This is the paper for continued speculation about alcoholic and non alcoholic fatty liver disease. Rats again, to begin with!
I was looking at the endotoxin levels (from Table 3) in the blood of rats on various feeding protocols. The correct level of endotoxin in your blood is zero. Endotoxin (gram negative bacterial wall components) belongs in your gut, not in your bloodstream (if endotoxin is in your blood stream one function of LDL cholesterol to mop it up. Hmmm, rosuvastatin triggers diabetes, possibly termed hepatic failure to respond to insulin correctly! Is there a link here?). It's a reasonable marker of increased intestinal permeability and a serious toxin in its own right. Ethanol increases endotoxin blood level markedly in combination with fish oil but only moderately if combined with saturated fat. So, apart from ethanol, fish oil has some influence on the increased intestinal permeability induced by alcohol. That's not too surprising if you read this paper about the effects of DHA in its own right. Now I have no idea of what 100 microM DHA means in terms of drinking a slug of fish oil or how real this cell culture system is in terms of the human gut but I can see that people going the grass fed meat route have less to be concerned about that those of us using fish oil as a supplement. And I'm sure that Christian will love the taurine link! I would suspect that the best time to take DHA might be in association with your main meal of the day if you are looking to "mimic" grass fed meat...
Endotoxin in the blood appears to be one of the best agents to convert a fatty liver to an inflamed fatty liver. The switch from fish oil to saturated fat lowers endotoxin somewhat (about halves it) but markedly reduces the products of the genes controlled by NF-kappaB. I was particularly looking at Table 4 where mRNA for COX-2, the inducible pro inflammatory cyclo-oxygenase enzyme, is reduced to zero and that for TNFalpha is markedly reduced. The mRNA for the housekeeping enzyme COX-1 is unaffected, as you would expect. It's also worth noting that fish oil/dextrose produces a zero level of mRNA for COX-2, rehabilitation fish oil somewhat in the absence of ethanol (or fructose?).
Staying on endotoxin but switching to humans and fructose, we are all well aware that fructose is "associated" with fatty liver and that this association is probably causal.
What I find far more interesting is that fructose is "associated" with increased plasma endotoxin in humans. I haven't found the intervention study to confirm a causal role of fructose on endotoxin uptake but, with the known effects of fructose on clinical NAFLD in humans, I think it will turn out to be the case. Fructose does seem to be the perfect replacement for alcohol if you want a tea-total cirrhotic liver. And be labelled a secret drinker by your hepatologist!
To summarise, both alcohol and fructose cause fatty liver. Both alcohol and fructose allow endotoxin from the gut to the bloodstream. PUFA (certainly omega 3, probably omega 6) enhance intestinal permeability effects. Endotoxin and the lipid peroxides from PUFA activate NF-kappaB. There is a cascade of inflammation in the liver as a consequence of this.
I think it is also worth noting (from table 3 again) that non haem iron is elevated in the livers of those rats with maximum endotoxin absorption. This is another aspect of the common hepatopathy of iron overload which needs thinking about.