Brain Energy and Aging: What Ketone Trials Actually Show

Brain energy may be more modifiable than age alone suggests. Higher metabolic syndrome severity tracks with a brain that appears to require more energy to shift between network states, while short trials in mild cognitive impairment (MCI) suggest ketones can change brain-fuel measures and some cognitive outcomes. Neither finding proves that a supplement prevents dementia. It does make cardiometabolic health and brain fuel use serious subjects, rather than wellness-world decoration.

Metabolic health may shape the brain's energy bill

The largest study in this group, Park et al.'s 2026 normative-modeling study, examined 25,697 UK Biobank participants ages 40 to 70. Using diffusion MRI, metabolic-syndrome phenotyping, network-control energy calculations, normative z-scores, and brain-age-gap modeling, the researchers asked whether metabolic syndrome severity related to the energetic cost of structural brain-network transitions.

Greater metabolic syndrome severity correlated with a higher estimated energy cost for those transitions and with markers interpreted as accelerated brain aging. Metabolic syndrome includes the familiar cluster of central obesity, high blood pressure, abnormal glucose regulation, and lipid abnormalities. The result links that cluster to an energetically less efficient brain phenotype at population scale.

It remains an association. The study cannot show that metabolic syndrome causes costly network transitions, or that lowering a person's metabolic syndrome score will reverse them. Shared causes, including lifestyle, illness, and factors that influence both vascular health and the brain, could explain part of the relationship. Still, the finding gives the brain a place in the usual conversation about glucose, blood pressure, and lipids.

Why glucose use and mitochondria are part of the story

Other human imaging research places impaired brain energy handling early in the Alzheimer's disease timeline. In cognitively normal adults followed longitudinally, Alhasan et al.'s 2025 FDG-PET study found that lower cerebral glucose metabolism predicted Alzheimer's disease progression pathways in competing-risks models. FDG-PET measures glucose uptake; it does not measure whether a dietary intervention can restore brain metabolism or avert dementia. Its value is prognostic: reduced glucose metabolism can appear before overt cognitive impairment in at least some trajectories.

A separate mechanistic human study, Terada et al. in 2022, linked mitochondrial complex I abnormalities to neurodegeneration and cognitive decline in Alzheimer's disease. Complex I helps mitochondria turn fuel into usable cellular energy. This supports bioenergetic failure as one plausible contributor to disease, but it does not establish that mitochondrial dysfunction starts the process or that a consumer supplement can correct it.

Together, these studies make a coherent case for testing brain-energy interventions. They do not identify one proven target. A brain scan can show a risk-related metabolic pattern; it cannot write a treatment plan.

Ketones have produced signals in mild cognitive impairment

The intervention evidence is more encouraging than definitive. In a six-month randomized controlled trial in people with MCI, Fortier et al. reported in 2021 that a ketogenic drink improved cognition. The trial tested alternative fuel delivery in a clinically relevant group already showing cognitive vulnerability. It did not test dementia prevention, long-term outcomes after treatment stopped, or effects in cognitively healthy adults.

A later report from the same clinical setting focused on white matter, the insulated nerve-fiber tracts that support fast communication between brain regions. Roy et al.'s 2024 secondary analysis found that a ketogenic supplement improved white-matter energy supply and processing speed in MCI. The pairing matters: increased ketone delivery to white matter provides a plausible route to a change in a specific cognitive domain. But processing speed is one outcome, not a declaration of broad cognitive recovery, preserved independence, or reduced dementia risk.

The newest trial added another ingredient. In the double-blind, six-month BREAK-AD randomized trial, Roy et al. in 2026 assigned participants with MCI in a 1:1 ratio to an active product or an isocaloric maltodextrin-and-mineral placebo. Each active sachet supplied 12 g of beta-hydroxybutyrate (BHB) and 550 mg of nicotinamide riboside chloride; participants took two sachets daily, for 24 g BHB and 1,100 mg nicotinamide riboside per day.

The active combination preserved fornix myelin density over six months. The investigators also reported associations between white-matter ketone uptake, myelin measures, and processing speed. Those are worthwhile biological signals, especially because the fornix is a white-matter tract involved in memory circuits. They remain surrogate outcomes. The trial did not establish reduced conversion to dementia, durable benefit after stopping treatment, or a definitive cognitive effect; the authors state that a larger sample is needed for definitive cognitive outcomes.

Why the supplement case remains unsettled

The combination trial cannot separate the effects of ketones from those of nicotinamide riboside. Nicotinamide riboside is an NAD+ precursor, and NAD+ participates in cellular energy metabolism, but this study tested the pair rather than either ingredient alone. A positive result for the combination is not evidence that nicotinamide riboside independently protects white matter.

Nor do favorable MCI trials settle the broader ketone question. A 2024 systematic review of medium-chain triglycerides for dementia-related diseases found a heterogeneous literature with mixed cognitive results, generally small samples, and limited durability data. MCTs raise ketones differently from a BHB-containing supplement, and studies have used different formulations, populations, and endpoints. Those differences help explain why a focused six-month MCI trial can show a signal while a broader review cannot support a blanket claim that ketones treat dementia.

The distinction is practical. Imaging markers can move without improving daily function. A statistically detectable change in processing speed may matter to patients, or it may be too small or too narrow to change how someone lives; the studies summarized here do not establish that answer. Likewise, a short-term MCI result cannot be assumed to apply to a healthy 45-year-old buying a ketone powder between meetings.

What to do with this evidence now

Treat brain energetics as a plausible research pathway, and treat ketone-plus-NAD+ products as experimental tools rather than established anti-aging therapy.

The most defensible action is also the least exotic: reduce cardiometabolic burden. Addressing elevated blood pressure, dysglycemia, central obesity, and abnormal lipids targets factors associated with higher brain network energy costs in the largest available study. That is not proof of brain rejuvenation. It is a sensible brain-health rationale for pursuing established cardiometabolic care.

For people with MCI, ketogenic strategies deserve discussion with a clinician who can account for medications, diabetes treatment, nutritional status, and the cause of cognitive symptoms. The trials offer a reason to study these interventions further, not a license to promise disease reversal. The current evidence supports an intriguing hypothesis: supplying alternative fuel may help a vulnerable brain compensate for impaired glucose use. It has not yet shown that this strategy prevents dementia.