It seems that any time the discussion of fatty livers developing on ketogenic diets comes around, a spate of confusing discussions tend to ensue. One of the confusing issues is that ketogenic reducing diets are effective at reducing hepatic fat levels. However, these studies usually start with a fatty liver and with weight loss comes hepatic fat loss. If there's a study out there where obese people are simply transitioned from a SAD to an equi-caloric ketogenic diet (85+% fat) demonstrating any change in liver fat accumulation, I'm not aware of it. However I believe the evidence of rodents on such diets is at least enough to be concerned that the change, if any, would be for the worse, not the better.
A lot of lean people also think, I can't have a fatty liver because I don't have a pot-belly so I don't have visceral fat. One of the reasons I thought the Eades' 6 Week Cure for the Middle Aged Middle book was such a disaster is that the first two weeks were supposedly devoted to detoxifying your liver. There were several references discussing just that, and a reduction in hepatic fat. I had high hopes for the book at the time, because I was expecting some science on how to specifically reduce visceral fat. I was disappointed that there was little if any discussion devoted to this topic. Let's look at the human abdomen. For starters, the liver is rather assymetrical. Additionally, although obesity leads to a higher prevalence of fatty liver, lean T1 diabetics have a rather high incidence of the disease. Roughly a quarter of those with fatty liver disease (FLD) have normal sized livers, so accumulation of hepatic fat does not necessarily increase the size of the liver. A normal adult liver has a span of 7-10 cm and an enlarged liver is over 2-3 cm larger. Still, this does not seem to be sufficient to cause a large belly. Also note the location of the liver. Most "bellies" are considerably lower.
Let's look at visceral fat depots. That mauve thing in the diagram on the right there is the liver. The "pot belly", as modeled by the fella on the left, is mostly omental visceral fat, and/or the front retroperitoneal visceral fat.
While the woman at left here is likely packing some visceral fat, that "spare tire" roll around the middle is evidence of considerable subcutaneous fat. While the guy above could probably squeeze some fat between a set of calipers, his tummy is likely relatively hard to the touch. In any case, if either had an enlarged liver, it would be the proverbial "drop in the bucket" of overall fat mass.
I think this is a very important distinction to be made in discussions of insulin resistance as IR is associated with both hepatic triglyceride content and visceral fat accumulation. Note I said associated as I'm not implying nor attempting to discuss causality here.
In my recent post, I discussed two studies implicating a ketogenic diet producing a fatty liver in mice, I also discussed these previously. In the second post, there is a table from the study showing that the triglyceride level of the keto-mice was higher than even the high-fat chow fed mice, despite growth and body mass similar to that of calorie restricted mice. In the second protocol from the same study, the results -- where some high fat fed mice were switched to keto diets and lost body mass to equal normal (low fat, ~12% usually) chow fed mice. However:
Paradoxically, the increase in glucose tolerance and correction of lipid profiles occurred in the context of increased hepatic triglyceride content compared with both HF and C groups
Thus, since many employ ketogenic diets for weight loss, from a metabolic profile similar to the HFD mice, if not worse*, the results of this study seem pertinent at least as cause for concern. Since most folks eating a "ketogenic" diet are eating nowhere near the 90+% fat diet, the cause for concern is somewhat mitigated ... but perhaps if you are one of those humans with the "mighty metabolism" exhibited for these mice (10 &15% higher energy expenditure than standard chow and calorie restricted chow respectively), in other words you truly are eating more and losing or not gaining or whatever, then perhaps you might want to get your liver tested. Elevated hepatic triglycerides occurred during growth of mice on KD and weight loss (then stabilized) after they were made obese on HFD. It may be that during the loss phase, liver triglycerides were reduced, but they weren't after weight stabilized.
But ... but ... my belly got flatter! (My own personal experience to the contrary aside ...) Here is where this blurring of the distinctions becomes important. Chances are if you gain or lose fat in your liver, we are not talking about a noticeable difference even if it is enlarged ... and it may well not be enlarged despite increased triglyceride content. You may have lost a lot of visceral fat around your organs ... or just subQ fat outside of the abdominal wall ... but the two are not one and the same. Therefore visceral or subcutaneous "slimness" is no reliable measure of your liver fat. In addition, you are likely to "feel great" for a very long time! The prevalence of some degree of FLD in Americans is shockingly high and it doesn't interfere with liver function for a good long time, certainly to any discernible degree. You may be looking at several years or even decades before it progresses to a deleterious state. Food for thought ...
* Coming from the SAD, being equivalent to the CAF diet, which was far worse than the standard rodent HFD, matters could well be worse for us humans.