I would like to start this discussion by offering some guidelines, based on the Framingham Heart Study (FHS), to help physicians who are interested in preventing atherothrombotic disease (ATD) events. While these guidelines are based on FHS precepts, I have had to modify some of them slightly to reflect advances in knowledge. The basic concepts of the FHS were true a half century ago and are still true today, even if they are ignored by many current researchers and practitioners.

First , let me state the obvious. You can not protect patients from ATD if you can not predict the population at risk of ATD, and the better you can predict, the better you can protect.

Second, as W.B.Kannel, MD pointed out in the 1960's, ATD events do not occur as "a bolt out of the blue." Such events occur in patients with ATD risk factors. Those with low levels of such risk factors are at low risk of ATD, with events occurring mainly in very old age.

Third, there are no risk factors unique to the ATD population. Those who develop ATD simply have more severe risk factors than those who do not develop early onset ATD. (Kannel was a bit ahead of his time when he said, in the 1970's, that an ATD event occurring in a patient represented a failure of the attending physician's professional competence. There is no safe level of any risk factor that will afford freedom from ATD. Indeed, risk factor analysis indicates that the risk factors occur in a spectrum of minimal risk to maximal risk, with the former being associated with ATD events in very old age, whereas the latter being associated with ATD events at an early age.)

Fourth, ATD risk factors do not operate in isolation. They are part of a milieu, interacting to produce ATD when the "mix is right." In other words, lower level risk factors acting together may cause ATD more so than a single risk factor of higher severity. Hence it pays to know and integrate all of the major ATD risk factors and not just to concentrate on a single risk factor.

Fifth, statistical averages do not have ATD events--patients do. Truth should never be sacrificed on the alter of statistic significance, which is why the Framingham Risk Score should not use total cholesterol (CT) and high-density lipoprotein cholesterol (HDL-c) as its predictors instead of the CT:HDL-c ratio, the latter being much more predictive than the former.

Sixth, the duration of time that a risk factor has been operative is key. One does not develop a risk factor today and have an ATD event tomorrow. Moreover, risk factors may worsen as the patient grows older. Hence it is important to know the level of any risk factor as early in life as possible. There are trajectories indicating the risk of an ATD event and as risk factors worsen over time they "jump" those trajectories, hastening the day of ATD reckoning. This is the basis of early and repeated testing of ATD risk factors. Moreover, it is true that damage done by untreated risk factors can be lessened when that risk factor is appropriately treated, but it is probably also true that the damage does not completely resolve--again stressing the need for early diagnosis and treatment.

Seventh, randomized controlled clinical trials are good for answering a specific question, but since they are selection biased, they can not be generalized to the population as a whole. Observational studies may have their problems, but they are still the best means of doing long-term investigations.

Well, I hope that this begins a productive conversation.

To continue this discussion, let me add that anyone can jump in whenever they have a question. In order to predict the population at risk of ATD, one needs a database. This is easy to do: simply get a full lipid panel, a 2 hr postprandial blood sugar level, perform a blood pressure measurement, and most importantly inquire about cigarette smoking status. You can predict the population at risk of ATD very easily with these measurements. But a word of caution: you have to know how your lab measures high-density lipoprotein cholesterol (HDL-c). Since we calculate low-density lipoprotein cholesterol (LDL-c) by the Friedewald formula and since the calculated LDL-c depends on how HDL-c is measured, this is very important.

We do evidence-based medicine to achieve the best results for our patients. Expert opinion is at the lowest level of evidence. In the 1980's and 1990's, multiple angiographic regression trials were performed to assess the effects of dyslipidemia therapy upon plaque. Angiographic stabilization/regression of plaque is associated with a marked drop in the subsequent occurrence of ATD events. The angiographic regression trials of this time frame measured HDL-c by the precipitation method. In POSCH, angiographic stabilization/regression of plaque could occur even if LDL-c rose so long as HDL-c rose even more. Conversely, angiographic progression of plaque could occur even if LDL-c fell, provided that HDL-c fell even more.

The thing is that in 1999 in my neck of the woods, the autoanalyzer manufacturers changed the technique by which HDL-c is measured from the precipitation method to the enzymatic method. The two methods give differing HDL-c results, the enzymatic methodology giving an HDL-c level about 0.25 mmoles/L (10 mg/dl) higher than did the precipitation method. That in turn means that the calculated LDL-c will be lower by the same amount. Any lipid ratio will be wildly different, depending upon which HDL-c measurement technique is used. In order to avoid these measurement differences, and to continue to base my treatment of the randomized clinical trials, I convert the HDL-c levels given by my lab to the precipitation-method equivalent. I reported this in the Journal of Clinical Lipidology in 2008.

No questions so far? I would like to add a few more comments. The importance of having evidence-based medicine was first stated (to my knowledge) by William W. Connor, MD, who pointed out some 50 years ago that we do not treat a blood factor to reach a certain goal, but rather to prevent disease. I have published a review of eight angiographic regression trials (the article is available on Research Gate) showing that it does make a difference as to the means by which the lipid goal of treatment is reached and that there is no safe level of LDL-c that eliminates the risk of ATD--so it is a question of what ATD risk is acceptable. If a ratio between LDL-c and HDL-c is used to gauge ATD risk , then a Cholesterol Retention Fraction (CRF, defined as [LDL-c minus HDL-c]/LDL-c of 0.59 or less is associated with a very low ATD risk. (More on this predictor in my next set of comments.) I have talked to a number of lipid experts and they say, off the record, that a lifetime risk of ATD of 15% or higher is a level at which treatment should begin. If we use this target as a marker for treatment initiation, then a CRF of 0.59 or less does not attain this threshold until the ninth of life. Whether or not all people should be treated to this level is an open question. More on this later.

To continue. It is time to talk about the lipid predictor. There are four major predictors, not counting triglycerides (TG) which only indirectly predict via high-density lipiprotein cholesterol (HDL-c) and they are the Cholesterol Retention Fraction, low-density lipiprotein cholesterol (LDL-c), non-HDL-cholesterol, and total cholesterol (CT). The Cholesterol Retention Fraction (CRF, defined as [LDL-c minus HDL-c]/LDL-c) is the newer kid on the block, but I have shown that it is better than the other predictors. (The article detailing the differences is available on Research Gate. This data has also been presented at the International Atherosclerosis Society symposia of 2006 and 2018. I believe that it has also been presented at a National Lipid Association symposium as well.)

The CRF represents the balance between the cholesterol entering the artery wall (LDL-c) and the cholesterol exiting the artery wall via reverse cholesterol transport (HDL-c). There has never been any argument that HDL-c is not protective against atherothrombotic disease (ATD) and that is why the CRF is better than LDL-c by itself--if only LDL-c is used, then everyone has to have their LDL-c lowered to 80mg/dl (2.0 mmoles/L) to be virtually protected against ATD. If LDL-c is calculated on the basis of the enzymatic method of measuring HDL-c, the the number is 70 mg/dl (1.75 mmoles/L). This article is also available on Research Gate. CT gives the treating physician no clue as to the levels of LDL-c or HDL-c, and hence no idea of the patient's ATD risk. Non-HDL-c is simply the backdoor approach to the CRF, but advocates for therapy when the only abnormality present is elevated TG.

The CRF has its limitations. As Roger Illingsworth,MD, pointed out to me almost 40 years ago, when LDL-c is very high, it can overwhelm the protection afforded by HDL-c. In my local lab, this level of LDL-c appears to be appears to be 170 mg/dl (4.25 mmoles/L), though if the enzymatic method of HDL--c is used, then the LDL-c threshold is 160 mg/dl (4.0 mmoles/). This is why I developed the concept of the Cholesterol Threshold (C Thr). The C Thr is exceeded whenever the CRF is 0.70 or higher or at any CRF if LDL-c exceeds 169 mg/dl, and treatment is therefore warranted. Additionally if LDL-c is also very low, the HDL-c is immaterial.

One more thought on this topic. All lipid predictors, consistent with Framingham precepts, are present, not as an all or none phenomenon, but rather as a spectrum of risk, with a higher intensity end and a lower intensity end. In the general population (that has not yet developed ATD) the levels of the predictor are present through out the spectrum, but concentrate toward the lower intensity end of the spectrum in those without clinical ATD, whereas in the ATD population the predictor tends to concentrate in the higher intensity end of the spectrum. If one divides the predictor into sextiles, then one notes a higher frequency of the predictor in the lower intensity end of the spectrum in those who have not yet developed ATD. This is characterized by a line descending from the northwest corner of the graph to the southeast corner. Conversely, if one graphs a similar line for the frequency of sextiles for ATD patients, the same line ascends from the southwest corner of the graph to the northeast corner. Putting the two graphs together, one notes an X-shaped picture, showing that one predicts the population with ATD better and better in fewer and fewer people. This is true for all predictors except the CRF, where the high-intensty frequency levels off in the higher sextiles (southeast corner of the graph) so that the CRF predicts the ATD population better as the predictor becomes more intense.

You know, I apologize for forgetting the Framingham Fraction (FF, defined as the CT:HDL-c ratio) as a lipid predictor. I don't know how I could have done that and I hope that Bill Castelli,MD, will forgive me. The CRF is better than the FF because HDL-c is the most difficult of the lipid moieties to measure, and hence the variability in HDL-c measurement makes reproducibility of any given FF problematical. Additionally the Ff follows the X-shaped pattern noted above.

A couple of things. We should discard the term hyperlipidemia in favor of the term dyslipidemia for the same reasonws that we should discard the term atherosclerotic disease in favor of the term atherothromboitic disease (ATD). In the former scenario, the term dyslipidemia describes the balance between LDL-c and HDL-c which is actually germane to the topic of ATD, whereas in the latter scenario the threshold for LDL-c would have to be lowered to 80 mg/dl or lower to be protective. The reason for the change to ATD is simply that one can not understand clinical ATD unless one considers thrombosis, either from a ruptured plaque or from plaque erosion.

Secondly, only rarely do ATD risk factors operate in isolation, as single agents. While it is true that treating ATD risk factors as isolated agents gives one more statistical power, this is not what happens in the real world. Truth should never be sacrificed on the alter of statistical power. Did you ever note that Bill Kannel,MD, went to his grave proclaiming the superiority of the FF (see previous comments) and yet the Framingham Risk Score uses CT and HDL-c as independent risk factors--and NOT as a ratio. I asked Peter Wilson,MD, about this and he told me what I have told you.

Finally, my next comments will expand on the interactions of ATD risk factors that prodece clinical ATD events.

Well, guess what! The internet server went down and the repairman fixed that, but disconnected me from the e-mail and yahoo.com would NOT let me sign back on. I finally got that resolved, but it cost me $110 and hours of time. So, I am sorry to have missed out on communications, but that's life.

When I started on this work some 45 years ago, I knew I needed an age-sex database to determine the differences between those who developed clinical ATD and those who did not. Since the other doctors in town were not interested in the primary prevention of ATD--and hence neither were patients--it was not long before I accumulated a relatively large database of ATD risk factors. Furthermore it was not helpful that the medications needed to do primary/secondary prevention of ATD were not very helpful either. By 1981, I had enough people in the general population database to separate out a subgroup of those who developed some clinical form of ATD, and enough people in the ATD database to check for any predictive measures. Cigarette smoking and hypertension were simple enough, but the lipids were a puzzle. Some ATD patients had events despite a low level of LDL-c or a high level of HDL-c. Then in 1981, an article was written entitled "Is the LDL:HDL ratio the best lipid predictor?" (I regret that the article is lost to me so I can not give credit where credit is due.) So I went back to my database, and lo and behold, those ATD patients with low LDL-c levels almost always had very low HDL-c levels and those patients with high HDL-c levels almost always had very high LDL-c levels. When this was not the case, the younger patients were cigarette smokers and the older patients had hypertension, with or without diabetes. The ratio was spot on. Then one morning in 1983 I was drowsing in bed when it occurred to me that what I really wanted to know was the best estimate of the amount of cholesterol accumulating within the artery wall. This I reasoned could best be determined by subtracting the cholesterol exiting the artery wall by reverse cholesterol transport (HDL-c) from the cholesterol entering the artery wall (LDL-c) and then dividing the difference by LDL-c. Thus was born the concept of the Cholesterol Retention Fraction (CRF), which is abnormal at CRF=0.70 or higher. However, as pointed out to me by the late Roger Illingsworth,MD, apo A-1 is unable to compensate for unlimited LDL-c, so I went back to my ATD database and found out that the ability of apo A-1 to compensate for LDL-c began to wane at LDL-c levels of 170 mg/dl (4.25 mmoles/L). This is the case when HDL-c is measured by the precipitation method. If HDL-c is measured by the enzymatic method, then the cut points are 160 mg/dl or 4.0 mmoles/L.

Finally back in 1981, review of my ATD database showed that only rarely did ATD risk factors operate in a vacuum. They were mostly present in various combinations. So I looked for any risk factor combination that might turn out to be predictive,but the only risk factor combination that had any predictive value was the CRF and systolic blood pressure--and then only when stratified by cigarette smoking. This led to the development of the graph noted in published articles and presented in article published in ResearchGate. More on this later.

Here I am again! In light of all of my preceding comments, the next step is obvious. Put all of the major risk factors into a single predictive tool, but keep it simple so that the doctors--but more importantly the patients--can understand it. Statistical analysis works best with independent risk factors, but unfortunately ATD risk factors are not independent of one another in most cases. True Familiar Hypercholesterolemia (HeFH or HoFH)and cigarette smoking can be isolated risk factors in younger patients, as can hypertension (with/without diabetes) in older patients--but in about 80% of ATD cases, it is a combination of less severe risk factors that is operative. That is why my predictive tool covers the waterfront. Using LDL-c instead of the CRF will work a somewhat, especially in HeFH or HoFH, but when mid-levels of LDL-c are involved, one must know the attendant HDL-c to determine the atherogenicity of LDL-c. This is best described in terms of a ratio--as the Framingham Risk Score showed. (See previous comments.) In patients from south-east Asia, where low HDL-c is the main lipid disorder, the CRF is especially important.

When the CRF is utilized as the lipid predictor, the predictions are the same in women as in men. This is not true when only LDL-c is used.

The risk factor graph is available in articles listed on ResearchGate. Look under the article: "Enhanced Prediction of the Population at Risk of Atherothrombotic Disease: Back to Framingham" or "Prediction of the Population at Risk of Atherothrombotic Disease: A Three Step Approach."

Before I go any further, I need to remind you that the lipid data are based on the calculated LDL-c via the Friedewald equation and the LDL-c value depends upon the HDL-c value. Most of my data were collected prior to 1999, which is when the manufacturers of auto-analyzers changed the methodology of HDL-c determination from the precipitation method to the enzymatic method. The results (of HDL-c determinations) between the two methodologies are not the same, with the results from the newer enzymatic method being on the order of 10 mg/dl (0.25 mmoles/L) higher than those obtained using the older (then gold-standard) method. That in turn means that the calculated LDL-c using the enzymatic method (of HDL-c measurement) will be on the order 0f 10 mg/dl (0.25 mmoles/L) lower using the newer methodology. This is bad enough, but the results are really bad when LDL-c and HDL-c are used in a ratio. It also means that the angiographic regression studies, from the 1990's, which I rely on for validation, are also compromised if referred to in terms of today's lab data. It would have been nice if the auto-analyzer manufacturers had let the medical community know about the change and given time for professional comment to avoid these problems.

But now on to the predictive tables. The basis of the tables to which I referred in my prior commentary is simply that 1) there is no set of ATD risk factors that guarantees freedom from ATD once the incubation period is over and 2) there are no ATD risk factors unique to the ATD population. Everyone has ATD risk factors, to a greater or lesser extent, which is why if one lives long enough and does not die of something else, one will eventually sustain an ATD event. Hence the prevention of ATD is actually not prevention but rather having the ATD event as late in life as possible. To predict when the inevitable ATD event will occur one needs to know the chief ATD risk factors for each and every patient. This data is then arranged as described in the tables. On then determines which CRF-SBP cohort best describes the patient in question and then predicts the average age of ATD onset based on the risk factor cohort. The only problem is that there is an incubation period over which cholesterol builds up in the artery wall and the only way to know where one is in the plaque-build-up time frame is to know when the dyslipdiemia began.

This is why I have advocated for the early lipid testing of people, beginning in childhood. Many children already have adult level lipids, even when young. (Cigarette smoking also starts in early teen-age years, while hypertension and diabetes tend to come along later in life.) To put things in the common parlance, you can't know where you are going if you don't know where you've been. Children who start with adult level lipids will manifest clinical ATD at a younger age than those who start with ideal lipid values. Lipids often deteriorate with age and the time to develop atherogenic lipid levels. Lipid trajectories can be drawn and trajectory hopping to higher trajectories can be expected. If a patient begins on a higher lipid trajectory, plaques will occur earlier in life. (The US Preventive Task Force Services is just plain wrong on its decision not to recommend lipid screening in children.) Incidentally, at time of puberty, hormones flush the body and make beneficial changes in lipids, at least in women, so it is best to get lipid testing prior to puberty. Moreover, cigarette smoking may have adverse effects on lipids, so it is also good to get lipid values before children start smoking. Finally, low levels of blood pressure, as often seen in children, may be protective against plaque formation. I can remember Bill Kannel, MD, saying that he had never seen plaque formation in the pulmonary artery circulation--and this can continue into adulthood--though this is NOT a reason to refrain from treating dyslipidemia.

As you set up your ATD prediction database, and you should do so for your own research purposes since your regional laboratory and my reference lab may differ in the results. About 40 years ago the New York City health department collected a large volume of blood and had the CDC determine the level of total cholesterol (CT) in the sample. The health department then divided the reference sample into 50 samples and then each sample into two sub-samples. The health department then sent out one of the sub-samples for CT measurement one week and the second sub-sample the next week--again for CT measurement. The samples were sent to 50 local laboratories. The CT results varied widely, though were consistent for each individual laboratory. Remember that CT is the most accurately measured lipid fraction and that HDL-c is the least accurately measured lipid fraction, so one can imagine that there may be a lot of variability between labs. My regional lab standardizes against CDC standards on a weekly basis. Perhaps your cab could do the same thing.

When you set up your spreadsheet and calculate the CRF, you will note that perhaps as much as 1% of CRF values may be negative. This is virtually always seen in people with low LDL-c and normal HDL-c, mostly children. Only two of my adult patients with negative CRF values developed ATD: the 60 year old was a cigarette smoker with a sky high blood pressure and the other was a 89 year old patient who had never smoked cigarettes. The former patient died from a cardiac arrest and the latter patient died in her 90's. So I don't think that this need be of concern.

One last comment for this session: Don't get lost in dogma. There are those who think that LDL-c is the be all and end all for the lipid portion of ATD prediction. This is not true. Remember that in (true) science, appeal to authority is not acceptable. When a hypothesis is presented, along with the supporting data, the hypothesis is tested to see if it is valid--it is not rejected out of hand just because it differs form your deeply held beliefs. The hypothesis is only rejected when it can not be replicated by independent researchers. This must be remembered in the case of the CRF and the change in HDL-c measurement in 1999. All of my data and the published angiographic regression studies I cited are based on the precipitation method of HDL-c measurement. Use of the enzymatic method of HDL-c measurement must be corrected for this change--or you could set up your own database using the enzymatic method, using the principles I have discussed.

A couple more comments. Lipid trajectories change over time--and hence the risk of ATD changes over time as well. It is useful to know how fast the patient is jumping the lipid trajectories, but this requires repeated lipid testing, which is usually not available. In other words, not only is it useful to know which lipid trajectory is present at baseline, but how fast the patient is jumping from one trajectory to another. Change in trajectory pattern is as important as the baseline trajectory and may occur over a relatively short period of time or a long period of time--it depends upon the patient. What this means is that the goals to be achieved amount to hitting a moving target, especially with respect to cigarette smoking. Current cigarette smoking shifts the incidence of ATD to younger ages for all lipid trajectories, whereas past smoking or never smoking status shifts the age of ATD onset to older categories. Likewise hypertension can shift the average age of ATD onset to the older years. I would like to suggest that the only way to follow this approach is to set up a lipid clinic that will remain intact for decades, so that a large database can be established. This can be done at one large clinical center or by many physicians who are interconnected via the net.

My thanks to Ehud Baron for his recommendation! Here is an easy way to describe interventional lipidology to your patients: I call it "The Dirty Carpet Theory of Atherothrombotic Disease." I ask patients to remember when their kids were young and playing outside in the dirt when dinner time came around. Mom calls the kids into the house to eat and they come in dragging in dirt on their shoes, dirtying up the brand new carpet. A mother, being house proud,. will of course get out her vacuum sweeper to clean up the carpet. If the vacuum cleaner is working well, she can clean up the dirt easily--especially if there is not much dirt. If the vacuum cleaner is not working well--say the bag is 90% full--or if there is a huge amount of dirt, the carpet will remain dirty despite Mom's best efforts. Imagine that the dirt is the "bad"cholesterol (LDL-c) and that the shoes (carrying the dirt) are apo-B particles. Imagine that the carpet is the arterial wall and that the vacuum cleaner is apo A-1 and that the dirt picked up by the vacuum cleaner is HDL-c. To keep the carpet clean, one needs to cut down on the incoming dirt or to enhance the vacuum cleaner. All patients can understand this scenario and a patient who understands what is going on with his/her lipids will likely stay on his/her treatment regimen.

When I noted my "Dirty Carpet Theory of Atherothrombotic Disease" I neglected to note that the children's shoes (that carry the dirt into the house) amount to apo B-100. Just to complete the picture.

BTW, I do not want anyone to think that I am ignoring LDL-c in favor of the CRF. LDL-c and the CRF go together, hand in hand--one enhances the other. You may remember that i described the finding that LDL-c of 170 mg/dl (4.25 mmoles/L) and higher overwhelms the ability of apo A-1 to remove cholesterol from the artery wall. On rare occasion the CRF will be elevated but LDL-c will be low (under 99 mg/dl or 2.5 mmoles/L). In such cases, in a primary care scenario I do not treat LDL-c. In a secondary case scenario, however, some benefit can be gained by bringing LDL-c down to levels below 80 mg/dl (2.0 mmoles/L).

Just as an addendum to my last comment, someone with a low LDL-c and high CRF may not be at risk of ATD at the precise moment, but LDL-c does rise with age while HDL-c, if anything, declines with age. Hence such patients need follow-up because the ATD will begin to tick when LDL-c rises.

I would like to note that I have published the case reports of two patients, both still taking their statins, who reached the ripe old age of 100 years. The older patient was 75 when she had her myocardial infarction under the care of another physician. No dyslipidemia treatment was offered and her chest pain returned after about 5 years after he CABGS. When her physician did nothing about her lipids, she came into my care and I started her on a potent statin as well as a CCB for the pain and a baby aspirin. She was followed by the CRF since her HDL-c was not low and she finally stopped her statin at age 104 years. She subsequently suffered mental problems and died at age 104 and 1/2 years. She had never smoked cigarettes. The second patient was found to have dyslipidemia at age 71 years; he was a cigarette smoker. He was put on high-intensity statin therapy and he quit smoking. At age 97 years or so, he developed heart failure and finally died at age 100 years of heart failure. He was followed by LDL-c levels. because his HDL-c was low. As far as I know, there are no other reports of people achieving such an advanced age, still taking statin therapy. The articles are available on ResearchGate

One last thought. Diabetes mellitus is said to be an ATD equivalent. This is true for macrovascular disease, mainly because diabetes is not just a disorder of blood sugar, but rather a mix of dyslipidemia and hypertension, and everything is made worse by cigarette smoking. Indeed, my diabetic patients who have given up cigarette smoking--or far better yet, have never smoked cigarettes--live well past the average lifespan for Americans. Indeed, I have had to initiate insulin therapy in people (who are not cigarette smokers) in their ninth decade of life after properly treating their dyslipidemia and hypertension. No diabetic patient of mine has ever gone on dialysis if I have taken care of them from the get-go. BTW knowledge of the total ATD risk milieu is absolutely essential to properly treat any individual ATD risk factor--if one wants to prevent ATD--as William E. Conner, MD, pointed out almost 50 years ago.

Also, given the propensity of diabetics to form thromboses, one could consider prophylactic administration of aspirin 81 mg daily to prevent clotting.

Finally, accurate control of diabetes is essential to prevent microvascular disease. But then microvascular disease is rare in my practice.

Well, I can't think of anything else to say--except that I am working on two papers. (Things move slowly once one retires and all of his IT people are gone to other jobs.) The first paper will be on debunking three ATD regression mantras: that the lowest LDL-c levels are the best, that it does not make a difference how LDL-c is lowered, and that consideration of HDL-c is immaterial. The second paper will sum up all of the things that I have published already into one paper.

But as for now, I think that I am finished with what I want to say and, as noted a while ago, I am open to any and all questions.

Well, since there are no questions, I guess that it's time to wrap things up. What I have shared with you are the results of over 50 years of working in the fields of interventional lipidology and preventive cardiology. Over the course of my professional life, I have sen most every thing in these fields, and I wanted to share my experience with you all. Remember that what is going on in these fields at present is dogma based on science, but not pure science. Pure science requires the examination of new hypotheses, testing them to determine if they can be verified. Pure science is not "My way or the highway." In any event, there single most important thing that you can to to prevent ATD remains (as I wrote about 25 years ago) getting people to stop smoking cigarettes--or better yet get them to never start smoking in the first place. That's one of the reasons why it is so important to talk to people at younger ages. The other thing is that by testing early in life, one can determine the lipid trajectory because the rate of ATD development depends upon which lipid trajectory on is on. People on the lower trajectories develop ATD at a much slower rate than do those of the higher trajectories. Since most people will jump form a lower trajectory to a higher one as they age, it is helpful to know where they started. Finally, the best lipid predictor is the ratio between LDL-c and HDL-c. (I prefer the Cholesterol Retention Fraction, or CRF)since Framingham can not use LDL-c due to Kannel's use of non-fasting blood during the first half of the Framingham Heart Study's existence. At any level of LDL0-c, a higher CRF portends an earlier age of ATD onset and a lower CRF portends a later age of aTD onset.

Always remember that we must base our treatments upon evidence, which is why we run trials. We may arrive at conclusions the than the one at which the trial's investigators arrived--and that is OK as long as the trial data supports the alternative conclusion. Also, remember that we must always stay within the same data systems