The human plasma proteome, the full collection of proteins measurable in the liquid portion of blood, sits at the intersection of genetics, metabolism and environment. Unlike DNA, which is essentially fixed from conception, the proteome responds dynamically to physiological states such as inflammation, insulin resistance, adipose tissue expansion and liver stress. Because many circulating proteins are secreted by adipose tissue or the liver and participate in cardiometabolic signaling, shifts in the plasma proteome are strong candidates for mediating the long-term consequences of childhood adiposity. Yet most previous studies captured only a single snapshot, typically in adulthood, leaving open the question of when in development these proteomic alterations first emerge and whether they are truly driven by fat mass or merely co-occur with it.

To address this, the research team analyzed plasma protein measurements at four timepoints across life, spanning childhood, adolescence and adulthood. Rather than relying on a single cohort or a single platform, the investigators combined multiple population-based birth cohorts in which participants had been followed from early life into adulthood, with blood samples collected, stored and later profiled using high-throughput affinity-based proteomic assays. These platforms, which measure thousands of unique proteins simultaneously using proximity extension assays or similar antibody-based technologies, allowed the researchers to quantify thousands of plasma proteins in the same individuals or closely matched cohorts at each life stage.

The central exposure was childhood adiposity, assessed through body mass index and other anthropometric indicators recorded during childhood and adolescence. The team then asked two complementary questions. First, is childhood adiposity associated with differences in plasma protein levels at the same timepoint, and does that association grow stronger as children age? Second, and more critically, does higher adiposity in childhood predict protein differences measured decades later, well after the original exposure, suggesting a durable biological imprint rather than a fleeting correlation?

The findings were striking. Higher childhood adiposity was associated with substantial proteomic differences detectable as early as childhood and adolescence, encompassing proteins involved in inflammation, leptin signaling, insulin action, extracellular matrix remodeling and lipoprotein metabolism. Among the most consistently affected proteins were well-established markers of cardiometabolic risk, including leptin, which scales tightly with fat mass, and inflammatory signaling molecules whose circulating levels are known to predict future cardiovascular events. The magnitude of these associations increased with age at measurement, indicating that the proteomic signature of excess adiposity intensifies as the metabolic consequences of obesity accumulate over time.

Importantly, many of the protein associations observed in childhood persisted into adulthood even after adjustment for adult body mass index. This persistence is a critical observation, because it suggests that childhood adiposity does more than simply track into adult overweight; it appears to leave a measurable molecular imprint on the circulating protein repertoire that is not fully explained by adult adiposity alone. In other words, a child who experienced excess body fat may carry a distinct proteomic profile decades later, even if they achieve a normal adult weight, pointing toward potential early-life biological embedding of cardiometabolic risk.

Associations alone cannot establish causation, and the researchers were careful to confront this limitation directly. Children with higher adiposity differ from their peers in diet, physical activity, socioeconomic circumstances and genetic liability, any of which could independently shape the proteome. To disentangle these possibilities, the study employed Mendelian randomization, a genetic epidemiological technique that uses naturally occurring genetic variants associated with adiposity as instrumental variables. Because genetic variants are randomly allocated at conception, they are largely immune to the confounding that plagues observational associations. When the team used genetic instruments for childhood body mass index to predict protein levels, they found that genetically influenced adiposity in childhood was linked to differences in many of the same proteins identified in the observational analyses, providing evidence consistent with a causal contribution of childhood fat mass to the plasma proteome.

The genetic analyses also illuminated which life stage matters most. By comparing the effects of genetic instruments for childhood adiposity with instruments for adult adiposity, the researchers could ask whether adult fat mass explains the protein associations or whether childhood-specific effects remain. For a subset of proteins, the childhood signal was independent of, and in some cases stronger than, the adult adiposity signal, supporting the idea that developmental windows exist during which excess adiposity exerts distinctive effects on circulating biology. This finding resonates with a broader literature showing that the timing of adiposity gain influences later disease risk more than body size at any single moment.

The study’s four-timepoint design also revealed trajectories. Some proteins showed associations that emerged in adolescence and strengthened through adulthood, consistent with cumulative metabolic injury. Others showed transient associations that attenuated with age, suggesting reversible responses to fat mass. A third group displayed persistent effects detectable decades after childhood, which the authors interpret as candidate mediators of the long-term disease burden attributable to early-life obesity. Proteins tied to inflammatory pathways, hepatic function and vascular biology featured prominently across these trajectories, aligning with the clinical observation that childhood obesity elevates risks of atherosclerosis, hypertension and metabolic liver disease well into midlife.

Technical rigor was a defining feature of the work. The investigators applied extensive quality control to the proteomic assays, adjusted for technical covariates such as batch effects and sample handling, and replicated key findings across independent cohorts and measurement platforms. Sensitivity analyses tested whether reverse causation, residual confounding or sample attrition could account for the results, and the consistency of protein associations across cohorts strengthened confidence in their robustness. The authors also acknowledged limitations, including the predominance of European-ancestry populations in available genetic reference data, the incomplete coverage of the proteome by current affinity-based platforms, and the possibility that some genetic instruments reflect correlated traits rather than adiposity itself.

The implications extend beyond observational epidemiology. Identifying which proteins are causally reshaped by childhood adiposity provides a starting point for understanding how early-life obesity is translated, at the molecular level, into adult disease. If specific circulating proteins prove to be genuine mediators, they could serve as early-warning biomarkers, allowing clinicians to identify children at elevated long-term risk before irreversible end-organ damage occurs. They could also expose therapeutic targets: drugs or interventions that normalize the proteomic consequences of excess adiposity during development might, in principle, interrupt the pathway from childhood obesity to adult cardiometabolic disease.

The work also contributes to an ongoing scientific debate about whether adult weight gain matters more or less than weight gained in childhood. Public health strategies have often focused on adult obesity, but accumulating evidence, including this proteomic analysis, suggests that adiposity during growth periods has biological consequences that are not simply undone by later weight normalization. The persistence of protein differences after accounting for adult body mass index argues for prevention efforts that begin in childhood, a message with significant relevance as childhood obesity rates continue to rise globally.

From a methodological standpoint, the study exemplifies a growing trend in life-course epidemiology: combining longitudinal cohorts, multiplexed proteomics and genetic causal inference to move from correlation toward mechanism. As proteomic platforms become cheaper and more comprehensive, and as large-scale genetic reference datasets expand across diverse ancestries, similar designs will be applied to other exposures, from early-life nutrition to environmental pollutants. The present findings establish childhood adiposity as a developmental exposure with quantifiable, durable and partly causal effects on the human plasma proteome, offering a molecular narrative for a clinical relationship that physicians have recognized for decades but could not previously explain at the protein level.

For researchers, the catalog of proteins altered by childhood adiposity represents a resource for mechanistic follow-up, including studies of adipose tissue biology, hepatic metabolism and vascular inflammation. For clinicians and policymakers, the results reinforce the urgency of early intervention. And for the thousands of cohort participants whose blood samples, collected decades ago in childhood, made the analysis possible, the work demonstrates the extraordinary scientific value of long-term longitudinal research in unraveling how the earliest chapters of biological life shape its later ones.

Subject of Research: Effects of childhood adiposity on the human plasma proteome across four lifecourse timepoints

Article Title: Evaluating the consequences of childhood adiposity on the human plasma proteome at four timepoints across the lifecourse

Article References: Dickson, P., Power, G. M., Niu, L., Holm, J.-C., Hansen, T., Turner, R. M., Gaunt, T. R., Davey Smith, G., & Richardson, T. G. (2026). Evaluating the consequences of childhood adiposity on the human plasma proteome at four timepoints across the lifecourse. Nature Communications. https://doi.org/10.1038/s41467-026-77097-9

Image Credits: AI Generated

DOI: 10.1038/s41467-026-77097-9

Keywords: childhood adiposity, plasma proteome, proteomics, Mendelian randomization, lifecourse epidemiology, obesity, cardiometabolic risk, inflammation, leptin, birth cohorts, protein biomarkers, Nature Communications

The investigation was built on the Danish Life Course (DANLIFE) cohort, a register-based resource encompassing all children born in Denmark since 1980. Because every Danish resident carries a unique personal identification number, researchers at the University of Copenhagen and collaborators could link national registries covering births, hospital contacts, psychiatric diagnoses, social welfare, and residential addresses into a single longitudinal record. The analytical sample comprised 1,235,519 individuals who had complete adversity information and survived to age 16. The team, led by Naja Hulvej Rod, conceptualised adversity along two dimensions: time, measured annually from birth to age 16, and layers, spanning the individual, the family, and the neighbourhood.

At the individual layer, the researchers captured three indicators. Perinatal adversity was defined as preterm birth before 37 weeks of gestation or being small for gestational age, below the tenth percentile on standardised growth curves. Physical-health adversity was operationalised as membership in the top 20 percent of cumulative emergency and inpatient hospital contacts between ages 0 and 15, drawn from the National Patient Registry. Mental-health adversity was any psychiatric diagnosis recorded before age 16 in the Danish Psychiatric Central Research Register. These markers reflect biological susceptibility and early disease burden that may compound later social exposures.

The family layer used group-based multi-trajectory modelling of annual adversity counts across three expert-identified dimensions: material deprivation, encompassing family poverty and parental long-term unemployment; loss or threat of loss, covering serious illness or death of a parent or sibling; and family dynamics, including maternal separation, foster-care placement, and parental or sibling psychiatric illness or substance abuse. Five distinct trajectory groups emerged: low adversity, early material deprivation, persistent material deprivation, loss or threat of loss, and high adversity, the last characterised by escalating hardship across all three dimensions simultaneously. Each individual was assigned to the trajectory with the highest probability of membership.

The neighbourhood layer broke new ground for cohort research of this scale. Denmark was divided into 1,885 small-area data zones of roughly 2,500 inhabitants each, nested within 98 municipalities, using residential coordinates and a clustering algorithm. Neighbourhood material deprivation was assessed annually through four indicators: the proportion of residents with low income, basic education only, unemployment, and overcrowded housing with more than one person per room. High neighbourhood deprivation was defined as falling in the top 20 percent for at least two of these indicators averaged across childhood, providing a granular portrait of the structural conditions surrounding each child as they grew up.

The results revealed a striking pattern of co-occurrence. Children in the high family-adversity group were far more likely than their peers to have been born small for gestational age, to receive a childhood mental-health diagnosis, and to be high users of hospital services, while those in the persistent material-deprivation group most often lived in deprived neighbourhoods. Adversity, in other words, clusters. A child facing poverty or parental illness is also more likely to face perinatal complications, health difficulties, and neighbourhood disadvantage, producing cascading patterns in which individual health, family conditions, and place act as both causes and consequences of one another across development and generations.

Each layer was independently associated with mortality when analysed separately using Cox proportional hazards models, complemented by Aalen’s additive hazards models to quantify absolute effects. Perinatal adversity carried a hazard ratio of 1.36, corresponding to 14 excess deaths per 100,000 person-years. A childhood mental-health diagnosis tripled the risk, with a hazard ratio of 3.00 and 68 excess deaths per 100,000 person-years, while high physical-health service use yielded a hazard ratio of 2.36. High family adversity showed the strongest single-layer association at a hazard ratio of 3.95, or 95 excess deaths per 100,000 person-years, and neighbourhood deprivation contributed a hazard ratio of 1.20. The leading causes of death were suicide and assault, accidents, and cancer, all socially patterned outcomes.

Crucially, the additive hazards analysis uncovered cross-layer interactions, meaning more deaths occurred than the sum of each layer’s separate effects would predict. Children exposed to both high family adversity and a mental-health diagnosis experienced 181 excess deaths per 100,000 person-years, of which 64 were attributable to the interaction itself. High family adversity combined with high physical-health service use produced 194 excess deaths per 100,000 person-years, with an estimated 106 due to interaction. Perinatal adversity plus high family adversity yielded 116 excess deaths, 22 attributable to interaction. No interaction emerged between family and neighbourhood adversity, though their combined independent effects still produced 89 excess deaths per 100,000 person-years. The highest cumulative risk appeared among the 17,810 individuals in the high family-adversity group who also experienced individual-layer adversity: a hazard ratio of 7.16, with 434 deaths in this group representing just 1.4 percent of the population. Even among children with low family adversity, individual and neighbourhood adversity together raised the hazard ratio to 1.82, showing that no layer of adversity is harmless in isolation.

The authors emphasise that these findings, derived from nearly complete national lifecourse data over four decades, empirically validate long-standing theoretical frameworks, from Bronfenbrenner’s ecological model to the bio-exposome concept, that had rarely been operationalised at such scale. Sensitivity analyses adjusting for parental education, restricting to birth years with unextrapolated neighbourhood data, and weighting for missing data all confirmed robustness. Limitations remain: registry data could not capture unreported abuse, undiagnosed illness, air pollution, racism, or bullying, likely leading to underestimation of effects, and results are conditional on survival to age 16. Denmark’s universal health care and strong social-security system may also make these estimates a lower bound for less supportive settings. The public-health implications, however, are clear. Because adversity at one layer amplifies harm at another, policies must operate across levels simultaneously: reducing preterm births, supporting families in crisis, expanding mental-health services, and tackling poverty, education, housing, and neighbourhood deprivation. Such integrated, multi-layered intervention offers the best hope of identifying highly vulnerable children early and breaking intergenerational cycles of disadvantage before they culminate in premature death.

Subject of Research: Multilayered childhood adversity and its association with mortality in young adulthood

Article Title: Multilayered childhood adversity and mortality: a population-based cohort study of 1.2 million individuals

Article References: Multilayered childhood adversity and mortality: a population-based cohort study of 1.2 million individuals. (n.d.). https://doi.org/10.1016/j.lanepe.2026.101863

Image Credits: AI Generated

DOI: 10.1016/j.lanepe.2026.101863

Keywords: childhood adversity, mortality, DANLIFE cohort, health inequality, adverse childhood experiences, neighbourhood deprivation, perinatal adversity, mental health, family adversity, population-based cohort, lifecourse epidemiology, social determinants