Weekly Issue — 2025-06-08

The case starts with IGNITE, a large observational study whose baseline analysis was published in the British Journal of Sports Medicine. Researchers measured cardiorespiratory fitness (CRF) directly — a graded treadmill test, not a self-report — in 648 older adults averaging nearly 70 years old, then ran a comprehensive neuropsychological battery and modelled five cognitive domains: episodic memory, processing speed, working memory, executive function and visuospatial ability. Higher CRF was associated with better performance across all five domains, even after controlling for the usual demographic suspects.

Two details from that paper deserve to be underlined. First, age and APOE4 status — the gene variant that raises Alzheimer's risk — didn't blunt the relationship, suggesting the fitness–cognition link held across people often written off as genetically fated. Second, the association was stronger in women, in people with fewer years of formal education, and in those taking beta-blockers, particularly for processing speed. That last group is interesting because beta-blockers blunt the heart-rate response to exercise; the fact that the cognitive correlation persisted there hints that what fitness does for the brain isn't reducible to one tidy mechanism.

IGNITE's baseline analysis is cross-sectional. It tells us that fitter older adults score better on cognitive tests at a single point in time; it cannot, on its own, prove that pushing your VO₂max up will pull your memory along with it. That's the gap a cluster of registered trials is now trying to close — and the reason this story is rated moderate rather than settled.

At Western University, a 26-week randomised trial is testing whether creatine supplementation and resistance training, alone and together, can move the needle on cognition, brain health and physical function in older adults with mild cognitive impairment. The 2×2 design — creatine or placebo crossed with weightlifting or a balance-and-tone control — is exactly what the field has been missing. Prior work has suggested resistance training helps the aging brain and that creatine, which naturally declines with age, supports muscle and bone. Whether the two stack in people whose cognition is already slipping is the open question.

A second trial, at Eastern Mediterranean University, is asking a more granular question: does how you move your hands matter? Finger movements activate a surprisingly large patch of cortex — sensorimotor strips, the supplementary motor area, Broca's, premotor and prefrontal regions — and asymmetrical, simultaneous two-hand movements have been shown to lift cognitive processes and cerebral blood flow more than one-handed work. The study compares bimanual drills, finger exercises and VR-mediated versions head-to-head in people with mild cognitive impairment. If a VR rig can deliver the same cortical activation as a clinician-led session, the implications for home practice — and for adherence — are obvious.

The third piece sits at the public-health end of the spectrum. Most exercise research has been done in urban, well-resourced cohorts. A trial out of the Universidad de Zaragoza is randomising 240 adults over 65 in rural areas to a 12-week, three-times-weekly multicomponent exercise program, followed by nine months of video-assisted sessions, with frailty, mental health and physical function as primary outcomes. It's the kind of pragmatic design that tells you whether a protocol survives contact with real lives, not just lab calendars.

The optimization-minded reader will already see the shape of the protocol. Cardiorespiratory fitness — the variable IGNITE measured — is built by sustained, structured aerobic work: zone-two volume, with intervals layered in. Resistance training, the lever being tested at Western, is what protects muscle mass, bone density and the metabolic environment your brain marinates in. The bimanual and VR work suggests that complex movement — coordination, novelty, asymmetry — may add something that pure cardio and pure lifting don't. And the rural-frailty trial is a reminder that consistency, delivered through whatever channel actually works for your life, is the variable that does most of the labour.

None of this is a prescription. The trials above are running; their endpoints aren't in. What the current evidence supports is a directional bet: a body kept fit across aerobic, strength and coordination domains appears to come with a sharper brain, and the mechanism is plausible enough that intervention trials are being funded to test it. That's the right register for this story — not a miracle, not a hedge, but a serious lead worth building habits around while the data mature.

For a desk that usually writes about skin barriers and posture, the throughline here is familiar: the most durable upgrades are the structural ones. Train the cardiovascular system that perfuses the brain. Build the muscle that defends metabolic health for decades. Add movement that is complex enough to demand cortical work. The face you present at 75 is partly a function of the work you do at 55; the mind behind it appears to follow the same rule.

The trial, listed on the U.S. clinical trial registry as NCT06948149, is the kind of design performance-science readers should recognize on sight: a 2×2 factorial that crosses a supplement (creatine vs. placebo) with an exercise mode (progressive resistance training vs. an active control of balance-and-tone classes). Four arms, one question: when an aging brain is already slipping, does the powder in the tub do anything on its own, does the barbell do anything on its own, and — the most interesting cell in the grid — do they do more together than apart?

For an audience used to thinking about creatine in milliseconds of phosphocreatine resynthesis, the cognitive framing requires a small mental gear shift. The neurons you are asking to remember a grocery list are not so different, energetically, from the type-II fibers you ask to clear a heavy single. Both run on ATP. Both can be bottlenecked when demand outruns supply. Creatine, in muscle, accelerates the recharge. The hypothesis under test here is whether the same chemistry helps in tissue that fires far more continuously and where the cost of an energy shortfall is measured not in a missed rep but in a missed name.

The case for taking creatine seriously in older brains rests on three pillars, none of them yet a load-bearing wall. The first is endogenous decline: creatine is naturally synthesized in the body and tissue stores fall with age, as the trial's own background notes. The second is the established benefit in older muscle and bone — creatine consistently shows up in the literature as a useful adjunct for lean mass and bone density when paired with training in older adults, the very substrate this trial is leveraging. The third, and the shakiest, is a thin but suggestive line of work hinting that creatine supplementation may nudge cognitive measures in older people, particularly under conditions of stress, sleep loss, or compromised baseline function.

That third pillar is exactly why an MCI cohort is the right place to push. Healthy young brains, like healthy young muscle, are creatine-replete and metabolically forgiving; an exogenous top-up has little headroom to work with. A brain operating with reduced reserve — the working definition of mild cognitive impairment — is the analogue of an athlete dropped into altitude. Supply constraints become visible. If creatine helps a brain anywhere, it should help here first.

The exercise arm matters just as much. Resistance training is one of the very few interventions with credible human evidence for improving cognition in older adults, and the trial's protocol doesn't cut corners: three 60-minute supervised sessions per week, progressive loading across major muscle groups, sustained over half a year. That is a real training stimulus, not a wellness gesture. The placebo-exercise arm — balance-and-tone classes — controls for the social, scheduling, and gentle-movement effects that any group exercise intervention inevitably carries, which is the kind of methodological hygiene that lets a result actually mean something.

The factorial structure is the elegant part. Comparing creatine-plus-training against placebo-plus-training isolates the creatine effect on top of a known cognitive intervention. Comparing creatine-plus-balance against placebo-plus-balance isolates creatine in the absence of meaningful loading — the closest this trial gets to asking whether a scoop a day, by itself, does anything for an MCI brain. And the contrast between the two combined arms tests for synergy: is the brain benefit of lifting amplified when the cell's energetic substrate is also topped up?

What the trial will not tell us, even at its best, is whether creatine prevents cognitive decline in people who do not yet have it, whether it changes long-term dementia trajectories, or whether a different dose, form, or duration would produce a different verdict. Twenty-six weeks is long enough to see meaningful cognitive change but short of the multi-year horizons on which neurodegeneration plays out. The outcomes — cognition, brain-health measures, and physical function — are appropriately broad, and the population is narrow on purpose. Read the eventual results as a sharply focused photograph, not a landscape.

The headline finding is large, and it is worth stating precisely. Across two randomized placebo-controlled trials totaling 1,852 non-diabetic adults with obesity, tirzepatide at 15 mg (or the highest tolerable dose) produced a mean percentage weight reduction 19.44% greater than placebo, with a 95% confidence interval of −22.48% to −16.41%. In absolute terms, that translated to roughly 17.55 kg more weight lost than placebo — about 39 pounds. For context, that is an order of magnitude beyond what diet-and-exercise trials reliably deliver at one year, and it is meaningfully larger than what older single-pathway GLP-1 agonists produced in comparable populations.

Individual trials are useful; pooled, pre-specified analyses of randomized trials are better. The authors of this review screened 281 articles and isolated only the studies that randomized non-diabetic adults to tirzepatide versus placebo — stripping out the diabetic populations that dominate the broader tirzepatide literature. That matters because glycemic status changes how the drug behaves and how patients respond. For the busy reader trying to triangulate whether the weight-loss numbers in the headlines apply to them, this is the apples-to-apples comparison that was missing.

The responder data is arguably more useful than the average. A mean is just a mean; what most people want to know is, "What are my odds of meaningful weight loss?" The meta-analysis reports that significantly higher proportions of tirzepatide-treated participants hit thresholds of ≥5%, ≥10%, ≥15%, ≥20% and ≥25% body-weight reduction compared with placebo. The ≥20% and ≥25% thresholds are particularly notable — territory historically reserved for bariatric surgery.

Weight loss is the headline, but the secondary outcomes are where the practical case for these drugs strengthens — or weakens, depending on what you value. The review reports improvements in glycemic and cardiometabolic parameters with tirzepatide versus placebo in this non-diabetic group, including lipid measures. That is consistent with what we'd expect when someone loses 15–20% of their body weight by any means, and it suggests the metabolic benefits are not gated on being diabetic to begin with.

The honest caveat: this is a two-trial pool. The effect estimates are precise enough to be confident in the direction and rough magnitude, but you should think of this as the strongest current read rather than the final word. More trials are running. The signal here is unlikely to evaporate, but the exact numbers will tighten.

Tirzepatide is not a free lunch. The meta-analysis found that significantly more participants on the drug experienced one or more adverse events compared with placebo. The adverse-event profile in the underlying trials has consistently been dominated by gastrointestinal complaints — nausea, diarrhea, constipation, vomiting — that tend to be worst during dose escalation and ease over time for most users, but not all. A meaningful minority discontinue because of side effects.

What the meta-analysis cannot tell you, because the trials it pools were not designed to answer it: what happens after you stop. The lean-mass question, the bone-density question, the rebound-weight-gain question, and the very-long-term safety question all remain genuinely open. If you are evaluating this drug, those unknowns belong on the same page as the 19.44% number.

Three things. First, if you have been telling yourself the weight-loss numbers you've seen for tirzepatide are inflated marketing, the randomized data in non-diabetic adults says otherwise — the effect size is real and it is large. Second, the cardiometabolic improvements suggest the benefit is not purely cosmetic; the same machinery that drops the weight appears to move the markers that matter for cardiovascular risk. Third, the adverse-event signal and the open long-term questions mean this is a conversation with a physician who can weigh your specific situation, not a checkout-cart decision.

The drug is not magic, and the data here is not a license for hype. But for the right person, working with the right clinician, the gap between tirzepatide and what came before is the kind of gap that genuinely changes the calculation.