10-Year CVD Risk Calculator

What the percentage is counting

Most risk tools on this site answer a narrow question: the chance of a blocked coronary artery. This page answers a wider one. The Framingham general cardiovascular profile, published by D'Agostino and colleagues in Circulation in 2008, folds five diagnoses into one number: coronary heart disease (heart attack, coronary death, angina), stroke and transient ischemic attack, heart failure, and peripheral artery disease, the leg cramping known as intermittent claudication. Ten years of follow-up, one percentage. The leg half of that count is often chased on exam with an ankle-brachial index, a cuff comparison between ankle and arm.

Breadth changes who looks risky. A 70-year-old with a normal lipid panel and a stiff, weakening heart scores higher here than on an artery-only calculator, because heart failure sits inside the endpoint. A 48-year-old with an LDL over 160 and clean arteries everywhere else scores lower than the headline figure on a cholesterol page, because the four inputs cannot see anything else.

That is also the honest limit. The CDC describes cardiovascular disease as a family of conditions, and this equation deliberately keeps the family together. If your main worry is a blocked artery specifically, the artery-only score on our ASCVD page is the better match. If you want one number that covers the whole circulatory system, this is it.

Four levers, measured on one 58-year-old woman

Take a 58-year-old woman with total cholesterol 225 mg/dL, HDL 55 mg/dL, systolic pressure 128 mm Hg untreated, no smoking, no diabetes. The equation puts her ten-year risk at 7.2%, inside the moderate band. Now push each of the four inputs, one at a time, and watch the output. The four levers are not the same size.

Diabetes and smoking dominate. A 40 mg/dL rise in total cholesterol, the kind that happens after two bad winters of eating, adds 1.5 points. Developing diabetes adds 6.6 points at this age, more than four times as much. The order matters when you decide where a change of habit will pay off most.

The arithmetic behind 7.15 percent

The Framingham Heart Study publishes every coefficient, so the whole calculation fits on one screen. For women the ten-year risk is 1 minus 0.95012 raised to the power of e raised to (sum of terms minus 26.1931). The terms are natural logs, and the log of blood pressure depends on whether you take medicine for it. Here is the woman from the last section, term by term.

The running total of 26.5653 minus the baseline constant of 26.1931 gives 0.3722. e to that power is 1.451, and 0.95012 to the power of 1.451 is 0.9285. One minus 0.9285 is the raw risk of 7.15%, which rounds to the 7.2% shown above. Every step is ordinary arithmetic: logs, one exponent, one subtraction.

Treatment status is not decoration. The same woman at the same 128 mm Hg while on pressure medicine scores 9.5% instead of 7.2%, because the equation treats a treated reading as evidence of higher underlying pressure. The number on the cuff is not the whole input.

Why diabetes and smoking outrank a cholesterol number

The reason sits in the log form of the equation. Each coefficient multiplies the natural log of the input, so big jumps in cholesterol buy less than you would expect. Moving total cholesterol from 225 to 265 multiplies the raw value by 1.18, and 1.20904 times the log of 1.18 is only 0.198 of log-hazard. The smoking flag, worth 0.52873 by itself, needs a cholesterol rise to 348 mg/dL to match it. Diabetes needs about 399 mg/dL, a value no lab reports as normal.

Two practical notes follow. First, lipid changes matter less at moderate levels than advertising suggests; a 30 mg/dL drop buys about 1.1 points for this woman, not the 5 points a supplement ad might imply. Second, HDL outruns total cholesterol per point moved: sliding HDL from 55 to 40 costs 1.7 points, more than the 1.5 points that a 40 mg/dL cholesterol rise costs. A panel where both numbers are bad is worse than the total alone, which is why the calculator takes HDL separately.

Age quietly beats all four levers. The same woman at 42 with identical numbers scores 3.4%; at 74 she scores 12.3% without a single input changing. Levers move the number you can move today, but age sets its ceiling.

Which lever to pull first, by risk band

This page splits results into four bands, and each band points at a different first move. The bands are ours, not a guideline's: the Framingham general score has no official treatment thresholds the way the pooled cohort equations do. Read them as a triage order, not a prescription.

The ordering also shifts with age. A 42-year-old with total cholesterol 195, HDL 62 and systolic 112 sits near 1.9%, and there starting to smoke costs 1.3 points while diabetes costs 1.8, small enough that one habit change can move the whole band. At 74, the same blood pressure rise buys 5.5 points and diabetes buys 10.7, which turns the conversation toward combinations rather than single moves.

A second profile: 64, treated, still smoking

A 64-year-old woman with total cholesterol 235 mg/dL, HDL 48 mg/dL, systolic 142 mm Hg on treatment, and current smoking scores 28.2%, the very high band. More than one in four people with her numbers would have an event inside ten years. Here is what each single change buys, and what all three together buy.

Notice what the last row does. No single change drops her out of the very high band. All four together land her at 8.9%, one band above moderate. That is the standard argument for stacking changes instead of hunting for one perfect fix, and it is also why the treatment conversation at 28% usually starts with a prescription rather than with willpower.

Framingham 2008 against PREVENT 2023

This page runs the Framingham general cardiovascular equation, the last broad-CVD score most clinicians know by name. The American Heart Association has since published PREVENT, a newer set of equations from 2023, and the two differ in ways that move the output.

What that means in practice: the same person can score 7.2% here and near 4% on PREVENT, and both can be right about the ranking. Absolute numbers drift with the cohort an equation was trained on. The four levers, though, hold their order across both models: diabetes and smoking sit on top, blood pressure next, lipids last. Use the absolute value as a rough band, not a contract.

What the score cannot see

The equation reads eight inputs and nothing else. Family history of early heart disease, high lipoprotein(a), a high coronary calcium score, chronic kidney disease, autoimmune conditions such as rheumatoid arthritis or lupus, and pregnancy complications such as preeclampsia all raise real risk, and none of them is an input. Anyone carrying one of those should read their result as a floor, not a ceiling, and bring the rest of the story to a care team.

Two structural limits matter as well. First, the score is for people free of cardiovascular disease at baseline; after a first heart attack or stroke the whole framing shifts to secondary prevention, where different and larger numbers apply. Second, the output describes a group of people who share your inputs. A score is a statement about a crowd, not about your particular arteries.

This page is general health information. It is not medical advice, it does not diagnose anything, and it does not tell anyone to start or stop a medicine. Decisions about treatment belong with a care team that can see the whole history.