Weekly Issue — 2025-12-28

The shift is captured in a 2025 clinical rationale paper in the European Journal of Preventive Cardiology, which lays out the case that Lp(a) meets the established criteria used to judge whether any new test belongs in routine population screening. The authors argue that measuring Lp(a) once is no longer a niche specialist's tool but a piece of preventive medicine that should sit alongside the standard lipid panel — a position now endorsed by the European and Canadian lipid guidelines and by the National Lipid Association.

What changed is not the particle itself — Lp(a) was first described in the 1960s — but the strength of the evidence linking it causally to disease. Large genetic studies using a method called Mendelian randomization, which exploits the random inheritance of gene variants to mimic a randomized trial, have shown that people who inherit gene variants producing higher Lp(a) levels go on to develop more atherosclerotic cardiovascular disease and more calcific aortic stenosis. That signal holds across ethnicities, which matters because Lp(a) distributions differ substantially between populations.

Lp(a) looks superficially like LDL — the familiar "bad cholesterol" — but with an extra protein, apolipoprotein(a), wrapped around it. That structural quirk appears to make it more atherogenic per particle than LDL, and it also seems to promote inflammation and possibly clotting in vessel walls. Crucially, your Lp(a) level is roughly 80–90% determined by the gene you inherited. It barely budges with diet, exercise, or the statins that lower LDL. For most people, the number you have at 25 is, give or take, the number you'll have at 65 — which is exactly why a single lifetime measurement is enough.

The rationale paper frames the clinical stakes bluntly: in people with markedly elevated Lp(a), the lifetime cardiovascular risk is comparable to that of untreated familial hypercholesterolemia — a well-known inherited cholesterol disorder that cardiology has taken seriously for decades. The difference is that familial hypercholesterolemia gets flagged because LDL is on every lab panel. Lp(a), historically, has not been.

If Lp(a) can't be lowered by today's standard tools, why measure it? The honest answer is that knowing the number changes the aggressiveness with which the modifiable risks are managed. A person with high Lp(a) and a borderline LDL is not in the same risk category as a person with the same LDL and a low Lp(a). Guidelines increasingly treat elevated Lp(a) as a thumb on the scale — a reason to push harder on LDL targets, blood pressure, smoking cessation, and to look more carefully at family members who may share the inherited risk.

The rationale paper argues that integrating Lp(a) into global cardiovascular risk assessment is likely to generate health-system savings by enabling earlier, more targeted prevention, while reinforcing a direction that a growing number of clinical guidelines and consensus statements are already moving in. That's a moderate-strength claim worth taking seriously, but it is a projection built on screening-criteria logic and modeling, not yet on a completed population-screening trial.

Two honest caveats belong in any sober account of Lp(a). First, the causal genetic evidence is strong, but the precise risk increase tied to any individual's specific level — and how to weigh it against other risk factors in a given patient — is still an area of active refinement. Second, drugs designed to lower Lp(a) directly, including several RNA-based therapies, are in advanced clinical trials but have not yet been shown in completed outcome trials to reduce cardiovascular events. That is the missing piece that would move Lp(a) from "risk marker worth knowing" to "risk factor we can directly treat."

Until those trial readouts arrive, the practical value of screening sits in stratification and family awareness rather than in a specific Lp(a)-lowering prescription. That is a meaningful but bounded benefit, and it is the one the current guidelines are endorsing.

Lp(a) is not a new discovery, and the guideline shift is not a revolution. It is, more accurately, a long-overdue alignment between what genetic epidemiology has been saying for years and what clinicians are now being asked to do in practice. A single blood test, taken once, surfaces a piece of inherited cardiovascular risk that has been sitting invisibly on millions of charts. The treatment landscape will keep evolving; the case for knowing the number, the rationale paper argues, no longer depends on waiting for it.

The most interesting recent move in the measurement camp is a deliberate simplification. Writing in Scientific Reports, a team led by Meyer and colleagues built a clinical aging clock from routine blood biochemistry across 59,741 healthy samples in a Southeast Asian cohort, using just seven biomarkers — the kind of panel that already sits in millions of electronic health records. The pitch is not novelty for its own sake. It is translatability: a clock that runs on what your doctor already orders.

What sets the work apart is less the seven inputs than the math wrapped around them. First-generation clocks have long suffered from a systematic skew — they tend to overestimate youth in older subjects and overestimate aging in younger ones, muddying the very concept of "age acceleration" that makes the tool clinically interesting. The authors introduce a correction method designed to neutralize that bias, and they argue it sharpens the link between age acceleration and downstream disease risk.

The authors report that their clock predicts both self-reported and physician-annotated ICD-coded health outcomes, with elevated hazard ratios associated with accelerated biological age. Just as importantly, they tested it under stress: the predictions remained stable in the presence of acute infections and transient immune activation, conditions that have tripped up earlier inflammation-heavy clocks. To address the perennial criticism that aging clocks are trained on one population and then quietly assumed to generalize, the team validated their approach against both NHANES and UK Biobank data, an honest test of multi-ethnic robustness.

None of this makes biological age a finished clinical tool. It does, however, move the conversation from "interesting biomarker" toward "plausible preventive instrument." The interpretability angle matters here. A clock that can point at organ-specific aging processes — rather than producing one inscrutable number — gives a clinician something to do with the result, which is the gap that has kept earlier clocks stranded in research papers and wellness apps.

Measurement is only half the story. The other half is what to do when the number is high. That is where geroprotectors come in — molecules proposed to maintain homeostasis by acting on the so-called hallmarks of aging: genomic instability, telomere attrition, mitochondrial dysfunction, cellular senescence, and the rest of the now-familiar list. The field has long had candidates; it has lacked an efficient way to find more.

A second 2025 paper, published in the Journal of Cheminformatics, applies structure-based machine learning to screen natural products for geroprotector potential. The authors frame the problem plainly: age-related diseases and syndromes are a growing burden on healthcare systems, pharmacological interventions targeting aging itself have been proposed for years, and machine learning is reshaping drug discovery by making the early stages faster, cheaper, and more systematic. Their screen is a step toward putting those three threads together.

The honest caveat: this is candidate identification, not clinical proof. A natural product flagged by a structure-based model has cleared the lowest bar in a long ladder that includes biochemical assays, cellular work, animal models, and — eventually, for any compound that survives — human trials. The value of ML screens is in narrowing where to look, not declaring what works.

Each paper on its own is a useful brick. Together they sketch the architecture of preventive longevity medicine as it might actually be practiced. A robust, low-cost clock turns a vague concept — "how is this person aging?" — into a number a clinician can track over time and across organ systems. A pipeline of ML-prioritized candidates, if any of them survive rigorous testing, would eventually give that clinician something to do with an unfavorable trend other than recommend the usual lifestyle changes.

It is worth being precise about the strength of the evidence. The clock paper is a well-validated methodological advance, not a randomized trial showing that acting on a clock reading improves outcomes. The geroprotector screen is preclinical computational work, not a demonstration that any specific natural product extends healthy lifespan in humans. The field's history is littered with promising mechanisms — senolytics, NAD precursors, rapalogs — that have moved through this same arc with mixed, unfinished, and sometimes disappointing human results.

What is genuinely new is the discipline. Bias-corrected, multi-cohort, infection-robust clocks built on the panels doctors already run. Structure-aware ML screens that respect the messy reality of natural-product chemistry. The hype-to-substance ratio in longevity research is finally moving in the right direction. The numbers on the page are not yet a prescription. They are, increasingly, a credible map.

If you're lean, lifting hard, and metabolically healthy, fatty liver probably isn't on your radar. It should be — at least conceptually. Metabolic dysfunction-associated steatohepatitis (MASH, formerly NASH) is the silent comorbidity riding shotgun with type 2 diabetes and visceral obesity, and there's still no widely approved drug for it. That's the gap the new combo work is targeting.

In a double-blind, placebo-controlled phase 2b study, 31 adults with type 2 diabetes and MASH fibrosis (stages F1–F3) who were already on a stable GLP-1RA — semaglutide, dulaglutide, or liraglutide — were randomized to add efruxifermin (an Fc-FGF21 analog) or placebo for 12 weeks. The primary endpoint was safety and tolerability, and the answer was clean: the addition was safe and well-tolerated, with mostly mild-to-moderate GI side effects and only one discontinuation for nausea. Secondary endpoints tracked liver fat fraction, fibrosis markers, and metabolic parameters.

Read that carefully. This is a 31-person, 12-week safety study, not a fibrosis-reversal blockbuster. But it's the proof-of-concept that the field needed: you can layer a second metabolic peptide on top of a GLP-1RA without the wheels falling off, in exactly the patient population most likely to need both. That's the template for the next decade of peptide medicine — combinations, not solos.

Here's where it gets relevant to anyone who actually trains. A 2025 comprehensive review maps the neural circuits and peripheral pathways activated by both endogenous and pharmacological GLP-1, with a specific focus on what happens when you layer aerobic training on top.

The mechanistic picture the authors describe is the kind of thing that makes evidence-minded lifters sit up. Per the review, elevated GLP-1 levels — whether from overexpression models or training-driven signaling — are associated with higher skeletal-muscle glycogen, a shift toward endurance-oriented muscle fibers, increased mitochondrial content, and improved glucose uptake. The review also catalogs GLP-1's documented roles in endurance, muscle recovery, fiber-type distribution, muscle mass, and energy efficiency, alongside effects on bile acids, short-chain fatty acids, L cells, and G-protein-coupled receptors.

Important caveat: this is a review article synthesizing mechanism and animal-plus-human data, not a head-to-head trial of "GLP-1 plus zone-2 cardio" in humans. The authors themselves flag outstanding questions and future challenges in optimizing GLP-1R agonists for cardiovascular disease management. Translation for the gym-floor reader: the biology rhymes, but don't conclude that adding a GLP-1 to your training block is a documented performance stack. It isn't. Yet.

This one matters because it explains why some people "can't keep it off." A 2025 review in the International Journal of Obesity walks through what happens to endogenous GLP-1 after metabolic bariatric surgery — Roux-en-Y gastric bypass and sleeve gastrectomy in particular — and why those dynamics may dictate long-term outcomes.

According to the review, postprandial GLP-1 rises sharply after these procedures, but fasting GLP-1 doesn't change much. Observational data link higher postprandial GLP-1 to more successful weight loss, and when patients regain weight, adding a GLP-1RA has produced significant additional loss in the studies surveyed. The authors hypothesize that maintaining higher basal-bolus GLP-1RA exposure may be a promising option for patients who plateau or rebound after surgery.

The takeaway isn't "surgery fails." It's that the body's own GLP-1 response appears to be one of the levers that determines durability — and when that lever weakens, pharmacology may be a legitimate rescue rather than a cop-out.

Anyone who's done a weekly injection knows the friction. Native GLP-1 has an extremely short half-life and gets shredded by gut proteases, which is why current agonists are subcutaneous and engineered for persistence. A recent protein-engineering paper takes a swing at both problems at once.

The authors computationally designed trivalent fusion proteins that combine a protease-resistant modified GLP-1, a DARPin domain that binds human serum albumin (a known half-life-extension trick), and an approved cell-penetrating peptide. Molecular dynamics simulations and docking flagged one candidate — mGLP1-DARPin-Pen — as the most stable. They cloned it into E. coli, expressed and purified it, confirmed it by SDS-PAGE and western blot, and showed high-affinity binding to human serum albumin. Insulin secretion assays in mouse cells supported the bioactivity premise.

This is early-stage preclinical work — bench validation, not a human trial. But it shows the direction the field is engineering toward: a long-acting GLP-1 you could plausibly swallow. That would change adherence, access, and the conversation around these drugs entirely.

Step back and the four threads tell one story. First-generation GLP-1s proved the receptor was a legitimate metabolic lever. The current wave is about extending that lever: combining it with FGF21 to hit the liver, pairing it with training to hit cardiovascular and skeletal-muscle endpoints, using it to rescue post-surgical rebound, and re-engineering it so you don't need a needle.

For the evidence-first lifter, the right posture is the same one that's served you on every other supplement cycle: respect the mechanism, read the trial size, and don't confuse "promising direction" with "settled protocol." The peptide frontier is real. It's just still a frontier.