“The great tragedy of Science – the slaying of a beautiful hypothesis by an ugly fact.” – Thomas H. Huxley
There has never been a greater need for a focus on the scientific method than in the realm of nutrition. Unfortunately for the health of all of us living in this world, nutrition science is one field that has forgotten how scientific hypotheses, theories, and facts arise.
It is not completely the fault of those in the nutrition science community, as determining mechanistic, causal pathways is easily confounded by the vast complexity that makes up any diet. The number of factors making up a single food, not to mention the interactions of all these different foods in a diet and the diet’s interaction with our complex, unique bodies, is enough to make anyone want to throw in the towel and say something too simplified like, “Well we could just tell them to stop avoiding saturated fat.” Still, there is a strong need for better attempts at following strict scientific practices, so that we make fewer mistakes when making recommendations on particular foods and diets.
In order to overcome these challenges, it is absolutely crucial that we stick to one very important piece of the scientific method, one that allows us to overcome major undertakings, including understanding which foods cause disease and why. This brings us to the importance of hypothesis testing.
Hypotheses are formed based on known information, and as more and more data is uncovered, we shape and reshape the hypothesis to include all the new data. If any data contradicts the hypothesis, then the hypothesis must be reshaped to incorporate it. This last part here is key, and is, unfortunately, the part that most often gets forgotten.
When discussing nutrition and pathophysiology, the formation of strong, valid hypotheses is key to helping us understand the complexity. It is (at this time) impossible to design clinical trials that show us, definitively, all the foods that pose dangers. Rather, we must rely on a variety of experiments and different types of data to give us a larger picture. When enough data piles up, we can piece it all together to find the answers to our biggest health questions.
But we must do this cautiously. We must follow strict scientific practices when we do so, so that we avoid giving incorrect, dangerous information to the public.
Now, out of all the components of food, there is one particular component that has gotten the most heat in the last several decades: saturated fat. This particular fat has played a central role in serious debate in nutrition science circles, as well as in public. Saturated fat, notorious for its disease-causing properties, has long been known as a key food to avoid.
However, as with many health topics in the last several years, saturated fat seems to be making a comeback as old science is revisited.
Just about everyone has realized that taking fat out of the diet is a sure method for causing a decline in health. However, out of this realization, several questions have come up: what types of fats are good, and which ones should we avoid? And if fats are now good, what about carbohydrates – are they really to blame? (for some concrete answers to the fats question, check out this article)
While we navigate these other questions, one strongly held belief remains steadfast: we all know that saturated fat is what is really bad, and we need to focus on avoiding it.
Or is it? Because while in the mainstream (e.g. information heard in doctor visits and government guidelines, along with the simple matter of popular believe) it is firmly held that saturated fat really is evil, today I am here to show you that the data doesn’t quite support this.
It is more important than ever to hold all of our “theories” about food up to scientific scrutiny, and saturated fat is no exception. Therefore, today I am going to discuss this hypothesis that fat, particularly saturated fat, is responsible for a number of diseases.
In doing so, I am not going to waste any time discussing the epidemiological data that initially sparked this hypothesis, as it is heavily flawed and has been thoroughly discussed elsewhere.* What I am going to do is dive straight into the studies that have higher merit – studies that help unravel the actual mechanisms by which saturated fat causes problems.
Another thing I am not going to talk about is saturated fat and heart disease. This is not because I want to ignore the effect saturated fat has (or doesn’t have) on heart disease – rather, the confusion in that department is so huge that it requires its own thorough discussion, which I provide for you in my article on atherosclerosis. Rather, today I will focus on how saturated fat is (or isn’t) the base of almost every major chronic disease, by talking about its role in insulin resistance.
To avoid getting lost in the immense complexity that is nutrition and pathophysiology, I am going to take you through one specific hypothesis: the hypothesis that consuming saturated fat causes insulin resistance. Why insulin resistance? Because insulin resistance is at the core of almost every single chronic disease plaguing us today and is additionally affecting an estimated half of Americans.
Insulin resistance is simply the inability of the body to tolerate glucose (i.e. glucose intolerance), and since glucose is the primary fuel source for most of this world, insulin resistance becomes problematic very quickly, in numerous ways. Therefore any knowledge we can attain that allows us to counteract this problem will be crucial to reversing this chronic disease epidemic.
To discuss this hypothesis, I will give you a brief background on where it comes from, along with the mechanisms relating saturated fat to insulin resistance. From here I will dive into several studies that will help us unravel the big picture of what is really going on with the whole story. Once we understand some of the data, we can decide whether the hypothesis is valid.
As usual, I am not here to tell you that one side of the saturated fat argument is correct. Rather, I am here to explain the data and draw conclusions given the big picture. This means that I am not here to refute the data from studies demonizing saturated fat, but I am going to spend a fair amount of time critiquing conclusions from these studies.
The Saturated Fat Hypothesis:
This hypothesis begins with observations that saturated fat correlates with disease. This observation is taken from studies dating back a few decades, and show that saturated fat is correlated with numerous diseases (1,2).*
This epidemiological perspective of saturated fat was coupled with the finding that high levels of plasma saturated fat (saturated fat circulating in the blood) are also linked to insulin resistance (3,4), giving the saturated fat hypothesis a stronger, somewhat mechanistic link. Given this data, proponents of the saturated fat hypothesis had what they needed. Since studies measuring levels of plasma saturated fat, along with studies injecting saturated fat directly into the bloodstream, tended to show corresponding higher levels of insulin resistance, therefore, they concluded, saturated fat must be dangerous.
However, further research unveiled more to the puzzle. Saturated fat circulating in the blood isn’t enough to cause insulin resistance. That fat must first be deposited into tissues, where it breaks down into toxic compounds. These toxic compounds, which include ceramides and diacylglycerols, can then impair the functioning of the insulin receptor (5). Note that the type of fat deposited in tissues matters, as saturated fats are more likely to lead to the production of these dangerous compounds, while monounsaturated fats are more likely to protect against their formation (6,7).
Findings such as these give stronger credit to the hypothesis that saturated fat causes insulin resistance. If saturated fat circulating through the body can get into cells and interfere with the functioning of insulin receptors, then it makes sense that saturated fat is dangerous. However, we have to be cautious, because there is a big leap from the dangers of circulating saturated fats to the statement that consuming saturated fats causes insulin resistance.
Known:
Increased consumption of SF —–> Increased Disease, including Insulin Resistance
Increased levels of plasma SF —–> Insulin ResistanceInferred:
Increased consumption of SF –????–> Insulin Resistance
Above is a layout of the data and the hypothesis. Remember, those that have taken the responsibility for your health have traditionally claimed that, given the first two known statements, the third must be true. Therefore, these health officials give the advice to cut back on saturated fat consumption. However, I am not willing to make that leap. I am going to explore the data for this last, crucial step, so that we don’t leave our health up to best guesses.
A brief refresher on insulin resistance:
If you are entirely unfamiliar with insulin and insulin resistance, I suggest going through one of my other articles on understanding insulin and macronutrients and insulin, where I discuss insulin more in-depth. Regardless, I will provide a brief refresher on insulin and specific mechanisms of insulin resistance.
Insulin is our body’s fat storage hormone. High levels of insulin mean energy storage. Only when insulin levels are low can energy (i.e. fats and glycogen) be released from cells and burned. Insulin resistance occurs when our metabolic system has been abused over years spent making poor dietary choices. It is marked by chronically high levels of insulin, along with a decreased ability for cells to respond to that insulin. This decreased ability to respond to insulin occurs when insulin receptors become incapable of functioning properly and thus fail to listen to the insulin signal. This causes the pancreas to pump out even more insulin to try to overcome this problem, leading to even higher insulin levels.
Insulin resistance is not a problem of one specific mechanism, but rather multiple problems that are all inter-related. However, the mechanisms causing the damage to the functioning of the insulin receptor appear to be key, so it is important to spend some time to understand these.
The primary mechanism of our focus today is the insulin receptor, as it is at the core of insulin resistance and its relationship with saturated fat. When the insulin receptor’s functioning is compromised, cells cannot respond to the insulin signal, causing insulin resistance.
You can think of insulin and the insulin receptor simply as a lock and key. Insulin binds to the cell via the receptor, causing a number of processes to occur, depending on that cell, but the result is always putting the body into fat storage mode. For example, insulin binding causes an adipocyte (fat cell) to deposit fat while also causing the prevention of the release of fat from the cell. Furthermore, in the muscle, the binding of insulin tells cells and mitochondria to take in glucose, and additionally to stop burning fat and to start burning glucose.
However, if the lock (the receptor) is damaged, insulin won’t work, and glucose won’t get taken care of. Therefore the pancreas must pump out more insulin to force cells to respond.
Saturated Fat, the Insulin Receptor, and Insulin Resistance
Now that you have a general understanding of insulin resistance, let’s discuss how saturated fat plays a role in it.
Let’s begin with the general model I introduced earlier, the one based on findings that higher levels of circulating fats cause insulin resistance because circulating fats can get stored in tissues and interfere with the functioning of the insulin receptor. Remember that this problem becomes worse when those fats are saturated.
My basic insulin-resistance model looks something like this:
A note on terminology: fats can circulate through the blood (plasma) either as triglycerides (3 fatty acids bound to a glycerol molecule) or as free fatty acids (FFAs). I will also be talking about the importance of fat cells (adipocytes) and fat tissue (adipose tissue), in contrast to non-adipose tissue, which includes skeletal muscle, liver, and pancreatic tissue.
This model is strongly supported in the literature (3-7) and consists of mechanisms that are very difficult to understand without a fair level of knowledge of biochemistry. If interested in the technical details of this model, I suggest (4).
For the purposes of this article, it will suffice to understand a few basic concepts relating fat (particularly saturated fat) stored in non-adipose tissue, and how it causes insulin resistance.
Problems with saturated fat are mainly a matter of toxic fatty acid derivatives, which interfere with the functioning of the insulin receptor and interact with other molecules, setting off inflammatory signaling pathways (3-7). More specifically:
1. Plasma saturated fat can result in the production of dangerous molecules, such as reactive oxygen species (ROS), diacylglycerol, and ceramides. These molecules, in turn, go off to wreak havoc on cells, including interfering with insulin receptors.
2. Plasma saturated fat correlates with higher levels of inflammatory markers such as TNF(a), along with the activation of immune cells such as monocytes and macrophages
3. Plasma saturated fat interferes with the functioning of mitochondria, making it more difficult for the cell to convert glucose to ATP (usable energy)
Taken together, a simple model looks something like this: Plasma saturated fat gets stored in tissue, which leads to insulin resistance when toxic by-products interfere with the functioning of the insulin receptor. Note that inflammation is also a key component, and often comes hand in hand with high levels of circulating saturated fats.
Given this strongly supported model, it is clear that to avoid insulin resistance, along with the numerous diseases it relates to, we need to avoid high levels of circulating fatty acids, particularly ones in the form of saturated fat.
The key question for us now is how do we do this. How do we avoid circulating saturated fat? How do we prevent circulating saturated fat from getting stored in non-adipose tissue? And how do we keep inflammation out of the equation?
This is where it gets interesting because this is where the debate comes in, as we attempt to answer these questions.
Avoiding Circulating (saturated) Fatty Acids
Let’s begin to answer this question with the proposed solution that is currently accepted by the public: To avoid circulating saturated fat, we should avoid the consumption of fat, and in particular, avoid saturated fat. This is the current message that we all know well.
The logic is simple: circulating levels of saturated fat cause insulin resistance, so to reduce saturated fat and insulin resistance we should avoid saturated fat. Let’s put this in the form of a testable hypothesis, and see how this plays out.
Given: Plasma saturated fat causes insulin resistance
Hypothesis: Consuming saturated fat causes insulin resistance.
Model:
First off, let’s look at how this advice has played out. Remember, this has been the advice for decades, and while people haven’t cut out saturated fat entirely, they have done a fair job of, overall, lowering intake. Looking at the current trends in saturated fat consumption and insulin resistance, we see a simple observation that, despite a reduction in saturated fat consumption, levels of insulin resistance keep rising, with a now estimated 50% of the population affected by some level of insulin resistance.
This is interesting, because if saturated fat really is the cause of these problems, it would seem logical that lowering our intake of it should result in some sort of decline in insulin resistance. However, what we see is the opposite, that insulin resistance rates are increasing while saturated fat levels are not rising, and are even slightly decreasing at the population level.
Since declining consumption of saturated fat does not seem to be helping the problem, what should we do? Well, the AHA suggests that the reduction in saturated fat is not enough, and that we need to reduce levels even more, down to around 5% of our calories. Then, maybe our health problems will go away.
I don’t know about you, but this answer seems silly to me. This solution isn’t working, so instead of trying even harder, maybe we should try to be a little smarter. Instead of following this method, let’s use some real data showing the exact mechanisms at work, and use this to formulate a logical plan.
To do this, let’s look at some studies showing how the consumption of saturated fat relates to levels of circulating saturated fat.
In a 2010 (8) randomized, cross-over, controlled study published in Lipids, researchers compared two isocaloric, carb-restricted diets, varying only in the amount of saturated (SFA) vs. unsaturated fats (UFA). This is a great study because it takes many of the variables, including carbohydrates, out of the equation, and allows us to examine only the effect of saturation on plasma fat levels. Even though the SFA group consumed about 2 times the amount of saturated fat as the UFA, here is what they found:
The researchers did find that plasma saturated fat was lower in the UFA group, compared to the SFA group. However, as the authors state, “the effect was less than might be expected given the nearly two-fold difference in dietary saturated fat…” Indeed even though the SFA group ate more than twice as much saturated fat, the difference in plasma saturated fat between the groups was very small (683 ± 203 mmol/ml compared to 600 ± 200 mmol/ml), and both groups varied from the baseline of total plasma SF being 1122 ± 707 mmol/ml.
Now, this is interesting, because our entire paradigm of nutritional information is based on the idea that consuming saturated fat causes health problems. However, here we see that exchanging saturated fat for unsaturated fat (more specifically, doubling the amount of saturated fat consumed), only slightly increased plasma saturated fat. While we could debate whether or not this slight increase is significant, I say we take a step back and look at the bigger picture. The amount of saturated fat was doubled, and yet the amount of saturated fat in the blood hardly changed. This is definitely not enough to account for the large rise in plasma saturated fat seen these past decades.
We don’t have to be in the low-carb paradigm to see similar findings. One interesting study (6) involved the consumption of a chocolate shake containing either saturated, monounsaturated, or polyunsaturated oil. In this study the only statistically significant differences in plasma FFAs were an increase in monounsaturated fat in the monounsaturated fat group and an increase in polyunsaturated fat in the polyunsaturated fat group. Consuming excess saturated fat did NOT cause a statistically significant change in plasma FFA saturated fat.
However, there was one clear finding, in that the saturated fat group did have impaired insulin sensitivity, which I will get back to later on.
So if changing the amount of saturated fat consumed does not change levels of circulating saturated fat, what does significantly change the amount of circulating saturated fat? From our first study, we see one glaringly obvious answer, and it is not the amount of saturated fat consumed. No, it is the amount of carbohydrate consumed.
Interestingly, what did significantly change the levels of circulating saturated fat was the shift from baseline to the carb-restricted diets, regardless of the fatty acid composition. Again, in our first study (7) we saw that shifting from a baseline diet (34.2 ± 18% carbohydrate, 17.2 ± 5.0% SF), to carb restricted diets (13.4 ± 2.6% SFA and 12.3 ± 2.5% UFA), led to plasma SFA levels that were almost halved.
Let’s bring in one more study to make this point clear:
In a 2009 study (9) examining obese, insulin-resistant subjects, researchers hypothesized that a low-carb (LC), high-fat diet would lead to increased triglycerides (fat) in the liver (IHTG), which would lead to insulin resistance. This hypothesis is based on the hypothesis described above: since circulating fat can get into tissue, break down and cause insulin resistance, then increasing fat intake might cause insulin resistance.
However, what they found was:
“Our results refute our original hypothesis that an LC diet will cause insulin resistance because of increased adipose tissue lipolytic rates and excessive free fatty acid release into the bloodstream. In fact, we found that LC intake rapidly caused a greater reduction in IHTG content, improvement in hepatic insulin sensitivity, and decrease in endogenous glucose production rate than consumption of an isocaloric low-fat diet.” (9).
If you need a translation, here is what happened:
The hypothesis that a low carbohydrate diet causes insulin resistance due to excessive FFAs (fat) released into the bloodstream was not supported.
Instead, they found improvements in metabolic factors across the board, including a decrease in triglycerides from the liver and improved insulin sensitivity.
Note that in this study, both diets (low-carb and low-fat) were calorie restricted, which likely accounts for improvements in both groups. However, the important finding for us, today, is that the high-fat group showed larger improvements than the high-carb group.
Now this is interesting, because we have been told that it is the high-fat content that makes us fat and insulin resistant, but here we see that with an increase in fat intake we actually see a decrease in the amount of fat in the liver, along with increased insulin sensitivity.
To put the nail in the coffin (actually, more realistically, help prevent those coffins altogether), let’s look at one more study. In a 2014 study (10), human subjects were put on a diet that incrementally increased carbohydrates while decreasing saturated fat intake. At the start of the study, they consumed a low-carb, high (saturated) fat diet, and as time went on, their carb intake increased and their saturated fat intake decreased. Given the saturated fat hypothesis, we would guess that as time went on (i.e. as saturated fat consumption decreased), the amount of saturated fat in the blood would decrease, along with insulin resistance. This was not the case.
What they found was that, in line with the previously discussed studies, the subjects had a large improvement in insulin sensitivity when initially put on the low-carb, high (saturated) fat diet. Almost doubling saturated fat intake resulted in no change in plasma saturated fat, and then decreasing that dietary saturated fat intake back to baseline also resulted in no change in plasma saturated fat.
Also note that, in this study, the only significant change in insulin sensitivity was a large improvement when subjects were initially put on the low-carb, high (saturated) fat diet.
Building a Scientifically-Supported Model
Taken altogether, these studies show that we need to revise the hypothesis that consuming saturated fat causes insulin resistance.
We know that circulating levels of saturated fat can cause insulin resistance, but now we also see that consuming saturated fat doesn’t inherently contribute to levels of circulating fat.
We also see this interesting dichotomy, where in one study (9) we see the addition of saturated fat causing insulin resistance, while in others (8,10), the restriction of carbohydrates and addition of fat, including saturated fat, corresponded with no change in plasma saturated fat and a decrease in insulin resistance! However, in the study in which insulin resistance was increased when saturated fat was added, it wasn’t due to higher levels of circulating FFAs.
This leaves us with many questions regarding how all this happens. How can consuming more fat lead to less insulin resistance in one situation but not another? And how come carbohydrates seem to increase levels of saturated fat in the blood, but consuming saturated fat does not.
Regardless, we know one thing: these findings seem to put a major hole in the hypothesis that saturated fat causes insulin resistance. Rather, it seems that the actual cause may not be the saturated fat, but rather, the carbohydrate consumed.
Until those questions are answered, there are still several key takeaways. First, the data supporting the demonization of saturated fat is not as strong as health officials have led us to believe. While there is some data pointing to the dangerous effects of saturated fat, we have to be cautious when drawing conclusions.
Overall, we see that the simplified model that consuming saturated fat causes insulin resistance is flawed. There is much more to the picture, and other factors, such as carbohydrate consumption, actually play a much larger role.
That being said, I am in no way, shape, or form arguing that you start loading up on the bacon and butter. There is still enough data to support the idea that we should be cautious about saturated fat consumption, particularly when saturated fat is consumed within a high-carb diet.
Rather, I am simply asking that you open your mind to the idea that there is much more going on, physiologically, than the simple model that says that consuming saturated fat causes insulin resistance.
In the actual model governing our physiology, there are numerous factors at work, and saturated fat is only one piece. While we have been led to believe that its role in the model is one of severe harm, it looks like, in reality, saturated fat plays many roles, and only under some conditions in certain circumstances causes harm.
To answer the lingering questions, and paint a more thorough picture of the role saturated fat plays in the metabolism, we have to take a step back and understand how our metabolism functions. Once we understand the bigger picture, we can piece together the role saturated fat really plays. That is the job of my next articles, so stay tuned.
References
1. Marshall JA, Bessesen DH, Hamman RF. High saturated fat and low starch and fibre are associated with hyperinsulinaemia in a non-diabetic population: the San Luis Valley Diabetes Study. Diabetologia1997;40:430–8
2.Parker DR, Weiss ST, Troisi R, Cassano PA, Vokonas PS, Landsberg L. Relationship of dietary saturated fatty acids and body habitus to serum insulin concentrations: the Normative Aging Study. Am J Clin Nutr 1993;58:129–36
3. Martins, A. R., Nachbar, R. T., Gorjao, R., Vinolo, M. A., Festuccia, W. T., Lambertucci, R. H., … Hirabara, S. M. (2012). Mechanisms underlying skeletal muscle insulin resistance induced by fatty acids: Importance of the mitochondrial function. Lipids in Health and Disease, 11(1), 30. http://doi.org/10.1186/1476-511X-11-30
4. Estadella, D., Da Penha Oller Do Nascimento, C. M., Oyama, L. M., Ribeiro, E. B., Dâmaso, A. R., & De Piano, A. (2013). Lipotoxicity: Effects of dietary saturated and transfatty acids. Mediators of Inflammation, 2013. http://doi.org/10.1155/2013/137579
5. Samuel, V. T. and, & and Gerald I. Shulman. (2013). Integrating Mechanisms for Insulin Resistance: Common Threads and Missing Links. Cell, 148(5), 852–871. http://doi.org/10.1016/j.cell.2012.02.017.Integrating
6. Giacca, C. X. A., & Lewis, A. C. G. F. (2006). Differential effects of monounsaturated , polyunsaturated and saturated fat ingestion on glucose-stimulated insulin secretion , sensitivity and clearance in overweight and obese , non-diabetic humans, 1371–1379. http://doi.org/10.1007/s00125-006-0211-x
7. Nolan, C. J., & Larter, C. Z. (2009). Lipotoxicity: Why do saturated fatty acids cause and monounsaturates protect against it? Journal of Gastroenterology and Hepatology (Australia), 24(5), 703–706. http://doi.org/10.1111/j.1440-1746.2009.05823.x
8. Forsythe, C. E., Phinney, S. D., Feinman, R. D., Volk, B. M., Freidenreich, D., Quann, E., … Volek, J. S. (2010). Limited effect of dietary saturated fat on plasma saturated fat in the context of a low carbohydrate diet. Lipids, 45(10), 947–962. http://doi.org/10.1007/s11745-010-3467-3
9. Kirk, E., Reeds, D. N., Finck, B. N., Mayurranjan, M. S., Patterson, B. W., & Klein, S. (2009). Dietary Fat and Carbohydrates Differentially Alter Insulin Sensitivity During Caloric Restriction. Gastroenterology, 136(5), 1552–1560. http://doi.org/10.1053/j.gastro.2009.01.048
10. Volk, B. M., Kunces, L. J., Freidenreich, D. J., Kupchak, B. R., Saenz, C., Artistizabal, J. C., … Volek, J. S. (2014). Effects of step-wise increases in dietary carbohydrate on circulating saturated fatty acids and palmitoleic acid in adults with metabolic syndrome. PLoS ONE, 9(11), 1–16. http://doi.org/10.1371/journal.pone.0113605
*For a full discussion of epidemiological data, I suggest Nina Teicholtz’s The Big Fat Surprise, Gary Taubes’s Good Calories, Bad Calories, or Sally Fallon Morell’s Nourishing Fats