Melatonin for Alzheimer’s: Why Dose and Duration May Matter

Melatonin may have a narrow role in Alzheimer's disease cognition, but the evidence does not support treating it as a general brain-protection supplement. A 2022 network meta-analysis separated melatonin trials by dose and treatment length and identified low-dose melatonin, taken for six to under 12 months, as its positive-finding treatment category in diagnosed Alzheimer's dementia. The paper's exact effect estimates, confidence intervals, rankings, and certainty assessments require confirmation before they can support clinical claims.

The distinction matters because older analyses often pooled very different protocols: a low nightly dose continued for months alongside higher-dose, short-term regimens. Those trials may all have tested melatonin, but they did not test the same intervention. The network analysis offers a plausible explanation for conflicting results. It does not settle the question of whether melatonin meaningfully improves life for people with Alzheimer's disease.

A hypothesis about dose and time

The 2022 network meta-analysis by Tseng et al. examined randomized, placebo-controlled trials in people with Alzheimer's dementia. It also included dose-stratified comparisons with approved Alzheimer's treatments: donepezil, galantamine, rivastigmine, memantine, and Namzaric.

Rather than combine every melatonin regimen into one estimate, the authors grouped treatment duration into four categories: under six months, six to under 12 months, one to under two years, and two years or longer. They defined low-dose melatonin as 2 to 2.9 mg per day and distinguished it from higher-dose regimens, such as 10 mg per day, whose effects may be dominated by pharmacological sleep promotion.

Within that framework, low-dose melatonin used for six to under 12 months was designated as the positive cognitive category. Higher-dose and short-term regimens were associated with null cognitive outcomes. The analysis therefore proposes dose and duration as effect modifiers: variables that may change the observed effect of treatment.

This is a useful hypothesis, not a finished answer. The available publication details do not establish the melatonin-specific MMSE effect size, its confidence interval, statistical significance, comparative rank against approved drugs, risk-of-bias assessment, or GRADE certainty. Without those details, readers should not treat the result as proof that a 2 to 2.9 mg regimen works or that it outperforms any established therapy.

Earlier null trials still count

Traditional pairwise meta-analyses that combined doses and durations found no significant cognitive benefit from melatonin in Alzheimer's disease. The network analysis does not erase those findings. It reframes them by arguing that a pooled estimate can obscure a regimen-specific effect.

A short-term 10 mg-per-day trial found no significant cognitive benefit, consistent with the network analysis's proposed pattern. That result may mean the dose, duration, or both were poorly suited to the effect the analysis proposes. It may also mean that the apparent low-dose, medium-term signal needs replication. One synthesis cannot distinguish those possibilities with certainty.

A prior subgroup analysis cited by Tseng et al. found an MMSE benefit in mild Alzheimer's disease but not in mild cognitive impairment (MCI). That population difference matters. A finding in diagnosed Alzheimer's disease does not automatically apply earlier in the disease process, when pathology, symptoms, and measurable outcomes may differ.

The mechanisms are plausible but unproven in people

Melatonin has credible biological links to sleep, circadian timing, oxidative stress, and neurodegeneration. Those links explain why researchers continue to study it. They do not identify the mechanism behind any human cognitive effect.

A 2023 mechanistic review by Reiter et al. describes a nighttime melatonin rhythm in cerebrospinal fluid and its proposed relationship to aquaporin-4-dependent glymphatic flow during sleep. Because this clearance system helps move brain waste products, including amyloid-beta and tau aggregates, age-related melatonin decline could plausibly matter. The review does not show that melatonin supplements increase amyloid-beta or tau clearance in living people with Alzheimer's disease or MCI.

Oxidative stress offers another rationale. A 2025 review of melatonin and Nrf2 signaling describes how melatonin may activate antioxidant transcription pathways in neurodegenerative disease. This is review-level mechanistic evidence, not a demonstration that Nrf2 activation improves cognition in Alzheimer's disease or MCI.

A 2025 mouse study found that melatonin improved glutamate homeostasis and synaptic potentiation through metabotropic glutamate receptor pathways in an Alzheimer's-like model. Animal results can generate a testable explanation; they cannot establish human dosing or clinical benefit. Likewise, a 2025 review of circadian disruption in neurodegeneration links impaired circadian pacemaking and reduced melatonin secretion with neurodegenerative disease, but does not show that restoring melatonin reverses cognitive decline.

The MCI trial asked a different question

The Alzheimer's analysis does not answer whether melatonin helps people with MCI. It excluded them. A 2021 trial protocol and its 2024 randomized, placebo-controlled feasibility trial tested 25 mg of melatonin daily for three months in adults with MCI.

The MCI investigators focused on feasibility, sleep outcomes, and cerebrospinal-fluid F2-isoprostanes, a biomarker of brain oxidative stress. The trial was designed as a feasibility study and was not powered to determine cognitive efficacy. Its value lies in showing that this type of pre-dementia biomarker study can be conducted; it does not establish that melatonin improves memory, daily function, or dementia risk.

Its protocol also falls outside the Alzheimer's network analysis's proposed positive category. Twenty-five milligrams is a pharmacological dose, and three months is short-term treatment. That mismatch means the MCI study cannot confirm or refute the low-dose, medium-term Alzheimer's hypothesis.

No adequately powered randomized trial in this evidence base demonstrates cognitive benefit from any melatonin regimen in MCI.

What the evidence supports—and what it does not

For diagnosed Alzheimer's disease, the research signal is specific: a network meta-analysis proposed that 2 to 2.9 mg per day for six to under 12 months may differ from short, higher-dose courses. The signal needs verification of the paper's detailed results and replication in contemporary, adequately powered trials before it can justify routine clinical use.

For MCI, prodromal Alzheimer's disease, healthy aging, and cognitive prevention, the evidence reviewed here does not support melatonin as a cognitive intervention. Plausible biology is a reason to run better trials, not a reason to assume protection.

The practical lesson is narrower than supplement marketing usually allows. Taking a consumer-dose melatonin product briefly for sleep does not resemble the regimen proposed by the Alzheimer's network analysis, and no evidence here shows that such use prevents cognitive decline.