Framingham Risk Score
Hướng dẫn chi tiết sắp ra mắt
Chúng tôi đang chuẩn bị hướng dẫn giáo dục toàn diện cho Máy tính Nguy cơ Framingham. Quay lại sớm để xem giải thích từng bước, công thức, ví dụ thực tế và mẹo từ chuyên gia.
The Framingham Risk Score (FRS) is the original and most historically influential cardiovascular risk prediction model. Developed by Wilson PWF and colleagues from data collected in Framingham, Massachusetts, and published in Circulation in 1998, it estimates the 10-year risk of developing a major cardiovascular disease (CVD) event — including angina pectoris, myocardial infarction, coronary insufficiency, or coronary heart disease death — in adults aged 30 to 74 years who are free of pre-existing cardiovascular disease at baseline. The Framingham Heart Study, established in 1948 and now spanning three generations of participants, is one of the longest-running and most cited prospective cohort studies in medical history. It identified most of the classical cardiovascular risk factors now taken for granted: cigarette smoking, elevated blood pressure, high cholesterol, obesity, diabetes, and physical inactivity. The 1998 Framingham Risk Score formalised these observations into a practical clinical scoring tool. Although the Pooled Cohort Equations (ASCVD, 2013) and more recent models have largely replaced the Framingham Score in primary prevention guidelines in the United States, the FRS remains the foundation of cardiovascular risk assessment used in many countries — including Canada (where it is mandated in national guidelines), the United Kingdom (modified as the QRISK framework), and numerous other national health systems worldwide. It is particularly valued for its simplicity, its long track record, and the depth of the epidemiological evidence underpinning it. Understanding the Framingham Score is essential for any clinician working in cardiovascular medicine, as it provides the conceptual basis for all subsequent risk prediction models.
The FRS uses separate point-scoring tables for men and women derived from Cox regression coefficients. Variables are converted to categorical point values based on ranges: Age (30–74), Total Cholesterol (categorised by mg/dL bands), HDL Cholesterol (categorised), Systolic Blood Pressure (with separate columns for treated vs untreated), Current Smoking (yes/no binary), and Diabetes (yes/no binary). Points for each variable are summed. The total point score is then converted to a 10-year CVD risk percentage using published sex-specific lookup tables. For example, in men a score of 6 points = 8% risk; a score of 14 points = 29% risk. The continuous equation form is: Risk = 1 − 0.88936 ^ exp(Coefficient Sum − 23.9388) for men; Risk = 1 − 0.95012 ^ exp(Coefficient Sum − 26.1931) for women, where coefficient sums use ln-transformed continuous variables weighted by published beta coefficients from Wilson et al. 1998.
- 1Step 1 — Establish eligibility: the patient must be aged 30–74 years with no known cardiovascular disease (no prior myocardial infarction, angina, coronary revascularisation, or stroke). The FRS is a primary prevention tool only.
- 2Step 2 — Collect the required variables: age, sex, total cholesterol (mg/dL), HDL cholesterol (mg/dL), systolic blood pressure (mmHg), whether the patient is currently on antihypertensive medication, whether the patient is a current smoker, and whether the patient has diabetes.
- 3Step 3 — Use the sex-specific point table to assign points for age. Points are awarded based on categorical ranges rather than continuous values; for example, in women, age 30–34 = −9 points and age 55–59 = +8 points, rising to +13 points for age 70–74.
- 4Step 4 — Assign points for total cholesterol (higher cholesterol = more points) and HDL cholesterol (higher HDL = fewer or negative points, reflecting its protective effect). Each value is matched to a published age-specific categorical range.
- 5Step 5 — Assign points for systolic blood pressure using the treated or untreated column as appropriate. Treated hypertension at a given SBP level carries more points than untreated, reflecting the higher baseline risk burden of patients requiring medication.
- 6Step 6 — Add binary points for current smoking (+4 for men, +3 for women) and diabetes (+3 for men, +6 for women). Note that diabetes carries a larger risk contribution for women than men in the Framingham equations, consistent with data showing diabetes largely eliminates the female cardiovascular survival advantage.
- 7Step 7 — Sum all category points, convert to 10-year CVD risk percentage using the published lookup table, and classify: Low <10%, Intermediate 10–20%, High >20%. Use the result in shared decision-making, noting that the FRS tends to overestimate risk in some contemporary populations and underestimate risk in South Asian and other high-risk ethnic groups not represented in the original cohort.
Excellent lipid and blood pressure profile; no modifiable risk factors present.
High HDL-C is strongly protective in the Framingham model, and this patient's favourable lipid profile combined with young age and no modifiable risk factors yields a very low 10-year risk. No pharmacological intervention is indicated; focus on sustaining a heart-healthy lifestyle.
Lipid-lowering and BP optimisation are the primary intervention priorities.
Moderately elevated total cholesterol, below-average HDL, and treated but inadequately controlled hypertension combine to place this patient in the intermediate risk band. Intensive lifestyle modification and statin therapy should be discussed. Achieving better BP control would significantly reduce his calculated risk.
Urgent multifactorial risk reduction required: statin, antihypertensive, and smoking cessation.
All six risk factors are adverse in this patient: advanced age, high total cholesterol, very low HDL, severe uncontrolled hypertension, active smoking, and diabetes. His 10-year risk exceeds 30%, meaning more than 1 in 3 similar men will have a major cardiovascular event within a decade without aggressive intervention. High-intensity statin therapy, antihypertensive medication, and smoking cessation support are all indicated simultaneously.
Diabetes carries a larger risk contribution for women than for men in the Framingham model.
Although this woman has no smoking history and a reasonable HDL level, the combination of diabetes (which contributes +6 points for women versus +3 for men), treated hypertension, and age near 60 produces intermediate 10-year risk. Statin therapy is appropriate, and optimal glycaemic control and blood pressure management should be emphasised.
Primary prevention consultations: guiding GP and cardiologist decisions about lipid-lowering and antihypertensive therapy in adults aged 30–74 without established cardiovascular disease.
National health policy: forming the basis of cardiovascular risk assessment in Canadian, UK (as QRISK), and many other international clinical guidelines, enabling standardised population-level screening thresholds.
Occupational health and workplace wellness programs: identifying employees at elevated cardiovascular risk to offer targeted lifestyle interventions, smoking cessation programs, and dietary counselling.
Health insurance risk stratification: actuarial models often draw on Framingham-based risk estimates to stratify policyholders and inform underwriting decisions for life and critical illness insurance products.
Medical education and clinical epidemiology: the Framingham Score is the canonical teaching example for cardiovascular risk prediction, multivariable Cox regression in clinical studies, and the conceptual distinction between absolute and relative risk in preventive medicine.
South Asian Patients — Risk Underestimation
People of South Asian ancestry (Indian subcontinent, Pakistan, Bangladesh, Sri Lanka) have approximately 1.5- to 4-fold higher cardiovascular event rates than White European populations at the same calculated Framingham Risk Score. This population was not represented in the Framingham cohort. Clinicians should consider upward adjustment of the calculated risk, use of South Asian-specific risk multipliers (e.g., multiply FRS by 1.4 as recommended in some Canadian provincial guidelines), or coronary artery calcium scoring to better characterise true risk.
Patients with Chronic Inflammatory Conditions
Conditions such as rheumatoid arthritis, systemic lupus erythematosus, psoriasis, and inflammatory bowel disease are associated with a 1.5- to 2-fold increase in cardiovascular risk beyond what risk scores predict. The Framingham Score does not include inflammatory biomarkers. In patients with these conditions, treat the calculated FRS as a likely underestimate and consider a risk-enhancement factor. Measurement of high-sensitivity CRP may provide additional risk discrimination.
HIV-Positive Patients
People living with HIV have approximately 1.5- to 2-fold higher cardiovascular event rates than HIV-negative individuals with the same Framingham risk profile. This excess risk is attributable to both the chronic inflammatory state of HIV infection and adverse metabolic effects of some antiretroviral medications. The FRS should be interpreted with caution in this group, and some clinicians apply a multiplier of 1.5 to the calculated risk to account for HIV-related cardiovascular risk augmentation.
Premature Menopause
Women who experience menopause before age 40 — whether surgical or natural — lose the cardioprotective effects of oestrogen earlier than average. The Framingham model does not include menopausal status as a variable. Women with premature menopause may have an underestimated Framingham risk, and some guidelines recommend treating their cardiovascular risk profile as equivalent to a woman who is approximately 5 years older than her chronological age.
Extreme Variable Values and Ceiling Effects
The Framingham point table uses categorical ranges for each variable. This means a patient with a systolic BP of 180 mmHg (treated) receives the same points as one with 160 mmHg (treated), potentially underestimating risk at extreme values. Similarly, total cholesterol of 300 mg/dL receives the same points as 280 mg/dL in some age-sex categories. For patients with extreme lipid or blood pressure values, complement the FRS score with direct LDL-C-based treatment thresholds and clinical judgement.
| 10-Year Risk | Category | Clinical Interpretation |
|---|---|---|
| <10% | Low | Major CVD event unlikely within 10 years; lifestyle counselling; statin not routinely recommended |
| 10–20% | Intermediate | Meaningful 10-year risk; statin therapy and lifestyle modification recommended; consider additional risk stratification |
| >20% | High | 1 in 5 or more will have a CVD event within 10 years; aggressive multifactorial risk factor management recommended |
| ≥30% (very high) | Very High | Risk approaching that of established CHD; treat as coronary risk equivalent; high-intensity statin and BP management |
What is the Framingham Risk Score?
The Framingham Risk Score is a validated cardiovascular risk prediction tool published by Wilson et al. in Circulation in 1998. It uses age, sex, total cholesterol, HDL cholesterol, systolic blood pressure, BP treatment status, smoking, and diabetes to estimate the 10-year probability of a major cardiovascular event including heart attack, angina, and coronary death.
How does the Framingham Score differ from the ASCVD Pooled Cohort Equations?
The Framingham Score predicts composite CVD risk including angina, whereas the ACC/AHA Pooled Cohort Equations (ASCVD) predict hard ASCVD events only — heart attack, coronary death, and stroke. The FRS uses a point table system developed in predominantly White cohorts, while the PCE uses separate Cox regression equations for four race-sex groups. The PCE is recommended by current US guidelines; the FRS remains widely used internationally, especially in Canada and parts of Europe.
What risk level triggers statin therapy according to the Framingham model?
Historically, an intermediate or high Framingham Risk Score (10–20% or >20%) was used to identify patients likely to benefit from statin therapy. Most current US guidelines now use the Pooled Cohort Equations with a 7.5% threshold, but many international bodies still use Framingham-derived thresholds. A 10% 10-year risk threshold is commonly used in countries where policy is based on the Framingham model.
Is the Framingham Score accurate for all racial and ethnic groups?
No. The original Framingham cohort was predominantly White, and the FRS is known to underestimate risk in South Asian populations and in some African American populations, while potentially overestimating risk in East Asian populations. For patients of South Asian ancestry, upward risk reclassification or use of coronary artery calcium scoring is recommended to avoid undertreating high-risk individuals.
Why does diabetes carry more points for women than for men in the Framingham Score?
Epidemiological data from the Framingham cohort and other studies show that diabetes attenuates the relative cardiovascular protection that women typically have compared to men at the same age. In effect, diabetes removes much of the female sex advantage, so it is assigned a higher point value (+6 for women vs +3 for men) to reflect this greater incremental risk increase in women.
What is the validated age range for the Framingham Risk Score?
The Framingham Risk Score was validated for adults aged 30 to 74 years. For adults younger than 30, the score is not validated and lifetime risk estimates are more appropriate for motivating lifestyle change. For adults older than 74, absolute risk may be underestimated as the published point tables do not extend beyond age 74.
Does the Framingham Score include stroke risk?
The original 1998 Wilson et al. version predicts coronary heart disease events — including angina, MI, and coronary death — but not stroke. Separate Framingham models exist for stroke prediction. The ACC/AHA Pooled Cohort Equations include both coronary events and stroke in a single calculation, which is one reason they are considered more comprehensive for primary prevention decision-making.
How should the Framingham Score be used alongside other risk assessment tools?
The FRS provides a valuable first estimate of 10-year risk, but clinical decision-making should incorporate additional risk-enhancing factors such as family history of premature CVD, high-sensitivity CRP, coronary artery calcium scoring, and ankle-brachial index. No single risk score captures all dimensions of individual cardiovascular risk, and shared clinician-patient discussion remains essential for translating a probability estimate into a personal treatment decision.
Mẹo Chuyên Nghiệp
The Framingham Risk Score is most powerful when used longitudinally — repeating it annually or every few years allows patients to see the trajectory of their risk and quantifies the benefit of interventions in concrete terms. Showing a patient that quitting smoking reduced their 10-year risk from 18% to 11%, or that losing 8 kg and controlling BP moved them from intermediate to low risk, is far more motivating than abstract lifestyle advice.
Bạn có biết?
The Framingham Heart Study, which gave rise to the Framingham Risk Score, began enrolling participants in 1948 and is now tracking its third generation of families. It coined the term 'risk factor' in the context of cardiovascular disease — a phrase so ubiquitous today that it is easy to forget it was entirely novel in the 1960s. Before Framingham, the prevailing medical view was that heart attacks were an inevitable consequence of ageing rather than something modifiable by lifestyle or medical intervention.
Tài liệu tham khảo
- ›Wilson PWF et al. Prediction of Coronary Heart Disease Using Risk Factor Categories. Circulation 1998.
- ›D'Agostino RB Sr et al. General Cardiovascular Risk Profile for Use in Primary Care. Circulation 2008.
- ›Kannel WB et al. Framingham Heart Study — 50 Years of Lessons. JAMA 2000.
- ›Greenland P et al. 2010 ACCF/AHA Guideline for Assessment of Cardiovascular Risk in Asymptomatic Adults. JACC 2010.
- ›Hippisley-Cox J et al. QRISK3 — a new cardiovascular disease risk calculator. BMJ 2017.