This video is complicated so I'll have to rewatch it to understand the details. But the implications are clear. This video is super. Thanks Dr. Greger!

Now I get it.

I still don't quite get it. I see why cholesterol and saturated fat don't correlate, because people naturally have different starting points of cholesterol. If only 4% of the population have hypercholestemia, why doesn't saturated fat correlate with heart disease?

I think what it means is that not everyone who eats saturated fat gets heart disease because enough people have naturally low cholesterol so that eating meat and dairy may not affect them in that way. Even though it's the number one killer?

Or is it that pretty much everyone in this country has heart disease???

I think the point is that when you are testing for significant correlation you are comparing the pattern with the noise to determine significance of the pattern. When the pattern is large compared with the noise, we say the correlation is significant. When the noise happens to be large, then it is much easier for the pattern to hide.

Add to this the fact that "normal" people are generally studied and that our population tends to have fairly homogenous behavior, the pattern gets hidden even better.

This good explanation and the excellent video itself get to the truth of the matter -- but how to get the average person to understand? The message that many people have solidified in their own minds recently is that there is no correlation between dietary saturated fat and heart disease. Therefore saturated fat intake doesn't really matter -- so why limit it if it improves the taste of food? The mainstream media has done their part to make a real mess of nutritional science, but why do editors of scientific journals allow for the publication of such deceptive studies?

That's because no real scientific study has proven it. Saturated fat from good sources are not unhealthy. Now if you're pouring 2 table spoons of coconut oil then eating red meat every meal then that is likely cause for concern. The studies used to argue against saturated fat are just like this one http://www.ncbi.nlm.nih.gov... so if you believe those than you believe this one that says vegetarian doesn't help all cause mortality. Just like not all carbs are equal, not all fat is equal. The world isn't black and white

Andrew: if you're serious about learning about the connection between SF and heart disease, I recommend reading some of Travis' posts and meta-analyses at the link below. He lays bare the few 'studies' which, due to massive methodological flaws, erroneously aren't able to find a link between SF and heart disease.

http://healthylongevity.blo...

Here's what has been proven:

- Increasing SF consumption increases LDL cholesterol. (see meta-analysis of 365 metabolic ward studies).

https://www.ncbi.nlm.nih.go...

- Increases in LDL cholesterol are linearly related to increases in the development of atherosclerosis (heart disease), see link below.

http://circ.ahajournals.org...

So increasing SF from below 10% of calories -> increases LDL to above 75 mg/dL -> causes atherosclerosis in the average person. The link has been proven. Studies funded by the American Egg Board and the dairy industry have tried to cast doubt on the mountain of evidence in support of the diet-heart connection. Don't let dollars beat science.

The link you provided for the study "Increasing SF consumption increases LDL cholesterol" no longer pulls up the study. I've been trying to find a number of studies about how health is adversely affected by dairy on NCBI and not a whole lot is pulling up for me. Are the studies being systematically removed?

Nicole: all 3 are working well for me at the moment. Maybe try again after clearing your internet cache?

That a statin study by definition, there were more people in group B taking statins so they had lower cholesterol than Group A in your referenced study. Statins have an effect of decreasing inflamation so the study becomes murky when trying to prove LDL reduction is the driver.

Mark: re your comment on statins, if you actually read the study, you'd find this section:

"When only those patients who were on statins were analyzed (n=49), there was still a significant positive linear relation between percent annual changes in P&M CSA versus LDL cholesterol (r0.45, P0.001) and a negative linear relation between annual changes in P&M CSA versus HDL cholesterol (r0.30, P0.05). An LDL value of 72.5 mg/dL was the cutoff at which regression analysis predicted no average annual P&M CSA increase."

In simple terms, even when you only looked at those on statins, the effect was still demonstrated and statistically significant. LDL above 75 is simply equivalent to heart disease (progression of atherosclerosis) in the average person.

"When the noise happens to be large, then it is much easier for the pattern to hide."

This is technically correct, and would be an issue if the cross-sectional studies were underpowered (ie their sample sizes were too small). But that does not appear to be the case, which means if a signal existed, it would have been through the noise.

That these studies were not underpowered and no signal was detected actually indicates the effect doesn't exist (in this case, between subjects).

Of course, the longitudinal studies show an effect does exist within-subjects.

Animal fats are ALWAYS going to be bad for the heart even in otherwise healthy people. Just because someone doesn't develop heart disease does not mean their hearts aren't suffering for their diet. Animal products are also detrimental for virtually every other single function of the human body. In no way does this suggest that animal saturated fat is ok for some people. It's not ok for anyone, or at lest any human or other herbivorous animal.

This sounds like a very biased comment. You apply to the religion that humans are only meant to eat plants? When there's nothing from reputable sources to back that up. It's easy to look at the world black and white. That's what extremists like extreme vegans and extreme panels do. Try to look at it more from an understanding of all angles. Learn to take value from everywhere. That doesn't mean you have to believe it or align to it. But rather than closing one ear to one side you open both. Or else you may just miss out from hearing something valuable.

Andrew: don't worry, not all vegans believe humans are herbivores. Greger himself has said that there's no evidence of harm from very occasional meat consumption. You may be interested to read this blog from a biologist and vegan stating that we're technically omnivorous. Personally, I don't see any reason to care what our ancestors ate: it's about finding what diet today makes us healthiest. The balance of evidence indicates that a whole-foods plant based diet, supplemented with B12, is the reigning champ right now.

https://veganbiologist.com/...

Remember that statistics are about averages and deviations from averages. Things that are too rare, for example a whole-plant based diet, wash out in the noise as 'outliers'. Within the normal bounds of US society: i.e. between 1 and 3 hamburgers of saturated fat a day, cholesterol is > 150 and as such heart-decease progression is random and a function of genetics and affluence (i.e. can afford a doctor to give you statins?).

Dr Greger, please respond to this issue. I am confused as well. Why don't the recent studies have power to detect association with CVD?

Here it is: re-quoting Walter Willett, chair of the Department of Nutrition at Harvard School of Public Health, here is the problem:

This paper is bound to cause confusion. A central issue is what replaces saturated fat if someone reduces the amount of saturated fat in their diet. If it is replaced with refined starch or sugar, which are the largest sources of calories in the U.S. diet, then the risk of heart disease remains the same. However, if saturated fat is replaced with polyunsaturated fat or monounsaturated fat in the form of olive oil, nuts and probably other plant oils, we have much evidence that risk will be reduced.

And the next line in that same article is:

"Dr. Willett emphasized that because this meta-analysis contains multiple serious errors and omissions, the study conclusions are misleading and should be disregarded."

The art of reading… such an undervalued skill…

If you had read the transcript you would have known that.

So… you post a quote out of context and I catch you red handed on trying to mislead and then reply with this?

Wow! You seriously have no self respect.

You can sure dish it out can't you? You don't make any sense to me.

It would be great to see a respectful exchange of ideas. We don't need to take a "side." But together we can help each other sift through the information.

Even plant oils that are saturated, such as virgin coconut oil?

I think avoiding sat fat is a good idea due to the link with insulin resistance but I am unconvinced by the cholesterol causes heart disease argument. Can anyone show me evidence that cholesterol is the main driver of heart disease that does not involve a statin study

Mark: we have strong evidence that higher LDL directly causes atherosclerosis.

http://www.onlinejacc.org/c...
http://circ.ahajournals.org...

Although some patients in the second study were indeed on statins as you've pointed out elsewhere, if you actually read the study, you'd find this section:

"When only those patients who were on statins were analyzed (n=49), there was still a significant positive linear relation between percent annual changes in P&M CSA versus LDL cholesterol (r0.45, P0.001) and a negative linear relation between annual changes in P&M CSA versus HDL cholesterol (r0.30, P0.05). An LDL value of 72.5 mg/dL was the cutoff at which regression analysis predicted no average annual P&M CSA increase."

In simple terms, even when you only looked at those on statins, the effect was still demonstrated and statistically significant. LDL above 75 is simply equivalent to heart disease (progression of atherosclerosis) in the average person.

Jeff

You are citing evidence from Statin trials. You simply cannot cite statin trials and lower LDL with reduction in heart disease. Statins are known to lower inflamation, they are not the best way to do this but they do lower it. What this means is that the lower LDL that comes with it may well be incidental. Evidence proves that people with higher LDL live longer. Also the first study boldly predicts that LDL is a main contributer in a multi factored disease. This is clearly not true, if it was a major player or the the player as many pro cholesterol folks would have it, we would see clear population relationships and and easy it easy to find many contradictions.
France and Russia have equal overall LDL levels but the CVD risk is chalk and cheese.

Any evidence has to be statin free to have any weight.

Did you really mean to say this "Evidence proves that people with higher LDL live longer" ?

Yes I did

Mark: are you aware of any studies that directly measure progression of atherosclerosis and do not involve any subjects on statins? If you are, please send links, I'd be interested to read. If there aren't any, then we must conclude that the best evidence we have to date indicates that LDL over 75 directly leads to atherosclerosis in the average person. Better studies are always nice to think about, but in the meantime we have to make decisions with the best quality evidence. As far as population-level trends, these provide hints but they simply pale in comparison to the strength of RCT's and direct measurement studies which strongly support the SF->LDL->CVD link. Epidemiological evidence is simply not strong enough to contend with RCT's and direct measurement studies.

Regarding mortality of seniors and its association with LDL, please use the link below to read the section of the study entitled "Is high LDL-C beneficial?". You'll see there that they don't dispute or address the connection to CVD, but rather state that it may confer protection to seniors (over 60) and by that method, the mortality effects of the CVD reduction are cancelled out by mortality impacts of other causes of death. Since CVD is a lifelong progressive disease, it would seem prudent to keep one's LDL below 75 until one's 60's, and then allow it to go higher if you have no warning signs of CVD, to allow the LDL to confer its apparent protection from other causes of death.

The authors themselves also noted that "Some of the participants in the study with high cholesterol may have started statins during the study, and therefore their high life expectancy could be due to them being on statins. Similarly, some of them may have started a healthy diet during the study, and this could have increased their life expectancy."

http://bmjopen.bmj.com/cont...

The video shows Vegetarians with Total to HDL ratio of 4.5, I was surprised by this as this is not a particularly good ratio. Total to HDL ratio has been touted as far more predictive than simple LDL. My LDL is around 3.0 mmol which would have Dr Greger alarmed and most doctors reaching for the Statins but my HDL is 1.8 giving me a ratio of around 2.75 which is why getting people worried about LDL readings is really unneccesary

Mark: the 3.0 mmol / L is equivalent to 116 mg/dl. According to this study (below), it is expected that someone with this LDL is currently adding to the plaque in their arteries and developing heart disease with every day that goes by. Just my two cents, but I would make efforts to lower my LDL if I were in this position. For a comparison point (genetic differences notwithstanding), since going to a plant-based diet with lots of exercise, my LDL is 55 (1.42) and HDL is (1.55). It can be done!

http://www.onlinejacc.org/c...
http://circ.ahajournals.org...

Best of luck.

Jeff, are you eating any fat, nuts, allowed oils, ?

April: yes, I have a few TBS of oil per week when cooking, flax every day, peanut butter every day (the processed stuff, gasp!), and tofu, those are my main fat sources. This is not a suggested regimen, just what I do right now :)

I still don't quite get it. I see why cholesterol and saturated fat don't correlate, because people naturally have different starting points of cholesterol. If only 4% of the population have hypercholestemia, why doesn't saturated fat correlate with heart disease?

I think what it means is that not everyone who eats saturated fat gets heart disease because enough people have naturally low cholesterol so that eating meat and dairy may not affect them in that way. Even though it's the number one killer?

Or is it that pretty much everyone in this country has heart disease???

I think the point is that when you are testing for significant correlation you are comparing the pattern with the noise to determine significance of the pattern. When the pattern is large compared with the noise, we say the correlation is significant. When the noise happens to be large, then it is much easier for the pattern to hide.

Add to this the fact that "normal" people are generally studied and that our population tends to have fairly homogenous behavior, the pattern gets hidden even better.

This good explanation and the excellent video itself get to the truth of the matter -- but how to get the average person to understand? The message that many people have solidified in their own minds recently is that there is no correlation between dietary saturated fat and heart disease. Therefore saturated fat intake doesn't really matter -- so why limit it if it improves the taste of food? The mainstream media has done their part to make a real mess of nutritional science, but why do editors of scientific journals allow for the publication of such deceptive studies?

Remember that statistics are about averages and deviations from averages. Things that are too rare, for example a whole-plant based diet, wash out in the noise as 'outliers'. Within the normal bounds of US society: i.e. between 1 and 3 hamburgers of saturated fat a day, cholesterol is > 150 and as such heart-decease progression is random and a function of genetics and affluence (i.e. can afford a doctor to give you statins?).

Here it is: re-quoting Walter Willett, chair of the Department of Nutrition at Harvard School of Public Health, here is the problem:

This paper is bound to cause confusion. A central issue is what replaces saturated fat if someone reduces the amount of saturated fat in their diet. If it is replaced with refined starch or sugar, which are the largest sources of calories in the U.S. diet, then the risk of heart disease remains the same. However, if saturated fat is replaced with polyunsaturated fat or monounsaturated fat in the form of olive oil, nuts and probably other plant oils, we have much evidence that risk will be reduced.

If you had read the transcript you would have known that.

You can sure dish it out can't you? You don't make any sense to me.

It would be great to see a respectful exchange of ideas. We don't need to take a "side." But together we can help each other sift through the information.