Charles Coughran <ccoughran(AT)ucsd.edu> contributed this section.

The best way to deal with problems associated with diabetes and exercise
begins with understanding of what goes on in the metabolic system of
normal people and what the differences are for diabetics. Only with
such understanding can you make intelligent choices about
pharmacological tactics. Relying on rules of thumb can cause more
problems it solves because of the wide variability of individual
responses and the wide variety of diseases that fall under the rubric
of diabetes. Not to mention, I have seen postings where the rules of
thumb were clearly misunderstood.

While the following is intended for those who take insulin, it may
assist those on oral medications as well. Exercise in this context
means extended aerobic activity, say a minimum of 20 minutes of
jogging. This is a somewhat simplified account but I think it captures
the most important aspects for exercise related bg control. Comments
encouraged.

When a normal person starts to exercise, the insulin output of his
pancreas goes down. At first blush, this seems backward since the
muscles are working hard and therefore require more glucose to be
transported from the blood into the cells. There are two reasons more
glucose can be transported with less available insulin. The first is
that during exercise insulin becomes much more efficient. The mechanism
of this effect is not fully understood, but it helps overcomes the
reduction in circulating insulin.

Second, exercise activates non-insulin mediated glucose transport
pathways. These pathways are not sufficient to handle the load in the
absence of insulin, but do increase the effective insulin efficiency.

When insulin levels decline relative to the counterregulatory hormones
-- glucagon, epinephrine, norepinephrine, growth hormone, and cortisol
-- the liver is stimulated to release stored glucose. The blood glucose
that is being transported into the cells is replaced by that from
hepatic stores. It is this hormonal balance system that keeps the
levels of blood glucose in the normal narrow range during exercise.

For those of us who inject insulin, the first problem is obvious. Our
circulating levels of insulin do not react to exercise. Absent any
correction, when the muscles demand glucose and insulin becomes more
efficient our blood glucose plummets and we become hypoglycemic. This
is the reason for a commonly encountered prohibition to not schedule
exercise when your insulin is peaking. The higher the level of
circulating insulin, the more pronounced the effect.

One solution is to reduce our circulating insulin levels by reducing
insulin intake. Here specific advice starts to be difficult due to the
wide variety of insulins, regimens, and individual variability. The
spectrum spans from a Type II who takes a little NPH to help his beta
cells out to a c-peptide free pumper. I have spoken to diabetic runners
whose tactics would put me in an ambulance, even though our situations
seem to be very similar. You see a lot of advice of the form, "reduce
your insulin 2 units for every hour of strenuous exercise". This kind
of advice ignores real world variability and is sometimes much worse
than useless.

Clearly, someone who takes one shot/day has a much more limited ability
to adjust circulating insulin levels than someone using multiple
injections or a pump.

The other approach is to increase blood glucose levels by eating
carbohydrates timed to arrive at the blood stream in the form of
glucose when it is needed. The easiest way to do that is usually to eat
fast acting carbohydrates during or immediately preceding exercise.
Again, there are rules of thumb around about so many grams of
carbohydrates for a particular length of exercise at some defined
level. Again, they seem to be swamped by individual and circumstantial
variability.

Some of us do a combination of both and pump up our bg levels somewhat
before exercise and reduce insulin levels to keep things on an even
keel.

The bottom line is to make careful adjustments and test, and test, and
test, to find out how things work for your particular body.

So much for too much insulin. What happens when the circulating insulin
level is too low? When levels are so low that even the increase in
insulin efficiency doesn't overcome the defect, glucose isn't
transported into the cells. Worse, since insulin levels are low the
liver continues to pump glucose into the blood. The result is bg levels
rise with exercise. The muscles get stressed due to lack of fuel and
the metabolism of fats kicks in, ketones start being produced and the
danger of ketosis or ketoacidosis looms. This is the basis for another
rule of thumb which is often misunderstood. The rule is usually stated
"don't exercise when your bg is above 240 mg/dl (13.3 mmol/l) and
ketones are present in the urine". This makes sense because those are
signs that you have inadequate insulin supplies -- that's how many of
us got diagnosed. Exercise in those circumstances will make things
worse, not better. On the other hand, if you are 300 mg/dl (16.7
mmol/l) because you just drank a large regular cola by mistake with
lunch, exercise is a great way to bring that bg down in a hurry. Why
your bg is elevated is just as important as the fact of the elevated
level when deciding whether or not exercise is contraindicated. The 240
is also a somewhat arbitrary number. Some people start throwing ketones
at significantly lower levels.

In short: avoid exercise if your insulin level is too low. Do exercise
if you are sure your insulin level is adequate but your blood glucose
is too high.

Exercise also produces effects at longer time scales. Sometime after
exercise, there is often a take up of blood glucose by the muscles to
replenish depleted stores. This most often occurs an hour or two after
exercise, but has been reported in the range of 1/2 hour to 48 hours.
Again, as is the case during exercise, artificially high insulin levels
will lead to hypoglycemia. The last rule of thumb is to watch for
hypoglycemia after exercise.

*SPECULATION BEGINS HERE* A problem some of us encounter from time to
time is a post exercise bg spike. Blood glucose readings will be
reasonable after exercise but sharply elevated a few hours later. It is
my speculation that this represents circulating insulin levels that
were adequate to deal with exercise induced blood glucose demand with
its attendant insulin efficiency increase, but too low to deal with the
post exercise demand when insulin efficiency has lowered somewhat. It
has been my experience that post exercise elevated bg levels respond to
much less insulin than would be required in a more normal situation. It
appears that insulin efficiency falls off after exercise at some rate
and you can be on the correct side of the curve during exercise and the
wrong side after. This hypothesis is the best of a couple I have come
up with. *SPECULATION ENDS HERE*

Regular exercise over time scales of weeks or months can reduce overall
insulin requirements. In addition, as muscles become trained and
improve their internal storage, it feeds back into the amount of
glucose demand present during exercise, and thus into the entire
control cycle.

Diabetes makes exercise, and almost everything else, harder. But, hey,
if it was easy it wouldn't be any fun :-)

There are two very good, readable books from which you can get more
information. The better is Campaigne and Lampman, _Exercise in the
Clinical Management of Diabetes_. Almost as good is _The Health
Professional's Guide to Diabetes and Exercise_ edited by Ruderman and
Devlin and published by the American Diabetes Association.