Just Read: Challenging the role of LDL vs Insulin Resistance in predicting heart disease, via Kaiser Permanente data

This paper (full citation below) caught my eye not just because of who tweeted it, but because the study performed was done at Kaiser Permanente.

It’s been receiving quite a bit of attention on social media in just the past few weeks, even though it was published in The Permanente Journal (@PermanenteJ) in 2015. So of course I took a look 🙂 .

Confirmation – Role of Triglycerides Not Fully Understood by Physicians

When I read the Just Read: Triglycerides and Cardiovascular Disease – American Heart Association Scientific Statement, I hedged in my post saying that I personally had a poor understanding of the role of triglycerides in predicting heart disease. That’s because I was trained on the central role of LDL (so-called “bad cholesterol”), actually to the point where I could do the Framingham equations in my head (which do not include triglycerides).

This paper sheds light in a simple statement that I am/was not alone:

Other components of the lipid panel provide information for assessing CVD risk, although these risk factors are not well understood by many physicians….When reviewing the lipid panel, physicians often address the LDL-c but neglect the triglyceride and HDL-c ratio

(This is a peer-reviewed euphemism for ‘most doctors are missing the boat’)

Insulin Resistance more predictive than LDL

…and the review of the medical records of 80,328 and 103,646 who were continuously enrolled at Kaiser Permanente Northern California showed that

  • those with an indirect measure of insulin resistance (triglyceride/HDL radio) were much more likely to be diagnosed or die from heart disease whether or not their LDL levels were high.
  • 68% higher, compared to the affect of high LDL cholesterol, calculated at 19%
  • Hypertension was also significant at 60% increased risk. This is especially interesting because one of the pathways proposed for hypertension is …. insulin resistance, which may affect the way the kidneys handle fluid and electrolyte balance

Why the low-carb high-fat diet community finds this interesting

I’m taking a very good educated guess by saying that the implication is that the pathway to reducing insulin resistance is … a low carb (and by extension high fat) diet. This is not discussed in the paper, and I don’t know if the authors endorse this connection or not. However, there’s good data to show that the SAD (standard american diet) which is high carbohydrate and lower fat, causes changes that create insulin resistance.

And, this has been validated by a much larger study, the PURE Study recently: Just Read: More Validation of Low(er) Carb High(er) Fat Diets: The PURE Study and Lipids

A few things (for me) to be aware of

There are several important comments by the authors that deserve follow-up (by me, others can too)

  • What are we supposed to measure? Right now, LDL is the number. I know NMR testing of lipid profiles are happening in studies and in practices now, but these are currently not widely recommended
  • The role of other markers like Apo B-100, which the authors state was not a predictor of heart attack from the Physicians Health Study
  • The impact of what was happening at Kaiser Permanente during the study period – a fairly, no, a very impressive program to place large numbers of patients on statins and hypertensive medications, with equally impressive results. The authors discount this idea, that if the study subjects were on statins this would affect the results: “However, the results would then suggest that clinicians should also focus on insulin resistance because it remains a powerful risk factor even after treatment of high LDL-c.”


It’s time for me to read ATPIV, 83 pages of the latest cholesterol recommendations, published in 2013. Just as I read the statement on Triglycerides, it’s important to get information from the source, in addition to the smart people behind the tweets.

Thanks to

Ivor Cummins (@FatEmperor) for the tip:


As I see it, the problem with high TGs is that they inhibit reverse cholesterol transport, as described here.
This depresses HDL levels, which is why TG/HDL in the paper is calculated from both high TG and low HDL.
All atherosclerotic cardiovascular disease risk can be seen as inhibition of RCT.
Smoking inhibits multiple RCT pathways; in familial hypercholesterolaemia, LDL cannot play its important role in RCT due to deficient ; inflammatory diseases like rheumatoid arthritis reversibly impair HDL function; and of course hyperinsulinaemia, by loading excess TGs onto HDL, inhibits efflux capacity. Hyperinsulinaemia also contributes to inflammatory milieus in the other states.

The elderly populations with high LDL can be explained by survivor bias; what allows survival is a freely functioning RCT system. (Note however that most CVD is diagnosed in the elderly age group, so survival with high LDL is a very common event – the norm – helping to explain why LDL is a weak predictor, a generator of false positives and negatives, while TG/HDL and low HDL are more specific ones.).

Helpful, thank you – I have been working to understand the TG interaction with diet – it takes some understanding! I will follow up on RCT. Thanks for stopping by, Ted

Ted Eytan, MD