Synthesis of Fat in the Liver

One of my hobby horses is the idea that you have to go through the glycogen stores in your liver before your body switches into fat burning mode and that this is one of the reasons cardio-style exercise is so ineffective for fat loss on a conventional low-fat diet. The opposite state, where the liver is completely full of glycogen, is also an interesting case. In this situation the liver starts manufacturing fatty acids from glucose. This is called de novo lipogenesis in the biology vocabulary. If you break this down from Latin to English, it is "generation of new fat."

There are three main stores of fat in the body: subcutaneous (under the skin), interstitial (in-between muscle fibres), and visceral (in and around the vital organs in the belly). Of the three types, visceral fat is dangerous to health while the others are relatively benign (Porter et al., 2009).

Visceral fat, typically measured by waist-to-hip circumference ratio, or more advanced imaging techniques, is a much better predictor of future diabetes or heart disease risk than the body mass index (Westphal, 2008). For example, diabetics who are relatively thin (i.e. have a low BMI) very often have what's called central obesity (Ruderman et al., 1998) or more colloquially, are "skinny fat." Fat tissue, as it happens is efficient at producing a wide variety of hormones such as adiponectin, leptin (e.g. Angulo et al., 2004), and resistin. For want of a better explanation, packing a lot of hormone-producing fat around the vital organs is bad juju.

I pose a couple of questions for the reader to ponder:

Why is visceral (belly) fat so contrary to good health? and,

What is it in our modern diet that is driving such an excess of visceral fat?

A distinct condition whereby fat deposits around the liver cause it to dysfunction is non-alcoholic fatty liver disease (see the New England Journal of Medicine review by Angulo (2002). As the name suggests, it is characterized by the appearance of fatty deposits in the liver tissue itself. Think fois gras. In addition to the formation of fat deposits, some of the more advanced forms of chronic liver disease feature the formation of fibrosis, which is the formation of scar tissue in the liver in response to repeated, chronic injury.

Adapted from Figure 1 in (Bataller and Brenner, 2005).

Funnily enough, this condition is tightly correlated with metabolic syndrome which is in turn associated with diabetes and many other debilitating conditions. Loria et al., 2005 (free) state that:

Given that metabolic syndrome and non-alcoholic fatty liver disease affect the same insulin-resistant patients, not unexpectedly, there are amazing similarities between metabolic syndrome and non-alcoholic fatty liver disease in terms of prevalence, pathogenesis, clinical features and outcome.

Loria does state that fatty liver disease does not cause metabolic syndrome, or vice versa. Since metabolic syndrome is a catch-all description of many symptoms, I think it would be fair to describe fatty liver disease as one potential component of metabolic syndrome.

The problem with fatty liver disease really appears to be the combination of insulin resistance (from ingesting too much glucose) and high circulating triglyceride levels. From the review by Petta et al., 2009, "In fact IR [RM: insulin resistance] is the key factor in the promotion of liver fat accumulation not only by inducing an increase of liver FFA [RM: free-fatty acid] influx, but also, via hyperinsulinemia, by stimulating the activity of enzymes implicated in de novo hepatic lipogenesis."

Incidentally non-alcoholic fatty liver disease was almost certainly what Morgan Spurlock was doing to himself with his soda-laden diet in the movie Super Size Me. I noticed when watching that movie that some of his doctors (2 of 3, IIRC) didn't know that the condition existed. Non-alcoholic fatty liver disease reached incidence levels of 20-25 % in an Italian population study (Bedogni et al., 2005).

Ok, so abdominal/visceral fat causes some combination of metabolic syndrome and/or fatty liver disease. So what causes people to preferentially deposit fat around their mid-section rather than elsewhere? In researching non-alcoholic fatty liver disease, I came across the following paragraph by Postic and Girard (2008, free access), which I think is instructive:

Insulin is essential for the maintenance of carbohydrate and lipid homeostasis. Insulin is secreted by pancreatic β cells in response to increased circulating levels of glucose after a meal. A large fraction of glucose absorbed from the small intestine is immediately taken up by hepatocytes [RM: liver cells], which convert it into glycogen. However, when the liver is saturated with glycogen (roughly 5% of liver mass), any additional glucose taken up by hepatocytes is shunted into pathways leading to synthesis of fatty acids, which will be esterified into TG [RM: triglycerides] to be exported to adipose tissue as very low-density lipoproteins (VLDLs). Insulin inhibits lipolysis [RM: fat burning] in adipose tissue by inhibiting hormone-sensitive lipase (HSL), the enzyme regulating FFA [free-fatty acid] release from adipose tissue (10). Therefore, from a whole-body perspective, insulin has a “fat-sparing” effect by driving most cells to preferentially oxidize carbohydrates instead of fatty acids for energy. Insulin also regulates glucose homeostasis at many sites, reducing hepatic glucose production (HGP) (via decreased glucose biosynthesis [gluconeogenesis] and glycogen breakdown [glycogenolysis]) and increasing the rate of glucose uptake, primarily into skeletal muscle and adipose tissue.

A very interesting review that hypothesized on a link between diabetes and fructose said the following (Johnson et al., 2009):

For example, very high doses of fructose (250 g/d x 7 d) cause insulin resistance in 1 wk (147), whereas slightly lower doses (216 g/d for 4 wk) only induce insulin resistance at sites where fructokinase is highly expressed (liver and adipocyte) (148), and even lower doses (100 g/d x 4 wk) result in no insulin resistance at all (149).

If you read through any significant amount of human biology on diet it's impossible to avoid the fact that the hormone system (and insulin and growth hormone in particular) is paramount in determining whether the body is in a state of fat gain or fat loss. It's only at the nutritional level that the facts become obscured by experimenting with too many variables at once.

If you will permit me an aside, most all of our actual information about diet and nutrition comes not from the 'top-down' approach of observation or intervention trials but from the 'bottom-up' approach of trying to establish the mechanics of human physiology. I like to call the 'bottom-up' approach the 'physicsification' of biology. Most properly the 'bottom-up' approach in biology can be described as the combination of biophysics, biochemistry, and genetics (bio-computer science).

In physics, one establishes base laws that govern a system, known as first principles, and then one gradually expands on the complexity until theory adequately matches experiment. Technically any other science can be described in terms of physics, but often we are stymied by excessive computational requirements or too many unknown, confounding factors. However, gradually scientists are slowly unraveling the secrets of biology.

The main advantage of having first principles is that it allows you to construct hypotheses that are likely true, and then test them. There are a lot of famous and successful predictions in physics. Observational nutritional science, not so much. For example, Einstein's general relativity predicted that light would bend (or 'lens') around strong gravitational objects like black holes; it does.

Now, back to the topic as to what drives visceral fat accumulation...

One potential source for abdominal fat is fats produced in the liver itself, most commonly by the conversion of carbohydrates to fat. Typically the total contribution of liver-synthesized triglycerides (de novo lipogenesis) to the total number of triglycerides in the blood stream (i.e. VLDL) is relatively small, on the order of 10 % (Marques-Lopes et al., 2001). This is too small a proportion to seriously be considered as a cause of obesity.

However, if you recall from the paragraph I quoted above, the liver only really starts to kick out a lot of lipids when you exceed its capacity for storing glycogen. A study by McDevitt et al. (2001, free access) specifically looked into the case of overfeeding versus not and what effect it had on fat synthesis in the liver. They found that with overfeeding by 50 % over basal metabolic rates, de novo lipogenesis increased 2-3 fold. Overfeeding on sugar (glucose-fructose) was uniformly worse than overfeeding on glucose, but only slightly.

A study of rats fed a diet of 60 % fructose versus conventional rat chow (Ackerman et al., 2005). After five weeks, the fructose-fed rats had 15 % higher blood pressure, 198 % higher blood triglycerides, and 90 % higher blood cholesterol levels. A similar study in overweight women found similar results: when fed 25 % of calories in the form of fructose for ten weeks resulted in a 140 % increase in circulating triglyceride levels (Stanhope and Havel, 2008). These rates of sugar consumption are consistent with soda pop intake for a significant hunk of the American populace.

One question is, why does fructose (and alcohol) intake result in visceral fat, and not the more benign sub-cutaneous or intra-muscular fat? I have one possible explanation that I like to term the 'circulatory fat deposition model.' When you ingest a toxin like fructose or alcohol, the body automatically increases circulation to the vital organs (and in particular the liver) so that it can be filtered out of the blood stream. Since any ingested substance will naturally diffuse to even concentration throughout the blood, this is the only way to preferentially increase the flux of toxin to the liver.

Fructose is well known to contribute greatly to post-meal triglyceride levels (Chong et al., 2007). The liver takes fructose and produces palmatic acid (i.e. a stable saturated fat) from it. It then releases that fat into the blood stream. Since the filtering of fructose isn't instant, the circulation in the body core is still heightened. As a result, the visceral fat tissues see a higher rate of triglyceride flux than the more benign skin or muscle fat (Note: flux in a scientific sense typically means mass or volume per second — put those Star Trek thoughts out of your mind). The visceral fat, which sees the most fabricated triglycerides floating on by, also happens to absorb the most. Hence fructose tends to promote visceral fat. On the other hand, if you ingest excess calories in the form of fat, it's not any more likely to deposit around the liver than it is your thighs, so it's not nearly so dangerous.

One sees a similar effect with amateur body-builders who ingest calorie-heavy shakes and energy drinks after or during exercise where their muscles are generating a lot of lactic acid. The body increases blood flow to those muscles to remove the lactic acid, but the fat deposits inside the muscle also see a much higher flux of fat and fat-building substrate as a result. This results in a characteristic thick and pasty muscle texture without a lot of functional power. Think of well-marbled beef steak.

If this hypothesis is true then combining dietary fat with any chemical that requires extensive liver processing (e.g. caffeine, artificial sweeteners) would also tend to result in visceral fat deposition. Oh look, a prediction. I did say something about those.

This information on de novo lipogenesis, and what we know of the fat-sparing properties of insulin, provides some support to the notion that carbohydrates and fats should not be mixed in meals. It's only when you eat an excess of glucose, or any fructose, that one can transform a pure carbohydrate meal into body fat. On the other hand, if you eat fats and carbohydrates in combination, the insulin response will prevent your body from burning the fat directly. Note that if you have a dysfunctional carbohydrate metabolism (i.e. metabolic syndrome) this precept probably does not apply. Of course this advice is only useful if you are capable of restricting your caloric intake on a pure carbohydrate diet.

Fats are satiating whereas carbohydrates most definitely are not. The hormonal reason for this is related to the fact that they each use a different mechanism for regulation. With insulin, as it ramps down, it promotes the production of ghrelin, one of the primary 'appetite' hormones. Fat metabolism doesn't appear to have a similar analogue, and as a result hunger on a high-fat diet lacks the ravenous component of the insulin roller coaster.

When you think about, there's plenty of reason to believe that carbohydrates promote over-eating. By in large, most of the plant carbohydrate sources our paleolithic ancestors would have access to all mature around the same time, late summer and fall. This is a time period when it is particularly advantageous for primitive man to pack on some fat to sustain him over the winter. On the other hand, for Joe 6-Pack with his year-round supermarket access, this doesn't work out so well.

The conclusions we can draw from this body of research are that one can safely ingest glucose regularly with the aim of not saturating the liver's glycogen storage capacity. The maximum reasonable glucose intake level will vary significantly from person to person depending on general activity level and overall health based on how insulin resistant they are. Where one gets into trouble is when you overfill your liver by eating too many calories, with a significant fraction of glucose calories, or significant fructose intake (likely in the form of sugar or corn syrup). This is likely to lead insulin resistance and liver dysfunction.