The truth about insulin

29 07 2010

It is generally assumed that because insulin is an anabolic hormone, more insulin means equals muscle mass.  If we look at the research we will find that insulin effects have been misunderstood (or overstated).

Insulin and skeletal muscle anabolism

Insulin has a role in muscle hypertrophy because it stimulates muscle protein synthesis (MPS) and inhibits muscle protein breakdown (MPB). But insulin’s effect on MPS is dependent on blood amino acid availability and it is achieved at minimal concentrations. For instance, ingestion of 100g of carbohydrates after resistance exercise has failed to show a positive net protein balance compared to placebo and the changes were minor compared to ingestion of amino acids.  Accordingly, postprandial stimulation of MPS is independent of changes in insulin.

With an adequate supply of amino acids, insulin’s maximal effects on MPS and MPB are achieved at 30mU/L. Further stimulation doesn’t increase MPS nor decrease MPB to a greater extent. When an ample amount of protein is ingested, the addition of carbohydrates in the post exercise environment doesn’t further increase MPS. After 60 min of resistance exercise, with the ingestion of only 0.3g/kg/h of protein insulin levels averaged 16.5 uU/mL during the 6 hour recovery period. This seems to support the findings by Rennie et al., who showed that increasing levels of insulin above 10-15uU/mL does not further enhance MPS nor reduce MPB.

A positive net protein balance is achieved by an adequate extracellular concentration of amino acids, a minimal amount of insulin and an energy surplus (approximately 0.7kcal/g of protein synthesized or 240kcal/day for an average person).

Insulin as a lipogenic hormone

Insulin is probably the most lipogenic hormone, so every diet must focus on insulin control. Some people argue that fat intake, independent of changes of insulin can inhibit HSL and lipolysis, so insulin’s role on body fat accretion isn’t that important.  Nevertheless, this is just one part of the story. Insulin, besides strongly inhibiting HSL, also increases glucose uptake and glycolisis in adypocites, FFA uptake by LPL, has long-term effects on the expression of lipogenic genes and possibly stimulates de novo fatty acid synthesis.

Fat loss/accretion is determined by the net balance between lipolysis and reesterification. During fasting, NEFA transcapillary flux (NEFAtf) across adypocites is strongly negative (outward flow) and becomes close to zero at 180-240min after the ingestion of 40g of fat before becoming strongly negative again. Following a mixed meal (carbohydrates, fat and protein), NEFAtf becomes positive after 60min and stays strongly positive until at least 300min. So, in abscence of insulin, fat can inhibit HSL but to a shorter and lesser degree than with insulin, and there is lack of reesterification. Besides, insulin’s action on NEFA reesterification seems to be dose-dependent and it’s not only related to glucose uptake, but to stimulation of ASP.

Insulin stimulation by different foods

This is probably one of the most controversial topics regarding insulin. Some studies have shown that certain protein-rich foods can stimulate insulin release equally or further than some carbohydrates. In this regard, leucine seems to have a powerful dose-dependent insulinogenic potential. There is also evidence that although reducing postprandial glycemia, the addition of fat to a carbohydrate meal increases postprandial insulin levels and doesn’t affect the insulin response to a protein only meal.

Before low carb advocates start to slit their wrists, these results have to be taken with caution. Most studies show the results after only one meal and in healthy people who follow a “normal” diet (that being rich in carbohydrates). Replacement of carbohydrate with protein reduces the postprandial insulin response, as seen during a low carb ketogenic diet (LCKD). Moreover, LCKD have shown to reduce fasting insulin levels, which follow a linear correlation with insulin release after a glucose load and probably after an insulinogenic stimuli. On the other hand, the glycemic load of a diet has a direct correlation with insulinemia.


We need insulin for a proper physiological function, but the amount needed isn’t that great. Trying to achieve more muscle mass by increasing carbohydrates and plasma insulin levels could not only be detrimental for body composition but also for longevity and health.

A low fat-high vegetable diet might increase CVD risk

20 07 2010

Another study that shows the “safety” of the proposed healthy guidelines. This one was published in the Journal of Arteriosclerosis, Thrombosis and Vascular Biology from the American Heart Association. The authors wanted to measure the levels of oxLDL and Lp (a) in a group of women after a dietary intervention. Women were told to follow one of two diets:

Group A: Low fat, high vegetable

Group B: Low fat, low vegetable

A crossover design was used, so all women followed both diets at the end of the study. From baseline:

  • Both groups increased their carbohydrate (3-4%) and protein (3%) intake.
  • Total fat intake was reduced (but only 5%).
  • Saturated fat was reduced 4% in group B and 5.5% in group A.
  • Monounsaturated fat decreased 1 and 2% in group A and group B, respectively.
  • Polyunsaturated fat increased 1% in group B and 3.5% in group A.
  • Dietary fiber increased from 21g to 25g in group B and a whopping 40g in group A.
  • All micro nutrients and antioxidants were higher in group A than in group B.


Plasma Lipids

Both diets lowered triglycerides. Total cholesterol was only reduced in Group A. Not surprisingly (at least for me) HDL was reduced with both diets, having the lowest value on the low fat-high vegetable diet. LDL cholesterol was reduced only in Group A but increased in Group B.

Plasma Antioxidant Levels

All antioxidants were increased in Group A, except for lycopene and α-tocopherol.

oxLDL and Lp (a)

Both markers increased in both groups:

OxLDL-EO6* Lp (a)*
Group A 19% 9%
Group B 27% 7%

* % of increase from baseline (median value).

The levels of oxLDL-EO6 and Lp (a) and the relative changes were strongly correlated in both groups. The authors concluded (my remarks):

In conclusion, we found that a diet traditionally considered to be anti-atherogenic (low in saturated fat and high in polyunsaturated fat and naturally occurring antioxidants) increased plasma levels of circulating oxidized LDL and Lp(a). The question of whether the changes observed in the present study are, in fact, pro-atherogenic or anti-atherogenic remains to be solved. (Italics added)

A clear bias towards the lipid hypothesis is seen, beginning with the title of the study (I wonder how the study would have been called if the results showed the opposite). But are these changes in fact, pro-atherogenic? Lets see:

  • LDL oxidation is an important factor in atherogenesis, as it attracts monocytes into the vascular intima and transform them into foam cells. This process contributes to cholesterol accumulation in arterial wall macrophages and promotes pro-inflammatory events that accelerate lesion development and plaque rupture (1). Because of this, LDLox is a strong marker for CVD (2).
  • Lp (a) is also a strong risk factor for CAD and its believed to be an independent one (3, 4).

So, it seems that increased levels of oxLDL and Lp (a) are in fact pro-atherogenic. But why do a “healthy” low fat-high vegetable diet would cause pro-atherogenic results? We have to consider that:

  • Saturated fat intake and Lp (a) are inversely correlated (so more saturated fat intake = less Lp (a)) (5).
  • Polyunsaturated fat intake increases LDL susceptibility to oxidation (6).
  • Antioxidants have not shown positive results in vivo reducing CVD (7)

The lack of effect of fiber and vegetables in reducing the risk of CVD might be due to the low fat intake. Studies like this one have shown that vegetable and fruit intake might reduce CVD and CAD risk only when combined with a high fat diet.

Even though the changes in fat composition between the baseline diet and the intervention diet weren’t that drastic, the results speak for themselves. The diet wasn’t that low fat either, so the results could be worst when following a recommended “healthy balanced” diet with a more drastic reduction in total fat intake and saturated fat.

Clarifying some misconceptions about short-term fasting

16 07 2010

Fasting is to most Dietitians as garlic is to vampires. And breakfast is a must if you want to lose fat (or be healthy). Omitting breakfast is supposed to make your metabolism sluggish and contribute to fat gain because:

a. It implies a fasting period, which reduces your metabolic rate.


b.  When you dont eat breakfast you are reducing the number of meals of the day and an increased meal frequency makes you burn more calories.

It has been repeatedly shown that the latter is a fallacy and that energy expenditure related to food intake is determined by overall calorie intake, not by the number of meals. You can read more about this here. But the myth of fasting = reduced metabolism = weight/fat gain is the one that bothers me the most.

Metabolism during short-term fasting

Because in real life most IFers (those who use an Intermittent Fasting approach) use fast periods between 12 and 36 hours, I will consider short-term fasting as a fasting period that lasts a maximum of 36 hours.

The “reduced metabolic rate” fallacy comes from the fact that during prolonged fasting metabolic rate is reduced in order to survive, which seems to be related to the degree of body fat depletion (less fat, less thermogenesis). But the same doesnt apply for shorter periods of fasting, as in IF or Alternate Day Fasting (ADF) protocols.

Surprisingly for many, quite the opposite happens during short fasts: energy expenditure is increased. For instance, one study showed a significant increase in Resting Energy Expenditure (REE), associated with an increased serum concentration of norepinephrine, an increased rate of lipolysis and ketogenesis, and a shift from glucose oxidation towards fatty acid oxidation (measured by RQ), during the first 48 hours of fasting. Others have shown a rise in REE after 72 hours of fasting, peaking at the 36h mark. In these subjects, both plasma adrenaline and noradrenaline levels increased during the study. In addition, adrenoceptor sensitivity seems to be increased during this period, so adrenergic-stimulated lipolysis is potentiated.

There seems to be a threshold around the 72h mark, where REE starts to fall. But most people fast 36 hours or less.

But isnt fasting going to eat my muscles?

The rise in plasma beta-hydroxybutyrate (bOHB) concentration during fasting has anti-proteolytic properties. There is also a lowering of free T3 which reduces muscle proteolysis. Accordingly, 40h of fasting has failed to show any marked increase in atrogenic genes (myostatin, atrogin-1, MuRF-1).

A very nice summary of short term fasting metabolism is done by MacDonald and Webber:

As stated in their paper:

The primary metabolic responses to fasting-starvation are summarized in Fig. 1 , and comprise a fall in peripheral (i.e. non-neuronal tissue) glucose utilization, an increase in fatty acid and ketone body utilization and decrease in proteolysis. These changes are mediated in part by reductions in plasma insulin and free triiodothyronine (T3) concentrations and by increases in plasma glucagon and adrenaline.
There is also evidence that 20h of fasting elicits a marked activation and an increase in mRNA content of the PDK4, LPL, UCP3 and CPTI genes in skeletal muscle, changes that are associated with an increased use of free fatty acids (UCP3, LPL, CPTI) and glucose sparing (PDK4). Interestingly, changes in these markers were affected differently by the type of meal eaten after the fast: transcription of PDK4 and LPL were further increased 1h after the ingestion of both a high carb (HC) or a high fat meal (HF) and transcription of CPTI remained elevated, while UCP3 remained elevated only after HF.

In Summary

During a short fasting period your body tries to spare glucose reducing the peripheral tissue glucose uptake and oxidation, with a concomitant rise in lipolyisis and ketogenesis. This means that your body uses less glucose and more free fatty acids* for energy. Because of the short duration of the fast, muscle protein is barely affected. On top of it, your energy expenditure goes up and your adrenoceptor sensitivity is through the roof, so any stimulant will work much better.

So next time someone ask you which is the best fat loss breakfast, you can be sure to answer “a big hot cup of strong coffee”.

* Although there is a degree of re-esterification occurring.

Sweetness can be more addictive than Cocaine

14 07 2010

While reading an article by Stephan Guyenet on super-stimuli (such as MSG and excitotoxins), I found the link to a study on a subject that had caught my attention a few weeks ago. In a recent discussion, someone mentioned seeing cases of depression when people began a low carbohydrate ketogenic diet (LCKD). Seems odd, as LCKD may have mood stabilizing properties and may be useful in the treatment of depression.

So what is happening? The first thing that came to my mind was that carbohydrate intake increases the release of serotonin in the brain, which in turn produces a feeling of wellbeing. By lowering or removing carbohydrates from the diet, a disruption of this pattern occurs and depression and adverse effects could appear  (similar to the abstinence syndrome*).

However, an important point is that the feeding pattern of overweight and obese people is not characterized by a high consumption of low glycemic carbohydrates and cardboard-like tasting  foods (such as “healthy” whole grains and cereals). On the contrary, excessive carbohydrate consumption is from sugary and sweet sources. And it seems that the sweet taste is the responsible for addiction to carbohydrates and sugars, not sugar per se.

Food reward shares brain pathways with other pleasurable activities such as sex and drug administration (as well as the same behavioral paradigm with other addictions, like binging, withdrawal, craving and cross-sensitization), and consists of two branches: sensory (gustatory pathway – mediated by receptors on the tongue and the mesolimbic dopamine system) and postingestive (mediated by the metabolic products of the food). When we eat a caloric sweet food (think sodas), we stimulate the two branches of the system, both the sensory and postingestive.

As our genome, the receptors responsible for sweet perception evolved in an environment poor in sugars and probably arent adapted to high concentrations of sweet tastants, like nowadays. This access to high amounts of sugar and sweet could be producing a supra-physiological reward, with the potential to override self-control and homeostasis mechanisms, creating addiction. A chronic intake of foods high in sugar and/or sweet flavors can cause exaggerated release of dopamine (DA) in the nucleus accumbens (NAc) in response to sweet, and a delayed acetylcholine (Ach) satiation response as shown in rats that are dependent on sucrose.

Moreover, behavioral cross-sensitization (defined as an increased locomotor response to a different drug or substance) between a high sucrose diet and low dose of amphetamine has been observed, and sugar and amphetamine seem to share the same mechanism to trigger hyperactivity. This suggests that addiction caused by sweetness occurs through a mechanism similar to other drug addictions: excessive and repeated increases in extracellular DA with subsequent brain adaptation due to chronic exposure, which in turn, raises the dose needed to exert the same effect. In addition, withdrawal of drugs such as morphine, nicotine and alcohol is accompanied by alterations in DA/ACh balance in the NAc; (DA goes down and ACh goes up) and the same effect is seen during withdrawal in sucrose-dependent rats.**

But the study that caught my attention was not referring specifically to sugar, but to sweetness. The authors allowed a group of rats to choose between saccharin-sweetened water (no calories) and intravenous cocaine (a highly addictive and harmful substance). 94% of the rats preferred the sweet taste of saccharin (preference that was also observed with sucrose). Surprisingly, this behavior was also seen in rats sensitized and addicted to cocaine, which, despite increasing the cocaine dose, continued to prefer saccharin. This preference was maintained even when increasing the “cost” for saccharin, so that rats were willing to “work harder” to get it.

Unlike natural sweeteners, artificial sweeteners with no (or very few) calories as saccharin, stimulate the reward system only at the sensory level, and not the postingestive; that is, there is a partial activation of the food reward pathways. The failure to activate the postingestive component may further fuel the food seeking behaviour (and cravings for sweetness). They may also activate the sweet taste receptor in a different manner than natural sweeteners.

Repeated exposure to sweet substances (both caloric and non-caloric) could trigger a vicious and addictive cycle, in which every time a larger amount of sweet is needed. Intense sweetness (even in the abscence of calories) may act as a supra-stimuli to cells, leading to severe cravings for sweet and difficulty to give up these foods ***.

So when going into a low carb approach, some people who are addicted to sugar and sweetness may show withdrawal symptoms, like depression. In this patients, cutting carbohydrates cold-turkey could lead to severe problems.

* Indeed, enhanced responding for sugar has been observed in sugar-dependent rats after abstinence, similar to other drugs of abuse.

** For a more thorough explanation, check this review.

*** This is very important during childhood, where repeated exposure to a particular taste could influence future preferences.