Why timing changes the story of melatonin in Alzheimer’s disease

A report that melatonin helps someone sleep, or even improves performance on a cognitive test, can sound like evidence that it slows Alzheimer’s disease. Those are related questions, but they are not the same outcome.

A 2024 review, Melatonin in Alzheimer’s Disease: Literature Review and Therapeutic Trials, brings together biological research, animal experiments and human clinical findings to examine that gap. Its overall picture is neither a proven treatment nor an empty hypothesis. Melatonin has plausible effects beyond sleep, and some studies report benefits, but the clinical findings are mixed and the studies do not use one consistent treatment schedule or one way of measuring change.

Sleep is only one part of the question

Melatonin is a hormone produced largely by the pineal gland in response to darkness. It helps coordinate daily rhythms, especially the sleep–wake cycle. Normal aging is associated with lower melatonin levels. The review reports that people with Alzheimer’s disease may have still lower levels, along with a weaker and less predictable rhythm. Melatonin loss in cerebrospinal fluid has also been observed early in the disease course and described as paralleling progression.

Those observations provide a reason to investigate melatonin. They do not show that the hormone changes cause Alzheimer’s disease, or that taking a supplement corrects the underlying disease process. They also leave open a practical question: is melatonin useful mainly because better sleep supports daytime function, or could it affect processes more directly related to neurodegeneration?

The review discusses several possible connections, but two examples show why researchers look beyond sleep. In a transgenic mouse model that overexpressed mutant amyloid precursor protein, melatonin treatment was associated with more soluble monomeric amyloid-beta 40 and less oligomeric amyloid-beta 40. The interpretation offered by the review is that melatonin may support clearance and inhibit aggregation in that model. The finding supports a possible role in delaying or preventing some processes, not in reversing established amyloid fibrils or proven disease damage in people.

The second example concerns oxidative and inflammatory stress. Melatonin can cross the blood–brain barrier and, in experimental systems, can affect free radicals, antioxidant enzymes and inflammatory responses. In rat brains exposed to amyloid-beta, one study described by the review found lower proinflammatory cytokine production and oxidative stress after melatonin treatment. A separate transgenic mouse study reported suppressed amyloid aggregation alongside antioxidant and anti-inflammatory effects. These are biological and animal findings, not demonstrations that a supplement slows Alzheimer’s disease in patients.

Why similar trials can disagree

The human evidence becomes more intelligible when the measurement itself is made visible. In a trial discussed by the review, people with Alzheimer’s disease and sleep disturbances received either 2.5 mg of extended-release melatonin or 10 mg. Actigraphy did not show improved sleep for either dose. Yet caregivers reported better sleep for the 2.5 mg group. An instrument recording movement and a caregiver describing observed sleep are not interchangeable measures. They may capture different parts of the same night, and they can produce different answers without either one being automatically useless.

Another trial described in the review used 2 mg of extended-release melatonin for 24 weeks alongside standard treatment for cognitive and sleep functioning. It reported improvements in cognitive performance and sleep maintenance, especially among patients who also had insomnia. A cognitive test asks how a person performs on particular tasks at particular times. It does not by itself show that amyloid, tau, neuronal loss or the long-term course of Alzheimer’s disease has changed.

The populations matter as well. The 24-week trial involved people with Alzheimer’s disease. Another study described in the review involved people with mild cognitive impairment, or MCI, who received immediate-release melatonin at bedtime for 9 to 18 months. That study reported better neuropsychological performance, sleep quality and wakefulness among those receiving melatonin than among the comparison group. MCI is not the same clinical state as diagnosed Alzheimer’s disease, so its findings can inform an early-intervention hypothesis without serving as a direct substitute for Alzheimer’s treatment evidence.

The review also notes that some studies report better sleep and day–night rhythmicity, while others do not. Even when cognitive improvement appears, researchers still have to ask whether the change follows better sleep, another biological action, or a mixture of both. The available examples therefore support careful interpretation rather than a single answer about efficacy.

The intervention changes with the clock

Melatonin is not simply a substance that can be compared by milligrams alone. Its effects depend on when it is taken, how quickly it is released and how long exposure lasts. The review describes the hormone as chronobiologically active: administration in the afternoon or evening can shift the body clock earlier, while morning administration can shift it later. The size and direction of the shift depend on timing and dose.

Release profile can change the exposure as well. The review cites evidence that a 2 mg extended-release tablet should remain intact or be halved to preserve extended-release behavior; quartering or crushing it produced effects more like an immediate-release preparation. That example makes the comparison problem concrete. Two studies may both describe “melatonin” while giving the body meaningfully different exposure patterns.

Age adds another layer. A small dose can produce blood levels near the range made naturally by the body, while commonly available supplements often begin at higher doses. Older adults can reach higher and more variable concentrations than younger adults, and higher doses may remain elevated into the following day. The review notes that this can contribute to daytime sleepiness and other disruption of the sleep–wake cycle.

These differences are plausible contributors to divergent findings, not a demonstrated explanation for every disagreement. Trials also vary in duration, participant characteristics, disease stage, coexisting sleep problems and the outcome used to judge success. A consistent research protocol would make results easier to compare, but the review does not establish one optimal dose, schedule, formulation or target population for Alzheimer’s disease.

The safety question extends beyond sleep

Melatonin’s activity outside sleep is part of its therapeutic appeal, but it also makes blanket claims about harmlessness unreliable. The review discusses a possible connection with insulin and glucose regulation. In rats with metabolic disturbances, melatonin treatment was associated with lower hyperinsulinemia and better learning and memory. Findings in women point in a less reassuring direction for some circumstances: studies described by the review found impaired glucose tolerance when melatonin was given in the morning or evening, including reduced insulin sensitivity in a study of post-menopausal women.

These results come from different species, populations and experimental conditions. They do not yield a simple rule for how melatonin affects metabolism in every person with Alzheimer’s disease. They do show why timing and proximity to meals may matter when researchers evaluate the treatment.

The review also discusses headache, dizziness, daytime sleepiness and hypothermia as possible adverse reactions. Diabetes, gastrointestinal conditions and liver conditions may be relevant when considering treatment, because a substance that influences several physiological systems may interact with a person’s existing health conditions. The article is not a dosing recommendation; it is a reminder that “available without a prescription” is not the same as “settled for this use.”

Why earlier treatment is being considered

Timing matters in two different senses. One is the time of day, which can alter circadian effects and exposure. The other is the point in the disease course. The review suggests that melatonin may be more useful before extensive Alzheimer’s-related neuropathology is established, because the proposed actions include supporting amyloid clearance, reducing aggregation and limiting oxidative stress rather than undoing damage that has already occurred.

This is why the review gives attention to mild cognitive impairment and prevention-oriented research. A study it describes by researchers at the University of Iowa was designed to test 5 mg of melatonin over nine months in older adults with and without MCI. It was intended to examine both cognitive function and Alzheimer’s-related biomarkers, including amyloid-beta and tau measures. The review described this as a pending randomized, placebo-controlled investigation, so its description supplies a research question rather than an outcome.

Putting cognitive measures and biomarkers in the same study could help clarify what kind of change melatonin produces. A better test score without a corresponding disease-related signal might suggest a functional or sleep-mediated effect. A biomarker change without meaningful cognitive improvement would raise a different question. Neither pattern alone would settle progression, but measuring them together is a step toward separating short-term performance from the longer course of disease.

What earlier treatment could mean for people at risk

The possibility raised by the review is that an inexpensive and accessible intervention might eventually give people with early cognitive difficulties more time before further decline. That would be a meaningful benefit only if longitudinal human research establishes a real effect on disease progression and identifies a regimen that patients can tolerate safely.

For now, melatonin’s mixed clinical findings, variable administration and unresolved biological pathway leave that possibility open rather than proven. The promise lies in testing whether early treatment can change the trajectory, not in treating improved sleep or a single cognitive result as evidence that it already does. The review supports further research, not self-treatment as a substitute for established Alzheimer’s care.

Key Sources

• Primary source: Melatonin in Alzheimer's Disease: Literature Review and Therapeutic Trials.
  DOI or Journal Page · Review or Meta-Analysis · 2024-01-01