Cardiovascular–kidney–metabolic disease is an umbrella term encompassing the interconnected disorders that dominate modern chronic illness: obesity, type 2 diabetes, hypertension, chronic kidney disease, atherosclerotic cardiovascular disease and heart failure. These conditions are increasingly understood not as separate entities but as stages of a single progressive syndrome in which metabolic dysfunction, vascular injury and declining kidney function reinforce one another. Decades of epidemiology have shown that the pathological seeds are planted early. Autopsy studies of young trauma victims performed by the Pathobiological Determinants of Atherosclerosis in Youth research group demonstrated that fatty streaks and intermediate atherosclerotic lesions are already present in the coronary arteries of adolescents, and that their extent tracks with classical risk factors measured during life. More recently, long-term cohort analyses published in the New England Journal of Medicine confirmed that cardiovascular risk factors recorded in childhood predict actual cardiovascular events in adulthood, not merely surrogate markers.

Despite this well-established evidence, identifying which children are genuinely at high risk has remained frustratingly difficult. Standard paediatric screening relies on measures such as body mass index, blood pressure, lipid panels and fasting glucose, yet many adults who suffer heart attacks or develop diabetes had unremarkable childhood profiles by these conventional metrics. Conversely, some children with elevated cholesterol or weight never progress to overt disease. The traditional risk variables are coarse instruments that capture only a fraction of the underlying biology. Proteomics, the large-scale measurement of the thousands of proteins circulating in blood plasma, offers a fundamentally richer view. Plasma proteins are the functional output of the genome, dynamically reflecting the activity of organs, immune processes, inflammation, vascular biology and metabolism in real time. A proteomic profile is therefore a molecular portrait of physiological state, one that can change with lifestyle and disease but also carry stable, trait-linked information.

In the new work, Landman and colleagues set out to determine whether the proteomic signatures associated with cardiometabolic disease traits in adults could already be detected in children, and whether those paediatric signatures carried prognostic weight. The researchers first characterized the relationship between plasma protein concentrations and CKMD-related phenotypes — including measures of adiposity, insulin resistance, blood pressure and lipid metabolism — in adult populations, establishing a reference map of the proteome–phenome relationships that accompany established disease and its antecedents. They then examined whether analogous protein patterns and phenotypic markers could be identified in children and adolescents under 20 years of age, testing the hypothesis that the disease-associated molecular architecture is not a late consequence of pathology but an early-emerging feature of predisposition.

The results were clear. The team identified a combined proteomic and phenotypic signature in children that mirrors the adult CKMD proteome and is strongly associated with the later development and clinical outcomes of the disease in adulthood. In practical terms, the same protein axes that encode cardiometabolic risk in middle-aged adults — involving inflammatory signalling, metabolic regulation, renal function and vascular remodelling — were detectable as coherent, measurable patterns in young subjects long before any clinical diagnosis was possible. The study, summarized in an accompanying Research Briefing in Nature Metabolism, demonstrates that these paediatric signatures are not merely statistical curiosities; they predict adult disease outcomes, providing a quantitative bridge between childhood biology and adult clinical events across a span of many decades.

Technically, this kind of longitudinal proteomic prediction requires sophisticated analytical machinery. High-throughput affinity-based platforms now allow thousands of plasma proteins to be quantified from small blood samples with high reproducibility, and machine-learning methods can distil these high-dimensional datasets into compact risk signatures that generalize across populations. The study’s integration of proteomic data with classical phenotypic traits represents an important methodological advance: by combining molecular and clinical dimensions, the model captures both the deep biology and the accessible measurements that a future paediatric screening programme would actually employ. The identification of proteins whose concentrations track modifiable risk, consistent with parallel work in children and adolescents with obesity published in Nature Communications, raises the possibility that these signatures are not destiny but sensitive readouts of early, reversible metabolic disturbance.

The implications for clinical practice are profound. Today, paediatric guidelines recommend targeted lipid and metabolic screening for children with family history or obesity, but coverage is incomplete and the tools are blunt. A validated proteomic risk score could, in principle, identify high-risk children from a routine blood draw, enabling earlier and more precisely targeted interventions — dietary counselling, structured physical activity, sleep and environmental measures, and in selected cases pharmacological therapy — at an age when arterial and metabolic damage is still minimal or reversible. The economic argument is equally compelling: cardiometabolic disease consumes an enormous share of global healthcare expenditure, and even modest improvements in the accuracy of childhood risk stratification could translate into substantial reductions in lifetime morbidity and cost.

Significant caveats remain before proteomic screening enters the paediatric clinic. The study establishes strong association, and while longitudinal links to adult outcomes are compelling, demonstrating that intervention guided by proteomic signatures actually improves clinical endpoints will require dedicated trials and independent replication across diverse populations. Proteomic assays are currently expensive relative to standard panels, and questions of ethical and psychological consequence — what it means to label a child as high-risk decades before disease — must be handled with care, particularly since many protein patterns appear modifiable with lifestyle change. Standardization across laboratories and ancestry groups, along with robust recalibration as children grow, will be essential for any scalable deployment.

Nevertheless, the study marks a conceptual milestone in preventive cardiology and metabolism. It confirms, at the molecular level, that cardiovascular–kidney–metabolic disease has detectable origins before age 20, and it transforms that observation from an epidemiological generality into an individual-level diagnostic prospect. Just as genomic medicine redefined how we think about inherited risk, paediatric proteomics now promises a dynamic, functional window into the trajectory toward chronic disease. If subsequent research validates and translates these signatures into routine practice, the era of waiting until middle age to diagnose the diseases that began in childhood may finally be drawing to a close — replaced by prevention that begins in the very decades when it can do the most good.

Subject of Research: Pediatric blood proteomic signatures that predict adult cardiovascular–kidney–metabolic disease outcomes

Article Title: Childhood proteomic signatures that predict adult cardiometabolic disease

Article References: Childhood proteomic signatures that predict adult cardiometabolic disease. (2026). Nature Metabolism. https://doi.org/10.1038/s42255-026-01588-8

Image Credits: AI Generated

DOI: 10.1038/s42255-026-01588-8

Keywords: proteomics, cardiometabolic disease, cardiovascular–kidney–metabolic disease, pediatric risk prediction, plasma biomarkers, Nature Metabolism, childhood obesity, atherosclerosis, type 2 diabetes, chronic kidney disease, prevention, machine learning