Dry Fasting — The Record

The record · dawn to dusk

The form with trials, and the findings that cut against it.

The only form of dry fasting with a substantial controlled human literature, recorded in full. 34 sections, 15 of them disconfirming, reported first.

Every finding below was read from a source opened in this run. The 15 disconfirming findings are reported first, before anything supportive, because a record that leads with its best result and files the awkward ones at the bottom is an argument wearing a record’s clothes.

D1. Serious harm in cirrhosis. Hospitalization and bleeding both rose sharply.

  • READ: Prasetya IB, Hasan I, Pribadi RR, Tahapary DL, Nababan SHH, Shatri H, Subekti I, Firdausa S, et al. "Ramadan fasting and adverse outcomes in cirrhosis: primary risk estimates with associated independent predictors." Scientific Reports 2026;16:12612. DOI 10.1038/s41598-026-43599-1.
  • Population and duration: 156 adults with cirrhosis (35.9% decompensated), Jakarta, Indonesia. Pre-Ramadan month (February 2025) versus Ramadan month (March 2025). Prospective cohort, intention to treat.
  • READ, measured outcomes:
  • Hospitalization 12.2% fasting versus 3.2% non-fasting. RR 3.80 (95% CI 1.45 to 9.92), p = 0.003.
  • Upper GI bleeding 10.9% versus 1.3%. RR 8.50 (95% CI 1.99 to 36.17), p < 0.001.
  • Hypoglycaemia on CGM (n = 52) 13.5% versus 0%. RR 15.00 (95% CI 0.88 to 256.0), p = 0.013 Fisher.
  • Decompensated subgroup (n = 56): hospitalization 25.0% versus 8.9%, RR 2.8 (1.08 to 7.25), p = 0.042; UGIB 23.2% versus 3.6%, RR 6.5 (1.54 to 27.49), p = 0.004.
  • Null: Child-Pugh deterioration RR 1.57 (0.83 to 2.96), p = 0.156. Poor quality of life RR 1.40 (0.64 to 3.06), p = 0.395.
  • Exact sentence relied on, READ: "Ramadan fasting in cirrhosis increases the risk of hospitalization, upper gastrointestinal bleeding, and hypoglycaemia."
  • Plasma volume: NOT corrected. NOT measured. READ, the authors' own hypothesis only: "Deprivation of fluid and nutritional intake during Ramadan Fasting may cause dehydration and metabolic stress." MEASURED: hydration status was not measured and no fluid analysis was performed.
  • Confounding: READ, no concurrent non-fasting control group; the comparison is one month against the next, so season and behaviour move with the exposure. Event counts are small and one hypoglycaemia CI runs 0.88 to 256.0. 14 non-completers were counted as having adverse outcomes, which inflates rates.
  • Grade B+ (human data, observational, outcome measured directly).

D2. A 25-hour total fast, no food and no water, raised emergency visits for renal colic.

  • READ: Zilberman DE, Drori T, Shvero A, Mor Y, Winkler HZ, Kleinmann N. "A single day fasting may increase emergency room visits due to renal colic." Scientific Reports 2021;11:6578. DOI 10.1038/s41598-021-86254-7.
  • Protocol: Yom Kippur. READ: "Food and liquid consumption is prohibited during this day for 25 h." This is a true dry fast and the closest controlled-calendar analogue to dawn-to-dusk fluid restriction, at longer duration.
  • Population: 11,717 renal colic ER visits, Sheba Medical Center, Israel, January 2012 to November 2019. Retrospective.
  • READ, measured outcomes: mean daily visits in the 3 days after the fast 6.66 ± 2.49 versus annual mean 4.1 ± 2.27, p < 0.001 (95% CI 1.57 to 3.31). Versus the week before 5.27 ± 2.656, p = 0.032. Versus September 5.06, p = 0.005. Versus October 4.78, p < 0.001.
  • Control comparison, READ: the 3 days after a non-fasting single-day holiday 10 days earlier were not significantly different from September or October (p = 0.207, p = 0.130), and not significantly different from the Yom Kippur window either (p = 0.285). MEASURED: the holiday control therefore did not separate fasting from season.
  • Exact sentence relied on, READ: "A single-day fasting may increase ER visits for RC."
  • MEASURED: the Discussion calls this an "almost twofold increase." 6.66 against 4.1 is 1.62 times. The paper overstates its own number.
  • Plasma volume: NOT corrected. NOT measured. MEASURED: this is an administrative-records study; no blood was drawn. READ: "The mechanism underlying this phenomenon is unknown."
  • Confounding: READ, individual fasting could not be confirmed and a 70% population fasting rate was assumed; no comorbidity records; warm September and October weather is acknowledged as a possible contributor.
  • Grade B+ (human data, outcome measured directly, observational).

D3. The largest kidney-function synthesis finds no consistent effect, and contains no randomized trials.

  • READ: Bello AK et al., corresponding author Pauly RP. "Impact of Ramadan fasting on kidney function and related outcomes in chronic kidney disease and kidney transplant recipients: a systematic review and meta-analysis." BMJ Open 2024 Nov 20;14(11):e085329. PMID 39572100. PMCID PMC11580238. DOI 10.1136/bmjopen-2024-085329.
  • Scope: 32 studies, 2,592 participants. 21 non-dialysis CKD, 11 kidney transplant. READ: all included studies were observational, 29 prospective and 3 retrospective. No RCTs were included. Two were abstracts only.
  • READ, pooled results:
OutcomekMD95% CII²
Creatinine, CKD, vs non-fasting arm50.82 µmol/L−7.22 to 8.8638%
Creatinine, transplant, vs non-fasting arm4−0.17 µmol/L−2.50 to 2.150%
Creatinine, CKD, post−pre1013.84 µmol/L−3.78 to 31.588%
Creatinine, transplant, post−pre6−5.40 µmol/L−10.78 to −0.020% (p = 0.049)
eGFR, CKD, vs non-fasting arm60.23 mL/min/1.73 m²−1.64 to 2.1021%
eGFR, transplant, vs non-fasting arm30.03 mL/min/1.73 m²−1.14 to 1.200%
eGFR, CKD, post−pre91.04 mL/min/1.73 m²−0.26 to 2.356%
  • Exact sentences relied on, READ: "We report inconsistent effects of Ramadan fasting on kidney function in CKD or kidney transplantation." And: "Results should be interpreted with caution due to the significant limitations of the underlying studies."
  • MEASURED: the one nominally significant pooled result, a 5.40 µmol/L fall in transplant creatinine, has an upper CI bound of −0.02, which touches zero, and it is a post-minus-pre comparison with no control arm. It is not evidence of benefit.
  • READ: proteinuria was reported in 13 studies, too few to pool, most non-significant. Quality of life was assessed in none.
  • Plasma volume: NOT corrected, NOT discussed. MEASURED: volume contraction appears only as a medication-related risk in the Discussion. For a review whose primary endpoint is a concentration (creatinine) measured in fluid-restricted people, this is the central unaddressed confound.
  • Grade B+ (meta-analysis of human observational data; not A, because no randomized trials were pooled).

D4. With a non-fasting comparison group, urea rose significantly in CKD fasters.

  • READ: "The effect of Ramadan fasting on renal functions in patients with chronic kidney disease." Turkish Journal of Biochemistry, 2019. DOI 10.1515/tjb-2018-0373. MEASURED: the page served carried no author list, volume, issue or page numbers. I will not derive them.
  • Population: 24 fasting (11 female, mean age 68 ± 13) and 55 non-fasting (29 female, mean age 69 ± 9) adults with stable CKD stage 3 or 4. Turkey. Assessed at baseline, week 1 after Ramadan, and months 3 and 6.
  • READ, measured outcomes (median, IQR):
  • Urea/BUN, fasting: 27 (24 to 35) to 32 (25 to 41) to 35 (25 to 44) to 33 (28 to 49) mg/dL, p = 0.015; 6-month versus baseline p = 0.004. Non-fasting group: 33 to 32 to 33 to 32, p = 0.937.
  • Creatinine, fasting: 1.76 to 1.84 to 2.08 to 1.78 mg/dL, p = 0.070. Non-fasting: p = 0.470. Null.
  • eGFR, fasting: 35 to 34 to 33 to 35 mL/min/1.73 m², p = 0.183. Change in eGFR −2.0 fasting versus −0.64 non-fasting, between-group p = 0.415. Null.
  • Deterioration (eGFR drop ≥25%): 12.5% fasting versus 7.3% non-fasting, p = 0.452. Null.
  • Sodium, potassium, uric acid, proteinuria: all null in the fasting group.
  • MEASURED: urea is the single marker that moved, and urea is the marker most sensitive to both dehydration and protein catabolism. Without a plasma-volume correction the two cannot be separated.
  • Plasma volume: NOT corrected, NOT measured, NOT discussed. MEASURED: haemoglobin was recorded (fasting 12.4, 12.5, 12.6, 12.4 g/dL) but was not used to adjust any renal value. Hydration status was not assessed. Patients were advised to drink 1.5 to 2 L after Iftar; adherence was not measured.
  • MEASURED, internal contradiction: Table 4's multivariate odds ratio for diabetes predicting eGFR drop is 0.080 (0.009 to 0.703), p = 0.023, which points to protection, while the text treats diabetes as a risk marker. Only 3 patients met the outcome, so the estimate is unstable.
  • Grade B+ (human data with a concurrent non-fasting comparison group).

D5. A controlled Iranian cohort found inflammation unchanged and urea higher in fasters.

  • READ: Dastghaib S, Siri M, Rahmani-Kukia N, Heydari ST, Pasalar M, Zamani M, Mokaram P, Bagheri-Lankarani K. "Effect of 30-day Ramadan fasting on autophagy pathway and metabolic health outcome in healthy individuals." Molecular Biology Research Communications 2025;14(2):115-127. DOI 10.22099/mbrc.2024.50105.1978. MEASURED: no PMID or PMCID appears in the text.
  • Design: controlled cohort, fasting group n = 24 versus non-fasting group n = 26, healthy adults 20 to 78 years, Shiraz, Iran. Blood drawn one day after Ramadan.
  • READ, measured outcomes:
  • IL-6: 3.5 ± 2.6 fasting versus 3.8 ± 2.3 pg/mL non-fasting, p = 0.29. Null.
  • hs-CRP: 0.8 ± 0.5 versus 1.1 ± 1.4, p = 0.40. Null.
  • BUN: 13.91 ± 3.62 versus 11.73 ± 3.86 mg/dL, p = 0.045. Higher in fasters. MEASURED: this is the only biochemical difference reported between the groups.
  • Glucose, triglycerides, cholesterol, HDL, LDL, creatinine, liver enzymes, electrolytes: measured, no significant differences reported.
  • CBC including haemoglobin, RBC, WBC, platelets: no significant changes, p > 0.05.
  • PBMC mRNA: BECN1 up 2.30-fold (p < 0.001); LC3II down about 16% (p < 0.001); P62 down about 15% (p < 0.001), from n = 3 assays.
  • Serum Beclin-1 and LC3 protein: p = 0.69 and p = 0.91. Null.
  • Plasma volume: NOT corrected, NOT measured, NOT discussed. MEASURED: the authors attribute the BUN rise to protein catabolism and do not consider daytime fluid restriction as an alternative cause.
  • MEASURED, internal contradictions: the Results text claims a "significant increase" in IL-6 and hs-CRP in the non-fasting group while every reported p-value is non-significant; the text describes an "upward trend" in Beclin-1 and LC3 at p = 0.69 and p = 0.91; the Discussion cites TFEB and AMPK/mTOR changes as "our findings" although neither was measured.
  • Confounding, READ: no within-subject baseline (single post-Ramadan draw); diet, sleep and physical activity not controlled; unequal sex distribution; convenience recruitment by social media; sample size calculated on hs-CRP, not on autophagy markers.
  • Grade B+ (human data with a concurrent non-fasting control group). The null inflammation result carries real weight because it is one of the very few concurrently controlled designs in the literature.

D6. C-reactive protein positivity doubled in post-menopausal women.

  • READ: Al Zunaidy NA, Al-Khalifa AS, Alhussain MH, Althwab SA, Mohammed MA, Faris ME. "The effect of Ramadan intermittent fasting on anthropometric, hormonal, metabolic, inflammatory, and oxidative stress markers in pre- and post-menopausal women: a prospective cohort of Saudi women." Frontiers in Nutrition 2024;11:1437169. DOI 10.3389/fnut.2024.1437169.
  • Population: 62 healthy Saudi women, 31 pre-menopausal (21 to 42 y), 31 post-menopausal (43 to 68 y). Qassim, Saudi Arabia, April to May 2022. Open-label pre-post, no control group.
  • READ, measured outcomes:
  • Qualitative CRP positive, post-menopausal: 5 (16.13%) rising to 10 (32.26%), p = 0.015. Pre-menopausal: 2 (6.45%) to 1 (3.23%), p = 0.071.
  • TNF-α fell: pre-menopausal 194.45 to 183.30 ng/mL, post-menopausal 123.43 to 117.19, time p = 0.021, group × time p = 0.048.
  • IL-10 rose: 30.31 to 39.73 pg/mL pre-menopausal, 21.46 to 21.56 post-menopausal, time p = 0.025.
  • IGF-1 null, p = 0.406. Total antioxidant capacity null, p = 0.482. MDA null, p = 0.147.
  • Systolic BP: not significant. Diastolic BP time p < 0.001, driven by post-menopausal 81.41 to 71.00 mmHg.
  • Fasting glucose null, p = 0.186.
  • Total body water, fat-free mass, hip circumference, waist-hip ratio: not significant.
  • MEASURED: CRP moved in the harmful direction in the older subgroup while TNF-α moved in the favourable direction in the same people. The inflammatory picture is not internally consistent, and the brand's customers are mostly in the older bracket.
  • Plasma volume: NOT corrected, NOT measured. MEASURED: the Discussion speculates that Ramadan fasting could affect blood sample concentration but reports no hemoconcentration data and applies no adjustment.
  • MEASURED, reporting defects: HDL-C is printed as rising from 1.62 to 4.75 mmol/L in pre-menopausal women, which is not physiologically plausible and is likely a typographical error; average fasting duration is printed as "350 h"; the text and abstract disagree on whether the second measurement was at the third or fourth week.
  • Grade B+ for the human data, C for the CRP result specifically (single isolated subgroup result on a qualitative, dichotomized marker in a paper with multiple printing errors).

D7. With a non-fasting control group in Germany, IL-1 beta rose in the fasting group.

  • READ: Khoshandam Ghashang S, Raha S, Hamdan I, Gutenbrunner C, Nugraha B. "A prospective controlled study on Ramadan fasting in the healthy young males in summer in Germany: effect on cytokines." Journal of Complementary and Integrative Medicine 2021;18(2):425-431. DOI 10.1515/jcim-2019-0209. MEASURED: no PMID shown on the page served.
  • Design: 50 healthy adult males, Germany, divided into a fasting group and a non-fasting group. Fasting about 18 to 19 hours per day, the longest daily fluid-restricted window in this record. Fasting group sampled at four time points, non-fasting group at two.
  • READ, measured outcomes: circulating IL-1β increased significantly from T2 to T4 and from T3 to T4 in the fasting group. IL-6, IL-8, IL-10, IL-12 and TNF-α were not reported as changed over time, and fasting versus non-fasting differences were insignificant at T1 and T3. TNF-α correlated with body weight, skeletal muscle mass and fat-free mass at T3; IL-12 correlated with skeletal muscle mass, fat-free mass and body water mass at T4.
  • MEASURED: IL-1β is pro-inflammatory. A significant rise in it, in the design with a real control group and the longest fast, runs against the anti-inflammatory claim.
  • Plasma volume: NOT mentioned. MEASURED: the text served gives no statement on plasma volume, hemoconcentration or hydration.
  • MEASURED: full text is paywalled. Only the abstract-level content was retrievable, so no numeric cytokine values, no methods detail and no limitations section could be read. Per the rule that a verdict may not rest on an unopened source, this is logged as a directional finding with its identifier, not as a quantified result.
  • Grade B+ (human data, concurrent control group), with the quantification open.

D8. In the flagship dawn-to-dusk proteome study, inflammatory markers did not move and homocysteine rose.

  • READ: Mindikoglu AL, Abdulsada MM, Jain A, Jalal PK, Devaraj S, Wilhelm ZR, Opekun AR, Jung SY. "Intermittent fasting from dawn to sunset for four consecutive weeks induces anticancer serum proteome response and improves metabolic syndrome." Scientific Reports 2020 Oct 27;10:18341. PMID 33110154. PMCID PMC7592042. DOI 10.1038/s41598-020-73767-w.
  • READ, the disconfirming half of this paper's own Table 1:
  • hs-CRP: not significant. IL-6: not significant. Both at end of week 4 and one week after.
  • Homocysteine ROSE +1.97 µmol/L, p = 0.0004 at end of week 4, then fell 0.63 below baseline at one week post, p = 0.029.
  • Insulin −2.0 µU/mL, p = 0.164. Null.
  • Glucose −6 mg/dL, p = 0.176. Null.
  • HOMA-IR −0.7, p = 0.130 at week 4. Null. It reached p = 0.035 only one week after the fast ended.
  • Triglycerides −16 mg/dL, p = 0.413. Null. One week post +5 mg/dL, p = 0.699.
  • HDL, LDL, total cholesterol: not significant.
  • Adiponectin and leptin: not significant.
  • MEASURED: elevated homocysteine is a cardiovascular risk marker. The paper reports a highly significant rise in it during the fast and, MEASURED, does not discuss the finding anywhere in the text I extracted.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. MEASURED, by direct extraction of the full text including the Discussion and "Strengths and limitations": the words plasma volume, hemoconcentration, hydration, dehydration and fluid shift do not appear anywhere in the paper. This is the single most important measurement in this strand. A 14-day-plus daily dry fast was run, blood concentrations were reported as biology, and the fluid question was never raised.
  • Confounding, READ: "The lack of caloric measurement by dietary assessment is one of the limitations of our study." MEASURED: sleep is not mentioned. Physical activity is not mentioned. There is no control group and the absence of one is not listed as a limitation. 9 of 14 were on antihypertensives, 7 on antidiabetics, 6 on statins; medication changes are not reported. n = 14, with 11 aged 60 or over.
  • Grade B+ for the human measurements; the null inflammatory results are the honest headline.

D9. Fasting glucose ROSE in the metabolomics cohort.

  • READ: Madkour M, Giddey AD, Soares NC, Semreen MH, Bustanji Y, Zeb F, Halwani R, Faris ME. "Ramadan diurnal intermittent fasting is associated with significant plasma metabolomics changes in subjects with overweight and obesity: A prospective cohort study." Frontiers in Nutrition 2022;9:1008730. DOI 10.3389/fnut.2022.1008730. MEASURED: PMCID PMC9868699 observed in the search result URL.
  • Population: n = 25 (7 female, 18 male), mean age 39.5 ± 10.0, overweight or obese, Sharjah UAE, Ramadan 2016, about 15 h/day. Pre-post, each subject own control.
  • READ, measured outcomes:
  • Fasting glucose 99 to 108.4 mg/dL, p = 0.036. A significant RISE.
  • BMI 29.09 to 28.53, p = 0.0001.
  • Waist 96.34 to 95.84 cm, p = 0.06. Null.
  • HDL-C 44.36 to 58.97 mg/dL, p = 0.0001. LDL-C 105.7 to 92.63, p = 0.0013.
  • Total cholesterol and triglycerides: not significant.
  • Cortisol decreased, log2 FC −1.143.
  • 27 of 89 metabolites differed at p < 0.05; 23 down, 4 up.
  • Ketone bodies: NOT measured. MEASURED: beta-hydroxybutyrate, acetoacetate and acetone do not appear in the results. The Discussion links traumatic acid to ketogenesis with no ketone data behind it.
  • Insulin, HOMA-IR, adiponectin, leptin and PCSK9 are listed in the Methods as measured; MEASURED, no results are reported for any of them.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. MEASURED: the paper notes daytime water abstention but never assesses hydration or discusses fluid-shift effects on metabolite concentrations. For an untargeted plasma metabolomics study this is a first-order defect, because every reported value is a concentration.
  • MEASURED, method violation inside the paper: the Methods define significance as adjusted p < 0.05 with |log2 FC| > 1; the Results use unadjusted p < 0.05 and most reported changes fall below |log2 FC| = 1. Table 3 ("decreased") contains positive fold changes (isovalerylcarnitine +1.431, benzoic acid +1.28) and Table 4 ("increased") contains a negative one.
  • Grade B+ for the human measurements. The glucose result is disconfirming and must not be omitted when glucose benefit is claimed.

D10. Weight and fat loss are real but small, partly lean tissue, and largely temporary.

  • READ: Fernando HA, Zibellini J, Harris RA, Seimon RV, Sainsbury A. "Effect of Ramadan Fasting on Weight and Body Composition in Healthy Non-Athlete Adults: A Systematic Review and Meta-Analysis." Nutrients 2019 Feb 24;11(2):478. PMID 30813495. PMCID PMC6412279. DOI 10.3390/nu11020478.
  • Scope: 70 publications, 90 comparison groups, 2,947 participants.
  • READ, pooled results:
  • Body weight −1.34 kg (95% CI −1.61 to −1.07), p < 0.001. At 2 to 5 weeks follow-up −0.59 kg (−0.99 to −0.20), p = 0.003.
  • Fat mass −0.98 kg (−1.05 to −0.92), p < 0.001. No significant change at follow-up.
  • Fat-free mass −0.66 kg (−0.75 to −0.57), p < 0.001. No significant change at follow-up. READ: the fat-free mass loss "was approximately 30% less than the loss of absolute fat mass."
  • Body fat percentage −1.07% (−1.55 to −0.59), p < 0.001.
  • Exact sentences relied on, READ: "there are consistent, albeit transient, reductions in weight and fat mass with the Ramadan fast." And: "All body composition parameters had returned to baseline values by this time."
  • MEASURED, derived arithmetic, flagged as derivation not as the authors' claim: 1.34 kg lost against 0.59 kg still absent at follow-up implies roughly 0.75 kg regained, about 56% of the loss.
  • MEASURED, the authors' own hydration argument: body water is invoked to explain why higher-BMI people lose more, on the reasoning that greater glycogen stores carry more water and "would be expected to lose more fluid in response to fasting." So the authors accept that part of the measured weight loss is water, and do not partition it. Plasma volume is not mentioned.
  • Confounding, READ: physical activity and energy intake not controlled; intake self-reported in all relevant studies; participants self-selected fasters, not randomized; body composition mostly by skinfold or bioimpedance, which the authors call relatively inaccurate, with only 3 studies using hydrodensitometry. Fasting duration did not predict weight change.
  • MEASURED, internal inconsistency: the Results say body fat percentage showed no significant change at follow-up, and the same paper later reports −0.94% (−1.73 to −0.15), p = 0.020, at follow-up.
  • Grade B+ (meta-analysis of predominantly observational human data).

D11. Confirmation that the fast measurably dehydrates people, from the authors who measured it.

  • READ: Trabelsi K, Rebai H, el-Abed K, Stannard SR, Khannous H, Masmoudi L, Sahnoun Z, Hakim A, Fellman N, Tabka Z. "Effect of Ramadan Fasting on Body Water Status Markers after a Rugby Sevens Match." Asian Journal of Sports Medicine 2011 Sep;2(3):186-194. PMID 22375238. PMCID PMC3289207. DOI 10.5812/asjsm.34748.
  • Population: 12 healthy male recreational rugby players, Sfax, Tunisia, Ramadan 2008. Repeated measures, no control group.
  • READ, measured resting values (pre-Ramadan / day 7 / day 30):
  • Plasma osmolarity 284.7 ± 3.2 / 291.3 ± 3.6 / 292.8 ± 2.1 mOsm/L. +2.3% at day 7 (p < 0.001) and +2.8% at day 30 (p < 0.001).
  • Haematocrit 43.5 ± 3.3 / 45.2 ± 2.2 / 45.4 ± 1.9 %. +3.9% day 7 (p = 0.04), +4.4% day 30 (p = 0.03).
  • Haemoglobin 14.6 ± 0.9 / 15.2 ± 1.0 / 15.3 ± 0.7 g/dL. +3.4% day 30 (p = 0.01).
  • Resting body mass fell 2.8% from pre-Ramadan to day 30, p < 0.001.
  • Total body water by impedance 51.3 / 52.8 / 51.5 L, not significantly different.
  • Exact sentence relied on, READ: "The present results show that Ramadan fasting induces dehydration at basal conditions."
  • Plasma volume: CALCULATED, NOT used to correct anything. READ: blood and plasma volume changes were determined by the Dill and Costill equation. MEASURED: the plasma volume figures reported are exercise-induced match deltas (−7.1%, −4.4%, −4.1%), not the Ramadan resting shift, and no measured marker was adjusted. MEASURED: urine osmolality, urine specific gravity and urine colour were not measured, and plasma sodium was used only inside the calculated osmolarity and never reported alone.
  • MEASURED: this is the key cross-reference for the whole strand. Haematocrit +4.4% and haemoglobin +3.4% at rest is hemoconcentration of roughly that order. Any uncorrected blood marker measured in the same state inherits an inflation of similar magnitude.
  • Confounding, READ: no control group; energy intake fell 15.4% and total water intake fell from 4.0 to 3.4 L/day, both self-recorded; maximum temperature was lower and humidity higher during Ramadan; the pre-Ramadan match started at 12:00 and the Ramadan matches at 16:00 after at least 12 h without food or water. MEASURED: the authors themselves suggest electrolyte shifts may have distorted the bioimpedance total body water result, so the unchanged total body water finding is unreliable.
  • MEASURED, table defect: the day-30 post-match body mass prints as "3.5 (14.9)", which contradicts the stated delta.
  • Grade B+ (human data, directly measured hydration markers).

D12. The Ramadan renal colic case-control was null during the fast, with a delayed rise after.

  • READ: Lidawi G, Majdoub M, Asali M, Avidor Y, Rub R. "Ramadan fasting and rates of admissions for renal colic due to urolithiasis: a retrospective case-control study." Journal of International Medical Research 2024 Jul 28;52(7):03000605241261916. PMID 39068526. PMCID PMC11287717. DOI 10.1177/03000605241261916.
  • Population: 2,781 adults with renal colic from urolithiasis, 1,014 fasting Muslims and 1,767 non-fasting non-Muslims, Hillel Yaffe Medical Centre, Israel, 2009 to 2019.
  • READ, measured outcomes: fasting Muslims, pre-Ramadan 305 (34.7%), Ramadan 318 (32.5%), post-Ramadan 391 (42.3%). Pre versus Ramadan p = 0.53. Null. Pre versus post p < 0.001 and Ramadan versus post p < 0.001.
  • Exact sentences relied on, READ: "did not lead to an increased incidence of RC admissions during the month of Ramadan." And: "There were increased RC admissions during the post-Ramadan month, which might indicate a delayed effect of fasting on RC."
  • MEASURED, contradiction: the abstract says there was no significant increase during Ramadan in either group, while Table 2 reports p = 0.003 for pre versus Ramadan in the non-fasting non-Muslim group (659 versus 575). MEASURED: the Methods claim matching for age and sex while Table 1 shows means of 50.6 against 54.3 years and differing sex proportions. No confidence intervals are reported for any admission comparison and seasonality is not modelled.
  • Plasma volume: not applicable, administrative records. MEASURED: no hydration measurement.
  • Grade C for the post-Ramadan rise (single unadjusted monthly count comparison, single centre), B+ for the null during-Ramadan result.

D13. The apex ketone meta-analysis explicitly excluded this fasting model.

  • READ: Ulupınar S, Terzi M, Malkoç N, Çamli A, İnce İ, Gençoğlu C, Özbay S. "Duration-Dependent Ketone Body Responses to Fasting: A Systematic Review, Meta-Analysis, and Meta-Regression." 2026. DOI 10.3390/ijms27219570. MEASURED: PMCID PMC13762781 observed in the search result URL; the journal name is not printed in the text served and the article ID ijms-27-09570 implies International Journal of Molecular Sciences volume 27. I will not state the journal as fact on an inference.
  • Scope: 30 studies, 44 effect sizes. Per-study n ranged 6 to 57. Total participants not reported.
  • READ, pooled results: fasting increased BHB g = 2.794, total plasma ketones g = 3.260, acetoacetate g = 1.731. Duration ≤3 days g = 1.768 (1.330 to 2.206); >3 days g = 3.780 (3.013 to 4.547). Meta-regression β = 0.300 standardized units per additional day (0.130 to 0.470).
  • The disqualifying facts, READ: eligibility required a minimum fasting duration of 24 hours, and religious fasting such as Ramadan and intermittent regimens such as time-restricted and alternate-day fasting were excluded. Water, mineralized water and non-caloric beverages were permitted in the included protocols.
  • MEASURED, therefore: this meta-analysis says nothing about dawn-to-dusk fasting. It reports no concentrations in mmol/L, only standardized effect sizes, and it has no estimate at 12, 14, 16 or 18 hours, which is the entire dawn-to-dusk range. Plasma volume is not mentioned; hydration appears only as a source of between-study variability.
  • READ: "Egger's regression test indicated significant funnel-plot asymmetry consistent with small-study effects."
  • Grade A for prolonged, water-permitted fasting of 24 hours or more. Grade: NOT APPLICABLE to dawn-to-dusk or fluid-restricted fasting.

D14. In diabetic kidney disease, the albumin-creatinine ratio change was null.

  • READ: Salwani D, Syukri M, Balqis U, Suhanda R. "Perbandingan Rasio Albumin Kreatinin Urin Sebelum Dan Setelah Puasa Pada Penyakit Ginjal Diabetik" (Comparison of Urinary Albumin Creatinine Ratio Before and After Fasting in Diabetic Kidney Disease). Journal of Medical Science (Jurnal Ilmu Medis RSUD dr. Zainoel Abidin) 2025;6(2):136-142. DOI 10.55572/jms.v6i2.246. Language: Indonesian, with English title and abstract.
  • Population: 37 adults with diabetic kidney disease and proteinuria ≥ +1, Banda Aceh, Indonesia. Fasting at least 7 consecutive days, 13 to 14 h/day. Prospective cohort.
  • READ, measured outcomes: urine ACR before fasting median 1,359 (range 10.57 to 12,442); after 7 days median 1,015 (7.89 to 9,764). Wilcoxon p = 0.140. Null. Baseline creatinine 1.44 ± 0.95 mg/dL, urea 42.24 ± 21.69 mg/dL, HbA1c 10.28 ± 2.82. MEASURED: before-versus-after serum creatinine and eGFR were not reported, only baseline values.
  • Exact sentences relied on, READ, Indonesian with literal English: "terdapat penurunan rasio albumin kreatinin namun tidak berbeda secara statistik" = "there is a decrease in the albumin creatinine ratio but it is not statistically different." And "tidak menunjukkan perbedaan yang bermakna secara statistik dengan nilai P 0,140" = "did not show a statistically significant difference with a P value of 0.140."
  • MEASURED, overreach: the Discussion concludes "puasa Ramadhan tidak memberikan dampak yang buruk terhadap responden dengan penyakit ginjal diabetik" ("Ramadan fasting does not have a harmful impact on respondents with diabetic kidney disease") from a non-significant result. A null does not establish safety.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. MEASURED: ACR is a ratio of two urine concentrations, which partly self-corrects for urine concentration, and this is the only outcome in the whole strand with any intrinsic protection against the hydration artifact. The authors do not note that.
  • MEASURED, defects: the study period is March to May 2024 in the abstract and March to July 2024 in the methods; 3 patients are said not to have returned yet n = 37 is analysed; ACR units are printed as mg in tables and mg/g in text.
  • Grade B+ (human data, observational, endpoint measured directly).

D15. Hormonal and biochemical pooling: minimal change, and a conclusion that outruns the data.

  • READ: Poursalehian M, Mohseni S, Shadman Z, Mohajeri-Tehrani M, Atlasi R, Khoshniat Nikoo M, Qorbani M, Larijani B. "Impact of Ramadan fasting on serum levels of major endocrinology hormonal and biochemical parameters in healthy non-athlete adults: A systematic review and meta-analyses." PLOS ONE 2024 May 23;19(5):e0299695. DOI 10.1371/journal.pone.0299695. MEASURED: no PMID shown on the page served.
  • Scope: 35 original articles, 1,107 participants (783 men, 271 women, 53 sex unknown), 1982 to 2021. READ: all used a pre-post design and "none of the studies included a control group of non-fasted participants."
  • READ, the only pooled estimate printed with a CI: morning cortisol, Hedges' g = −2.14 (95% CI −4.54 to 0.27), p = 0.08, described by the authors as "a large insignificant decrease."
  • READ: "No significant relationship was found between pre- and post-Ramadan hormonal levels of T3, T4, TSH." And, contradicting the abstract: "A significant reduction was, however, observed regarding free T4 values."
  • MEASURED: T3, T4, TSH, FT3, FT4, testosterone, LH, FSH, prolactin, PTH, calcium and phosphorus pooled values appear only in figures and could not be read. Insulin, glucose, HOMA-IR, leptin, adiponectin, creatinine, urea, uric acid, sodium, osmolality, lipids and CRP were not pooled at all; glucometabolic markers were excluded on the grounds that Faris et al. had already reviewed them.
  • MEASURED, reporting defects: the abstract says effects were expressed as odds ratios while the results use Hedges' g; the abstract's "no significant relationship" is printed with "P-value<0.05"; the cortisol conclusion about circadian disruption rests on a non-significant pooled estimate whose CI crosses zero.
  • READ, the authors' conclusion: "Ramadan Intermittent Fasting results in minimal hormonal changes and is a safe practice for healthy individuals." MEASURED: a corpus with no control group anywhere in it cannot establish safety; it can only fail to detect harm.
  • Plasma volume: NOT discussed, NOT corrected.
  • Grade B+ (meta-analysis of uncontrolled human pre-post data).

2.1 The count

MEASURED, by direct extraction of every source opened in this sweep: of 20 primary studies and syntheses examined, ZERO corrected any reported biomarker for plasma volume change. Three calculated a plasma volume change and then did not use it. Seventeen did not measure or mention it at all.

SourceMeasured hemoconcentration?Corrected a biomarker?
Khalfoun 2024, Medicina, DOI 10.3390/medicina60071143YES, Dill and CostillNO
Trabelsi 2011, PMID 22375238YES, Dill and CostillNO
Hammouda 2013, DOI 10.1371/journal.pone.0079873YES, Dill and Costill cited in MethodsNO, and the values were never reported
Mindikoglu 2020, PMID 33110154NO, term absent from the paperNO
Mindikoglu 2022, PMID 36506940NO, term absent from the paperNO
Mindikoglu 2025, DOI 10.1186/s12014-025-09547-3NONO
Madkour 2023, DOI 10.1038/s41598-023-43862-9NONO
Madkour 2022, DOI 10.3389/fnut.2022.1008730NONO
Al-Jafar 2024, PMID 38301826Total body water by impedance, used as a covariate onlyNO
Almeneessier 2019, DOI 10.1371/journal.pone.0226034NONO
Dastghaib 2025, DOI 10.22099/mbrc.2024.50105.1978NONO
Al Zunaidy 2024, DOI 10.3389/fnut.2024.1437169NO, speculated in Discussion onlyNO
Ghashang 2021, DOI 10.1515/jcim-2019-0209NONO
Darzabi & Hejazi 2020, DOI 10.5812/jjcdc.100079NONO
Ziaee 2007, Iran J Endocrinol Metab 9(1):47-53NONO
Turkish CKD 2019, DOI 10.1515/tjb-2018-0373Haemoglobin recorded, not appliedNO
Larijani 2003, Med J Malaysia 58(5):678-680NONO
Salwani 2025, DOI 10.55572/jms.v6i2.246NONO (the ACR ratio partly self-corrects)
Bello 2024 meta-analysis, PMID 39572100NONO
Fernando 2019 meta-analysis, PMID 30813495Body water invoked verballyNO
Faris 2025 umbrella review, DOI 10.1016/j.clnesp.2025.03.006NONO

2.2 How large the artifact is

  • READ, Trabelsi 2011: resting haematocrit rose 4.4% and haemoglobin 3.4% by day 30 of Ramadan, with plasma osmolarity up 2.8% (all p < 0.05).
  • READ, Khalfoun J, Zouhal H, Triki R, Jribi W, Saeidi A, Almaqhawi A, Clark CCT, Laher I, Ben Abderrahman A. "Ramadan Intermittent Fasting and Plasma Volume Variations in Individuals with Different Body Weights." 2024. DOI 10.3390/medicina60071143. MEASURED: PMCID PMC11278820 observed in the search result URL; the journal is not printed in the text served, and the DOI prefix and figure labels indicate Medicina. I will not assert the journal on an inference. 32 men aged 20 to 40, Tunisia, Ramadan 2019, 15 to 16 h/day, four groups of 8 by BMI.
  • Plasma volume change T0 to T1 by group: normal weight +0.3 ± 1.9%, overweight −0.1 ± 2.4%, obese −1.9 ± 1.0%, severe obesity −2.8 ± 3.4%. T1 to T2: obese −4.1 ± 1.7%, severe obesity −4.2 ± 4.2%.
  • Haemoglobin, severe obesity: 17.2 to 17.5 to 17.7 to 17.4 g/100 mL. Haematocrit, obese: 49.3 to 49.8 to 50.4 to 49.8%.
  • MEASURED, dose-response: plasma volume contraction scaled with body mass. Normal-weight men barely moved; men with severe obesity lost about 4% of plasma volume. The brand's customer base skews toward higher body mass, which is the group where the artifact is largest.
  • MEASURED, defect: the abstract and Results describe "increases in ΔPV" in the obese and severely obese groups while Table 4 reports negative values, and the Discussion and Conclusions describe decreases. The paper contradicts itself on the direction of its own primary outcome.
  • MEASURED, the consequence, stated as a derivation and not as a measured result: a 2 to 4% plasma volume contraction inflates any uncorrected plasma concentration by roughly 2 to 4%. That is small against a 74-fold proteomic change. It is not small against an 8 mmHg blood pressure fall's biochemical companions, a −16 mg/dL triglyceride change, a 0.82 µmol/L pooled creatinine difference, a 13.84 µmol/L pooled creatinine difference whose CI is −3.78 to 31.5, or a 2.18 mg/dL BUN difference. Several of the pooled renal estimates in this strand are of the same order as the uncorrected artifact.

2.3 One result that hemoconcentration alone cannot explain

  • READ, Ziaee V, Yousefi R, Ahmadinejad Z, Shaikh H, Rezaei M, Behjati MJ. "The effect of Ramadan fasting on serum osmolarity, some electrolytes and hematological parameters." Iranian Journal of Endocrinology and Metabolism 2007;9(1):47-53. Language: English with a Persian abstract, Iranian journal. MEASURED: no DOI or PMID stated.
  • Population: 80 medical students (41 male, 39 female), mean age 22.7.
  • READ, direction only, MEASURED that the page served carried no numeric values or p-values: serum osmolarity increased significantly in both sexes; sodium increased significantly in both sexes; RBC count increased significantly in both; potassium decreased in males; glucose decreased in both; weight decreased in both; haemoglobin and haematocrit DECREASED in males. READ: "We found no significant changes in levels of urea and creatinin in serum."
  • Exact sentences relied on, READ: "Ramadan fasting leads to increased serum osmolarity, RBC count and some Hgb indices in both genders." And: "Hgb, Htc decreased after Ramadan Fasting in male subjects."
  • MEASURED, why this matters: osmolarity and sodium rising while haemoglobin and haematocrit fall is not the signature of simple hemoconcentration, which would raise all four. This result argues that something beyond fluid contraction is also occurring, and it is the one data point in this record that partially rescues the sodium and osmolarity findings from being dismissed as pure artifact.
  • MEASURED: numeric values are in a PDF I could not open on an available channel. Logged with its identifier as a directional finding, not a quantified one.
  • Grade C (single study, direction only, numbers unread).

2.4 A controlled design that put the artifact in front of the reader and still did not correct

  • READ, Darzabi T, Hejazi K. "The Effects of Ramadan Fasting on Electrolytes Index, Serum Osmolarity and Body Composition in Fasting and Non-Fasting Students: A Quasi-Experimental Study." Jundishapur Journal of Chronic Disease Care 2020;9(2):e100079. DOI 10.5812/jjcdc.100079. MEASURED: no PMID stated.
  • Design: quasi-experimental with a non-fasting control group. Fasting n = 15, non-fasting n = 14, healthy inactive male students, mean age 21.2 ± 1.7, BMI 25.0 ± 2.1. One Ramadan, May to June.
  • READ, measured outcomes. Significant within the fasting group and null within the non-fasting group: serum osmolarity, sodium, potassium, urea, uric acid, creatinine, glucose, weight, BMI, waist, hip and waist-hip ratio. Albumin not significant in either group.
  • Significant between-group interaction: osmolarity p = 0.01, sodium p = 0.03, urea p = 0.04, uric acid p = 0.005.
  • Not significant between groups: weight, BMI, waist, hip, waist-hip ratio, glucose, creatinine, potassium.
  • MEASURED, the reading that matters: the only markers that separated fasters from non-fasters are the four that fluid contraction would move. Weight, glucose and creatinine did not separate. That pattern is what a hemoconcentration artifact looks like, and the paper does not test for it.
  • Plasma volume: NOT corrected, NOT discussed. MEASURED: osmolarity was calculated from sodium, urea and glucose rather than measured directly, and it is labelled mg/dL, which is not a unit of osmolality.
  • MEASURED, serious defect: the table headers read "Post-Test" then "Pre-Test", and under those labels osmolarity, sodium, urea, uric acid, creatinine and weight move in the opposite direction from what the text reports. The headers are probably transposed; the authors do not say so. Direction is therefore not safe to quote from this paper.
  • MEASURED, overreach: the conclusion "fasting in Ramadan does not affect health if one follows certain health recommendations" is not tested by the design, which measured no health outcome and no fluid intake.
  • Grade B+ for the controlled design, C for any individual direction given the header defect.

3.1 Ketone bodies, and how fast they rise

FINDING, MEASURED: there is no witness-level human measurement of beta-hydroxybutyrate during a dawn-to-dusk fluid-restricted fast in anything I could open. The claim that ketones rise fast in dry fasting is NOT supported by measured human data in this literature.

What exists:

  • READ, Ulupınar 2026, DOI 10.3390/ijms27219570: ketones rise with fasting duration, BHB g = 2.794, slope β = 0.300 per day. But this meta-analysis required ≥24 h fasts and explicitly excluded Ramadan and all intermittent regimens, and the included protocols permitted water. Grade A for prolonged water-permitted fasting. Not applicable here.
  • READ, Al-Jafar R, Climaco Pinto R, Elliott P, Tsilidis KK, Dehghan A. "Metabolomics of Ramadan fasting and associated risk of chronic diseases." American Journal of Clinical Nutrition 2024;119(4):1007-1014. PMID 38301826. DOI 10.1016/j.ajcnut.2024.01.019.
  • n = 72 paired samples of 140 recruited, London, 2019, adults intending to fast >20 days. NMR metabolomics.
  • Acetone INCREASED, β = 0.10 (95% CI 0.04 to 0.16), p = 0.001, FDR 0.019. Acetate DECREASED, β = 0.22 (0.27 to 0.18), p < 0.001.
  • MEASURED, the critical limitation: READ, visit 2 blood was drawn "within 8 to 12 d after the end of Ramadan." The measurement is not during the fast, and not even during Ramadan. READ: fasting hours per day are not reported, and time of day and fasting state at the draw are not reported.
  • MEASURED: the Nightingale platform measures four ketone bodies. Results for 3-hydroxybutyrate and acetoacetate are not reported. Their changes are unknown.
  • Exact sentence relied on, READ: "Ramadan fasting is associated with short-term favorable changes in the metabolic profile."
  • Plasma volume: NOT corrected, NOT measured, NOT discussed. Total body water was measured by bioimpedance and used only as a covariate in a secondary model. Weight fell 1.7 kg (95% CI 1.2 to 1.9) and body fat 1.1%.
  • Confounding, READ, adjusted for: age, sex, mosque, day of second measurement, and changes in waist, fat-free mass and BMI, or alternatively total body water and fat percentage. READ, could not be isolated: smoking (too few smokers to adjust), diet, energy intake, physical activity, sleep, medications, chronic disease status, ethnicity, socioeconomic status, blood sampling conditions, and 68 non-returners who differed by age.
  • MEASURED: the disease risk scores are predictions from metabolite profiles, not measured disease. READ, the authors' own statement: the study "does not provide evidence" of lower disease risk.
  • Grade B+ for the acetone change. C for any during-fast ketone inference drawn from it, because it was not measured during the fast.
  • MEASURED, not measured anywhere: Madkour 2022 (DOI 10.3389/fnut.2022.1008730) did not measure ketones. Dastghaib 2025 did not measure ketones. Mathew/Suhre (J Transl Med 2014;12:161, DOI 10.1186/1479-5876-12-161), an 11-man dawn-to-dusk metabolomics pilot run at the Islamic Centre Munich in 2009 with 13 to 18 h daily fasts, measured about 200 metabolites and does not mention 3-hydroxybutyrate, acetoacetate or acetone anywhere.
  • MEASURED, a source I could not open: "Intermittent fasting during Ramadan increases the body's ability to produce ketone bodies and promote gut Bacteroides and Firmicutes" exists only on ResearchGate on the channels available to me. No verdict rests on it and no claim is taken from it.

Receipt for this outcome: covered, and the honest answer is a gap. Dawn-to-dusk ketone kinetics in humans are unmeasured in the retrievable literature.

3.2 Insulin and glucose

  • READ, Faris MAI, Abdelrahim DN, El Herrag SE, Khaled MB, Abu Shihab K, AlKurd R, Madkour M. "Cardiometabolic and obesity risk outcomes of dawn-to-dusk, dry intermittent fasting: Insights from an umbrella review." Clinical Nutrition ESPEN 2025;67:127-145. DOI 10.1016/j.clnesp.2025.03.006. 20 meta-analyses pooled. This is the apex synthesis for this exact fasting model.
  • Fasting blood glucose, general population, n = 16,106: Hedges' g = −0.23 (95% CI −0.33 to −0.13), p < 0.01.
  • MEASURED: no pooled estimate is reported for insulin, HOMA-IR or HbA1c in the content served, although the Methods list them. Their pooled values are unknown to me.
  • Plasma volume: NOT discussed. MEASURED: the review defines dry fasting as abstaining from all food and drink including water, and then analyses concentration-based outcomes without addressing fluid shift.
  • Grade B+ (umbrella review of meta-analyses of observational human data; not A, because no randomized trials of the fast exist to pool).
  • READ, Larijani B, Zahedi F, Sanjari M, Amini MR, Jalili RB, Adibi H, Vassigh AR. "The Effect of Ramadan Fasting on Fasting Serum Glucose in Healthy Adults." Med J Malaysia 2003;58(5):678-680. Malaysian journal, Iranian cohort, published in English.
  • n = 115 (67 male, 48 female), ages 15 to 45, Rey, Iran, Ramadan 1423, about 11.5 ± 0.5 h/day, all fasted at least 25 days.
  • Fasting serum glucose: baseline 88.4 ± 9.0 mg/dL, day 14 75.4 ± 15.3, day 28 62.9 ± 7.7, p < 0.0001. Females 89.7 to 67.5 to 65.7. Males 87.5 to 81.1 to 60.8. Calorie intake 1,470 ± 794 to 1,191 ± 521 kcal.
  • Exact sentences relied on, READ: "Our study shows a significant decrease in fasting serum glucose during Ramadan." And: "It seems that any change in blood glucose during Ramadan is slight and unlikely to adversely affect normal healthy subjects."
  • MEASURED, the artifact that governs this result: baseline blood was drawn after a 12-hour overnight fast, while the Ramadan samples were drawn just before the evening meal. The sampling conditions differ, so part of the 25.5 mg/dL fall is the difference between an overnight fast and a 12-hour daytime fast, not an effect of Ramadan. Plasma volume: NOT corrected, NOT measured, NOT discussed.
  • Grade B+ for the measurement, C for attributing the fall to the fast, given the sampling mismatch.
  • Disconfirming, READ, Madkour 2022, DOI 10.3389/fnut.2022.1008730: fasting glucose ROSE 99 to 108.4 mg/dL, p = 0.036. See D9.
  • READ, Mindikoglu 2020, PMID 33110154: insulin −2.0 µU/mL p = 0.164, glucose −6 mg/dL p = 0.176, HOMA-IR −0.7 p = 0.130. All null during the fast. HOMA-IR reached p = 0.035 only one week after the fast ended.
  • READ, Al Zunaidy 2024, DOI 10.3389/fnut.2024.1437169: fasting glucose null, p = 0.186.
  • READ, Almeneessier 2019, DOI 10.1371/journal.pone.0226034: serum glucose at 15:30 not significantly different across baseline, fasting outside Ramadan and Ramadan, p = 0.4. Null.
  • READ, Dastghaib 2025: glucose measured, no significant difference against the non-fasting control group.

MEASURED, the honest summary: the only pooled glucose estimate is a small favourable one (g = −0.23). The primary studies are split: one large drop confounded by sampling mismatch, one significant rise, and four nulls. Insulin and HOMA-IR are null in every primary study opened. No study corrected for plasma volume.

3.3 Body weight and fat mass

  • READ, Faris 2025 umbrella review, DOI 10.1016/j.clnesp.2025.03.006, healthy subjects: body weight g = −0.33 (−0.37 to −0.29), p = 2.63 × 10⁻⁵²; waist circumference g = −0.30 (−0.38 to −0.23), p = 5.73 × 10⁻¹⁵; body fat percentage g = −0.26 (−0.37 to −0.14), p = 6.81 × 10⁻⁶. Grade B+.
  • READ, Fernando 2019, PMID 30813495: weight −1.34 kg (−1.61 to −1.07); fat mass −0.98 kg; fat-free mass −0.66 kg; body fat −1.07%. At 2 to 5 weeks, weight −0.59 kg and all body composition parameters back to baseline. See D10. Grade B+.
  • READ, Correia JM, Santos I, Pezarat-Correia P, Silva AM, Mendonca GV. "Effects of Ramadan and Non-ramadan Intermittent Fasting on Body Composition: A Systematic Review and Meta-Analysis." Frontiers in Nutrition 2021;7:625240. DOI 10.3389/fnut.2020.625240. 66 articles, 60 meta-analysed, including 27 randomized controlled trials (overwhelmingly in the non-Ramadan arms).
  • Ramadan, pre-post: body weight SMD −0.234 (−0.341 to −0.127), p < 0.001; BMI −0.212 (−0.353 to −0.072), p = 0.003; fat mass kg −0.376 (−0.618 to −0.135), p = 0.002; fat mass % −0.197 (−0.312 to −0.082), p = 0.001.
  • Ramadan, versus control: body weight −0.353 (−0.651 to −0.054), p = 0.02; fat mass % −0.533 (−1.025 to −0.04), p = 0.034. BMI −0.439 (−1.017 to 0.138), p = 0.136, NULL. Fat mass kg −0.286 (−0.985 to 0.413), p = 0.423, NULL.
  • Fat-free mass: NULL in every Ramadan analysis. Pre-post −0.142 (−0.324 to 0.041), p = 0.129. Versus control −0.203 (−0.599 to 0.192), p = 0.314.
  • MEASURED, the sensitivity result that matters most: READ, the Ramadan BMI pre-post and Ramadan fat mass % pre-post results are no longer significant when weak-quality studies are excluded. Heterogeneity was often above I² = 70%.
  • Non-Ramadan intermittent fasting was consistently stronger: body weight pre-post −1.845 (−2.213 to −1.477); fat mass kg −1.963 (−2.243 to −1.682). READ: "non-Ramadan IF appears to be more effective in improving overall body composition."
  • Plasma volume: NOT corrected. READ, dehydration IS addressed: Ramadan fasting reduced body weight "despite being accompanied by dehydration," and "coercive dehydration" is listed among factors that may affect body composition changes during Ramadan.
  • Grade A for intermittent fasting body composition in general (27 RCTs pooled). Grade B+ for the Ramadan subgroup specifically, where only 6 of 28 body-weight-versus-control comparisons were Ramadan and the designs are observational.

MEASURED, the honest summary: weight loss is small (about 1.3 kg), about 40% of the tissue lost is lean, it is largely regained within 2 to 5 weeks, the per-protocol effect is weaker than non-Ramadan intermittent fasting, two of the pooled Ramadan estimates collapse when weak studies are dropped, and no analysis partitions the water component despite one meta-analysis stating the fast is accompanied by dehydration.

3.4 Blood pressure

  • READ, Faris 2025 umbrella review, DOI 10.1016/j.clnesp.2025.03.006: diastolic blood pressure, n = 3,456, Hedges' g = −0.26 (95% CI −0.38 to −0.15), p = 9.18 × 10⁻⁶. MEASURED: no significant pooled systolic blood pressure estimate is reported, although SBP is listed among the outcomes examined. Grade B+.
  • READ, Mindikoglu 2020, PMID 33110154: systolic −8 mmHg (p = 0.023) and diastolic −8 mmHg (p = 0.002) at end of week 4, from baseline 139/81 mmHg. One week after the fast both had lost significance: systolic −4 mmHg p = 0.431, diastolic −2 mmHg p = 0.275. MEASURED: the effect did not outlast the fast. 9 of 14 participants were on antihypertensives and medication changes are not reported. Grade B+, n = 14, no control group.
  • READ, Al-Jafar 2024 companion paper: Al-Jafar R, Yuqi W, Elliott P, Tsilidis KK, Dehghan A. "The dietary changes during Ramadan and their impact on anthropometry, blood pressure, and metabolic profile." Frontiers in Nutrition 2024;11:1394673. DOI 10.3389/fnut.2024.1394673. Reporting earlier LORANS results: systolic −7.3 (p < 0.001), diastolic −3.4 (p < 0.001), weight −1.6, BMI −0.60, waist −1.95 (p = 0.011), fat mass −1.24, total body water −1.45 (p < 0.001). MEASURED: units and confidence intervals are not given for these, and they are cited from a prior report rather than newly analysed. Nutrient change explained only 14% of the systolic change.
  • MEASURED: this paper also documents what else changed. Dietary glucose +216%, fructose +175%, sugars +62%, lycopene +368%, starch −21%, proline −28%, glutamic acid −30%, all FDR-significant, while total energy intake changed 0%, mean difference 4.2 kcal (−115 to 123), p = 0.973. Energy was constant and composition was transformed. Any claim that the benefit comes from the fasting window rather than from the diet that replaced the normal one is not supported.
  • Plasma volume: NOT measured, NOT discussed. Total body water fell 1.45 units, which is a hydration signal the authors do not carry into the interpretation.
  • READ, Al Zunaidy 2024: systolic BP not significant. Diastolic time p < 0.001, driven by post-menopausal women 81.41 to 71.00 mmHg.

MEASURED, the honest summary: diastolic blood pressure falls, and that is the one blood pressure result with a pooled estimate behind it. Systolic has no significant pooled estimate in the apex umbrella review and was null in one cohort. The single clearest systolic fall (−8 mmHg) came from n = 14 with no control group and disappeared within a week of the fast ending.

3.5 Lipids

  • READ, Faris 2025 umbrella review, DOI 10.1016/j.clnesp.2025.03.006: total cholesterol, healthy subjects, n = 9,314, g = −0.16 (−0.25 to −0.07), p = 6.27 × 10⁻⁴. Triglycerides, healthy, n = 9,020, g = −0.16 (−0.24 to −0.08), p = 6.87 × 10⁻⁵. Triglycerides, general population, n = 16,688, g = −0.15 (−0.22 to −0.08), p < 0.01. MEASURED: no pooled HDL-C, LDL-C or VLDL-C estimate is reported in the content served. Grade B+.
  • READ, Madkour MI, Islam MT, Tippetts TS, Chowdhury KH, Lesniewski LA, Summers SA, Zeb F, Abdelrahim DN, AlKurd R, Khraiwesh HM, AbuShihab KH, AlBakri A, Obaideen K, Faris ME. "Ramadan intermittent fasting is associated with ameliorated inflammatory markers and improved plasma sphingolipids/ceramides in subjects with obesity: lipidomics analysis." Scientific Reports 2023;13:17322. DOI 10.1038/s41598-023-43862-9. MEASURED: PMCID PMC10576029 observed in the search result URL.
  • n = 57 (40 male, 17 female), age 18 to 58, mean 38.4 ± 11.2, overweight or obese without diagnosed disease, Sharjah UAE, Ramadan 1438 (June to July 2016), about 15 h/day. Prospective cohort, no control group.
  • Lipids: total cholesterol unchanged. LDL-C decreased, p = 0.013. HDL-C increased, p = 0.001. Total triglycerides decreased, p = 0.038. Total diacylglycerols decreased, p = 0.004.
  • Sphingolipids: sphingosine and sphinganine decreased (p ≤ 0.001); sphingosine-1-phosphate and sphinganine-1-phosphate decreased (p ≤ 0.001); sphingomyelin C17, C22, C24 decreased (p = 0.001, 0.021, 0.035) while C14, C16, C18, C20, C24:1 were not significant; dihydrosphingomyelin C20, C22, C24, C24:1 decreased while C16 and C18 were not significant. Ceramides and dihydroceramides: NO significant change, p ≥ 0.11.
  • MEASURED, against the title: the paper is titled "improved plasma sphingolipids/ceramides" and ceramides did not change. That is a title the data does not carry.
  • MEASURED: exact numeric values for the inflammatory markers and sphingolipids appear only in figures and could not be read; only p-values were retrievable.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. READ: blood was drawn 8 to 10 h into the fast at both time points, but the text does not state whether timing was matched.
  • Confounding, READ: total calories did not change significantly, but protein fell (p = 0.002), sugar rose (p = 0.001) and cholesterol intake fell (p = 0.001); physical activity was not objectively measured and the authors cite literature that Ramadan reduces it; sleep and meal timing were not measured; significant weight loss occurred, which could drive the lipid changes independently.
  • Grade B+.
  • READ, Hammouda O, Chtourou H, Aloui A, Chahed H, Kallel C, Miled A, Chamari K, Chaouachi A, Souissi N. "Concomitant Effects of Ramadan Fasting and Time-Of-Day on Apolipoprotein AI, B, Lp-a and Homocysteine Responses during Aerobic Exercise in Tunisian Soccer Players." PLoS ONE 2013;8(11):e79873. DOI 10.1371/journal.pone.0079873.
  • n = 15 male professional soccer players, age 17.3 ± 0.3, Tunisia, Ramadan 2010. Tested before Ramadan, week 2 and week 4, each at 07:00 and 17:00.
  • Resting values moved favourably by end of Ramadan: HDL-C 1.06 to 1.16 mmol/L (p < 0.05 morning); LDL-C 1.33 to 1.26; apo-B 0.55 to 0.51; Lp-a evening 0.13 to 0.10; hs-CRP evening 0.84 to 0.60 mg/L; homocysteine 15.82 to 14.83 µmol/L morning and 16.82 to 14.40 evening, both significant.
  • MEASURED: Hammouda's homocysteine FELL while Mindikoglu's ROSE by 1.97 µmol/L at p = 0.0004. The two human results on homocysteine point in opposite directions. Neither corrected for plasma volume. Homocysteine is not settled in this literature and must not be claimed in either direction.
  • Performance costs, READ: Yo-Yo test distance at 17:00 fell from 2043.64 ± 533.5 m before Ramadan to 1756.36 ± 503.57 at week 2 and 1625.45 ± 562.07 at week 4, both significantly different from before Ramadan. Energy intake fell from 3,302 ± 709.8 to 2,693 ± 517.13 kcal. Body mass fell 67.22 to 64.80 kg.
  • Plasma volume: CALCULATED, NOT applied. READ: "Plasma volume changes were estimated based on resting and postexercise haematocrit and hemoglobin," citing Dill and Costill (J Appl Physiol 1974;37:247-248). MEASURED: no sentence describes any correction, all biomarkers are reported as raw means, and the plasma volume, haematocrit and haemoglobin values are never reported in the Results or tables at all. This is the clearest single instance in the strand of a study that measured the artifact and then buried it.
  • Confounding, READ: the authors acknowledge no control group and that biochemistry was sampled only at rest and 3 min post-exercise.
  • MEASURED, defect: Table 1 gives baseline body mass as 67.22 ± 16.2 kg against 69.1 ± 4.2 kg in the subject description, and the SD is implausible for a 15-man squad.
  • Grade B+.

MEASURED, the honest summary: total cholesterol and triglycerides fall with pooled support. HDL and LDL have no pooled estimate in the apex review. The lipidomics paper's named ceramide claim is not supported by its own data. Every lipid value in this strand is an uncorrected plasma concentration measured in fluid-restricted people.

3.6 Inflammatory markers and cytokines

The strongest design in the whole strand sits here, and it is the one that separates fasting from Ramadan.

  • READ, Almeneessier AS, BaHammam AA, Alzoghaibi M, Olaish AH, Nashwan SZ, BaHammam AS. "The effects of diurnal intermittent fasting on proinflammatory cytokine levels while controlling for sleep/wake pattern, meal composition and energy expenditure." PLOS ONE 2019 Dec 10;14(12):e0226034. DOI 10.1371/journal.pone.0226034.
  • Design: within-subject, four periods, 12 healthy non-obese non-smoking males aged 25.1 ± 2.5, BMI 23.4 ± 3.5, King Saud University, Riyadh, 9 April to 29 May 2018. Periods: adaptation, fasting outside Ramadan (FOR), non-fasting baseline (BL), and Ramadan. Each lab stay about 24 h with serial blood draws through an indwelling cannula. Dawn-to-sunset abstention from food and drink; the printed sunset and dawn times imply roughly 14 to 15 h/day.
  • What was actually controlled, READ: lab energy intake fixed at 35 kcal/kg/24 h with fixed macronutrient proportions in every period; bedtime fixed at 23:00 in every period; one week of pre-enrolment wrist actigraphy confirming regular timing and morning chronotype; home sleep monitored by actigraphy throughout; energy expenditure measured continuously by SenseWear Pro armband in lab and at home; lab light held at about 50 lux from 18:00, under 30 lux at the pre-dawn meal, under 1 lux during sleep.
  • Measured outcomes, READ. Only IL-1β, IL-6 and IL-8 were measured; TNF-α was NOT measured.
  • IL-1β: significantly lower than baseline at 06:00, 11:00, 22:00, 02:00 and 04:00 during FOR, and at 22:00, 02:00 and 04:00 during Ramadan (p < 0.05).
  • IL-6: significantly lower than baseline at 06:00, 22:00, 02:00 and 04:00 during both FOR and Ramadan (p < 0.05). Cosinor amplitude during Ramadan significantly lower than baseline, p = 0.038.
  • IL-8: significantly lower at 11:00, 02:00 and 04:00 during FOR, and at 02:00 and 04:00 during Ramadan (p < 0.05).
  • Null: cosinor acrophase unchanged for all three cytokines in both fasting periods; lab sleep parameters; home sleep duration; FOR energy expenditure; body weight (69.3, 69.6, 69.1 kg, p = 0.9); serum glucose (p = 0.4).
  • MEASURED, the finding that matters most: the cytokine reduction appeared in fasting outside Ramadan as well as in Ramadan, and READ, the authors describe the reduction as "more obvious" during FOR. The anti-inflammatory signal therefore tracks the fasting itself and not Ramadan's social, dietary and sleep package, which is genuine support for the mechanism. MEASURED: but the paper reports no direct statistical comparison between FOR and Ramadan, so "more obvious" is a pattern description and not a tested result.
  • Where the control broke, READ: home control of Ramadan meals largely failed, with only 2 of 12 participants following the food protocol during Ramadan. Period order was fixed (FOR, then BL, then Ramadan) with no counterbalancing, so the results are confounded with season and order. The baseline comparator differed from the fasting periods in meal timing, sleep timing and wake time: fasting periods involved a split night with waking for the pre-dawn meal and a later final rise at 07:45 against 07:00. Home bedtime and rise time were significantly delayed during Ramadan. Home energy expenditure fell significantly during Ramadan. Blood draws through the cannula reduced sleep efficiency in every period.
  • Plasma volume: NOT corrected, NOT measured, NOT discussed. MEASURED: stable body weight is the only indirect hydration indicator and body weight does not measure plasma volume.
  • MEASURED: no numeric cytokine concentrations or cosinor values were retrievable; they sit in figures and in Table 3. The direction and the p-values are witness-level; the magnitudes are not, and no magnitude is claimed here.
  • Grade B+ (human data, within-subject, with the strongest confounder control in the strand; not A, because the fast was not randomized and the period order was fixed).
  • Against it, READ: Mindikoglu 2020 hs-CRP and IL-6 both null (D8). Dastghaib 2025 IL-6 p = 0.29 and hs-CRP p = 0.40, both null against a concurrent control group (D5). Ghashang 2021 IL-1β rose significantly in the fasting group (D7). Al Zunaidy 2024 CRP positivity doubled in post-menopausal women, p = 0.015 (D6).
  • For it, READ: Madkour 2023 IL-6 and TNF-α both decreased, p = 0.001, and IL-10 increased, p = 0.001 (3.5). MEASURED: hs-CRP, IL-1β, adiponectin and leptin were not measured in that study. Al-Jafar 2024 glycoprotein acetyls (GlycA), an NMR inflammation marker, decreased, β = 0.07 (0.10 to 0.04), p < 0.001, FDR 0.006 (3.1), measured 8 to 12 days after Ramadan ended. Hammouda 2013 hs-CRP evening 0.84 to 0.60 mg/L, significant (3.5).

MEASURED, the honest summary: the inflammatory evidence is genuinely split. Six human datasets show reduction in at least one marker (IL-6, TNF-α, IL-8, IL-1β, GlycA, hs-CRP), four show null or adverse movement (hs-CRP null, IL-6 null twice, IL-1β up, CRP positivity up). The two studies with a concurrent non-fasting control group give opposite answers: Dastghaib null, Ghashang IL-1β up. The best-controlled study in the strand supports reduction and also shows the effect is larger when the fast is done outside Ramadan. No study corrected for plasma volume, and every cytokine value is a plasma concentration.

3.7 The circulating immune cell proteome

This is the only outcome where dawn-to-dusk dry fasting has a dedicated, named human literature. It is one laboratory group.

  • READ, Mindikoglu AL, Park J, Opekun AR, Abdulsada MM, Wilhelm ZR, Jalal PK, Devaraj S, Jung SY. "Dawn-to-dusk dry fasting induces anti-atherosclerotic, anti-inflammatory, and anti-tumorigenic proteome in peripheral blood mononuclear cells in subjects with metabolic syndrome." Metabolism Open 2022 Nov 1;16:100214. PMID 36506940. PMCID PMC9731888. DOI 10.1016/j.metop.2022.100214.
  • n = 14 (8 men, 6 women), mean age 59 ± 16, all with metabolic syndrome by Grundy criteria, 12 of 14 with NAFLD. READ: "strict dry fasting without any liquid or food intake," dawn to dusk, more than 14 h/day for 29 days, Ramadan 2019. Sampled before, at the end, and one week after.
  • READ, PBMC proteome, end of fasting versus baseline, log2 fold change and p:
  • Up: TUBB4B +6.08 (0.014), F10 +5.15 (0.004), MYL9 +4.92 (0.036), HMGA1 +4.36 (0.044), LSP1 +4.25 (0.004), GLUD2 +4.24 (0.009), H2AFX +4.06 (0.035), MECP2 +3.54 (0.011), HIST1H1B +3.16 (0.048), CFHR2 +3.08 (0.041), TMEM40 +2.74 (0.027), APOL1 +2.52 (0.044), ACTR3B +2.39 (0.041), HSPA8 +2.01 (0.047).
  • Down: APOH −3.68 (0.0496), CALM3 −3.01 (0.015), CALM1 −2.96 (0.015), CALM2 −2.94 (0.015), RAB1A −2.88 (0.044), CD36 −2.63 (0.040), FLNC −2.37 (0.041), APOB −1.01 (0.008).
  • One week after versus baseline, up: H2AFV +6.91 (0.004), H2AFZ +6.77 (0.004), SOD2 +4.68 (0.010), IGLL5 +2.81 (0.044), CP +1.84 (0.042), KRT77 +1.67 (0.045), ATP1A3 +1.25 (0.048). Down: PPIF −4.33 (0.019), GNG5 −3.85 (0.048), ATP5C1 −2.57 (0.029), FLOT2 −2.01 (0.039), ABCB11 −1.93 (0.047), APOB −1.18 (0.0004).
  • MEASURED: APOB was the single persistent finding, lower at both time points, strongest one week after at p = 0.0004. It is the most defensible item in the whole proteomic set.
  • Pathways, READ: lipid and atherosclerosis 6/32 (p = 6.014e-4); neurodegeneration-multiple diseases 5/76 (p = 0.004); C-type lectin receptor signalling 4/8 (p = 1.064e-5). One week after versus end of fast: chemical carcinogenesis-ROS 3/41 (p = 0.019); Huntington disease 3/57 (p = 0.025); bile secretion 2/6 (p = 0.004); cholesterol metabolism 2/14 (p = 0.023).
  • MEASURED, the statistical problem: no multiple-testing correction is described for approximately 900 proteins. The threshold is nominal p < 0.05 with a 1.5-fold cutoff. At that threshold about 45 proteins would be expected to pass by chance alone, and 22 passed at the end-of-fast time point. The proteomic list is not distinguishable from noise on the reported statistics. APOH at p = 0.0496 is at the boundary.
  • MEASURED: the paper measured no clinical outcome. The weight, BMI, waist and blood pressure reductions it cites are from the 2020 companion paper. Glucose, lipid panel and inflammatory marker results are not reported. The correlations between protein changes and metabolic, lipid, hepatic, adiposity, oxidative stress and inflammation markers sit in Supplementary Tables S4 and S5, which were not in the text served.
  • MEASURED, the naming problem: the title asserts "anti-atherosclerotic, anti-inflammatory, and anti-tumorigenic." These are pathway-level inferences from protein abundance, not measured atherosclerosis, inflammation or tumour outcomes. The paper measured no plaque, no cytokine and no tumour.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. MEASURED: the words plasma volume, hemoconcentration, hydration status and dehydration do not appear. READ: iFOT normalization is within-sample, and the paper does not discuss whether it accounts for volume change. PBMC protein abundance is normalized within the cell pellet, which gives it more protection from plasma volume artifact than any serum measure in this strand. The paper does not make that argument, so it cannot be credited to the authors.
  • Confounding, READ, addressed: inclusion and exclusion criteria (recent cardiovascular events, alcohol or substance use, active cancer, infection); paired within-subject design. READ, not controlled: no control group, no randomization, and medication use, diet composition, physical activity, sleep and hydration are not reported as controlled. The authors call for randomized controlled trials.
  • Grade C for any individual protein in this set, on the multiple-testing ground. Grade B+ for the fact that the PBMC proteome changes measurably during a 29-day dawn-to-dusk dry fast, and B+ for APOB specifically, which persisted and strengthened.
  • READ, the companion serum proteome: Mindikoglu 2020, PMID 33110154. End of week 4 versus baseline, up: AP5Z1 +6.201 (p = 0.005, about 74-fold), VPS8 +5.730 (0.043), INTS6 +4.234 (0.041), CALU +4.135 (0.012), IGFBP5 +4.008 (0.015), POLRMT +3.355 (0.020), KIT +2.891 (0.033), CROCC +2.742 (0.043), PIGR +2.684 (0.041). Down: SRGN −6.286 (0.009), NIFK −5.048 (0.019), CAMP −5.020 (0.012), CD109 −3.977 (0.027), POLK −2.987 (0.026). One week post, up: PRKCSH +6.191 (0.031), CALU +5.026 (0.007), CALR +3.958 (0.020), IGFBP4 +3.728 (0.033), SEMA4B +3.386 (0.020), HIST1H2BB +2.656 (0.044), HIST1H2BD +2.620 (0.044), HIST1H2BA +2.565 (0.020). Down: PLAC1 −7.071 (0.032), CAMP −3.743 (0.038).
  • MEASURED, instability: PRKCSH was −0.642 at p = 0.336 at the end of week 4 and +6.191 at p = 0.031 one week later. A protein moving from null to 73-fold in seven days, in n = 14 with no multiple-comparison correction, is a noise signature.
  • READ, the authors' own finding: "There was no significant correlation between log2 fold changes in the selected proteins" and changes in weight, waist or BMI.
  • Plasma volume: NOT corrected, NOT measured, NOT mentioned. MEASURED: unlike the PBMC paper, these are serum concentrations measured in people who had taken no fluid for more than 14 h, and they carry the hemoconcentration artifact directly and uncorrected.
  • Grade C for any individual serum protein. Grade B+ for the fact of measurable serum proteome change.
  • READ, the third paper in the series: Mindikoglu AL, Eckel-Mahan K, Opekun AR, Alzubaidi MM, Crochet ZR, Jalal PK, Jung SY. "Decreased proliferation of HepG2 liver cancer cells in vitro and exhibited proteomic changes in vivo in subjects with metabolic syndrome and MASLD who performed four-week dawn-to-dusk dry fasting." 2025. DOI 10.1186/s12014-025-09547-3. MEASURED: PMCID PMC12186377 observed in the search result URL; the journal is not printed in the text served and the DOI prefix suggests Clinical Proteomics. I will not assert the journal on an inference.
  • n = 4 metabolic syndrome/MASLD (3 men, 1 woman, mean age 67 ± 6) and n = 4 healthy (all male, mean age 32 ± 6). Proteomics on 3 versus 3 after removing one outlier from each group. Dry fasting over 14 h daily for 4 weeks.
  • READ: HepG2 viability fell in 6 of 7 MetS/MASLD serum assays (p from 0.0002 to 0.067) but ROSE in one MetS/MASLD subject (S4, p = 0.0002) and rose in one healthy subject (p = 0.001), and was null in three healthy assays.
  • READ, proteomics, MetS/MASLD: CD248 +8.124 (p = 0.001), LRP1 +1.401 (0.031), LYVE1 +1.054 (0.029), DPP4 +0.937 (0.027), B2M −0.977 (0.033). Healthy: CSTA −12.325 (0.0003), PFN1 −12.865 (0.012), FABP5 −12.127 (0.010), LCN2 −11.271 (0.002), FLNA −10.331 (0.0498), YWHAE −9.769 (0.026), FSTL4 −9.016 (0.0002), CTSC −8.542 (0.001), CD93 −2.567 (0.041), CD109 −1.321 (0.039), CHMP4A +0.661 (0.045). READ: EGFR and APOA1 network nodes decreased in MetS/MASLD and increased in healthy subjects.
  • MEASURED: log2 fold changes of −12.9 and +8.1 from n = 3 per group are not credible as serum protein changes over four weeks. READ, the authors' own limitation: this is a pilot with no prior power calculation, and they estimate at least 15 matched samples would be needed. Age, sex and lifestyle could not be accounted for and the two groups differ substantially in both age and sex composition. Only HepG2 cells were tested, with no dose escalation. READ: the MTT assay reflects mitochondrial activity, not cell death.
  • MEASURED, defect: the text calls CHMP4A a "2-fold" increase while log2 0.661 is about 1.6-fold.
  • Plasma volume: NOT corrected, NOT measured. MEASURED: hydration appears only as background and as a safety recommendation, and the single osmolality statement cited (normal plasma and urine osmolality) comes from a different study, reference 38, not from this cohort. READ: the end-of-study draw time of day is not stated. MEASURED: serum drawn after a fasting period with no fluid intake was applied to cells at whatever concentration it came out at. Hemoconcentration could affect both the proteomic and the viability results, and the paper does not address this.
  • Grade C. n = 3 per group, implausible effect magnitudes, a contradicted direction in two of eight subjects, and an in vitro endpoint standing in for a clinical one.

MEASURED, the honest summary: three papers from one group are the entire human dawn-to-dusk dry-fasting proteome literature. All three are pilots with n = 14 or smaller, none has a control group, none applies multiple-testing correction, none corrects for plasma volume, and the one that applied serum to cancer cells had n = 3 per group with fold changes that are not physiologically credible. The PBMC work is the most defensible because protein abundance is normalized inside the cell pellet rather than per volume of plasma. APOB, down at both time points and strongest at p = 0.0004 one week after, is the single most defensible result. Everything else in this outcome is C-grade.

3.8 Kidney function and hydration markers

Kidney function

  • READ, Bello 2024, PMID 39572100: inconsistent effects, no RCTs, every pooled creatinine and eGFR estimate crossing zero except one transplant creatinine fall with an upper bound of −0.02. See D3. Grade B+.
  • READ, Tarabeih M, Qaddumi J, Hamdan Z, Hassan M, Jebrin K, Khazneh E, Bahar S, Ahmed N, Sawalha R, Sawalmeh O. "Increasing Overnight Fluid Intake and Kidney Function During Ramadan Fasting: A Randomized Controlled Trial." Transplantation Proceedings 2022. DOI 10.1016/j.transproceed.2022.10.059. MEASURED: PMID 36549977 observed in the search result URL. This is the only randomized trial in the strand.
  • Design: 58 healthy adults aged 18+, no diabetes, hypertension, heart or kidney disease, randomized by SPSS into two groups of 29. Hydrated group asked to drink 2 to 3 L overnight from sunset to dawn; control group continued usual intake, averaging about 620 mL. Measured on Ramadan days 1, 10, 20 and 30, and 30 days after. May to August 2018.
  • READ, measured outcomes: the control group showed a higher change of 0.214 mg/dL in mean serum creatinine (95% CI 0.191 to 0.238, p < 0.001) and a higher change of 6.16 mg/dL in mean urea (95% CI 5.29 to 7.03, p < 0.001). eGFR by Cockcroft-Gault averaged 150.1 mL/min/1.73 m² in the hydrated group against 108.9 in controls.
  • Exact sentences relied on, READ: "Ramadan fasting was not associated with a permanent increase in serum creatinine or urea." And: "these changes were only temporary and disappeared when Ramadan was concluded." And: "We also found no clinically adverse effects on renal function tests in both groups."
  • MEASURED, why this is the single most important study in the strand: it is a randomized demonstration that the creatinine and urea rise seen during Ramadan is a FLUID-INTAKE effect, not a kidney effect. Give people 2 to 3 L overnight and the rise is largely abolished; leave them at 620 mL and creatinine rises 0.214 mg/dL and urea 6.16 mg/dL. That is direct randomized evidence that the renal markers in every uncorrected observational Ramadan study are measuring hydration.
  • Plasma volume: NOT measured, NOT corrected. MEASURED: haematocrit appears only in the baseline table. Urine osmolality, urine specific gravity and serum osmolality were NOT measured, and READ, the authors list the lack of urine analysis as a limitation.
  • MEASURED, defect: the Results text and Table 2 give a 6.16 mg/dL urea group difference while Table 3 gives 12.84 mg/dL. READ, the authors acknowledge a small sample and a one-month post-Ramadan follow-up.
  • Grade A for the effect of overnight fluid volume on creatinine and urea during Ramadan fasting (randomized, endpoint measured directly). This is the only grade A on a dawn-to-dusk question in this strand, and what it establishes is that the renal signal is hydration.
  • READ, Turkish CKD 2019, DOI 10.1515/tjb-2018-0373: urea rose p = 0.015 in fasters, creatinine and eGFR null, with a non-fasting comparison group. See D4. Grade B+.
  • READ, Dastghaib 2025: BUN higher in fasters, p = 0.045, with a concurrent control group. See D5. Grade B+.
  • READ, Salwani 2025, DOI 10.55572/jms.v6i2.246: urine ACR null in diabetic kidney disease, p = 0.140. See D14. Grade B+.
  • READ, Darzabi & Hejazi 2020, DOI 10.5812/jjcdc.100079: urea and uric acid showed significant between-group interaction against non-fasting controls (p = 0.04, p = 0.005) while creatinine did NOT (p = 0.80). See 2.4. Grade B+ design, C per-marker.
  • READ, Ziaee 2007: "We found no significant changes in levels of urea and creatinin in serum." Grade C.
  • READ, Al-Jafar 2024, PMID 38301826: the renal failure metabolic risk score was not significant, 0.72 (95% CI 3.34 to 1.89). MEASURED: minus signs appear lost in the text served so the sign cannot be confirmed. Grade C.
  • Clinical event evidence: renal colic up after a 25-hour total fast (D2, Grade B+); null during Ramadan with a post-Ramadan rise (D12).

MEASURED, the convergent reading across every kidney result: urea and BUN move; creatinine and eGFR do not. Urea is the marker most sensitive to dehydration and to protein catabolism. The one randomized trial shows that giving 2 to 3 L overnight abolishes most of the urea and creatinine rise. The most defensible conclusion is that the renal signal in dawn-to-dusk fasting is a hydration signal, not evidence of kidney injury, and equally not evidence of kidney benefit.

Hydration markers: osmolality, creatinine, eGFR, urine specific gravity

  • Plasma / serum osmolality. READ, Trabelsi 2011, PMID 22375238: resting plasma osmolarity 284.7 to 292.8 mOsm/L, +2.8% at day 30, p < 0.001, with the authors' own conclusion "Ramadan fasting induces dehydration at basal conditions." Grade B+. READ, Darzabi & Hejazi 2020: serum osmolarity changed significantly in fasters only, with a significant between-group interaction, p = 0.01, though direction is unsafe to quote because the table headers appear transposed. Grade C. READ, Ziaee 2007: serum osmolarity increased significantly in both sexes, direction only. Grade C. MEASURED: in Trabelsi and Darzabi, osmolarity was CALCULATED from 2×[Na] + urea + glucose, not measured by osmometry. No study in this strand measured osmolality directly by osmometer.
  • Serum creatinine. Covered above. Pooled estimates null; the randomized fluid trial shows the observational rise is a hydration artifact.
  • eGFR. READ, pooled: 0.23 mL/min/1.73 m² (−1.64 to 2.10) in CKD against a non-fasting arm, and 0.03 (−1.14 to 1.20) in transplant. Null. READ, Turkish CKD: null, p = 0.183. READ, Tarabeih randomized: 150.1 hydrated against 108.9 control, by Cockcroft-Gault. MEASURED: that is a 41 mL/min/1.73 m² gap produced purely by overnight drinking volume, which is a further demonstration that eGFR in this setting tracks hydration.
  • Urine specific gravity. MEASURED: NOT MEASURED IN ANY STUDY OPENED IN THIS SWEEP. Zero of twenty sources reported it. Trabelsi 2011 measured plasma osmolarity, haematocrit and haemoglobin and explicitly did not measure urine osmolality, urine specific gravity or urine colour. Tarabeih 2022 lists the absence of urine analysis as a limitation. The single most direct, cheapest and least artifact-prone hydration marker is absent from this entire literature. This is a real gap in the record and must be stated as such, not filled.
  • Total body water. READ, Trabelsi 2011: 51.3 / 52.8 / 51.5 L by impedance, not significantly different, and MEASURED, the authors themselves say electrolyte shifts may have distorted the estimate, so the result is unreliable. READ, Al-Jafar 2024 Frontiers: total body water fell 1.45 (p < 0.001) after Ramadan, units not given. READ, Al Zunaidy 2024: total body water not significant. MEASURED: three studies, three different answers, all by bioimpedance, which is the method most sensitive to the electrolyte shifts the fast itself causes.

PART 4. WHAT EACH STUDY COULD AND COULD NOT ISOLATE

MEASURED, by extraction across the corpus. Ramadan moves sleep, meal timing, intake composition, activity and season together.

ConfounderIsolated anywhere?Evidence
Fasting versus Ramadan's social packageYES, onceAlmeneessier 2019 ran fasting OUTSIDE Ramadan as a separate period; cytokines fell in both, more obviously in FOR. The only true separation in the strand.
Sleep/wakePartially, onceAlmeneessier 2019: bedtime fixed at 23:00, pre-enrolment and home actigraphy. But fasting periods had a split night and a later rise (07:45 against 07:00), and home bedtime and rise were significantly delayed during Ramadan. Not mentioned at all in Mindikoglu 2020, 2022, 2025; Madkour 2022, 2023; Fernando 2019; Dastghaib 2025 (not controlled); Al-Jafar 2024 (not adjusted).
Meal compositionAttempted, failedAlmeneessier 2019 fixed lab intake at 35 kcal/kg/24 h; only 2 of 12 followed the food protocol at home during Ramadan.
Total energy intakeMeasured, and it does not changeAl-Jafar 2024 Frontiers: energy change 0%, mean difference 4.2 kcal (−115 to 123), p = 0.973, while dietary glucose rose 216% and fructose 175%. Madkour 2023: total calories unchanged but protein fell (p = 0.002) and sugar rose (p = 0.001). Hammouda 2013: energy fell 3,302 to 2,693 kcal. Trabelsi 2011: energy fell 15.4%. Mindikoglu 2020: not measured at all, and the authors list this as a limitation.
Physical activityMeasured, and it fallsAlmeneessier 2019: home energy expenditure fell significantly during Ramadan. Madkour 2023: not objectively measured; authors cite literature that Ramadan reduces activity. Fernando 2019: not controlled. Hammouda 2013: Yo-Yo distance fell from 2043.6 m to 1625.5 m at 17:00.
Hydration and plasma volumeNEVERZero of twenty sources corrected a biomarker. See Part 2.
Season and orderNEVERAlmeneessier 2019 used a fixed period order with no counterbalancing. Mindikoglu 2020 and 2022 were fixed to May to June 2019. Zilberman 2021 acknowledges warm September and October weather. Lidawi 2024 does not model seasonality.
MedicationNEVERMindikoglu 2020: 9 on antihypertensives, 7 on antidiabetics, 6 on statins; changes not reported. Al-Jafar 2024: 8 with diabetes, 15 with hypertension, not adjusted.
Concurrent non-fasting control group4 of 20Present: Dastghaib 2025, Turkish CKD 2019, Ghashang 2021, Darzabi & Hejazi 2020. Absent: all others, including every Mindikoglu paper and all three large meta-analyses of primary Ramadan data. Poursalehian 2024, 35 studies: "none of the studies included a control group of non-fasted participants."
Nocturnal prayer, social and psychological factorsNEVERAlmeneessier 2019 names these as unmeasured.

PART 5. LANGUAGE SWEEP RECEIPT

Languages searched: 10 of 10 required. English, Arabic, Turkish, Persian, Malay, Indonesian, Urdu, French, German, Russian. Every one is named below with what it returned, including the empty ones.

LanguageQuery returnedPrimary research obtained
EnglishFull yield. 20 primary studies and syntheses opened and extracted.YES. Every source in Parts 1 to 4.
Arabic (صيام رمضان وظائف الكلى الكرياتينين الأسمولية الجفاف)EMPTY for primary research. Returned lay health press (youm7.com ×2, alkhaleej.ae, aawsat.com), clinical-advice portals (altibbi.com, drnermeen.com, ekshef.com), a radio health programme page (mc-doualiya.com) and one religious polemic page. No journal article, no data.NO. MEASURED: the Arabic-language primary literature on this subject is published in English-language journals by Arab authors. Faris, Madkour, Trabelsi, Almeneessier, BaHammam, Al Zunaidy, Al-Jafar, AlKurd, Khaled, Abdelrahim and Abu Shihab are all already in this record. The Arabic literature is captured; it is simply not written in Arabic.
Turkish (Ramazan orucu böbrek fonksiyonu kreatinin ozmolalite dehidratasyon)PRODUCTIVE. Returned one primary study plus Turkish nephrology guidance pages (tekinakpolat.com ×2, istahed.org.tr, turkhipertansiyon.org, dralihatay.com) and one Index Copernicus PDF on kidney patients and fasting.YES. "The effect of Ramadan fasting on renal functions in patients with chronic kidney disease," DOI 10.1515/tjb-2018-0373, extracted in full as D4. One of only four studies in the strand with a non-fasting comparison group.
Persian (روزه رمضان عملکرد کلیه کراتینین اسمولالیته کم آبی)PRODUCTIVE. Returned Ziaee 2007 at ijem.sbmu.ac.ir (serum osmolarity, electrolytes, haematological parameters) and an Iranian study on resistance training plus fasting with creatinine clearance and GFR in inactive women with abdominal obesity (ebnesina.ajaums.ac.ir article-1-658-fa), plus Persian creatine-supplement pages that are off topic. A second Persian-language systematic review surfaced on the Urdu query: journals.sbmu.ac.ir/jrrh/article/view/42393, on fasting with light physical activity and haematological and biochemical indices.YES. Ziaee 2007 extracted as 2.3, the one result in the strand whose pattern hemoconcentration alone cannot explain. MEASURED: the two further Persian items were located but not opened; no claim rests on them.
Malay (kesan puasa Ramadan terhadap fungsi buah pinggang kreatinin dehidrasi kajian Malaysia)EMPTY for Malay-language primary research. Returned one Malaysian clinical-advice article by a nephrologist (mmgazette.com), one Malaysian Federal Territory Mufti ruling on breaking the fast for chronic illness (muftiwp.gov.my), and Indonesian items already captured.NO Malay-language primary study. One Malaysian-journal study obtained in English: Larijani 2003, Med J Malaysia 58(5):678-680, extracted in 3.2.
Indonesian (puasa Ramadan fungsi ginjal kreatinin osmolalitas dehidrasi)PRODUCTIVE. Returned Salwani 2025 (urine ACR in diabetic kidney disease, rsudza.acehprov.go.id), "Puasa Ramadhan dan Pengaruhnya terhadap Progresifitas Penyakit Ginjal Kronik" (jurnal.usk.ac.id JKS), and Indonesian health reviews on ResearchGate.YES. Salwani 2025, DOI 10.55572/jms.v6i2.246, extracted in full with verbatim Indonesian sentences as D14. MEASURED: the JKS CKD-progression article was located but not opened; no claim rests on it.
Urdu (رمضان روزہ گردے کے افعال کریٹینین پانی کی کمی تحقیق مطالعہ)EMPTY for primary research. Returned Islamic magazine essays on the physical and psychological benefits of fasting (tarjumanulquran.org ×2, suayharam.com, naseemathar.blogspot.com), an Urdu pharmaceutical-trade article, and an entertainment site. No journal article, no data. One Persian systematic review surfaced incidentally and is logged under Persian.NO. MEASURED: no Urdu-language primary research on this subject was reachable on any available channel.
French (jeûne du Ramadan fonction rénale créatinine osmolalité déshydratation)PARTIALLY PRODUCTIVE, then blocked. Returned one French-language item, "Ramadan et diabétique insulinonécessitant : Attention aux reins," Annales d'Endocrinologie, ScienceDirect PII S0003426617307138, which warns about kidneys in insulin-requiring diabetics during Ramadan. ScienceDirect refused the fetch (client_error) and under channel law I did not return to it. The remaining French-query results were English-language papers, several already in this record.PARTIAL. The French-language item is logged with its identifier and journal. MEASURED: it was not opened, so no finding is taken from it and no verdict rests on it.
German (via "prospective controlled study Ramadan summer Germany cytokines")PRODUCTIVE. Returned two German-based studies.YES, two. Ghashang 2021, DOI 10.1515/jcim-2019-0209, Hannover, 50 men, 18 to 19 h/day, with a non-fasting control group, extracted as D7. And the Munich dawn-to-dusk metabolomics pilot, Mathew/Krug/Skurk/Halama/Suhre, J Transl Med 2014;12:161, DOI 10.1186/1479-5876-12-161, 11 men at the Islamic Centre Munich, extracted in 3.1 for the finding that it measured about 200 metabolites and mentions no ketone body at all. Both published in English.
Russian (пост Рамадан функция почек креатинин осмолярность обезвоживание)EMPTY. Returned only general Russian clinical-laboratory reference pages on creatinine and kidney-function panels (helix.ru ×2, fdoctor.ru, doctoryurov.com, aidiagme.com), one Russian urology article on sodium hypochlorite and urinary lithogenic properties, and one endocrinology dissertation PDF. Nothing on Ramadan, dawn-to-dusk fasting or dry fasting.NO. MEASURED: no Russian-language research on this subject was found.

MEASURED, the structural finding from the language sweep: this literature is written in English by authors from Arabic-, Turkish-, Persian-, Malay- and Indonesian-speaking countries. Searching the source languages recovered three additional primary studies that the English queries had not surfaced (Turkish CKD 2019, Ziaee 2007, Salwani 2025), and two of those three have a non-fasting comparison group, which makes them disproportionately valuable. The Arabic and Urdu queries returned religious and lay health writing only. The Russian query returned nothing on the subject.


PART 6. CHANNELS

MEASURED, logged for audit. Under channel law, each blocked channel was abandoned at the first refusal and never retried.

ChannelStatusAction
WebSearchWorking throughoutPrimary discovery instrument
WebFetch on publisher sites (nature.com, frontiersin.org, journals.plos.org, mdpi-hosted, degruyterbrill.com, brieflands.com, mbrc.shirazu.ac.uk, e-mjm.org, rsudza.acehprov.go.id, staff.najah.edu, ijem.sbmu.ac.ir, spiral.imperial.ac.uk, translational-medicine.biomedcentral.com, ajcn via spiral)WorkingPrimary extraction instrument
pubmed.ncbi.nlm.nih.govreCAPTCHA on first requestAbandoned immediately, never retried
www.ebi.ac.uk (Europe PMC REST API)Proxy refused, HTTP 403 connect_rejectedAbandoned
europepmc.org (web)client_errorAbandoned
pmc.ncbi.nlm.nih.govWorked for 10 fetches, then reCAPTCHAAbandoned at the captcha; switched to publisher sites and the Imperial repository
semanticscholar.orgclient_errorAbandoned
journals.mums.ac.irrobots.txt unreachableAbandoned
sciencedirect.comclient_error on two separate articlesAbandoned; cost one French-language item and one glucometabolic meta-analysis
nmcd-journal.comclient_errorAbandoned; cost one cardiometabolic meta-analysis with systolic BP
thieme-connect.comanti-bot interstitialAbandoned
researchgate.netNot attempted; known to blockNo claim rests on any ResearchGate-only item

MEASURED, what the blocked channels cost, stated as a known gap rather than filled: 1. Faris et al., "A systematic review, meta-analysis, and meta-regression of the impact of diurnal intermittent fasting during Ramadan on glucometabolic markers in healthy subjects," ScienceDirect PII S0168822720304769. Pooled insulin and HOMA-IR estimates remain unread. 2. "Does four-week consecutive, dawn-to-sunset intermittent fasting during Ramadan affect cardiometabolic risk factors in healthy adults? A systematic review, meta-analysis, and meta-regression," Nutrition, Metabolism and Cardiovascular Diseases, PII S0939-4753(21)00223-4. A pooled systolic blood pressure estimate remains unread. 3. "Ramadan et diabétique insulinonécessitant : Attention aux reins," Annales d'Endocrinologie, PII S0003426617307138. French-language renal warning remains unread. 4. Faris et al., "Ramadan Fasting Exerts Immunomodulatory Effects: Insights from a Systematic Review," PMCID PMC5712070, Frontiers in Immunology 2017;8:1144, DOI 10.3389/fimmu.2017.01144. Located, not opened. No claim rests on it. 5. The Ramadan ketone-bodies and gut-microbiome paper, available only on ResearchGate. No claim rests on it.


# RECEIPT

Outcomes covered: 9 of 9

#Outcome requiredStatusBest gradePlasma volume corrected anywhere?
1Ketone bodies and how fast they riseCOVERED, and the finding is a GAP. No measured BHB during a dawn-to-dusk fast exists in the retrievable literature. Acetone rose (β 0.10, p = 0.001) measured 8 to 12 days AFTER Ramadan. The apex ketone meta-analysis excluded Ramadan and required ≥24 h fasts.B+ for acetone post-fast; A for ≥24 h water-permitted fasting, which is a different exposureNO
2Insulin and glucoseCOVEREDB+ (pooled FBG g = −0.23)NO
3Body weight and fat massCOVEREDA for intermittent fasting generally (27 RCTs); B+ for Ramadan specificallyNO
4Blood pressureCOVEREDB+ (pooled DBP g = −0.26; no significant pooled SBP)NO
5LipidsCOVEREDB+ (pooled TC and TG g = −0.16)NO
6Inflammatory markers and cytokinesCOVEREDB+ (Almeneessier 2019, best-controlled design)NO
7Circulating immune cell proteomeCOVEREDC per protein; B+ for APOB and for the fact of changeNO (PBMC normalization gives partial intrinsic protection, not claimed by the authors)
8Kidney function: creatinine, eGFRCOVEREDA (Tarabeih 2022, the only randomized trial, on overnight fluid volume)NO
9Hydration markers: osmolality, urine specific gravityCOVERED. Osmolality: +2.8%, p < 0.001, always CALCULATED and never measured by osmometer. Urine specific gravity: NOT MEASURED IN ANY OF THE 20 SOURCES.B+ for plasma osmolalityNO

Languages searched: 10 of 10

English (full yield), Arabic (empty for primary research), Turkish (1 primary study, with control group), Persian (1 primary study extracted, 2 more located), Malay (empty in Malay; 1 Malaysian-journal study in English), Indonesian (1 primary study extracted), Urdu (empty), French (1 item located, channel refused, not opened), German (2 primary studies, one with control group), Russian (empty).

Plasma volume correction: 0 of 20

MEASURED: not one source corrected any reported biomarker for plasma volume change. Three calculated it and did not apply it (Khalfoun 2024, Trabelsi 2011, Hammouda 2013, the last never even reporting the values). Seventeen did not measure or mention it.

Randomized trials of dawn-to-dusk fasting itself: 0

MEASURED. The one RCT randomized overnight fluid volume inside the fast.


# WHAT DAWN-TO-DUSK FASTING IS AND IS NOT SHOWN TO DO

Shown, with pooled human support (grade B+):

  • Lowers body weight by about 1.34 kg, of which roughly 0.66 kg is lean tissue, with most of the weight regained within 2 to 5 weeks. Fat-free mass change is null in every pooled Ramadan analysis.
  • Lowers waist circumference and body fat percentage, by small standardized effects (g = −0.30 and −0.26).
  • Lowers diastolic blood pressure (g = −0.26).
  • Lowers total cholesterol and triglycerides (g = −0.16 each).
  • Lowers fasting blood glucose by a small standardized effect (g = −0.23), against primary studies that are split, one of which measured a significant RISE.
  • Measurably raises plasma osmolality (+2.8%), haematocrit (+4.4%) and haemoglobin (+3.4%) at rest. The authors who measured it concluded the fast "induces dehydration at basal conditions."
  • Changes the circulating immune cell proteome measurably, with APOB reduced at both the end of the fast and one week after (p = 0.0004).

Shown, with randomized human support (grade A), and this is the only grade A on the fast's own terrain:

  • The rise in serum creatinine and urea during Ramadan is largely a function of overnight fluid intake, not of kidney damage. 2 to 3 L overnight versus about 620 mL abolished most of the difference (creatinine 0.214 mg/dL, urea 6.16 mg/dL, both p < 0.001).

NOT shown:

  • That ketones rise, or how fast. There is no measured beta-hydroxybutyrate value during a dawn-to-dusk fluid-restricted fast anywhere in the retrievable literature. The apex ketone meta-analysis excluded this fasting model by design and required fasts of 24 hours or more with water permitted.
  • That inflammation falls reliably. Six datasets show a fall in at least one marker; four show null or adverse movement. hs-CRP was null in two studies, IL-6 null in two, IL-1β rose significantly in one, and CRP positivity doubled in post-menopausal women (p = 0.015). The two designs with concurrent control groups disagree with each other.
  • That insulin or insulin resistance improves during the fast. Insulin and HOMA-IR were null in every primary study opened; HOMA-IR reached significance only after the fast ended.
  • That systolic blood pressure falls. No significant pooled estimate exists in the apex umbrella review, and the clearest single fall (−8 mmHg) came from n = 14 with no control group and was gone within a week.
  • That ceramides improve. The paper titled for ceramide improvement measured no significant ceramide change (p ≥ 0.11).
  • That any named proteomic change is real. No multiple-testing correction was applied across approximately 900 proteins in the PBMC work or in the serum work; one protein moved from p = 0.336 to a 73-fold change in seven days.
  • That it is anti-cancer, anti-atherosclerotic or anti-tumorigenic. Those words appear in paper titles. No plaque, no tumour and no disease incidence was measured in any of them; the cancer risk figures are metabolite-derived predictions, and the authors state the work "does not provide evidence" of lower disease risk.
  • That it is safe for people with liver cirrhosis. The opposite was measured: hospitalization RR 3.80 and upper GI bleeding RR 8.50.
  • That it is free of renal event risk. A 25-hour total fast raised emergency visits for renal colic by about 1.6-fold, and a Ramadan case-control found a significant rise in the month after.
  • That it preserves kidney function better than not fasting. Every pooled creatinine and eGFR estimate against a non-fasting arm crosses zero. The synthesis's own words: "inconsistent effects."
  • That any of the measured blood changes are real rather than concentration artifacts. Zero of twenty sources corrected a single biomarker for plasma volume. Plasma volume contracts by 2 to 4% during the fast, and by the most in people with the highest body mass. Several pooled renal estimates are of the same order as that uncorrected artifact.

The one honest mechanistic win, and it is worth having: Almeneessier 2019 ran the same men through fasting OUTSIDE Ramadan and through Ramadan, with intake fixed at 35 kcal/kg/24 h, bedtime fixed, sleep actigraphed and energy expenditure instrumented. IL-1β, IL-6 and IL-8 fell below baseline in both fasting periods, and the authors describe the fall as more obvious when the fast was done outside Ramadan. That is the only design in this literature that separates the fast from Ramadan's sleep, social and dietary package, and it points the effect at the fasting itself. It is n = 12 healthy young men, period order was fixed, home meal control failed for 10 of 12 during Ramadan, no magnitude is readable, and plasma volume was never corrected.

The governing caution for any customer-facing use of this strand: Dawn-to-dusk fasting is the only dry-fasting model with a real controlled human literature, and that literature is observational, uncontrolled in most instances, and uniformly uncorrected for the fluid shift the fast itself causes. The effects that survive scrutiny are small, mostly temporary, and concentrated in diastolic blood pressure, total cholesterol, triglycerides, waist and weight. Harm is documented and specific: cirrhosis, renal colic, hypoglycaemia in diabetes. The single strongest piece of evidence in the entire strand is a randomized trial proving that the renal numbers people cite are measuring how much water was drunk overnight.


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.