SCORE2-Diabetes (10-year CVD risk)

European 10-year risk of fatal and non-fatal cardiovascular disease in people with type 2 diabetes aged 40–69, without prior ASCVD or severe target-organ damage. Includes age at diagnosis, HbA1c, and eGFR. For people without diabetes, use SCORE2 / SCORE2-OP.

Age (years)

Sex

Risk region (ESC)

Match to ESC country grouping (same regions as SCORE2).

Systolic BP (mmHg)

Lipid unitsmmol/Lmg/dL

Total cholesterol (mmol/L)

HDL cholesterol (mmol/L)

Age at T2D diagnosis (years)

HbA1c unitsmmol/mol (IFCC)% (NGSP)

HbA1c (mmol/mol)

eGFR (mL/min/1.73 m², creatinine-based)

Current smoker

Disclaimer: Educational tool only. SCORE2-Diabetes is for a defined population (T2D, 40–69, no ASCVD or severe target-organ damage per guideline intent). It does not replace clinical judgment or national ESC guidance.

SCORE2-Diabetes — how it works

Overview

SCORE2-Diabetes extends the SCORE2 competing-risk framework to people with type 2 diabetes aged 40–69 and without prior atherosclerotic CVD, using pooled European and related cohorts. The same four ESC geographical recalibration scales as SCORE2 (ages under 70) are applied after forming an intermediate 10-year risk.

Primary reference: SCORE2-Diabetes Working Group, Eur Heart J 2023 (doi:10.1093/eurheartj/ehad260). Coefficients in this app match the supplementary specification as implemented in RiskScorescvd (15_SCORE2-Diabetes_func.R).

Predictors

Age (years), sex, current smoking, systolic blood pressure
Total and HDL cholesterol (mmol/L in the published model; mg/dL converted with the usual factor)
Age at type 2 diabetes diagnosis
HbA1c (IFCC mmol/mol in the model; NGSP % converted via the ADA master equation)
eGFR from creatinine (mL/min/1.73 m²), enter on the log scale in the equation
ESC risk region (recalibration)

Computation (conceptual)

Build a sex-specific linear predictor from SCORE2-style terms (age, smoking, SBP, lipids) with diabetes fixed to “present”, plus diabetes duration (via age at diagnosis), HbA1c, log(eGFR) with a quadratic eGFR term, and selected age interactions.
Map to an uncalibrated cumulative incidence using the same baseline constants as SCORE2 for ages under 70.
Apply region-specific calibration: 1 − exp(−exp(scale₁ + scale₂ × ln(−ln(1 − raw)))).
Express as a percentage rounded to one decimal place.
Assign 2023 ESC diabetes guideline bands to that rounded percentage: <5%, 5–<10%, 10–<20%, ≥20%.

Intended population

Ages 40–69 with type 2 diabetes and no known ASCVD or severe target-organ damage, aligned with the original model scope. Ages 70+ use SCORE2-OP for non-diabetic risk estimation; SCORE2-Diabetes itself is not defined for 70+ in the published algorithm implemented here.

What is SCORE2-Diabetes?

SCORE2-Diabetes (Systematic Coronary Risk Evaluation 2-Diabetes) is a risk prediction model developed to estimate 10-year risk of fatal and non-fatal atherosclerotic cardiovascular disease (CVD) in people living with type 2 diabetes mellitus in European clinical and public-health contexts. It builds on the structure of SCORE2—the ESC-aligned tool for apparently healthy individuals without diabetes—but replaces the diabetes indicator with a richer set of diabetes-related information so that risk reflects both traditional vascular risk factors and key diabetes-specific drivers such as glycaemic exposure and kidney function.

The output is typically expressed as a percentage to one decimal place, then mapped to four risk bands used in contemporary ESC guidance for people with diabetes. The model incorporates competing-risk considerations in its development so that the estimated risk targets cardiovascular events rather than all-cause mortality alone.

Why SCORE2-Diabetes exists

Type 2 diabetes confers a substantially higher background risk of myocardial infarction, stroke, and other atherosclerotic events than matched populations without diabetes. Standard SCORE2 equations were developed and validated chiefly for people without diabetes; applying them directly to diabetes populations can misrepresent risk because diabetes changes both the baseline hazard and the relative importance of several risk factors.

SCORE2-Diabetes was derived using large-scale individual participant data from multiple cohorts across several countries, focusing on people with type 2 diabetes and without prior clinical cardiovascular disease at baseline. The result is a single integrated framework that:

Preserves interpretability alongside SCORE2 (same broad conceptual structure and European regional calibration philosophy).
Adds terms that capture duration of exposure to diabetes (via age at diagnosis), glycated haemoglobin (HbA1c), and estimated glomerular filtration rate (eGFR), which jointly modulate vascular risk in diabetes.
Produces risk estimates that can be recalibrated to ESC geographical risk regions, reflecting differences in underlying CVD rates across Europe.

Who the model is intended for

The published SCORE2-Diabetes algorithm implemented in clinical tools is specified for adults aged 40 to 69 years with type 2 diabetes. The development cohorts and reporting emphasise people without known atherosclerotic cardiovascular disease at the time of assessment and without features that define very advanced organ damage in the sense used by prevention guidelines (for example, certain manifestations of severe target-organ damage that place patients in different management pathways).

SCORE2-Diabetes is not a substitute for diagnosing diabetes, staging chronic kidney disease in full detail, or deciding acute treatments. It is a prognostic aid for medium-term CVD event risk in a defined ambulatory population. People without diabetes should use SCORE2 (ages 40–69) or SCORE2-OP (older adults per guideline scope), not SCORE2-Diabetes.

Inputs you need for the calculator

To compute a SCORE2-Diabetes estimate, the following information is required:

Age (years) — must fall within the model’s defined range (40–69).
Sex — male or female, because coefficients are sex-specific.
ESC risk region — low, moderate, high, or very high CVD mortality region, matching the ESC’s country groupings used for SCORE2 recalibration. This choice adjusts the final risk to regional epidemiology.
Smoking status — current smoker versus not currently smoking (former and never smokers are handled together as “not current” in the same way as in reference implementations).
Systolic blood pressure — in mmHg.
Total cholesterol and HDL cholesterol — in mmol/L in the primary specification; calculators may accept mg/dL and convert using the conventional factor.
Age at diagnosis of type 2 diabetes — used as a proxy for exposure duration and lifetime glycaemic/cardiovascular load; it must not exceed current age.
HbA1c — in IFCC units (mmol/mol) in the model; some interfaces allow entry as NGSP percentage with standardised conversion.
eGFR — creatinine-based estimated GFR in mL/min/1.73 m², reflecting the prognostic role of diabetic kidney disease and related pathways.

Together, these variables feed a linear predictor that is transformed into a baseline cumulative incidence and then regionally calibrated so the final percentage reflects European practice.

How the risk score is constructed (conceptual overview)

Although full algebraic detail is lengthy, the computation follows a transparent sequence:

Linear combination of risk factors. Age, smoking, systolic pressure, lipids, and diabetes-related variables are entered using pre-specified centring and scaling (for example, age relative to a reference such as 60 years in five-year units, cholesterol relative to common clinical anchors, and blood pressure relative to typical systolic values). Interaction terms allow the effect of some factors to change with age, which improves calibration across the adult span.
Diabetes-specific extensions. Beyond classical factors, the model adds contributions from age at diabetes diagnosis (interacting with diabetes status), HbA1c, and log-transformed eGFR, including a quadratic component for eGFR so that risk increases more steeply as kidney function worsens. These terms capture pathways such as hyperglycaemia-mediated injury, hypertension–volume–RAAS interplay, and chronic kidney disease as a risk amplifier.
Transformation to an uncalibrated 10-year cardiovascular risk. The linear predictor is mapped through a sex-specific survival-type transformation consistent with the SCORE2 family for ages under 70, yielding an intermediate risk estimate before regional adjustment.
Regional calibration. The intermediate estimate is adjusted using ESC region-specific calibration parameters so that predicted event rates align with observed epidemiology in low, moderate, high, and very high risk settings. This step is essential when comparing patients across countries or when applying thresholds that were validated in European datasets.
Rounding. The final risk is expressed as a percentage rounded to one decimal place, matching published reference calculators and reducing spurious precision.

Interpreting the percentage and risk bands

The numeric output is the estimated probability (in percent) of experiencing fatal or non-fatal atherosclerotic CVD over the subsequent 10 years, under the assumptions of the model and the quality of the entered data.

Contemporary ESC diabetes guidance maps the rounded SCORE2-Diabetes percentage into four communication bands used for prevention discussions:

Rounded 10-year SCORE2-Diabetes risk	Typical label
Below 5%	Low risk
5% to below 10%	Moderate risk
10% to below 20%	High risk
20% or higher	Very high risk

These bands are intended to support shared decision-making about lifestyle, blood pressure targets, lipid-lowering intensity, and glucose-lowering choices with documented cardiovascular benefit in appropriate patients. They are not mandatory treatment rules by themselves: clinicians integrate comorbidity, frailty, bleeding risk, patient preference, and local formulary guidance.

Strengths and limitations to keep in mind

Strengths include alignment with ESC prevention philosophy, explicit regional calibration, incorporation of HbA1c and eGFR (which materially change risk in diabetes), and a transparent connection to the widely taught SCORE2 framework.

Important limitations include:

Scope of age: The primary algorithm is defined for 40–69 years; extrapolation outside that range is not supported by the published specification.
Condition-specific exclusions: The model targets type 2 diabetes without prior ASCVD in the development narrative; using it in people with established cardiovascular disease or in type 1 diabetes without careful thought may be inappropriate.
Input quality: Office versus home blood pressure, non-fasting lipids, and assay differences for HbA1c can shift estimates. eGFR equations vary by creatinine assay and demographics; inconsistent laboratory standards reduce precision.
Ethnicity and migration: European calibration may not transfer perfectly to all ancestries or regions outside the development data.
Dynamic risk: The score is a snapshot; intensified treatment that lowers BP, lipids, or HbA1c should improve prognosis, but the model does not automatically simulate future treatment effects unless used in a decision-analytic way.

Always treat calculator output as educational and adjunctive to clinical judgment, national guidance, and multidisciplinary diabetes and cardiology care.

CalcMD.

Systematic Coronary Risk Evaluation 2-Diabetes (SCORE2-Diabetes)

Learn how SCORE2-Diabetes estimates 10-year fatal and non-fatal CVD risk in adults 40–69 with type 2 diabetes using ESC regions, lipids, BP, smoking, HbA1c, eGFR, and age at diagnosis.