Duration-dependent effects of water-only fasting on blood lipids: a systematic review, meta-analysis, and threshold meta-regression.
Level 3 - non-randomized controlled study
Systematic review and meta-analysis of pre-post observational intervention studies
PubMed 41994097 · doi:10.3389/fnut.2026.1772246
What was done
A systematic review, meta-analysis, and piecewise threshold meta-regression evaluated the effects of water-only fasting on blood lipids across varying durations. PubMed, Scopus, and Web of Science (2000–2025) were searched for human studies reporting pre-post lipid measurements. Thirty-two studies met eligibility criteria. Effect sizes were calculated as Hedges' g using random-effects models, with subgroup analyses (≤3 days vs. >3 days) and piecewise threshold meta-regression to evaluate duration-dependent responses. Publication bias was assessed using funnel plots and Egger's tests.
What was found
Water-only fasting produced lipid-specific and duration-dependent adaptations: - HDL decreased significantly overall (Hedges' g = -0.233), with no change in fasts ≤3 days, clear reductions in fasts >3 days, and a decline threshold within the first ~3 days. - LDL increased significantly overall (g = 0.489) across all duration subgroups, with progressive elevation up to ~10 days followed by attenuation or partial reversal. - Total cholesterol increased significantly overall (g = 0.343), with larger effects in fasts >3 days and a nonlinear threshold at ~5 days marking stabilization or modest decline. - Triglycerides showed no significant overall effect (g = -0.039), with reductions in fasts ≤3 days and increases in fasts >3 days (threshold at ~2.5 days). - VLDL exhibited small, non-significant changes (g = 0.203) with substantial heterogeneity and limited data. - Publication bias was detected for LDL and total cholesterol, but not for HDL, triglycerides, or VLDL.
Why it matters
This review demonstrates that water-only fasting does not uniformly improve lipid profiles, but instead causes duration-dependent elevations in LDL and total cholesterol along with multi-day declines in HDL. These findings provide empirical thresholds to guide metabolic monitoring during therapeutic fasting protocols.
Limits
The review included pre-post fasting studies without control comparator groups. Total participant sample size and demographic characteristics were not stated in the abstract. Evidence of publication bias was present for LDL and total cholesterol, and VLDL analysis was limited by sparse data and high heterogeneity.
Cited by
- supports Fasting can cause a transient increase in LDL cholesterol due to increased lipid flux from the liver.