This article is built from my weekly livestream, Neurofeedback and Chill, where I teach a topic and biohack live with the audience. This week I walked through new brain aging research out of Stony Brook University, explained why your brain's metabolic health starts shifting decades before any symptoms appear, and demonstrated a form of neurofeedback that trains blood flow directly. I taught gerontology at UCLA for over a decade, so this topic sits close to my work.
Why Does Brain Aging Start at 44 Instead of 70?
For years the story about aging put all the action at the end of life. You lived well, you aged gracefully, and decline arrived in your seventies or eighties. The new data complicate that timeline.
Researchers at Stony Brook ran a retrospective study on brain connectivity across more than 19,000 people (Mujica-Parodi et al., 2025). They found that communication between brain regions changes in specific age cohorts, and it changes in a stepwise way rather than smoothly. The first measurable drop in function shows up around age 44. From 44 to 59 there is a moderate decline. From 60 onward the decline gets steeper, until it levels off again near 90.
That pattern is an S-shaped curve with break points, not a straight line. The transition points sit at roughly 44, 67, and 90. The work was published in PNAS in March 2025 (Mujica-Parodi et al., 2025), a collaboration that pulled in scientists from Massachusetts General Hospital, Mayo Clinic, Oxford, and Memorial Sloan Kettering.
The reason this matters for you: the first inflection happens in midlife, while you still feel fine. If you are between 40 and 59, you are sitting inside the window where the brain is starting to strain metabolically but the neurons are still healthy enough to respond to support.
What Drives the Decline? Neuronal Insulin Resistance
When the researchers looked at what was pushing the curve down, they kept landing on the same mechanism. Inflammatory factors tied to insulin resistance were driving the change, not just riding along with it (Mujica-Parodi et al., 2025).
Here is the mechanism. Your neurons carry glucose transporters in their membranes. The insulin-sensitive ones, the GLUT4 transporters, pull blood sugar into the cell when insulin signals them. The mitochondria then burn that glucose to make ATP. That is the standard fuel pathway.
In insulin resistance, the signaling flattens out. Oscillation is life. When a signal like insulin or cortisol stays high and static instead of varying across a healthy range, the tissue stops responding. The transporters stop sucking sugar in efficiently. Now you have circulating glucose the cell cannot use well, and that extracellular sugar starts to do damage.
That damage is glycation, sometimes described as the tissue rusting. Sugar binds to proteins and creates advanced glycation end products, which drive oxidation and inflammation. We have known for fifteen-plus years that insulin resistance is a major factor in Alzheimer's, to the point that some researchers started calling it type 3 diabetes (de la Monte & Wands, 2008). The same oxidative process shows up in vascular dementia, in Lewy body pathology, and in the arteries as atherosclerosis. What is new is the timeline. The metabolic strain starts measurably eroding brain function at 44, decades before cognitive symptoms appear.
Loss of variable range is the through-line here. The same principle explains why chronically elevated cortisol drives hippocampal atrophy and drops plasticity (Sapolsky, 2000), and why you can lose most of your dopamine neurons before Parkinson's symptoms surface. The body can tune around a working range. Take away the oscillation and the system fails.
Why Do Ketones Help the Brain in Midlife?
After the cohort study, the Stony Brook team ran a feeding experiment. They gave groups in their 30s, 40s, 60s, and 90s metabolic challenges, either glucose or exogenous ketones, and watched brain function (Mujica-Parodi et al., 2025).
Glucose restored brain function in the youngest group, the people in their 30s. As age went up, glucose did progressively less. That confirmed the cohort findings.
The ketone results were the interesting part. The group that recovered the most brain function from exogenous ketones was the 44-to-59 cohort. The younger people did not change much, and the oldest group did not change much either. There is a midlife sweet spot for recovering metabolic flexibility with ketones.
The mechanism comes down to a second transporter. Ketones enter the neuron through the monocarboxylate transporter, which does not age the same way the insulin-sensitive glucose transporter does. Even after your insulin system has eroded from years of high, flat signaling, the ketone pathway stays largely intact for decades. Ketones let you sidestep the oxidative stress and inflammation that come with a high-sugar, high-starch diet.
Ketones do more than provide backup fuel. They are anti-inflammatory, and they get incorporated into structural components during repair, leaving those tissues somewhat resistant to further oxidation. I think of the repaired tissue a bit like kintsugi pottery, where gold thread runs through the mended cracks and makes the object more durable than before.
One honest caveat on the evidence. The strong signal in the Stony Brook work is specific to the midlife window, and the broader trial literature on exogenous ketones for cognition shows only modest overall effects, not support for taking them as prevention before age 44. That distinction matters when you decide whether this applies to you.
What Can You Actually Do in the 40-59 Window?
This is my call to action: learn your own metabolic health and learn to take control of it during the years when it pays off the most. The goal is compression of morbidity, squeezing the decline into the last short stretch of life so you keep quality and function for as long as possible. Dementia cases are projected to roughly triple by 2050 (GBD 2019 Dementia Forecasting Collaborators, 2022), so individual metabolic control is not a small thing.
You have three levers for managing energy flux, and each one is a different kind of partitioning.
Timing Your Food: Fasting Windows
The first lever is time, which is fasting. A 16:8 or 18:6 schedule, where you eat within a six- or eight-hour window and fast the rest, lets insulin rise during the eating window and drop fully outside it. That restored variability is the point.
Push the longer fasting window toward the end of the day if you can. Lower blood sugar overnight supports your circadian rhythm and lets you produce more growth hormone during sleep. I have written more on this in my piece on strategic fasting and on the minimum viable morning routine for circadian health.
Cyclical fasting works better than fasting every single day, especially if you are active or building muscle. A note on hormones: women in their 20s and 30s adapt to caloric stress so fast that aggressive fasting can crush metabolism and hormones. Post-menopausal women have more buffer and can run more aggressive protocols if they want.
Managing Calories Without Crashing Your Metabolism
The second lever is amount, which is dieting. There is a gotcha. Moderate caloric restriction tells the metabolism to slow down to match the lower intake, so it only works for a while, and younger bodies adapt fast.
My advice: run a meaningful deficit on some days, but do not stay in deficit at the weekly or two-week level. Refeed often enough to send the body the signal that resources are available. A constant restricted state combined with hermetic stress is a poor way to stay metabolically healthy. Deep, sustained fasting earns its place for specific reasons, like seizure control in epilepsy or spiritual practice, but not as a default weight-loss strategy.
People in their 40s and 50s should be building dense bone and strong muscle. That muscle becomes a metabolic sink that carries you through the next two or three decades.
Macronutrient Partitioning and Ketones
The third lever is what you eat, the macronutrient mix. Inside your eating window, prioritize protein. Protein is the signal for muscle protein synthesis and metabolic repair. Your body can convert protein to glucose through gluconeogenesis, but that is an on-demand process that fires when energy runs low, not a response to eating extra protein.
A point people miss about fuel: about 85 percent of the energy you burn at any moment is free fatty acids, not glucose or ketones. When you talk about going keto, you are working with the last 15 percent, dropping glucose and bringing ketones up as both a fuel and a signaling molecule.
If you train hard, do not run low-carb every single day. Endurance and heavy resistance work will strip the glycogen out of your liver and muscle, and then you bonk: shaky, foggy, no energy. Time the carbs you do eat for right after your hard workout, when depleted muscle and liver pull them straight in without spilling into the bloodstream. To measure ketosis, I have found a breath acetone meter the most useful tool, more than blood or other breath devices.
For the broader metabolic screening picture, Dale Bredesen's work reported that people with significant symptomatic dementia recovered meaningful function on multi-factor metabolic protocols (Bredesen, 2014). His programs now track a long list of metabolic factors, including methylation markers, homocysteine, hormones, and genetic variants. Managing the oxidative stress of blood sugar is powerful, and it is not the only factor.
What Is HEG Neurofeedback and Why Did I Demo It?
Since the whole topic is about delivering fuel and blood flow to the brain, I ran HEG instead of my usual EEG demonstration. HEG stands for hemoencephalography. It trains the vascular dynamics, the blood flow, rather than the electrical signal.
I use the passive infrared version, PIR HEG, invented by Dr. Jeff Carman. The headband holds two germanium lenses that pick up waves of heat coming off the front half of the brain. It is passive, so it emits nothing, and it is insensitive to movement, which makes it easy to use with people who cannot sit still, including kids on the spectrum.
The signal works like the BOLD response in fMRI. About one and a half to two seconds after neurons fire to do something, blood flow loads into that region. With HEG, when I focused or shifted into a positive mood, perfusion to my frontal lobe surged a couple seconds later, the tone rose, and the ball climbed. Unlike EEG neurofeedback, which is mostly involuntary, HEG has a voluntary component. You are doing vascular pumping.
In my own use and what the small HEG literature suggests, PIR HEG has been applied to migraines, brain fog, post-viral fatigue, concussion recovery, and to support social function in non-verbal autism. Carman's early work reported reductions in migraine activity within sessions (Carman, 2004). I also run it before EEG beta training, because warming up the brain's perfusion first seems to make the EEG session hit harder. I avoid HEG before alpha training.
A note on the near-infrared version, NIR HEG. It emits red light, usually somewhere around 800 nanometers, deep enough to penetrate the skull, and measures a much more focal signal. One audience member reported the near-infrared device was intensely stimulating with effects within seconds. Biofeedback learning takes minutes to develop, so my best guess is that was a direct red-light therapy effect, a localized metabolic surge in the mitochondria rather than a feedback effect. I stick with passive receivers for exactly this reason: fewer side effects and effects that come from a learning loop you can control. If you want the background on red light penetration depth, see my article on photobiomodulation.
If you want to understand the electrical side of brain training, my guides on whether neurofeedback is legitimate and on biohacking brain fog cover the mechanisms in more depth.
Where to Start This Week
The contrast that should motivate you is one you already know. Some 80-year-olds are frail, slow, and cognitively dim. Others are sharp, energetic, and quick. We seem to have real influence over which of those we become, and a lot of that leverage sits in the metabolic choices you make in your 40s and 50s.
If you are in or near the 40-to-59 window, start by learning your own metabolic numbers and your own response to food timing. Try a fasting window you can sustain, prioritize protein inside it, refeed instead of grinding yourself into a chronic deficit, and build muscle and bone now while it counts. If you want to watch your brain change over months and years, a QEEG brain map gives you a measurable baseline. Pick one thing this week, a single habit you have been meaning to start, and do it.