The core pharmacological problem is the oral route itself. Melatonin is rapidly absorbed from the gastrointestinal tract, but it undergoes extensive first-pass metabolism in the liver, leaving very little of the swallowed dose intact in circulation. A systematic review cited in the paper estimated average oral bioavailability at roughly 15 percent, but individual measurements vary enormously. In a landmark crossover study in healthy volunteers by Andersen and colleagues, a 10 milligram oral dose produced a median bioavailability of only about 2.5 to 3 percent, a peak plasma concentration near 3,550 picograms per millilitre at around 41 minutes, and an elimination half-life of approximately 54 minutes. That variability is compounded by real-world factors: age, smoking, caffeine intake, oral contraceptives, comorbidity, critical illness, inflammation, and interacting medications all reshape how a given dose translates into exposure. Two patients taking the same tablet may experience strikingly different internal drug levels.

Intravenous administration tells a very different story. Bypassing the gut and the liver’s first-pass machinery entirely, IV melatonin achieves complete systemic availability and far higher, more predictable concentrations. Healthy-volunteer studies show that a 10 milligram IV dose has an elimination half-life of about 39 minutes, a volume of distribution near 1.2 litres per kilogram, and clearance of roughly 0.022 litres per minute per kilogram, producing an immediate supraphysiological peak followed by rapid first-order elimination. Doses as high as 100 milligrams intravenously have been tolerated in controlled volunteer settings without measurable psychomotor impairment. Yet the review is emphatic that this pharmacokinetic advantage does not equal therapeutic superiority. Higher exposure is simply a different exposure condition; whether that exposure engages any clinically meaningful target remains an empirical question that no study has yet answered for most proposed indications.

The analgesic literature illustrates how mechanistic plausibility can mislead. Preclinical work implicates MT1 and MT2 receptors, opioidergic and GABAergic systems, NMDA signalling, ion channels, redox pathways, and inflammatory mediators in melatonin’s pain-modulating effects. Human findings, however, are inconsistent. Most tellingly, a randomized controlled study using a human inflammatory pain model found no significant analgesic, anti-hyperalgesic, or peripheral anti-inflammatory effect even after 10 or 100 milligram IV doses that generated concentrations far above physiological nocturnal levels. The authors treat this as a cautionary lesson: mechanistic plausibility plus high systemic exposure still cannot predict clinical benefit, because the right target may simply not be present, or the exposure may not reach the relevant tissue at the relevant time.

To bring quantitative rigour to the discussion, the team built simple one-compartment pharmacokinetic models using ordinary differential equations. For immediate-release oral melatonin, the model assumes first-order absorption and elimination and reproduces the short-lived concentration spike observed empirically. Under explicit assumptions, the model suggests that a hypothetical target concentration of 500 picograms per millilitre would map to an oral dose of roughly 1.4 milligrams, with allometric scaling pushing that estimate toward 1.7 milligrams in a 90-kilogram individual. For IV administration, where bioavailability is effectively complete, the same target concentration corresponds to a theoretical infusion rate calculated from published clearance values. The authors stress repeatedly that these are hypothesis-generating illustrations, not validated dosing recommendations; no clinically validated concentration threshold exists for any non-circadian indication, and the models deliberately ignore circadian phase, formulation-specific release kinetics, and individual pharmacodynamic variability.

The preclinical literature adds another layer of complexity. Recent in vitro studies from 2020 onward report that pharmacological concentrations of melatonin can reduce cancer cell proliferation and migration through NF-kB/MAPK and FAK/PD-L1 pathways, protect mitochondrial function, modulate SIRT1/PGC-1alpha, NRF2, GPX4 and NAD+ metabolism, and influence autophagy and ferroptosis in reproductive and other cell types. Animal models extend the picture: in a mouse Parkinsonian model, melatonin reduced MPP+-induced dopaminergic injury via heat shock protein 70 and autophagy, while other experimental work paradoxically found that enhanced melatonin signalling worsened motor deficits and that melatonin antagonism improved outcomes. Metabolic, inflammatory, sepsis, and pulmonary models similarly report heterogeneous, exposure-dependent effects. The unifying translational limitation is that the doses and concentrations used in these systems often cannot be reproduced by conventional oral regimens in humans, making the preclinical-clinical gap less a failure of biology than an unresolved pharmacokinetic-pharmacodynamic problem.

Against this backdrop, the clinical evidence organizes itself into a clear hierarchy. Sleep and circadian outcomes remain the domain where melatonin’s pharmacology, physiology, and trial results align most coherently. Across intensive care patients, night-shift workers, people with mild cognitive impairment, children with cerebral palsy or paediatric neurological disorders, and patients with glaucoma-associated sleep disturbance, oral melatonin has shown benefits consistent with its established chronobiotic actions, though results are not uniform and efficacy depends critically on administration time relative to circadian phase. Formulation matters too: immediate-release products suit short phase-shifting signals, while prolonged-release preparations better approximate nocturnal secretion patterns.

Outside the circadian sphere, the picture grows decidedly murkier. Perioperative trials examining preoperative anxiety, anaesthetic sparing, postoperative pain, and delirium prevention yield inconsistent results, with some studies reporting reductions in analgesic requirements and others showing no superiority over placebo; one randomized trial found oral melatonin did not reduce anxiety before elective hernia repair, and another reported no reduction in delirium severity among hospitalized older adults. Cardiovascular and renal studies offer encouraging but preliminary signals, including improvements in endothelial function and N-terminal pro-B-type natriuretic peptide levels in heart failure, and mixed results for acute kidney injury prevention in patients receiving polymyxin B. Critical-care and COVID-19 trials range from pilot studies to pragmatic quasi-experimental designs with incomparable dosing regimens. Most of these studies never measured circulating melatonin, making it impossible to distinguish genuine pharmacodynamic failure from inadequate exposure, poor formulation, or interindividual variability.

Neurological applications demand particular restraint. In Parkinson’s disease, human trials have mainly examined sleep-related endpoints: a small randomized trial improved subjective sleep quality without improving motor dysfunction, and a trial of prolonged-release melatonin for REM sleep behaviour disorder missed its primary endpoint while reporting mild headache, fatigue, and morning sleepiness. Combined with the contradictory animal findings, the review concludes that generalized claims of neuroprotection are unjustified and that disease-specific biology must take precedence over assumptions drawn from melatonin’s pleiotropic reputation.

The review’s central contribution is ultimately conceptual: it shifts the interpretive framework from milligrams administered to exposure achieved. Future studies, the authors argue, should prospectively link administered dose, formulation, route, timing, achieved plasma and where possible tissue concentrations, target engagement, and clinically meaningful outcomes, rather than treating a nominal melatonin dose as a transferable quantity across indications. Until such pharmacokinetic-pharmacodynamic data exist, oral melatonin retains its clearest role in sleep and circadian medicine, intravenous melatonin remains an investigational research tool for controlled high exposure, and the hormone’s broader therapeutic promise, however biologically intriguing, stays exactly that: a promise awaiting rigorous, exposure-aware validation.

Subject of Research: Pharmacokinetics, dose-exposure modelling, and clinical evidence for melatonin's non-circadian therapeutic applications

Article Title: Melatonin beyond chronobiology: integrating pharmacokinetics, dose-exposure forecasting, and clinical evidence

Article References: Pandolfi, S., Paone, F. M., Metalla, M., Ghezzi, C., Bjørlund, G., & Chirumbolo, S. (2026). Melatonin beyond chronobiology: integrating pharmacokinetics, dose-exposure forecasting, and clinical evidence. Molecular Biology Reports, 53(1), Article 1601. https://doi.org/10.1007/s11033-026-12765-z

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12765-z

Keywords: melatonin, pharmacokinetics, circadian rhythms, bioavailability, intravenous melatonin, dose-exposure forecasting, chronobiotic, sleep medicine, neuroprotection, perioperative medicine, anti-inflammatory, clinical trials