Midlife labs

Routine general health check-ups do not make people live longer. A short list of cheap, specific blood markers does change what you do, provided you read them against guideline lines and your own trend rather than the "normal" flag. Most of the tighter "optimal" targets sold on top have never been tested against outcomes.

Once a year, have a fasting morning blood draw for apoB, fasting insulin and glucose, HbA1c, hs-CRP, ferritin with transferrin saturation, uric acid, a blood count and a metabolic panel. Measure Lp(a) once in your life. Test hormones, vitamin D and B12 only when you have a reason, and skip the marketed extras: NMR particle panels, IgG food panels, salivary cortisol, provoked heavy-metal tests and whole-body MRI. This is a screening plan for well adults, not a workup for symptoms or a substitute for your clinician's judgement.

Why more testing is not better

The general health check has been tested in trials, and it fails. In a Cochrane review, checks made little or no difference to deaths from any cause (11 trials, about 233,000 adults) or from cancer (8 trials, about 139,000 adults), and probably none to cardiovascular deaths [1]. A 2021 review agreed, while noting that checks do pick up more high blood pressure and depression and get more people to proven screens [2].

Every extra test in a healthy adult carries costs:

  • Overdiagnosis. Finding a real abnormality that would never have harmed you. The standard example is thyroid cancer: a modelling analysis of US registry data estimated that 72–94% of papillary thyroid cancers diagnosed between 1991 and 2019 were overdiagnosed, and for most of that period the usual response was surgery, often followed by lifelong thyroid hormone [3]. The 2025 American Thyroid Association guidelines made active surveillance a formal, if conditional, option for small (1 cm or less), low-risk papillary cancers [4].
  • Cascades. One ambiguous result leads to imaging, biopsies and their complications; a 95%-specific test for a 1%-prevalence finding gives more false positives than true ones.
  • Unneeded prescriptions. A TSH of 5.2, a ferritin in the upper quarter, or one high LDL in a stressful week can start a lifelong drug that does no good and has side effects.
  • Anxiety. After a false-positive mammogram, women report worry about breast cancer for up to 3 years, more after invasive follow-up, and are slightly less likely to return for their next screen [5]. That evidence is from breast screening, but any borderline flag can start the same cycle.

The answer is a small panel that answers real questions, read against guideline cut-offs and your own trend over time.

ApoB: count the particles, not the cholesterol

What drives plaque is the number of cholesterol-carrying particles that get into the artery wall, not the mass of cholesterol inside them [6]. Each of these particles carries exactly one apolipoprotein B (apoB) molecule, so an apoB test is a direct particle count. The standard LDL cholesterol test measures cholesterol mass instead. ApoB predicts cardiovascular risk at least as well as LDL cholesterol, and better when the two disagree.

LDL cholesterol and apoB disagree in around 1 in 5 people [7]; working out LDL with the newer Martin/Hopkins equation shrinks that gap markedly [8]. The risky mismatch, a "normal" LDL hiding many small particles, is most common with insulin resistance, high triglycerides or belly fat. The 2026 US guideline still treats LDL and non-HDL cholesterol as the main targets and gives apoB a supporting role [9]. For a baseline workup, apoB is probably worth more than the LDL on the same panel, though the case that acting on apoB rather than LDL improves outcomes rests on observational cohorts.

No guideline gives one apoB target for everyone; each scales it to risk (mg/dL; 100 mg/dL = 1.0 g/L):

GuidelineapoB goal, or level at which to step up treatment
European (2019)Below 100 at moderate risk, below 80 at high risk, below 65 at very high risk [10]
US National Lipid Association (2024)Consider stepping up above 90 at borderline-to-intermediate risk, above 70 at high risk, above 60 at very high risk [11]
US ACC/AHA (2026)Below 100, with LDL below 100, when the coronary calcium score is 1–99; consider below 55 in very-high-risk heart disease once LDL and non-HDL goals are met [12]

How to act on the result is covered in Lipid management.

Catching insulin resistance early

Glucose and HbA1c (a three-month average of blood sugar) are late signals. In 6,538 British civil servants, insulin sensitivity fell steeply over the five years before a diabetes diagnosis, while the sharp rise in fasting glucose began only about three years out [13]. As tissues stop responding to insulin, the pancreas pushes out more, so fasting insulin can probably flag insulin resistance while glucose still looks normal.

HOMA-IR combines the two fasting values: glucose in mmol/L × insulin in μIU/mL, divided by 22.5. Insulin tests are not standardised between labs, so track the trend in one lab [14]. No "optimal" value has been tested against outcomes, and the targets below 1.0 sold online sit under the median of about 1.3 in one large group of lean, healthy adults [15]. The outcome data point at the high end instead: in a US cohort without diabetes, a HOMA-IR above 2.8 went with about 64% higher mortality than 1.4 or below, though only in people with a body mass index below about 25 [16]. So treat a HOMA-IR of about 2 or more as worth acting on, and a lower value as reassurance rather than a number to push down.

For HbA1c, aim for 5.0–5.4%, not the floor. In 11,092 middle-aged US adults without diabetes followed for about 15 years, an HbA1c below 5.0% went with about 48% higher mortality than 5.0–5.4%, even after people with anaemia were excluded [17]. Pooled data from 28,681 adults aged 50 and over found no clear excess below 5.0%, and in the cohort where it was most pronounced (US NHANES), it faded once anaemia, liver markers and other factors were accounted for [18]. A very low HbA1c is a prompt to look for anaemia or liver disease.

A continuous glucose monitor does not belong on the panel. In people without diabetes, evidence that it improves any hard outcome is thin, brands disagree, and the usual result is anxiety over normal post-meal spikes (see Continuous glucose monitors).

Inflammation, iron and uric acid: both ends matter

hs-CRP (high-sensitivity C-reactive protein) is the best-validated blood marker of low-grade chronic inflammation. In a cohort of 27,939 healthy women followed for 30 years, the top fifth for hs-CRP had about 70% more major cardiovascular events than the bottom fifth, a stronger signal than LDL cholesterol gave (about 36%) and independent of it [19]. The 2026 US guideline counts it: if you have no cardiovascular disease and your 10-year risk is borderline (3% to under 5%), an hs-CRP of 2 mg/L or more on two occasions, with no other explanation, can justify a high-intensity statin [20].

As a rough guide, under 1 mg/L is the lowest-risk range, though no target below the guideline line has been tested against outcomes; 1–3 mg/L is worth tracing to metabolic and lifestyle causes; 3 or more on two draws weeks apart is high, and worth considering tighter lipid and metabolic control. A fresh infection, a hard workout in the last 24–48 hours or an injury all push it up, so repeat any high value several weeks later before acting. High hs-CRP often travels with insulin resistance [21], so improving the metabolic picture is a sensible first step; recheck before adding anything aimed at inflammation itself.

Ferritin (iron stores) has a lab range so wide, about 10 to over 300 µg/L, that it hides signal at both ends, though no narrower "optimal" window has been tested in a trial:

  • Low. Below 15 µg/L iron stores are empty and below 30 they are low [22]; with infection or inflammation, under 70 can still mean deficiency [23]. It is common in premenopausal women, distance runners and long-term plant-based eaters. A value under 30 usually means iron deficiency, something to raise with a clinician, who may look for a cause, especially in men and postmenopausal women.
  • High. Over 23 years in a Danish population, a ferritin of 200 µg/L or more went with about 10% higher mortality and 600 or more with about 50% higher, although the same paper's pooled analysis found no excess for the top quarter or third across studies [24]. Above 150 µg/L in women who menstruate, or 200 in men and other women, check transferrin saturation and hs-CRP before calling it overload, because inflammation also raises ferritin [25]. High ferritin with a normal hs-CRP and a transferrin saturation of 45% or more points to genuine iron loading, and US liver guidance advises a test for the haemochromatosis gene [26]. More in Iron and aging.

Uric acid tracks mortality in a U-shape. In 375,163 Korean adults, both the lowest levels (under about 210 µmol/L in men and 150 in women) and the highest (about 565 or more in men, 505 in women) went with more deaths [27]; a 2025 analysis saw the same U-shape in US adults but not in a two-year study of older Chinese adults [28]. Very low values in older adults often reflect malnutrition or muscle loss. High uric acid causes gout, but lowering it has not prevented heart attacks or strokes: among 5,721 people aged 60 or over who already had heart disease and no gout, 11.0% on allopurinol and 11.3% on usual care had a heart attack, stroke or cardiovascular death over about five years [29]. Rheumatology guidance conditionally advises against urate-lowering drugs without gout [30]. Flag either extreme and look at diet (fructose, alcohol, purines); there is no outcome-tested window.

Kidneys, liver and blood count: cheap, but read them properly

A blood count and a metabolic panel are the cheapest, best-validated organ checks in medicine. Three of their numbers are easy to misread.

eGFR (estimated kidney filtration) is calculated from creatinine, which also reflects muscle and diet: high protein, creatine supplements, heavy strength training, vegetarian or keto diets and changes in muscle mass all shift it without any change in the kidneys. Above about 60 mL/min/1.73 m², eGFR alone says little. In general-population cohorts, mortality was flat between 75 and 105 and rose only lower down, about 18% higher at 60 than at 95 [31]. A drift below 90 with age is common and is not kidney disease without other signs of damage, and a very high eGFR is not automatically good news: it can reflect low muscle mass. International kidney guidance adds a cystatin C test when creatinine may mislead, and a urine albumin-to-creatinine ratio (uACR) for anyone with diabetes, high blood pressure or heart disease, because albumin in the urine can flag damage while eGFR is still normal [32].

ALT and AST (liver enzymes) can hint at fatty liver disease, now called metabolic dysfunction-associated steatotic liver disease (MASLD) and closely tied to insulin resistance [33]. A normal result does not rule it out: the enzymes are often normal even with advanced scarring. US liver guidance treats ALT or AST above about 30 U/L as a possible sign of liver injury, below most labs' upper limit. With prediabetes, type 2 diabetes or two or more metabolic risk factors, it recommends a FIB-4 score (from age, AST, ALT and platelets) every 1–2 years; 1.3 or more (above 2.0 after 65) should lead to a liver scan, a specialised blood test or referral [34].

The blood count can hint at B12, folate or iron problems, but late and unreliably. A raised red-cell size caught only about 1 in 6 cases of low B12 in screened populations [35], and was absent in a third of people with pernicious anaemia [36]. If you suspect a deficiency, test B12 or ferritin directly.

Lp(a) once, APOE almost never

Lp(a) is an LDL-like particle with an extra protein attached. More than 90% of your level is inherited; it is fairly stable from about age five, so one test is enough, barely moved by diet or exercise, and not lowered by statins, which in some studies have raised it [37]. Roughly 1 in 5 people has 125 nmol/L (50 mg/dL) or more, which carries about 1.4 times the cardiovascular risk of the median. The 2026 US guideline and the National Lipid Association recommend measuring it once in every adult [38].

A high result changes little about which drugs you take. The first outcome trial of a drug built to lower Lp(a), pelacarsen, lowered Lp(a) but missed its main goal of reducing heart attacks and strokes in topline results released in September 2026 [39]. Filtering it from the blood (apheresis) is approved in the US only for a narrow group with an Lp(a) of 130 nmol/L or more, familial high cholesterol and artery disease [40]. What a high result does change is how hard you pull the other levers: apoB as low as tolerated, tight blood pressure, closer follow-up. Details in Lipid management.

APOE ε4 is the largest common genetic risk factor for late-onset Alzheimer's and modestly raises cardiovascular risk. Don't test for it as a healthy adult. It predicts poorly: about half of people clinically diagnosed with Alzheimer's carry no ε4 copy [41], though most with amyloid-confirmed disease do [42]. Even with two copies, found in about 2% of people, lifetime risk of Alzheimer's dementia is put at up to about 60% by age 85, so many never develop it [43]. No preventive treatment exists for people without symptoms, so the result delivers anxiety without an action; the things that help (aerobic exercise, sleep, blood-pressure control, hearing, social contact) apply either way. In some countries it can also affect long-term care and disability cover.

The one place APOE belongs is before an anti-amyloid antibody (lecanemab or donanemab) for mild cognitive impairment due to Alzheimer's or early Alzheimer's. People with two copies get far more brain swelling and small brain bleeds on these drugs: on lecanemab, symptomatic swelling occurred in 9% of people with two ε4 copies, 2% with one and 1% with none [44]. EU licences for both drugs, and the UK licence for lecanemab, exclude people with two copies [45]; US labels say testing should be done but allow treatment without it [46].

Tests that need a reason: B12, vitamin D and hormones

B12 earns a place in a few groups. Diabetes guidelines advise periodic checks for long-term metformin users, especially after four to five years of use, at 1,500 mg a day or more, with another risk factor for deficiency, or with anaemia, nerve symptoms or kidney disease [47]. In long-term follow-up of a large diabetes-prevention trial, each year of metformin raised the odds of low B12 by about 13% [48]. Strict vegans without supplements or fortified foods, and long-term users of stomach-acid drugs (PPIs or H2 blockers) are also at risk, but UK guidance tests them only with a suggestive symptom or sign. A total B12 above about 258 pmol/L makes deficiency unlikely and below 133 confirms it; in between, a methylmalonic acid or holotranscobalamin test helps decide [49]. See B vitamins.

Vitamin D is not a routine screen. The 2024 Endocrine Society guideline found no trial evidence that routine testing helps, including in people with obesity or darker skin, and no clear evidence for an optimal blood level, so it suggests against routine testing without a specific medical reason [50]. A bone-density or malabsorption work-up is such a reason. See Vitamin D.

Hormones are where consumer "longevity" panels overreach most:

  • Testosterone: test for symptoms (low libido, fatigue, lost morning erections, body-composition change, persistent low mood), with two morning fasting draws because levels swing through the day. Chasing high-normal levels in men without symptoms has no outcome support and carries real risks: thickened blood, worsening of hidden prostate disease and higher blood pressure. See Testosterone therapy.
  • Estradiol, FSH and progesterone in women rarely help outside a symptomatic perimenopause workup, fertility evaluation or adjusting menopausal hormone therapy, and the decision to start that therapy rests on symptoms and timing, not labs.
  • TSH (thyroid): a mildly raised TSH in a well midlife adult is usually transient. In 737 adults aged 65 or over, treating mildly underactive thyroid did not improve symptoms, tiredness or thinking; the trial was too small and short to test heart attacks or strokes and included no midlife adults [51]. Test with a reason: fatigue, weight or mood change, cold intolerance, high cholesterol, fertility or pregnancy. More in Thyroid management.
  • DHEA-S falls with age but doesn't change management in healthy adults, and DHEA supplements are controversial. A random cortisol is useless as a "stress" test.

Biological-age clocks: interesting, not a decision tool

DNA-methylation clocks try to measure ageing itself. Second-generation clocks such as GrimAge and PhenoAge predict death and disease in groups, and DunedinPACE was built to measure the pace of ageing: a score of 1.0 means one biological year per calendar year, 0.85 means 15% slower [52].

The problem is noise. DunedinPACE gives nearly the same number when one blood sample is run twice, but that is lab precision, not stability in you: re-sampled within hours or days, around meals, stress or a change of altitude, most clocks were only moderately stable and DunedinPACE was among the less stable [53]. No agreed standard yet says when a clock is accurate enough to guide one person's care [54].

A slower clock predicts, but doesn't prove, a longer or healthier life. Only a few randomised trials have moved one, all as secondary analyses with small effects: omega-3 by about 1% on DunedinPACE in adults over 70 [55], two years of calorie restriction by 2–3% (but not on PhenoAge or GrimAge) [56], and semaglutide by about 9% in a small trial in people with HIV-related fat gain [57].

If you test anyway, use a provider that names its published clock and runs it in an accredited lab, and keep the same provider, sample type, time of day and fasting state: in some clocks the same person's estimate differed by almost 30 years between a cheek swab or saliva and blood [58]. Testing more often only samples more noise, and don't change treatment on the result alone.

Tests to refuse

NMR particle counts when you have apoB. The two carry nearly the same information: in 27,673 healthy women followed for 11 years, the top fifth on either measure had about 2.5 times the cardiovascular risk of the bottom fifth [59], and a review of 25 studies preferred apoB for its availability, standardisation and cost [60].

IGF-1 as a longevity marker. It is essential for diagnosing pituitary disorders, not for tracking ageing. Both ends carry risk: across 12 studies, low IGF-1 went with about 27% higher mortality and high IGF-1 with about 18% higher [61], a pattern repeated in about 381,000 UK adults [62]. No trial shows that nudging it in either direction helps, and genetic studies suggest lifelong higher IGF-1 modestly raises the risk of a few cancers, about 8% higher colorectal cancer risk per standard deviation [63].

Salivary "adrenal stress" panels. "Adrenal fatigue" is not a medical diagnosis. A review of 58 studies found no evidence that it exists; salivary cortisol rhythms failed to separate fatigued from healthy people in 16 of 26 studies [64]. The Endocrine Society says the tests sold to diagnose it "are not based on scientific facts" [65]. Real adrenal disorders need specific, validated workups.

IgG food-sensitivity panels. Food-specific IgG mostly reflects normal exposure to, and tolerance of, foods you eat often [66], and Canadian, US and European allergy societies advise against these tests [67], [68]. The elimination diets they prompt narrow nutrition and strain social eating for no proven gain. The one signal is in irritable bowel syndrome, from a sham-controlled trial that used a purpose-built assay rather than a consumer panel [69]; that is a question for a gastroenterologist.

Provoked heavy-metal urine tests. Clinics give a chelating drug, then collect urine. Chelators raise urinary metals in healthy people too, no reference range exists for provoked urine, and some labs compare it with the normal range, which makes results look falsely high. The American College of Medical Toxicology says the test is not validated, has no demonstrated benefit and may be harmful [70]. The chelation that often follows can kill through dangerously low blood calcium with edetate disodium, can strip zinc, copper and iron, and can damage the kidneys with some agents. Ordinary metal screening is only worth it with a work or known exposure.

Whole-body MRI as a screen. In about 9,000 people without symptoms it found confirmed cancer in roughly 1.6%, with frequent incidental findings and no evidence that it improves outcomes [71].

The panel at a glance

TestAim for, or flagWhen
ApoBScaled to risk: below about 90–100 mg/dL at moderate risk, 70–80 at high, 55–65 at very highYearly
Fasting insulinWithin your lab's range, trended in one labYearly
HbA1c5.0–5.4%Yearly
HOMA-IRBelow about 2Yearly (calculated)
hs-CRPUnder 1.0 mg/L (lowest risk; not an outcome-tested target)Yearly
FerritinAbove 30; below 150 µg/L (women who menstruate) or 200 (men and other women)Yearly, with transferrin saturation
Uric acidNeither extremeYearly
Blood count, metabolic panelALT/AST under about 30 U/L; eGFR stableYearly; add uACR with diabetes, high blood pressure or heart disease, and FIB-4 with metabolic risk factors
B12Above 258 pmol/LPeriodically on long-term metformin, especially after four to five years; otherwise only with symptoms plus a risk factor
Lp(a)125 nmol/L or more means higher riskOnce

Two scans are worth knowing about. A coronary calcium score (a quick, non-contrast CT) is for when a statin decision is genuinely borderline, in men from 40 and women from 45. If the score is zero and you would rather avoid a statin, deferring and rescanning in 3–7 years is reasonable, unless you have very high LDL (4.9 mmol/L or more), diabetes after 40, smoke, or have a strong family history of early heart disease. Any calcium means plaque, and the guideline recommends lipid-lowering treatment, particularly at a score of 100 or more [72]. The cancer and other screens that do earn their place (mammography, colorectal, HPV, lung CT for heavy current or recent smokers aged 50–80, one-time hepatitis C and, in men aged 65–75 who ever smoked, aortic aneurysm) are covered in Cancer screening cadence.

Further reading

  • Krogsbøll LT et al. General health checks in adults for reducing morbidity and mortality from disease. Cochrane Database Syst Rev 2019.[73]
  • Liss DT et al. General Health Checks in Adult Primary Care: A Review. JAMA 2021.[74]
  • De Oliveira-Gomes D et al. Apolipoprotein B: Bridging the Gap Between Evidence and Clinical Practice. Circulation 2024.[75]
  • Blumenthal RS, et al. 2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Dyslipidemia: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. J Am Coll Cardiol 2026.[76]
  • Ridker PM et al. Inflammation, Cholesterol, Lipoprotein(a), and 30-Year Cardiovascular Outcomes in Women. NEJM 2024.[77]
  • Mora S et al. Lipoprotein Particle Profiles by NMR Compared With Standard Lipids and Apolipoproteins in Predicting Incident CVD in Women. Circulation 2009.[78]
  • Belsky DW et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife 2022.[79]
  • Stott DJ et al. Thyroid Hormone Therapy for Older Adults with Subclinical Hypothyroidism (TRUST). NEJM 2017.[80]
  • Aroda VR et al. Long-term Metformin Use and Vitamin B12 Deficiency in the DPP/DPPOS. J Clin Endocrinol Metab 2016.[81]
  • Cadegiani FA, Kater CE. Adrenal fatigue does not exist: a systematic review. BMC Endocr Disord 2016.[82]
  • American College of Medical Toxicology. ACMT Recommends Against Use of Post-Chelator Challenge Urinary Metal Testing. J Med Toxicol 2017.[83]
  • Francis DO et al. Overdiagnosis of Papillary Thyroid Cancer. JAMA Netw Open 2026.[84]

— § —