Is an increase in heart scan score GOOD?

In response to an earlier Heart Scan Blog post, I don't care about hard plaque!, reader Dave responded:

Hello Dr Davis,

Interesting post about hard and soft plaque. I recently had a discussion with my GP regarding my serious increase in scan score (Jan 2006 = 235, Nov 2007 = 419).

After the first scan we started aggressively going after my LDL, HDL and Trig...196,59,221

And have them down to 103, 65, 92 - we still have a way to go to 60/60/60 [The Track Your Plaque target values]-

So the increase is a surprise, but my doctor said that the increase could in part be cause some of the soft plaque had been converted to hard plaque and the scan would show that conversion.

Dave's doctor then responded to him with this comment:

"Remember that although your coronary calcium score has gone up, this does not mean that you are at greater risk than you were a year ago. Remember that the most dangerous plaque is the not-yet calcified soft plaque, which will not show up on an EBT [i.e., calcium score]. It is only the safe, calcified plaque that can be measured with the EBT. [Emphasis mine.] For your score to go up like it did, while your lipids came down so much, what had to happen was that lots of dangerous unstable plaque was converted to stable, calcified plaque. There are no accepted guidelines for interpreting changes in calcium scores over time, because the scores tend to go up as treatment converts dangerous plaque to safer plaque. We do know that aggressively lowering LDL reduces both unstable and stable plaque, and we know that risk can be further lowered by adjuvant therapy such as I listed above."


This bit of conventional "wisdom" is something I've heard repeated many times. Is it true?

It is absolutely NOT true. In fact, the opposite is true: Dave's substantial increase in heart scan score from 235 to 419 over 22 months, representing a 78% increase, or an annualized rate of increase of 37%. This suggests a large increase in his risk for heart attack, not a decrease. Big difference!

Dr. Paulo Raggi's 2004 study, Progression of coronary artery calcium and risk of first myocardial infarction in patients receiving cholesterol-lowering therapy in 495 participants addresses this question especially well. Two heart scans were performed three years apart, with a statin drug initiated after the first scan, regardless of score.

During the period of study, heart attacks occurred in 41 participants. When these participants were analyzed, it was found that the average annual increase in score over the three year period was 42%. The average annual rate of increase in those free of heart attack was 17%. The group with the 42% annual rate of increase--all on statin drugs--the risk of heart attack was 17.2-fold greater, or 1720%.

The report made several other important observations:

--20% of the heart attack-free participants showed reduction of heart scan scores, i.e., reversal. None of the participants experiencing heart attack had a score reduction.
--Only 2 of the 41 heart attacks occurred in participants with <15% per year annual growth, while the rest (39) showed larger increases.
--The intensity of LDL reduction made no difference in whether heart attacks occurred or not. Those with LDL<100 mg/dl fared no better than those with LDL>100 mg/dl.

Dr. Raggi et al concluded:

"The risk of hard events [heart attack] was significantly higher in the presence of CVS [calcium volume score] progression despite low LDL serum levels, although the interaction of CVS change and LDL level on treatment was highly significant. The latter observation strongly suggests that a combination of serum markers and vascular markers [emphasis mine] may constitute a better way to gauge therapeutic effectiveness than isolated measurement of lipid levels."

This study demonstrates an important principle: Rising heart scan scores signal potential danger, regardless of LDL cholesterol treatment. Yes, LDL reduction does achieve a modest reduction in heart attack, but it does not eliminate them--not even close.

These are among the reasons that, in the Track Your Plaque program, we aim to correct more than LDL cholesterol. We aim to correct ALL causes of coronary plaque, factors that can be responsible for continuing increase in heart scan score despite favorable LDL cholesterol values.

So, Dave, please forgive your doctor his misunderstanding of the increase in your heart scan score. He is not alone in his ignorance of the data and parroting of the mainstream mis-information popular among the statin-is-the-answer-to-everything set.

Just don't let your doctor's ignorance permit the heart attack that is clearly in the stars. Take preventive action now.

Comments (30) -

  • Anonymous

    11/20/2007 5:41:00 PM |

    Dr Davis,

    What should Dave do?  He appears to have improved his LDL:HDL ratio as well as his total C to HDL ratio substantially, but his CAC score jumped significantly.  Maybe look at other risk factors?

    The info here gives no indication of median blood pressure for Dave.  LP(a)?  No indication of particle sizes. But, which of these or others would be most likely to be Dave's downfall in attempting to mitigate a future hard endpoint?

    I don't ask this lightly, I myself am trying to follow the TYP program and keep my high-for-my-age 29 CAC score from growning.  But, I'm frankly not looking forward to my rescan in about a year.  I'm a bit worried about the, "What if my scan shows a dramatic increase?  What then?"

    Thank you for the valuable information you provide.


  • Dr. Davis

    11/20/2007 11:17:00 PM |

    I would urge Dave to follow all the principles of the Track Your Plaque program, including:

    1) Fish oil to provide minimum 1200 mg EPA + DHA per day

    2) Correction of all concealed lipoprotein patterns such as IDL and Lp(a)

    3) Vitamin D raised to 50 ng/ml--crucial!

    4) Normalization of blood pressure, including during exericse.

    5) Normal blood sugar (<100 mg/dl).

    Further efforts might be required, depending on the long-term effects on rate of plaque growth.

  • Ross

    11/21/2007 3:41:00 AM |

    My question is: how repeatable do you think the scores are on the CT scan?  Are they bulletproof (+/- 5% no matter where measured), consistent by analyst (+/- 5% with the same doctor analyzing the scan), or...?  

    I am currently visiting my brother in law, who is an FP doctor with a private practice.  One of his professional friends, a cardiologist who seems a cut above (thinks stenting is a cop-out), recently told him that he only trusted two centers in the mid-Ohio region to score a 16-slice CT scan accurately, and that even then, the variability was still too high for his taste.  Two numbers within 20% were within his expected error bars and weren't different enough to indicate any change to him.  Two different scan centers?  He wouldn't even compare the two scan scores.

    In my own job (software), I've had to manage human-measured numbers over and over again.  One observation keeps coming up: a single value doesn't mean much without an understanding of the accuracy of that value.  I really am curious about how you estimate confidence intervals on CT scan scores.

  • Dr. Davis

    11/21/2007 3:55:00 AM |

    Hi, Ross--

    Excellent questions.

    Several thoughts:

    1) 16-slice scanners are, unfortunately, prone to wider error in heart scan scoring, perhaps as much as 20%. The variation in scoring on an EBT or 64-slice device is far less.

    2) Variation from scan to scan, when expressed as percent, depends to a great degree on the score itself. Lumping all scores together, variation should be no more than 8-9%. However,a low score of, say 2, then repeated at 4 means 100% variation. However, the same absolute difference of 2 but with a score of 1002 and repeated at 1004 is <1% variation. Therefore, higher scores assume much less percent variation, usually <5%.

    3) Variation among different reading physicians tends to be a minor issue, since much of the scoring is done by standard criteria determined by software, not the human eye. The only real source of human variation comes from disputable areas, such as the mitral valve (which can sometimes encroach into the coronary area and appear like plaque) and the mouth of arteries, which can be debated as being in the aorta or in the coronary arteries themselves. However, these disputable areas are issues in <5% of scans.

  • Tom

    11/21/2007 4:30:00 AM |

    It's interesting that a 29 year old is able to track his plaque. I'm in my 60's now and recently found your site AFTER bypass surgery and a calcium score >700 via a 64 slice scan.
    In reading past comments, those of us having had the heart procedure are now unable to follow our progress via the cac score. Until this post I had hoped to use your recommended blood tests for indication of progress, but if LDL reduction achieves a modest risk reduction, we are left without a specific guide.
    Question: Was the progress in blood tests in dave's case a result of statins ?

  • Dr. Davis

    11/21/2007 12:46:00 PM |

    That's why lipoproteins are so important--they provide other indicators. In my experience, people who have LDL cholesterol as the sole cause of heart disease are a very small minority. The vast majority of people have multiple causes beyond LDL.

    Also, about 50% of people can still get a heart scan score after bypass surgery if you find a center willing to do a detailed analysis. You will need to ask.

    Also, I don't know what Dave did, since he is a reader and everything he posted is above. Are you there, Dave?

  • Dr. Davis

    11/21/2007 5:41:00 PM |

    Hi, Paul--

    I think your doctor might be confusing heart scans with CT coronary angiograms. She is right in saying that CT angiograms (using X-ray dye) require a lot of radiation; 100 chest x-rays worth with present technology.

    However, a plain heart scan to generate a heart scan score requires 4 chest x-rays worth on an EBT device, 8-10 on an 64-slice multi-detector device.

    See the Track Your Plaque Special Report, Radiation and Heart Scans: The Real Story at

  • Anonymous

    11/21/2007 6:01:00 PM |

    Regarding repeatability, there is a 2005 study by Serukov, Bland, and Kondos that shows that the repeatability is a function of the square root of the calcium score, and that volume score is more repeatable than Agatston score. The reference is

    “Serial Electron Beam CT Measurements of Coronary Artery Calcium: Has Your Patient's Calcium Score Actually Changed?” Alexander B. Sevrukov, J. Martin Bland and George T. Kondos, American Journal of Roentgenology 2005; 185:1546-1553

    In this report, the standard deviation of the difference between two sequential calcium scored is

    SDAG130 = 2.515 *sqrt(avg score)
    SDVol130 = 1.758 *sqrt(avg score)

    This results in the following values, where SDA is the standard deviation for the Agatston score and SDV is the standard deviation for the volume score.


    These values show why many people use 15% as a breakpoint - only if the score has changed by more than 15% can it be said that the change is real. And this is only true for scores above 200 or so.


  • Anonymous

    11/21/2007 7:17:00 PM |

    My cardiologist told me that EBT scanning is not recommended for anyone under the age of 30. Is this true? If so, how do I (29 years) reliably know that I am at risk?

    I discovered your blog recently. Since I have a very bad family history of diabetes, high blood pressure, and cholesterol, I decided to visit a cardiologist last month so that I can request for an EBT scan. He said that I'm too young for that, and has instead asked me to take a Carotid IMT and Stress test - are these tests reliable enough to provide insight on my risk? Could these tests return "false positive" values?

    I had found during a blood test I did this July only to find that my triglycerides were at 600!! The other cholesterol values were bad too - totalC-HDL-LDL-Tri (255-31-Not measurable-600)

    Since then I have found your blog, lost around 25 lbs and did a VAP recently (I asked for NMR and all I got from doctors - what? What the heck is that?) So I settled for a VAP, since they knew about it.

    I did a VAP along with a comprehensive blood test and the measures that came up high were.

    LIPID related:
    Total LDL-C Direct:130 (Normal<130)
    Real LDL-C:110 (N<100)
    Sum Total LDL-C: 130 (<130)
    Remnant LIPO (IDL+VLDL3): 30 (<30)
    HDL-2:9 (>10)
    VLDL3: 14 (<10)

    Non-LIPID related high values:
    Uric Acid: 8.3  (4.0-8.0)
    Fasting Glucose: 104 (65-99)
    Creatine Kinase Total: 631 (<=200)

    LP PLA2 is normal: 164 (115-245)
    HBA1C suggests prediabetic: 5.7 (Normal <6%)

    Due to my very high value of CK Total, I researched online and found that this can increase due to high exercise, and I had it repeated after taking rest, and it returned normal results. My doctor was really surprised about this and initially hesitant to fractionise my CK. I feel empowered that I am able to take charge of my health and preventative care with the
    information that is available online (of course, one needs to tread that carefully and make an informed decision due to various conflicting opinions out there).

    Sorry for the long post, Doc. I have a newfound awareness of my health thanks to your blog, and am very much interested in knowing your inputs. I just hope that more physicians in our country follow your noble path and understand the true value and empowerment of preventive care.

    - Philip

  • Dr. Davis

    11/21/2007 8:09:00 PM |

    Hi, Philip--

    In general, 29 is very young, perhaps too young, unless there is an outstanding family history (e.g., father with heart attack at age 37). Although your lipid/lipoproteins are concerning, it would be highly unusual to have anything but a zero heart scan score at your age.

  • Dr. Davis

    11/21/2007 8:14:00 PM |

    Hi, Harry--
    Thanks for the help!

  • Neelesh

    11/22/2007 4:51:00 AM |

    Hi Dr. Davis,
      I've just bought the Track Your Plaque book, waiting for its arrival. I've had a heart attack a year back.I'm 30 years old with no family history, non-alcoholic, non-smoker and vegetarian.
    The event was attributed to ectatic arteries(Type-III) and a very high level of LP(a)- between 120-130. The standard lipid profile was also marginally higher. If I had not insisted for an LP(a) test after reading Dr Agatston's South Beach Heart Program, I would have never found the LP(a) factor.
       I was stented during the hospitalization and now I'm wondering how effective the heart scan will be, given that the accuracy reduces  with stented arteries (


  • Dr. Davis

    11/22/2007 2:35:00 PM |

    Hi, Neeleesh--

    I do advocate heart scanning in people with stents, but I generally suggest that only the unstented arteries be scored. It's imperfect, excluding the most diseased artery, but it's proven a useful compromise, leaving you with two "scorable" arteries.

    The study you cite, however, is not about heart scans, it's about CT coronary angiography, a study that yields "percent blockage" sort of information, not an index of plaque.

    Beyond Lp(a), you should strongly consider vitamin D normalization.  By your first name, I take it you are from India/Pakistan or similar background, an ethnic origin that is associated with severe vitamin D deficiency.

  • Neelesh

    11/22/2007 3:00:00 PM |

    Thanks Dr. Davis. And yes, I'm from India.

  • wccaguy

    11/22/2007 3:13:00 PM |

    Dr. Davis,

    I found your answer to Neeleesh to be interesting in the extreme.  I have a  follow up question to it.

    I don't have specific references for the two facts I have heard but couldn't reconcile:

    1   India has high coronary artery disease incidence.

    2   Your answer to Neeleesh states that vitamin d levels are low in India and Pakistan.  And that would help much to explain the high rate of coronary artery disease in these countries.

    3   And yet India is close to the equator and so vitamin d levels should be relatively high because of sun exposure right?

    The question then is this:  What is the cause of the low vitamin d level in those countries?


  • Dr. Davis

    11/22/2007 4:00:00 PM |

    It is interesting, isn't it?

    I believe part of the explanation is that, the darker your skin complexion, the more you are "protected" from intense and prolonged sun exposure. But, activation of 7-hydrocholesterol to 25-OH-vitamin D3 may require many hours more exposure. Thus, a fair skinned person might activate D within minutes, while a dark skinned individual might require hours.

    Another factor that has not been thoroughly explored but has potential for yielding enormous insights: Vit D receptor genotypes. That is, vitamin D deficiency may express itself in different ways in different populations. Some might get colon cancer, others multiple sclerosis, others coronary disease.

    I believe that the dark-skinned phenomenon becomes especially an issue when migrating to sun-deprived climates such as the northern U.S.

  • wccaguy

    11/22/2007 6:12:00 PM |

    Hi Doc,

    Your explanation makes sense.

    I did a quick google search and found experts on the problem in India attributing it to the increasing extent to which Indians were staying indoors and not "being active."

    But the vitamin D issue throws the whole question of "activity" into question doesn't it?  It might not be the activity per se but instead the amount of sunlight reduction.

    And if, per your explanation, darker skinned people need more time in the sun than lighter skinned people for Vitamin D3 to be "activated" then than a decrease in sunlight would have more effect on darker skinned people than lighter skinned people.

    Very interesting...  And perhaps INCREDIBLY good news!!!

    Because it means that there might be a cheap effective treatment for the coronary disease epidemic in India.

    Does all that make sense?

  • wccaguy

    11/22/2007 6:19:00 PM |

    Just to follow up one more point on this D3 question...

    I guess what we need to do is find a study which shows a correlation between degree of skin pigmentation and Vitamin D3 activation?

    (I'm not sure if the word "degree" is the right word, but perhaps the question is understood anyway?)

    Answering that question would certainly set up the basis for a scientific study right?

  • Dr. Davis

    11/23/2007 12:56:00 AM |

    Yes, it does. It could serve as the basis for a tremendously interesting study.

  • Dr. Davis

    11/23/2007 1:09:00 AM |

    There are indeed a few studies that document this effect, e.g., Factors that influence the cutaneous synthesis and dietary sources of vitamin D (abstract viewable at Arch Biochem Biophys. 2007 Apr 15;460(2):213-7.)

    However, I am not aware of any study that examines the effect of vitamin D supplementation specifically in this population that tracks coronary atherosclerosis. One British study  in Bangladeshi adults did demonstrate dramatic reduction in inflammatory markers with vit D replacement (Circulating MMP9, vitamin D and variation in the TIMP-1 response with VDR genotype: mechanisms for inflammatory damage in chronic disorders? at  ).

  • Dave K

    11/24/2007 12:21:00 AM |

    Hi Dr Davis,

    Sorry - I have been offline for a couple of days.  Interesting discussion.  I will try and add some detail lipid info.

    July 2007 Blood work showed

    My Lp(a) is 7
    IDL = 10
    HDL-2 = 15
    HDL-3 = 50
    VLDL C = 18
    VLDL1+2 = 7

    Currently taking fishoil 1700 mg of DHA+EHA
    Vitamin D 800mg - just incresed to 2000
    Baby Aspirin
    Just started Zetia after getting this last scan result
    Eat basic South Beach phase 3
    BMI - 27
    Glucose is 105
    Exercise 4X week...

    Blood pressure high-normal but I don't know about during exercise.  Cardilogist scheduled me for a stress test after this volume increase.

    I have not has a blood test for Vit D.

    Also - I had an angiograham after the first scan because I was having chests pains .... it turned up that I had no blockages whatsoever.  So we judged the chest pains as non cardiac.

    So I am following your list pretty close.  I guess I just have to wait to see how these changes do.  How long would you wait for another scan?

    Not sure what else to add - your website says to consider L-arginie...

    I do have a specific question.  In the scan report it shows where the calcium was found.  Don't know the software, but there was one spot where it showed in the early report that it didn't show in this report (of course there was several new areas) - could that have actually been a reversal at that spot?

  • Dr. Davis

    11/24/2007 1:25:00 AM |

    Small LDL and a deficiency of large HDL, along with modest excess weight, high blood sugar, high blood pressure all suggest you are (or were) likely over-dependent on processed carbohydrates like wheat products. Your pattern would likely respond vigorously to reduction or elimination of these foods and weight loss. Niacin can help this pattern. In our experience, normalization of vitamin D is crucial.

  • Dave K

    11/26/2007 5:51:00 AM |

    Dr Davis,

    Few more data ....

    Some of the treatments have only been for the last 6 months or so.  The Statin was first (of course) and it took almost a year to get something I could tolerate.  The we talked about Vit D (700) and fish oil (800 Omega 3).  After a full Lipid scan around 9 months ago - we decided to add more fish oil.  So the full dosage I listed is only 6 months old or so.

    Also - I love my red wine and I know the number says two glasses and i rarely do two - so its three or four ... which might be my next step....

    From your last response, I assume the VLDL and IDL levels are the ones you would target hardest at this point.

    Don't do a lot of sugar or wheat... Do eat Oatmeal everyday with rasins or blueberries.

    Oh and my other question was with this kind of increase how long would you wait for the next scan?

  • Dr. Davis

    11/26/2007 12:08:00 PM |


    I generally recommend waiting a year after all identifiable causes have been corrected. However, given your minimal doses of vit D, I usually have my patients wait at least six month after vitamin D blood levels are corrected.

  • Dave

    11/26/2007 8:01:00 PM |

    Dr Davis,

    Thank you ... keep up the great work and I'll keep reading... and tracking.


  • G

    11/27/2007 12:39:00 AM |

    Neeleesh and DR. D,

    This Canadian physician appears to have a lot of indepth awareness of the diff phenotypes. He suggests (in the author's response) that D2 may not work as well in East Indians (may worsen glycemic control) versus D3 (the more biologically active vitamin D). Very fascinating!!
    Repletion of vitamin D with vitamin D2 is common
    practice, and vitamin D2 can be used safely when monitored
    to achieve normal levels of 25(OH)D. This might
    take 2 to 3 months, as discussed in your letter and in my
    paper, because the half-life is about 2 weeks. Using vitamin
    D3 (1000 to 5000 IU) daily, depending on the level
    of deficiency, will also achieve this goal. I also agree
    that the goal is to achieve levels of 25(OH)D higher than
    100 nmol/L, preferably 100 to 125 nmol/L.
    My concern regarding vitamin D2 is that it is a synthetic
    analogue and might interact with the vitamin D
    receptor differently in various cell systems. It has been
    reported that vitamin D3 might improve glycemic control.
    7 Vitamin D2 has been reported to cause worsening
    of glycemic control in people of East Indian descent.8
    Is this because of vitamin D receptor polymorphism, or
    because of enhanced 24-hydroxylase enzyme activation,
    or is it due to how vitamin D2 interacts with the receptor?
    Until this has been sorted out, I feel safest using
    vitamin D3. There are about 2000 synthetic analogues
    of vitamin D. The search is on for one that can cross the
    blood-brain barrier to treat certain types of brain cancers
    without causing hypercalcemia.9 But then again,
    what other effects would this compound have? There
    are still so many unknowns.
    The first step is to recognize that most Canadians
    do not get enough vitamin D, especially in the winter
    months, because of where we live. This recognition
    might reduce the need for expensive drugs to treat
    various conditions and might improve the well-being of
    many Canadians.
    An ounce of prevention is worth a pound of cure.
    —Gerry Schwalfenberg MD CCFP
    Edmonton, Alta
    by e-mail

    here's the orig article which is one of the most excellent summaries I've seen so far -- great minds think alike -- they advise > 50ng/ml like DR. Davis as well!

  • Neelesh

    11/27/2007 4:05:00 AM |

    Interesting study indeed. Thanks for the information. I guess I have a lot of things to discuss with my cardiologist next week. Smile

  • chickadeenorth

    12/2/2007 11:16:00 PM |

    Hi to Gerry Schwalfenberg MD CCFP, do you know any Dr In Edtmn who practices Track your Plague, if so could you suggest names to help me. I live out by Jasper and need a skilled Dr in this treatment program, I would travel to Edtmn.Many thanks.
    (hope its ok for me to ask this here)

  • cadoce66

    4/5/2008 8:37:00 PM |

    hi my aunts 63 yrs and she underwent an angioplasty with a medicated stent .. Shes on PLAVIX and her artery was 90% blocked and she had an evolving AWMI...
    Please advise what she should taketo prevent another blockage or heart attack!

  • buy jeans

    11/3/2010 10:34:10 PM |

    So, Dave, please forgive your doctor his misunderstanding of the increase in your heart scan score. He is not alone in his ignorance of the data and parroting of the mainstream mis-information popular among the statin-is-the-answer-to-everything set.