The transition from genomics to phenomics in personalized population health.

Modern health care faces several serious challenges, including an ageing population and its inherent burden of chronic diseases, rising costs and marginal quality metrics. By assessing and optimizing the health trajectory of each individual using a data-driven personalized approach that reflects their genetics, behaviour and environment, we can start to address these challenges. This assessment includes longitudinal phenome measures, such as the blood proteome and metabolome, gut microbiome composition and function, and lifestyle and behaviour through wearables and questionnaires. Here, we review ongoing large-scale genomics and longitudinal phenomics efforts and the powerful insights they provide into wellness. We describe our vision for the transformation of the current health care from disease-oriented to data-driven, wellness-oriented and personalized population health.

MalaCards: an amalgamated human disease compendium with diverse clinical and genetic annotation and structured search.

The MalaCards human disease database (http://www.malacards.org/) is an integrated compendium of annotated diseases mined from 68 data sources. MalaCards has a web card for each of ∼20 000 disease entries, in six global categories. It portrays a broad array of annotation topics in 15 sections, including Summaries, Symptoms, Anatomical Context, Drugs, Genetic Tests, Variations and Publications. The Aliases and Classifications section reflects an algorithm for disease name integration across often-conflicting sources, providing effective annotation consolidation. A central feature is a balanced Genes section, with scores reflecting the strength of disease-gene associations. This is accompanied by other gene-related disease information such as pathways, mouse phenotypes and GO-terms, stemming from MalaCards' affiliation with the GeneCards Suite of databases. MalaCards' capacity to inter-link information from complementary sources, along with its elaborate search function, relational database infrastructure and convenient data dumps, allows it to tackle its rich disease annotation landscape, and facilitates systems analyses and genome sequence interpretation. MalaCards adopts a 'flat' disease-card approach, but each card is mapped to popular hierarchical ontologies (e.g. International Classification of Diseases, Human Phenotype Ontology and Unified Medical Language System) and also contains information about multi-level relations among diseases, thereby providing an optimal tool for disease representation and scrutiny.

Rational confederation of genes and diseases: NGS interpretation via GeneCards, MalaCards and VarElect.

BACKGROUND: A key challenge in the realm of human disease research is next generation sequencing (NGS) interpretation, whereby identified filtered variant-harboring genes are associated with a patient's disease phenotypes. This necessitates bioinformatics tools linked to comprehensive knowledgebases. The GeneCards suite databases, which include GeneCards (human genes), MalaCards (human diseases) and PathCards (human pathways) together with additional tools, are presented with the focus on MalaCards utility for NGS interpretation as well as for large scale bioinformatic analyses. RESULTS: VarElect, our NGS interpretation tool, leverages the broad information in the GeneCards suite databases. MalaCards algorithms unify disease-related terms and annotations from 69 sources. Further, MalaCards defines hierarchical relatedness-aliases, disease families, a related diseases network, categories and ontological classifications. GeneCards and MalaCards delineate and share a multi-tiered, scored gene-disease network, with stringency levels, including the definition of elite status-high quality gene-disease pairs, coming from manually curated trustworthy sources, that includes 4500 genes for 8000 diseases. This unique resource is key to NGS interpretation by VarElect. VarElect, a comprehensive search tool that helps infer both direct and indirect links between genes and user-supplied disease/phenotype terms, is robustly strengthened by the information found in MalaCards. The indirect mode benefits from GeneCards' diverse gene-to-gene relationships, including SuperPaths-integrated biological pathways from 12 information sources. We are currently adding an important information layer in the form of "disease SuperPaths", generated from the gene-disease matrix by an algorithm similar to that previously employed for biological pathway unification. This allows the discovery of novel gene-disease and disease-disease relationships. The advent of whole genome sequencing necessitates capacities to go beyond protein coding genes. GeneCards is highly useful in this respect, as it also addresses 101,976 non-protein-coding RNA genes. In a more recent development, we are currently adding an inclusive map of regulatory elements and their inferred target genes, generated by integration from 4 resources. CONCLUSIONS: MalaCards provides a rich big-data scaffold for in silico biomedical discovery within the gene-disease universe. VarElect, which depends significantly on both GeneCards and MalaCards power, is a potent tool for supporting the interpretation of wet-lab experiments, notably NGS analyses of disease. The GeneCards suite has thus transcended its 2-decade role in biomedical research, maturing into a key player in clinical investigation.

VarElect: the phenotype-based variation prioritizer of the GeneCards Suite.

BACKGROUND: Next generation sequencing (NGS) provides a key technology for deciphering the genetic underpinnings of human diseases. Typical NGS analyses of a patient depict tens of thousands non-reference coding variants, but only one or very few are expected to be significant for the relevant disorder. In a filtering stage, one employs family segregation, rarity in the population, predicted protein impact and evolutionary conservation as a means for shortening the variation list. However, narrowing down further towards culprit disease genes usually entails laborious seeking of gene-phenotype relationships, consulting numerous separate databases. Thus, a major challenge is to transition from the few hundred shortlisted genes to the most viable disease-causing candidates. RESULTS: We describe a novel tool, VarElect ( http://ve.genecards.org ), a comprehensive phenotype-dependent variant/gene prioritizer, based on the widely-used GeneCards, which helps rapidly identify causal mutations with extensive evidence. The GeneCards suite offers an effective and speedy alternative, whereby >120 gene-centric automatically-mined data sources are jointly available for the task. VarElect cashes on this wealth of information, as well as on GeneCards' powerful free-text Boolean search and scoring capabilities, proficiently matching variant-containing genes to submitted disease/symptom keywords. The tool also leverages the rich disease and pathway information of MalaCards, the human disease database, and PathCards, the unified pathway (SuperPaths) database, both within the GeneCards Suite. The VarElect algorithm infers direct as well as indirect links between genes and phenotypes, the latter benefitting from GeneCards' diverse gene-to-gene data links in GenesLikeMe. Finally, our tool offers an extensive gene-phenotype evidence portrayal ("MiniCards") and hyperlinks to the parent databases. CONCLUSIONS: We demonstrate that VarElect compares favorably with several often-used NGS phenotyping tools, thus providing a robust facility for ranking genes, pointing out their likelihood to be related to a patient's disease. VarElect's capacity to automatically process numerous NGS cases, either in stand-alone format or in VCF-analyzer mode (TGex and VarAnnot), is indispensable for emerging clinical projects that involve thousands of whole exome/genome NGS analyses.

Non-redundant compendium of human ncRNA genes in GeneCards.

MOTIVATION: Non-coding RNA (ncRNA) genes are increasingly acknowledged for their importance in the human genome. However, there is no comprehensive non-redundant database for all such human genes. RESULTS: We leveraged the effective platform of GeneCards, the human gene compendium, together with the power of fRNAdb and additional primary sources, to judiciously unify all ncRNA gene entries obtainable from 15 different primary sources. Overlapping entries were clustered to unified locations based on an algorithm employing genomic coordinates. This allowed GeneCards' gamut of relevant entries to rise ∼5-fold, resulting in ∼80,000 human non-redundant ncRNAs, belonging to 14 classes. Such 'grand unification' within a regularly updated data structure will assist future ncRNA research. AVAILABILITY AND IMPLEMENTATION: All of these non-coding RNAs are included among the ∼122,500 entries in GeneCards V3.09, along with pertinent annotation, automatically mined by its built-in pipeline from 100 data sources. This information is available at www.genecards.org. CONTACT: Frida.Belinky@weizmann.ac.il SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A framework towards digital twins for type 2 diabetes.

Introduction: A digital twin is a virtual representation of a patient's disease, facilitating real-time monitoring, analysis, and simulation. This enables the prediction of disease progression, optimization of care delivery, and improvement of outcomes. Methods: Here, we introduce a digital twin framework for type 2 diabetes (T2D) that integrates machine learning with multiomic data, knowledge graphs, and mechanistic models. By analyzing a substantial multiomic and clinical dataset, we constructed predictive machine learning models to forecast disease progression. Furthermore, knowledge graphs were employed to elucidate and contextualize multiomic-disease relationships. Results and discussion: Our findings not only reaffirm known targetable disease components but also spotlight novel ones, unveiled through this integrated approach. The versatile components presented in this study can be incorporated into a digital twin system, enhancing our grasp of diseases and propelling the advancement of precision medicine.

Generally-healthy individuals with aberrant bowel movement frequencies show enrichment for microbially-derived blood metabolites associated with reduced kidney function.

Bowel movement frequency (BMF) has been linked to changes in the composition of the human gut microbiome and to many chronic conditions, like metabolic disorders, neurodegenerative diseases, chronic kidney disease (CKD), and other intestinal pathologies like irritable bowel syndrome and inflammatory bowel disease. Lower BMF (constipation) can lead to compromised intestinal barrier integrity and a switch from saccharolytic to proteolytic fermentation within the microbiota, giving rise to microbially-derived toxins that may make their way into circulation and cause damage to organ systems. However, the connections between BMF, gut microbial metabolism, and the early-stage development and progression of chronic disease remain underexplored. Here, we examined the phenotypic impact of BMF variation in a cohort of generally-healthy, community dwelling adults with detailed clinical, lifestyle, and multi-omic data. We showed significant differences in microbially-derived blood plasma metabolites, gut bacterial genera, clinical chemistries, and lifestyle factors across BMF groups that have been linked to inflammation, cardiometabolic health, liver function, and CKD severity and progression. We found that the higher plasma levels of 3-indoxyl sulfate (3-IS), a microbially-derived metabolite associated with constipation, was in turn negatively associated with estimated glomerular filtration rate (eGFR), a measure of kidney function. Causal mediation analysis revealed that the effect of BMF on eGFR was significantly mediated by 3-IS. Finally, we identify self-reported diet, lifestyle, and psychological factors associated with BMF variation, which indicate several common-sense strategies for mitigating constipation and diarrhea. Overall, we suggest that aberrant BMF is an underappreciated risk factor in the development of chronic diseases, even in otherwise healthy populations.

The relationship between 11 different polygenic longevity scores, parental lifespan, and disease diagnosis in the UK Biobank.

Large-scale genome-wide association studies (GWAS) strongly suggest that most traits and diseases have a polygenic component. This observation has motivated the development of disease-specific "polygenic scores (PGS)" that are weighted sums of the effects of disease-associated variants identified from GWAS that correlate with an individual's likelihood of expressing a specific phenotype. Although most GWAS have been pursued on disease traits, leading to the creation of refined "Polygenic Risk Scores" (PRS) that quantify risk to diseases, many GWAS have also been pursued on extreme human longevity, general fitness, health span, and other health-positive traits. These GWAS have discovered many genetic variants seemingly protective from disease and are often different from disease-associated variants (i.e., they are not just alternative alleles at disease-associated loci) and suggest that many health-positive traits also have a polygenic basis. This observation has led to an interest in "polygenic longevity scores (PLS)" that quantify the "risk" or genetic predisposition of an individual towards health. We derived 11 different PLS from 4 different available GWAS on lifespan and then investigated the properties of these PLS using data from the UK Biobank (UKB). Tests of association between the PLS and population structure, parental lifespan, and several cancerous and non-cancerous diseases, including death from COVID-19, were performed. Based on the results of our analyses, we argue that PLS are made up of variants not only robustly associated with parental lifespan, but that also contribute to the genetic architecture of disease susceptibility, morbidity, and mortality.

Microbial community-scale metabolic modelling predicts personalized short-chain fatty acid production profiles in the human gut.

Microbially derived short-chain fatty acids (SCFAs) in the human gut are tightly coupled to host metabolism, immune regulation and integrity of the intestinal epithelium. However, the production of SCFAs can vary widely between individuals consuming the same diet, with lower levels often associated with disease. A systems-scale mechanistic understanding of this heterogeneity is lacking. Here we use a microbial community-scale metabolic modelling (MCMM) approach to predict individual-specific SCFA production profiles to assess the impact of different dietary, prebiotic and probiotic inputs. We evaluate the quantitative accuracy of our MCMMs using in vitro and ex vivo data, plus published human cohort data. We find that MCMM SCFA predictions are significantly associated with blood-derived clinical chemistries, including cardiometabolic and immunological health markers, across a large human cohort. Finally, we demonstrate how MCMMs can be leveraged to design personalized dietary, prebiotic and probiotic interventions aimed at optimizing SCFA production in the gut. Our model represents an approach to direct gut microbiome engineering for precision health and nutrition.

Proteomic changes induced by longevity-promoting interventions in mice.

Using mouse models and high-throughput proteomics, we conducted an in-depth analysis of the proteome changes induced in response to seven interventions known to increase mouse lifespan. This included two genetic mutations, a growth hormone receptor knockout (GHRKO mice) and a mutation in the Pit-1 locus (Snell dwarf mice), four drug treatments (rapamycin, acarbose, canagliflozin, and 17α-estradiol), and caloric restriction. Each of the interventions studied induced variable changes in the concentrations of proteins across liver, kidney, and gastrocnemius muscle tissue samples, with the strongest responses in the liver and limited concordance in protein responses across tissues. To the extent that these interventions promote longevity through common biological mechanisms, we anticipated that proteins associated with longevity could be identified by characterizing shared responses across all or multiple interventions. Many of the proteome alterations induced by each intervention were distinct, potentially implicating a variety of biological pathways as being related to lifespan extension. While we found no protein that was affected similarly by every intervention, we identified a set of proteins that responded to multiple interventions. These proteins were functionally diverse but tended to be involved in peroxisomal oxidation and metabolism of fatty acids. These results provide candidate proteins and biological mechanisms related to enhancing longevity that can inform research on therapeutic approaches to promote healthy aging.

Microbial community-scale metabolic modeling predicts personalized short chain fatty acid production profiles in the human gut.

Microbially-derived short chain fatty acids (SCFAs) in the human gut are tightly coupled to host metabolism, immune regulation, and integrity of the intestinal epithelium. However, the production of SCFAs can vary widely between individuals consuming the same diet, with lower levels often associated with disease. A systems-scale mechanistic understanding of this heterogeneity is lacking. We present a microbial community-scale metabolic modeling (MCMM) approach to predict individual-specific SCFA production profiles. We assess the quantitative accuracy of our MCMMs using in vitro, ex vivo, and in vivo data. Next, we show how MCMM SCFA predictions are significantly associated with blood-derived clinical chemistries, including cardiometabolic and immunological health markers, across a large human cohort. Finally, we demonstrate how MCMMs can be leveraged to design personalized dietary, prebiotic, and probiotic interventions that optimize SCFA production in the gut. Our results represent an important advance in engineering gut microbiome functional outputs for precision health and nutrition.

Multiomic signatures of body mass index identify heterogeneous health phenotypes and responses to a lifestyle intervention.

Multiomic profiling can reveal population heterogeneity for both health and disease states. Obesity drives a myriad of metabolic perturbations and is a risk factor for multiple chronic diseases. Here we report an atlas of cross-sectional and longitudinal changes in 1,111 blood analytes associated with variation in body mass index (BMI), as well as multiomic associations with host polygenic risk scores and gut microbiome composition, from a cohort of 1,277 individuals enrolled in a wellness program (Arivale). Machine learning model predictions of BMI from blood multiomics captured heterogeneous phenotypic states of host metabolism and gut microbiome composition better than BMI, which was also validated in an external cohort (TwinsUK). Moreover, longitudinal analyses identified variable BMI trajectories for different omics measures in response to a healthy lifestyle intervention; metabolomics-inferred BMI decreased to a greater extent than actual BMI, whereas proteomics-inferred BMI exhibited greater resistance to change. Our analyses further identified blood analyte-analyte associations that were modified by metabolomics-inferred BMI and partially reversed in individuals with metabolic obesity during the intervention. Taken together, our findings provide a blood atlas of the molecular perturbations associated with changes in obesity status, serving as a resource to quantify metabolic health for predictive and preventive medicine.

Lifespan-extending interventions induce consistent patterns of fatty acid oxidation in mouse livers.

Aging manifests as progressive deteriorations in homeostasis, requiring systems-level perspectives to investigate the gradual molecular dysregulation of underlying biological processes. Here, we report systemic changes in the molecular regulation of biological processes under multiple lifespan-extending interventions. Differential Rank Conservation (DIRAC) analyses of mouse liver proteomics and transcriptomics data show that mechanistically distinct lifespan-extending interventions (acarbose, 17α-estradiol, rapamycin, and calorie restriction) generally tighten the regulation of biological modules. These tightening patterns are similar across the interventions, particularly in processes such as fatty acid oxidation, immune response, and stress response. Differences in DIRAC patterns between proteins and transcripts highlight specific modules which may be tightened via augmented cap-independent translation. Moreover, the systemic shifts in fatty acid metabolism are supported through integrated analysis of liver transcriptomics data with a mouse genome-scale metabolic model. Our findings highlight the power of systems-level approaches for identifying and characterizing the biological processes involved in aging and longevity.

A metabolomic signature of the APOE2 allele.

With the goal of identifying metabolites that significantly correlate with the protective e2 allele of the apolipoprotein E (APOE) gene, we established a consortium of five studies of healthy aging and extreme human longevity with 3545 participants. This consortium includes the New England Centenarian Study, the Baltimore Longitudinal Study of Aging, the Arivale study, the Longevity Genes Project/LonGenity studies, and the Long Life Family Study. We analyzed the association between APOE genotype groups E2 (e2e2 and e2e3 genotypes, N = 544), E3 (e3e3 genotypes, N = 2299), and E4 (e3e4 and e4e4 genotypes, N = 702) with metabolite profiles in the five studies and used fixed effect meta-analysis to aggregate the results. Our meta-analysis identified a signature of 19 metabolites that are significantly associated with the E2 genotype group at FDR < 10%. The group includes 10 glycerolipids and 4 glycerophospholipids that were all higher in E2 carriers compared to E3, with fold change ranging from 1.08 to 1.25. The organic acid 6-hydroxyindole sulfate, previously linked to changes in gut microbiome that were reflective of healthy aging and longevity, was also higher in E2 carriers compared to E3 carriers. Three sterol lipids and one sphingolipid species were significantly lower in carriers of the E2 genotype group. For some of these metabolites, the effect of the E2 genotype opposed the age effect. No metabolites reached a statistically significant association with the E4 group. This work confirms and expands previous results connecting the APOE gene to lipid regulation and suggests new links between the e2 allele, lipid metabolism, aging, and the gut-brain axis.

Genome-microbiome interplay provides insight into the determinants of the human blood metabolome.

Variation in the blood metabolome is intimately related to human health. However, few details are known about the interplay between genetics and the microbiome in explaining this variation on a metabolite-by-metabolite level. Here, we perform analyses of variance for each of 930 blood metabolites robustly detected across a cohort of 1,569 individuals with paired genomic and microbiome data while controlling for a number of relevant covariates. We find that 595 (64%) of these blood metabolites are significantly associated with either host genetics or the gut microbiome, with 69% of these associations driven solely by the microbiome, 15% driven solely by genetics and 16% under hybrid genome-microbiome control. Additionally, interaction effects, where a metabolite-microbe association is specific to a particular genetic background, are quite common, albeit with modest effect sizes. This knowledge will help to guide targeted interventions designed to alter the composition of the human blood metabolome.

Heterogeneity in statin responses explained by variation in the human gut microbiome.

BACKGROUND: Statins remain one of the most prescribed medications worldwide. While effective in decreasing atherosclerotic cardiovascular disease risk, statin use is associated with adverse effects for a subset of patients, including disrupted metabolic control and increased risk of type 2 diabetes. METHODS: We investigated the potential role of the gut microbiome in modifying patient responses to statin therapy across two independent cohorts (discovery n = 1,848, validation n = 991). Microbiome composition was assessed in these cohorts using stool 16S rRNA amplicon and shotgun metagenomic sequencing, respectively. Microbiome associations with markers of statin on-target and adverse effects were tested via a covariate-adjusted interaction analysis framework, utilizing blood metabolomics, clinical laboratory tests, genomics, and demographics data. FINDINGS: The hydrolyzed substrate for 3-hydroxy-3-methylglutarate-coenzyme-A (HMG-CoA) reductase, HMG, emerged as a promising marker for statin on-target effects in cross-sectional cohorts. Plasma HMG levels reflected both statin therapy intensity and known genetic markers for variable statin responses. Through exploring gut microbiome associations between blood-derived measures of statin effectiveness and adverse metabolic effects of statins, we find that heterogeneity in statin responses was consistently associated with variation in the gut microbiome across two independent cohorts. A Bacteroides-enriched and diversity-depleted gut microbiome was associated with more intense statin responses, both in terms of on-target and adverse effects. CONCLUSIONS: With further study and refinement, gut microbiome monitoring may help inform precision statin treatment. FUNDING: This research was supported by the M.J. Murdock Charitable Trust, WRF, NAM Catalyst Award, and NIH grant U19AG023122 awarded by the NIA.

Risk factors for severe COVID-19 differ by age for hospitalized adults.

Risk stratification for hospitalized adults with COVID-19 is essential to inform decisions about individual patients and allocation of resources. So far, risk models for severe COVID outcomes have included age but have not been optimized to best serve the needs of either older or younger adults. Models also need to be updated to reflect improvements in COVID-19 treatments. This retrospective study analyzed data from 6906 hospitalized adults with COVID-19 from a community health system across five states in the western United States. Risk models were developed to predict mechanical ventilation illness or death across one to 56 days of hospitalization, using clinical data available within the first hour after either admission with COVID-19 or a first positive SARS-CoV-2 test. For the seven-day interval, models for age ≥ 18 and < 50 years reached AUROC 0.81 (95% CI 0.71-0.91) and models for age ≥ 50 years reached AUROC 0.82 (95% CI 0.77-0.86). Models revealed differences in the statistical significance and relative predictive value of risk factors between older and younger patients including age, BMI, vital signs, and laboratory results. In addition, for hospitalized patients, sex and chronic comorbidities had lower predictive value than vital signs and laboratory results.

Manifestations of Alzheimer's disease genetic risk in the blood are evident in a multiomic analysis in healthy adults aged 18 to 90.

Genetics play an important role in late-onset Alzheimer's Disease (AD) etiology and dozens of genetic variants have been implicated in AD risk through large-scale GWAS meta-analyses. However, the precise mechanistic effects of most of these variants have yet to be determined. Deeply phenotyped cohort data can reveal physiological changes associated with genetic risk for AD across an age spectrum that may provide clues to the biology of the disease. We utilized over 2000 high-quality quantitative measurements obtained from blood of 2831 cognitively normal adult clients of a consumer-based scientific wellness company, each with CLIA-certified whole-genome sequencing data. Measurements included: clinical laboratory blood tests, targeted chip-based proteomics, and metabolomics. We performed a phenome-wide association study utilizing this diverse blood marker data and 25 known AD genetic variants and an AD-specific polygenic risk score (PGRS), adjusting for sex, age, vendor (for clinical labs), and the first four genetic principal components; sex-SNP interactions were also assessed. We observed statistically significant SNP-analyte associations for five genetic variants after correction for multiple testing (for SNPs in or near NYAP1, ABCA7, INPP5D, and APOE), with effects detectable from early adulthood. The ABCA7 SNP and the APOE2 and APOE4 encoding alleles were associated with lipid variability, as seen in previous studies; in addition, six novel proteins were associated with the e2 allele. The most statistically significant finding was between the NYAP1 variant and PILRA and PILRB protein levels, supporting previous functional genomic studies in the identification of a putative causal variant within the PILRA gene. We did not observe associations between the PGRS and any analyte. Sex modified the effects of four genetic variants, with multiple interrelated immune-modulating effects associated with the PICALM variant. In post-hoc analysis, sex-stratified GWAS results from an independent AD case-control meta-analysis supported sex-specific disease effects of the PICALM variant, highlighting the importance of sex as a biological variable. Known AD genetic variation influenced lipid metabolism and immune response systems in a population of non-AD individuals, with associations observed from early adulthood onward. Further research is needed to determine whether and how these effects are implicated in early-stage biological pathways to AD. These analyses aim to complement ongoing work on the functional interpretation of AD-associated genetic variants.

From taxonomy to metabolic output: what factors define gut microbiome health?

Many studies link the composition of the human gut microbiome to aberrant health states. However, our understanding of what constitutes a 'healthy' gut ecosystem, and how to effectively monitor and maintain it, are only now emerging. Here, we review current approaches to defining and monitoring gut microbiome health, and outline directions for developing targeted ecological therapeutics. We emphasize the importance of identifying which ecological features of the gut microbiome are most resonant with host molecular phenotypes, and highlight certain gut microbial metabolites as potential biomarkers of gut microbiome health. We further discuss how multi-omic measurements of host phenotypes, dietary information, and gut microbiome profiles can be integrated into increasingly sophisticated host-microbiome mechanistic models that can be leveraged to design personalized interventions. Overall, we summarize current progress on defining microbiome health and highlight a number of paths forward for engineering the ecology of the gut to promote wellness.

Towards early risk biomarkers: serum metabolic signature in childhood predicts cardio-metabolic risk in adulthood.

BACKGROUND: Cardiovascular diseases may originate in childhood. Biomarkers identifying individuals with increased risk for disease are needed to support early detection and to optimise prevention strategies. METHODS: In this prospective study, by applying a machine learning to high throughput NMR-based metabolomics data, we identified circulating childhood metabolic predictors of adult cardiovascular disease risk (MetS score) in a cohort of 396 females, followed from childhood (mean age 11·2 years) to early adulthood (mean age 18·1 years). The results obtained from the discovery cohort were validated in a large longitudinal birth cohort of females and males followed from puberty to adulthood (n = 2664) and in four cross-sectional data sets (n = 6341). FINDINGS: The identified childhood metabolic signature included three circulating biomarkers, glycoprotein acetyls (GlycA), large high-density lipoprotein phospholipids (L-HDL-PL), and the ratio of apolipoprotein B to apolipoprotein A-1 (ApoB/ApoA) that were associated with increased cardio-metabolic risk in early adulthood (AUC = 0·641‒0·802, all p<0·01). These associations were confirmed in all validation cohorts with similar effect estimates both in females (AUC = 0·667‒0·905, all p<0·01) and males (AUC = 0·734‒0·889, all p<0·01) as well as in elderly patients with and without type 2 diabetes (AUC = 0·517‒0·700, all p<0·01). We subsequently applied random intercept cross-lagged panel model analysis, which suggested bidirectional causal relationship between metabolic biomarkers and cardio-metabolic risk score from childhood to early adulthood. INTERPRETATION: These results provide evidence for the utility of a circulating metabolomics panel to identify children and adolescents at risk for future cardiovascular disease, to whom preventive measures and follow-up could be indicated. FUNDING: This study was financially supported by the Academy of Finland, Ministry of Education of Finland and University of Jyv€askyl€a, the National Nature Science Foundation of China (Grant 31571219), the 111 Project (B17029), the Shanghai Jiao Tong University Zhiyuan Foundation (Grant CP2014013), China Postdoc Scholarship Council (201806230001), the Food and Health Bureau of Hong Kong SAR's Health and Medical Research Fund (HMRF grants 15162161 and 07181036) and the CUHK Direct Grants for Research (2016¢033 and 2018¢034), and a postdoctoral fellowship from K. Carole Ellison (to T.W.). The UK Medical Research Council and Wellcome (Grant ref: 217065/Z/19/Z) and the University of Bristol provide core support for ALSPAC. NFBC1966 received financial support from University of Oulu Grant no. 24000692, Oulu University Hospital Grant no. 24301140, ERDF European Regional Development Fund Grant no. 539/2010 A31592. This work was supported by European Union's Horizon 2020 research and innovation programme LongITools 874739.