Polygenic Score (PGS) ID: PGS000330

Predicted Trait
Reported Trait Type 2 diabetes (T2D)
Mapped Trait(s) type 2 diabetes mellitus (MONDO_0005148)
Released in PGS Catalog: Sept. 18, 2020
Download Score FTP directory
Terms and Licenses
PGS obtained from the Catalog should be cited appropriately, and used in accordance with any licensing restrictions set by the authors. See EBI Terms of Use (https://www.ebi.ac.uk/about/terms-of-use/) for additional details.

Score Details

Score Construction
PGS Name PRS_T2D
Development Method
Name LDpred
Parameters ρ = 0.3; LD radius = 4000; LD reference panel = 2,690 Finnish individuals [autosomal variants only]
Variants
Original Genome Build hg19
Number of Variants 6,437,380
Effect Weight Type NR
PGS Source
PGS Catalog Publication (PGP) ID PGP000100
Citation (link to publication) Mars N et al. Nat Med (2020)
Ancestry Distribution
Source of Variant
Associations (GWAS)
European: 100%
898,130 individuals (100%)
Score Development/Training
European: 100%
21,813 individuals (100%)
PGS Evaluation
European: 26.7%
African: 20%
East Asian: 20%
South Asian: 20%
Hispanic or Latin American: 6.7%
Additional Asian Ancestries: 6.7%
15 Sample Sets

Development Samples

Source of Variant Associations (GWAS)
Study Identifiers Sample Numbers Sample Ancestry Cohort(s)
GWAS Catalog: GCST009379
Europe PMC: 30297969
898,130 individuals European NR
Score Development/Training
Study Identifiers Sample Numbers Sample Ancestry Cohort(s) Phenotype Definitions & Methods Age of Study Participants Participant Follow-up Time Additional Ancestry Description Additional Sample/Cohort Information
21,813 individuals,
47.3 % Male samples
European
(Finnish)
FINRISK National registries were used to assess disease incidence. Follow-up ended at first-ever diagnosis of the disease of interest, death or at the end of follow-up on December 31, 2018, whichever came first. Disease endpoints are defined in Table S9. Mean (Age At Baseline) = 48.0 years Used to select optimal threshold of ρ for all subsequent analyses. FINRISK surveys from 1992, 1997, 2002 and 2007

Performance Metrics

Disclaimer: The performance metrics are displayed as reported by the source studies. It is important to note that metrics are not necessarily comparable with each other. For example, metrics depend on the sample characteristics (described by the PGS Catalog Sample Set [PSS] ID), phenotyping, and statistical modelling. Please refer to the source publication for additional guidance on performance.

PGS Performance
Metric ID (PPM)
PGS Sample Set ID
(PSS)
Performance Source Trait PGS Effect Sizes
(per SD change)
Classification Metrics Other Metrics Covariates Included in the Model PGS Performance:
Other Relevant Information
PPM000897 PSS000441|
European Ancestry|
21,030 individuals
PGP000100 |
Mars N et al. Nat Med (2020)
Reported Trait: Incident type 2 diabetes C-index: 0.845 age, sex, BMI, history of stroke or CHD, parental history of diabetes, SBP, DBP, HDL, triglycerides, FINRISK cohort, genotyping array/batch, 10 ancestry PCs 10-year risk
PPM000892 PSS000441|
European Ancestry|
21,030 individuals
PGP000100 |
Mars N et al. Nat Med (2020)
Reported Trait: Incident type 2 diabetes HR: 1.7 [1.63, 1.78] C-index: 0.763 age, sex, FINRISK cohort, genotyping array/batch, 10 ancestry PCs 10-year risk
PPM000887 PSS000448|
European Ancestry|
135,300 individuals
PGP000100 |
Mars N et al. Nat Med (2020)
Reported Trait: Type 2 diabetes (incident and prevalent cases) HR: 1.74 [1.72, 1.77] genotyping array/batch, 10 ancestry PCs, stratified by sex
PPM021416 PSS011737|
European Ancestry|
109,021 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.79 [1.75, 1.83] AUROC: 0.726 [0.721, 0.73] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021444 PSS011738|
European Ancestry|
38,941 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes HR: 1.94 [1.8, 2.09] C-index: 0.767 [0.75, 0.783] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021474 PSS011735|
African Ancestry|
44,346 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.27 [1.23, 1.31] AUROC: 0.72 [0.714, 0.727] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021495 PSS011736|
Hispanic or Latin American Ancestry|
33,652 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.58 [1.53, 1.64] AUROC: 0.768 [0.761, 0.775] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021519 PSS011740|
East Asian Ancestry|
1,149 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes OR: 1.29 [1.11, 1.51] AUROC: 0.619 [0.577, 0.661] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021544 PSS011741|
South Asian Ancestry|
852 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes OR: 1.32 [1.12, 1.57] AUROC: 0.66 [0.619, 0.701] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021586 PSS011743|
African Ancestry|
6,871 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.31 [1.21, 1.42] AUROC: 0.716 [0.696, 0.735] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)
PPM021611 PSS011745|
East Asian Ancestry|
1,432 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.43 [1.12, 1.83] AUROC: 0.714 [0.658, 0.77] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)
PPM021626 PSS011749|
South Asian Ancestry|
6,992 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Type 2 diabetes OR: 1.62 [1.51, 1.73] AUROC: 0.724 [0.709, 0.738] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)
PPM021663 PSS011744|
East Asian Ancestry|
1,350 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes HR: 1.58 [1.13, 2.22] C-index: 0.695 [0.611, 0.779] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)
PPM021687 PSS011748|
South Asian Ancestry|
5,685 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes HR: 1.28 [1.17, 1.4] C-index: 0.622 [0.6, 0.645] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)
PPM021563 PSS011739|
Additional Asian Ancestries|
870 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes OR: 1.45 [1.22, 1.72] AUROC: 0.674 [0.632, 0.717] Age at baseline, sex (PGS adjusted for 20 PCs prior to model fitting)
PPM021642 PSS011742|
African Ancestry|
6,019 individuals
PGP000656 |
Ritchie SC et al. medRxiv (2024)
|Ext.|Pre
Reported Trait: Incident type 2 diabetes HR: 1.17 [1.06, 1.29] C-index: 0.646 [0.619, 0.672] Age at baseline, sex assessment centre (PGS adjusted for 20 PCs prior to model fitting)

Evaluated Samples

PGS Sample Set ID
(PSS)
Phenotype Definitions and Methods Participant Follow-up Time Sample Numbers Age of Study Participants Sample Ancestry Additional Ancestry Description Cohort(s) Additional Sample/Cohort Information
PSS011735 T2D cases were ascertained based on a combination of hospital diagnosis codes, prescription medication, and lab results from blood tests occurring prior to baseline assessment. Participants were considered controls if they had no history of any diabetes diagnoses, T2D medication, or abnormal glucose or HbA1c lab results. Participants with T1D or uncertain diabetes status were excluded from analysis
[
  • 5,663 cases
  • , 38,683 controls
]
African unspecified Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations AllofUs
PSS011736 T2D cases were ascertained based on a combination of hospital diagnosis codes, prescription medication, and lab results from blood tests occurring prior to baseline assessment. Participants were considered controls if they had no history of any diabetes diagnoses, T2D medication, or abnormal glucose or HbA1c lab results. Participants with T1D or uncertain diabetes status were excluded from analysis
[
  • 4,033 cases
  • , 29,619 controls
]
Hispanic or Latin American Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations AllofUs
PSS011737 T2D cases were ascertained based on a combination of hospital diagnosis codes, prescription medication, and lab results from blood tests occurring prior to baseline assessment. Participants were considered controls if they had no history of any diabetes diagnoses, T2D medication, or abnormal glucose or HbA1c lab results. Participants with T1D or uncertain diabetes status were excluded from analysis
[
  • 10,069 cases
  • , 98,952 controls
]
European Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations AllofUs
PSS011738 T2D was defined using ICD-10 codes E10-E14, G59.0, G63.2, H28.0, H36.0, M14.2, N08.3, or O24.0-O24.3. Participants with any diabetes history were excluded from analysis. Incident diabetes events were treated as incident T2D
[
  • 726 cases
  • , 38,215 controls
]
European Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations INTERVAL
PSS011739 Incident T2D was ascertained through a combination of linkage to national healthcare records, self-reported medical history at follow-up assessment (either diagnosis from a primary care physician or current diabetes medication usage), or with blood biomarker concentrations indicative of diabetes following the American Diabetes Association criteria (fasting glucose ≥ 7 mmol/L or HbA1c ≥ 6.5% or random blood glucose ≥11 mmol/L)
[
  • 187 cases
  • , 683 controls
]
South East Asian
(Malay Singaporean)
Ancestry label assigned based on Malay being the majority reported ethnicity within the genetic cluster SingaporeMEC
PSS011740 Incident T2D was ascertained through a combination of linkage to national healthcare records, self-reported medical history at follow-up assessment (either diagnosis from a primary care physician or current diabetes medication usage), or with blood biomarker concentrations indicative of diabetes following the American Diabetes Association criteria (fasting glucose ≥ 7 mmol/L or HbA1c ≥ 6.5% or random blood glucose ≥11 mmol/L)
[
  • 205 cases
  • , 944 controls
]
East Asian
(Chinese Singaporean)
Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations SingaporeMEC
PSS011741 Incident T2D was ascertained through a combination of linkage to national healthcare records, self-reported medical history at follow-up assessment (either diagnosis from a primary care physician or current diabetes medication usage), or with blood biomarker concentrations indicative of diabetes following the American Diabetes Association criteria (fasting glucose ≥ 7 mmol/L or HbA1c ≥ 6.5% or random blood glucose ≥11 mmol/L)
[
  • 194 cases
  • , 658 controls
]
South Asian
(Indian Singaporean)
Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations SingaporeMEC
PSS011742 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 395 cases
  • , 5,624 controls
]
African unspecified Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS011743 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 766 cases
  • , 6,105 controls
]
African unspecified Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS011744 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 37 cases
  • , 1,313 controls
]
East Asian Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS000441 National registries were used to assess disease incidence. Follow-up ended at first-ever diagnosis of the disease of interest, death or at the end of follow-up on December 31, 2018, whichever came first. Disease endpoints are defined in Table S9.
[
  • 1,346 cases
  • , 19,684 controls
]
,
47.3 % Male samples
Mean (Age At Baseline) = 48.0 years European
(Finnish)
FINRISK FINRISK surveys from 1992, 1997, 2002 and 2007
PSS011745 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 76 cases
  • , 1,356 controls
]
East Asian Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS011748 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 513 cases
  • , 5,172 controls
]
South Asian Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS011749 Prevalent T2D status at baseline was adjudicated from a combination of retrospective hospital episode records, self-reported history of diabetes, and baseline medication using the Eastwood et al. algorithms. Incident T2D cases were ascertained following the Eastwood et al. algorithms on the basis of ICD-10 diagnosis coding E11 in either the hospital inpatient or death registry data
[
  • 1,253 cases
  • , 5,739 controls
]
South Asian Ancestry label assigned based on genetic similarity with 1000 Genomes reference panel superpopulations UKB
PSS000448 National registries were used to assess disease incidence. Follow-up ended at first-ever diagnosis of the disease of interest, death or at the end of follow-up on December 31, 2018, whichever came first. Disease endpoints are defined in Table S9.
[
  • 17,519 cases
  • , 117,781 controls
]
,
43.7 % Male samples
Mean (Age At Baseline) = 59.2 years
Sd = 16.6 years
European
(Finnish)
FinnGen