All worlds
ELEVEN WORLDS · 11 · DIAGNOSTICS

Diagnostics & BiomarkersEvidence 2026

Before any stack, method or protocol makes sense for you, you need a measurement. This world shows the biomarkers with the most robust evidence, where the optimal range diverges from the lab range — and how we turn measuring, intervening and re-measuring into a closed loop.

Reviewed

What this world covers

This world is different from the ten before it. The other worlds are about what you take or do — substances, methods, protocols. Here it's about what makes everything else meaningful in the first place: measuring. Diagnostics and biomarkers are the feedback loop of biohacking. Without them you optimize blind.

This world bundles the diagnostic values with the best evidence and the greatest practical leverage. Four axes structure the field: metabolism (blood glucose, HbA1c, fasting insulin and HOMA-IR, continuous glucose monitoring), cardiovascular risk (lipid profile, ApoB, Lp(a)), silent inflammation and methylation (hs-CRP, homocysteine) and the function and status markers (thyroid panel, liver values, kidney values, complete blood count, ferritin). On top of that come the dynamic markers from wearables and tracking — VO2max, resting heart rate, heart rate variability, body temperature, cycle — which don't come from the lab but serve the same purpose: feeling becomes data.

Why measuring comes before optimizing

There is one principle this world places above all others: You can't optimize what you don't measure. The other ten worlds tell you what works on average across a study population. This world tells you what works for you — and that can only be found out through measurement.

The most common mistake in biohacking is to start with the intervention instead of the measurement. Someone takes an expensive supplement for six months, "because studies" — without ever having measured the marker it's supposed to move, neither before nor after. Whether it changed anything for them remains a matter of faith. Yet the solution is simple and costs a fraction of the supplement budget: measure, intervene, re-measure. Only this closed loop turns a hypothesis into a result.

The principle works in the other direction too. Many effects attributed to expensive substances are in truth the correction of a measurable deficit — a vitamin D level that's too low, a homocysteine value that's too high, an unrecognized insulin resistance. Whoever measures first finds these levers and saves themselves the hunt for effects that aren't happening in their own body at all.

The most important markers

Metabolic axis: HbA1c, insulin & HOMA-IR

Metabolism is the axis where it shows earliest whether an optimization course is on track. HbA1c reflects the average blood sugar of the past roughly three months via the glycation of red blood cells — a robust, well-established value. But it reacts late.

The earlier marker is fasting insulin and the HOMA-IR calculated from it. Insulin often rises for years before blood sugar derails, because the body initially keeps glucose in check with more insulin. Anyone who measures only glucose or HbA1c doesn't see this silent phase. Continuous glucose monitoring (CGM) rounds out the picture with the dynamics: how blood sugar responds to individual meals, stress and sleep. Studies on personalized nutrition show that this response varies strongly between individuals — what's harmless for one person drives a spike in another.

Cardiovascular axis: ApoB instead of LDL-C

For cardiovascular risk the better marker has prevailed: ApoB. Every artery-damaging lipoprotein carries exactly one ApoB molecule, so ApoB measures the actual particle count — and thus what burdens the arterial wall. Classic LDL cholesterol only estimates the cholesterol content and can underestimate the risk when the two values diverge, for instance with insulin resistance or high triglycerides. The full lipid profile stays useful, above all the ratio of triglycerides to HDL as a rough insulin-resistance indicator. Lp(a) is largely genetically determined and should be measured at least once in a lifetime.

Inflammation & methylation: hs-CRP and homocysteine

Silent, low-grade inflammation is a common thread through almost all chronic diseases. High-sensitivity CRP (hs-CRP) makes it measurable: under 1 mg/l is considered favorable, over 3 mg/l elevated. Homocysteine is both an inflammation and a methylation marker — elevated values are associated with cardiovascular and cognitive risks and respond well to the B vitamins B12, folate and B6. An optimal range sits closer to under 8 µmol/l, considerably narrower than the lab range.

Function & status: thyroid, liver, kidney, blood count

The thyroid panel needs more than just TSH: only free T4, free T3 and, if suspected, the antibodies paint the full picture, because a normal TSH can mask weak T4-to-T3 conversion. Liver values — especially GGT — are not only an alcohol but also a metabolic and oxidative-stress marker. Kidney values (creatinine, eGFR, ideally cystatin C) make sure that interventions don't overload the excretory system. The complete blood count plus ferritin uncovers anemias, iron status and a range of silent problems.

Dynamic markers: VO2max, resting heart rate, HRV

Not every biomarker comes from the lab. VO2max — cardiorespiratory fitness — is one of the strongest single predictors of all-cause mortality there is, and directly trainable on top. Resting heart rate and heart rate variability (HRV) from the wearable map the state of the autonomic nervous system: a resting heart rate rising over several days or a falling HRV signals strain, poor recovery or an oncoming infection. Body temperature and cycle tracking provide further layers of context — especially for women, whose biomarkers fluctuate systematically across the cycle.

Reference range is not optimal range

Perhaps the most important concept of this world: the lab reference range and the optimal range are not the same thing. The reference range is statistically defined — usually the interval in which 95% of a reference population falls. But that population is the average of the general public, including many people with suboptimal metabolism.

"Within the normal range" therefore only means "not conspicuous compared to the average" — not "associated with the best outcome". The optimal range, by contrast, is derived from outcome studies: where risk is lowest and function highest. For homocysteine, hs-CRP, HbA1c, ferritin or ApoB, this range often sits noticeably narrower than the lab range. This is exactly the gap biohacking plays in — and exactly the one the platform makes explicitly visible for every marker.

How we rate evidence

With markers, too, we distinguish strictly by evidence density — and above all between two questions that are often confused:

  1. Predictive value — How well does the marker predict a hard outcome? (e.g. VO2max and mortality, ApoB and cardiovascular events)
  2. Causal lever — Does shifting the marker also improve the outcome? (supported by Mendelian randomization and intervention studies, e.g. for ApoB)
  3. Surrogate caution — Some markers correlate with health without their targeted shifting guaranteeing the benefit. We flag that.

A marker that only predicts an outcome is not automatically a marker you should treat. We make this distinction explicit, instead of reflexively declaring every conspicuous value a treatment target.

Most common measurement mistakes

Diagnostics fails more often on methodology than on the marker:

  • Not measured fasting where fasting values are needed (glucose, insulin, lipids are otherwise distorted).
  • Tested right after hard training — this temporarily raises liver values, CK and CRP and triggers false alarms.
  • Single values over-interpreted instead of looking at the trend across several measurements.
  • Thyroid judged by TSH only, missing the conversion level.
  • LDL-C used where ApoB is discordant and therefore more informative.
  • Every possible marker chased instead of consistently following the few with real leverage.

What does not belong in this world

The boundary is clear: this world is the measurement layer, not the intervention layer. The substances and methods you use to move markers live in their own worlds:

  • Omega-3, vitamin D, B vitamins for value correction → World 01 (Vitamins & Minerals)
  • Fasting, cold, sauna as interventions → World 02 (Methods)
  • Thyroid and hormone therapy → World 03 (Hormones)
  • Longevity markers in the context of rapamycin & co. → World 06 (Longevity)

The geroscience biomarkers and epigenetic clocks (Horvath, GrimAge, DunedinPACE as aging rate) sit in a special zone: they are measurement instruments and therefore belong here, but their interpretation overlaps heavily with the longevity world. We treat them as a diagnostic tool — with the explicit note that they are surrogates, not the goal itself.

How Biohacking AI operationalizes this

On the platform this world is the common thread that grounds all the others. Four tools interlock:

  1. The biomarker tracker turns single values into a trend curve and links every change to whatever you adjusted in the meantime — so the effect of an intervention becomes visible instead of believed.
  2. The studies database shows, for each marker, which interventions demonstrably move it in human studies — and cleanly separates "shifts the value" from "improves the outcome".
  3. The forum collects anonymized panels and before-and-after trajectories — moderated, with no sales agenda, with optimal instead of just reference ranges.
  4. The coach reads your values in the context of your stack, flags what's worth a re-test, and says explicitly where the evidence for a recommendation is too thin.

The goal is not "measure more values." The goal is: know the few markers with real leverage, follow them over time and anchor every optimization to a measurement — instead of to a feeling. Diagnostics doesn't replace a medical workup; the platform helps you understand and track your values, not diagnose yourself.

How we operationalize it

The platform for this world

Biomarker tracker with trend curves

Log your blood work and wearable data — the platform plots a trend curve instead of a snapshot and links every change to whatever you adjusted in your stack or lifestyle in the meantime.

Every marker linked to studies

For each biomarker you see which interventions demonstrably move it in human studies — with evidence level, effect size and the distinction between 'shifts the value' and 'improves the outcome'.

Forum for lab values & protocols

Reference ranges are one thing, lived n=1 trajectories are another. In the diagnostics forum you share anonymized panels and before-and-after values with others tracking the same markers — moderated, with no sales agenda.

Coach interprets in context

The coach reads your values against optimal ranges and against your current stack — tells you which marker is worth a re-test, what's connected and where the evidence for a recommendation is still too thin. Diagnosis remains a physician's job.

Substances & topics

What is curated in Diagnostics & Biomarkers

21 topics under continuous study monitoring. Each links to its full evidence overview.

FAQ

Frequently asked questions

Which blood values should I measure at minimum as a biohacker?
A sensible baseline panel covers four axes: metabolic (fasting glucose, fasting insulin for HOMA-IR, HbA1c), cardiovascular (ApoB, full lipid profile, Lp(a) once), silent inflammation (hs-CRP, homocysteine) and organ function plus status (complete blood count, ferritin, liver values with GGT, kidney values, vitamin D, thyroid with TSH/fT3/fT4). That's around 15 values that together map 80% of the relevant levers. More important than breadth is repetition: a baseline value, then re-tests 3–6 months apart, so you see trends instead of random fluctuations.
ApoB or LDL cholesterol — which is more informative?
ApoB is the more precise marker. Every atherogenic lipoprotein — LDL, VLDL, IDL, Lp(a) — carries exactly one ApoB molecule. ApoB therefore counts the actual number of artery-damaging particles, while LDL-C only estimates the cholesterol content within them. In many people the two agree, but with insulin resistance, high triglycerides or small dense LDL particles they diverge (discordance) — and then LDL-C underestimates the risk. Large cohorts and Mendelian randomization studies support ApoB as causally closer to the action. Orientation: < 80 mg/dl in primary prevention, < 65 mg/dl at elevated risk.
HOMA-IR: How do I detect insulin resistance early?
HOMA-IR is calculated from fasting glucose and fasting insulin: (glucose in mg/dl × insulin in µU/ml) / 405. The decisive advantage: fasting insulin often rises for years before HbA1c or fasting glucose become conspicuous — the body initially keeps blood sugar normal by producing more insulin. Values above roughly 2.0 point to beginning insulin resistance, while a metabolically favorable range sits closer to 1.0–1.5. Anyone who measures only glucose or HbA1c systematically misses this early phase. That's exactly why fasting insulin belongs in every serious baseline panel.
What does a CGM actually tell you even without diabetes?
A continuous glucose monitor (CGM) shows how your blood sugar responds to specific meals, stress, sleep and exercise — information a single fasting value can't provide. Studies on personalized nutrition (Zeevi 2015, the PREDICT series) show that the glucose response to the same meal varies strongly between individuals. It's practically useful for seeing postprandial spikes and glycemic variability and adjusting meals deliberately. Caveat: the outcome evidence in the metabolically healthy is still young, and not every small spike is pathological. A CGM is a learning and tracking tool, not a substitute for diagnosis.
VO2max — why is it considered the single best predictor of life expectancy?
Cardiorespiratory fitness, measured as VO2max, correlates more closely with all-cause mortality in large cohorts than almost any other single marker. A widely cited analysis (Mandsager et al., JAMA 2018, over 120,000 people) found no upper limit of benefit — higher fitness was consistently associated with lower mortality, and the jump from the lowest into the next fitness quartile was dramatic. Unlike a blood value, VO2max is directly changeable through training, which makes it the most rewarding marker for interventions. Estimates from wearables are coarser than a spiroergometry test, but work well for tracking your own trend.
Thyroid values: Is TSH enough or do I need the full panel?
TSH alone can mislead. It's the control signal from the pituitary gland, not the thyroid hormone itself. A normal TSH doesn't rule out impaired conversion of T4 into the active T3, and especially when symptoms are present the full picture is worth it: TSH plus free T4, free T3 and — if autoimmunity is suspected — the antibodies TPO and Tg. The TSH range often discussed as optimal (roughly 1–2) is narrower than the broad lab range. Important: values fluctuate over the course of the day, so measure in the morning and consistently. Interpretation belongs in a physician's hands — this world helps you request the right values in the first place and track them over time.
HRV and resting heart rate from the wearable — how reliable are the values?
The absolute number is less reliable than your own trend. Heart rate variability (HRV) and resting heart rate depend heavily on the measurement method, time of day and sleep phase — a wearable HRV value can't be compared one-to-one with that of another device or a chest-strap measurement. Within the same device and under the same conditions (e.g. the nightly average), though, the trends are meaningful: an HRV falling over several days or a rising resting heart rate often signals strain, an oncoming infection or poor recovery. Use them as a trend signal for recovery and training, not as an absolute medical value.
Reference range vs. optimal range — why doesn't 'normal' mean 'optimal'?
The lab reference range is statistically defined: usually the interval in which 95% of a reference population falls. But that population is the average — including many people with suboptimal metabolism, excess weight or beginning disorders. 'Within the normal range' therefore only means 'not conspicuous compared to the average', not 'associated with the best outcome'. The optimal range, by contrast, is derived from outcome studies — where risk and function are most favorable. For markers like homocysteine, hs-CRP, HbA1c or ApoB, this optimal range often sits noticeably narrower than the lab range. Making this gap explicitly visible is the core task of this world.

Measure first — then optimize.

Enter your latest blood work and see in two minutes which markers are in the optimal range and which you should keep an eye on.

Waitlist
More worlds