Health Biomarkers for Longevity: What's Actually Worth Tracking

This article covers biomarkers studied as markers or risk factors for healthspan and mortality in the general adult population; it does not address the interpretation of any individual person's results.

The short answer

A handful of biomarkers have decades of evidence linking them to healthspan and the risk of dying earlier. The strongest are not exotic. They are apolipoprotein B (ApoB) or LDL cholesterol, blood pressure, a measure of blood sugar such as HbA1c or fasting glucose, cardiorespiratory fitness (often expressed as VO2 max), grip strength, waist circumference, and high-sensitivity C-reactive protein (hsCRP). These are the markers that show up repeatedly in large cohort studies and, in several cases, in randomised trials where changing the number changed the outcome.

Much of what is sold in consumer longevity panels sits well below this tier of evidence. A test can be biologically interesting, accurately measured, and still tell you very little about how long or how well you are likely to live. The useful distinction is not "good marker" versus "bad marker" but how strong the evidence is, and whether moving the number actually moves your risk.

Why some biomarkers matter and most do not

It helps to separate two things that often get blurred together.

• A risk marker predicts an outcome. People with a higher value tend, on average, to fare worse. That is association, and it can still be confounded by other factors.
• A causal, modifiable target is one where lowering or raising it in a trial changes the outcome. This is a much higher bar, and far fewer biomarkers clear it.

ApoB and LDL cholesterol clear the higher bar. Mendelian randomisation studies, which use genetic variants as a natural experiment to reduce confounding, support a causal role for ApoB-containing lipoproteins in atherosclerotic cardiovascular disease [Richardson 2025], and lipid-lowering trials show that reducing these particles reduces events. Blood pressure clears it too, as we will see. Many trendy markers have never been tested this way, so we simply do not know whether chasing the number does anything.

For more on why prediction and proof are not the same thing, see our guide on how to read a clinical trial. It is the single most useful lens for judging any biomarker claim.

The biomarkers with the strongest evidence

ApoB and LDL cholesterol: the lipid story

LDL cholesterol has been one of the most studied cardiovascular risk factors for decades. ApoB is a refinement: because each atherogenic lipoprotein particle carries exactly one ApoB molecule, ApoB counts the number of particles that can lodge in an artery wall, which standard cholesterol measures do not. A large body of work, summarised in recent reviews, finds ApoB to be at least as accurate a marker of cardiovascular risk as LDL cholesterol or non-HDL cholesterol, and often more so, particularly when the two measures disagree [Marston 2025].

What sets this marker apart is the consistency across study types. Observational cohorts, genetic studies and randomised lipid-lowering trials all point the same direction, and the genetic evidence supports causation rather than mere association [Richardson 2025]. Targets are individualised by overall risk, so a single universal number is not appropriate, which is exactly why results should be interpreted with a clinician.

Blood pressure: the clearest case of a modifiable target

Blood pressure is the textbook example of a biomarker that, when changed, changes outcomes. In the SPRINT trial, adults at increased cardiovascular risk were randomised to a systolic target below 120 mmHg or below 140 mmHg. The intensive group had substantially lower rates of major cardiovascular events and lower all-cause mortality [SPRINT 2015], and the longer-term follow-up reported a durable mortality benefit [SPRINT 2021]. The trial also recorded more of certain adverse events in the intensive arm, which is part of why targets are set individually rather than pushed as low as possible for everyone.

This is what strong evidence looks like: a randomised intervention moving the biomarker and the hard outcome at the same time.

Blood sugar: HbA1c and fasting glucose

HbA1c reflects average blood glucose over roughly the preceding two to three months, while fasting glucose is a snapshot. Both are associated with cardiovascular disease and mortality, but the relationship is not a simple straight line. In the MESA cohort, both very low and elevated values were linked to worse outcomes, producing a J-shaped or U-shaped curve [Rhee 2019]. That nuance matters: it means lower is not automatically better, and that interpretation depends on context such as whether someone has diabetes.

As a longevity marker, glucose control sits in a strong-but-nuanced tier. The association is robust; the optimal range is population- and person-dependent rather than a single aspirational figure.

Cardiorespiratory fitness and VO2 max

Cardiorespiratory fitness, the body's capacity to deliver and use oxygen during sustained effort, is one of the most powerful predictors of mortality yet measured. A meta-analysis of healthy adults found that each one-unit (1 MET) increase in fitness was associated with a meaningful reduction in all-cause mortality and cardiovascular events [Kodama 2009]. A very large cohort undergoing treadmill testing found fitness inversely associated with all-cause mortality with no observed upper limit of benefit, and the least-fit group carried strikingly higher risk [Mandsager 2018].

The strength here is in the size of the effect and the consistency across populations. The caveat is that much of the data is observational, so part of the benefit may reflect underlying health rather than fitness alone, though the gradient is steep enough that fitness is widely treated as a genuine target. If you are weighing how to measure it, our piece on VO2 max: lab versus watch covers what consumer devices can and cannot tell you.

Grip strength

Grip strength is a simple, cheap proxy for overall muscular strength and, by extension, for biological resilience. In the PURE study of nearly 140,000 adults across 17 countries, every 5 kg lower grip strength was associated with a 16% higher risk of death from any cause and a 17% higher risk of cardiovascular death [Leong 2015]. It outperformed systolic blood pressure as a predictor of all-cause mortality in that analysis.

The important qualification is that grip strength is a marker of underlying health, not necessarily a direct lever. Squeezing a dynamometer harder is not the intervention; the strength it reflects is built through broader physical activity. It is a useful gauge precisely because it is so easy to measure and so consistently predictive.

Waist circumference and central adiposity

Where fat sits matters more than weight alone. Waist circumference and waist-to-hip ratio capture central (abdominal) adiposity, which is more closely tied to cardiometabolic risk than body mass index. An individual-participant meta-analysis of more than 82,000 people found that measures of central adiposity were associated with cardiovascular disease mortality, with waist-to-hip ratio showing the strongest association of the measures compared [Czernichow 2011].

As a longevity marker, central adiposity is well evidenced and almost free to assess with a tape measure. Like grip strength, it is best read as a window into metabolic health rather than a number to be gamed.

High-sensitivity C-reactive protein (hsCRP)

hsCRP measures low-grade systemic inflammation and is the most extensively validated inflammatory marker for cardiovascular risk. It predicts events even in people whose cholesterol is well controlled, capturing what is sometimes called residual inflammatory risk. Crucially, the causal step has partial trial support: in CANTOS, an anti-inflammatory drug that lowered hsCRP without changing cholesterol reduced cardiovascular events [Ridker 2017], and the size of the hsCRP reduction tracked the size of the clinical benefit [Ridker 2018]. That moves inflammation from association toward a plausible target, though the specific therapy was not adopted for broad prevention.

A quick comparison

Where consumer panels overreach

Many direct-to-consumer longevity panels bundle dozens of analytes, advanced lipid subfractions, niche hormones, micronutrient screens and proprietary "scores". Some of these are legitimate research tools. The problem is the gap between what is measured and what is known.

A few patterns are worth treating with caution:

• Markers with prediction but no proof. Plenty of biomarkers correlate with ageing or disease yet have never been shown to change outcomes when altered. Tracking them can feel productive while telling you little actionable.
• High biological variability. Some values swing widely day to day or with recent illness, sleep or exercise, so a single reading is easy to over-interpret.
• Composite "biological age" scores. These can be informative as research outputs, but methods vary and consumer versions are not interchangeable. We unpack this in how to test your biological age and what epigenetic clocks measure.

None of this means the newer markers are worthless. It means the evidence is younger and thinner, and they should not crowd out the boring, well-validated basics.

The over-testing trap

More data is not automatically better health. Testing a large panel guarantees that some values will fall outside the reference range by chance alone, which can trigger anxiety, repeat tests and downstream investigations that carry their own risks and costs. Chasing an individual number can also distract from the broad behaviours that move several of these markers at once.

A more measured approach is to focus on the small set of biomarkers with the strongest evidence, interpret them as a pattern rather than in isolation, and remember that the goal is healthspan, not a perfect spreadsheet. Our overview of healthspan versus lifespan and the broader longevity hub put this in context.

What this adds up to

The biomarkers worth the most attention are the ones with the deepest evidence: ApoB or LDL cholesterol, blood pressure, a blood-sugar measure, cardiorespiratory fitness, grip strength, waist circumference and hsCRP. A subset of these, blood pressure and lipids most clearly, are not just predictors but modifiable targets where intervention has been shown to change outcomes. The rest are strong markers of underlying health that are best improved through behaviour rather than gamed directly.

The newer, flashier panels are not necessarily wrong, but they are far less established, and treating them as equivalent to the validated basics is where consumer longevity testing tends to lose the plot. Track less, track what is proven, and read the numbers with a clinician who can place them in the context of you.

Related Proco pages

• Healthspan vs lifespan
• How to test your biological age
• What epigenetic clocks measure
• VO2 max: lab vs watch

Sources

Lab values are only meaningful in the context of your full clinical picture, so interpret any results with a qualified clinician rather than in isolation.