NMR metabolomics: how it works, the 249-biomarker panel, cohorts and value in prevention

In just a few years, NMR metabolomics has moved from the research laboratory into prevention check-ups. What exactly does it measure, what data is it based on, and what can you reasonably expect from it for your own health? This article answers point by point, drawing on the key publications, and clarifies what this technology does not do.

What is metabolomics?

Metabolomics is the study of all the small molecules, called metabolites, present in a biological sample: sugars, lipids, amino acids and products of energy metabolism. Together, these molecules make up the metabolome. After genomics (genes) and proteomics (proteins), metabolomics is the layer closest to what the body is actually doing at a given moment: it integrates the effects of genes, diet, physical activity, sleep, medications and the environment.

This is what makes it interesting for prevention: the genome does not change, whereas the metabolome shifts within a few weeks in response to lifestyle habits. It provides information about a current state and a trajectory, not just about a predisposition.

A distinction is made between targeted metabolomics, which precisely quantifies a defined list of molecules, and untargeted metabolomics, which explores as many signals as possible to discover new markers. Prevention panels belong to the first category: the same molecules, the same units and the same methods from one sample to the next, which is a prerequisite for comparing a result with reference cohorts and following a person over time.

How does NMR work on a blood sample?

Nuclear magnetic resonance (NMR) is a physical technique that places the sample in a strong magnetic field and measures the response of atomic nuclei, mainly hydrogen nuclei, to a radiofrequency pulse. Each molecule produces a characteristic signature, a spectrum, whose intensity is proportional to its concentration. It is the same physical principle as MRI, applied to a tube of plasma or serum rather than to the whole body.

Three properties explain the success of NMR in population metabolomics:

• It is quantitative and reproducible. The signal is proportional to the amount of the molecule; measurements are comparable from one laboratory and one year to the next, which makes it possible to build reference databases of hundreds of thousands of samples.
• It does not alter the sample and requires little preparation, allowing large measurement volumes at a controlled cost.
• It sees lipoproteins in detail. NMR distinguishes particles by class and size and quantifies the cholesterol, triglycerides and phospholipids in each, information that routine laboratory testing does not provide.

Its limitation is lower sensitivity than mass spectrometry: NMR measures molecules that are abundant in the blood, not trace compounds. For prevention, this trade-off makes sense, because the best-established risk markers are precisely abundant molecules.

The Nightingale Health platform combines this NMR spectroscopy with machine learning algorithms that translate the raw spectrum into concentrations for each biomarker: automated quantification that is consistent across the millions of samples analyzed.

What does the Nightingale Health 249-biomarker panel measure?

The Nightingale Health panel (Helsinki, Finland) quantifies 249 metabolic biomarkers from a single blood sample. It covers lipoproteins by class and size, fatty acids, amino acids, glycolysis markers, ketone bodies and low-grade inflammation, measured by GlycA.

Specifically:

• Detailed lipoproteins: particle concentration, average size and composition (cholesterol, cholesteryl esters, free cholesterol, triglycerides, phospholipids) for each subclass, from VLDL to HDL, plus apolipoproteins A1 and B.
• Standard lipids: total cholesterol, LDL, HDL, non-HDL and triglycerides, in the same units as a routine blood test.
• Fatty acids: total, saturated, monounsaturated, polyunsaturated, omega-3, omega-6, docosahexaenoic acid and linoleic acid, and their relative proportions.
• Glycolysis and energy metabolism: glucose, lactate, pyruvate and citrate, plus ketone bodies (acetoacetate, 3-hydroxybutyrate, acetone).
• Amino acids: alanine, glutamine, glycine, histidine, isoleucine, leucine, valine, phenylalanine and tyrosine.
• Other: creatinine, albumin and GlycA, an integrated marker of chronic inflammation.

According to Nightingale Health, the platform has analyzed more than two million samples, has given rise to more than 900 peer-reviewed publications and is used by more than 200 research organizations in more than 30 countries, in fields ranging from cardiovascular disease and diabetes to liver, kidney and lung disease, aging and cancer. For the meaning of each family of markers, see our guide to blood biomarkers.

Which cohorts underpin NMR metabolomics?

NMR metabolomics draws its value from the cohorts in which it has been applied, first and foremost UK Biobank: a biomarker is meaningful for prevention only if we know, from a large number of people followed for years, which health events it is associated with.

UK Biobank is a cohort of about 500,000 British volunteers aged 40 to 69 at recruitment, between 2006 and 2010, followed for at least thirty years. It links genomic data (whole-genome sequencing of all 500,000 participants in 2023), imaging, health records and, for NMR metabolomics, a first wave of about 120,000 participants released in 2021, followed by an extension to the entire cohort. More than 9,000 peer-reviewed publications had drawn on it as of November 2023.

Other cohorts have adopted the same panel, which makes it possible to check that the associations observed in the United Kingdom hold elsewhere: the Estonian Biobank (200,000 samples), the Mexico City Prospective Study (150,000), the BELIEVE study in Bangladesh (100,000), the Finnish National Biobank (45,000), as well as ARIC in the United States, the Rotterdam Study, INTERVAL, PREDIMED Plus and the China Kadoorie Biobank. A marker validated in several populations is more robust than one validated in a single population.

What do the key publications show?

The key publications show that an NMR metabolomic profile from a single blood sample is associated with the risk of many diseases and with mortality, and that it adds predictive information to traditional risk factors for several of them. None of them concludes that it has diagnostic capability.

Two readings of this table are both correct. First, the blood metabolome carries real, reproducible risk information that goes beyond what could be measured routinely before. Second, this information is probabilistic, established at the population level, and translating it for an individual requires a doctor. The claim that “metabolomics detects diseases years in advance,” which one sometimes reads, is not what these studies say; they show that the metabolomic profile improves the prediction of 10-year risk for several diseases.

What does NMR metabolomics offer, and what does it not do?

NMR metabolomics offers a richer risk estimate than traditional factors alone, an overall picture of metabolism from a single blood sample, and a reproducible point of comparison for tracking a trajectory. It does not make a diagnosis, does not replace a standard blood test or organized screening programs, and does not remove the need for a clinical examination.

What it offers

• A complement to risk scores. Scores such as SCORE2 rely on a handful of variables; metabolomics adds dozens more (detailed lipoproteins, inflammation, fatty acids). We deliberately use the phrase “more information” rather than “more accurate,” in the absence of a direct comparative study in French practice.
• A systemic view. Research on UK Biobank suggests that the same metabolic profile is associated with very different diseases; a single analysis provides information on several risk axes at once.
• Tracking over time. Because the measurement is standardized, two samples taken a year apart are comparable; since the metabolome responds to lifestyle habits, a change can be documented, in either direction, without any being promised. Metabolic age is one way of summarizing it.

What it does not do

• It does not diagnose. A biomarker associated with a disease in a cohort does not mean that a person has or will have that disease.
• It does not measure everything. No complete blood count, no vitamins, no tumor markers, no hormones: the standard blood test is still needed for what it does well.
• It does not replace clinical assessment. Medical history, blood pressure, waist circumference, smoking, sleep, symptoms: most of the risk estimate remains clinical.
• It cannot be interpreted on its own. A panel of 249 values will statistically contain a few atypical results with no significance; read without a doctor, they cause worry and trigger unnecessary tests.

How can you get an NMR metabolomic analysis in practice?

In France, NMR metabolomics remains mainly a research tool; a few prevention services offer it as an option, under medical responsibility, with interpretation by a doctor. It is not reimbursed by French National Health Insurance (Assurance Maladie), unlike the official prevention check-up, which is fully covered (100%) at four key ages and is the first step to consider.

Before using it, three questions arise: would a result change a decision for me (if not, the test is not useful); who interprets and signs the report; and will my regular doctor (médecin traitant) be informed? These are the criteria of personalized health properly understood. Professionals will find more details on our pages Sokrate for doctors and Sokrate for laboratories.

Key takeaways

• Metabolomics studies the small molecules in the blood; NMR quantifies them simultaneously and reproducibly, without destroying the sample.
• The Nightingale Health panel measures 249 biomarkers: detailed lipoproteins, fatty acids, amino acids, glycolysis, ketone bodies and inflammation (GlycA).
• Its value rests on cohorts such as UK Biobank (about 500,000 participants) and on key publications (Deelen 2019, Buergel 2022, Julkunen 2023, Karjalainen 2024).
• It refines the estimate of 10-year risk for several diseases; it does not make a diagnosis and replaces neither the standard blood test nor clinical assessment.
• A result should be interpreted with a doctor and fit within the covered care pathway.

What Sokrate lets you do

Sokrate relies on the Nightingale Health NMR metabolomics platform to offer, as an option to its prevention check-up, an analysis of 249 biomarkers from a single blood sample, interpreted by a doctor and explained in plain language: an estimate of your risks, priorities and modifiable levers; the report is sent to your regular doctor unless you object. Results are expected within 7 days. The prevention check-up itself, prepared online (Sokrate 360 questionnaire) and signed by a doctor, remains fully covered (100%) by French National Health Insurance when carried out by an authorized professional. See how the Sokrate pathway works or join the waitlist.