Dry Fasting — The Record

Limits register

4. SECTION C: WHO MUST NOT DO THIS AT ALL

Documented harms are kept apart from theorized ones. Every row names the source it was read from.

Fifteen populations named by the commission. Each row carries the evidence found behind it. Where the governing text is a position statement rather than a study, that is stated and the row is graded NG. Where the evidence is thin or absent, that is stated rather than filled in.

C0. The governing professional-body text that covers nine of these fifteen at once

FieldContent
GradeNG (professional body position statement, international)
IdentifierISBN 978-2-930229-99-7
CitationDiabetes and Ramadan: Practical Guidelines, 2021 update of the 2016 first edition. International Diabetes Federation (IDF) and DAR International Alliance, Brussels, Belgium. January 2021. Editor: Mohamed Hassanein.

READ, verbatim, the three recommendation sentences: > "Individuals who are in the high risk category should not fast."

> "Those at the moderate risk level are advised not to fast."

> "Those at the low risk level should be able to fast."

Moderate-risk individuals who choose to fast anyway "would need to be cautious and discontinue fasting if any problems arise." Individuals previously deemed "very high risk" receive the same recommendations as the high-risk group.

MEASURED, the risk bands: low risk 0 to 3, moderate risk 3.5 to 6, high risk above 6.

MEASURED, the point values (Table 2), which is how populations below reach "should not fast":

Risk elementPoints
Hypoglycaemia unawareness6.5
eGFR under 306.5
Unstable macrovascular disease6.5
Pregnant and not within targets6.5
Impaired cognition or frailty6.5
Recent severe hypoglycaemia5.5
Physical labour, highly intense4
eGFR 30 to 454
Multiple episodes of hypoglycaemia weekly3.5
Over 70 with no home support3.5
Pregnant and within targets3.5
Multiple daily mixed insulin injections3
DKA or HONC within the last 3 months3
Basal bolus or insulin pump2.5
Once-daily mixed insulin2
Stable macrovascular disease2
eGFR 45 to 602
HbA1c above 9%2
SMBG indicated but not done2
Basal insulin1.5
Type 1 diabetes1
Diabetes duration 10 years or more1
Glibenclamide1
Fasting hours 16 or more1
Gliclazide MR, glimepiride or repaglinide0.5
Other therapy, not sulfonylurea or insulin0

READ, the pregnancy footnote verbatim: > "Pregnant and breastfeeding women have the right to not fast regardless of whether they have diabetes."

Two gaps in that instrument, recorded because they are load-bearing:

1. SGLT2 inhibitors are not named anywhere in Table 2. Under the table's own logic they fall in the "other therapy" line at 0 points. The chapter does not state this. Any claim that IDF-DAR scores SGLT2 inhibitors is not supported by the text I read, and the absence is published here rather than filled in. 2. The score bands have arithmetic holes. Scores between 3 and 3.5, and any decimal value between 6 and the next whole number, fall outside every band as written. The chapter does not say how to classify them.

C1. Pregnancy

FieldContent
GradeA (umbrella review of 13 reviews, several containing meta-analyses)
IdentifierPMID 39785103, PMC12093917, DOI 10.1002/ijgo.16127
CitationAl-Taiar A, et al. "Impacts of Ramadan fasting during pregnancy on pregnancy and birth outcomes: An umbrella review." International Journal of Gynaecology and Obstetrics, epub 9 Jan 2025, print June 2025.

MEASURED, as reported by the included reviews (these are the source reviews' estimates, not new pooled analyses by the umbrella authors):

OutcomeSource reviewEffect95% CIp
Birth weight, 3rd trimesterOng 2023 (MA)−60 gnot reported< 0.007
Birth weight, 2nd + 3rd trimestersOng 2023 (MA)−80 gnot reported0.03
Birth weight, one study n = 402Oosterwijk 2021lower mean in fasting groupnot reported0.024
Birth weightGlazier 2018 (MA)SMD 0.030.00 to 0.05not reported
Birth weightChen 2023 (MA)+0.01 kg−0.06 to 0.08not reported
Low birth weightGlazier 2018 (MA)OR 1.050.87 to 1.26not reported
Low birth weightChen 2023 (MA)OR 1.370.74 to 2.53not reported
Preterm birthChen 2023 (MA)OR 0.930.60 to 1.44not reported
Preterm birthGlazier 2018 (MA)OR 0.990.72 to 1.37not reported
Gestational age at birthChen 2023 (MA)+0.05 weeks−0.22 to 0.31not reported
Fasting blood glucoseChen 2023 (MA)−2.73 mg/dL−5.23 to −0.220.03
Biparietal diameter, 1st trimesterOng 2023 (MA)−0.20 mm−0.37 to −0.030.02
Biparietal diameter, 2nd trimesterOng 2023 (MA)−0.20 mm−0.38 to −0.020.03
Biparietal diameter, 3rd trimesterOng 2023 (MA)−0.21 mm−0.39 to −0.020.03
Biparietal diameter, 2nd + 3rdOng 2023 (MA)+0.39 mm0.15 to 0.640.002
Femur length, 2nd trimesterOng 2023 (MA)−0.20 mm−0.34 to −0.060.006
Femur length, 2nd + 3rdOng 2023 (MA)+0.17 mm0.06 to 0.290.003
5-minute Apgar, one study n = 215Ong 20239.00 ± 0.01 vs 8.92 ± 0.53not reported0.044
Amniotic fluid index, 2nd trimester, n = 80Ong 2023 (MA)reduced AFI, no numeric estimatenot reportednot reported
Placental weightGlazier 20181 of 3 studies found lower weight, no pooled estimatenot reportednot reported
Gestational weight gainChen 2023 (MA)+0.5 kg−1.27 to 0.26not reported
Gestational diabetesChen 2023 (MA)no significant difference, no estimate given
Pre-eclampsiaChen 2023 (MA)no significant effect, no estimate given

READ, the conclusion verbatim: > "There is little evidence that Ramadan fasting during pregnancy can negatively impact pregnancy and birth outcomes."

Both halves published, neither suppressed. The umbrella review's verdict is that evidence of harm is little. Inside the same review, third-trimester birth weight was measured 60 g lower (p < 0.007), second-and-third-trimester birth weight 80 g lower (p = 0.03), biparietal diameter lower at every trimester, and amniotic fluid index reduced in one second-trimester study. The signs are inconsistent: the 2nd+3rd-trimester biparietal and femur estimates are positive while the single-trimester ones are negative. The review reports no effect sizes for stillbirth, miscarriage, congenital anomalies, neonatal death, or small-for-gestational-age, and reports none for long-term offspring outcomes.

The governing text is not the study. IDF-DAR, verbatim: "Pregnant and breastfeeding women have the right to not fast regardless of whether they have diabetes." Pregnant and not within targets scores 6.5, which alone exceeds the high-risk threshold of 6, and the high-risk recommendation is "should not fast." NHS North East London lists "Women who are pregnant or breastfeeding or menstruating" among groups for whom "Fasting is not considered compulsory." The Russian fasting-therapy manual lists "период беременности и лактации" (pregnancy and lactation) as a relative contraindication.

Exposure caveat. Every pregnancy figure above is from a dawn-to-sunset intermittent dry fast with overnight rehydration. No study of pregnancy during a continuous multi-day dry fast was located in any language.

C2. Breastfeeding

FieldContent
GradeB+ (prospective comparison, 55 infants) for infant growth. NG for the exemption.
IdentifierNo DOI, PMID or PMC appeared on the document. Journal record: J Fasting Health 2013;1(2):66-69.
CitationHaratipour H, Sohrabi MB, Ghasemi E, Karimi A, Zolfaghari P, Yahyaei E. "Impact of Maternal Fasting During Ramadan on Growth Parameters of Exclusively Breastfed Infants in Shahroud, 2012." Published 25 Dec 2013.

MEASURED: 55 infants (abstract gives 20 case and 35 control; the Methods section gives 25 and 30, so the paper is internally inconsistent and that is recorded). 23 infants (41.8%) male. Mean age 3.43 ± 1.38 months case versus 2.31 ± 1.45 months control, not significantly different.

Weight (kg), case versus control, at start of Ramadan through 3 months after: 4.45 ± 2.15 / 4.27 ± 2.11; 4.95 ± 2.84 / 5.12 ± 3.01; 6.27 ± 3.24 / 6.05 ± 3.27; 7.43 ± 5.14 / 7.65 ± 5.38; 8.23 ± 5.28 / 8.24 ± 5.35. Between-group p values NS for weight, length and head circumference.

READ, verbatim: > "All growth parameters increased during the study period (P<0.05), with the same rate of increase for both groups (P=0.125)."

> "fasting of breastfeeding mothers does not have any influence on growth indices of the infants."

Measured gap stated: milk volume and feeding frequency were not measured or reported in that study. Multiple further internal inconsistencies are present (sample sizes differ between abstract and methods, the measurement schedule is described two different ways, Table 1 totals do not match the group sizes, and the Results text says two months while Table 2 covers three). Reliability caveat attached accordingly.

The governing text, again, is not the study. IDF-DAR grants breastfeeding women the right not to fast regardless of diabetes status. NHS NEL lists breastfeeding among the non-compulsory groups. The Russian manual lists lactation as a relative contraindication.

C3. Type 1 diabetes

FieldContent
GradeA (systematic review and meta-analysis, 9 studies, 458 adolescents)
IdentifierDOI 10.1186/s12902-025-01835-1. PMID not shown on the page I opened.
CitationSafari O, Shafiee A, Heidari A, Nafarzadeh F, Aminzadeh D, Esmaeilpur Abianeh F, Amini MJ, Bakhtiyari M, Bahadori Monfared A. "Ramadan fasting among adolescents with type 1 diabetes: a systematic review and meta-analysis." BMC Endocrine Disorders 2025;25:45. Published 18 Feb 2025.

MEASURED, pooled random-effects event rates per 100 observations:

READ, verbatim: > "Meta-analysis showed no significant reduction in HbA1c levels post-Ramadan (SMD: -0.12; 95% CI: -0.38 to 0.14),"

> "Ramadan fasting among adolescents with T1DM does not significantly alter HbA1c levels,"

The authors also conclude that the high incidence of hypoglycaemia calls for vigilant glucose monitoring and tailored management during fasting.

Recorded exactly as found. A pooled hypoglycaemia rate of 50.79 per 100 and a pooled DKA rate of 11.02 per 100 sit inside a paper whose stated conclusion is that fasting does not significantly alter HbA1c. The confidence intervals on hyperglycaemia (3.11 to 88.09) and DKA (0.36 to 30.78) are extremely wide, which is a precision limitation and is stated here rather than hidden by quoting the point estimate alone.

Governing texts: IDF-DAR scores type 1 diabetes at 1 point and, READ verbatim from Chapter 2, "People with T1DM have been categorised as high and very high risk" for specific Ramadan complications; Chapter 1 lists T1DM among high-risk groups. The Russian fasting-therapy manual lists "инсулино-зависимый сахарный диабет I типа" (insulin-dependent type 1 diabetes) as an absolute contraindication.

C4. Type 2 diabetes, and specifically on insulin or SGLT2 inhibitors

FieldContent
GradeB+ for the underlying multinational observational study. See the caveat: I did not open it.
IdentifierDOI 10.3389/fendo.2016.00032 for the secondary report I did open.
Citation of the source I openedAlmalki MH, Alshahrani F. "Options for Controlling Type 2 Diabetes during Ramadan." Frontiers in Endocrinology 2016;7:32. Published 18 April 2016.

MEASURED, as reported in that secondary source, citing the EPIDIAR study:

Rule 25 compliance, stated plainly: I did not open EPIDIAR itself. The 7.5-fold and fivefold figures above are read off a 2016 narrative review that cites EPIDIAR. EPIDIAR is an unopened source and no verdict in this register rests on it. The secondary source is also internally inconsistent: it describes 7.5-fold as a risk ratio in one place and as a "hypoglycemia increase" in another, and it separately states that patients "had at least one episode of severe hyperglycemia requiring hospitalization, 4% of which occurred during Ramadan," which is a different measure. It reports no type 1 figures, no events per 100 people per month, and no absolute incidence rates. Those inconsistencies are published here.

SGLT2 inhibitors, the specific hazard: euglycemic DKA with fluid withheld. See D3 below. The governing instruction is NHS NEL, READ verbatim: SGLT2 inhibitors are "ideally, don't start within one month of or during Ramadan. If stable on it, do not stop it."

Insulin. IDF-DAR scores multiple daily mixed insulin at 3, basal bolus or pump at 2.5, once-daily mixed at 2, basal at 1.5. NHS NEL, READ verbatim: "Insulins require a dose reduction (e.g. short-acting/pre-mixed by 25-50%) and/or a change to timings."

C5. Chronic kidney disease

FieldContent
GradeB+ (prospective cohort, 60 patients) for the ARF signal. NG for the thresholds that govern practice.
IdentifierJournal record: Saudi J Kidney Dis Transpl 2015 Mar;26(2):320-4. DOI and PMID not shown on the page where I read this record.

MEASURED (Mbarki H, Tazi N, Najdi A, Tachfouti N, Arrayhani M, Sqalli T): 60 CKD patients, 35 female, mean age 45.6 ± 15.8. 3 developed acute renal failure in the first week. 4 developed acute renal failure at the end of the month. See A4b above for the full row including the Arabic-versus-English summary discrepancy.

The governing numeric threshold, MEASURED from IDF-DAR Table 2: eGFR under 30 scores 6.5 points, which on its own exceeds the high-risk threshold of 6, and the high-risk recommendation is "should not fast." eGFR 30 to 45 scores 4; eGFR 45 to 60 scores 2.

Russian fasting-therapy manual, absolute contraindication, verbatim: > "хроническая недостаточность функции печени и почек любого генеза" > > "chronic liver or kidney failure of any origin"

Turkish-language nephrology review, Demir ME, "Böbrek hastaları ve oruç ibadeti" (Ramadan fasting and patients with kidney diseases), Journal of Health Sciences and Medicine 2018;1(2):45-46. No DOI present on the document. READ, verbatim Turkish with translation: > "Ramazan ayında ki uzun açlık ve susuzluk sürelerinin böbrek sağlığını tehdit etme riski her zaman vardır." > > "The risk that long fasting and thirst periods during Ramadan threaten kidney health is always present."

> "Yine su tüketiminin yetersiz olması yavaş yavaş ve ilerleyici bir şekilde böbrek fonksiyonlarının azalmasına yol açabilir." > > "Likewise, insufficient water intake may gradually and progressively lead to a decline in kidney function."

That article is a narrative review (Derleme). It carries no measurement, so it is tagged READ throughout and graded NG.

A Turkish measured study exists and was partially opened. İslam M, "Impact of Ramadan fasting on eGFR in patients with late stage chronic kidney disease," Journal of Health Sciences and Medicine 2023;6(5):1016-1021, DOI 10.32322/jhsm.1330331. MEASURED: 192 stable stage 3 to 5 CKD patients, self-controlled longitudinal observational design, about 14 to 15 hours of fasting per day for one month. The eGFR and creatinine values, p values, adverse events and conclusion were not obtainable from the page I opened; the abstract was truncated and the p values garbled. Logged as an open debt in Section 9, not filled in.

A counter-finding, named and carried: a prospective observational study titled "Fasting in Ramadan is not associated with deterioration of chronic kidney disease" (PMID 28098105) exists. I did not open it because the hosts serving it returned reCAPTCHA or client errors. It is logged as an unopened source in Section 9. No verdict here rests on it, and its title is recorded so the counter-direction is not hidden.

C6. History of kidney stones

FieldContent
GradeB+ (systematic review of 10 observational studies, 9,906 fasting participants). Not graded A: it is a systematic review without pooled estimates.
IdentifierDOI 10.5489/cuaj.6664. PMID not shown.
CitationKirubarajan A, Lam ACL, Khan S, Yau M, Golda N, Buckley R. "The association between renal stones and fasting: A systematic review." Canadian Urological Association Journal 2021;15(3):E169-74. Published online 7 Aug 2020.

MEASURED: 10 observational studies, 9,906 fasting participants (9 on Ramadan fasting, 1 on a 12-hour fast).

READ, verbatim: > "with the majority (7/9) finding that renal colic incidence was unaffected by the month of fasting"

> "Based on the available evidence, it is unlikely that fasting significantly increases the risk of renal stones."

Measured gaps in that review, stated: urine volume, urine osmolality, urinary citrate and urinary oxalate were not reported in it.

The lithogenic mechanism, measured directly in stone formers.

FieldContent
GradeB+ (comparative study, 66 patients with stone history) with a reliability caveat
IdentifierNo DOI present. Journal record: Romanian Journal of Military Medicine 2021;Vol. CXXIV, No. 4, pages 525-532.
CitationAsadi M, Nazari H. "Evaluation of fasting impact during Ramadan month on patients with a history of kidney stones in a military hospital." Nephrology and Urology Research Center, Baqiyatallah University of Medical Sciences, Tehran.

MEASURED, non-fasting versus fasting:

READ, verbatim excerpts: > "fasting during Ramadan has contradictory effects on urinary stone formation factors"

> "there is not enough evidence in the present study that Ramadan fasting precipitates risk factors"

This is the hard number that matters and it is published whole. 24-hour urine volume in fasting stone formers was 604 mL against 1031 mL, a 41% lower urine volume at p = 0.001. Urinary citrate, the principal stone inhibitor, was also significantly lower in the fasting group. The same paper's own conclusion is that the effects are contradictory and that its evidence does not establish precipitated risk. Both are recorded.

Reliability caveat, stated: that paper carries several internal inconsistencies. The Abstract gives the study period as 2014-2015 while the Methods give May 2016 and May 2017. The text gives p = 0.12 and 0.14 for age and gender while Table 1 gives 0.118 and 0.452. The Results state that uric acid saturation showed no difference "(P=0.035)", which conflicts with the reported significance threshold. The Discussion in places compares "fasting" with "fasting" where it appears to mean non-fasting, so the direction of some comparisons cannot be verified from the text.

Russian fasting-therapy manual, relative contraindication, verbatim: > "мочекаменная болезнь (при оксалурии, фосфатурии, уратурии)" > > "urolithiasis, with oxaluria, phosphaturia, or uraturia"

C7. Cardiovascular disease

FieldContent
GradeNG (position statements and clinical guidance). No measured cardiovascular outcome study of dry fasting was located.

MEASURED from IDF-DAR Table 2: unstable macrovascular disease scores 6.5 points, which alone exceeds the high-risk threshold of 6, where the recommendation is "should not fast." Stable macrovascular disease scores 2.

READ from IDF-DAR Chapter 1: high-risk groups include people with comorbidities such as cardiovascular disease, stroke and chronic kidney disease, and the chapter notes that "many of whom will decide to fast against medical and religious advice."

READ from NHS North East London, the cardiovascular "should not fast" example given verbatim: "Moderate - severe heart failure".

Russian fasting-therapy manual. Absolute contraindications include "недостаточность кровообращения (легочно-сердечная недостаточность) III степени" (grade III circulatory insufficiency, cor pulmonale) and "тромбофлебиты и флеботромбозы" (thrombophlebitis and phlebothrombosis). Relative contraindications include "стойкие нарушения сердечного ритма и проводимости" (persistent cardiac rhythm and conduction disorders) and "варикозное расширение вен" (varicose veins).

The thrombosis row (A6) is the measured link. In 311 Ramadan fasters, varicose veins were present in 86.7% of the 15 VTE cases versus 38.0% of non-cases, p < 0.001, and venous reflux in 50% versus 12.7%, p < 0.001. A Russian clinical manual written decades earlier lists varicose veins and phlebothrombosis as contraindications to fasting therapy. The two are independent sources pointing the same way. They are not the same stream and are counted separately.

C8. People taking diuretics

See D4. MEASURED governing text, NHS North East London, verbatim: > "Drugs including diuretics may contribute to dehydration/acute kidney injury (AKI)."

Patients on them "may need to review with a clinician to consider dose changes/alternatives."

MEASURED, denominator stated: studies located in this sweep that measured outcomes in diuretic users specifically during dry fasting: 0. The quantified evidence for diuretics in this register is the triple-whammy nephrotoxicity cohort (D2), which did not involve fasting.

C9. People taking lithium

See D1. MEASURED: in 250 patients with bipolar disorder (131 fasting, 119 not), mid-Ramadan serum lithium was 0.81 ± 0.02 versus 0.84 ± 0.03, p = 0.052, and renal function did not differ. That is a null result at 14 hours, not a null result for a multi-day dry fast. NHS NEL lists lithium only inside the mental-health "should not fast" reason "Risk of electrolyte imbalance (e.g. lithium) or medication out of range."

C10. People taking NSAIDs

See D2. MEASURED: in a nationwide cohort, adding NSAIDs to RAS inhibitors plus diuretics raised AKI incidence from 9.6 to 20.0 per 1,000 person-years, IRR 2.08 (95% CI 1.58 to 2.74). NHS NEL's only NSAID instruction concerns dosing interval, not withdrawal.

C11. People taking ACE inhibitors or ARBs

See D2. READ, NHS North East London, verbatim: patients "should not stop taking these medications unless they are specifically asked to do so by their clinician." That guidance follows COVID-19 era advice from several cardiology and renal bodies. The measured AKI risk from the RAS-inhibitor-plus-diuretic-plus-NSAID combination is in D2.

C12. Eating disorder history

FieldContent
GradeB+ (national cross-sectional survey, over 2,700 adolescents and young adults). Effect estimates not obtained: see the caveat.
IdentifierDOI 10.1016/j.eatbeh.2022.101681
CitationGanson KT, Nagata JM, et al. "Intermittent fasting: Describing engagement and associations with eating disorder behaviors and psychopathology among Canadian adolescents and young adults." Eating Behaviors, 2022. Canadian Study of Adolescent Health Behaviors.

MEASURED, as reported in the University of Toronto release of 9 November 2022:

Stated plainly. Those numbers were read off the publisher institution's press release, not off the paper. The paper itself could not be opened: both hosts serving it returned client errors. No odds ratios, confidence intervals, p values or betas were obtained. The direction of association is recorded; the magnitude is not measured and is not stated. Logged as an open debt in Section 9.

The Cleveland Clinic consumer article (Section 6) names eating disorders among the harms of dry fasting. That is a dietitian-interview consumer article, graded NG, and it is not evidence of effect.

Exposure caveat. This evidence concerns intermittent fasting generally, not dry fasting. No study of dry fasting and eating disorder outcomes was located in any language.

C13. Children

FieldContent
GradeA for adolescents with type 1 diabetes (see C3). NG for the age thresholds that govern practice.

MEASURED, C3 restated because it is the only graded pediatric number in this register: pooled hypoglycaemia 50.79 per 100 (95% CI 30.36 to 71.10) and DKA 11.02 per 100 (95% CI 0.36 to 30.78) in 458 adolescents with type 1 diabetes across 9 studies, with one included study reporting 19.7% DKA and dehydration in the fasting group versus none in controls.

Russian fasting-therapy manual, relative contraindication, verbatim: > "ранний детский (до 14 лет) и старческий (после 75 лет) возраст" > > "early childhood under 14 and old age over 75"

MEASURED, denominator stated: studies located in this sweep measuring dehydration, electrolyte or renal outcomes in healthy children during dry fasting: 0. The pediatric evidence located is entirely in children with type 1 diabetes or adrenal insufficiency.

C14. Older adults

FieldContent
GradeNG (position statements and clinical guidance). No measured dry-fasting outcome study in older adults was obtained.

MEASURED from IDF-DAR Table 2: impaired cognition or frailty scores 6.5 (alone above the high-risk threshold of 6, where the recommendation is "should not fast"); over 70 with no home support scores 3.5.

READ from IDF-DAR Chapter 1: high-risk groups include "elderly individuals with diabetes".

Russian fasting-therapy manual: old age after 75 is a relative contraindication (verbatim above).

READ from NHS North East London: increased dehydration risk on longer, warmer days "can particularly affect people with existing medical problems such as diabetes, high blood pressure, heart disease/lung disease, pregnant or elderly."

Open debt, stated: a study titled "The effect of Ramadan on elderly patients presenting to the emergency department" exists at PMID 38692594. I did not open it: PubMed and PMC both returned reCAPTCHA, and the Europe PMC mirror returned a client error. It is logged as an unopened source in Section 9. No verdict rests on it.

C15. Anyone working in heat

FieldContent
GradeA for the hypohydration-plus-heat thresholds (NATA position statement with graded recommendations). B+ for the measured fluid requirement in heat. NG for the labour scoring.

MEASURED from IDF-DAR Table 2: highly intense physical labour scores 4 points, moderate scores 2, and fasting hours of 16 or more scores 1. Four points puts a person two-thirds of the way to the high-risk threshold on labour alone.

MEASURED from NATA (DOI 10.4085/1062-6050-52.9.02): thermoregulation compromised above 1% body mass loss (SOR B); exertional heat illness risk rises above 3% (SOR B); cardiovascular stress increased by hypohydration (SOR A); core temperature rises about 0.15°C to 0.20°C per 1% of body weight lost; during exercise, thirst signals at around 2% "appear to be absent or ignored."

MEASURED from the UAE construction cohort (DOI 10.1186/1745-6673-3-21): workers in WBGT above 27.5°C drank 5.44 L per 12-hour shift and stayed euhydrated, with urine specific gravity averaging 1.012.

Unopened source, named, no verdict resting on it: "The health risks of occupational stress in islamic industrial workers during the Ramadan fasting period," PMID 1819340. I have its title and PMID from a search listing only. I did not open it and I do not have its journal, year, volume or pages, so no claim is made from it. A second paper, "Effect of fasting during Ramadan on thermal stress parameters" (Manjunath, Aravindhakshan), was located but every host serving it returned a client error. Both are logged in Section 9.



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.