Weekly Issue — 2026-03-15

Here's the line most people get wrong at the squat rack: "It works because you eat less, so you lose weight, so your blood sugar drops." Half right. A 2025 mediation analysis pooled three SURPASS randomized trials — 2,831 participants on tirzepatide versus placebo or semaglutide — and asked how much of the drug's HbA1c reduction was actually caused by the weight loss. The answer, depending on the comparator and the background therapy: surprisingly little. Against placebo as monotherapy, only roughly 12–27% of the A1c difference was mediated by weight loss; on a background of insulin, it climbed to 25–45%. The rest is the molecule doing molecule things — incretin signaling, insulin sensitivity, hepatic glucose handling.

Against semaglutide, the picture flips: 54–71% of tirzepatide's incremental A1c advantage was weight-loss dependent, which fits the clinical impression that dual GIP/GLP-1 agonism just drags more fat off than GLP-1 alone. Translation for the curious lifter: tirzepatide isn't just a better appetite suppressant. It's a better glucose drug independent of what the scale says — and most of its edge over semaglutide is the extra pounds.

Type 1 is a different beast. Beta cells are gone or going; insulin is the floor, not the ceiling. So the question of whether GLP-1s belong in the T1D toolkit has been live for years. A 2025 systematic review pulled together 25 RCTs and gave the cleanest answer yet. Versus placebo on top of insulin, GLP-1 RAs delivered a mean A1c reduction of 0.23%, a 3.93 kg weight drop, and a 5.74 unit/day cut in total insulin dose. Time-in-range didn't budge meaningfully, and the odds of severe hypoglycemia were similar between groups — important, because that's the failure mode everyone fears.

The honest read: this is a modest, real benefit for the right T1D patient, not a beta-cell rescue. The same analysis noted that adding a GLP-1 did not prevent progressive C-peptide loss, and glucagon counter-regulation to hypoglycemia stayed intact. You're not regrowing islets. You're trimming weight, shaving insulin, and nudging A1c — useful, not miraculous. And this is a clinician's decision, not a DIY one.

Here's the part of the conversation that matters in our world. If a drug strips 15–20% body weight, what's coming off — and what's staying on? A 2025 meta-analysis specifically zoomed in on T2D patients with sarcopenia across nine RCTs and 1,089 participants. The fat-side findings were unambiguous: GLP-1 RAs significantly reduced BMI, fat mass, and body fat ratio. Visceral fat moved in the right direction. Body weight overall trended down without reaching the threshold the authors set.

The muscle picture, per the authors' framing, is more Dionysus-and-Apollo than victory parade: signals are mixed, heterogeneity is real, and the dataset is small enough that anyone shouting either "GLP-1s spare muscle" or "GLP-1s gut muscle" is overstating what we know. The practical takeaway for the gym-literate reader: if a peptide is doing the calorie-deficit heavy lifting, the resistance training, the protein floor, and the sleep all matter more, not less. None of that is new science. It's just the part the influencers skip.

Two more 2025 entries pushed the GLP-1 story further off-label. The first: a retrospective database analysis of more than 2.5 million diabetic patients comparing metformin, GLP-1 agonists, sulfonylureas, and short-acting insulin for Alzheimer's risk. The headline result raised eyebrows — metformin use was associated with a statistically significant increased likelihood of AD diagnosis compared to GLP-1 use (HR 2.228), undercutting the long-running assumption that metformin was the neuroprotective champ. Diabetes itself raised AD risk versus non-diabetics. Retrospective data can't prove causation — channeling bias and confounding-by-indication are live concerns — but the GLP-1-favoring signal lines up with mechanistic work on incretin signaling in the brain. Worth watching, not worth prescribing on.

The second: a mechanistic mouse study showed liraglutide preserved endothelium-dependent vasodilation in ophthalmic arteries of septic mice, blunting oxidative stress markers and NOX2 overexpression. This is preclinical, in a single vascular bed, in a sepsis model — about as far from a clinical recommendation as research gets. But it adds to a coherent picture: GLP-1 receptor activation has vascular and anti-inflammatory effects that aren't explained by glucose or weight. Whether that translates to humans, in any tissue that matters, is the next decade's question.

The GLP-1 class is doing more than suppressing appetite — that part of the hype is grounded. Tirzepatide's glycemic muscle isn't all about the scale. The T1D add-on case is real but modest. The muscle question deserves serious tracking, not panic. The Alzheimer's and vascular signals are intriguing and early. None of it makes these drugs casual supplements. They're prescription medications with real effects, real side effects, and real unknowns — and the smartest thing a lifter can do with this evidence is read it carefully, train hard, eat enough protein, and bring any actual clinical question to an actual clinician.

Here's the setup. Researchers tracked 1,959 Chileans aged 60 and up between 2017 and 2019, watching how their cognitive states changed over time. Then they ran the data through something called a multistate model — basically a way of estimating not just how long people live, but how many of those years they spend with their thinking still sharp. That metric has a name: cognitive-impairment-free life expectancy, or CIFLE for short. Think of it as the brain-health version of "healthy years" on top of total years.

The big finding: higher education was associated with an average increase of 4.7 years of cognitive-impairment-free life, and 4.5 extra such years specifically for women. That is not a rounding error. That is a meaningful slice of someone's later life spent recognizing their grandkids, managing their own finances, and finishing the crossword.

The theory behind all this is called cognitive reserve. The shorthand: your brain builds up a kind of buffer over a lifetime of learning, problem-solving, and staying engaged. When age-related wear and tear shows up later, a bigger buffer means the damage takes longer to show on the outside. Education is the classic proxy for that buffer — not because a diploma is magic, but because years of school tend to track with a lifetime of mentally demanding work, reading, and social complexity.

What's new here isn't the theory. It's the math. Until recently, "cognitive reserve helps" was the kind of thing experts would nod about without being able to tell you, in years, what "helps" means. This study tries to answer that question directly — at least for one country, in one time window.

Here's the part I keep thinking about. Education is largely set by the time you're an adult — you can't retroactively go to more school in your 30s and 40s in any meaningful biological sense. So what about the people who didn't get those years? Are they just out of luck?

Apparently not. The researchers found that physical activity added 5.5 years of cognitive-impairment-free life expectancy for women with lower educational attainment — enough that the authors describe it as a "powerful compensatory effect" and a mechanism for narrowing the gender gap in brain-healthy years. In other words: movement seems to be doing some of the same protective work that schooling does, and it's available later in life.

Why might that be? The leading explanations are unglamorous and overlapping: better blood flow to the brain, lower vascular risk, improved sleep, less inflammation, more social contact when you walk with a friend or take a class. None of these are miracle mechanisms. They're the same boring levers that protect the heart — which, it turns out, are also the levers that protect the thing the heart is feeding.

One more piece of the study deserves attention, because it complicates the cheerful headline. The same paper reports that older adults in regions outside Chile's metropolitan area faced up to a 30% higher risk of cognitive impairment. National averages, the authors note, mask very different local realities.

That is a polite way of saying: your zip code is doing some of the heavy lifting that we like to credit to personal choices. Access to clinics, walkable neighborhoods, clean air, schools, social services — all of these shape both the education line and the physical-activity line in the data. Cognitive reserve isn't only a personal project. It's also, partly, an infrastructure project.

A few honest caveats, because this is one study and the evidence here is moderate, not settled. The data come from a two-year window in a single country, which is short for a question about decades-long brain aging. The protective factors are observational associations — the researchers can't prove that going for walks causes the extra years; people who exercise also tend to differ in income, diet, social connection, and baseline health. And the model estimates life expectancy from current patterns, which always smuggles in assumptions.

What the study does well is give a shape to something that's usually hand-waved. "Education and exercise protect your brain" is true but slippery. "Roughly 4 to 5 extra years of clear thinking" is something you can actually plan around.

The takeaway I'm sitting with: cognitive reserve isn't a vibe. It's a buffer you can measure, and at least in this snapshot of nearly 2,000 lives, the things that grow it are stubbornly familiar. Keep learning. Keep moving. And maybe pay a little more attention to the neighborhoods and systems that make those two things easier — or harder — for the people aging around you.

Metabolic syndrome is less a single disease than a cluster — a group of cardiometabolic risk factors that tend to travel together: elevated blood pressure, elevated fasting glucose, higher triglycerides, lower HDL cholesterol, and extra weight carried around the middle. None of those things sound dramatic on their own. Stacked, they meaningfully raise the long-term odds of type 2 diabetes and cardiovascular disease, which is why clinicians treat the constellation as a flag rather than a footnote.

The new Cureus review describes metabolic syndrome as a significant and growing global public health challenge, and notes that until now a comprehensive quantitative synthesis of its worldwide prevalence and the lifestyle behaviors tied to it had been surprisingly thin on the ground. That is the gap this paper tries to fill.

The headline figure from the 2026 meta-analysis is a pooled global prevalence of 32%, drawn from ten observational studies and 54,709 participants. That is a striking number, and worth saying out loud: in the populations these studies sampled, roughly a third of adults met criteria for the syndrome.

But — and this is the part that gets cropped out of most headlines — the 95% confidence interval ran from 22% to 43%. Heterogeneity between studies, measured by the I² statistic, was 98.72%, which is about as high as that number gets. In plain English: the studies varied enormously in where they were conducted, how they defined the syndrome, and who they included. The 32% is the best central estimate the math will give you, but the honest read is that prevalence is high, varies wildly by population, and is climbing in step with the lifestyle patterns of modern adulthood.

That uncertainty is not a reason to dismiss the finding. It is a reason to hold it with the right amount of grip — firmly enough to act on, loosely enough to keep questioning.

The most consistent signal across the included studies was about movement. Physical inactivity was consistently associated with increased odds of metabolic syndrome, with a representative odds-ratio range of roughly 1.25 to 2.15 across the studies the authors examined. Adherence to recommended physical activity levels, by contrast, tracked in the protective direction.

An odds ratio is not a prophecy. It is a way of saying: in these populations, people who moved less were meaningfully more likely to meet the criteria. It does not prove that a single morning walk reorganized anyone's triglycerides, and the review is explicit that this is observational evidence — associations pulled from real-world data rather than controlled experiments. That distinction matters, because observational studies cannot fully untangle the chicken-and-egg problem of whether inactivity drives metabolic dysfunction, whether metabolic dysfunction makes movement feel harder, or — most likely — both, in a loop.

Still, the direction and consistency of the finding are the kind of thing clinicians and researchers take seriously. When ten different studies in different settings keep landing in the same general neighborhood, the neighborhood is worth taking a walk through.

The review itself does not single out perimenopausal women, and we are not going to put words in its mouth. But the cluster it describes — creeping waist circumference, drifting blood pressure, glucose that does not behave the way it did at 28 — is a cluster a lot of women between 35 and 50 will recognize from their own annual labs. Hormonal shifts in this window change how the body stores fat, how it handles insulin, and how it responds to the same workout that used to work fine. None of that is a personal failing; it is biology recalibrating.

What the Cureus meta-analysis adds to that lived experience is a quantitative reminder that the modifiable inputs still matter, possibly more than ever. The studies it pooled were observational, the heterogeneity was high, and the evidence rating here is honestly moderate — but the through-line is clear enough to be useful.

The review does not hand readers a regimen, and neither will we. What it offers is a map: metabolic syndrome is common, it is rising, and the behaviors most strongly tied to it in the pooled data are the ones you already suspected. Sitting less. Moving more, in whatever form you will actually do for years rather than weeks. Knowing your own numbers — waist circumference, fasting glucose, blood pressure, triglycerides, HDL — well enough to spot a trend before it becomes a diagnosis.

If two or three of those numbers have been quietly drifting on your last few annual physicals, that is the conversation worth having with your primary care clinician. Not a panic conversation. A map-reading one.