Sunday, June 27, 2010

Exercise and blood glucose levels: Insulin and glucose responses to exercise

The notion that exercise reduces blood glucose levels is widespread. That notion is largely incorrect. Exercise appears to have a positive effect on insulin sensitivity in the long term, but also increases blood glucose levels in the short term. That is, exercise, while it is happening, leads to an increase in circulating blood glucose. In normoglycemic individuals, that increase is fairly small compared to the increase caused by consumption of carbohydrate-rich foods, particularly foods rich in refined carbohydrates and sugars.

The figure below, from the excellent book by Wilmore and colleagues (2007), shows the variation of blood insulin and glucose in response to an endurance exercise session. The exercise session’s intensity was at 65 to 70 percent of the individuals’ maximal capacity (i.e., their VO2 max). The session lasted 180 minutes, or 3 hours. The full reference to the book by Wilmore and colleagues is at the end of this post.

As you can see, blood insulin levels decreased markedly in response to the exercise bout, in an exponential decay fashion. Blood glucose increased quickly, from about 5.1 mmol/l (91.8 mg/dl) to 5.4 mmol/l (97.2 mg/dl), before dropping again. Note that blood glucose levels remained somewhat elevated throughout the exercise session. But, still, the elevation was fairly small in the participants, which were all normoglycemic. A couple of bagels would easily induce a rise to 160 mg/dl in about 45 minutes in those individuals, and a much larger “area under the curve” glucose response than exercise.

So what is going on here? Shouldn’t glucose levels go down, since muscle is using glucose for energy?

No, because the human body is much more “concerned” with keeping blood glucose levels high enough to support those cells that absolutely need glucose, such as brain and red blood cells. During exercise, the brain will derive part of its energy from ketones, but will still need glucose to function properly. In fact, that need is critical for survival, and may be seen as a bit of an evolutionary flaw. Hypoglycemia, if maintained for too long, will lead to seizures, coma, and death.

Muscle tissue will increase its uptake of free fatty acids and ketones during exercise, to spare glucose for the brain. And muscle tissue will also consume glucose, in part for glycogenesis; that is, for making muscle glycogen, which is being depleted by exercise. In this sense, we can say that muscle tissue is becoming somewhat insulin resistant, because it is using more free fatty acids and ketones for energy, and thus less glucose. Another way of looking at this, however, which is favored by Wilmore and colleagues (2007), is that muscle tissue is becoming more insulin sensitive, because it is still taking up glucose, even though insulin levels are dropping.

Truth be told, the discussion in the paragraph above is mostly academic, because muscle tissue can take up glucose without insulin. Insulin is a hormone that allows the pancreas, its secreting organ, to communicate with two main organs – the liver and body fat. (Yes, body fat can be seen as an “organ”, since it has a number of endocrine functions.) Insulin signals to the liver that it is time to take up blood glucose and either make glycogen (to be stored in the liver) or fat with it (secreting that fat in VLDL particles). Insulin signals to body fat that it is time to take up blood glucose and fat (e.g., packaged in chylomicrons) and make more body fat with it. Low insulin levels, during exercise, will do the opposite, leading to low glucose uptake by the liver and an increase in body fat catabolism.

Resistance exercise (e.g., weight training) induces much higher glucose levels than endurance exercise; and this happens even when one has fasted for 20 hours before the exercise session. The reason is that resistance exercise leads to the conversion of muscle glycogen into energy, releasing lactate in the process. Lactate is in turn used by muscle tissues as a source of energy, helping spare glycogen. It is also used by the liver for production of glucose through gluconeogenesis, which significantly elevates blood glucose levels. That hepatic glucose is then used by muscle tissues to replenish their depleted glycogen stores. This is known as the Cori cycle.

Exercise seems to lead, in the long term, to insulin sensitivity; but through a fairly complex and longitudinal process that involves the interaction of many hormones. One of the mechanisms may be an overall reduction in insulin levels, leading to increased insulin sensitivity as a compensatory adaptation. In the short term, particularly while it is being conducted, exercise nearly always increases blood glucose levels. Even in the first few months after the beginning of an exercise program, blood glucose levels may increase. If a person who was on a low carbohydrate diet started a 3-month exercise program, it is quite possible that the person’s average blood glucose would go up a bit. If low carbohydrate dieting began together with the exercise program, then average blood glucose might drop significantly, because of the acute effect of this type of dieting on average blood glucose.

Still exercise is health-promoting. The combination of the long- and short-term effects of exercise appears to lead to an overall slowing down of the progression of insulin resistance with age. This is a good thing.


Wilmore, J.H., Costill, D.L., & Kenney, W.L. (2007). Physiology of sport and exercise. Champaign, IL: Human Kinetics.


Chris Kresser said...


Doug McGuff mentioned a few studies in Body By Science indicating that high intensity strength training (to failure) has a greater impact on insulin sensitivity than lower intensity training. Have you come across this in your research on this subject?

Anonymous said...

I work out in fasted state (12-14 hours) using a Body by Science program. Dr. McGuff explains the cori and kreb cycles in his book. The last 3 months my BG has continued to drop and now averages 87 or better. I notice on workout days it tends to run a few points higher but never that much higher. I think fasting also cause it to be a couple (2-3 points) higher. By eating a paleo diet and lifting twice a week I have lost 60 pounds and droped my BG from 135+ to 87 in 6 months.

Ned Kock said...

Hi Chris.

Higher intensity training will have a bigger impact on growth hormone release in response to exercise. This in turn may lead to more fat burning, and thus higher adiponectin levels. High adiponectin levels are associated with insulin sensitivity.

Kindke said...

I've noticed with myself that HIIT training puts me in severe danger of hypoglycemia, as noted by the extreme breathing and light headedness,

Meanwhile I function fine on resistence training or cardio ( I only do "cardio" when commuting to work on my bicycle )

However if I eat before working out, HIIT doesnt give me hypoglycemia.

Jack C said...


I recently came across some information related to exercise and lipase that may be of interest. You probably know that exercise results in increased HDL and decreased TG, but may not know why.

The three major members of the lipase family are lipoprotein lipase (LPL), hepatic lipase (HL) and endothelial lipase (EL).

EL, which was discovered in 1998, is unique in that it is the only lipase secreted by endothelial cells, and its preferred substrate is HDL. LPL works mainly and triglycerides, and HL works on both TG and HDL.

EL secretion is activated by inflammation which results in increased catabolism of HDL and thus lower HDL levels. EL secretion is suppressed by exercise and the omega-3 fats DHA and EPA (among other nutrients)which results in increased HDL concentration.

In contrast, inflammation down regulates synthesis of LPL thereby decreasing TG catabolism and increasing TG, while exercise increases LPL resulting in increased hydrolysis of TG and therefor lower TG levels.

Thus, exercise increases HDL and lowers TG while a sedentary lifestyle has the opposite effect. Similarly, an inflammatory diet (high carb, omega-6) diet will reduce HDL and increase TG, while an anti-inflammatory diet does the opposite.

Much of this information came from PMID 16980590.

Ned Kock said...

Very interesting, thanks.

Anonymous said...

So it seems like high intensity interval training, first thing in the morning might be a good way to lose subcutaneous fat!

So if I understand correctly,

The Pancrease releases the Insulin horomone that tells the liver that there is too much (>5g) sugar in the blood (hyper), and that it needs to stop producing sugar, and get rid of the sugar that's present (store it in the liver, or turn it to fat).

The opposite scenario happens during exercise or low-blood sugar... as muscles use their local glycogen, they produce the lactate hormone, which stops the liver from storing glycogen and instead gets it to release the stored sugar into the blood.

The problem I have, is the sentence,

"resistance exercise leads to the conversion of muscle glycogen into energy, releasing lactate in the process. Lactate is in turn used by muscle tissues as a source of energy"

If lactate is a by-product when muscle burn's it's local glycogen, how could it also burn it? I would think that none of it would ever get to the liver!


Ned Kock said...

Hi Chuck.

It is our body's way of re-using resources. Glycogenolysis leads to lactate production, which is also used by muscle as a source of energy. The remaining lactate goes into circulation, reaching the liver and being used as fuel for gluconeogenesis.

An analogy would be a car using carbon dioxide emissions as fuel, and creating more fuel in the process!

Ned Kock said...

Btw, lactate is not a hormone, it is an acid.

Anonymous said...

An analogy would be a car using carbon dioxide emissions as fuel, and creating more fuel in the process!

I don't know, that seems like a pretty bad thermodynamic analogy. Perhaps a better one would be reclaiming some of the heat lost to do work, like a turbocharger.

Anonymous said...

alright, so exercise raises blood glucose and causes temporal insulin resistance - which is what i 'notice' it when i train in fasted state: hunger simply disappears - so, my question is what would happen if one eats immediately after an exercise, especially something that raises blood glucose? and, thus, wouldn't it be better to wait (maybe even by sleeping) some time after an exercise to let BG levels go down?

Ned Kock said...

Hi Anon.

I definitely think that one should wait a bit before eating after exercising, but not too much. Protein synthesis is dramatically increased after exercise, particularly within a 2 h window. Therefore, waiting a bit (e.g. 30 minutes) before eating seems to make sense.

Anonymous said...

so how does this fit in with the bodybuilder mentality of taking whey protein (fast absorbing protein) along with high GI carbs to shuttle nutrients (aminos) into muscles for extra muscle growth?

surely insulin will already be higher after a workout , plus protein spike insulin a little too, so is this un-necessary?

Eddie Watts
(ps great blog, have now read all of your posts since you were mentioned on MDA)

DePuy Pinnacle Lawsuit said...

This is a good read, quite technical but very informative. More than that, I know that author only wants something good out of this.

Cep Agaric said...

I am hubungan diabetes darah tinggi grateful to find your blog from the search, when I visit here I get very nice information

Anonymous said...

HI have reactive hypoglycemia (not diabetic) and have found that if I eat more protein, it is more or less being controlled.

Can I still do high intensity work outs? I am used to working out hard and find it so sad that my sugar levels may drop post workout?

Seeker said...

what on earth?

This guy just says it's "largely incorrect".

Well show us a study of 50 people in a controlled experiment.

Just saying "largely incorrect" without some scientific control, is hilarious. You mean you thought this before, so people show you they are healthy good glucose control, but their readings skyrocket after exercise?

And your response is WHAT???? That it cant be? Almost that.

I just five minutes ago tested my blood sugar after a very brisk bike exercise. I was just curious, I am not diabetic, and I seldom get over reading of 120 even after eating. I had not eaten in two hours.

I expected a reading of 100 or 90. But I got the highest reading I can remember getting on this meter. I tested it again, and used another meter. Again it was significantly higher.

So please don't tell me it's "largely incorrect". And get out of the whole advising peole about health until you learn the value of facts, instead of retreating to your own bs.