Everyone is selling you a brain hack for aging. Red light on your skull. Rapamycin in your morning stack. Exogenous ketones, cold plunges at dawn, NMN capsules at three dollars a pop. Most of it is noise. Some of it is early-stage science dressed up as clinical advice. A small number of things have real evidence accumulating behind them, and those are the ones worth walking through.
I'm Dr. Andrew Hill, a cognitive neuroscientist trained at UCLA in gerontology and developmental neuroscience. I've analyzed over 25,000 brains. What follows is seven categories of brain aging intervention, ranked by what the data actually supports, with the mechanism for each and an honest read on whether the evidence is strong enough to act on.
When should you start working on brain aging?
Most of this works longitudinally, not as an acute fix. The interventions that help your 60s, 70s, and 80s are the ones you start in your 30s, 40s, and 50s. If you are reading this in your 40s, you are getting a jump on it.
The goal in gerontology is compression of morbidity. You want to slide into the end of life with a strong machine and push any functional decline into the last short stretch, rather than spending decades slowly losing function. A handful of these interventions look trajectory-modifying, not symptom-modifying. That distinction matters, because most of what we know how to do in brain health changes symptoms without changing the underlying slope. The ones that bend the trajectory give you more healthy years, and that is the prize. We are not talking about solving aging or anti-senescence singularity claims here. Quality of years, not just years.
Does resistance training protect the aging brain?
Resistance training is the strongest item on this list. The 2011 Erickson study showing exercise grows hippocampal volume is the one everyone cites. Here is the update.
A 2025 meta-analysis in Frontiers in Psychiatry combined 17 randomized controlled trials, more than 740 adults over 60. Resistance training significantly improved working memory, verbal learning, and spatial memory. The effect sizes were moderate and not subtle: 0.44 for working memory, 0.63 for spatial. A 2026 Bayesian network meta-analysis of 38 studies and over 4,000 people put resistance training head-to-head against aerobic exercise. Resistance came in at 0.62, aerobic at 0.58. Roughly the same.
You cannot skip the weight room because you are a good walker or you own a standing desk. You need to load the machine with enough resistance to produce hormetic stress, the environmental press that triggers a healing response. The mechanism runs through several pathways: BDNF, lactate signaling, IGF-1, and vascular adaptation. This is especially useful for older adults who cannot tolerate high-impact aerobic work, because squats, deadlifts, and rowing are easier on the joints.
The practical dose from these studies: twice a week, 45 minutes, progressive load, with a 12-week minimum to see cognitive benefit. One caveat: effects on processing speed and executive function washed out in the larger meta-analysis, which often happens when you pool studies, because the effect is not uniform across individuals. Walking and some aerobic work help too. If you want the broader framing on training cognitive capacity, I've covered it in Biohacking Intelligence.
Does when you eat change your brain age?
Time-restricted eating, sometimes called intermittent fasting in the biohacking world, looks like it affects measurable brain age. The early-window version, where you stop eating in the mid-afternoon, outperforms the simple 16:8 restriction.
A 2025 randomized controlled trial in Frontiers in Aging studied 80 men aged 50 to 70 with metabolic syndrome on an early 8-to-10-hour feeding window. After only four weeks, MRI showed a reduced brain age gap, with increased gray matter in the hippocampus, thalamus, and substantia nigra, the dopamine-generating region, plus improved recall. A companion paper in Frontiers in Nutrition found increased fractional anisotropy, a marker of white matter integrity, with greater connectivity in the anterior thalamic radiations. White matter remodeling in four weeks is structural brain change, not metabolic reshuffling. A separate 2025 RCT ran 12 weeks of early time-restricted eating in adults averaging 59 with mild cognitive impairment. MoCA scores and recognition memory improved, and adherence was very high even in that population.
If you are metabolically healthy, a 14:10 or 16:8 window can make sense. My standard recommendation: do not eat in the first hour after waking, let your body burn off the cortisol and blood sugar that woke you up, then stop eating mid-afternoon so you have six to eight hours of clearance before sleep. That improves circadian entrainment and likely supports growth hormone release.
One important caveat on sex differences. The early-window study was all men. Reproductively cycling women are far more sensitive to caloric restriction, drops in dietary protein, and underfeeding. If you are a premenopausal woman, aggressive early time-restricted eating at a week-long deficit is likely too much. Cycle it instead. Run one or two heavier fasting days and replenish across the week so your body does not adapt too hard, too fast. Postmenopausal women and men tolerate the hormetic demand better and tend to land in a sweet spot for this around 50 and above. The full protocol logic lives in Strategic Fasting.
Can neurofeedback help with cognitive aging?
Neurofeedback is operant conditioning for your brain. We place a few sensors on the scalp, measure the brain regions involved in switching attention, sustaining focus, and producing deep sleep, and reward the brain with audio and visual feedback when it shifts in the right direction. You cannot feel your brain waves, so it works as an involuntary, delayed steering process. A few sessions in, the brain figures out what is being asked and the shift becomes more obvious. If you want the deeper mechanics, see Is Neurofeedback Legitimate?.
The workhorse protocol is sensorimotor rhythm training. SMR sits on the sensorimotor strip. You have seen it in a cat on a windowsill: liquid body, laser focus. During sleep it shows up as sleep spindles, which help you stay asleep through your partner moving or a dog barking. SMR helps pump the brakes on seizure activity and reduces fidgetiness.
A 2022 Dutch study trained SMR in adults with insomnia. When they separated out the people who actually changed their SMR, the gains went well beyond sleep: medium-sized improvements in attention, faster reaction time, fewer omissions on a continuous performance task, better working memory, and better Stroop and trail-making performance. The cognitive gains were independent of the sleep improvement. SMR moved cognition and sleep on their own tracks. That matters for aging, because sleep and attention tend to decline together. After 50, SMR on the motor strip may be your minimum viable protocol. Half of what I interleave into client sessions is still some form of SMR. More on the protocol itself in SMR Neurofeedback.
Alpha training shows effects too. A 2024 BMC Geriatrics study trained alpha in elders with age-related hearing loss, mean age 76. The group increased left frontal alpha power and improved on the Mini Mental Status Exam, on digit span forward, and, notably, on speech perception in background noise, with word recognition moving from fair to good. Alpha supports sensory gating, and training it up lets people use the auditory tissue they still have. Uncorrected hearing loss drives social withdrawal and understimulation, so this is a meaningful target. The mechanics of the alpha rhythm are in Decoding Alpha Waves.
A 2025 NeuroImage study added mechanistic depth. In patients aged 60 to 70 with mild cognitive impairment and memory complaints, neurofeedback changed how brain networks communicated, moving beyond symptom scores. Network entropy increased. Higher entropy and richer EEG microstate dynamics correlate with intelligence, so the brain was getting more complex, more youthful in its dynamics. And 93% improved their MoCA scores. A 2025 Frontiers in Aging Neuroscience paper looked at peak alpha frequency, your brain's idling speed. Faster peak alpha relative to your age correlated with better MoCA scores. On a QEEG you can also see that the language-dominant left hemisphere needs to stay synchronous within its alpha range, and when it desynchronizes you get word-finding trouble and tip-of-the-tongue states that feel like a short-term memory problem but are really a timing problem.
If you are over 50, this is the argument for getting a QEEG brain map. It shows your peak alpha frequency, your SMR frequencies, and how well you regulate sleep, and it tells you which interventions to prioritize.
Which supplements actually help brain aging?
Most supplements sold for brain aging do little. Three are worth your attention.
Omega-3s, especially DHA. A 2025 Scientific Reports meta-analysis combined 58 randomized controlled trials. Around 2 grams a day of total omega-3s with roughly half a gram of DHA produced a solid effect across multiple cognitive domains. They move several systems modestly, which is exactly what you want when you are trying to bend a trajectory over decades. Sourcing matters. Flavored omega-3s tend to be more oxidized than unflavored, and refrigerated products are less oxidized than shelf-stable ones. There is a newer choline-bound form, LPC-DHA, that crosses the blood-brain barrier at higher doses, but I have seen a few clients develop significant brain fog on it and the oxidative quality data is thin, so caution is warranted there for now.
Magnesium L-threonate. A 2025 double-blind, placebo-controlled study of 100 people in Frontiers in Nutrition using Magtein improved performance on the NIH cognitive battery, with working memory, episodic memory, and reaction time gains. The headline was an estimated 7.5-year reduction in cognitive age, alongside improved HRV and lower resting heart rate. The caveat is the sample was 18 to 45, not elders, so the result is promising but provisional for older adults.
Creatine. This is the most underappreciated of the three. A 2024 meta-analysis and a 2026 review converged: creatine helps cognition, not just muscle. It acts as a bioenergetic buffer, giving neurons more ATP reserve, which matters because aging neurons run with reduced metabolic headroom. The standard dose is 5 grams a day, cheap and well tolerated, though my own stomach does not handle it well. The gap is that there are only two active RCTs in adults over 60 right now, the Western University trial and the CREAGE trial. I expect creatine to become a commonly recommended compound for elders over the next few years.
Supplements fill gaps. They do not substitute for eating well, training, or sleeping.
What about the emerging longevity tools?
These are runners-up. The biology is interesting, the human outcome data is thinner.
Transcranial photobiomodulation (red light). I see it help some clients with brain fog and metabolic complaints, and do nothing for others. A modest, variable effect. More detail in Brain Biohacking with Photobiomodulation.
GLP-1 agonists (semaglutide, liraglutide). A 2025 JNNP study found neuroprotective signals against Alzheimer's, Parkinson's, and general decline, with reduced neuroinflammation, improved insulin signaling, and possible neurogenesis on imaging. A 2026 Baylor study showed metformin acting in the hypothalamus through a protein called Rap1, reframing these as brain drugs rather than only blood-sugar drugs. Human cognitive outcome data is sparse. If you are on a GLP-1 for metabolic reasons you may get the brain benefit along the way, but starting one purely for cognition is getting ahead of the evidence.
Urolithin A. This mitophagy activator prevented age-related cognitive decline in mice with early supplementation but not late, and a 2024 human trial showed improved mitochondrial function in immune cells at 1,000 mg a day. No cognitive outcome data yet. If it works, it is a start-early compound.
NAD+ precursors (NR, NMN). A pilot RCT gave 1 gram a day of nicotinamide riboside to adults with mild cognitive impairment. Blood NAD+ rose, epigenetic aging markers improved, but cognition did not move. A larger Norwegian trial, NAD-AGE, at 2,000 mg a day is ongoing with early promising signals. The biology is compelling, the cognitive payoff unproven. Background on the energetics is in NAD+ and brain metabolism.
The through-line on all four is metabolic and mitochondrial. Over the next few years we will learn which of these matters most and for whom.
What basics still win for the aging brain?
Sleep. Glymphatic clearance happens mostly during deep sleep. A 2016 Nature Communications study showed that increased non-REM duration moves amyloid and tau from the brain into the bloodstream for elimination in humans, and poor sleep meant poor clearance. Aim for 7 to 8 hours, treat any apnea, and protect deep sleep. Consistency of schedule matters more than the exact hours. The full protocol is in Biohacking Sleep.
Deliberate heat. A long-running Finnish study by Laukkanen followed over 2,000 men for decades. Four to seven sauna sessions a week was associated with a 66% lower dementia risk compared to once a week. The proposed mechanisms are vascular: blood pressure reduction, hormetic stress adaptation, heat shock proteins, and possibly BDNF. A 2025 study clarified that controlled heat helps but involuntary heat stress raises risk. The protocol: 15 to 20 minutes at 170 to 210°F, several times a week. Traditional dry sauna outperforms infrared, where you appear to need roughly double the time for a similar benefit.
Social connection. A 2022 UK Biobank study found social isolation associated with a 1.26 times higher dementia risk, partially mediated by frontotemporal gray matter loss. A 2025 meta-analysis found an effect size of 0.44 for isolation on cognitive decline. Social engagement is cognitively demanding work: reading faces, tracking conversations, managing complexity across time. It is also emotionally regulating. Isolation is neurotoxic at measurable, structural levels.
The hierarchy for protecting your brain after 50
Sort out exercise and sleep first. They are non-negotiable infrastructure. Add some early time-restricted eating and consider neurofeedback, both with good current evidence. Use supplements to fill specific gaps, with omega-3s, magnesium L-threonate, and creatine leading the pack. Watch the emerging tools, but do not bet on them yet.
One thing ties this together: knowing where your brain actually is. A QEEG shows your peak alpha frequency, your SMR frequencies, and how well you regulate sleep, and it tells you which of these interventions to prioritize. Treat it the way you treat a lipid panel. Get a baseline, learn how your machine runs against people your age, and act on what you find.