Americans Spend 14 Years Sick. Longevity Sells the Cure

Begin with the measurement. The Global Burden of Disease programme at the Institute for Health Metrics and Evaluation reported in July 2026 that people worldwide now spend 14.5% of their lives in poor health, up from 13.6% in 1990 ✓ Established Fact [1]. Life expectancy rose over that generation, but the years added were not added evenly: the morbidity gap grew by almost two years. Women carry more of it than men, 12.1 years against 9.3 [1]. The distribution across rich countries is the part the industry rarely quotes.

The United States has the largest national morbidity gap in the world at 14 years, followed by Australia at 13.9 and Canada at 13.7 ✓ Established Fact [1]. These are not poor health systems by global standards. They are the health systems that spend the most per capita, and they produce the longest tails of medicated, monitored, functionally limited old age. The gap is a wealth phenomenon before it is a medical one.

A second team reached the same conclusion by a different route. Armin Garmany and Andre Terzic, working from World Health Organization data across 183 member states, found the healthspan-lifespan gap rising from 8.5 years in 2000 to 9.6 years in 2019, a 13% increase in under two decades ✓ Established Fact [2]. The United States again sat at the extreme, 12.4 years, roughly 29% above the global average, and the gap for women ran 2.4 years wider than for men [2]. Two datasets, two methods, one direction of travel.

Against that measured problem stands a market of roughly the size one would expect. Global dietary supplement revenue sat between 196 and 218 billion dollars in 2025 depending on scope, growing at close to 9% a year, with North America taking about 36% of it [22]. Longevity clinics have spread from the United States to Switzerland, Singapore and Dubai, bundling genomic sequencing, multi-omics profiling, advanced imaging, immune assessment, microbiome analysis and epigenetic testing into annual memberships ✓ Established Fact [17]. Fountain Life prices its APEX membership between 19,500 and 21,500 dollars a year; Human Longevity packages can pass 50,000 dollars once concierge access, hormone therapy and peptide protocols are added, and insurance almost never covers any of it [23].

The research capital is larger still and differently motivated. Altos Labs launched in 2022 with three billion dollars, the largest biotech launch on record, to pursue partial epigenetic reprogramming [28]. Hevolution Foundation, created by Saudi royal decree and operating from Riyadh, aims to deploy up to one billion dollars a year into ageing biology and has already committed more than 400 million dollars across some 200 laboratories [27]. The XPRIZE Healthspan competition put 101 million dollars on the table and in August 2026 named 20 finalists, ten of them taking a million dollars each, for trials that must restore muscle, cognition and immune function in adults aged 50 to 90 by at least ten years [19].

The asymmetry that governs everything that follows is this. The problem is measured to a decimal place by public institutions publishing in the open literature. The remedies sold against it are measured, when they are measured at all, in trials of a few dozen people, in surrogate markers rather than outcomes, and frequently by parties with a financial position in the result [24].

This report grades the field on a single axis: what would survive a hostile reading of the evidence. Three bands emerge. Interventions with large randomised or dose-response human data and hard endpoints. Interventions with coherent mechanisms, animal data and small human trials that have not yet produced an outcome. And interventions sold at scale whose human evidence is either absent or actively contrary. The third band is the most profitable.

One more feature of the market deserves stating at the outset. The morbidity gap is widest in the countries that spend most on health — the United States at 14 years, Australia at 13.9, Canada at 13.7 [1] — and the supplement market is concentrated in the same places, with North America taking about 36% of global revenue [22]. The industry is not selling into medical scarcity. It is selling into medical abundance that has stopped converting spending into function, which is a harder problem and a better business.

In January 2023 Carlos Lopez-Otin and colleagues published an expanded hallmarks framework in Cell, raising the original nine to twelve: genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled macroautophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication, chronic inflammation and dysbiosis ✓ Established Fact [4]. The paper is the intellectual backbone of almost every commercial longevity claim made since.

What gives the framework its rigour is the admission criteria. To count as a hallmark, a process must appear with age, must accelerate ageing when experimentally accentuated, and must offer the opportunity to decelerate, stop or reverse ageing when therapeutically targeted [4]. The third criterion is the load-bearing one, and in humans it is the one with the least direct evidence. A hallmark qualifies on animal and cellular demonstrations; it does not qualify on a human trial with a mortality endpoint, because no such trial has been completed for any hallmark.

The geroscience hypothesis follows from the framework. If a single upstream process drives cardiovascular disease, dementia, cancer and frailty in parallel, then an intervention against that process should shift all of them together, which is a far larger prize than any disease-specific drug. That is the argument that moved three billion dollars into Altos Labs and a billion a year of intended Saudi funding into geroscience [28] [27]. It is a good argument. It is not yet a demonstrated one.

The most serious internal challenge comes from demography rather than biology. In Nature Aging in October 2024, Jay Olshansky and colleagues examined three decades of mortality in the eight longest-lived populations plus Hong Kong and the United States and found that improvement in life expectancy has decelerated since 1990 ◈ Strong Evidence [3]. Their projection is stark: without interventions that slow ageing itself, survival to age 100 is unlikely to exceed 15% for women and 5% for men in this century [3].

The reasoning is that the twentieth century's gains came from removing early and mid-life causes of death — infection, maternal mortality, cardiovascular events in middle age — and those gains are largely banked. What remains is the damage of ageing itself, which no current clinical tool addresses. Olshansky's recommendation is not despair but redirection: stop selling radical life extension and work on healthspan, the years lived in function [3].

The translation problem is the field's structural weakness. Interventions that extend lifespan in mice — caloric restriction, rapamycin, senolytics — do so in genetically uniform animals housed in sterile conditions, fed controlled diets, and measured over a two-year lifespan. Humans are outbred, live in noise, and take 80 years to produce an endpoint. A trial that measures human lifespan directly is not fundable, which is why every human geroscience trial to date has measured something else.

That constraint has a financial consequence that recurs throughout this report. The interventions easiest to finance are the ones with a proprietary molecule at the centre; the interventions with the best existing human evidence have no owner. The TAME trial, designed to test metformin against a composite of age-related disease endpoints, needs roughly 75 million dollars and has never enrolled a participant, because metformin is off-patent and no sponsor can capture the return ✓ Established Fact [9].

The field's most serious attempt to escape the endpoint trap is structural rather than scientific. XPRIZE Healthspan reframes the question as a one-year functional test: restore muscle, cognition and immune function in adults aged 50 to 90 by at least ten years, with a twenty-year target, and do it within twelve months of treatment [19]. Twenty finalist teams entered trials in 2026 with a grand prize of up to 81 million dollars in 2030 [19]. Whether or not a team wins, the competition will generate the first comparable body of one-year functional trial data the field has ever had.

The PEARL trial is the most rigorous test of rapamycin in normally ageing adults yet published. Investigators randomised 114 adults aged 50 to 85 to weekly 5 mg or 10 mg compounded rapamycin or placebo and followed them for 48 weeks [5]. The primary endpoint was change in visceral adiposity. It did not move: p = 0.942 ✓ Established Fact [5]. In a field that has spent a decade describing rapamycin as the most promising geroprotector available, the flagship human trial failed its own pre-registered test.

The secondary findings are real and narrow. Women on the 10 mg weekly dose gained lean tissue mass at 24 and 48 weeks, with a mean difference of 6.19 at the later timepoint and p = 0.018, and reported meaningfully less pain, with p below 0.001 [5]. General health scores improved in the 5 mg group. Adverse events were comparable across arms, though gastrointestinal symptoms were more common on rapamycin than placebo [5]. That is a safety signal and a sex-specific functional signal, not a demonstration of slowed ageing.

Caloric restriction has the cleanest randomised human dataset in the field. CALERIE assigned 220 non-obese adults to 25% caloric restriction or an unrestricted diet for two years, and Waziry and colleagues then applied DNA methylation clocks to the samples ✓ Established Fact [6]. DunedinPACE, which measures rate of ageing rather than accumulated age, slowed by roughly 2 to 3%. PhenoAge and GrimAge showed no significant change at all [6].

The honest reading of that result cuts both ways. A 3% slowing of DunedinPACE is associated with roughly 15% lower mortality risk in independent cohorts, which would be a large public health effect if it transferred [6]. But two of three clocks showed nothing, the participants achieved closer to 12% restriction than the prescribed 25%, and the endpoint is a biomarker whose own reliability is the subject of section four. CALERIE is the strongest evidence the field has, and it is a 3% move in one of three surrogate measures.

Senolytics occupy a thinner band again. The foundational human study of dasatinib plus quercetin was an open-label pilot in nine patients with diabetic kidney disease: a three-day oral course reduced adipose and skin senescent cell abundance eleven days later, with fewer p16 and p21 positive cells and less macrophage infiltration in fat ◈ Strong Evidence [8]. That is a genuine mechanistic demonstration in humans, and it is the entire randomised-adjacent foundation for a class of compounds now sold and prescribed on the promise of clearing senescent cells.

Metformin is the cheapest candidate and the least tested. The TAME trial has been designed, peer-reviewed and publicly advocated for a decade; it requires about 75 million dollars, has no commercial sponsor because the drug is off-patent, and has not enrolled a participant or listed a site ✓ Established Fact [9]. The field that raised three billion dollars for reprogramming in a single funding round [28] has not raised 75 million for the trial that would test its cheapest hypothesis.

The most interesting recent entrant was not designed as a longevity drug at all. Modelling from the SELECT cardiovascular outcomes trial estimated that semaglutide adds about 1.9 years of life expectancy and 2.0 event-free years in adults with cardiovascular disease and overweight but no diabetes ◈ Strong Evidence [26]. A separate study reported a 9% slowing of DunedinPACE on semaglutide, with improvement in mortality-linked PCGrimAge markers [25]. Whether that constitutes an effect on ageing or an effect on obesity and inflammation is genuinely contested ⚖ Contested [24].

What would settle these questions is not mysterious. It is a multi-centre randomised trial of a few thousand people over five to six years against a composite of incident age-related disease, which is precisely what TAME was designed to be and what the XPRIZE trials approximate on a shorter horizon [9] [19]. Until one of those reports, every claim in the commercial market rests on a 114-person trial that missed its endpoint [5], a 220-person trial that moved one surrogate [6], and nine patients with kidney disease [8]. That is the honest inventory.

Epigenetic clocks read DNA methylation at selected sites and map the pattern onto an age estimate. The prominent third-generation clocks are PhenoAge and GrimAge, which estimate accumulated biological age, and DunedinPACE, which estimates the current rate of ageing. They are legitimate research instruments: CALERIE used DunedinPACE as an endpoint, and the resulting 3% figure is the field's headline finding [6].

Sehgal and colleagues, publishing in Aging Cell, did what consumer testing companies do not: they took repeated measures from the same people under ordinary short-term perturbations, including meals, acute stress and pollution exposure [7]. Epigenetic age estimates fluctuated substantially. Most clocks reached only moderate reliability at best ✓ Established Fact [7]. A number sold as a readout of how fast your body is ageing turns out to be sensitive to what you did that morning.

The second finding is more damaging than the first. Technical reproducibility did not predict biological reliability. Clocks that returned near-identical values when the same sample was run twice were often the ones whose values shifted most when the same person was sampled twice — GrimAgeV2 and DunedinPACE were among the technically most robust and the biologically most fragile [7]. A vendor can advertise 99% reproducibility, be telling the truth, and still be selling a number that will not hold from one Tuesday to the next.

Reliability is not an abstract statistical property here; it propagates directly into conclusions. In the same analysis, clocks with higher reliability produced more stable associations with cognitive outcomes and more consistent responses to interventions, while less reliable clocks produced variable or outright misleading results [7]. Every published intervention effect measured on a fragile clock inherits that fragility, including some that the industry cites as proof of concept.

The clinical ethics literature has arrived at the same place from a different direction. Writing in the AMA Journal of Ethics in December 2025, clinicians warned that biological age results are readily misunderstood by patients, can cause psychological harm, and supply a veneer of objectivity to age-based discrimination ◈ Strong Evidence [32]. Their recommendation to companies selling these tests is straightforward: be far more transparent about reliability and limitations than current marketing is [32].

There is a defensible use for these instruments and it is not personal. As a group-level endpoint in a randomised trial, where noise averages out across hundreds of participants, a clock can detect a small shift that no feasible mortality study could — which is exactly what CALERIE did [6]. As an individual readout sold back to the person who paid for it, the same instrument is being used at a precision it does not have.

The distinction matters commercially because the consumer test is the entry point of the funnel. A number that says you are five years older than your birth certificate creates the demand for the supplement stack, the clinic membership and the peptide protocol. The instrument that generates the anxiety is the one whose reliability is weakest, and the interventions sold to fix the number are the ones with the least outcome evidence [17] [32].

There is a validated personal panel and it is unglamorous. The eight components of the cardiovascular health score — diet, activity, nicotine exposure, sleep, weight, lipids, glucose and blood pressure — are measured in any clinic, cost almost nothing, and carry an association with 8.1 additional years of life expectancy at age 50 [13]. They are the biomarkers with outcome data attached. The reason they do not anchor the longevity dashboard is not that they measure less; it is that they cannot be sold as proprietary.

The physical activity evidence is the most replicated finding in preventive medicine. Pooled cohort data published through the American Heart Association found that adults meeting the recommended 150 to 300 minutes of moderate activity a week had 20 to 21% lower all-cause mortality, and that 75 to 150 minutes of vigorous activity delivered 19% ✓ Established Fact [11]. Two to four times the recommended volume produced further reductions, with no signal of harm at the top of the range [11].

Resistance training is the underrated half. A meta-analysis in the American Journal of Preventive Medicine found that any amount of resistance training was associated with 15% lower all-cause mortality, 19% lower cardiovascular mortality and 14% lower cancer mortality ✓ Established Fact [10]. The dose-response is non-linear and peaks early: maximum benefit, a 27% reduction, at around 60 minutes a week, with the advantage diminishing beyond that [10]. Combined with aerobic activity, all-cause mortality ran 40% lower [10].

The composite measure is more striking than either component. Using the American Heart Association's Life's Essential 8 framework, adults with high cardiovascular health scores gained an estimated 8.1 years of life expectancy at age 50 against those scoring poorly: 7.5 years for men and 8.9 for women ✓ Established Fact [13]. Life expectancy rose roughly two years for every ten-point increment in the score, and 42.6% of the gain came from reduced cardiovascular death alone [13]. The remaining 57.4% came from everywhere else, which is precisely the multi-disease effect geroscience promises.

Sleep completes the set and is the most commonly traded away. A 2025 meta-analysis in GeroScience pooling 79 cohort studies through October 2024 found that sleeping under seven hours a night carried 14% higher all-cause mortality against a seven to eight hour reference, and that nine hours or more carried 34% higher mortality ✓ Established Fact [12]. The long-sleep association is partly reverse causation from underlying illness; the short-sleep association is harder to explain away.

Put the numbers side by side and the hierarchy is not close. Two years of 25% caloric restriction moved one surrogate biomarker by 3% [6]. The flagship rapamycin trial moved its primary endpoint not at all [5]. Cardiovascular health scoring is associated with 8.1 years of life at 50 [13], and an hour a week of resistance training with a 27% mortality reduction [10]. The interventions with the largest measured effects are the ones with no molecule, no membership and no margin.

The standard objection is that this evidence is observational and therefore confounded — healthy people exercise, rather than exercise making people healthy. The objection has force and does not survive contact with the detail. The associations are dose-responsive, consistent across populations and decades, present after adjustment for baseline health, and supported by randomised trials of exercise on intermediate outcomes. They are the strongest causal case available short of a mortality trial nobody will run [11] [10].

There is a second objection, more honest and more interesting: that behaviour is hard and pills are easy, so a pill with a 3% effect is worth more than an intervention with a 27% effect that people do not adopt. That is a real argument about implementation, and it is the argument the industry should be making. It is not the argument in the brochure, which claims the pill works better.

For scale, take the most effective pharmaceutical in the adjacent space. Modelling from the SELECT trial put semaglutide at about 1.9 additional years of life expectancy in adults with cardiovascular disease and overweight [26] — a major clinical result, achieved in a high-risk population, on a drug with billions behind it. The behavioural composite is associated with 8.1 years in the general population at age 50 [13]. The most successful metabolic drug of the decade delivers roughly a quarter of what the free intervention is associated with, and is prescribed with far more confidence.

Marco Demaria, writing in Aging in October 2025, described the model precisely: clinics combine genomic sequencing, multi-omics profiling, advanced imaging, full-body scans, immune assessment, microbiome analysis and epigenetic testing, and convert the output into individualised regimens that may include exercise and nutrition prescriptions alongside nutraceuticals, hormone replacement and more experimental therapies [17]. His verdict on the sector is that most clinics are pretty much disconnected from academic geroscience and from clinical geriatrics [17].

The pricing is public and consistent. Fountain Life lists its APEX membership between 19,500 and 21,500 dollars a year; Human Longevity packages pass 50,000 dollars once concierge access, hormone therapy and peptide protocols are bundled in; insurance almost never covers any of it [23]. Fountain Life reports that roughly 14% of members turn up an actionable, potentially life-threatening finding [23]. That figure is the sector's strongest argument and also its central problem.

It is a problem because a finding is not the same as a life saved. In September 2025 the Canadian Association of Radiologists issued a policy statement opposing whole-body MRI as a screening tool in asymptomatic individuals outside specific evidence-based indications, on the basis that no high-quality study demonstrates improved long-term outcomes or reduced morbidity or mortality ✓ Established Fact [14]. The American College of Radiology reached the same conclusion in 2023 [14]. No cancer society, radiology society or major medical body recommends the practice in healthy adults without risk factors [14].

The mechanism of harm is the incidental finding. Between 15 and 30% of adult diagnostic imaging studies contain at least one incidental abnormality, and more than 60% of patients with such a finding undergo further imaging, most of which yields little clinically useful information ✓ Established Fact [31]. Each step carries cost, radiation in some modalities, biopsy risk and durable anxiety. In a screened population that is mostly well, the cascade is the dominant outcome rather than the exception.

Further along the same corridor sit the regenerative offerings. A 2026 analysis in the Proceedings of the National Academy of Sciences documented hundreds of patients harmed by unproven stem cell interventions, including blindness, tumour formation and life-threatening infection, and argued that the existing US framework is already inadequate — before draft guidance issued in September 2025 on expedited review of regenerative products ◈ Strong Evidence [16]. Clinicians describe a parallel market of unapproved injections marketed as next-generation cell therapy [29].

The supplement layer has a measurable harm baseline that predates the longevity framing entirely. Geller and colleagues, using national surveillance data in the New England Journal of Medicine, estimated 23,000 emergency department visits a year in the United States attributable to dietary supplement adverse events, producing about 2,154 hospital admissions ✓ Established Fact [18]. Weight-loss and energy products drove a large share of cardiac presentations in adults aged 20 to 34 [18].

At the top of the market the protocol has become the product. Bryan Johnson spends about two million dollars a year on a personal regimen of more than 30 clinicians, over 100 daily supplements and continuous diagnostics, and has since commercialised it as a consumer line [30]. The commercial logic is exact: the protocol does not need to be shown to work in order to be sold, because the evidence standard for a supplement is not an outcome trial. It is a label [20].

None of this makes the clinic model worthless, and the distinction matters. Coronary calcium scoring in appropriate patients, lipid and glucose panels, blood pressure measurement, cardiorespiratory fitness testing and structured exercise prescription are all validated, and a clinic that delivered only those would be practising good preventive medicine [13]. The problem is the bundle: validated diagnostics are sold in the same package as whole-body screening the radiology profession opposes [14] and therapies Demaria describes as unproven or risky [17], at a single price that makes no distinction between them.

The American baseline was set in 1994. The Dietary Supplement Health and Education Act placed supplements under a food rather than a drug framework, requiring no pre-market approval of safety or efficacy and shifting the burden onto the regulator to prove harm after the fact ✓ Established Fact [20]. Everything downstream — the scale of the market [22], the harm baseline [18], the marketing latitude — follows from that single statutory choice.

The NMN episode shows the machinery in operation. The FDA had excluded nicotinamide mononucleotide from the supplement definition in 2022 on the grounds that it had entered drug investigation first. Industry petitioned, then litigated. In two letters dated 29 September 2025 the agency reversed itself, concluding that evidence of prior marketing defeated the race-to-market exclusion, and NMN returned to lawful sale ✓ Established Fact [20]. No new safety or efficacy data decided the question.

The European Union runs the opposite logic. Novel foods require authorisation before sale, and NMN has never received it. In May 2026 the European Food Safety Authority issued a positive safety opinion on beta-NMN with a proposed safe intake of 300 mg a day, which is a scientific finding rather than an authorisation; the Commission decision remains outstanding ✓ Established Fact [21]. Nicotinamide riboside chloride, the competing NAD precursor, cleared the same process in 2017 [21].

Japan occupies a third position and paid for it publicly. The Foods with Function Claims system, launched in April 2015, allows manufacturers to label health benefits on evidence they assemble themselves, without government approval [15]. In March 2024 red yeast rice supplements from Kobayashi Pharmaceutical were linked to roughly 3,000 health hazard events, 212 hospital admissions and five deaths from acute kidney injury ✓ Established Fact [15]. The Lancet's assessment was that the absence of a unified consultation system for health foods contributed to the delay in identifying the problem [15].

The clinic layer is regulated by geography rather than by rule. Demaria records the sector spreading across the United States, Switzerland, Singapore and Dubai [17], each jurisdiction applying its own standard to the same bundle of scans, biomarkers and off-label prescriptions. Where the regulator has the clearest authority is not the product but the claim: US advertising law requires competent and reliable scientific evidence for health claims regardless of whether the product needed approval to exist [33].

Advertising law is the binding constraint the sector consistently underestimates. The Federal Trade Commission requires competent and reliable scientific evidence behind health and disease claims, has acted repeatedly against unsubstantiated anti-ageing marketing, and treats undisclosed paid endorsement as a separate violation [33]. A supplement can be lawfully sold and unlawfully advertised in the same transaction, and the second exposure is the larger one.

The structural conclusion is that no jurisdiction regulates longevity as a category. Each regulates a component — the ingredient, the device, the scan, the claim — while the product being sold is the assembly of all of them into a narrative about time. That narrative is not a regulated object anywhere, which is precisely why the market forms around it [17] [20].

A coherent regime would separate three things the current rules conflate: whether a substance is safe to sell, whether it does what the label implies, and whether the clinical service built around it has an evidence base. Japan's post-2024 reforms moved on the first after five deaths [15]; EFSA's opinion on NMN addressed the first without touching the second [21]; US advertising law addresses the second but only after the product is already in the market [33]. The third — the clinic, the panel, the protocol — is governed almost nowhere, which is where the money has gone [17].

The first band is established. Physical activity at guideline volumes, resistance training at about an hour a week, adequate sleep, and the composite of cardiovascular health behaviours: pooled cohort evidence in hundreds of thousands of people, dose-responsive, with mortality endpoints [11] [10] [12] [13]. Effect sizes run from 14 to 40% on mortality and up to 8.1 years on life expectancy at 50. Nothing else in this report approaches that.

The second band is mechanistically strong and clinically incomplete. The hallmarks framework is coherent and productive [4]. Caloric restriction moved a validated surrogate by 2 to 3% in a proper randomised trial [6]. Senolytics cleared senescent cells in humans, in nine of them [8]. Rapamycin is safe at intermittent low dose over a year and produced sex-specific functional gains while missing its primary endpoint [5]. GLP-1 agonists show both modelled life-year gains and a biomarker signal [26] [25].

The third band is sold hardest. Consumer biological age testing at the precision implied by quarterly retesting [7] [32]. Whole-body MRI screening in people without symptoms or risk factors [14] [31]. Unapproved regenerative injections [16] [29]. NAD precursor supplementation as an ageing intervention rather than a biochemical one, where the leading proponents on both sides hold commercial positions [24] [21].

The contested ground is narrower than either camp admits. Both sides accept the hallmarks framework. Both accept that CALERIE moved DunedinPACE. The disagreement is about what a surrogate move licenses: whether a 3% shift in a methylation index is early evidence of a rate change or a measurement artefact inside an instrument with moderate reliability [6] [7]. That is an empirical question with an answer, and the answer requires trials that have not been funded.

The ownership test explains the funding pattern better than the evidence does. Rank the interventions by quality of human evidence and the order is exercise, sleep, cardiovascular risk management, caloric restriction, then everything else. Rank them by capital deployed and the order inverts almost exactly. The variable that predicts investment is not effect size but whether the effect can be enclosed in a patent, a membership or a subscription [9] [28] [23].

For a reader deciding what to do, the hierarchy follows from the numbers rather than from any position on geroscience. Guideline physical activity and about an hour of weekly resistance work carry the largest measured effects available [11] [10]. Sleep between seven and eight hours is the next [12]. Cardiovascular risk factors managed to guideline is the composite that produces the 8.1-year figure [13]. Everything after that is a wager on a surrogate, priced accordingly.

The longevity market is not a fraud and it is not a science. It is a sector selling a real and widening problem [1] [2] a set of remedies graded from excellent to absent, at prices that do not track the grading. The field's own leading demographer suggests the honest goal is healthspan rather than lifespan [3], and the best-evidenced route to healthspan has been known for thirty years and costs nothing to license [13]. That is the finding the sector is structurally incapable of selling.

The gap is real, widening and measured by institutions with no product to sell [1] [2]. The response with the best evidence is decades old, free, and associated with effects an order of magnitude larger than anything in the commercial band [13] [10] [11] [12]. The response with the most capital behind it has not yet completed a randomised human trial with a hard endpoint [4] [5]. Both things are true at once, and any honest account of the longevity market has to hold them together rather than choose one.