction, which complicates any net-angiogenesis reading |
| 40.4 | 8 weeks of moderate CR lowered VEGF by 11 +/- 6 percent overall in 50 obese adults, but the same chapter states the fall occurred only in the normoglycemic subgroup | CR | human single-arm | H3 | C | Korybalska 2017, IntechOpen, doi:10.5772/intechopen.72624 | MEASURED, and self-contradictory within the source |
| 40.5 | Serum VEGF correlates positively with anthropometric measures in obese humans, cross-sectional | n/a observational | human | H2 observational | C | Miyazawa-Hoshimoto et al., READ through Korybalska 2017, NOT opened | READ, secondhand; retrieval debt |
| 40.6 | Hypoxia raised VEGF in differentiated mouse 3T3-F442A adipocytes. Hypoxia, not fasting | n/a | mouse cell line | C1 | C | Lolmede et al. 2003, READ through Korybalska 2017, NOT opened | READ, secondhand; retrieval debt |
| 40.7 | Ten years of CR in rhesus macaques was described as showing a similar lack of VEGF-driven proliferation. Direction unclear in the secondary source | CR | primate in vivo | A1 | C | Csiszar et al., READ through Korybalska 2017, NOT opened | READ, secondhand and ambiguous; retrieval debt |
| 40.8 | Fasting raises HIF signaling and ketone availability, therefore VEGF should rise. Hops: fasting to HIF (another slice, not measured here), HIF to VEGF (broken by Halberg 2009 in vivo). Inference rejected | n/a | hops shown | M | C | derivation stated; Halberg 2009 READ through Korybalska 2017 | inference, shown and NOT accepted |
Conflict left unresolved, named on both sides
- Does caloric restriction lower human VEGF? Side A: yes, where weight loss exceeds 5 percent (Cancer Research 2016, H1, P<0.0001, two arms). Side B: not in glucose-intolerant obese adults (Korybalska 2017, H3, same direction of intervention, opposite result in a subgroup), and weight loss does not always lower it (Dalmas 2011, cited and unopened). Unresolved.
- Pro- versus anti-angiogenic direction. VEGF (pro-angiogenic) and PEDF (anti-angiogenic) both fell in the same trial, in the same participants. What that means for net angiogenesis is not resolved by anything located here, and no net-angiogenesis measurement was made in any human source opened.
- Is any of this fasting? Every human row in this node is CR. Unresolved because untested.
Open questions
- VEGFR1, VEGFR2, VEGFR3, FGF2, ANGPT1, ANGPT2, TIE2 and THBS1 under any fasting modality in humans: nothing located. Only the ligand VEGFA has human intervention data, and only under CR.
- Acute-exercise VEGF, which the founder's own role line invites, was not opened in this sweep.
- Full citation for the one H1 source in this node is incomplete. See debts.
← 39. Integrin / FAK adhesion signaling All 41 nodes 41. Immune-checkpoint signaling (T-cell co-inhibition) →
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.