Sunday, May 26, 2013

Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 3

The following questions still need to be answered to solve the paradox:

  • Why are some people overweight yet insulin sensitive? 
  • Why are some people normal weight yet insulin resistant?

Adipose Tissue


The number and size of adipocytes is perhaps the most important factor.  Large adipocyte cell size is associated with insulin resistance in obesity and also in those with a normal BMI (see Part 1 and 2) [1].  The number and size of adipocytes is probably largely driven by genetics and whether there has been any previous weight gain and weight loss.  Adipose tissue dysfunction can also increase adipocyte cell size, but I suspect it more likely just contributes to the vicious cycle rather than start it


The location of adipocytes is another important factor.  Abdominal fat is associated with insulin resistance [1] and partially explains why women have lower rates of insulin resistance (lower ferritin and higher estrogen levels and are probably some other reasons why women have lower rates of insulin resistance)

Other Organs

The Liver

Elevated liver triglycerides (such as in NAFLD) increases liver insulin resistance and doesn’t require obesity to occur [2].  Choline deficiency [3], high refined sugar in the context of a bad diet [2], and LPS* [4] are some of the things can cause NAFLD

* Although LPS is probably a cause of obesity as well (via SOCS3)


BMI sometimes isn’t an accurate predictor of body composition.  Two people could have the same BMI, yet different amounts of muscle and fat tissue.  Having a normal BMI with low muscle mass and high fat mass is called ‘skinny fat’ and is probably a feature of most of the insulin resistant normal weight.  Remember that insulin resistant normal weight has larger adipocytes [1], so unless they have fewer fat cells, they would have more fat stores

Exercise is probably another important factor as it can dramatically improve insulin sensitivity with (but usually only a small amount) or without weight loss [5] [6].  Some of this can probably be attributed to a change in body composition (lowering fat, increasing muscle) [7], some to short term post-workout effects (like GLUT4 (glucose transporter) not requiring insulin) [8], and some to longer term signalling (like AMPK) [9]

The Pancreas

Some of the ‘insulin resistance’ in people with a normal weight may actually be from a relative insulin deficiency caused by autoimmune attacks on the pancreas [2]

Other Factors

Elevated Lipolysis

Things that elevate lipolysis (ghrelin, stress hormones, growth hormone, testosterone, etc) are probably going to increase insulin resistance via FFAs without much effect on weight.  For example:

  • Acromegaly is a disorder of growth hormone hypersecretion where the people affected tend to weigh less and have more insulin resistance [10]. 
  • Anabolic steroids (usually testosterone analogues) tend to lower body fat, increase muscle mass, yet increase insulin resistance*. 
  • Psychological stress increases insulin resistance [11] [12].  In addition to increasing lipolysis, stress hormones such as adrenaline and cortisol also directly decrease glucose uptake by cells and increase liver glucose output**.

* However, testosterone replacement therapy reduces insulin resistance in T2D, hypogonadal men [13]

** Increasing liver glucose output doesn’t increase insulin resistance but would give someone a higher fasting glucose measurement

*** I’ve had a suspicion for a while that in some people stress promotes weight gain and in others promotes weight loss.  This doesn’t sound completely contradictory when you consider that the list of symptoms for depression includes overeating and loss of appetite


Inflammation is an obvious factor seeing as how pro-inflammatory cytokines increase insulin resistance by increasing lipolysis, JNK, SOCS3, PTP1B and IKKβ and by lowering adiponectin and AMPK (see Part 1).  The problem is that SOCS3 and PTP1B also increase leptin resistance and so inflammation promotes weight gain as well.  So it probably depends on the location of the inflammation (brain, liver or adipocytes).  Also, as changes in insulin signalling generally seem to be faster than changes in weight perhaps people who are insulin resistant due to inflammation and are currently at a normal weight are likely inflamed and should expect further weight gain from inflammation.


Some people may be more or less sensitive to things like the effects of inflammatory signals (for example) on insulin vs. leptin receptors, the homeostatic control of body fat, etc, based on various genetic polymorphisms, mutations and/or epigenetic factors

Sunday, May 19, 2013

Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 2

Adipocytes and Weight Gain 

Adipocytes have three general responses to an increasing level of triglyceride storage 

  • Hypertrophy (getting larger)
  • Hyperplasia (dividing, therefore more cells)
  • Letting someone else deal with the problem (muscles, liver, etc) 

Hypertrophy, filling up and overflowing, caused the insulin resistance (beginning with the adipocytes) and other problems I mentioned in part 1.  Letting someone else deal with the problem just increases insulin resistance in the other tissue(s).  Hyperplasia doesn’t increase insulin resistance, it just forestalls it (assuming you’re going to keep gaining weight). 

It’s well known that weight gain generally increases insulin resistance and inflammation.  While someone is gaining weight there’s some hypertrophy and some hyperplasia.  The hypertrophy (not the hyperplasia) increases insulin resistance and inflammation.  Remember that large adipocyte cell size is a characteristic of insulin resistant obesity [1] [2] 

* One of the features of insulin resistant obesity is something called adipose tissue dysfunction, which is generally described as a hypersecretion of signalling molecules that promote insulin resistance, inflammation and atherosclerosis and less secretion of protective signalling molecules (such as adiponectin).  One component of adipose tissue dysfunction is an inability to generate new adipocytes (adipogenesis), which in turn leads to larger adipocytes (assuming weight gain) and the host of problems mentioned in part 1 [2]. 

Adipocytes and Weight Loss 

It’s also well known that weight loss generally reduces insulin resistance and inflammation, and often quite a large effect.  This seems to mainly be due to weight loss reducing adipocyte cell size but not adipocyte cell number as much [1]. 

This can present an interesting situation: people who were previously obese then have lost that weight and now have a normal weight have a smaller adipocyte cell size than weight matched controls, which promotes insulin sensitivity.  European Americans who were previously overweight were more insulin sensitive than weight matched controls.  However, this relationship was the opposite for African Americans [3], which suggests a strong genetic influence in all of this. 

Adipocytes and Leptin Signalling 

If you’re a European American I wouldn’t suggest that you try to improve your insulin sensitivity by gaining weight first and then losing it.  The reason is that leptin release is largely determined by adipocyte cell size [4], so in the case formerly obese very insulin sensitive there isn’t as much leptin released compared with weight matched controls. 

To use an extreme to help make the point see the table below:

While leptin is a pro-inflammatory cytokine, it does also help you maintain a stable body weight.  This can help explain how: 

  • Simple calorie restriction without changing anything else is so difficult to sustain, as eating less and reducing adipocyte cell size will substantially lower leptin release
  • People who were previously overweight have a harder timing maintaining their weight
  • Weight loss from obesity is more difficult as you approach your target weight (which may be especially true for people with insulin sensitive obesity* and may also explain a very anecdotal observation I’ve made that women -who more often have insulin sensitive obesity - tend to have a harder time than men in getting to an maintaining their target weight*)

Assuming the above is correct this would mean:


* Speculations

Monday, May 13, 2013

Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 1

A Paradox? 

My approach to chronic disease has been to look for underlying pathologies.  Obesity and insulin resistance (IR) share a lot in common, including underlying pathologies.  They are both responses to energy overload, seem to be mainly caused by mitochondrial dysfunction and/or inflammation, and SOCS3 and PTP1B increase both leptin resistance (obesity) and insulin resistance [1].  So it makes sense then that most people who are overweight are also insulin resistant, and measures that cause weight loss also improve insulin sensitivity.

But what about those who are overweight but are insulin sensitive* and those who have a normal weight but are insulin resistant?  Surely this is a paradox.  Robb Wolf has said something like ‘there are no paradoxes in biology, just an incomplete understanding’.  So the purpose of this blog post, like pretty much all the others, is to improve my understanding

* Roughly 25% of people with a BMI greater than 35 (obese) are insulin sensitive [3] 

Differences between Insulin Resistant and Insulin Sensitive Obesity 

Rather than just being an academic exercise, this is an important question as people with insulin resistant obesity have a higher risk of disease than people with insulin sensitive obesity [4] [5].  A good place to start for an explanation would be to find out about the differences between them.  People with insulin resistance obesity have: 

  • More abdominal fat (abdominal obesity/apple-shaped) [4]
  • Higher markers of oxidative stress and inflammation in adipose tissue and plasma [3] [4]
  • Lower AMPK, PGC-1α [4] and adiponectin [3]
  • High intramyocellular lipids (IMCL)* and hepatic lipids [6]
  • Larger but fewer adipocytes [3] [7]
  • Higher HIF-1α (a signal of hypoxia) [8]
  • Impaired adipogenesis (growth of new adipocytes) [3]
  • Macrophage infiltration in adipocytes [3] and higher ratio of M1 (pro-inflammatory) to M2 (anti-inflammatory) macrophages [9] 

* Many athletes have very high IMCLs yet are insulin sensitive.  But athletes don’t have much body fat and have a really good aerobic metabolism.  The difference seems to be that athletes store IMCLs in such a way as to promote efficient utilisation (probably smaller droplets?) [6] 

** The level of adiponectin and macrophages in adipose tissue strongly predicts insulin resistant obesity (r2=0.98) [3] 

Insulin Resistant Obesity 

In insulin resistant obesity what appears to happen most of time is this: adipocytes keep filling up with triglycerides and get larger until they can’t store anymore, and in the process they become progressively more insulin resistant.  Insulin resistance is a signal of energy overload and impairs insulin’s actions of depositing triglycerides and suppressing FFA release from adipocytes, which leads to high circulating triglycerides and FFA [4]. 

The saturated FFAs (mainly palmitate) bind to TLR4 on macrophages and activate them, which is followed by NFκB release, an inflammatory cascade and macrophage infiltration in adipocytes.  Reactive oxygen species (ROS) are also released in response to energy overload and oxidative stress can initiate a similar inflammatory cascade.  TNF-α (activated by NFκB) increases lipolysis, the first of many vicious cycles [10] 

Some of the mechanisms whereby FFA, ROS and inflammation increase insulin resistance are by reducing the activity of insulin sensitisers such as AMPK and adiponectin and activating JNK [11].  AMPK [4] and adiponectin [9] have other beneficial effects such as reducing oxidative stress and inflammation (another vicious cycle). 

While the abdominal adipocytes (in particular) are filling up and getting larger hypoxia (low oxygen) may result from the vascular not being able to keep up and/or from the adipocytes becoming too large for oxygen diffusion to work well [10].  HIF-1α is a signal of hypoxia, is elevated in adipose tissue in people with obesity (more so in insulin resistant obesity) and reduces after surgery-induced weight loss.  HIF-1α promotes inflammation, macrophage infiltration, mitochondrial dysfunction and endoplasmic reticulum stress, reduces adiponectin and may cause adipocyte cell death which leads to further FFA release [8] 

Meanwhile the elevated FFAs and triglycerides in circulation are deposited in the liver and muscles (not the blood vessels*).  Elevated triglycerides in the muscles/liver increase muscle/liver insulin resistance [6]. 

* For information on cardiovascular disease and atherosclerosis follow the links on this post 

** See the flowchart below for a summary of the mechanisms involved

Sunday, May 5, 2013

Debunking the Mediterranean Diet

The Debunking 

The Mediterranean diet is one of America’s slower growing diet fads.  The main idea behind it is that the key to longevity and optimal health is to abandon our modern diets, which make us ill, and move far back in time to the people of Crete more than 60 years ago and eat like them. 

This idea was started in 1945 by Ancel Keys and has become more popular since then (Google: ‘books on the Mediterranean diet’).  The language makes references to nutritionism, and plant-based diet.  The diet seems primarily targeted at women as there are many images of skinny women near water 

Lots of red wine! 

This idea broken down into 4 parts: 
  1. Our modern diet today makes us chronically ill and is high in saturated fat, cholesterol, salt and high GI carbs
  2. We need to abandon these modern diets and move back in time to the 1940s and eat more like the people of Crete over 60 years ago
  3. We know what these diets were like and they had a lot of red wine, they were mainly plant based and that was supplemented with fish and chicken.  But it definitely did not contain much animal fat and red meat
  4. If we emulate this diet we will improve out health and enable us to live longer 

All that red wine! 

The Mediterranean diet as it is promoted in popular books, on TV, on self-help websites, and in the overwhelming majority of popular news articles has virtually no basis in diets of the Mediterranean 

Myth 1: People in Crete consumed large quantities of red wine.  Quite the opposite, at most they consumed an average of about 30 grams (one once) per day.  (Calculated from these data: people in Crete ate about 2500 calories per day and 1% of those came from wine, beer and spirits [1], a glass of wine contains 85 calories per 103 grams [2]) 

Myth 2: Mediterranean peoples did not eat much red meat, animal fats and in some portrayals total fat.  The Cretan diet in in 1940s was higher in fat than the SAD (38% [1] vs. ~30-35%).  And it seems Mediterranean peoples may have eaten more animal foods and animal fats than what we are told* [3].  Also low fat dairy is a modern first world invention.  Many Cretans were hungry** [1], so they probably didn’t make low fat cheese then throw the fat away. 

Myth 3: Mediterraneandiet foods are what people in the Mediterranean have traditionally eaten.  This is not true.  Most food pictures in Mediterraneandiet promotional images are from non-organic and genetically modified plants or animals***.  Also in this recipe (first one I came across) the black pepper and basil came from India, not the Mediterranean (tut tut). 

Mediterranean diets were regionally and seasonally variable and they did a lot more activity than we did.  But it’s almost impossible for us now to eat this sort of diet.  Grain agriculture, as it’s currently practiced, is not sustainable 

What dietary lessons can we learn from real Mediterranean diets? 

  • Diversity is important 
  • Eat fresh foods in season, when they are ripe and most nutritious
  • Eat whole foods, not processed foods 

* Although that article may have just as much selection bias as Keyes, et al 

** If you want a diet to be sustainable why would you base it off an eating pattern where 72% were dissatisfied with their diet (which is probably mainly due to there not being enough calories) and many wanted more meat. 

*** Not that I'm an organic and anti-GMO zealot.  I'm just applying the same stupid criteria that is used to say that: 'since Paleo isn't 100% perfect reenactment, therefore it is a completely invalid approach to diet'

Not Debunked 

Just so everyone knows, this isn’t a real criticism of the Mediterranean diet, but rather a tongue in cheek post on Christina Warriner’s ‘Debunking the Paleo Diet’.  I don’t consider this post or her talk to debunk any diet as neither has used biological mechanisms or data from clinical trials.  Robb Wolf has commented on the talk (here and here) so I won’t add much, just a few things: 

3:10: “Humans have no known anatomical, physiological, or genetic adaptations to meat consumption”.  What about our requirement for vitamin B12, K2 and LCO3s 

10:30: She discusses how we have selectively bred plants to increase their size and calories and reduce their toxins, seeds and fibrous bits as if that’s a bad thing.  But also the idea that plants were toxic and fibrous kind of contradicts her earlier assertion that we didn’t eat much meat.  If plants were too difficult for ancient humans to extract calories from they would have turned to something else, animal foods perhaps...

14:20: ‘To look at a Paleo diet lets go 7,000 years back in time’…Really? 

Throughout the talk she made references that allude to Paleolithic peoples having a harder time finding calories, which is true, but is a popular idea often taken to extremes beyond the boundaries of common sense.  I compiled the list below from various sources.  While not all of these were said by Christina Warriner, how could hunter-gatherers have possibly survived assuming: 

  • They maintained a high level of activity and muscle mass (comparatively)
  • Meat was scarce and lean and getting it required a lot of effort
  • Plants were more toxic, more fibrous, had more seeds and less sweet
  • They often went hungry (thrifty gene hypothesis) and missed meals (the intermittent fasting idea)
  • They lacked central healing (therefore used more energy to warm up)
  • They had (more) parasites that stole calories
  • Food was blander so people ate less
  • Amylase activity was lower
  • There was less food processing, therefore food was more difficult to digest 

Something has to give 

* By the way, I have bought Marlene Zuk’s ‘Paleofantasy’ and plan to read it after my exams in June. 

** Some credit for this post should go to Jamie Scott who inspired this post with the following tweets