MESA Study: Track Your Plaque-Lite?

The long-awaited data analyses from the Multi-Ethnic Study of Atherosclerosis (MESA) are finally making it to press.

The MESA Study is an enormously ambitious and important study of 6800 people, 45 to 84 years old, that includes white, black, Hispanic, and Chinese participants from six communities around the U.S. (Forsyth County, NC; Northern Manhattan and the Bronx, NY; Baltimore and Baltimore County, Md; St Paul, Minn; Chicago, Ill; and Los Angeles County, California.) Participants had no history of heart disease at enrollment. All underwent a heart scan (either EBT or multi-detector heart scans) at the start. It is therefore the largest prospective study involving heart scans ever performed. It is, not unexpectedly, yielding some fascinating observations relevant to the Track Your Plaque program. The MESA study is, incidentally, funded by the non-commercial, publicly-funded National Heart, Lung, and Blood Institute and is therefore presumably free of commercial bias.

Among the most recent publications is Risk factors for the progression of coronary artery calcification in asymptomatic subjects: Results from the Multi-Ethnic Study of Atherosclerosis (MESA) In this analysis of 5700 of the MESA participants, a repeat heart scan was obtained an average of 2.4 years after the first. Conventional risk factors for heart disease were obtained at the start (see below for details under Measurement of Covariates.)

After analyzing the data and risk factors assessed, such as age, sex, race, blood pressure, body mass index (BMI), presence of diabetes, blood sugar, and family history of heart disease, two questions were asked:

1) What risk factors predict heart scan scores?

2) What risk factors predict progression (i.e., increase) in heart scan scores?

(The second question is particularly relevant to us and the Track Your Plaque experience.)

The MESA analysis showed that essentially all the risk factors assessed correlated with both the initial heart scan score, as well as the rate of progression. No surprises here.

But the most eye-opening finding was that the conventional risk factors assessed explained only 12% of the variation and progression in heart scan scores (coefficient of determination, or R squared, = 0.12.) In other words:

--Conventional risk factors like LDL cholesterol, diabetes, and excess weight explain only a tiny fraction of why someone develops coronary atherosclerotic plaque as represented by a heart scan score.

--The great majority of risk for a high heart scan score remains unexplained by conventional risk factors.

--The great majority of risk for progressive increase in heart scan scores also remains unexplained by conventional risk factors.

In light of the MESA analysis, it's no surprise that strategies like reducing LDL cholesterol with statin drugs fails to prevent most heart attacks. It's no surprise that conventional prevention programs that talk about "knowing your numbers," eating a "balanced" or low-fat diet, etc., fail miserably to prevent the vast majority of heart attacks and heart procedures.

MESA confirms what we've been saying these past few years: If you want control over coronary heart disease, you won't find it in Lipitor, a low-fat diet, and other limited conventional notions of risk. Correction of conventional risk factors like cholesterol and blood pressure are, in a word, a failure. I wouldn't even call the conventional approach Track Your Plaque-Lite. They don't even come close.

If conventional risk factors can explain only 12% of the reason behind heart disease, we've got to look elsewhere to understand why you and I develop this process.

Measurement of Covariates
Information on demographics, smoking, medical conditions, and family history was collected by questionnaire at the initial examination. Height and weight were also measured at the baseline examination, and blood was drawn for measurements, including lipids, inflammation, fasting glucose, fibrinogen, and creatinine. Resting blood pressure was measured 3 times in the seated position, and the average of the last 2 measurements was used in the analysis. Medication use was determined by questionnaire. Additionally, the participant was asked to bring to the clinic containers for all medications used during the 2 weeks before the visit. The interviewer then recorded the name of each medication, the prescribed dose, and frequency of administration from the containers.

Copyright 2008 William Davis,MD

Comments (11) -

  • Mike

    1/4/2008 3:52:00 PM |

    The part of the study that caught my attention was "Current and former smokers had higher incident CAC rates than never smokers, but this difference was not statistically significant once other risk factors were considered. "

  • Barry

    1/4/2008 5:14:00 PM |

    Dr Davis,

    I recently came across your blog and it peaked my interest. I have been tracking my lipids for several years now.

    Let me give you a little background. My dad had by-pass surgery when he was in his forties. He spent the rest of his life watching his diet and lipids and blood pressure. In 2005 he died from coronary failure a month after turing 71.

    Because of my family history, and lipid levels (LDL 130, HDL <40), my physician wanted to start me on advicor. I resisted and tried to moderate my lipids through diet and exercise. I used the low fat diet approach and got nowhere. So I started on Advicor late in 2003. By March of 2004 my profile was right where my doctor wanted it: LDL 76, HDL 41, TriG 98. I continued on Advicor and had my lipid profile and liver enzymes checked every 6 months. I started keeping my scores in a spreadsheet and have been tracking them ever since.

    After taking Advicor for a while I started reading about the side effects, about the additional predictors for heart disease, and the limitations of the physician "approved" diets.

    After doing all this reading, I wanted to try once again to moderate my lipids through diet. In January 2006 I stopped taking advicor and changed my diet so as to reduce carbohydrate. I started eating cheese omlets cooked in butter for breakfast. I did not avoid red meat. I stopped eating rice, potatoes, and white flours, etc. As a result of these changed my HDL went up from 40 to 50. My TriG went down to a low of 73. Unfortunately, my LDL went up to 130-140. So after about a year I went back on Advicor but kept my diet similar (except I went back to whole grain cereal for breakfast). The other change I made last year is I started on an aggressive exercise program. I exercise 4-5 times a week for 50-60 minutes, keeping my heart rate at the 80% level. I also got a bicycle and started biking on the weekends (that has been quite fun and rewarding - I did an MS-150 ride this past October). I've only lost a few pounds 210 to 195. At 6'3" my BMI is barely in the "normal" range. I have recently switched Doctors due to other factors, but my new doctor wants to keep me on Advicor.

    My questions are these. Am I just fooling myself into a false sense of security by taking Advicor and monitoring my lipids? Should I continue this plan or make some mods?

    I anticipate that you will recommend a CT scan. Are they costly? Covered by insurance? Require a Dr's referral? What levels of radiation exposure do they impose?

    I look forward to hearing your reply.

  • Anonymous

    1/4/2008 6:50:00 PM |

    thats very interesting. The "conventional" risks only explain about 12% of plague increase and yet its well documented that statins reduce heart attack by somewhere between 30 and 40%.

  • Dr. Davis

    1/5/2008 1:55:00 AM |

    Hi, Barry-

    Thanks for your interest. However, I cannot answer direct medical questions, as good as they are.

    I would invite you to look at the website that this blog accompanies,

    Using conventional cholesterol as your index of risk is a fool's game. I could introduce you to hundreds of people who've had heart attacks and bypass surgery who thought they were being well served by conventional cholesterol. In my view, it is a model-T of medical testing.

  • Harry35

    1/5/2008 8:38:00 PM |

    I didn't find much value in this report because it doesn't look at percent changes in CAC, which seems to me to be the most important parameter to judge if plaque is growing at an unacceptable rate. Another thing about it, they didn't consider lipoprotein subclasses, but haven't most or all of the MESA subjects had NMR testing? Combining Otvos' NMR data on the MESA population with the CAC percent change data from this study could give some real insight into how and why plaque progresses. It looks like no one is going to do such a study. Is there any way a member of the public can get the raw data from the MESA studies, so we can do our own analysis? After all, MESA was funded by public funds, wasn't it?

  • Dr. Davis

    1/5/2008 9:30:00 PM |

    To my knowledge, NMR lipoprotein analysis was not performed as part of this study, or at least used for this type of analysis to predict events or progression of calcium scoring.

  • Harry35

    1/7/2008 12:09:00 AM |

    Yes, NMR results were not used in the study, but the study was done on 5756 subjects who participated in the MESA test. Mora did a study of the NMR results 5538 MESA participants, but didn't look at CAC progression (Atherosclerosis, 2007 May;192(1):211-7) Isn't it a pretty good possibility that these studies were both done on the same basic group of MESA participants? If so, the data is out there to do a study of the effect of lipoprotein subclasses on CAC progression, it just hasn't been done yet.

    Is it possible that the MESA data could be made available to us, or is the data totally  controlled by the people who collected it?

  • Dr. Davis

    1/7/2008 3:09:00 AM |

    Hi, Harry--

    I do believe that access is obtainable, though I am unsure how restricted. Go to for application information. I would like to do the same when some research time has been freed up.

  • Harry35

    1/8/2008 1:46:00 AM |

    Dr. Davis, from reading the MESA website, it looks like the MESA people aren't going to release the data except to qualified researchers working for established institutions or companies, and who will publish their results in a peer-reviewed journal. They aren't interested in making the data available to the general public, where it could be misinterpreted, thereby casting doubt on the overall MESA project.

    That leaves people like me out, because I just want to play with the data and look for possible correlations that haven't yet been investigated, like the effect of lipoprotein subclasses on CAC progression. Unless some established research group becomes interested in doing such a study and publishing the results, it probably won't get done. It's a bit frustrating, because it looks like the data is all there to do such a study, but there is no way to get them to release the data.

    It will probably take a fairly significant expenditure of time and analysis to investigate this and publish the results, so it will probably take a full time research group to do it. With all the other things you are doing, it may be difficult for you to take the time to do this kind of study. Perhaps with your contacts with clinical research people who might be interested in investigating this, (e.g., Otvos, Raggi, Budoff, Rumberger?) you could get the ball rolling.

    Also, it looks like a person can't get the data unless they specify exactly what they are looking for, which sort of rules out looking for correlations that haven't been previously suspected. There is something to be said for shoveling all the data into a stepwise multiple regression program and seeing what pops up as the most significant parameters, but with the tight controls they have on releasing data, this isn't likely to happen, which is unfortunate for those of us who have a personal interested in prevention.

  • Dr. Davis

    1/8/2008 2:46:00 AM |

    Hi, Harry--

    All is not lost. It may take a while, but there may be some possibilities.

    I've been giving it some thought. I can't give it high priority right now, given the need to get some of our own data analyzed and published. But I believe it is something we should explore in future.

  • buy jeans

    11/3/2010 9:52:13 PM |

    --Conventional risk factors like LDL cholesterol, diabetes, and excess weight explain only a tiny fraction of why someone develops coronary atherosclerotic plaque as represented by a heart scan score.