Low expectations

The Framingham Risk Calculator is a standard method used by many physicians to predict risk for heart attack or death from heart disease over a 10-year period. Low-risk is defined as <10% risk of heart attack or cardiac death over 10 years; high-risk is defined as 20% or more over 10 years; intermediate-risk is in between.

Let's put it to the test:

Amy is a 53-year old businesswoman. She is 5 ft 4 inches, weighs 150 lbs. Her father had a heart attack in his early 50s followed by the usual list of hospital procedures including bypass surgery at age 60.

What is Amy's risk for heart attack or death from heart disease over the next 10 years? If we enter her data into the Framingham risk calculator, the following result is returned:

Information about your risk score:
Age: 53
Gender: female
Total Cholesterol: 198 mg/dL
HDL Cholesterol: 74 mg/dL
Smoker: No
Systolic Blood Pressure: 120 mm/Hg
On medication for HBP: No
Risk Score: 1% Means 1 of 100 people with this level of risk will have a heart attack in the next 10 years.

So, according to the Framingham calculation, Amy has <1% risk for heart attack or death from heart disease over the next 10 years. Most primary care physicians would, at most, prescribe a statin drug and talk about a reduction in saturated fat.

Thankfully, Amy didn't fall for that bit of conventional mis-information. She instead got a CT heart scan, principally because of her father's history. Her score: 117. At age 53, this put her into 90th percentile, in the worst 10% of scores for women in her age group (50-55). By heart scan criteria, her risk for heart attack is probably more like 4-5% per year, or approximately 40-50% over the next 10 years.

Let's do just a bit more math. If Amy hadn't known about her heart scan score and no preventive action was taken, the expected progression of her heart scan scores would likely be:

Start: 117
Year 1: 152
Year 2: 198
Year 3: 257
Year 4: 335
Year 5: 436
Year 6: 567
Year 7: 737
Year 8: 958
Year 9: 1245
Year 10: 1618

In fact, given Amy's starting heart scan score of 117, it is highly unlikely that she survives the next 10 years without heart attack or a fatal heart event. Yet the Framingham risk calculator puts Amy's risk at less than 1%. Could anything be more wrong?

The folly of the Framingham calculator was highlighted by a recent publication from the large Multi-Ethnic Study of Atherosclerosis (MESA), in which 3600 women (45-84 years), all of whom fell into the "low-risk" category by the Framingham calculator--just like Amy--were tracked over approximately 3 3/4 years. This study generated several observations:

1) 30% of the "low-risk" women had positive heart scan scores.
2) 5% of the "low-risk" women had scores of 300 or greater (very significant for a woman). 8.6% of these women experienced a cardiovascular event like heart attack or death over the period. Women with a heart scan score of 300 or greater had a 22-fold greater event risk compared to women with zero heart scan scores.
3) Women with heart scan scores of 1 to 299 had a cardiovascular event risk of approximately 5-fold greater risk over the period.

Across the U.S., 90% of women younger than 70 years old fall into the Framingham "low-risk" category. Yet this fiction is accepted as the prevailing standard, along with LDL and total cholesterol, for determination of risk in women and men.

In my view, using the Framingham risk calculator is a misguided, misleading path, one that will mis-classify a substantial number of women who could otherwise be spared from heart attack and catastrophe.

By the way, Amy is also the Track Your Plaque program record holder (by percentage drop), with a 63% drop in heart scan score over a 15 month period.

Comments (6) -

  • lungdoc

    12/14/2007 4:21:00 AM |

    I just recently signed up to TYP. My CT score today was 794!
    I am 50 year old diabetic. My LDLs had been <70 and HDLs>50 for years. Recently found vit D was only 25 and now 66 after supplementation. HGB AIC had been 6-7 for years. I have now started concentrating on bringing it down. My most recent HS CRP was 1.3. My small ldl-p was 542 and homocysteine is 15.4
    What's my next step?

  • Dr. Davis

    12/14/2007 12:22:00 PM |

    The next step is to enter your data into the Track Your Plaque program generator to generate basic comments.

    Also, I invite you to participate in the Track Your Plaque Forum for more detailed conversations.

    I assume fish oil is part of your program, an absolute requirement along with with your vitamin D.

    You may benefit spefically from efforts to reduce small LDL.

  • Anonymous

    12/14/2007 5:29:00 PM |

    And perhaps we should do a bit more math in order to prevent terrorizing the readership (something most would not even think about because they're too frightened by the hype). If you do the simple math, you will see that only 15 women out of 3600 had a cardiac event over 3+ years of follow up. That means 4 out of 1000 low Framingham Risk women will have an event. That sounds like a pretty low risk to me.

    3600 total low risk

    5% with calcium score of 300 or more=180 women

    8.6% of the 180 women had an event in 3.75 yrs.=15 women in almost 4 years

    15 women out of the entire 3600 low Framingham risk group=0.004 or 4 per 1000 women in 4 years or 1/1000 per year.

    So with low Framingham risk, the calcium  score increases your relative risk but the ABSOLUTE risk remains very low.

  • Dr. Davis

    12/15/2007 12:11:00 AM |

    Yes, you are right: the absolute risk was low in the four year period of observation.

    Actually, there were 18 coronary events (defined as "CHD" in the study) and 24 events if strokes were included ("CVD" events).

  • MAC

    12/15/2007 3:51:00 PM |

    Dr. Davis,
        Do you Have any opinion regarding myeloperoxidase (MPO)?


    "One fascinating aspect of this study is that this marker of inflammation precedes by nearly a decade the development of clinical coronary disease,"

  • Dr. Davis

    12/15/2007 5:05:00 PM |

    Hi, MAC--

    I think myeloperoxidase holds promise as a marker for risk. What we do about it remains an open question.

    We are planning a Track Your Plaque Special Report on unique markers of risk for 2008 that will include a discussion about myeloperoxidase.