Longevity & AgingResearch PaperOpen Access

Fasting Rewires the Brain — But Carries Hidden Psychiatric Risks

A 2025 review of 87 studies finds prolonged fasting boosts mood and cognition in healthy adults while triggering psychiatric crises in vulnerable individuals.

Saturday, May 30, 2026 0 views
Published in Nutrients
Cross-section illustration of a glowing human brain with ketone molecules and neural synapses lighting up during a fasting state

Summary

A 2025 narrative review synthesized 87 studies (39 human, 48 preclinical) on the psychological and psychiatric effects of prolonged and intermittent fasting. In metabolically healthy adults, supervised fasting and time-restricted eating (TRE) produced modest reductions in depressive symptoms, perceived stress, and anxiety, alongside small executive-function improvements. Neurobiologically, β-hydroxybutyrate (BHB) suppresses NLRP3 inflammasome activity, upregulates BDNF, enhances mitochondrial biogenesis, and recalibrates the HPA axis. Religious fasting (Ramadan, Orthodox Great Lent) showed similar mood benefits mediated partly by spiritual meaning. However, vulnerable individuals — including those with bipolar disorder, eating disorders, or psychosis risk — faced mood destabilization, manic or psychotic decompensations, and behavioral relapse. Methodological heterogeneity and short follow-up limit conclusions. The authors call for standardized psychiatric safety monitoring in future fasting trials.

Detailed Summary

Why this matters: Prolonged and intermittent fasting has moved from niche wellness practice to mainstream health intervention, yet its effects on the brain and mental health remain poorly characterized. This 2025 review addresses a critical gap — clarifying when fasting may serve as a neuropsychiatric tool and when it becomes a trigger for psychiatric harm.

What was studied: Authors Bonaccorsi and Romeo conducted a narrative synthesis of 87 studies (39 human, 48 preclinical) published January 2010 through June 2025, sourced from PubMed, Scopus, and PsycINFO. Human studies encompassed RCTs, cohort studies, pre-post investigations, and observational data covering TRE, alternate-day fasting, supervised multi-day Buchinger-type programs, and religious fasting (Ramadan and Eastern Orthodox Great Lent). Preclinical studies focused on neurobiological mechanisms in rodent models. Validated psychiatric instruments (PHQ-9, GAD-7, STAI-S, BDI-II, BAI, PSS) were required for human inclusion.

Key results: In metabolically healthy adults, short-term TRE and supervised prolonged fasting were associated with small-to-modest reductions in depressive symptoms and perceived stress, with a 2023 systematic review of 15 RCTs reporting Hedges g = 0.32 for depression. Supervised Buchinger programs consistently reduced STAI-S scores, lowered morning salivary cortisol, and increased high-frequency heart-rate variability — a marker of parasympathetic tone. Neurobiologically, BHB emerged as a pleiotropic signaling metabolite: it inhibits NLRP3 inflammasome-mediated IL-1β release via HCA2 receptors on microglia, acts as a class I histone deacetylase inhibitor to upregulate BDNF and PGC-1α, augments hippocampal GABAergic tone, and drives AMPK/sirtuin-1-mediated mitochondrial biogenesis. Gut microbiome remodeling — enriching butyrate-producing taxa like Roseburia and Faecalibacterium — further supports blood-brain barrier integrity and dampens neuroinflammation. Religious fasting showed comparable affective benefits, moderated by cultural context and perceived spiritual meaning. Adverse outcomes included mood destabilization, anxiety exacerbation, and rare psychotic or manic decompensations in at-risk individuals; individuals with eating disorder phenotypes showed increased food preoccupation and relapse risk.

Implications: For clinicians, fasting may be a valuable adjunct for mood and stress management in carefully screened, metabolically healthy adults under supervision. The neurobiological mechanisms — particularly BHB-driven anti-inflammatory and neuroplastic pathways — provide a compelling rationale for therapeutic applications. Cultural and spiritual dimensions of religious fasting represent an underexplored moderating variable with real clinical significance.

Caveats: Most human studies were small (n = 20–100), relied on self-report rather than clinician-rated instruments, and lacked follow-up beyond a few months. Methodological heterogeneity precluded meta-analysis. Observational data more frequently identified psychiatric harms than RCTs, suggesting selection bias. The authors strongly recommend future trials incorporate HDRS-17, CGI-S/CGI-I, standardized adverse-event tracking, and prospective psychiatric safety monitoring with ≥6–12-month follow-up.

Key Findings

  • Supervised prolonged fasting produced small but consistent reductions in depression and perceived stress (Hedges g ≈ 0.32 across 15 RCTs).
  • BHB inhibits NLRP3 inflammasome, upregulates BDNF, and enhances GABAergic tone — key anti-inflammatory and neuroplastic mechanisms.
  • Bipolar, psychosis-prone, and eating-disorder individuals faced elevated risk of manic episodes, psychotic breaks, and behavioral relapse.
  • Religious fasting (Ramadan, Orthodox Great Lent) modestly improved mood and stress, with spiritual meaning acting as a cultural moderator.
  • No current fasting trials use standardized clinician-rated psychiatric safety monitoring, a critical gap the authors urge future studies to address.

Methodology

Narrative synthesis of 87 studies (39 human, 48 preclinical) from PubMed, Scopus, and PsycINFO (2010–2025). Human studies required validated psychiatric/psychological instruments; quality was assessed via Cochrane RoB-2 (RCTs), Newcastle–Ottawa Scale (observational), and SYRCLE (animal). Quantitative meta-analysis was not performed due to heterogeneity in fasting protocols and outcome measures.

Study Limitations

Most human studies were small, short-term, and relied on self-report rather than clinician-administered psychiatric scales, limiting reliability and generalizability. Significant heterogeneity in fasting protocols, populations, and outcome measures precluded meta-analysis and cross-study comparison. Rare but serious adverse events (mania, psychosis) were primarily documented in observational rather than controlled settings, introducing ascertainment bias.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.