A 12-day medically supervised fast produced large but almost entirely reversible metabolic improvements, and — contrary to longevity-hype expectations — did not uniformly lower epigenetic age. Instead, the aging clocks moved in scattered, clock-specific directions, and people who already looked “biologically younger” at baseline were the ones who responded most strongly to the fast.
Fasting has become common in the longevity movement, sold on the promise that going without food flips ancient cellular switches — autophagy, AMPK, ketosis — that wind back the biological clock. A new study from a Swiss-German-led team puts that promise under a molecular microscope, and the picture is more sobering than the marketing.
Researchers followed 32 healthy adults through a 12-day inpatient “Buchinger” fast (roughly 200–250 kcal/day of juices and broth), sampling blood before, during, and one month after. The metabolic response was textbook and dramatic: ketones climbed more than ten-fold, weight fell about 6 kg, and glucose, insulin, LDL cholesterol and blood pressure all dropped sharply. The catch: by the one-month follow-up, after people returned to normal eating, almost all of those gains had evaporated back toward baseline. Fasting, in other words, produced a powerful state — not a durable trait.
The novel part is what the DNA said. The team ran twelve different “epigenetic clocks” — algorithms that read chemical methylation marks on DNA to estimate biological age and mortality risk. If fasting rejuvenated the body, the clocks should have ticked backward in unison. They didn’t. During the fast, several chronological-age clocks actually ticked slightly up (a known stress response), the pace-of-aging clock DunedinPACE ticked down, and at one month the clocks fragmented into system-specific patterns — heart-related markers up, musculoskeletal markers down — with enormous person-to-person variation. At the single-gene level, fasting left almost no fingerprint at all during the fast itself.
The most provocative finding flips the usual causal arrow. Rather than fasting making people younger, the people who were already epigenetically younger at baseline lost the most weight. That hints these clocks may capture something real about metabolic adaptability — a measure of how flexibly your biology can switch fuels under stress — rather than acting as a dial that fasting can simply turn.
The authors are admirably restrained: a small sample, no control group, blood-only sampling, and a mere one-month follow-up mean this is a careful mapping exercise, not proof that fasting extends human life. What it does deliver is a useful reality check — composite aging clocks can blur genuinely opposite biological signals, and short, intense interventions may not register on them the way enthusiasts assume.
Actionable Insights
The honest take-home: a 12-day fast delivers large but transient cardiometabolic benefits and no demonstrated durable reduction in biological age. Effect sizes below are within-subject change vs. baseline; Cohen’s d uses pooled SD (a standardized magnitude), and reversal status is from the one-month follow-up.
- Weight loss: −5.9 kg by day 12 (≈ −7.5% of body weight; 95% CI 5.2–6.6 kg). Standardized within-subject effect d_z ≈ 2.9 (very large). But at one month only ~−2.8 kg (≈ −3.5%) was retained — roughly half regained. [Confidence: High]
- Insulin sensitivity: HOMA-IR fell −62% (1.78 → 0.67; d ≈ 0.97), insulin −49% (d ≈ 1.22). Fully reverted by M+1 (HOMA 1.68, non-significant). [Confidence: High]
- LDL cholesterol −28% (d ≈ 1.15) and triglycerides −35% (d ≈ 0.86) during fasting; both back to baseline at M+1. [Confidence: High]
- Blood pressure: systolic −8% (d ≈ 0.67). Not re-measured at one month, so durability is unknown. [Confidence: Medium]
- Biological age: No clock showed a robust, durable rejuvenation. The only longevity-relevant directional signal was a transient DunedinPACE decrease during the fast. [Confidence: Medium]
Practical translation for a biohacker: use prolonged fasting as a metabolic reset/diagnostic, not as a one-shot anti-aging therapy. Any lasting cardiometabolic benefit will depend on what you do in the weeks after refeeding, not the fast itself.
Source:
- Open access Paper /Preprint: Effects of long-term fasting: longitudinal epigenetic responses in humans
- Institutions: ETH Zurich (Lab of Nutrition and Metabolic Epigenetics) — lead; Buchinger Wilhelmi Clinic, Überlingen; Genknowme (Epalinges); University of Lausanne; UCLA (Steve Horvath); King’s College London.
- Countries: Switzerland (lead), Germany, USA, United Kingdom.
- Venue / “Journal”: bioRxiv — a preprint server. The manuscript states it “was not certified by peer review.”
- Impact Evaluation: The impact score of this venue is **Not Applicable / 0 (bioRxiv is a preprint server with no Journal Impact Factor or CiteScore)