OK, this is the direct link to figure 1 of Krauss' commentary paper. I just want to run through what seems to be happening and some of the consequences. Best to have the picture open alongside Hyperlipid.
Top left is the nucleus, then there's the stack of endoplasmic reticulum, then a newly synthesised lipid particle stuck on the outer surface of the ER. There is an interesting initial particle labelled I receiving "lipids" (another post there) which goes on to become that unlabelled particle on the outer surface of the ER (call it a nascent lipoprotein). The nascent lipoprotein has two arrows leading away from it representing two metabolic options. The upper arrow goes to a particle labeled III. On this pathway the particle receives saturated fatty acids which, at step V, stop it being destroyed. Destruction is termed PERPP, and the line from PERPP has a flat end meaning its blocked, and the legend "+SFAs" is shown as doing the blocking (by the little curved arrow). Got it? This SFA loaded particle is exported as an IDL (bad) or a large LDL (good). Luckily the IDL appears, on this diagram, to convert to the large form of LDL. But IDL may only be bad if secreted post prandially, I don't know, life is so complex in the ad hoc world of the lipid hypothesis!
Summary so far: Basic particle plus SFA gives the "good" version of "bad" cholesterol. Don't you just love these terms.
Now go back to that nascent lipoprotein and follow the alternative pathway shown by the downwards arrow. This leads to particle labeled IV, having accepted a lipid droplet via the curved arrow. Remember the lipid droplet. It might matter later. Anyway, there is then a dashed arrow showing possible secretion of this particle. If secreted it becomes a large VLDL particle.
While large LDL particles are good guys, large VLDLs are not. In fact they are the precursor to the evil incarnate particle, the small dense LDL. Like Darth Vader, only without the pre death conversion to the good side. Secrete and die.
Fish oils to the rescue! There is an arrow from the PERPP "destructaparticle" system pointing straight to this evil particle, joined by a little "n-3 PUFA-ox" which destroys the evil large VLDL before it ever gets secreted. Less of the large VLDL means less of the small dense LDL. More joy on blood results. Not only do triglycerides drop, it's the evil fraction of triglycerides which drop.
To summarise this pathway: Basic particle plus lipid droplet gives you bad VLDLs and small dense LDLs, unless aborted by omega 3 lipoxidation product (malondialdehyde is claimed but I'll get on to signaling molecules eventually).
All of that is pretty straight forward but it doesn't give us any insight in to anything except how to improve lab numbers.
Here are the nitty gritty questions.
1. Where did the SFAs come from? This is easy. Just follow Krauss' refs and you will find it's diet. Another post there needs writing.
2. Where did the lipid droplet come from? Well, lipid droplets in the liver vary from normal physiological amounts through to hepatic lipidosis. Hepatic lipidosis is a routine feature of the metabolic syndrome. We know that fructose is converted to lipid as rapidly as possible in the liver. We know that insulin inhibits the release of all VLDLs from the liver. Combining fructose with a glucose source (to raise insulin) seems a good way to generate hepatic lipid and block it's release. The bigger the dose and particularly the more continuous the ingestion, the more lipid droplets are likely to form. Sucrose or HFCS would do the job nicely. So might alcohol. Alcohol is interesting. Low doses improve insulin sensitivity and high doses do the opposite. The histpathology of non alcoholic hepatic steatosis is indistinguishable from alcoholic steatosis. They are the same condition. Keeping those lipid droplets in your liver seems a good way to get hepatic lipidosis and subsequent cirrhosis.
3. Where are the lipid droplets going? Well, if you dose up on fish oils the answer is nowhere! Lipid droplets should be off-loaded as small dense LDL precursor particles, the large VLDLs. You're not going to release them if you're on high dose fish oils! So you are trading the drop in "bad" LDLs for a rise in hepatic lipidosis. Are you going to mangle your liver to make Krauss happy? Yes?
Now let's go back and look at the high dose fish oil study from back in 1991.
Here are the triglyceride effects of 30ml fish oil a day for three weeks. Without vitamin E some omega 3s still get to the liver and trigs (probably large VLDLs in this case) drop from 2.6 to 2.0mg/dl. Modest and not statistically significant. Have a wash out period and do it again, but this time preserve the omega 3s with vitamin E. The drop is 48% this time, p<0.01, enough to warm the cockles of a cardiologist's heart.
Remember, these absent triglycerides should have contained lipid droplets which the liver wants rid of. I find it fascinating to see that, in the post washout/pre omega 3+vitamin E session, that the trigs were up at 3.4mg/dl, quite a bit higher than the 2.6 at the start of the study. Is this the liver off loading lipid droplets retained during the first section of the study? With trigs down at 1.8mg/dl by the end of the high dose vitamin E section, how much hepatic steatosis is going on?
Now look at this table, especially insulin and glucose. You'll have to click to enlarge.
All of these FBG values are scarily high, so these volunteers are on the edge of diabetes. What happens when you load up on fish oils? Insulin: No significant change at any time. FBG; both fish oil sessions show increased FBG! For low dose vitamin E section the change had p<0.05, for the vitamin E protected phase it had p<0.001.
That smells of insulin resistance to me and hepatic lipid overload is the easiest explanation.
Aha! So the Greenland eskimo, who refuse to die of cancer on 15g/d of EPA+DHA, must all have been dropping like flies from hepatic cirrhosis. Or type 2 diabetes. Apparently not. This dose of fish oil appears to be fine, just so long as you are not making lipid droplets in the first place. That means no sugar and no excessive alcohol. Remember modest dose alcohol improves insulin sensitivity, so zero alcohol is not needed. The Greenland eskimo were VERY low carb.
So where does that leave fish oil supplements? I think if you have a problem with alcohol they are very bad news and you should be absolutely minimising all forms of PUFA, unless you really want cirrhosis. If you are eating to the ADA or AHA sucrose ladened guidelines and already have "mysterious" raised liver enzymes, you will make an already appalling job worse.
If you are LC and low PUFA in the first place, or even just eating a diet which doesn't generate hepatic lipidosis (minimal sucrose), I think there are advantages to modest dose fish oils for long term changes in insulin sensitivity (another post needed). Dropping your triglycerides in the short term is not one of them, unless you are in to treating numbers, in which case; come back torcetrapib, all is forgiven!
PS Let's just clarify: There is no "good" or "bad" bad cholesterol. You can fuel your metabolism with saturated fat, and the "good" bad cholesterol goes up as a marker. It's the saturated fat which is really good. Or you fuel your metabolism on sucrose, which raises "bad" bad cholesterol. Stuff the cholesterol. It's a marker you are being evil to your metabolism by eating sucrose, which is what does the damage.