The Framingham Crap Shoot

The Framingham risk score is a risk-assessment tool that has become the basis for heart disease prediction used by practicing physicians.

The Framingham system determines that:

· 35% of the adult population in the U.S., or 70 million, is deemed “low-risk.” Low-risk is defined as the absence of standard risk factors for heart disease; low-risk persons have no more than a 1-in-20 chance (5%) of dying from heart disease in the next 10 years. Physicians are advised by the American Heart Association (AHA) and its experts that no specific effort at risk reduction is necessary.

· 25%, or approximately 50 million, U.S. adults are deemed “high-risk,” based on the presence of 2 or more risk factors. High-risk persons experience a 20%-30% likelihood of heart attack in the next 10 years. People at high-risk are candidates for preventive efforts according to the guidelines set by the Adult Treatment Panel-III (Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults; ATP-III) for cholesterol-reducing statin drug treatment and for “lifestyle-modifying” advice.

· The remaining 40% of the adult population, or 80 million people, are judged “intermediate-risk,” with the likelihood of heart attack between 5-20% over the next 10 years. This group should receive preventive advice and might be considered for statin drug treatment.

Let’s do some arithmetic. By the above scheme, the low-risk population will experience 3,500,000 heart attacks over the next decade, or 350,000 heart attacks per year.

The intermediate-risk population (without preventive treatment) will experience 8,000,000 heart attacks over the 10-year time period, or 800,000 per year.

The high-risk population, the group most likely to receive standard advice on diet, exercise, and be prescribed statin cholesterol drugs, will have their risk reduced by 35% by preventive efforts over the 10-year period. This means that heart attacks over 10 years will be reduced from 12,500,000 to 8,125,000 by standard prevention efforts, or reduced to 812,500 heart attacks per year.

These numbers are no secret. They are well known facts that have simply come to be accepted by the medical community. In other words, the standard approach to heart attack prediction makes the fact that two million people will succumb to cardiovascular events in the next year no mystery. This exercise in prediction is coldly accurate when applied to a large population.

The problem is that this approach cannot reliably distinguish which individuals will have a heart attack from those who will not.

From 100 people chosen at random, for instance, the numbers game played above will not confidently identify who among those 100 will have a heart attack, who will not, who will develop anginal chest pains and end up with stents or bypass surgery, or who will die. We just know that some of them will. Some people at high risk will have a heart attack, some people at intermediate risk will have a heart attack, some people at low risk will have a heart attack.

For any specific individual (like you or me), it’s a crap shoot.

That's why precise individual measurement of cardiovascular risk is required for real risk assessment, not applying broad statistical observations and forcing them to conform to the unique life of a specific individual, particularly risk calculators with as few risk parameters as the Framingham risk score.

Comments (2) -

  • katherine

    11/24/2008 3:31:00 PM |

    just came across this...thoughts?

  • John

    11/29/2008 3:33:00 AM |

    Good article. One of the most common mistakes made by health conscious individuals is the idea that if study X says A is bad or good then one should adjust one's lifestyle accordingly.

    If only it were that simple. A statistical analysis of a group is applicable to that group, not necessarily the specific individuals within that group; let alone those outside the studied group. We cannot determine individual risk on the basis of such studies yet time and again these studies are mentioned as "evidence" that we must accept. Sadly, too many health promoters, who should know better, tend to make the same logical error.

    Very pleased to see that this error is highlighted here.