Chris Masterjohn on saturated fat

[MarcR]

I love the nuance Dr. Masterjohn brings to this topic as he discusses the particular properties and functions of short-, medium-, and long-chained saturated fats. For example, I learned that coconut oil is 60% medium-chained according to the usual definition - 8, 10, or 12 carbon atoms - but only the 8- and 10-carbon fats can be used for energy when insulin is elevated. 8- and 10-carbon fats comprise only 15% of coconut oil and 100% of MCT oil, which explains why MCT oil packs a wallop.

The bit about coconut and MCT oil is just one example of the article's merit - I recommend it to anyone who is interested in fatty acid metabolism:

Saturated Fat Does a Body Good

[Julie G]

fascinating article, Marc. Thanks for sharing. Learning about the Kitivan's high SFA intake feels relevant given their high propensity of ApoE4. When I learn about the benefits of SFA, however, I feel conflicted knowing the general effect it has upon our population's advanced lipids... Do you share my qualms?

[marthaNH]

It all makes sense to me, too, but my own body's apparent over-reaction to the stuff just makes me scared. I don't try to eliminate it -- don't think you could -- but I can't bring myself to embrace it again.

[MarcR]

When I learn about the benefits of SFA, however, I feel conflicted knowing the general effect it has upon our population's advanced lipids... Do you share my qualms?

I don't because I am skeptical of the conventional wisdom that low serum lipids are a surrogate for mortality risk.

On the contrary, I have qualms about depriving my cells of the benefits of a generous supply of lipids. (For an extreme example, consider Russ' cautionary tale.) Here's an 8-year study of 100,000+ people aged 50+ without preexisting diabetes or heart disease that shows much lower all-cause mortality among people with high total, HDL, and LDL serum cholesterol as long as triglycerides are not also high:

The current study aimed to test whether having levels of TC and its subfractions within the recommended target goals at baseline provided a survival difference in a general population of older adults. The most striking finding was that compared with the reference levels, high TC, HDL-C, or LDL-C levels were associated with lower mortality in the elderly and this was the case for even very high levels. The finding that high TC or LDL-C levels were associated with a lower mortality is contrary to the general assumption that there is a higher mortality among subjects with high lipoprotein levels. Our findings could seem controversial. However, most studies performed in older adults show an inverse association between TC and mortality [11,14] and a recent study demonstrated an inverse association between TC and non-cardiovascular mortality in a population free of CVD and statin use at baseline [13]. These findings were significant from the age of 65 years and were largely due to an inverse association with non-HDL-C. Our study demonstrates the same significant inverse association between high levels of TC and its subfractions and mortality in all age groups from as early as 50 years.

The evidence for a causal relationship between hyperlipidaemia and development of CVD is overwhelming [2,24]. However, our data indicate that a major part of the general population aged 50 + has lipoprotein levels above the recommended levels and that those groups without CVD or diabetes at measurement have a survival benefit compared with the groups with the recommended low levels.

The whole paper is worthwhile, and I particularly encourage careful review of the four results tables.

I understand high triglycerides to be a marker for high insulin, so this study and others like it reinforce for me the importance of keeping insulin low and trusting that my body will maintain appropriate supplies of the various lipid subfractions as long as I continue to provide it with ample, high-quality, unprocessed, non-insulinogenic food while maintaining my metabolic flexibility through fasting and exercise.

Given the mortality reduction associated with the low triglycerides / high cholesterol scenario, I think my ApoE4-typical lipid production response to saturated fats may well be a sign that I benefit disproportionately from a generous supply of non-oxidized serum lipids.

[Julie G]

I admire your certainty, Marc. I'm generally a fan of the HFLC approach for our population. It's restored my health, but getting there with primarily SFA scares me. I've seen too many of my E4 friends, who eat a high SFA diet, end up with high calcium scores or coronary blockages. I know it's difficult to make a direct correlation, but that alone gives me pause. The study you cite is impressive, but not conclusive. And, of course, ApoE status wasn't taken into account.

We have lots of evidence that higher cholesterol is not only associated with longevity, but also better cognition, especially in the elderly. We also have some evidence that high cholesterol in middle age IS associated with worsened cognition as one ages, but that's never been separated from glycemic biomarkers so it needs to be interpreted with caution. The notion that low TGs are associated with better mortality is also no surprise. Low TGs are correlated with better glycemic markers and subsequent overall health. FWIW, I have little qualms about high TC with low TGs as long as LDL-P and/or ApoB is also low. I may be overly cautious, but until we learn more I plan to heed advanced lipids. To my knowledge, high levels there have NOT been associated with improved mortality.

Given the mortality reduction associated with the low triglycerides / high cholesterol scenario, I think my ApoE4-typical lipid production response to saturated fats may well be a sign that I benefit disproportionately from a generous supply of non-oxidized serum lipids.

The assumption that using unsaturated fat leads to oxLDL has been debunked by many here including myself. My oxLDL is in the lowest quartile and I keep SFA below 10% and overall fat intake at 60-65% of my total calories.

This whole area is understudied and wide open for debate and I certainly don't claim to have definitive answers. Thanks for sharing your perspective.

[MarcR]

Thanks for weighing in, Julie. I always appreciate your perspective.

I'm a long way from certainty; in fact, I perceive myself to be less certain than most of us here on this topic. As I see it, mainstream thought in these forums runs something like this:

- A lipid profile consisting of low triglycerides, low total cholesterol, and high HDL (and low calculated LDL) is safe. As long as triglycerides are low, high total cholesterol, high calculated LDL, and high LDL-P may or may not be safe.
- Eating saturated fat raises serum cholesterol for E4 carriers.
- Therefore, minimizing saturated fat is safer for E4 carriers.

I think of myself as less certain because I am not convinced that low total cholesterol and low calculated LDL are safe. The paper referenced above is pretty compelling evidence. I still do agree with others that high total cholesterol and high calculated LDL may or may not be safe - I would just add, "either". I'm genuinely uncertain, but since we all gotta eat, we all have to choose a strategy. I'm placing my bet on a variety of fats consumed ad libitum. I am not trying to suppress saturated fat because it seems possible that I will do more harm than good by doing so.

On another topic, when I mentioned "non-oxidized serum lipids", I did not intend to convey any connection between unsaturated fat consumption and oxLDL - I was just referencing my understanding that lipid oxidation occurs primarily in high-insulin environments.

[circular]

Such an interesting thread. Marc count me in your uncertain camp. I had moved toward lower coconut oil to help protect against too high LDL-P now that I'm in keto, but yesterday began to wonder if coconut products may provide me with some balance, since I'm not losing my small extra weight in keto, by sending some of my consumed calories straight through the liver for energy, instead of into fat storage (sorry for the run-on sentence).

[Julie G]

As I see it, mainstream thought in these forums runs something like this:
-A lipid profile consisting of low triglycerides, low total cholesterol, and high HDL (and low calculated LDL) is safe. As long as triglycerides are low, high total cholesterol, high calculated LDL, and high LDL-P may or may not be safe.
-Eating saturated fat raises serum cholesterol for E4 carriers.
-Therefore, minimizing saturated fat is safer for E4 carriers.

My take may be a little different. I don't think TC or LDL matter. From my perspective, high is fine, even preferable as long as LDL-P and/or ApoB are below the 50th percentile. I suspect that the increased availability of cholesterol may even help address the reduced lipid-carrying efficiency of ApoE4. It's when particle numbers go above the 50%ile that I get nervous. More particles typically mean smaller particles that tend to be associated with oxidation and insulin resistance. Can someone have an LDL-P of 2,000 and low levels of small LDL-P or sdLDL and low glycemic markers? I haven't seen that. My conclusion: SFA raises LDL-P and ApoB for E4 carriers, therefore minimizing it may be safer for E4 carriers. I fully acknowledge that I may be wrong and I love that you're challenging what you perceive to be our groupthink- never good. Teasing out these nuances is vitally important.

Masterjohn's article was eye-opening in many ways and makes me rethink my relationship with SFA. I'm among the few who can get away with more and still yield good advanced lipids. I'm considering it.

[Tincup]

x Can someone have an LDL-P of 2,000 and low levels of small LDL-P or sdLDL and low glycemic markers?

J - On her first Gundry test - before she adopted his Matrix diet, Theresa had LDL-P 2380, smaill LDL-P of 1382 sdLDL of 34 or 31 (slightly above Gundry's target of 30 mg/dL- depending on the lab). She also had an A1C of 5.0, fasting insulin of 5, fasting glucose of 85 and a EBCT calcium score of 3.8.

Did this improve following Matrix? Yes, LDL-P 1671, smaill LDL-P of 702 sdLDL of 20, A1C of 4.4, fasting insulin of 3 and fasting glucose of 81. Prior to Matrix, she ate a vegan diet. So the improvement had nothing to do with eliminating animal saturated fat. In fact total fat consumption probably went up. Her net carb consumption did drop, she also eliminated grains & legumes she'd eaten as a vegan. She'd added the 20g/day of protein from white fish, shellfish or omega 3/pastured eggs.

A LCHF prescribing & practicing doc I know (E2/2) has an LDL-P of 1700, stellar glucose/insulin metrics (he uses a modified Kraft test on his patients) and has an EPCT calcium score of 0.

[MarcR]

Another data point - in March of 2014 I had LDL-P 1,846, small LDL-P <90, TC 273, LDL-C 189, HDL-C 70, HDL-P 38.6, trigs 72, LP-IR <25, FBG 75, HbA1c 5.1. The LDL-P is just a hair shy of your 2,000 cutoff, and the small LDL-P and glycemic markers are low. Note the concordance between LDL-P and LDL-C - they are within half a quintile of one another.

(I don't have a more recent NMR, but here's my data since that time: July 2014 TC 215, LDL-C 120, HDL-C 86, trigs 47, FBG 78; April 2015 TC 253, LDL-C 170, HDL-C 68, trigs 75, FBG 89, HbA1c 4.6.)