Sunday, June 30, 2013

Why Do We Care About Diet, Lifestyle and Health?

Why Do We Care? 

I think there’s a worthwhile question to ask ourselves: why do we care about diet, lifestyle or health? 

To answer this question we need to back to basics and indulge in a little philosophy. 

The cliché: ‘what is the meaning of life?’ doesn’t really ask ‘what is the definition of life?’, because that would be fairly simple, but rather asks ‘what is the purpose or significance of life?’.  The purpose of life is quite simple too, it’s reproduction: the passing of our DNA to the next generation.  Beyond that is that there are no other inherent purposes and no inherent significance.  If we want purpose and significance we have to create our own (but then to go into semantics I think that would be better described as ‘the purpose/significance in life’ rather than ‘the purpose/significance of life’). 

Purpose can be defined as: 

1.      The reason for which something exists or is done, made, used, etc.
2.      An intended or desired result; end; aim; goal. 

I’ll going to repeat myself: beyond reproduction there are no other inherent purposes in life.  I think this could be a reason why humans are inherently religious, because religion tries to add purpose to our lives.  The problem is most religious purposes assume an afterlife and at least one deity that you essentially ‘suck up’ to (if they are omniscient they know your intentions).  I’m an atheist who doesn’t believe in free will so that idea makes no sense to me.  Morality is another purpose on the Wikipedia page for ‘the Meaning of Life’.  I like morality, but to what end?  Is life merely a checklist of: ‘yes I have been honest, compassionate, forgiving, etc’? 

Sometimes I prefer to take a simple and reductionist approach with otherwise complex problems to cut through all the noise.  I see the goal in life as being to make this equation positive and then to make it as high as possible: 

Quality of Life   x   Years Lived 

Quality of life isn’t about lots of money, a big house, a fast car, etc (although if that’s what makes you legitimately happy, then go for it).  Rather it’s about how you feel on average at each moment of every day.  Years lived simply multiplies the average quality of life 

* In this basic model the goal of morality would be to improve other people’s quality of life

** If you’re feeling utilitarian simply multiply the equation by the number of people.  Utilitarianism is a very logical set of ethics that’s very difficult to do ‘perfectly’.  One criticism could be that it neglects individual freedom, so if you’re feeling libertarian you can add that condition. 

What’s The Relevance? 

What does this have to do with diet, lifestyle or health?  Everything. 

The only logical reason I see why anyone would care about either is to improve their quality of life and/or years lived (assuming quality of life is positive).  In this respect diet, lifestyle and health are means to an end, and not the end in themselves.  There are many lessons to taken from this:

  • Chasing any diet should not interfere with health and chasing health should not interfere with quality of life.
  • You are not a slave to your diet, your diet is a slave to you.  Paleo or any other diet is a tool, a means to end, rather than end in itself.
  • Don’t eat just to improve your health, eat delicious foods that make you feel good for the whole day (and don’t starve yourself to lose weight).  Don’t exercise just to burn calories or improve your health, exercise to improve self-confidence.  Don’t just reduce stress and improve sleep just to improve your health, reduce stress and improve sleep to improve your life
  • It’s important to also realise that health has more inputs that just diet and lifestyle.  And that quality of life has many more inputs than just health.  In our goal to improve quality of life we shouldn’t ignore these other factors. 

There are probably others, but I’ll leave it there.  In the modern world it’s easy to get distracted from our long term goals.  We should try to actively remember them and make decisions with them in mind

* I like Robb Wolf’s phrases of ‘health, performance and longevity’ and ‘look, feel and perform’ as a guide to how you’re going because all of them are ultimately inputs into quality of life (except longevity) (whereas stuff like BMI and cholesterol, not so much) 

Further Reading:
(1) The Meaning of Life - Wikipedia
(2) There’s more to health than food, and there’s more to life than health

Sunday, June 23, 2013

Obesity is now a Disease

There’s been a lot of talk about the American Medical Association having recently classified obesity as a disease. 

Is Obesity a Disease? 

I Googled some definitions.  From biology online: 

An abnormal condition of an organism which interrupts the normal bodily functions that often leads to feeling of pain and weakness, and usually associated with symptoms and signs. 

A pathologic condition in which the normal functioning of an organism or body is impaired or disrupted resulting in extreme pain, dysfunction, distress, or death. 

And from Wikipedia: 

The term disease broadly refers to any condition that impairs normal function, and is therefore associated with dysfunction of normal homeostasis. 

There appears to be two main points among these definitions: 

  • Diseases impair/interrupt normal (physiological) function
  • Diseases have signs and symptoms 

Obesity, which is defined as a BMI ≥ 30, can impair normal physiological function most notably in insulin resistant obesity where large adipocyte cell size leads to elevated free fatty acids and hypoxia, which in turn causes inflammation, insulin resistance and fat deposition in the liver*.  Although if adipocyte cell size remains normal (insulin sensitive obesity) there might not be much impaired function, besides athleticism. 

As for signs and symptoms this is where it gets telling.  A sign of obesity could be large waist circumference, but that’s really defining obesity in a more valid way than BMI (which by the way is good enough in most, but not all contexts) and is a more reliable predictor of insulin resistant obesity.  Obesity is the symptom 

About a year ago I wrote a post titled ‘Obesity: a Symptom of an Underlying pathology’ (so you can already see where I stand).  When I changed the format of the blog I broke it into 5 different posts, but the most important one is Leptin Resistance, where provide I evidence for the following:

Mitochondrial dysfunction >> ER stress >> PTP1B >> leptin resistance >> obesity
LPS and other factors >> inflammation >> SOCS3 >> leptin resistance >> obesity

The point is: I think labelling obesity as a disease is closer to the truth than it not being labelled as a disease, but I see obesity more like a symptom rather than a disease.  Paul Jaminet once made a comment along the lines of ‘calorie restricting your way out of obesity treats the symptom (excess weight) but not the disease’**.  If you want to treat obesity then brainstorm and target the underlying causes.

* By the way it’s generally: obesity >> insulin resistance, not the other around

** In a way this is similar to going on a low carb diet for type 1 diabetes.  Sure, you’re post-prandial blood glucose is normal, but your pancreas still doesn’t produce insulin.  The difference here being that you can reverse obesity but perhaps not type 1 diabetes past a certain point

The Feedback

From what I’ve seen some critics of this decision are CICOers and low carbers, but no doubt many people who think food reward is the dominant cause in obesity would be against this decision as well.  Which isn’t surprising, simply because these models of obesity are where obesity is caused by too much food, too many carbs or too much hyperpalatable/rewarding* food.  Particularly in their most simplified versions each case doesn’t lend itself towards discussions related to disease or underlying pathologies as we commonly think of them: such as organ/gland atrophy, immune dysfunction, etc.

The motivations behind the decision are likely financially motivated and people are right to anticipate new/more drugs and surgery for people with obesity, which probably will lower an individual’s sense of responsibility.  But arguing against the truth (assuming for a moment it is) for the sake of increasing other people's motivation to lose weight is an error Walter Willett made recently.  I bet some people who are against the labelling of obesity as a disease (purely on the grounds of drugs/personal responsibility) are probably also critical of Walter Willett's actions, which is a double standard.

* There is a difference between palatability and reward.  Palatability = tastes good.  Reward = promotes further consumption

Will Things Get Better

As for the question of: will classifying obesity as a disease make things better or worse?  I’m sitting on the fence.  It’s one of those questions that no one could know the answer to, and one where people can make reasonable arguments for each side of the debate, for example:

  • Good: people stop treating obesity as simple CICO and get serious looking at underlying pathologies that may be causing their obesity and hindering their weight loss attempts.  People stop being blamed for obesity 
  • Bad: pharmaceutical and surgical options become more available and people feel powerless to reverse the 'disease of obesity' (replacing blame with hopelessness), which both lead to poor diet and lifestyle choices and more money spent on drugs/surgery

Even if drugs come onto the market is that necessarily bad?  Sometimes drugs can have an appropriate therapeutic target and actually change the course of a disease.  I’ve noticed that some researchers are looking into PTP1B inhibitors*, which could reduce leptin resistance and insulin resistance.  If the goal is to improve people’s health, improve their quality of life and reduce healthcare cost and if there’s a drug that would do that with minimal side effects (because most people aren’t going to make a major dietary change), then isn’t that a good thing?

* I haven’t noticed anything for SOCS3 inhibitors, which might be because a major function of SOCS3 is to negatively regulate pro-inflammatory cytokines

Further Reading:
(1) Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 1
(2) Leptin Resistance
(3) Top Science Journal Rebukes Harvard's Top Nutritionist
(4) Is Obesity a Disease?

Sunday, June 16, 2013

Mitochondrial Dysfunction and Parkinson's Disease

If you would like basic some information on mitochondria, oxidative phosphorylation, etc, take a look at mitochondrial dysfunction 

Parkinson’s Disease 

Parkinson’s disease (PD) is a neurodegenerative disease that results from the death of dopamine producing (dopaminergic) neurons in the substantia nigra (SN).  Like many neurotransmitters, dopamine has many effects on brain function such as motivation, reward and controlling movement.  Symptoms of Parkinson’s disease include tremors, difficulty walking and later on, cognitive impairment.  PD can be caused by a few rare gene mutations, but is generally considered to have no known cause. 

The standard treatment for PD is L-DOPA (the precursor of dopamine) because dopamine can’t cross the blood brain barrier, but L-DOPA can.  But the dopaminergic neurons need to convert L-DOPA to dopamine and become less able to make the conversion as the disease progresses 

* PD symptoms appear when about 60-80% of the dopaminergic neurons in the SN die 

Mitochondrial Dysfunction in Parkinson’s Disease 

Mitochondrial dysfunction has been seen in people with PD and animal models of PD:

  • People with PD have defects and lower activity in complex 1 of the electron transport chain.  Complex 1 defects increase the generation of free radicals and reduce ATP production [1]
  • People with PD have higher lactate levels (suggests a poor aerobic metabolism) [2]
  • People with PD have lower levels of CoQ10 in their serum and platelets [3]
  • Glutathione depletion precedes neuron death in the substantia nigra [1] [4]

Mitochondrial genetics are associated with PD.  Mutations in mtDNA or DNA related to mitochondrial function (such as PINK1 of the PINK1/LTEN/AMPK pathway and parkin) can cause PD and can account for some of the heritability (which is only 15%).  PINK1 and parkin cooperate to maintain mitochondrial homeostasis and overexpressing PINK1 is neuroprotective [1].  For some PD related genes see this table 

In addition, a toxin called MTPT (a synthetic opiate) and a pesticide called rotenone (which some legumes contain by the way) impair mitochondrial function, which results in Parkinson’s like symptoms and pathology in the substantia nigra neurons [1].  MTPT and rotenone are often used as animal models of PD 

Why the Dopaminergic Neurons of the Substantia Nigra? 

The dopaminergic neurons in the SN are particularly vulnerable to oxidative stress and mitochondrial dysfunction for the following reasons:
  • They have low levels of antioxidants [5]
  • They contain a lot of dopamine, melanin and lipids that are vulnerable to oxidation [5]
  • The SN has about 4.5 times the concentration of microglia than the rest of the brain, which are kind of like macrophages in the central nervous system* [6]
  • They use a lot of energy because they are large neurons that have many synapses [4] 

* Which is probably why the dopaminergic neurons of the substantia nigra, but not other dopaminergic neurons, are more vulnerable [5] 

The Mitochondria as a Therapeutic Target for Parkinson’s Disease 

Supporting mitochondrial function can improve measures of Parkinson’s disease in animal models:

  • MPTP caused an animal model of PD model in mice (decreased dopamine levels and increased oxidised glutathione and malondialdehyde).  The combination of creatine and CoQ10 protected against most of the MPTP damage [7]
  • Rotenone caused an animal model of PD in rats (decreased neurons in substantia nigra, dopamine levels, ATP in the substantia nigra and glutathione, SOD and catalase activity, and increased lipid peroxides, protein carbonyls and immune activity).  Carnitine and alpha lipoic acid protected against most of the rotenone damage [8]
  • AMPK protects against mitochondrial dysfunction and dopamine loss in an unusual animal model of PD, flies with either a LRRK2 mutation or parkin gene KO [9] 

And in humans: supplementing coenzyme Q10 slowed down Parkinson’s disease symptoms and the benefit was greatest in those who had the highest dose (1200 mg) [3] 

Other Factors 

I don’t think mitochondrial dysfunction is the only factor in PD.  Just very briefly, some other factors in PD include: 

  • Polymorphisms in pro-inflammatory cytokines [5]
  • LPS, where a single large injection of LPS resulted in a progressive loss of dopaminergic neurons: a 23% loss at 7 months and 47% loss at 10 months [6]
  • 20% lower BDNF mRNA in the SN [10]

Sunday, June 9, 2013

Mitochondrial Dysfunction and Type 2 Diabetes

If you would like basic some information on mitochondria, oxidative phosphorylation, etc, take a look at mitochondrial dysfunction 

Mitochondrial Dysfunction in Type 2 Diabetes

Mitochondrial dysfunction has been found in type 2 diabetics (T2Ds) and in animal models of T2D:

  • T2Ds have impaired metabolic flexibility (associated with insulin resistance), and impaired mitochondrial function* [1]
  • Among diabetics, mtDNA mutations are associated with a lower glucose stimulated insulin secretion (GSIS) [2]
  • Mitochondrial function is impaired in the heart of an animal model of T2D [3]
  • In animal models of T2D, oxidative stress and mitochondrial dysfunction precede T2D [2] 

And mitochondrial genetics are associated with T2D: 
  • Mice with impaired mtDNA expression specifically in β cells are hyperglycemic and have impaired GSIS [2]
  • Polymorphisms in mtDNA are associated with T2D [4]
  • A type of diabetes called mitochondrial diabetes is due to inherited mtDNA mutations [5]
  • About a 1/3 of people with Friedreich́s ataxia (caused by a mtDNA mutation) develop T2D [6] 

* Impaired mitochondrial function (-12.5%) was independent of VO2 max (a measure of fitness) and was a weak predictor of RER (r2=0.19).  RER was only 3 points higher in T2Ds [1] 

Why the Pancreatic β Cells? 

The pancreatic β cells have an abnormal glucose metabolism that is ideal for a sensor, but makes them especially vulnerable to oxidative stress and mitochondrial dysfunction 

Normal Cells
Pancreatic Beta Cells
Glucose transporters (GLUT4) are packaged and sent to the cell membrane in response to insulin
The beta cell glucose transporters (GLUT1) aren't dependent on insulin, so glucose can freely diffuse into the cell at all times, and the glucose content in the cytosol is proportional to blood glucose
Oxidative phosphorylation is determined by the energy demands of the cell
Oxidative phosphorylation is determined by the availability of glucose
Glycolysis inhibits further glycolysis.  For example G6P (an intermediary of glycolysis) inhibits glucokinase (the enzyme that catalyses the first step of glycolysis)
Glucokinase is not inhibited by G6P
Pyruvate dehydrogenase (PDH), which puts pyruvate into the citric acid cycle, strongly decrease in activity in response to high glucose levels
The activity of PDH is only decreased by 22% in response to high glucose levels
Can metabolise glucose to lactate
Can’t metabolise glucose to lactate.  Glucose can only be metabolised through oxidative phosphorylation

In addition, pancreatic beta cells have low levels of antioxidant enzymes (30% of the amount of SOD as the liver and 5% of the catalase and glutathione peroxidase) [8] and insulin secretion is highly energy dependent, which generally means more OXPHOS and more ROS [2] 

A Basic Model of How Mitochondrial Dysfunction Causes Type 2 Diabetes 

Pancreatic β cells can be forced to into more oxidative phosphorylation than they require.  Oxidative phosphorylation produces ROS, but pancreatic β cells have low levels of antioxidants, making them are vulnerable to oxidative stress. 

Energy overload and oxidative stress can impair mitochondrial function by inhibiting the activity of the electron transport chain.  Oxidative stress can also damage proteins, lipids and DNA.  mtDNA can’t be repaired and β cells have a poor capacity for DNA repair.  The damage from oxidative stress is a vicious cycle where macromolecule damage leads to poor function, which leads to worse function, which leads to more ROS and more oxidative stress.  The resulting mitochondrial dysfunction impairs insulin secretion and initiates apoptosis.  Both β cell dysfunction and β cell loss seem necessary for T2D to develop [2] [6] [9] 

One of the few ways β cells can protect themselves is by increasing uncoupling protein 2 (UCP2) levels, which is up regulated in T2Ds [6].  UCPs reduce the opportunity for superoxide production but also reduce ATP generation.  This is a double edged sword because while UCP2 protects the mitochondria, ATP is necessary for insulin secretion [7]. 

* While T1D and T2D have different underlying pathologies, they both involve β cell dysfunction and death, which interestingly seems to be largely mediated by impaired mitochondrial function and mitochondrial apoptotic signalling pathways [2] 

** 10 minutes of intense oxidative stress to β cells damaged mitochondria, increased mitochondrial ROS production and increased apoptosis, and reduced oxygen consumption, ATP production and insulin secretion for days.  But it also increased gene expression for SOD, catalase and UCP2, which made them more resilient to exposure 3 weeks later.  A kind of mitochondrial hormesis [10] 

Insulin Resistance and Type 2 Diabetes 

So far it seems that too much glucose >> T2D, and that the ‘pancreatic burnout’ theory of T2D is right, in a way.  But there’s another piece of the puzzle.  As we know, glucose normally stimulates insulin release and insulin happens to be a pro-growth, anti-apoptotic hormone. 

In β cells, activation of the insulin receptor is essential for growth [7].  β cell insulin receptor knock out (βIRKO) mice have increased apoptosis and decreased proliferation of β cells, reduced β cell mass and impaired glucose tolerance [11].  To use another severe animal model, liver insulin receptor knock out (LIRKO) mice are hyperglycemic and hyperinsulinemic, but don’t develop T2D because they compensate for liver insulin resistance by increasing beta cells* [12] 

So under healthy conditions it seems that:
High glycemic load >> high insulin release >> compensatory growth of β cells 

In humans [2] and animal models [13] [14] insulin resistance precedes** T2D.  With insulin resistance, the β cells won’t receive the pro-growth signals of insulin while being inundated with lots of glucose that they must metabolise through oxidative phosphorylation.  Given the adaptability of β cells to oxidative stress and the pro-growth effects of insulin, it would seem like T2D could only develop in the presence of insulin resistance and/or persistent oxidative stress 

* So if insulin resistance is so bad, why don’t LIRKO mice develop T2D?  I’m not sure, perhaps because their insulin resistance isn’t caused by an underlying pathology (such as chronic inflammation) so their β cells remain insulin sensitive.  That being said I don’t have long term data of LIRKO mice but I suspect that while they may not develop T2D, they aren’t particularly healthy either.  Also see Evelyn's comment about NEFA 

** For information on insulin resistance see Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 3 and Stephan Guyunet’s seven part series (see Part 1, 2, 3, 4, 5, 6, 7)

Sunday, June 2, 2013

Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Summary and Conclusions


Obesity and insulin resistance are strongly associated.  Many of the mechanisms (mitochondrial dysfunction/ER stress and inflammation) and signalling proteins (SOCS3 and PTP1B) cause both leptin resistance (obesity) and insulin resistance.  However, there are some people who are overweight but insulin sensitive and others who are normal weight but insulin resistant, which seems like a paradox.  This discordance only accounts for a small number, but it does exist

There are several differences between insulin resistant and insulin sensitive obesity.  People who have insulin resistant obesity have abdominal obesity, increased oxidative stress and inflammation, lower AMPK, PGC-1α and adiponectin, more liver and muscle fat accumulation, larger and fewer adipocytes, hypoxic adipocytes and macrophage infiltration in adipocytes.

Insulin resistant obesity mainly occurs due to overflowing adipocytes where elevated free fatty acids spill into the bloodstream and initiate an inflammatory cascade, resulting in insulin resistance.

Therefore people who are overweight and have smaller but a greater number of adipocytes are probably protected from insulin resistant obesity.  However, since leptin is released in proportion to adipocyte size these people may find it more difficult to lose weight.  This may be part of the explanation why losing the last bit of weight is more difficult than the first

I put forward some explanations of the paradox in Part 3, which are summarised in the table below

Insulin Sensitive
Insulin Resistant
Normal Weight
Probably Healthy
↔ Adipocyte Number
↔ Adipocyte Size
Some Excess Abdominal Fat
↑ Adipocyte Size
Low Muscle Mass
Not Exercising
Elevated Lipolysis
Type 1.5 Diabetes
↑ Adipocyte Number
↔ Adipocyte Size
Are Exercising
Recent Weight Loss
↑ Adipocyte Size

See Overweight but Insulin Sensitive and Normal Weight but Insulin Resistant: Part 1, 2 and 3


The similarity of mechanisms between insulin resistance and leptin resistance and the positive feedback loops involved suggest (to me) that in most, but not all cases insulin sensitive obesity and insulin resistant normal weight are simply intermediary states and two different pathways to the same endpoint - insulin resistant obesity (assuming no change in diet or lifestyle).  The differences between insulin resistance and leptin resistance just explain which path was taken.  Being insulin sensitive and overweight or insulin resistant and normal weight, while better than insulin resistant obesity, is sign that something isn’t right and needs fixing.

However, the exception to this rule is if a formally overweight/obese person has lost a significant amount of weight, but not all of it.  In this context they are still overweight, but have regained insulin sensitivity due to weight loss.  Their adipocytes may be greater in number, but of a normal size, potentially explaining why they are currently insulin sensitive and also why they have difficultly losing the last bit of weight.

Part 2 and part 3 contain a fair bit of speculation at times.  Don’t take this stuff as gospel

Further Reading and Other Stuff:
(1) Not all fat people get diabetes, and not all diabetics are fat
(2) Think skinny people don’t get type 2 diabetes? Think again.
(3) Fat Tissue Insulin Sensitivity and Obesity

As I was researching this post I came across a few papers: two discussing the similar role that insulin and leptin have in the hypothalamus [1] [2], and one on how men respond more strongly to the appetite suppressing effects of insulin, while women respond more strongly to the appetite suppressing effects of leptin [3].  Also check out this blog post by Todd Becker

I was considering doing a post on insulin resistance but while writing the last three blog posts I’ve covered a fair bit on the topic.  If you would like to learn more about insulin resistance Stephan Guyenet has a seven part series on the subject (see Part 1, 2, 3, 4, 5, 6, 7)

In part 1 Stephan mentions he is sceptical that mitochondrial dysfunction has a role in insulin resistance and obesity.  He uses studies of knock-out mice to back this up.  The knockout mice all have impaired oxidative phosphorylation (OXPHOS).  That doesn't necessarily mean they have mitochondrial dysfunction.  Inhibition of OXPHOS would reduce ROS production in the ETC, and this is one way by which uncoupling proteins are actually protective.  So long is the inhibition of OXPHOS is not mediated by pathological mechanisms like ROS and inflammation, and has the opportunity to signal feedback mechanisms (which is likely in a knockout mouse) then I'm not surprised there was no insulin resistance

Many studies find that mitochondrial dysfunction is associated with insulin resistance and is a cause of T2D.  There are fewer studies on mitochondrial dysfunction as a cause of insulin resistance, but they exist [4] and the mechanisms are very plausible.

As for mitochondrial dysfunction >> obesity: while I don’t have direct evidence for it, there’s evidence to support each step of this chain: mitochondrial dysfunction >> endoplasmic reticulum stress >> PTP1B >> leptin resistance >> obesity

Perhaps mitochondrial dysfunction ought to be better defined.  Maybe instead of less ATP production (because UCPs lower ATP production and reduce oxidative stress) mitochondrial dysfunction should be defined as satisfying these three criteria: lower ATP production, mitochondrial oxidative stress and mitochondrial DNA mutations.