Sunday, March 25, 2012

Leptin and the Body Fat Setpoint

About Leptin

The 'body fat setpoint' is a theory based on observations that: (1) people and other animals maintain a fairly constant weight despite not counting calories and day-to-day variations in calorie intake and expenditure; and (2) after periods of overfeeding or underfeeding people will return to roughly their previous weight (see next section).

Leptin is a homeostatic controller of stored body fat and the main driver behind the body fat setpoint.  Leptin is released by our fat cells (adipocytes) roughly in proportion to stored body fat levels.  Leptin can enter the brain, specifically the arcuate nucleus of the hypothalamus*, where it activates POMC neurons and inhibits AgRP/NPY neurons, both of these actions promote a decrease in appetite (food intake) and an increase in energy expenditure (see below)

* A part of the brain that is responsible for homeostasis of other things such as steroid hormones, body temperature and the circadian rhythm)

How Leptin Regulates Body Fat

Following overfeeding there is an increase body fat, which leads to an increase in leptin.  The higher leptin levels result in a decrease to appetite and an increase to energy expenditure, so that when the overfeeding period ends people eat less and burn more calories, which creates necessary calorie deficit for weight loss.  This weight loss continues until the body fat setpoint is reached, which is when leptin returns to baseline levels


Ad Libitum means ‘at one’s pleasure’, or how much one chooses to eat when food is freely available

Following underfeeding (calorie restriction) there is an decrease body fat, which leads to an decrease in leptin.  The lower leptin levels result in an increase to appetite and a decrease to energy expenditure, so that when the underfeeding period ends people eat more and burn fewer calories, which creates necessary calorie surplus for weight gain.  This weight gain continues until the body fat setpoint is reached, which is when leptin returns to baseline levels


This explains why it’s so difficult to maintain weight loss caused by deliberate calorie restriction

“With such a powerful system in place to keep body fat mass in a narrow range, a major departure from that range implies that the system isn't functioning correctly. In other words, obesity has to result from a defect in the system that regulates body fat, because a properly functioning system would not have allowed that degree of fat gain in the first place.” – Stephen Guyenet 

This is also why wild animals and hunter-gatherers pretty much remain lean all year round and is another refutation of the thrifty gene hypothesis.  It’s normal for this system to work. 

Defects in Leptin Signalling >> Obesity

Examples of this system not working are mice with a genetic mutation where they can’t produce leptin (ob/ob) and mice with nother genetic mutation where they don’t have leptin receptors (db/db).  Both of these mice are very obese.

ob/ob left, wild type right

But these mutations are extremely rare.  Most human obesity isn't caused by these mutations and is characterised by very high leptin levels (due to having lots of body fat) and a lack of significant weight loss in response to leptin administration.  This suggests that the defect in most human obesity in leptin resistance (LR) (analagous to insulin resistance) [3].  When mice are fattened through diet induced obesity (DIO)* mechanisms of leptin resistance, such as various signals that inhibit the signal transduction of leptin [3], inflammation [4] and apoptosis [5] of hypothalamic neurons, occurs before obesity, suggesting leptin resistance causes obesity, but also maintains obesity.

So then, what causes leptin resistance?

* Obesity researchers need a way of inducing obesity in animals quickly and reliably.  Many studies use a special diet called ‘high fat diet’ or ‘HFD’, which is high in fat but made from refined ingredients.  They use it because mice/rats consistently eat many more calories on this diet [6].  Note the difference between ‘high fat diet’ (a name) and ‘a high fat diet’ (a description of a diet that can be generalised).  There’s enough evidence to suggest humans can lose weight on high fat diets and another way of explaining HFD is ‘diet induced obesity’ (DIO), so I’ll use DIO instead of HFD. 

Further Reading:
(1) Why Do We Eat? A Neurobiological Perspective. Part III
(2) The Body Fat Setpoint

Sunday, March 18, 2012

Carbs, Fat and Fructose

Low Fat and Low Carb 

Low carb (LC) diets are often based on the role of insulin in fat synthesis (de novo lipogenesis), esterification (fatty acids >> triglycerides) and storage, while inhibiting β-oxidation.  The most basic argument is that carbs increase insulin and insulin increases fat storage, which is the carbohydrate-insulin-hypothesis (CIH) of obesity. 

Low fat (LF) diets don’t rely on any scientific mechanism besides the FLTD and the main argument is that fat has more calories per gram than carbs or protein, therefore high fat foods have more calories (not necessarily) and reduced fat foods have fewer calories.  Just like CICO, LF completely neglects the hormonal regulation of appetite. 

The CIH suggests a LF,HC diet would be the most obesogenic, while the LF,CICO suggests a LC,HF diet would be the most obesogenic.  However, people on both LC and LF diets spontaneously reduce their calories and lose more weight than those on calorie restricted diets [1].  This suggests neither macronutrient is obesogenic per se and that there are other factors involved. 

One likely factor is that LC and LF diets limit junk foods.  Junk foods are generally low in protein and nutrients, but high in fat, carbs and toxins such as artificial trans fats.  Junk foods are too high in carbs for LC and too high in fat for LF and so are avoided on those diets.  When LC or LF dieters replace junk food they are probably improving food quality by increasing protein and nutrient density and reducing artificial trans fats.  Protein is very satiating and low protein diets (<15%) can increase total calorie intake [2], multivitamin and multimineral supplements have a modest effect on weight loss [3] and artificial trans fats can increase weight gain in an isocaloric diet [4]. 

Evidence against the CIH of obesity:

  • Insulin isn’t the only hormone involved in fat storage.  Acylation stimulating hormone has the same functions as insulin on adipocytes and is released in response to dietary fat [5]
  • Insulin doesn’t lock fat away and increase appetite (the insulin >> cellular starvation argument).  Insulin is a satiety hormone and satiety from protein correlates with the insulin released [6]
  • High fasting insulin doesn’t significantly block fat oxidation and reduce metabolic rate in obesity.  People with obesity have equal or more free fatty acids than lean people because the obese have insulin resistant adipocytes and high fasting insulin is associated with a higher metabolic rate, independent of weight [6]
  • Hyperinsulinemia isn’t a causal factor in obesity.  Diet induced obesity in mice without certain genes for inflammation increases weight gain, but not insulin resistance and hyperinsulinemia [7].  Liver-specific insulin receptor knock out (LIRKO) mice are insulin resistant and have hyperinsulinemia, but actually weigh slightly less than normal mice [8] 

Low Fructose 

The CIH seems to have merged its message with low fructose.  The idea now being that fructose is metabolised like ethanol.  Fructose increases liver insulin resistance, which in the context of a high carb diet results in hyperinsulinemia and high insulin blocks leptin, increasing leptin resistance.  The new CIH/LFruc calls fructose a toxin. 

Fructose is metabolised like alcohol: Fructose and ethanol metabolism are very different and ethanol isn’t a carbohydrate.  Fructolysis is actually very similar to glycolysis as both involve the splitting of fructose/glucose into 2 pyruvate (pyruvate can enter the mitochondria and help form acetyl-CoA) and fructose/glucose can be converted to glycogen during fructolysis/glycolysis.  The important difference is that phosphofructokinase regulates glycolysis, but fructolysis doesn’t have an equivalent regulator [9].  Alcohol dehydrogenase converts ethanol to acetaldehyde (which produces oxidative stress and is highly toxic [10]), and acetaldehyde dehydrogenase converts acetaldehyde to acetic acid (vinegar), which gets converted to acetate and finally helps form acetyl-CoA.  Fructose metabolism is very different to alcohol, fructose can be converted to liver glycogen and does not produce highly toxic metabolites such as acetaldehyde. 

High insulin blocks leptin, causing leptin resistance: LIRKO mice refute this argument as they have extremely high circulating insulin levels, but weigh slightly less than normal mice [8].  Insulin actually increases leptin transport across the blood brain barrier [11]. 

Fructose is a toxin: Ethanol and acetaldehyde are toxins, but fructose isn’t necessarily.  The relationship between fructose and toxicity is likely a threshold effect.  Fructose probably only becomes a problem when the liver mitochondria become overwhelmed by excess acetyl-CoA (energy overload).  Until then fructose can refill liver glycogen and be metabolised in a similar way as glucose.  The same thing is true for other sources of calories – they aren’t generally a problem unless someone is overfeeding.  The difference between fructose and other calories is that the threshold effect is probably more sensitive with fructose because it has no negative regulator of its metabolism (such as phosphofructokinase) and a lot of it (~50%) is metabolised by the liver* [9].

Much of the discussion on the harmful effects of fructose has been extrapolated from studies using superphysiological doses of fructose (~60% of total calories).  At normal doses in humans fructose doesn’t substantially increase de novo lipogenesis**, deplete ATP to uric acid, increase blood pressure or weight gain [9]. 

Low carb diets work, but it’s not because of insulin.  The high refined sugar SAD is obesogenic, but not because fructose is a toxin, fructose metabolism is like ethanol, or because hyperinsulinemia increases leptin resistance. 

* When we consider that fructose toxicity depends on conditions and context, we can better understand how some people do fine on a fruitarian diet.  Even with fibre, which slows the rate of fructose entering the liver, fruitarian livers deal with high fructose load, probably much higher than those on the SAD.  Calorie restriction and/or endurance exercise seem to be common among fruitarians.  Calorie restriction reduces the load and both increase the rate of aerobic metabolism [12] [13] and mitochondrial biogenesis [14] [15].  But also, there is more the fruit than fructose.  Fruit contains many nutrients and other beneficial compounds. 

** If one of the reasons fructose is bad and fattening is because de novo lipogenesis can convert it to fat, then dietary fat should be even worse 

Further Reading:
(1) The Carbohydrate Hypothesis of Obesity: a Critical Examination
(2) Is Insulin Resistance Really Making Us Fat?
(3) Fate of fructose: Interview with Dr. John Sievenpiper

Sunday, March 11, 2012

Calories In, Calories Out

Calories In, Calories Out is Correct

The first law of thermodynamics (FLTD) states that energy is neither created nor destroyed and is conserved in a closed system.  Therefore if we take in more energy (calories) than we get rid of we will store/incorporate the remainder (often as body fat).  In other words:

Food (calories in) – BMR x Activity (calories out) = Change in stored energy

The actual equation is a lot more complex (see [1]).  The FLTD can be used to suggest that the reason people become overweight is from a calorie surplus by eating too much and exercising too little, and that the key to weight loss is to produce a calorie deficit by eating less and exercising more, because we can control food intake and activity level.  And this suggestion is completely correct [2]

Calorie Restriction Alone isn’t a Very Effective Weight Loss Strategy

We live in the real world, not metabolic wards.  People can go on diets and exercise to lose weight, but in response to dieting and weight loss there is:

  • An increase in appetite
  • A decrease in satiety from gut signals
  • A reduction in basal metabolic rate
  • Less of a drive to be physically active
  • A loss of lean body mass [3]

Willpower is required for calorie restriction, but willpower has its limits and most diets fail [4].  There's a reason why expressions like ‘yo-yo dieting’ exist.

More evidence against simple calorie restriction comes from an interesting experiment.  Rats were put on a regular diet or a fattening diet.  The rats on the fattening diet ate more calories and gained weight.  A third group of rats were placed on the fattening diet but they could only eat as many calories as the control group.  This third group gained 60% more weight than the rats on the regular diet [5].  Other examples include: low protein diets (<15%) can increase total calorie intake [6] and artificial trans fats can increase weight gain in an isocaloric diet [7]

An alternative strategy is to adopt a diet/lifestyle plan where you spontaneously adopt a negative energy balance with as few of the adverse side effects as possible.  For example: people on low carb, low fat and Mediterranean diets spontaneously reduced their calories and lost more weight than simple calorie restriction diets [8]

This isn’t a problem with CICO or the FLTD, but rather with the delivery of the information and the implementation

Beyond CICO

Body fat, appetite and satiety are all regulated by a number of hormones (see the table in [9]).  The role of these hormones is to ensure you eat the right amount of calories, not too much, not too little, and maintain a healthy weight.  Calories aren’t ‘bad’ or ‘unhealthy’.  High calorie foods/meals are supposed to make you full, not fat.  What is bad is when calories produce less satiety than they should or feeling hungry despite having 385,000 excess stored calories (50kg to lose).

Very few people want to be overweight, so what isn’t considered when someone says 'just eat less and exercise more' is why people who are overweight eat too much and/or exercise too little?  To borrow an analogy from Gary Taubes:

“Say instead of talking about why fat tissue accumulates too much energy, we want to know why a particular restaurant gets so crowded…If you asked me this question — why did this restaurant get crowded? — and I said, well, the restaurant got crowded (it got overstuffed with energy) because more people entered the restaurant than left it, you’d probably think I was being a wise guy or an idiot”…Of course, more people entered than left, you’d say. That’s obvious. But why?”

Sunday, March 4, 2012

The Thrifty Gene Hypothesis

The thrifty gene hypothesis (TGH) is one of the most pervasive ideas surrounding obesity.  It suggests that humans have historically lived in alternating periods of feast and famine.  So those with thrifty genes, which promote food seeking behaviours, efficiency in using calories and the storage of excess calories to fat, were better able to survive periods of famine and reproduce.  Therefore in the modern environment with access to unlimited food and no need for physical activity those of us who have thrifty genes are the ones that get fat.

The TGH was originally used to explain how the Pima have higher rates of obesity/T2D.  The TGH suggests that because the Pima have been living a traditional lifestyle up until more recently than those of European descent, therefore the Pima have been subject to feast and famine for longer and so more of them would have the thrifty genes that produce obesity in a perpetually food abundant environment [1]. 

The TGH is based on several key premises that aren’t supported: 

Humans have historically lived in alternating periods of feast and famine: Pacific islanders such as the Nauruans (the most obese and diabetic nation) and Hawaiians are particularly vulnerable to obesity and T2D when they adopt a western diet and lifestyle, leading people to believe these people carry more of the thrifty gene.  However, Pacific islanders live in a tropical climate and generally have an abundance of food year round (the Kitavans are a perfect example of this: “It is obvious from our investigations that lack of food is an unknown concept, and that the surplus of fruits and vegetables regularly rots or is eaten by dogs.”), and previously there were no reports of obesity/T2D among them*, which is true of other hunter-gatherers as well.  Thrifty genes should actually be more common in those of European descent because of European winters and famine being more a feature of early agriculture with a heavy reliance on harvest time and on a small number of crops.  Something like the Irish potato famine just simply would not happen to hunter-gatherers [2] [3]. 

Famines apply a selection pressure that favours thrifty genes: During periods of famine, infectious disease and plant toxins kill more people than starvation.  This is likely due to micronutrient deficiencies and hunger driving people to eat rotten or poisonous food.  Also, famines mostly affect the mortality rates of the young (<5) and the old (>60), the wrong demographics for a selection pressure.  The old are past their reproductive years and childhood obesity is very recent [3].  Also, it has been calculated that if thrifty genes conferred a very modest survival advantage, then almost everyone would would have thrifty genes, therefore it's unlikely that thrifty genes conferred any positive selection pressure [4]

In our modern environment those with thrifty genes will overeat, get fat and stay fat because the famine never comes: If this were true then people with thrifty genes (those who are or will become overweight/obese) would gain weight and not be able to lose weight on ad libitum diets, but this is not the case.  Obese Hawaiians lost weight on an ad libitum diet of traditional foods [4] (probably high in fruit, starchy vegetables and fish).  People on low carb or low fat diets done ad libitum spontaneously reduce calories and outperform calorie restricted diets for weight loss [5] 

Therefore the overweight/obese likely have thrifty genes: Evidence actually suggests pre-obese people don't have a lower metabolic rate [6].  Many people with obesity actually use calories much less efficiently.  They tend to have low numbers of mitochondria and a slow, inefficient electron transport chain [7], which when coupled with higher numbers of glycolytic enzymes and lower numbers of aerobic enzymes, increases the wasteful process of fermenting glucose and excreting of lactic acid [8].  They tend to have more fast-twitch muscle fibres [9], which use energy less efficiently than the slow-twitch muscle fibres [10]. 

How does thrifty genes and overweight/obesity convey any advantage when having higher body fat:

  • Requires more energy at rest and to move
  • Makes one slower and less athletic
  • Produces more pro-inflammatory cytokines and other cellular signalling (such as low adiponectin) that promotes poor health and disease [11]
  • Probably reduces sexual selection

An alternative explanation as to why the Pima/islanders/etc have higher rates of obesity is their lack of adaptation to new foods which may include alcohol, milk, fruit (fructose malabsorption), seeds, refined sugar and nutrient poor foods.  (The list depends entirely on the traditional culture) and other factors like poverty. 

The TGH is often used to justify weight loss approaches.  CICOs can use it to say we get fat because of too much feast and to lose weight we need to replicate famine (calorie restriction).  Low (fatters/carbers/fructosers) can use it to say (fat/carbs/fructose) was a scarce resource found only in the warmer months** so the (dietary fat/insulin from carbs/IR>LR from fructose) increases body fat and locks the fat from (fat/carbs/fructose) into adipocytes until winter.

* Some people use the Venus figurines as evidence that hunter-gatherers regularly experienced obesity (during times of plenty).  But I think it is more likely that ancient artists recreated the extraordinary - the people the figurines were modelled from had rare genetic mutations such as the ob/ob gene. 

** Many of these theories assume European seasonality, whereas the tropics – where humans originally came from – experience much different seasonality. 

*** The Thrifty Phenotype Hypothesis (TPH) is very different to the TGH.  The TPH suggests famine and a scarcity of nutrients (low birth weight) or T2D mothers delivering excess amino acids and glucose (high birth weight) to the baby in utero compromises normal metabolism and predisposes them to later T2D [12].  The effect of the TPH is interesting, but it’s a long way from being a major factor in T2D or obesity

**** It’s surprising that something so entrenched in the public mindset as the thrifty gene hypothesis attracts so little attention in academia.  Searching for ‘thrifty gene’ on PubMed gives me 51 free studies, ‘thrifty gene hypothesis’ only 20, but searching for ‘type 2 diabetes mitochondrial dysfunction’ gives me 252