Know Your Numbers: The Blood & Age Tests Worth Doing

Large Mendelian randomization studies, prospective cohorts and meta-analyses — in one pooling of 233,455 people apoB was the strongest lipid risk marker, about 12% better than LDL-cholesterol — indicate that apoB is the predominant lipid trait driving…

Dose used in studies: Once to find out where you stand, then repeat about 8-12 weeks after any deliberate change - a statin, a real diet change, meaningful weight loss. Fasting is not required for ApoB. Stick to one lab, because between-lab differences are larger than the changes you are looking for.

What the evidence actually says: ApoB counts atherogenic particles rather than the cholesterol inside them, and where the two disagree the particle count is the better guide to risk - that part is well supported and is why several lipid guidelines now list it as a preferred target. Whether a mailed finger-stick delivers that number reliably is a separate question, and the honest answer is partly. A dried blood spot ApoB assay has been shown to correlate closely with a plasma reference method, but that work was done on laboratory-prepared spots from venous blood with a haematocrit correction applied, not on cards collected at a kitchen table and posted in July - collection technique, spot volume and haematocrit are all real sources of error nobody is controlling at your end. Treat a home ApoB as a screening number that tells you roughly which band you are in, and if it lands high, the next step is a venous panel through a clinician rather than a second kit.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests · ApoB

Large meta-analyses, prospective cohorts, and Mendelian randomization studies establish elevated Lp(a) as a causal, independent risk factor for coronary heart disease, myocardial infarction, ischemic stroke, peripheral artery disease, and calcific aortic…

Dose used in studies: Once in adult life. Lp(a) is set by the LPA gene you were born with and moves very little afterwards, so one good result is the result. Repeat only if the units or the lab change and you cannot line the numbers up.

What the evidence actually says: The analyte is about as solid as anything on this site: LPA gene variants that raise Lp(a) track coronary disease in a dose-dependent way, the European Atherosclerosis Society treats a raised level as an independent causal risk factor, and in the 30-year follow-up of the Women's Health Study a single baseline Lp(a) still carried predictive weight three decades on. But this is not a standalone Lp(a) test - it is a multi-analyte heart panel that also returns ApoB and hs-CRP, so it overlaps two other tests on this list, and Lp(a) reporting still has a real standardisation problem: mg/dL and nmol/L are not interchangeable, and apo(a) isoform size biases mass-based assays. The part the marketing skips is that there is no licensed Lp(a)-lowering drug today - olpasiran and its class lower the number impressively, but the cardiovascular outcome trials have not reported. So a high result changes only how hard you and a clinician manage everything else you can actually move, which is worth knowing once, because the number itself will not change.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests · Lp(a)

Higher fasting insulin and HOMA-IR reflect greater insulin resistance, a central feature of metabolic syndrome that precedes and predicts type 2 diabetes and, in large cohorts and meta-analyses, is associated with increased risk of cardiovascular disease,…

Dose used in studies: Fasting 10-12 hours, and only ever compared against your own previous result from the same laboratory. Twice a year at most. Treat the trend as the signal and ignore the absolute HOMA-IR figure printed on the report.

What the evidence actually says: HOMA-IR is a reasonable idea with a real problem underneath it. The formula itself has been in use since 1985 and correlates acceptably with clamp measures at a population level, but it is built on a fasting insulin value, and insulin has no reference measurement procedure - when 12 commercial assays from 9 manufacturers were compared against isotope-dilution mass spectrometry, differences ran from roughly -298 to +303 pmol/L and only one assay agreed fully with the reference, despite all of them claiming traceability to the same WHO standard. An earlier eight-assay comparison found median insulin concentrations differing by a factor of 1.8, which makes any published HOMA-IR cut-off essentially assay-specific: 2.4 from one lab and 2.4 from another are not the same finding, and a threshold you read online may not apply to your report at all. Note also that this product is a 7-in-1 metabolic panel rather than a dedicated insulin test - and if the question you are actually asking is whether your glucose metabolism is drifting, an HbA1c through a clinician is cheaper, standardised and interpretable.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests

Across large prospective cohorts and meta-analyses, higher hs-CRP is consistently associated with increased risk of coronary heart disease, stroke, cardiovascular and non-vascular mortality, and with aging-related states such as frailty and dementia; a…

Dose used in studies: Only when you are well. Check it twice, a couple of weeks apart, and take the lower value - a single high reading is far more likely to be a cold than a signal. Do not test within 2-3 weeks of any infection, injury, vaccination or hard unaccustomed exercise.

What the evidence actually says: hs-CRP has earned its place: it separated risk independently of LDL cholesterol in the classic NEJM prediction work, and in 30-year follow-up of the Women's Health Study it was still carrying predictive information decades after a single measurement. The measurement travels better than most on this list - CRP has been validated from finger-stick dried blood spots against automated analysers, with stability, haematocrit and precision formally checked, and remote self-collection in older adults has been shown to be reliable. The weakness is not the assay, it is the biology: CRP is an acute-phase protein that swings enormously with any infection or inflammation, so one number in isolation is close to meaningless and this product bundles it with vitamin D rather than selling it alone. Also worth saying plainly - a genuinely raised hs-CRP in a well person has no specific consumer fix attached to it; it is information you take to a clinician, not an instruction.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests · hs-CRP

Across large cohorts and meta-analyses, cystatin C-based eGFR is at least as accurate as creatinine-based eGFR and is more strongly associated with death, cardiovascular events, and end-stage kidney disease—in one ~90,750-person consortium meta-analysis it…

Dose used in studies: Not a routine self-monitoring test. Reasonable once if you have a specific reason to think a creatinine-based eGFR is misleading you - unusually high or low muscle mass, or a creatinine result you are trying to make sense of. Otherwise this belongs to whoever is already following your kidney function.

What the evidence actually says: Cystatin C genuinely outperforms creatinine alone. It is less dependent on muscle mass, the combined creatinine-cystatin C equation estimates GFR more accurately than either marker by itself, and where the two estimates disagree the cystatin C-based number is the one that tracks mortality and kidney outcomes - a large meta-analysis of that discordance makes the point clearly, and KDIGO now recommends cystatin C as a confirmatory test. That is exactly why it is a poor fit for a consumer kit: cystatin C earns its keep as a second, confirmatory measurement read alongside a creatinine, a urine albumin-to-creatinine ratio and your history, and an eGFR from a mailed sample with none of that context is a number without a denominator. It is also affected by things other than filtration - thyroid status, corticosteroids, body composition - so an abnormal result here leads directly to a clinician and a proper panel, which is a fair argument for starting there.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests

Large prospective cohorts and meta-analyses consistently link higher serum GGT, even within the reference range, to increased all-cause and cardiovascular mortality, incident hypertension, type 2 diabetes, metabolic syndrome and fatty liver, and some…

Dose used in studies: Not a tracking test. Reasonable once if you have a concrete reason - regular alcohol intake you want an objective read on, a medication known to affect the liver, or a metabolic picture you are already working on. Interpreting it in isolation is the mistake.

What the evidence actually says: GGT is a real risk marker, not a curiosity - pooled across 19 cohorts and more than 9 million people, those in the top third had roughly 1.6 times the all-cause mortality of the bottom third. It is also about as non-specific as a blood test gets - it rises with alcohol, with a long list of ordinary medications, with metabolic and fatty liver disease, and with obesity, and it does not tell you which of those it is - head-to-head work shows that direct markers such as phosphatidylethanol outperform indirect markers like GGT for identifying alcohol-related liver injury. That is the whole difficulty with buying it alone: a raised GGT gives you a question rather than an answer, and the next step is a full liver panel and a conversation with a clinician who knows what you drink and what you take. If you already know the answer to those questions, the test is unlikely to change what you do.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests

Iron status has a U-shaped relationship with health: iron deficiency (low ferritin) is a leading cause of anemia and, even without anemia, is associated with fatigue that iron repletion can improve, while in older adults it has been linked to more severe…

Dose used in studies: Reasonable once if you have a specific reason - heavy periods, a plant-based diet, blood donation, unexplained fatigue - and again 3 months after starting iron, no sooner. Not a number to track routinely if you have no reason to suspect a problem.

What the evidence actually says: Ferritin is the standard first-line marker of iron stores and a genuinely low value is informative. The trap is that ferritin is also an acute-phase reactant, so it rises with infection, inflammation, obesity and liver disease - which means a normal or even high ferritin does not rule out iron deficiency in anyone with an inflammatory burden, and this is common enough to matter. The thresholds are shakier than they look, too: the WHO cut-offs of 12 and 15 micrograms per litre were set by expert opinion, and when NHANES data were used to derive thresholds anchored to actual iron-deficient erythropoiesis they came out substantially higher, so a result reported as normal against the old cut-off may not be. A ferritin on its own, without a CRP to interpret it against and without a full blood count, is a number you will end up taking to a clinician anyway - and iron is not a supplement to start on a home result, because taking it when you do not need it does harm.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests

Across large prospective cohorts and meta-analyses, lower grip strength is consistently associated with higher all-cause and cardiovascular mortality, incident cardiovascular disease, disability, and, in some studies, cognitive decline and depression, often…

Dose used in studies: Three squeezes per hand, seated, elbow at 90 degrees, best value recorded - and the same chair, same arm position, same time of day every time. Monthly is plenty. What matters is your own trend over years, not where you sit against a population chart.

What the evidence actually says: Grip strength is one of the most durable prognostic markers in epidemiology: in PURE across 17 countries and in half a million UK Biobank participants, lower grip tracked all-cause and cardiovascular mortality, and it did so after adjustment and across very different populations. This specific device also clears the bar that most consumer instruments fail - the Camry digital dynamometer has been tested head-to-head against the Jamar, the reference instrument, with excellent reliability and correlations of 0.97-0.99, so a cheap unit is genuinely measuring the thing. Two honest caveats - grip strength is a marker of overall muscle and health rather than a target in itself, and squeezing a dynamometer more often will not move it, since the intervention that moves it is resistance training, which is graded separately and is where the actual benefit lives. The second is that this is the site's second grip dynamometer: it measures exactly what the other one measures, the other is graded strong, and nothing in the evidence says a 200 kg-capacity unit tells a middle-aged adult anything a 90 kg one does not - buy one, and let price decide which.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests · Grip strength · Blood pressure

In prospective cohort studies and pooled analyses spanning dozens of cohorts, higher blood levels of EPA + DHA have been consistently associated with lower all-cause and cardiovascular mortality - across ten cohorts an Omega-3 Index near 8% carried roughly…

Dose used in studies: Once to find your baseline, then again about 4 months after changing oily fish intake or starting a supplement - red cell fatty acids turn over slowly and anything sooner is reading noise. Most people sit around 4-5%; the level associated with lowest risk in the cohort work is roughly 8% or above.

What the evidence actually says: This is the one test on the list where the analyte and the method were developed together and both hold up. Pooled across 17 prospective cohorts, higher blood omega-3 levels tracked lower total and cause-specific mortality, the Framingham data point the same way, and the index adds predictive information on top of the pooled cohort equation - and unlike a serum fatty acid, red cell membrane composition is a stable several-month average rather than a snapshot of last night's dinner. Two things to hold in mind anyway - the evidence is observational, and people with a high Omega-3 Index differ from people with a low one in many ways, and the supplement trials designed to test whether raising it changes outcomes have come out mixed rather than clean, so a low number is a reason to eat more oily fish and not proof that a capsule will buy you the associated risk. And the index is a proprietary measure run by the laboratory whose founder developed it and co-authors much of the supporting literature - that is disclosed rather than hidden, but you should know it is not an independently standardised assay with an external reference method behind it.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests · Omega-3

Across large studies, dysregulated cortisol patterns track with worse health: a meta-analysis found flatter daily cortisol slopes are associated with poorer mental and physical health, the Whitehall II cohort linked flatter salivary slopes to higher…

What the evidence actually says: Salivary cortisol is a legitimate assay - late-night salivary cortisol is one of the recommended first-line tests for suspected Cushing syndrome, and its diagnostic accuracy has been characterised across 139 studies - but that is a specific endocrine question, asked by a clinician about a patient with reasons to suspect it, and it is not what this product is selling. Consumer 'stress testing' is a different claim, and there is no established evidence that a four-point daily cortisol curve in a well person measures psychological stress, predicts anything about their health, or identifies a treatable state such as 'adrenal fatigue', which is not a recognised diagnosis. The measurement is also unusually fragile in exactly the setting it is sold for: the day-curve depends on sample timing nobody is verifying, and the cortisol literature itself found that fewer than 6% of published studies electronically confirmed when samples were actually taken, with timing error producing frankly erroneous curves. Graded preliminary rather than as a measurement tool because the analyte is real but the consumer claim built on it has essentially no supporting outcome evidence - it is the weakest thing in this category and we would rather say so than file it alongside tests that work.

Full evidence page → · Check price on Amazon ↗ · More Diagnostics & Epigenetic Tests