10.1 Hazards and populations covered, done / total
A hazard or population counts as done when the accessible channels were exhausted and the finding was recorded, including where the finding is a measured zero. It does not mean strong evidence was found. The grade column in each row says what was found.
| Block | Item | Done | Grade of best evidence found |
|---|---|---|---|
| (a) Harms | Acute kidney injury | ✅ | B+ both directions (OR 0.65 lower in fasters; 3/60 ARF in CKD) |
| (a) Harms | Rhabdomyolysis | ✅ | C (CK >33,000 U/L, n=1) + NATA SOR C |
| (a) Harms | Electrolyte and sodium disturbance | ✅ | B+ (direct measurement in a dry fast: K p<0.001, Na p=0.129) |
| (a) Harms | Hypernatremia | ✅ | B+ × 2 independent streams, 2 languages (49.5%; 42-60%) |
| (a) Harms | Thrombotic events | ✅ | B+ (4.8% of 311, OR 10.64, no control group) |
| (a) Harms | Syncope | ✅ | Measured zero for the hard endpoint. B+ proxy (dizziness p=0.039) |
| (a) Harms | Death | ✅ | Measured zero peer-reviewed. 1 media report, graded NG |
| (b) Survival span | How long without water, and the physiology across it | ✅ | B+ (VSED median 7 days, end-of-life) + A (NATA thresholds) |
| (c) Must not | Pregnancy | ✅ | A (umbrella review) + NG (IDF-DAR 6.5 pts, NHS, Russian manual) |
| (c) Must not | Breastfeeding | ✅ | B+ (55 infants, NS) + NG (IDF-DAR right not to fast) |
| (c) Must not | Type 1 diabetes | ✅ | A (hypoglycaemia 50.79/100) + NG (Russian absolute contraindication) |
| (c) Must not | Type 2 diabetes, insulin or SGLT2i | ✅ | C (EDKA case, pH 6.92) + NG (IDF-DAR, NHS) |
| (c) Must not | Chronic kidney disease | ✅ | B+ (3/60 then 4/60 ARF) + NG (eGFR<30 = 6.5 pts → "should not fast") |
| (c) Must not | History of kidney stones | ✅ | B+ (urine volume 604 vs 1031 mL, p=0.001) + B+ (RR 1.267) |
| (c) Must not | Cardiovascular disease | ✅ | NG (IDF-DAR 6.5 pts unstable; NHS "moderate-severe heart failure"); no measured CV outcome study of dry fasting |
| (c) Must not | On diuretics | ✅ | Measured zero during fasting. NG governing text |
| (c) Must not | On lithium | ✅ | B+ null at 14 h (p=0.052) |
| (c) Must not | On NSAIDs | ✅ | B+ (IRR 2.08) but measured without fasting |
| (c) Must not | On ACE inhibitors or ARBs | ✅ | B+ (same cohort) + NG ("should not stop ... unless ... asked") |
| (c) Must not | Eating disorder history | ✅ | B+ direction only, no effect estimates obtained |
| (c) Must not | Children | ✅ | A in T1D adolescents. Measured zero in healthy children |
| (c) Must not | Older adults | ✅ | NG (frailty 6.5 pts, >70 no support 3.5 pts, Russian >75) |
| (c) Must not | Anyone working in heat | ✅ | A (NATA >1% and >3% thresholds) + B+ (5.44 L / 12 h requirement) + NG (labour 4 pts) |
| (d) Interactions | Lithium toxicity | ✅ | B+ null |
| (d) Interactions | NSAID nephrotoxicity | ✅ | B+ positive, quantified, no fasting |
| (d) Interactions | SGLT2i and euglycemic ketoacidosis | ✅ | C positive, in a fasting context |
| (d) Interactions | Diuretics | ✅ | Measured zero |
| (d) Interactions | Anticoagulants | ✅ | A, significant post-fast only (RR 1.69) |
| (e) Exemption | Ramadan medically supervised exemption | ✅ | NG × 4 sources, 4 languages, 4 jurisdictions |
| (f) Position | Regulatory or professional body statements | ✅ | Measured zero regulatory. 6 texts ranked, best is A (NATA, on hypohydration not fasting) |
HAZARDS AND POPULATIONS: 27 of 27 done. COMMISSIONED SECTIONS: 6 of 6 done (a, b, c, d, e, f). LANGUAGES: 10 of 10 searched. 8 of 10 yielded. 1 located-unopened (Persian). 1 empty (Urdu).
10.2 Measured zeros, stated as zeros
Silence is the failure, so every zero is printed with its denominator rather than left out.
| Question | Studies located | Languages searched |
|---|---|---|
| Controlled trials of a continuous dry fast longer than 24 hours | 0 | 10 |
| Peer-reviewed reports of death attributed to dry fasting | 0 | 10 |
| Studies measuring syncope as an endpoint during dry fasting | 0 | 10 |
| Studies measuring diuretic users during a dry fast | 0 | 10 |
| Studies measuring the triple-whammy combination during a dry fast | 0 | 10 |
| Studies measuring lithium during a continuous multi-day dry fast | 0 | 10 |
| Studies of dehydration or renal outcomes in healthy children during dry fasting | 0 | 10 |
| Studies of dry fasting and eating-disorder outcomes | 0 | 10 |
| Studies of pregnancy during a continuous multi-day dry fast | 0 | 10 |
| Regulatory agency position statements on dry fasting | 0 | 10 |
| Professional body position statements naming dry fasting as their subject | 0 | 10 |
| Clinical guidance documents addressing dry fasting as a wellness practice | 0 | 10 |
10.3 Open debts, named and not hidden
1. İslam M 2023, DOI 10.32322/jhsm.1330331. 192 stage 3-5 CKD patients fasting 14-15 h/day is measured. The eGFR and creatinine values, their p values, the adverse events and the conclusion were not obtained: the abstract on the page was truncated and the p values garbled. Open. 2. Ganson & Nagata 2022, DOI 10.1016/j.eatbeh.2022.101681. Direction of association obtained from the University of Toronto release. No odds ratios, confidence intervals, p values or betas obtained. Both hosts serving the paper returned client errors. Open. 3. NATA position statement, DOI 10.4085/1062-6050-52.9.02. The converted text was garbled at two figures: the clinical severity bands (mild about 1-5%, moderate about 5-10%, severe above 10%) and the core-temperature coefficient (0.15-0.20°C per 1% body weight). Both must be confirmed against the original PDF before either is published on a customer surface. Every other NATA figure came through clean. Open. 4. Klatte 2023, handle fub188/37553. Body weight and 24-hour urine volume p values not confirmed; the document text was cut off before them. Open. 5. Nutrients 2022;14(1):148, DOI 10.3390/nu14010148. A PubMed link on that page includes the number 35011024. It was not confirmed as the PMID and is therefore not stated as one. Open. 6. Pooled DKA rate for SGLT2 inhibitor users during Ramadan. Not obtained. The systematic review (DOI 10.1055/s-0044-1791484) is behind an automated-request block and the meta-analysis (PMC12897491) behind reCAPTCHA. Open. 7. Persian study on osmolarity and serum electrolytes during Ramadan fasting, Iranian Journal of Endocrinology and Metabolism, article ID 375. Located by title and journal. Both hosts returned robots-unreachable. Authors, year, volume and pages not obtained. Open, and this is the most consequential open debt in the strand: it is the only located non-English study whose primary outcome is exactly the mechanism in question. 8. Publisher-side inconsistency, BMC Nephrology 2022, DOI 10.1186/s12882-022-02674-1. The Results text gives the adjusted OR as 0.66 and Table 2 gives 0.65. Recorded, not resolved. Both figures are printed in A4a rather than one being chosen.
10.4 Unopened sources, named so that no verdict rests on them
Rule 25: a verdict may not rest on an unopened source. Each of the following was located and not opened. Nothing in Sections A through G rests on any of them.
1. EPIDIAR itself. Its figures appear in C4 only as reported by DOI 10.3389/fendo.2016.00032, and that secondhand status is stated in the row. 2. PMID 28098105, "Fasting in Ramadan is not associated with deterioration of chronic kidney disease: A prospective observational study." A counter-finding, named so the counter-direction is visible. 3. PMID 38692594, "The effect of Ramadan on elderly patients presenting to the emergency department." 4. PMID 1819340, "The health risks of occupational stress in islamic industrial workers during the Ramadan fasting period." Title and PMID only. Journal, year, volume and pages not obtained, so this does not yet meet the publisher-article-record standard. 5. "Effect of fasting during Ramadan on thermal stress parameters" (Manjunath, Aravindhakshan). Every host returned a client error. 6. Journal of Family and Community Medicine 2016;23(2), "Ramadan fasting is not usually associated with the risk of..." Another counter-finding, named. 7. PMID 39572100 / PMC11580238, "Impact of Ramadan fasting on kidney function and related outcomes in chronic kidney disease and kidney transplant recipients: a systematic review and meta-analysis." This is a grade-A source on a core population and it was not opened. 8. DOI 10.1055/s-0044-1791484, SGLT2 inhibitor Ramadan systematic review. 9. PMC12897491, SGLT2 inhibitor Ramadan meta-analysis. 10. PMID 17679744, "Fasting the month of Ramadan by Muslims: could it be injurious to their kidneys?" 11. PMID 39636701, "Voluntarily Stopping Eating and Drinking: Systematic Review of Empirical Studies." A grade-A source directly on Section B and it was not opened. 12. Emara et al. 2022 (199 haemodialysis patients) and Ayu Islami & Nugroho 2024 (44 patients). Their figures appear in E4 only as cited by the Indonesian article, and that is stated in the row. 13. The Persian article in 10.3 item 7.
10.5 Channel receipt
Channel law was applied: the instant a channel blocked, throttled, captcha'd, refused or stalled, it was abandoned for the rest of the run and never retried.
Abandoned on first block, never retried: eutils.ncbi.nlm.nih.gov (proxy refused, HTTP 403); www.ebi.ac.uk Europe PMC REST (proxy refused, HTTP 403); pmc.ncbi.nlm.nih.gov (reCAPTCHA); pubmed.ncbi.nlm.nih.gov (reCAPTCHA); europepmc.org (client error); www.thieme-connect.com (automated-request block); onlinelibrary.wiley.com (client error); www.sciencedirect.com (client error); journals.lww.com (client error); journals.sagepub.com (client error); www.tandfonline.com (client error); www.semanticscholar.org (client error); www.researchgate.net (client error); ijem.sbmu.ac.ir (robots unreachable); www.sid.ir (robots unreachable).
Channels that carried the work: WebSearch (57 queries, standard and extended); link.springer.com; academic.oup.com; mdpi.com; frontiersin.org; media.idf.org; dergipark.org.tr; cuaj.ca; revistamedicinamilitara.ro; nata.org; annfammed.org; health.clevelandclinic.org; primarycare.northeastlondon.icb.nhs.uk; muftiwp.gov.my; bio.wikireading.ru; applications.emro.who.int; medrxiv.org; eajem.com; legit.ng; eurekalert.org; nephro.blog; journal.stikespemkabjombang.ac.id; tawassol.ma; refubium.fu-berlin.de (only the jsessionid URL form answered; the plain bitstream form returned a client error); pmc.ncbi.nlm.nih.gov (three successes before its first reCAPTCHA, after which it was abandoned).
Direct literature-API access was unavailable for this entire run. Both PubMed E-utilities and the Europe PMC REST service were refused by the egress proxy at HTTP 403 on the first attempt. Every identifier in this file was therefore read off a rendered page, not resolved through an API, which is why several rows state "PMID not shown on the page I opened" rather than supplying one.
10.6 What this register does not establish
Stated plainly, without apology and without inflation in either direction.
1. No controlled human trial of a continuous multi-day dry fast exists. Everything graded A or B+ in this file was measured on a 11-to-18 hour fast with overnight rehydration. 2. The single hospital-characterized continuous dry fast is n = 1, a conference abstract, and does not report its patient's outcome. 3. Three of the hazards named by the commission return a measured zero for dry fasting specifically: syncope as a hard endpoint, death in the peer-reviewed record, and diuretic interaction. 4. Two of the mechanisms most often asserted for dry fasting are contradicted by the only direct measurements located. Plasma osmolality fell and hematocrit fell during a measured 19-day dry fast (p < 0.001 each). Serum lithium fell rather than rose (p = 0.052). 5. Three grade-A sources directly on commissioned questions were located and not opened (10.4 items 7, 11 and the SGLT2 meta-analysis), and one directly on-target non-English primary study was located and not opened (10.3 item 7). Those four are the highest-value remaining work in this strand. 6. Several load-bearing sources carry internal inconsistencies that were recorded rather than smoothed: the BMC Nephrology adjusted OR (0.65 vs 0.66), the Romanian military medicine stone paper (study period, p values, comparison direction), the J Fasting Health breastfeeding paper (sample size, measurement schedule, table totals), the Frontiers secondary report of EPIDIAR (7.5-fold described two ways), and the IDF-DAR score bands (arithmetic gaps at 3 to 3.5 and above 6).
END OF STRAND 4.
Compiled 2026-11-22. 27 of 27 hazards and populations done. 6 of 6 commissioned sections done. 10 of 10 languages searched, 8 yielding, 1 located-unopened, 1 empty. 12 measured zeros printed with their denominators. 8 open debts named. 13 unopened sources named. 15 channels abandoned under channel law, 25 channels carried the work.
Limits live in receipts. Nothing above was softened, and nothing that cut against the hazards was suppressed.
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.