Measure: Les analyses de sang qui changent une décision

What It Is

A blood panel is a set of measurements drawn from a single sample. Its value depends almost entirely on the measurements chosen and the reason each was ordered. A small, well-chosen set returns numbers you act on: whether to lower a cholesterol particle, whether blood sugar is drifting, whether the kidneys strain before a symptom shows. A large, undirected set, the kind marketed as a full-body or executive panel, can return 50 or more numbers in people with no symptoms. Most were never going to change anything, and some set off a chain of follow-up even when nothing is wrong. The task is separating the markers that guide a decision from those that only generate follow-up.

How to Read a Result

1A Reference Range Is Not the Same as Optimal

A laboratory reference range is simply where most results from a broadly healthy sample fall, usually the middle 95 percent. It is not the level of lowest risk, and not a hard line between well and unwell. LDL cholesterol sits well inside the reference range at levels trials still show benefit from lowering. A value can read as normal on the report and still call for action.

2One Value Is a Snapshot, Retest Before Acting

Blood markers vary from day to day and with the laboratory, the time, illness, hydration and the last meal. hs-CRP alone can shift 30 percent or more between two mornings, enough to cross a threshold. A mildly off value is a reason to retest, and the trend across repeats carries more meaning than any single draw.

3A Number Is a Marker, Not the Whole Person

A result gains its meaning from context: your age, symptoms, family history, the other markers and the direction it is moving. Read without that context, a single number produces both false alarms and false reassurance.

How It Works

Most useful markers are indirect readings of a process you cannot see. LDL cholesterol and ApoB track the same population of fatty particles that lodge in artery walls. LDL cholesterol measures how much cholesterol those particles carry; ApoB counts how many particles there are, since each carries a single ApoB molecule. When the particles are small and cholesterol-poor, common in insulin resistance and high triglycerides, the two numbers disagree. The particle count is then the better guide to risk.

Blood-sugar markers track an unseen process too. HbA1c measures the fraction of hemoglobin with glucose bound to it, accumulated over the red cell's ~120-day lifespan. A single draw therefore averages the past two to three months, with no fasting needed.

The Markers That Change a Decision

Heart and Lipids

Heart and lipid markers are the highest-value group: the evidence reaches cause, and the actions it identifies lower risk.

• LDL cholesterol is the anchor lipid marker. Across 26 randomized trials in about 170,000 people, each 39 mg/dL (1.0 mmol/L) fall in LDL cut heart attacks and strokes by roughly 22 percent. Genetic, epidemiological and trial evidence converge: LDL helps cause arterial disease.
• ApoB refines the reading by counting particles, and earns its place when it disagrees with LDL cholesterol.
• Lp(a) is the underused one: an inherited LDL-like particle, roughly 90 percent set by genes, stable across adult life, and measured a single time to classify risk for good. It adds cardiovascular risk on top of your other lipids, independent of them, yet the standard panel omits it and it is still ordered too rarely.
• Blood pressure is not drawn from blood, but it is among the most decision-changing numbers you can track. Reading it correctly at home matters more than almost any lifestyle change made in response, and the current US threshold for high blood pressure is 130/80 mm Hg. See self-measures for the technique and high blood pressure for what moves it.

Blood Sugar and Metabolism

HbA1c is the core blood-sugar marker: 5.7 to 6.4 percent flags prediabetes, 6.5 percent or above diabetes, early enough to act. Fasting glucose adds a same-morning snapshot. HOMA-IR, built from fasting insulin and fasting glucose, estimates how hard the body works to keep glucose in check, and can move before HbA1c or fasting glucose do. Labs report insulin on different scales, so a fixed cut-off does not transfer; read HOMA-IR as your own trend across repeats.

A drifting HbA1c or HOMA-IR is mostly a diet signal: ultra-processed foods supply roughly 57 percent of the calories US adults eat, and that intake tracks with insulin resistance. It is also the lever you hold. In the Diabetes Prevention Program, diet and 150 minutes of weekly activity cut progression to diabetes by 58 percent, against 31 percent for metformin.

Inflammation

hs-CRP is the practical inflammation marker. It rises with any infection, injury or flare, so a value drawn when someone is unwell shows that illness and is repeated once they are well. Read at baseline, above 3 mg/L marks higher cardiovascular risk and below 1 mg/L lower, though the independent effect is modest. It can tip a borderline decision but will not settle one by itself.

Organs, Thyroid and Iron

Four more tests earn their place when there is a reason to look:

• eGFR and urine albumin are two kidney tests, one from blood and one from urine, and each predicts long-term risk on its own. An eGFR below 60 for three months, or albumin leaking into urine, marks kidney disease even when the other reads normal, so both belong on the panel.
• Liver enzymes, ALT and AST, screen for liver strain; a single mildly raised value, under roughly 2 times the lab's upper limit, means little on its own and is read in context.
• TSH is the right first test when symptoms point to a thyroid problem, such as fatigue, weight change or cold intolerance. A result outside roughly 0.4 to 4.0 mIU/L warrants a closer look, though checking it routinely in people who feel well has a weaker case.
• Ferritin, read with a full blood count, is the best single test for iron deficiency, with a value below 30 ng/mL the usual flag. Low ferritin in a menstruating woman with fatigue is a different signal from the same value in a well older adult. Ferritin also rises with inflammation, which can mask a true deficiency.

The Full-Body Panel and What the Trials Show

The full-body panel orders dozens of markers at once in someone with no symptoms and no specific reason. Panels like this have been put to large randomized trials. Pooled across 17 randomized trials and 251,891 people, the general health check left total mortality and cardiovascular and cancer death unchanged while adding diagnoses and prescriptions.

The tests themselves are accurate; the trouble is what an abnormal result sets off. In a low-risk person, most abnormal-looking values are meaningless variation or findings that would never have caused harm. Each can still lead a healthy person into further tests, scans and procedures that carry their own risk. This is the ground of the Choosing Wisely campaign and of preventive-services reviews of routine screening.

In 17 trials covering 251,891 people, the full panel changed no death rate while adding diagnoses to healthy lives. A marker earns its draw only when a plausible result would change what you do.

The Research & Studies

Everything here is based on the research we have collected and checked, sorted into groups and ordered with the strongest evidence first. Click any claim to open the studies behind it.

Cholesterol And Lipids

Each 39 mg/dL (1.0 mmol/L) drop in LDL cholesterol cuts heart attacks and strokes about 22%Strong

In plain terms

The more you lower LDL cholesterol, the fewer heart attacks and strokes, and it does not matter much how you lower it. That is why LDL is the lipid number that guides a decision.

In detail

The Cholesterol Treatment Trialists pooled individual data from 26 statin trials totaling about 170,000 participants. Major vascular events fell by a proportional 22% (relative risk 0.78, 95% CI 0.76 to 0.80) for every 39 mg/dL (1.0 mmol/L) reduction in LDL cholesterol, and the relationship held across baseline risk and across the size of the reduction. Because the benefit scales with the absolute LDL drop, LDL is a marker that reads back a modifiable risk, not a fixed characteristic.

Who this may not transfer to:Trials pooled men and women across many countries; the proportional benefit per unit of LDL lowering applies broadly to adults deciding how far to lower LDL.

The study · 1

Cholesterol Treatment Trialists Collaboration, efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials · Lancet 2010;376(9753):1670-1681

LDL cholesterol causes arterial disease, and the harm builds up over a lifetimeStrong

In plain terms

LDL is not just linked to heart disease, it helps drive it, and the damage builds up over a lifetime. A high number earlier in life matters more than the same number caught late.

In detail

The consensus panel drew on Mendelian randomization (people who inherit LDL-lowering gene variants have correspondingly lower lifetime cardiovascular risk), long-running cohorts, and randomized trials, and found all three lines converge on causation. The genetic data show the effect is a function of cumulative LDL exposure, area under the curve of concentration times time, which is why lowering LDL earlier yields more benefit per unit than lowering it late.

Who this may not transfer to:Genetic and trial evidence spans large mixed adult populations; the cumulative-exposure principle applies to anyone carrying a raised LDL earlier in life.

The study · 1

Ference et al., low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel · Eur Heart J 2017;38(32):2459-2472

Measurement And Diagnosis

Higher HbA1c predicted later diabetes and heart disease across 11,092 adults, from one drawStrong

In plain terms

HbA1c reflects your average blood sugar over the past few months in one draw, no fasting needed. A rising value flags diabetes risk early, while there is still time to act.

In detail

Dans l'étude Atherosclerosis Risk in Communities, 11,092 adultes noirs et blancs sans diabète ont eu une HbA1c mesurée initialement et ont été suivis pendant environ 15 ans. Une HbA1c plus élevée était associée de façon graduée au diabète diagnostiqué ainsi qu'à la maladie coronarienne, à l'AVC et au décès, et pour les critères cardiovasculaires, l'HbA1c était au moins aussi informative que la glycémie à jeun. L'HbA1c reflète la glycation au cours des deux à trois mois précédents, si bien que, contrairement à une glycémie à jeun, elle ne dépend pas d'une seule matinée ni du fait que la personne ait jeûné.

Who this may not transfer to:The cohort was US Black and White adults without diabetes; HbA1c transfers as a fasting-free read of recent blood sugar, with the hemoglobin caveats noted.

The study · 1

Selvin et al., glycated hemoglobin, diabetes, and cardiovascular risk in nondiabetic adults · N Engl J Med 2010;362(9):800-811

Un eGFR inférieur à 60 et une albumine urinaire en hausse prédisent chacun de façon indépendante le risque de décès à long termeStrong

In plain terms

Deux tests discrets, l'un sanguin (eGFR) et l'autre urinaire (albumine), signalent des troubles rénaux et un risque de décès plus élevé bien avant l'apparition de symptômes. Ensemble, ils en disent plus que chacun pris isolément.

In detail

Le Chronic Kidney Disease Prognosis Consortium a regroupé 14 cohortes en population générale, soit environ 105,872 participants. La mortalité toutes causes augmentait une fois le DFG estimé tombé sous 60 mL/min/1.73 m2, et le rapport albumine/créatinine urinaire était un prédicteur continu et indépendant de la mortalité toutes causes et cardiovasculaire sur toute la plage. Comme l'eGFR et l'albuminurie capturent des aspects différents des lésions rénales, la stadification actuelle utilise les deux axes ; un eGFR normal avec une albumine en hausse marque tout de même un risque supplémentaire.

Who this may not transfer to:General-population cohorts of mixed-sex adults; eGFR and albumin transfer as long-term risk markers once confirmed on repeat testing.

The study · 1

Chronic Kidney Disease Prognosis Consortium (Matsushita et al.), association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis · Lancet 2010;375(9731):2073-2081

Les bilans de santé larges n'ont pas modifié la mortalité chez 251,891 personnes (risque relatif 1.00)Strong · no effect

In plain terms

Proposer à des personnes en bonne santé un bilan large comportant de nombreux tests ne les a pas aidées à vivre plus longtemps ni à éviter le décès par cancer ou maladie cardiaque. Trouver davantage de choses ne s'est pas traduit par de meilleurs résultats.

In detail

La revue Cochrane a regroupé 17 essais randomisés, 251,891 participants, comparant des bilans de santé généraux (multiples tests de dépistage et mesures de facteurs de risque) aux soins habituels chez des adultes non sélectionnés. La mortalité toutes causes confondues n'était pas modifiée (risque relatif 1.00, IC à 95 % 0.97 à 1.03), avec un niveau de preuve élevé, de même que la mortalité par cancer et la mortalité cardiovasculaire. Les bilans ont bien augmenté le nombre de diagnostics et de prescriptions ; on a donc trouvé et traité davantage, sans bénéfice mesurable en termes de décès ou de maladie. Ce constat porte sur des panels larges et non ciblés chez des personnes sans symptômes, et non sur un test spécifique prescrit pour une raison précise.

Who this may not transfer to:Trials enrolled unselected mixed-sex adults; the no-effect result applies to broad routine screening of the well, not to a targeted test ordered for a reason.

The study · 1

Krogsboll et al., general health checks in adults for reducing morbidity and mortality from disease · Cochrane Database Syst Rev 2019;(1):CD009009

ApoB, the particle count, reads cardiovascular risk better than LDL cholesterolModerate

In plain terms

ApoB counts how many artery-clogging particles you carry, not just how much cholesterol is in them. When the two disagree, the particle count is the better guide to risk, and it is still ordered too rarely.

In detail

Sniderman and colleagues pooled studies comparing lipid markers and found ApoB outperformed both LDL cholesterol and non-HDL cholesterol as a predictor of cardiovascular events. Marston and colleagues, analyzing UK Biobank participants, found the risk of myocardial infarction tracked the concentration of ApoB-containing particles, not their cholesterol or triglyceride content; once ApoB was accounted for, the associations of LDL cholesterol and triglycerides largely fell away. ApoB is most useful when it diverges from LDL cholesterol, as it often does in insulin resistance, high triglycerides or diabetes.

Who this may not transfer to:Marker-comparison studies and the UK Biobank cohort are mixed-sex adults; the particle-count advantage matters most in insulin resistance, high triglycerides or diabetes.

The studies · 2

Sniderman et al., a meta-analysis of low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, and apolipoprotein B as markers of cardiovascular risk · Circ Cardiovasc Qual Outcomes 2011;4(3):337-345

Marston et al., association of apolipoprotein B-containing lipoproteins and risk of myocardial infarction in individuals with and without atherosclerosis · JAMA Cardiol 2022;7(3):250-256

Each 3.5-fold higher Lp(a) carries about 1.13 times the coronary risk, and it is inherited so you measure it onceModerate

In plain terms

Lp(a) is a cholesterol particle you are mostly born with. A high level adds heart risk on top of your other numbers, and because it barely changes through life you only need it measured once.

In detail

The Emerging Risk Factors Collaboration pooled individual data from about 126,000 participants in long-term prospective studies and found lipoprotein(a) associated continuously with coronary heart disease and stroke, a relative risk near 1.13 (95% CI 1.09 to 1.18) per 3.5-fold higher concentration after adjustment for conventional risk factors. Genetic work found LPA gene variants that raise lipoprotein(a) are associated with coronary disease, which supports a causal reading. Levels are roughly 90% heritable and stable across adulthood, so a single measurement classifies a person for life.

Who this may not transfer to:Pooled cohorts and genetic data are mixed-sex adults; because Lp(a) is inherited, one measurement carries across a person's adult life.

The studies · 2

Emerging Risk Factors Collaboration, lipoprotein(a) concentration and the risk of coronary heart disease, stroke, and nonvascular mortality · JAMA 2009;302(4):412-423

Clarke et al., genetic variants associated with Lp(a) lipoprotein level and coronary disease · N Engl J Med 2009;361(26):2518-2528

Chaque écart-type supplémentaire de hs-CRP ajoute environ 1.23 fois le risque coronarien après ajustement completModerate

In plain terms

La hs-CRP est une mesure générale de l'inflammation de bas grade, et des niveaux plus élevés s'accompagnent d'un risque cardiaque un peu plus élevé. Elle affine le tableau à la marge et n'est pas utilisée seule.

In detail

L'Emerging Risk Factors Collaboration a regroupé les données individuelles de 160,309 personnes sans antécédent de maladie vasculaire. Le rapport de risque ajusté pour l'âge et le sexe pour la maladie coronarienne était de 1.63 (IC à 95 % : 1.51 à 1.76) par 1 écart-type supplémentaire du log de la protéine C-réactive, s'atténuant à 1.37 (1.27 à 1.48) après ajustement pour les facteurs de risque conventionnels et à 1.23 (1.07 à 1.42) après ajustement supplémentaire pour le fibrinogène. La contribution indépendante est modeste, ce qui explique pourquoi la hs-CRP est utilisée pour reclasser les personnes proches d'un seuil de décision, et non comme dépistage principal.

Who this may not transfer to:Pooled cohorts of mixed-sex adults without prior vascular disease; the modest independent signal transfers as a refiner near a decision threshold.

The study · 1

Emerging Risk Factors Collaboration, C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis · Lancet 2010;375(9709):132-140

Les preuves sont insuffisantes pour recommander le dépistage thyroïdien chez les adultes sans symptômes, tandis que la TSH est le premier test dès l'apparition de symptômesModerate · mixed

In plain terms

Il n'y a pas d'argument solide pour vérifier la thyroïde chez des personnes qui se sentent bien. Lorsque des symptômes évoquent la thyroïde, la TSH est le bon premier test.

In detail

La Task Force a examiné le dépistage des troubles thyroïdiens chez les adultes asymptomatiques et non enceintes et a émis une déclaration I, concluant que les preuves étaient insuffisantes pour évaluer l'équilibre entre bénéfices et risques, en grande partie parce que les anomalies infracliniques sont fréquentes, se résolvent souvent d'elles-mêmes, et que leur traitement chez des personnes sans symptômes n'a pas montré de bénéfice clair. Il s'agit d'une déclaration sur le dépistage des personnes en bonne santé, non sur le diagnostic : chez une personne présentant de la fatigue, un changement de poids, une intolérance au froid ou d'autres signes évocateurs, la TSH avec T4 libre réflexe est le test de première intention standard.

Who this may not transfer to:The verdict covers asymptomatic non-pregnant adults; pregnancy and symptomatic testing are separate situations it does not cover.

The study · 1

LeFevre, US Preventive Services Task Force, screening for thyroid dysfunction: US Preventive Services Task Force recommendation statement · Ann Intern Med 2015;162(9):641-650

Le HOMA-IR, à partir de l'insuline et de la glycémie à jeun, donne une lecture précoce de la résistance à l'insulineEmerging

In plain terms

L'insuline à jeun, combinée à la glycémie à jeun dans une valeur appelée HOMA-IR, peut suggérer que votre corps travaille plus fort pour contrôler la glycémie avant que les tests habituels ne changent. C'est un indice précoce utile, pas un test définitif.

In detail

Le HOMA-IR se calcule comme l'insuline à jeun multipliée par la glycémie à jeun, divisée par une constante, un modèle calibré sur le clamp hyperinsulinémique-euglycémique. Il est raisonnablement corrélé à la résistance à l'insuline mesurée par clamp et peut augmenter alors que l'HbA1c et la glycémie à jeun sont encore normales, ce qui en fait un signal précoce intéressant. Ses limites sont pratiques : les dosages d'insuline ne sont pas bien standardisés entre laboratoires, la variation quotidienne de l'insuline à jeun est large, et il n'existe pas de seuil unique accepté, si bien qu'il est interprété comme une tendance et une fourchette approximative, non comme un seuil précis.

Who this may not transfer to:Derived from a model calibrated to clamp studies; read as a personal trend over time, not compared against a fixed cut-off between people.

The study · 1

Wallace, Levy and Matthews, use and abuse of HOMA modeling · Diabetes Care 2004;27(6):1487-1495

Les preuves sont insuffisantes pour recommander le dépistage de la vitamine D chez les adultes sans symptômesEmerging · mixed

In plain terms

Vérifier systématiquement la vitamine D chez des personnes qui se sentent bien n'a pas clairement démontré son utilité. Les preuves ne sont pas encore là, et les experts ne s'accordent même pas sur le seuil considéré comme bas.

In detail

Dans sa recommandation de 2021, l'US Preventive Services Task Force a jugé les preuves insuffisantes (déclaration I) pour évaluer l'équilibre entre bénéfices et risques du dépistage de la carence en vitamine D chez les adultes asymptomatiques, vivant en communauté et non enceintes. Une difficulté centrale est qu'il n'existe pas de seuil consensuel pour la carence, si bien qu'une même personne peut être qualifiée de carencée ou de suffisante selon le laboratoire, et les essais de dépistage-et-traitement chez des adultes non sélectionnés n'ont pas montré de bénéfice clair sur la santé. Le dosage reste raisonnable lorsqu'il existe une raison spécifique, comme une affection ou un médicament affectant la vitamine D.

Who this may not transfer to:The verdict covers asymptomatic community-dwelling non-pregnant adults; it does not extend to people with a specific reason to test.

The study · 1

US Preventive Services Task Force, screening for vitamin D deficiency in adults: US Preventive Services Task Force recommendation statement · JAMA 2021;325(14):1436-1442

Go Deeper

• Continuous glucose monitoring: the wearable way to watch blood sugar, and what it adds over a blood draw.
• Self-measures: what you can track at home without a blood draw, including how to read blood pressure correctly.
• Insulin and glucose: the physiology behind HbA1c, fasting glucose and HOMA-IR.
• Cholesterol and lipids: a fuller reading of the lipid panel and what moves it.

The Chinese Medicine View

Chinese medicine has no blood marker and does not reach for one. Instead of one variable at a time, the tradition reads a pattern, drawing many signs into a picture of a whole person in a particular state:

• tongue and pulse
• complexion and sleep
• digestion, temperature, and energy through the day

A single sign read alone is treated as a fragment. The classical texts warn that one finding out of context can mislead the practitioner.

The tradition also holds a distinction between root and branch, ben and biao. A number on a report is usually a branch, a visible manifestation. The practitioner's questions reach for the root behind it: the constitution and imbalance producing the reading, and why this person in this season shows it. A diagnosis such as Spleen failing to transform Dampness names that root, a whole-person pattern the practitioner then treats.

Cautions For This Practice

Extra restraint

Everything to be aware of is here, in one place. This practice suits most healthy people; a few situations call for real care.

Un résultat limite ou fortuit déclenche des cascades d'examens supplémentaires, préjudice rapporté par 991 médecins

Ganguli et collègues ont interrogé un échantillon national de 991 médecins américains sur les cascades de soins consécutives à des découvertes fortuites. Presque tous en avaient fait l'expérience, et une majorité a rapporté des préjudices en résultant, y compris un préjudice physique lié aux procédures en aval, un préjudice psychologique et de l'anxiété pour les patients, une perte de temps et un coût financier, ainsi qu'une frustration personnelle. Plus un panel mesure de marqueurs chez une personne sans symptômes, plus il produit de résultats limites et fortuits, et chacun d'entre eux peut déclencher une cascade comportant ses propres risques, indépendamment de l'existence réelle d'un problème.Ganguli et al., cascades of care after incidental findings in a US national survey of physicians

Read Results Against Your Full History

A result is information to bring to a clinician familiar with your full history, weighed together with the other numbers before anything is concluded.

One Abnormal Value Is Not a Diagnosis

Blood markers vary between days, laboratories and circumstances. A single value just outside the range calls for a repeat, and a diagnosis rests on several readings in context.

Do Not Change Treatment Off a Single Number

A home or one-off reading is a reason to ask a question before starting, stopping or adjusting a medication. Take the results and the trend to a prescriber before acting on a value alone.

Start slow, be smart, read the research, and consult a professional if you have any concerns. This is here to inform your choice, not make it for you.

Common Questions

Lp(a) is largely inherited, so what can I do about a high result?

For now, no widely approved treatment lowers Lp(a) with a proven change in outcomes. A high value still matters. It sharpens your overall risk picture and points you at the risk factors you can change. That is where prevention acts.

My clinician will not order ApoB or Lp(a). Can I get them another way?

Yes. Both are available through direct-to-consumer lab testing, where you order and pay for the test yourself without a clinician's requisition. Take any abnormal result to a clinician familiar with your history, since what you do about a number is a decision to make together.