5 Steps to prevent Heart Disease

Tuesday, May 8, 2018

@DaveKeto Challenge did not ask me to prove LDLc is the cause of atheroma

If you say that such a level of LDL-c causes heart disease, you are strictly speaking probably not correct. It depends on everything else. It is quite possible to have a considerably higher LDL and NOT get CHD. After all, there is no correlation between LDL per se and individual levels of atheroma. Not on imaging, CAC or post-mortem. Lots of studies.

My response to this criticism of my claim that I answered @DaveKeto lipid challenge.

I don't think I said a LDLc of 105 caused atheroma.  I said high HDLc and low TG did not protect from atheroma.  I said since these patients developed atheroma they need to get their LDLc down to 70.   A better biomarker is non-HDLc which was 116 which is surprising considering how high the HDLc was.  This number needs to get down below 80 with statin.  Remember Tim Russert died with LDLc 68 but non-HDL was 118 despite aggressive treatment.  Thus these people developed CAD with non-HDLc 116, trials have shown that getting their non-HDLc less than 80 will prevent many future events.  I am not trying to prove the certainty of the Germ Theory as Koch did.  I am trying to save lives with what has been proven to work in a convergence of evidence including the best level of evidence: large trials powered to show better mortality in double blinded random control trials.
Gearoid writes he have evidence that LDLc not correlated with CAC.  I ask you to show me those studies.  First, we know LDLc is a poor biomarker compared to non-HDLc and ApoB.  Second, we know calcium score goes up with statins but it has to be adjusted with CAC volume.  Using old studies that don't take this new information into account and ignoring the new studies is cherry picking. 


  1. Hi Brian,
    If the CAC score goes up with statin, for no extra risk because plaque is stabilised, then the CAC score under statin is no longer the measure it was in an untreated population.
    Similarly, if LDL goes down under statin, associated with reduction in event rates, because HMGR has been inhibited, LDL is no longer the same measure it was in untreated population, because LDL can associate with two patterns - an HMGR-stimulated pattern (HMGR is activated by insulin and leptin) for which highest TG/HDL is a proxy - we would expect statin to be protective here as it inhibits HMGR - and an HMGR quiescent pattern (HMGR is inhibited by glucagon and cortisol) for which lowest TG/HDL is a proxy. In this pattern, the statin may be superfluous, which is what we see in the 4S analysis.
    These people still have some events, albeit a much lower rate, but lowering LDL with statin is not preventing them.
    So we can say - yes, the statin works. And it lowers LDL across the board. But it doesn't work for everyone, and TG/HDL, not LDL, is predicting whether it will work.

    1. That is, fasting TG/HDL as an untreated value.
      If it is a drug treated value, then, as with CAC, all bets are off - Goodhart's law applies.