No announcement yet.

Relative and Absolute risk reduction: Aren't both misleading?

  • Filter
  • Time
  • Show
Clear All
new posts

  • Relative and Absolute risk reduction: Aren't both misleading?

    One of the first things I see when statin effectiveness comes up is the argument about relative and absolute risk reduction.

    The "relativists" will look at a risk reduction of 3% to 1.5% and say "That's 50%", while the "absolutist" will say "That's only 1.5%".

    I understand the absolutist position and I have a tendency to give it more credence, but is it all that clear? Over what time frame? If the reduction is 3% to 1.5% of x years, might it be 6% to 3% after x times 2 years? And therefore the relativist position be that it's still 50%, but the absolutist would now have to concede that it's up to 3%. Or maybe not, there may be diminishing returns over time, which is sort of my point: it's much more complicated statistically than either side admits. The truth, I believe is somewhere in between. And I don't mean take the average of 50% and 1.5% and say that's it, either. Rather the truth (I believe, and subject to revision, given evidence) is somewhere between them qualitatively: somewhere between "Useless" (1.5%, or whatever) and "Wonder Drug" (50%, or whatever.)

    By lowering the cholesterol level standard that make one "eligible" for statins (I love that - like winning the Publisher's Clearinghouse Sweepstakes - you are :"eligible" to take statins), I believe we're going the wrong way - we should be improving the reduction by better identifying who will truly benefit, rather than scattershot approach of just increasing the number of people who take statins (or any medication, for any condition); the scattershot approach will perhaps pick up larger numbers, but probably lower percentages. It does have the advantage of increasing drug company profits, so is clearly the better choice (said with irony.)

    And a more honest assessment of side effects are in order; I've seen some claims that "1 in 10,000" people have adverse effects, a ridiculous number on the face of it but actually put forth by some. I have queried, non-scientifically, many colleagues who are on statins and a very high percentage report they stopped taking them because of the adverse effects. Probably a higher percentage than the population as a whole, because people who can't take them tend to be more vocal than people who don't have effects, or at least don't recognize effects.

    I guess this is actually three topics, but maybe under the heading of "Lying with statistics." There is a long joke about applicants for the job of Soviet mathematician that goes on for some time but the punchline is that the guy who gets the job is the one who answers "What do you need it to be?" when asked by the communist party interviewer "How much is 2+2?" This is what is happening when we look at medical statistics.

    Any thoughts on this?

  • #2
    Yeah some thoughts. First, if LDL-C doesn't cause cardiovascular disease, then what is that cholesterol and those plaques doing that are found in the arteries of people who died of heart disease and stroke? If that isn't what is causing cardiovascular disease then what is? Well it really is cholesterol and plaques that cause the cardiovascular disease, and then the question becomes how come with all of these people with cardiovascular disease are having normal LDL-C? Follow the biology. What do a lot of these people with normal LDL-C have in common is decades of insulin resistance causing cardiovascular inflammation and endothelium dysfunction. If you get enough of that long enough, then a "normal" LDL-C will lead to varying amounts of cardiovascular disease by the time you are 50/60 (maybe younger depending upon your genetics and lifestyle). Secondly many people with metabolic syndrome, and it is a lot of people in the US now, have LDL-C discordance with LDL-P values as LDL-C doesn't rise as much as LDL-P given increasing metabolic syndrome. That means that those people are mislead by the "normal" LDL-C value to start with because it is really LDL-P that counts as far as how much cholesterol gets deposited in the arteries.

    Next, more people die or have life-limiting results from suffering heart attacks and strokes than other causes. I keep noticing that people keep focusing all of their ire on LDL-C. They don't seem to understand how inflammation plays into cardiovascular disease, and how statins significantly lower inflammation. For many people, especially those with "normal" LDL-C and a healthy diet/lifestyle, lowering inflammation could be enough to keep them healthy for a few more years than would otherwise occur. It may well have been by luck that statins do so well on reducing inflammation as that wasn't appreciated until later. Lowering of inflammation is so important that it needs to be emphasized in every conversation on statins and LDL-C. Follow the biology.

    If you still don't believe that LDL-C causes cardiovascular disease, look at one of the PCSK9 inhibitor studies. The patients had been on a high dose statin for a long period, but there was still a residual risk. The PCSK9 inhibitors powerfully lowered the LDL-C level, sometimes as low as 20 mg/dL, and the results showed a reduction in heart attacks in as little as three years. This short timeframe is amazing as most drug studies need at least 5-6 years to show a measurable effect. We could have a discussion on PCSK9 inhibitors, but that would be too long, and frankly those are so expensive that few people will be taking that family of drugs until the patents expire. Without a doubt, those studies did prove that lowering LDL-C does reduce the risk of heart attack.

    Another thing that people keep bringing up is the 10 year risk models. The problem with a 10 year risk model is that many of us will live 30-40 years after we get that first inkling that there is a real cardiovascular problem going on. The risk goes up exponentially as we age, and that seemingly low risk climbs to a significant level and is what eventually catches up to us (on a population basis). The final thing in this area is that medical guidance on treatment is generally based upon population models, but that doesn't mean that it will be perfect for all of us (different genetics and lifestyle can make a difference).

    Yes, there are many people saying that they have side effects with taking statins, and I was one of them. When some people talk about "statins", they somehow group all of the statins with all of the individual doses of each of those statins together as some monolith to be fought. There was a time when simvastatin came off patent when a lot of people were prescribed a high dose. The reason for the high dose was the less than optimum lifestyle that so many people were/are leading. For the most part people don't want to make serious, long-term changes to their lifestyles, and so off to the primary care doctor to get a high dose of the statin of the day. Some of the people taking it noted not feeling good with a very few having muscle breakdown issues. I was one of the not feeling good folks. So, what did a lot of people do, they stopped taking the statins. Well, they and their doctor typically didn't explore lowering the dose or choosing another statin. They just stopped taking "statins" told their family/friends of their bad experience, especially after seeing some articles on statins causing problems, and felt vindicated when the "LDL-C doesn't cause cardiovascular disease myth" arose. What did I do in response to my statin side effect? I lowered my dose and the side effect went away. I got most of the benefit, and from what I can see little drawback. I will say that my liver enzyme ALT increased a bit right after I started taking a statin, and that is also probably typical. Most side effects of statins can be managed with careful choosing of statins and doses. There is perhaps 1-2% of people who absolutely cannot take any statin on the market at this time. That leaves a lot of people who could. Even a very small dose statin will give a meaningful lowering of cardiovascular inflammation.