The causality of insulin resistance

There seem to be two large somewhat competing hypotheses trying to explain the causality of insulin resistance (as measured at a whole body level). The lipotoxicity hypothesis, explains to us how insulin sensitivity is reduced in tissues when too much fat builds up in the specific tissue cells; likely caused by high serum levels of free fatty acids and triacylglycerols. The other hypothesis is the inflammation hypothesis, which seeks to explain reductions in insulin sensitivity by high levels of inflammation, possibly caused by stress in general, endoplasmatic stress or dietary fatty acid composition and more.

Both inflammation and cellular lipid overload correlate with insulin resistance and the metabolic syndrome and looking at the literature it seems that both mechanisms have to be a part of a causal chain.

The big unanswered question here is the direction of causality. Overweight, atherogenic dyslipidemia, inflammation and other factors appear in concert with insulin resistance. But what comes first? Does obesity cause insulin resistance or is there some underlying factor causing both insulin resistance and obesity? And which tissues are the first to become insulin resistant? 
Clever scientists have succeeded in creating animal models that are extremely good at storing energy as fat. The funny thing is that this ability seems to protect against insulin resistance. The consequence of this research is the notion that if you are really good at getting fat, you are protected from insulin resistance. The existence of lean insulin resistant individuals would support this notion and would also cast doubt on overweight causing insulin resistance.

There is more to support the above mentioned. It has been proposed that insulin resistance develops because of an imbalance of fat distribution between the tissues. Consistent with this hypothesis is the observation that some obese individuals have few manifestations of the metabolic syndrome. Normoglycemic and normolipidemic obese individuals display improved postprandial fat storage compared with lean subjects. Presumably, the more efficient adipose tissue fat-storing capacity the better the protection against lipotoxicity in nonadipose tissues with reduced risk of insulin resistance.

The pool of FFAs in fat cells is released into the circulation in relation to its size and the greater total fat mass of adipose tissue in obese individuals result in elevated fatty acid flux to nonadipose tissues. Although this is commonly accepted at indisputable, it has proven difficult to find that overweight and obese individuals consistently have higher serum FFA levels. Insulin resistance is not likely caused by overweight in itself. We can conclude this way because there are lean insulin resistant people and as mentioned, many overweight people are also insulin sensitive. There must be some other cause.

Non alcoholic fatty liver also correlates with insulin resistance. Fructose intake has been found to be associated with insulin resistance. Children with non-alcoholic fatty liver disease have been found to have diets high in fructose in addition to low activity patterns.

Fructose is indeed an interesting factor with respect to insulin resistance. It is also a likely candidate for a causal factor
Gross et al examined nutrient consumption in the United States between 1909 and 1997, and found a significant correlation between the prevalence of diabetes and corn syrup. Both high fructose corn syrup and sucrose contribute to high intakes of fructose. The use of HFCS in the US has apparently increased by 1000% between 1970 and 1990.

Fructose is treated and metabolized differently from glucose in the liver. It is extremely lipogenic and makes the liver churn out large amounts of triacylglycerol leading to high levels of VLDL and small dense LDL. This fat producing ability is thought to contribute to cellular lipid overload and insulin resistance.

Chronic fructose consumption also reduces the adipocyte derived hormone, adiponectin, which also seem to contribute to insulin resistance. In addition, despite its low glycemic index chronic ingestion of fructose actually seem to stimulate hyperinsulinemia.

Dietary macronutrients
The metabolic syndrome consists of a cluster of risk factors for diabetes and cardiovascular disease. These are overweight (now commonly measured as waist circumference), hypertriacylglycerolemia, reduced HDL levels, increased blood pressure and increased fasting glucose. Insulin resistance is generally a common feature of the syndrome and a likely causative agent for several of these factors.

Volek and Feinman made a funny observation in 2005. They found that the factors that define the metabolic syndrome are the same factors that are greatly improved by carbohydrate restriction. They proposed that the metabolic syndrome may well be defined by the response to carbohydrate restriction.

Carbohydrate restriction is still not recommended as the standard treatment for the metabolic syndrome despite the fact that improvements in these factors don’t even require weight loss (as it does with low fat/calorie diets) on a low carbohydrate diet.

If carbohydrate intake can improve all the factors of metabolic syndrome and insulin resistance is an important part of this syndrome. And if we know how carbohydrates in general contribute to high levels of triacylglycerols and FFA, especially fructose, then dietary carbohydrate is indeed a very likely causative factor for insulin resistance.

There is a very strong relationship between “central” fat distribution and insulin resistance. Robert Eckel claims that the sum of the evidence indicate that the metabolic syndrome (read insulin resistance) begins with excess central adiposity. By this logic, whatever cause central obesity is then what we should focus on.

A side note to central adiposity is the interesting findings of the Womens Health Initiative. After 7 years of following a diet with more fruit, vegetables and grains and less fat the 19 541 person intervention group had  increased their waist circumference. The depressing findings of what happens when women are advised to follow the national guidelines for dietary intake, lead the authors to the tragicomic conclusion:

A low-fat eating pattern does not result in weight gain in postmenopausal women.

Numerous clinical and experimental studies have linked stress to metabolic disorders. The obvious culprit is cortisol and subsequent hypercortisolemia. Cortisol has particularly strong effects on visceral fat. Giving corticosteroids in the drinking water of mice result in rapid and dramatic increases in weight gain, increased adiposity, elevated plasma leptin, insulin and triacylglycerol levels, hyperphagia, and decreased locomotion.

As visceral fat storing is a trademark of the insulin resistance metabolic syndrome, and all the factors of the syndrome are greatly improved by dietary carbohydrate restriction it is also likely that the combination of stress and a high carbohydrate diet sets the stage for insulin resistance.

The causality
No matter the mechanisms, there are likely conclusions to be drawn. Although it seems that carbohydrate intake and composition, physical activity level and stress all contribute to insulin resistance in some manner, one factor seems to stand out. I would be quick to remove low physical activity level from my list, because I believe this is likely an effect of a fat tissue that is reluctant to release stored energy. Its reluctance is likely (for the most part) caused by high insulin levels caused by dietary carbohydrates. Exercising an overweight person on a high carbohydrate diet is like exercising an anorectic person. The energy is not there to be used. Carbohydrate intake must first be manipulated in order to increase fat release and oxidation.

If excessive carbohydrate consumption causes obesity, insulin resistance and metabolic syndrome (and it likely does), and all of these factors are improved or cured by removing dietary carbohydrates, then dietary carbohydrates are a strong, if not the strongest candidate for the causal role. This conclusion is reached by looking at the physiology only, and including evolutionary and epidemiological data, I believe would only serve to strengthen this hypothesis.   

My proposed chain of causality is something like this:

Stress and high levels of FFA and triacylglycerol caused by high carbohydrate intake, stresses and pressures the fat tissue to grow. Inflammation occurs as a consequence, and the liver becomes insulin resistant because the inflammatory substances impair insulin signalling in the liver. The liver then does not decrease its glucose production in response to insulin. Increased glucose levels cause more inflammation by causing fat cells to grow even more. The cells grow due to hyperinsulinemia caused by high liver glucose output. The fat cells then become insulin resistant and do not reduce FFA production in response to insulin. This increases FFA levels which makes tissues fill up with fat (among them the pancreas, which results in abnormal insulin output) and become insulin resistant as well. All this is happening while the individual is getting more overweight and diabetic, and inflamed to the point of near combustion. 

Colpo, Eades and black swans.

There is a blogging battle going on. Dueling, are Anthony Colpo and Michael Eades. It seems Eades has Colpo cornered, and in the last round Eades pulled the «Popper’s black swan» card. Inspired the aged but beautiful writings of Popper, I thought of some other nice quotes fitting for the nutrition debate. Enjoy!

Here are some from T. Kuhn:

«No process yet disclosed by the historical study of scientific development at all resembles the methodological stereotype of falsification by direct comparison with nature.«

«They will devise numerous articulations and ad hoc modifications of their theory in order to eliminate any apparent conflict

«If any and every failure to fit were ground for theory rejection, all theories ought to be rejected at all times.«

«It makes a great deal of sense to ask which of two actual and competing theories fits the facts better.«

«… that the man who continues to resist after his whole profession has been converted has ‘ipso facto’ ceased to be a scientist.«

«…scientific training is not well designed to produce the man who will easily discover a fresh approach.«

And a last one from Popper quoting J. Black:

«A nice adaptation of conditions will make almost any hypothesis agree with the phenomena. This will please the imagination, but does not advance our knowledge.«

The trouble with insulin resistance, part III

I was in way over my head starting to ramble on about insulin resistance hoping I would identify a causal link. But I feel I should finish what I’ve started.

The trouble with insulin resistance is that it is so bloody hard to point to the exact cause at a cellular and molecular level. I’ve looked at how insulin resistance could in theory be caused by a lipid overload which through its effect on insulin signaling reduces the effect of insulin. In addition to lipid overload there is the inflammation hypothesis.

It is commonly accepted that chronic inflammation associated with obesity induces hepatic insulin resistance. Not long ago clever scientists discovered that the fat tissue was not simply a waste basket for surplus energy, but also an important endocrine organ. Many of the inflammatory substances thought to induce insulin resistance originate from the fat tissue.

Several inflammatory substances has been looked at with interest, among them are, CRP (C-reactive protein) TNF-α (tumor necrosis factor alfa), IL-1, IL-6 (interleukin 1 and 6) leptin and resistin. These are so called cytokines which are protein signalling molecules much like hormones. Other substances thought to play a part in an inflammatory led insulin resistance include; I-kappa-B-kinase beta (IKK-β), nuclear factor-κB (NF-κB), c-Jun N-terminal kinase (JNK), CC chemokine receptor 2 (CCR2) and macrophage migration inhibitory factor (MIF). And the list goes on. It is long and incomprehensible. But the big picture is clear. Obesity, insulin resistance and inflammation go hand in hand.

Some of the above substances are thought to inhibit insulin signalling in liver cells. This would then be classified as a hepatic insulin resistance. An insulin resistant liver in turn has an impaired suppression of glucose production by insulin. The result is hyperglycaemia.

Then why is the fat tissue so inflamed?

One theory claims that fat cells produce inflammatory cytokines because of endoplasmatic reticulum stress. Stress caused by the expanding of the fat cell as it fills up with excess energy thus promoting adipose tissue production of JNK, which inhibit insulin signalling. The stress also makes macrophages (white blood cells) travel into fat cells to a far larger degree than in lean people.

The source of many of the inflammatory signals originating from the adipose tissue is thought to be macrophages. Obesity-associated adipose tissue inflammation is characterized by an infiltration of macrophages into the adipose tissue. Macrophages also incorporate into artery walls and play an important part in atherosclerosis. According to Taubes quoting Anthony Ferrante, 5% of the adipose tissue in lean individuals is macrophages. In obese, the number comes close to 50%.

Ketogenic or carbohydrate restricted diets have been shown to decrease inflammation in both human and animal models. Although the mechanisms are still unclear the reduced inflammation is in thought to in part be because of reduced levels of reactive oxygen species (ROS). Hyperglycemia, as occurs in insulin resistance and type-2 diabetes, contributes to a breakdown in cellular function that leads to overproduction of reactive oxygen species. It is in fact a rather consistent feature common to all cell types that are damaged by hyperglycemia.

One way that carbohydrate restriction may decrease inflammation is through an altered fatty acid profile. Low-carbohydrate diets may result in profound alterations in PUFA (poly-unsaturated fatty acids), particularly affecting arachidonate, an omega 6 PUFA. When incorporated into membranes arachidonate is commonly assumed to have a deleterious effect on the overall inflammatory balance because of its conversion to proinflammatory eicosanoids (e.g., prostaglandin E2, thromboxane A2, leukotrienes B4).

In contrast, eicosanoids derived from long omega 3 PUFA have less inflammatory effects.

Less oxidative stress should result in better preservation of arachidonate since free radicals play a part in its metabolism. Inflammatory cytokines are known to increase production of free radicals which in turn initiate arachidonic acid release and breakdown.

Most evidence indicates that restricting dietary carbohydrate positively impacts inflammation.

Another part of the inflammation puzzle may lie in AGE (not age, but Advanced Glycation End products). These are proteins which have gotten attached to glucose probably because of hyperglycemia. These modified circulating proteins can then bind to AGE receptors and activate them, thereby causing the production of inflammatory cytokines.

Funny thing here is that mitochondria produce ROS. Researchers claim this is mostly because of an increased FFA flux with accompanying increased FFA oxidation by the mitochondria. The increased oxidation causes mitochondrial overproduction of ROS. Thus, it seems that too much fatty acids in the blood will both cause a lipid overload in tissues and cause a greater production of reactive oxygen species all of which will increase insulin resistance.