Founder's components: JAK1/2/3, TYK2, STAT3/5, SOCS.
DRY status
DRY: none located for any node component. Ligand-level DRY evidence exists and is not a substitute. Al lami et al. 2024 (30-day dawn-to-dusk dry fast) measured cytokines that signal through this node and reported decreases: IL-15 (p = 0.014), IL-22 (p = 0.025), IL-12p40 (p = 0.038). No JAK, TYK2, STAT or SOCS protein, transcript or phosphorylation state was measured in that study or in any other dry-fast study located. Papagiannopoulos et al. 2013, Papagiannopoulos-Vatopaidinos et al. 2020, Koppold et al. 2021 and Mähler et al. measured no component of this node. Queries run: the four DRY queries listed under node 1, plus "fasting STAT5 growth hormone pulsatility human liver IGF-1 resistance", "fasting SOCS3 STAT3 hypothalamus leptin signaling mouse food deprivation", "FGF21 inhibits growth hormone signaling STAT5 liver fasting Inagaki". Empty for DRY.
Disconfirming, nulls and contradictions first
- No study located, in any species and under any fasting modality, measured a JAK, TYK2, STAT or SOCS protein. This node is unmeasured in the fasting literature reached in this run. Every row below is ligand-level or inferred. The node must not be written as modulated by fasting. [M, grade C]
- The single most-measured ligand for this node is flatly null in humans. IL-6 signals through JAK1/STAT3. Mulas et al. 2023 found IL-6 unchanged in 5 of 5 human trials that measured it. Lin et al. 2025 found IL-6 unchanged at 12 months in both a TRE arm and a calorie-restriction arm. Khalafi et al. 2025, pooling 21 studies and 839 participants, found no significant effect of intermittent fasting on IL-6. [H1, grade B+ for the null]
- The two meta-analyses contradict each other on IL-6. Khalafi et al. 2025 report IL-6 as not significant across 21 studies. Lü et al. 2025, pooling 8 randomized trials and 573 participants in metabolic syndrome, report IL-6 reduced by a weighted mean difference of -0.58, 95 percent CI -1.08 to -0.08, p = 0.022. Both are 2025 systematic reviews with meta-analysis, both searched to 2024 or 2025, and they disagree. Reported unresolved; neither is selected as the winner.
- The one mechanism that would wire fasting to STAT5 could not be opened. The FGF21-to-growth-hormone- signaling link was located as a titled source and the host was JavaScript-gated on both the article and the PDF path. No row is written from it. Logged as retrieval debt D25.
Graded rows
| # | Modality | System | Tier | Grade | Finding | Source |
|---|---|---|---|---|---|---|
| 3.1 | TRE and IF-ADF (pooled) | Human, 5 trials measuring IL-6 | H1 | B+ | IL-6 unchanged in 5 of 5 trials. LIGAND LAYER ONLY. Authors' verbatim conclusion: "intermittent fasting has little or no effect on key circulating pro-inflammatory cytokines". | Mulas et al. 2023, Front Nutr 10:1146924, DOI 10.3389/fnut.2023.1146924 |
| 3.2 | Mixed IF (pooled) | Human, 21 studies, 839 participants | H1 | B+ | No significant effect on IL-6. Time-restricted feeding held the highest probability ranking for IL-6 change, but that ranking was NOT statistically significant. | Khalafi et al. 2025, Nutrients 17(15):2388, PMID 40805975, DOI 10.3390/nu17152388 |
| 3.3 | Mixed IF (pooled) | Human, 8 RCTs, 573 participants, metabolic syndrome | H1 | B+ | IL-6 reduced, WMD -0.58, 95% CI -1.08 to -0.08, p = 0.022. CONTRADICTS ROW 3.2. GRADE assessed, Cochrane RoB 2, no publication bias by Begg's test. | Lü et al. 2025, J Health Popul Nutr 44(1):293, PMID 40826125, DOI 10.1186/s41043-025-01039-2 |
| 3.4 | DRY (30 d dawn-to-dusk) | Human, increased BMI, single arm | H3 | C | JAK/STAT-family cytokines decreased: IL-15 p = 0.014, IL-22 p = 0.025, IL-12p40 p = 0.038. LIGAND LAYER. No JAK or STAT measured. | Al lami et al. 2024, Metabolism Open 21:100274, PMID 38455231, DOI 10.1016/j.metop.2024.100274 |
| 3.5 | RAMADAN (1 month) | Human, n=34 men | H3 | C | IL-2 significantly decreased. IL-2 signals through JAK1/JAK3 and STAT5. LIGAND LAYER. No numbers retrieved. | Rahbar et al. 2019, Int J Prev Med 10(1):7, PMID 30774841, DOI 10.4103/ijpvm.ijpvm_252_17 |
| 3.6 | RAMADAN (1 month) | Human, n=40 fasting vs n=28 control | H2 | B | IL-6 significantly lower during Ramadan than at baseline, p < 0.001, with no change in controls. CONTRADICTS ROWS 3.1 and 3.2 at the ligand layer. Note the modality differs: Ramadan restricts fluid as well as food. Absolute values not retrieved. | Aksungar et al. 2007, Ann Nutr Metab 51(1):88-95, PMID 17374948, DOI 10.1159/000100954 |
| 3.7 | none (no fasting tested) | Not stated in abstract; MeSH lists mice | M | C | Age-related increases in NF-kB, IL-1beta, IL-6, TNF-alpha, COX-2 and inducible nitric oxide synthase attenuated by calorie restriction. Review, no numbers retrieved, species not stated in the abstract. Included for node 9 and listed here only because IL-6 is a JAK/STAT ligand. | Chung et al. 2001, Ann N Y Acad Sci 928(1):327-335, PMID 11795524, DOI 10.1111/j.1749-6632.2001.tb05662.x |
Open questions
- Human PBMC and adipose material has been collected under defined fasting protocols (Elliott 36 h, Defour 26 h, Han 24 h) and in none of the retrieved records was a STAT phosphorylation state reported. The node is closable with existing sample banks and has not been closed.
- Whether the Ramadan IL-6 decrease (row 3.6) and the TRE IL-6 null (rows 3.1, 3.2) are a modality difference, a fluid-restriction difference, a population difference or a measurement-timing difference. Both were retrieved; neither explains the other.
← 2. PI3K / AKT / mTOR All 41 nodes 4. Hedgehog →
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.