SEARCH · PubMed Health

Results for “Polygenic Risk Score”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 73 records · Page 4Linked to original sources

PubMed2026

Machine Learning-Driven Prediction of Coronary Artery Disease Risk Based on UK Biobank Plasma Proteomics.

BACKGROUND: Coronary artery disease (CAD) is a leading global cause of mortality, yet the predictive accuracy of conventional risk models is limited. Here, we integrate conventional risk factors, polygenic risk scores, and large-scale proteomics to develop a unified model for enhanced CAD risk prediction. METHODS: Using data from UK Biobank, participants with plasma proteomics and genetic risk data were included after excluding prevalent CAD. Participants from England were split into training (n=32 330) and internal validation (n=13 857) sets, and Scotland/Wales participants formed an external validation set (n=5775). Incident CAD was ascertained from linked health records. A 202-protein proteomic risk score was derived by least absolute shrinkage and selection operator Cox regression, and CatBoost models were trained using conventional risk factors alone and with incremental addition of polygenic risk scores and protein proteomic risk scores; Shapley Additive Explanations-guided forward selection identified a compact protein panel. RESULTS: Across cohorts, the median age was 58 years and ∼45% were men. Protein proteomic risk score was dose-dependently associated with CAD risk. Compared with conventional risk factors alone, integrating polygenic risk scores and protein proteomic risk scores improved discrimination, with the area under the curve increasing from 0.750 (95% CI, 0.732-0.767) to 0.789 (95% CI, 0.772-0.805) in internal validation and from 0.717 (95% CI, 0.683-0.750) to 0.762 (95% CI, 0.732-0.791) in external validation. A 9-protein panel (GDF15 [growth differentiation factor 15], MMP12 [matrix metalloproteinase 12], NPPB [natriuretic peptide B], PGF [placental growth factor], REN [renin], ADGRG2 [adhesion G-protein coupled receptor], ACE2 [angiotensin-converting enzyme 2], CDCP1 [CUB domain-containing protein 1], CXCL17 [C-X-C motif chemokine ligand 17)]) captured most proteomic predictive information. CONCLUSIONS: Our findings demonstrate that integrating conventional risk factors, polygenic risk scores, and proteomic data improves CAD risk prediction. This study highlights the utility of proteomics in precision cardiovascular medicine and simplified risk stratification tools.

Humans↗

PubMed2025

Cumulative Genetic Risk for Asthma Contributes to Disease Severity in Children with Asthma living in Urban Environments.

BACKGROUND: Childhood-onset asthma is highly heritable, with nearly 200 risk loci identified in genome-wide association studies. Aggregated polygenic risk scores can be used to quantify genetic predisposition to asthma, but their power to predict asthma severity in multi-ancestral groups has not been determined. OBJECTIVE: Our aim was to examine the predictive power of biobank-derived asthma polygenic risk scores in children with asthma living in urban environments. METHODS: We generated polygenic risk scores for asthma, derived from a large-scale genome-wide association meta-analysis, in four multi-ancestry asthma study cohorts of children living in urban environments. We assessed genetic predictions across different subphenotypes of asthma and tested for associations between genetic asthma risk and measures of asthma severity. RESULTS: Genetic asthma prediction was significantly stronger for more symptomatic asthma phenotypes (P<0.001). Polygenic risk scores were significantly higher in difficult-to-control vs. easy-to-control asthma (P=0.02). Genetic risk was also significantly associated with more frequent exacerbations (P=0.03), higher blood eosinophil levels (P=0.01), and lower lung function (P<0.001). CONCLUSION: Cumulative genetic risk for asthma is associated with disease severity and exacerbation risk in children with asthma living in urban environments.

Journal Article↗

PubMed2026

Baseline Computed Tomography Coronary Angiography and Polygenic Risk Profiles in Adults With Type 2 Diabetes: A Cross-Sectional Analysis From the VOLTAIRE Study.

AIMS: To characterise baseline clinical, anatomical, and genetic cardiovascular risk profiles in participants enrolled in the VOLTAIRE (Evaluation of Polygenic Scores and CT Imaging in Risk Factor Modification in Patients with Type 2 Diabetes) study and examine concordance across these domains. METHODS: This analysis included adults with T2D who completed baseline computed tomography coronary angiography (CTCA) and polygenic risk score (PRS) assessment prior to randomisation in the VOLTAIRE study. Coronary atherosclerosis was evaluated using coronary artery calcium (CAC) score and CTCA-derived stenosis severity. Clinical risk was assessed using the New Zealand Society for the Study of Diabetes 5-year cardiovascular risk calculator. Polygenic risk for coronary artery disease was assessed using a genome-wide PRS and categorised into tertiles. RESULTS: Among 126 participants with T2D (mean age 57.5 ± 8.7 years; 62.7% male), coronary atherosclerotic burden was highly heterogeneous: 34.9% had CAC = 0, whereas 19.8% had CAC ≥ 400. Moderate-to-severe coronary stenosis (≥ 50%) was present in 40.5% of participants overall, including 20.4% of those classified as low clinical risk. PRS distribution was variable (low 37.3%, intermediate 35.7%, high 27.0%). Overlap between anatomical, genetic, and clinical domains was limited, with only 8.7% of participants classified as high risk across all three. CONCLUSIONS: Substantial heterogeneity and limited overlap exist between anatomical, genetic, and clinical cardiovascular risk measures in T2D. These findings support a multimodal approach to risk assessment integrating imaging and genetic profiling. TRIAL REGISTRATION: https://www. CLINICALTRIALS: gov; ID: NCT07091162.

Aged↗

PubMed2026

Genetic Risk Factors for Kidney Function in Individuals with Type 1 Diabetes.

KEY POINTS: Previous research has identified polygenic risk scores that are associated with low eGFR and albuminuria in the general population. We observed that these eGFR and albuminuria polygenic risk scores were associated with eGFR and albuminuria, respectively, in type 1 diabetes. Associations were independent of glycemic control and suggest shared genetic kidney risk factors between type 1 diabetes and the general population. BACKGROUND: Genetic risk factors underlying kidney disease in type 1 diabetes (T1D) remain poorly understood. We examined whether previously established polygenic risk scores (PRS) for eGFR and albuminuria are associated with these measures in adults with T1D in the Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications study. METHODS: We applied eGFR and albuminuria PRS derived in general population cohorts to 1304 DCCT/Epidemiology of Diabetes Interventions and Complications participants with genome-wide genotyping. We tested PRS associations with eGFR and urine albumin excretion rate (AER) as well as incident eGFR <60 ml/min per 1.73 m 2 , AER ≥30 mg/24 h, and AER ≥300 mg/24 h. For consistency, PRS values were linearly transformed so higher scores corresponded to higher eGFR and AER. We also examined associations of kidney outcomes with rs55703767 in COL4A3 , which has previously been associated with CKD in T1D. RESULTS: At DCCT baseline, participants had a mean age of 27 years; 53% were male. 49% of participants were randomized to intensive versus conventional glucose-lowering therapy. Participants were followed for median of (first-third quartiles) 35 (33-37) years. The eGFR PRS was significantly associated with continuous eGFR (per one SD higher PRS 2.72 ml/min per 1.73 m 2 higher [95% confidence interval (CI), 2.05 to 3.40]) and incident eGFR <60 ml/min per 1.73 m 2 (hazard ratio [HR]=0.82 [95% CI, 0.73 to 0.92]), but not consistently with albuminuria. There was no association with quantitative AER (2.42 mg/24 h [95% CI, -1.86 to 6.89]) or sustained AER ≥30 mg/24 h (HR=1.03; [95% CI, 0.94 to 1.14]). The albuminuria PRS was significantly associated with incident AER ≥30 mg/24 h (HR=1.12 [95% CI, 1.02 to 1.22]) but not continuous eGFR (0.49 ml/min per 1.73 m 2 higher [95% CI, -0.23 to 1.21]) or incident eGFR <60 ml/min per 1.73 m 2 (HR=0.96 [95% CI, 0.85 to 1.08]). Associations were similar in analyses stratified by DCCT treatment group assignment. rs55703767 was associated with lower incident macroalbuminuria in the overall cohort (HR=0.77 per minor allele [95% CI, 0.59 to 0.99]), and upon stratification by DCCT treatment group assignment, only within the conventional and not intensive glucose-lowering therapy group. CONCLUSIONS: PRS associated with eGFR and albuminuria in the general population were associated with corresponding measures in adults with T1D. The results suggest shared genetic risk factors for kidney disease between T1D and the general population but different genetic risk factors for albuminuria and eGFR in T1D. CLINICAL TRIALS REGISTRATION NUMBERS: NCT00360893 , NCT00360815 .

Adult↗

PubMed2026

Large-Scale Plasma Proteomics Enhances Prediction of Liver-Related Events Among Individuals With Prediabetes and Type 2 Diabetes: A Prospective Cohort Study in the UK Biobank.

OBJECTIVE: To develop a protein risk score (ProRS) for predicting liver-related events (LREs) in patients with diabetes and compare its predictive performance with the Fibrosis-4 Index (FIB-4) and an established polygenic risk score. RESEARCH DESIGN AND METHODS: This prospective cohort study included 13 516 individuals with prediabetes and type 2 diabetes (T2D) from the UK Biobank. Cox proportional hazards models and LASSO regression were applied to identify proteins associated with incident LREs and construct the ProRS. Predictive performance was assessed using Harrell's C-index, time-dependent area under the receiver operating characteristic curve, net reclassification improvement and integrated discrimination improvement. RESULTS: Over a median follow-up of 13.5 years, 171 (1.3%) incident LREs occurred. We identified 877 proteins associated with LRE risk, primarily enriched in inflammatory signalling, extracellular matrix remodelling and complement/coagulation cascades. In the training set, we developed a 24-protein ProRS (C-index, 0.842; 95% CI 0.797-0.884) that stratified individuals into low-, medium- and high-risk groups, with 10-year cumulative incidences of LREs of 0.2%, 1.2% and 14.2%, respectively. Compared with the low-risk group, the hazard ratio for LREs was 57.1 (95% CI 31.9-102) in the high-risk group. In the internal validation set, the ProRS model (C-index, 0.876; 95% CI 0.827-0.920) accurately predicted both short- and long-term LREs and outperformed FIB-4 index (C-index, 0.733; 95% CI 0.657-0.807) and polygenic risk score (C-index, 0.636; 95% CI 0.564-0.706). CONCLUSIONS: The protein risk score demonstrated superior performance compared with the FIB-4 index and the polygenic risk score in predicting incident LREs among individuals with prediabetes and T2D. The score allows stratification of individuals according to liver-related risk, though external validation in multi-ethnic cohorts is warranted.

Humans↗

PubMed2025

Polygenic Susceptibility in Peripartum, Alcohol-Induced, and Cancer Therapy-Related Cardiomyopathies.

IMPORTANCE: Rare monogenic variants linked to nonischemic dilated cardiomyopathy (DCM) are enriched among individuals with secondary cardiomyopathies, such as peripartum (PPCM), alcohol-induced (ACM), and cancer therapy-related (CCM) cardiomyopathies. However, it remains unclear whether a polygenic predisposition to DCM also contributes to these conditions. OBJECTIVE: To assess the association of a DCM polygenic score with PPCM, ACM, and CCM, and to evaluate the contributions of monogenic and polygenic susceptibilities to these secondary cardiomyopathies. DESIGN, SETTING, AND PARTICIPANTS: This was a retrospective genetic association analysis of data from the Mass General Brigham (MGB) Biobank (n = 42 137, 2008-2025), with replication in the UK Biobank (n = 295 160, 2005-2010), FinnGen (n = 417 950, 2017-2025), and the Veterans Affairs Million Veteran Program (n = 516 066, 2011-2025). In MGB Biobank, medical records were reviewed to ascertain secondary cardiomyopathy cases and antecedent clinical risk factors. EXPOSURES: DCM polygenic risk score and DCM monogenic variants. MAIN OUTCOMES AND MEASURES: The primary outcomes were the association of the DCM polygenic risk score with PPCM, ACM, and CCM and the prevalence of monogenic variants and a high polygenic score among individuals with cardiomyopathy. RESULTS: The mean (SD) age in the MGB Biobank was 55.7 (17.0) years at enrollment, and 24 551 (58.3%) were female. Across the 4 study cohorts, 3414 individuals with secondary cardiomyopathy were identified, including 70 with PPCM, 2281 with ACM, and 1063 with CCM. The DCM polygenic score was associated with PPCM (odds ratio [OR], 1.82 per SD; 95% CI, 1.43-2.30), ACM (OR, 1.56; 95% CI,1.34-1.82), and CCM (OR, 1.64; 95% CI,1.24-2.15) (all with P < .001). Monogenic variants were enriched but present in 7 of 113 individuals with medical record-reviewed cardiomyopathy in MGB, while 66 had a high polygenic score, which conferred an approximately 3-fold increased odds of cardiomyopathy. Most individuals with cardiomyopathy lacked antecedent clinical risk factors. CONCLUSIONS AND RELEVANCE: In this cohort study, individuals with PPCM, ACM, and CCM were enriched for monogenic DCM variants and a high DCM polygenic score, suggesting a shared genetic susceptibility influenced by distinct environmental precipitants. These findings support a shared genetic architecture between secondary cardiomyopathies and DCM, although additional work with larger numbers of individuals with cardiomyopathy is needed to confirm these findings.

Humans↗

PubMed2025

The Role of Genomic-Informed Risk Assessments in Predicting Dementia Outcomes.

INTRODUCTION: By integrating genetic and clinical risk factors into genomic-informed dementia risk reports, healthcare providers can offer patients detailed risk profiles to facilitate understanding of individual risk and support the implementation of personalized strategies for promoting brain health. METHODS: We constructed an additive score comprising the modified Cardiovascular Risk Factors, Aging, and Incidence of Dementia Risk Score (mCAIDE), family history of dementia, APOE genotype, and an AD polygenic risk score in NACC and ADNI, and assessed its association with progression to all-cause dementia. RESULTS: 81% of participants had at least one high-risk indicator for dementia, with each additional risk indicator linked to a 34% increase in the hazard of dementia onset. DISCUSSION: We found that most participants in memory and aging clinics had at least one high-risk indicator for dementia. Furthermore, we observed a dose-response relationship where a greater number of risk indicators was associated with an increased risk of incident dementia.

dementia risk scores↗

PubMed2025

Post-traumatic stress disorder and REM-sleep behavior disorder: exploring genetic associations and causal links.

OBJECTIVE: To explore potential genetic and/or causal associations between Post-Traumatic Stress Disorder and neurodegeneration-related isolated/idiopathic rapid-eye-movement sleep behavior disorder. METHODS: We conducted polygenic risk score, genetic correlation, and Mendelian randomization analyses using the latest genome-wide association studies summary statistics and individual genotyping data. Next, a blinded observer examined dopamine transporter imaging binding status-a marker of neurodegeneration-in patients with isolated/idiopathic rapid-eye movement sleep behavior disorder, with (N = 6) and without Post-Traumatic Stress Disorder (N = 32). RESULTS: Polygenic risk scores for Post-Traumatic Stress Disorder were associated with isolated/idiopathic rapid-eye-movement sleep behavior disorder, with each standard deviation increase linked to 14.7% higher odds (odds ratio = 1.15, 95% confidence interval: 1.04 to 1.26, p = 0.005). However, genetic correlation was weak, and Mendelian randomization did not support a potential causal relationship. The proportion of individuals with abnormal dopamine transporter imaging binding status was significantly higher in the Post-Traumatic Stress Disorder group compared to those without the disorder (p=0.01, X2 = 6.62). INTERPRETATION: Polygenic risk scores analysis identified an association between Post-Traumatic Stress Disorder and neurodegeneration-related isolated/idiopathic rapid-eye-movement sleep behavior disorder, consistent with the result from the small exploratory substudy. The lack of strong genetic correlation or causation may reflect limited sample size. Further research with larger and more diverse cohorts is crucial to clarify the genetic, biological and physiological mechanisms underlying this association.

Journal Article↗

PubMed2026

Phenotypic and genotypic risk factors for uveal melanoma in a high ambient UV radiation environment.

BACKGROUND: Most evidence regarding risk factors for uveal melanoma (UM) derives from case-control studies prone to recall bias, and its rarity has limited prospective research. To address these gaps, we applied a population-based case-control design incorporating polygenic risk scores and Mendelian randomization to explore genetic and phenotypic determinants of UM risk. METHODS: The study was conducted in Queensland, Australia. Incident UM cases diagnosed between 2011 and 2022 were recruited through specialist ocular oncology clinics. Controls were participants in the QSkin Sun and Health Study, a prospective cohort of 43,794 adults aged 40-69 at baseline (2010-2011). Demographic, phenotypic, and sun exposure factors were harmonized across studies. Polygenic risk scores were calculated for pigmentation traits and nevi characteristics using genome-wide association study datasets, and Mendelian randomization was used to assess potential causal relationships with UM risk. RESULTS: Among 485 UM cases and 43,724 controls, several phenotypic traits were strongly associated with UM risk: male sex, blue or light eye color, inability to tan, freckling, and high nevus density. A family history of cutaneous melanoma and a personal history of keratinocyte cancer were also associated with higher risk. Polygenic risk score analyses confirmed significant associations for eye color and freckling. Mendelian randomization analyses supported causal relationships between lighter eye color, freckling propensity, reduced tanning ability and UM risk. CONCLUSIONS: These findings provide genetic evidence supporting a causal role for pigmentation-related traits in UM susceptibility. IMPACT: This evidence may help refine risk assessment and inform counselling for individuals with suspicious ocular lesions.

Journal Article↗

PubMed2025

Blood Pressure Genetics in Han Taiwanese With Cross-Trait Analysis in East Asians: Insights Into Comorbidities, All-Cause Mortality, and Cardiovascular Mortality.

BACKGROUND: Hypertension is a major health burden in East Asia. However, the genetic architecture and clinical implications of blood pressure (BP) traits remain underexplored beyond European-focused studies. This large-scale study aimed to investigate hypertension, systolic BP, and diastolic BP, to uncover genetic links to comorbidities and mortality in Han Taiwanese individuals. METHODS: This large-scale study used China Medical University Hospital biobank data and conducted genome-wide association studies on 25 523 hypertension cases and 47 522 controls, plus 66 236 individuals for systolic BP and 66 152 for diastolic BP. Cross-trait genetic correlations were assessed across 5 East Asian biobanks. Mendelian randomization and polygenic risk scores were applied to assess causality and predict clinical outcomes. RESULTS: We identified 8 loci and 36 genes for hypertension, 7 loci and 17 genes for systolic BP, and 9 loci and 26 genes for diastolic BP. ATP2B1 and FGF5 were common to all BP traits, implicating calcium signaling and vascular remodeling pathways. Cross-trait analyses showed shared genetic liability between BP traits and cardiovascular and metabolic comorbidities. Phenome-wide association studies confirmed strong associations with circulatory diseases. Mendelian randomization analyses demonstrated that elevated BP causally increases the risk of unstable angina pectoris. Polygenic risk scores predicted significantly higher risks and earlier onset of unstable angina pectoris, all-cause mortality, and cardiovascular mortality among individuals in the top polygenic risk score quintiles. CONCLUSIONS: Our findings highlight the genetic basis of BP and comorbidities in East Asians, suggesting that BP genetic risk may inform future approaches to early risk assessment and prevention.

Aged↗

PubMed2026

Arrhythmia and cardiomyopathy risk in Taiwan with complementary biobank evidence on thyroid genetic susceptibility: an integrative population-based framework.

BACKGROUND: Arrhythmia-induced cardiomyopathy (AiCM) is a potentially reversible cause of ventricular dysfunction; however, only a subset of patients with arrhythmia develop cardiomyopathy. Emerging evidence suggests that endocrine factors, particularly thyroid dysfunction with genetic susceptibility, may contribute to inter-individual variability in arrhythmia-related myocardial outcomes. METHODS: We performed a dual-cohort population-based study using the National Health Insurance Research Database (NHIRD, 2000-2015) and the Taiwan Biobank (TWB). In NHIRD, we examined the association between newly diagnosed arrhythmia and incident cardiomyopathy using Cox proportional hazards models. In TWB, genome-wide data, thyroid-stimulating hormone (TSH), polygenic risk scores (PRSs), lifestyle factors, and metabolic comorbidities were analyzed using multivariable regression and interaction models to assess determinants of thyroid dysfunction. RESULTS: In the NHIRD cohort, arrhythmia was associated with a significantly increased risk of incident cardiomyopathy (adjusted hazard ratio (aHR): 2.49, 95% CI: 1.94-2.96), with atrial fibrillation showing the strongest association among arrhythmia subtypes. In the TWB cohort, a higher thyroid polygenic risk score was strongly associated with thyroid dysfunction (adjusted odds ratio (aOR): 6.64, 95% CI: 5.86-7.52). The association between genetic susceptibility and thyroid dysfunction was further modified by metabolic and lifestyle factors, including diabetes, hyperlipidemia, and dietary patterns. Genome-wide analysis identified multiple loci associated with thyroid-stimulating hormone regulation, consistent with a polygenic architecture of thyroid endocrine traits. CONCLUSION: Arrhythmia was associated with an increased risk of cardiomyopathy in a nationwide cohort, while thyroid genetic susceptibility was strongly associated with thyroid dysfunction in a biobank cohort and modified by metabolic and lifestyle factors. These findings provide complementary population-level evidence of parallel cardiovascular and endocrine-genetic associations. Because the two cohorts were not individually linked, causal inference cannot be established. The results support a systems-level framework of endocrine-cardiac interaction and suggest that integrated clinical and genetic risk assessment may help identify individuals who warrant closer monitoring.

arrhythmia↗

PubMed2025

Polygenic Prediction of Peripheral Artery Disease and Major Adverse Limb Events.

IMPORTANCE: Peripheral artery disease (PAD) is a heritable atherosclerotic condition associated with functional decline and high risk for limb loss. With growing knowledge of the genetic basis for PAD and related risk factors, there is potential opportunity to identify individuals at high risk using polygenic risk scores (PRSs). OBJECTIVE: To develop a novel integrated, multiancestry polygenic score for PAD (PRS-PAD) and evaluate its risk estimation for PAD and major adverse limb events in 3 populations. DESIGN, SETTING, AND PARTICIPANTS: This longitudinal cohort study was conducted among individuals with genotyping and electronic health record data in the UK Biobank (2006-2021), All of Us (AoU, 2018-2022), and the Mass General Brigham Biobank (MGBB, 2010-2023). Data were analyzed from July 2023 to February 2025. EXPOSURES: PRS-PAD, previously published PAD polygenic scores, and clinical risk factors. MAIN OUTCOMES AND MEASURES: The primary outcomes were PAD and major adverse limb events, defined as a surrogate of major amputation and acute limb ischemia. RESULTS: The study populations included 400 533 individuals from the UK Biobank (median [IQR] age, 58.2 [45.0-71.4] years; 216 215 female participants [53.9%]), 218 500 from AoU (median [IQR] age, 53.6 [37.7-65.0] years; 132 647 female participants [60.7%]), and 32 982 from MGBB (median [IQR] age, 56.0 [32.0-80.0] years; 18 277 female participants [55.4%]). In the UK Biobank validation cohort, PRS-PAD was associated with an odds ratio [OR] per SD increase of 1.63 (95% CI, 1.60-1.68; P < .001). After adjusting for clinical risk factors, the OR for the top 20% of PRS-PAD was 1.68 (95% CI, 1.62-1.74; P < .001) compared to the remainder of the population. Among PAD cases without a history of diabetes, smoking, or chronic kidney disease (n = 3645), 1097 individuals (30.1%) had a high PRS-PAD (top 20%). In incident disease analysis, PRS-PAD improved discrimination (C statistic, 0.761), which was nearly equivalent to the performances of diabetes (C statistic, 0.760) and smoking (C statistic, 0.765). Among individuals with prevalent PAD, high PRS-PAD was associated with an increased risk of incident major adverse limb events in the UK Biobank (hazard ratio [HR], 1.75; 95% CI, 1.18-2.57; P = .005), MGBB (HR, 1.56; 95% CI, 1.06-2.30; P = .02), and AoU (HR, 1.57; 95% CI, 1.06-2.33; P = .03). CONCLUSIONS AND RELEVANCE: This cohort study develops a new PRS that stratifies risk of PAD and adverse limb outcomes. Incorporating polygenic risk into PAD care warrants further investigation to guide screening and tailor management to prevent major adverse limb events.

Humans↗

PubMed2024

Prediction of metabolic syndrome using machine learning approaches based on genetic and nutritional factors: a 14-year prospective-based cohort study.

INTRODUCTION: Metabolic syndrome is a chronic disease associated with multiple comorbidities. Over the last few years, machine learning techniques have been used to predict metabolic syndrome. However, studies incorporating demographic, clinical, laboratory, dietary, and genetic factors to predict the incidence of metabolic syndrome in Koreans are limited. In the present study, we propose a genome-wide polygenic risk score for the prediction of metabolic syndrome, along with other factors, to improve the prediction accuracy of metabolic syndrome. METHODS: We developed 7 machine learning-based models and used Cox multivariable regression, deep neural network (DNN), support vector machine (SVM), stochastic gradient descent (SGD), random forest (RAF), Naïve Bayes (NBA) classifier, and AdaBoost (ADB) to predict the incidence of metabolic syndrome at year 14 using the dataset from the Korean Genome and Epidemiology Study (KoGES) Ansan and Ansung. RESULTS: Of the 5440 patients, 2,120 were considered to have new-onset metabolic syndrome. The AUC values of model, which included sex, age, alcohol intake, energy intake, marital status, education status, income status, smoking status, dried laver intake, and genome-wide polygenic risk score (gPRS) Z-score based on 344,447 SNPs (p-value < 1.0), were the highest for RAF (0.994 [95% CI 0.985, 1.000]) and ADB (0.994 [95% CI 0.986, 1.000]). CONCLUSIONS: Incorporating both gPRS and demographic, clinical, laboratory, and seaweed data led to enhanced metabolic syndrome risk prediction by capturing the distinct etiologies of metabolic syndrome development. The RAF- and ADB-based models predicted metabolic syndrome more accurately than the NBA-based model for the Korean population.

Humans↗

PubMed2024

Distinguishing different psychiatric disorders using DDx-PRS.

Despite great progress on methods for case-control polygenic prediction (e.g. schizophrenia vs. control), there remains an unmet need for a method that genetically distinguishes clinically related disorders (e.g. schizophrenia (SCZ) vs. bipolar disorder (BIP) vs. depression (MDD) vs. control); such a method could have important clinical value, especially at disorder onset when differential diagnosis can be challenging. Here, we introduce a method, Differential Diagnosis-Polygenic Risk Score (DDx-PRS), that jointly estimates posterior probabilities of each possible diagnostic category (e.g. SCZ=50%, BIP=25%, MDD=15%, control=10%) by modeling variance/covariance structure across disorders, leveraging case-control polygenic risk scores (PRS) for each disorder (computed using existing methods) and prior clinical probabilities for each diagnostic category. DDx-PRS uses only summary-level training data and does not use tuning data, facilitating implementation in clinical settings. In simulations, DDx-PRS was well-calibrated (whereas a simpler approach that analyzes each disorder marginally was poorly calibrated), and effective in distinguishing each diagnostic category vs. the rest. We then applied DDx-PRS to Psychiatric Genomics Consortium SCZ/BIP/MDD/control data, including summary-level training data from 3 case-control GWAS ( N =41,917-173,140 cases; total N =1,048,683) and held-out test data from different cohorts with equal numbers of each diagnostic category (total N =11,460). DDx-PRS was well-calibrated and well-powered relative to these training sample sizes, attaining AUCs of 0.66 for SCZ vs. rest, 0.64 for BIP vs. rest, 0.59 for MDD vs. rest, and 0.68 for control vs. rest. DDx-PRS produced comparable results to methods that leverage tuning data, confirming that DDx-PRS is an effective method. True diagnosis probabilities in top deciles of predicted diagnosis probabilities were considerably larger than prior baseline probabilities, particularly in projections to larger training sample sizes, implying considerable potential for clinical utility under certain circumstances. In conclusion, DDx-PRS is an effective method for distinguishing clinically related disorders.

Journal Article↗

PubMed2026

Distinguishing different psychiatric disorders using DDx-PRS.

Despite great progress on case-control polygenic prediction, an unmet need remains for a method that genetically distinguishes clinically related disorders (e.g., schizophrenia (SCZ) versus bipolar disorder (BIP) versus major depressive disorder (MDD) versus controls). We introduce differential diagnosis-polygenic risk score (DDx-PRS), which jointly estimates the posterior probabilities of each diagnostic category (e.g., SCZ = 50%, BIP = 25%, MDD = 15%, control = 10%) by modeling variance-covariance structure across disorders, leveraging case-control polygenic risk scores and prior clinical probabilities for each diagnostic category. We applied DDx-PRS to Psychiatric Genomics Consortium SCZ, BIP, MDD and control data, including summary-level training data from three case-control genome-wide association studies (n = 41,917-173,140 cases; total n = 1,048,683) and held-out test data from different cohorts with equal numbers for each diagnostic category (total n = 11,460). DDx-PRS was well calibrated and well powered (consistent with simulations) and produced comparable results to methods that require tuning data. True diagnosis probabilities in the top deciles of predicted diagnosis probabilities were considerably larger than prior baseline probabilities, implying appreciable potential for clinical utility in certain settings.

Humans↗

PubMed2026

Meta-ERS: an exposome-based risk score using non-genetic factors to guide osteoporosis prevention.

BACKGROUND: Osteoporosis is influenced by both genetic and environmental factors, yet the relative contribution of the exposome remains unclear. This study aimed to systematically identify non-genetic exposures related to osteoporosis and develop an exposome risk score (ERS) to evaluate individual osteoporosis susceptibility. METHODS: We conducted an exposome-wide analysis of 477,792 UK Biobank participants to identify key exposures associated with osteoporosis. The selected exposures were combined into a weighted Meta-ERS and validated in the Scotland/Wales cohort. The Meta-ERS was further compared with polygenic risk scores (PRS) and linked to plasma proteomics to explore underlying biological pathways. RESULTS: We identified 41 independent non-genetic exposures spanning socioeconomic status, mental health, sleep, diet, smoking, physical activity, environment, and marital status, with socioeconomic status and mental health emerging as the most significant drivers. Based on the identified exposures, we constructed eight domain-specific exposure risk scores and integrated them into a weighted Meta-ERS. The Meta-ERS (R2 = 5.1%; Proportion of Chi-Square = 14.3%) demonstrated an ability to explain osteoporosis variation that was on par with polygenic risk scores (R2 = 4.8%; Proportion of Chi-Square = 12.0%). Importantly, modifying unfavorable exposures mitigated the negative effect of PRS on osteoporosis, particularly among high PRS individuals (1.5- to 1.8-fold greater absolute risk reduction than in those with low PRS). Proteomic analyses further revealed potential mechanisms through which the exposome influences osteoporosis, including hormonal regulation, inflammation, ossification, muscle development, lipid metabolism, and accelerated bone aging. Among these, growth/differentiation factor 15 was identified as a key mediator protein, with a mediation proportion of 13.13%-36.52%. CONCLUSIONS: The Meta-ERS facilitates the quantification of individual osteoporosis risk and identifies modifiable exposures for targeted prevention. Its application can enable personalized risk stratification and guide lifestyle or environmental interventions.

Aged↗

PubMed2025

Polygenic Contributions to Lithium Augmentation Outcomes in Unipolar Depression.

IMPORTANCE: Lithium augmentation is an effective treatment for patients with major depression after inadequate antidepressant response, but therapeutic outcomes vary considerably between individuals. Molecular studies may provide novel insights into treatment prediction and guide personalized therapy. OBJECTIVE: To investigate the association of polygenic risk scores (PRS) for schizophrenia (SCZ), major depressive disorder (MDD), and bipolar disorder (BIP) with clinical outcomes after lithium augmentation. DESIGN, SETTING, AND PARTICIPANTS: This cohort study analyzed prospectively assessed treatment outcomes in patients who underwent lithium augmentation. Disorder-specific PRS were calculated using well-powered genome-wide association study summary statistics. Participants were recruited from 13 psychiatric hospitals, primarily in the greater Berlin area, between 2008 and 2020. They were patients with MDD who showed inadequate response to at least 1 antidepressant, a baseline score of 12 or more on the 17-item Hamilton Depression Rating Scale (HAMD-17), adequate treatment duration (≥4 weeks), and no diagnostic or co-medication changes. Data analysis was conducted between June 2022 and November 2023. EXPOSURE: Polygenic risk scores for MDD, SCZ, or BIP. MAIN OUTCOMES AND MEASURES: Response was defined as a 50% or greater reduction in HAMD-17 score, remission as a HAMD-17 score of 7 or less. Cox proportional hazards models, adjusted for ancestry, demographic, and clinical covariates, were used to estimate hazard ratios (HRs) for favorable outcomes. RESULTS: Among 193 patients (mean [SD] age, 49.5 [13.4] years; 118 [61.1%] female and 75 [38.9%] male), higher BIP-PRS were associated with both response (HR, 1.29; 95% CI, 1.02-1.63; P = .03) and remission (HR, 1.52; 95% CI, 1.14-2.04; P = .004), explaining 2.51% and 4.53% of the variability in treatment outcomes, respectively. Individuals in the highest tertile of the BIP-PRS distribution had a 2.02-fold (95% CI, 1.15-3.53) higher likelihood of response and a 2.26-fold (95% CI, 1.17-4.36) higher chance of remission compared with those in the lowest tertile. Additionally, lower MDD-PRS was associated with better response to lithium augmentation (HR, 0.81; 95% CI, 0.66-1.00; P = .048; Nagelkerke R2 = 1.99%). No significant associations were observed between SCZ-PRS and response (HR, 1.00; 95% CI, 0.80-1.24; P = .97) or remission (HR, 1.12; 95% CI, 0.85-1.48; P = .42). CONCLUSIONS AND RELEVANCE: Individuals carrying a higher polygenic burden for BIP and lower polygenic risk for MDD are more likely to benefit from lithium augmentation. Our findings suggest that disease-related PRS may aid in developing treatment prediction models for lithium augmentation response in depression, potentially informing clinical decision-making.

Humans↗

PubMed2025

Patient stratification by genetic risk in Alzheimer's disease is only effective in the presence of phenotypic heterogeneity.

Case-only designs in longitudinal cohorts are a valuable resource for identifying disease-relevant genes, pathways, and novel targets influencing disease progression. This is particularly relevant in Alzheimer's disease (AD), where longitudinal cohorts measure disease "progression," defined by rate of cognitive decline. Few of the identified drug targets for AD have been clinically tractable, and phenotypic heterogeneity is an obstacle to both clinical research and basic science. In four cohorts (n = 7241), we performed genome-wide association studies (GWAS) and Mendelian randomization (MR) to discover novel targets associated with progression and assess causal relationships. We tested opportunities for patient stratification by deriving polygenic risk scores (PRS) for AD risk and severity and tested the value of these scores in predicting progression. Genome-wide association studies identified no loci associated with progression at genome-wide significance (α = 5×10-8); MR analyses provided no significant evidence of an association between cognitive decline in AD patients and protein levels in brain, cerebrospinal fluid (CSF), and plasma. Polygenic risk scores for AD risk did not reliably stratify fast from slow progressors; however, a deeper investigation found that APOE ε4 status predicts amyloid-β and tau positive versus negative patients (odds ratio for an additional APOE ε4 allele = 5.78 [95% confidence interval: 3.76-8.89], P<0.001) when restricting to a subset of patients with available CSF biomarker data. These results provided no evidence for large-effect, common-variant loci involved in the rate of memory decline, suggesting that patient stratification based on common genetic risk factors for progression may have limited utility. Where clinically relevant biomarkers suggest diagnostic heterogeneity, there is evidence that a priori identified genetic risk factors may have value in patient stratification. Mendelian randomization was less tractable due to the lack of large-effect loci, and future analyses with increased samples sizes are needed to replicate and validate our results.

Alzheimer Disease↗