BG (blood glucose)

My meter measures BG in mg/dL (that is, milligrams of glucose per deciliter of blood). Outside the United States, however, BG is seldom measured in those units. If your meter gives readings in mmol/l, divide my results by 18 to get equivalent values. That is, if my reported result is 90, for you it's equivalent to 5.

There is an important distinction to be made between fasting tests and post-prandial tests. Both are measured with the same meter, but there's a world of difference between them.

Fasting glucose

A fasting glucose test is made first thing in the morning, as soon as you get out of bed and before you do anything else. Don't eat breakfast, don't have a cup of coffe, don't take a shower, don't get dressed, don't do anything -- just take a glucose reading. Once you've got that on the record, you can do the rest of your morning routine. BG has a natural tendency to climb over the course of the morning; for purposes of comparing one day's results to another, it's best if you do your fasting test at about the same time every day.

It used to be said that the normal range for fasting BG was 70 to  109. However, the normal range for BG (like the normal range for a lot of other medical statistics) is currently being revised downard. Nowadays the normal range for BG is often said to be 65 to 99. I tend to go with that lowered range; any time I get a fasting test over 99 I feel that I've done something wrong, and had better do something about it quickly.

Post-prandial glucose

This means glucose measured after a meal. So how long after a meal should you measure it, and what should the result be? There doesn't seem to be a lot of agreement on those points.

I was told to test one hour after a meal, and to try to keep the results below 150 mg/dL. The one-hour rule reflects a basic fact about digestion: it's at the one-hour point that the carbohydrates in a meal are sufficiently digested to drive your BG as high as it's likely to go. This is true for everyone, diabetic or not, but for people with diabetes the peak tends to be higher.

However, some doctors think the most important issue is not how high your BG goes, but how long it stays high. They advise their patients to test 2 hours after a meal, not 1 hour. And how high is BG allowed to be at that point? Opinions seem to vary on this point, but the value I've heard quoted most often is 140 mg/dL.

However, I'm continuing as I started: for me, a post-prandial BG test happens 60 minutes after I finish eating a meal, and I want the result to be under 150. 

Special note: low readings in early 2008

From mid-February to early April of 2008 I was using a batch of test strips which seemed to be biased to read about 6 points low. My fasting average fell from 90 to 84 as soon as I started using those test strips. The manufacturer denies that this is possible (full disclosure: they're lying). When I ran through that batch and started a new one, my BG results increased, as I expected.

However, in June I managed to drive my fasting average downward, almost as low as it had been when I was using that odd batch of strips (but legitimately this time, as far as I can tell). Maybe those falsely low readings gave the inspiration I needed to achieve some genuine ones.


How accurate are glucose meters? It's hard to find out. Most manufacturers treat this information as secret. I can't imagine what gives them that right, but if you try to find out anything about the issue you will discover almost no useful information. The "Specifications" section of the manual that came with my glucose meter provides such fascinating information as how much the thing weighs and what kind of batteries it uses. Important facts, no doubt, but considering that the only purpose of the thing is to measure the concentration of glucose in human blood, I was curious as to how accurately it does that. The manufacturer has decided, for whatever reason, that anyone who wants to know this should mind his own damned business.

Once, during a phone call to the customer support line for the meter, the person on the other end of the line blurted out, very much inadvertantly, a comment indicating that the accuracy spec was plus or minus 16%. That sounds so imprecise that I'm not surprised the manufacturer doesn't want to publicize it.

However, there are other issues to consider here, which I understand better than most patients because I work for a company that makes test equipment:

I guess the short answer (unsatisfying though it may be) is that the meter is not terribly accurate on any one test, but that the measurement errors that occur in glucose testing tend to vary randomly, and to disappear during averaging. The bottom line is that you have to take averages and long-term trends a lot more seriously than the ups and downs of daily testing.

Blood pressure

For a long time I have been mystified by the emphasis that is placed on reducing blood pressure as a means of preventing heart attacks. I could see how high blood pressure could lead to an aneurysm, but how could it lead to a heart attack (which is caused by arterial blockage, after all)?

I was finally enlightened on this subject by a cardiologist friend, who said that high blood pressure places a constant physical stress on the arteries, and that this results in inflammation of the arterial walls. Inflammation of those particular tissues (whether it's caused by high blood pressure, overproduction of insulin, overproduction of stress hormones, or anything else) leads to arterial trauma. Arterial trauma leads to blood clots which can cause a heart attack. In other words, simply by maintaining a high level of blood pressure, you are wounding your arteries and setting them up to fail.

In just about any drug store these days, you can buy a highly automated device for testing your blood pressure at home. You don't need to develop the professional skills of a nurse or doctor to use these things; they give you unambiguous numbers on an LCD display.

A blood pressure test result is actually two numbers: the first is the "systolic" pressure (the elevated pressure that occurs in the middle of a pulse) and the second is the "diastolic" pressure (the reduced pressure that occurs between pulses). The second number is always lower than the first.

It used to be said that your blood pressure was okay if the sytolic was below 140 and the diastolic was below 90 (although diabetes patients were urged to do better than that if at all possible). Apparently those golden days will never come again; the current thinking is that the systolic should be below 120 and the diastolic should be below 80. The revised targets are not easy to hit; sometimes I can do it, and sometimes I can't.

It has often been claimed that the lower number (the diastolic number) matters a lot more than the higher one. Most experts are impatient with this idea; they say it is a misleading generalization, and that both numbers are important. Which is too bad, because I usually do better on the lower number than on the higher one.


I measure weight in pounds; if you measure it in kilograms, multiply my results by 0.45.

People who exercise a lot tend to see a lot of weight-fluctuations from day to day -- fluctuations which have very little to do with gain or loss of body fat. The real issue is usually gain or loss of water. I seem to be especially prone to dehydration during prolonged exercise, apparently because I to perspire more than a normal human. If I'm doing a long trail-run or a long bike ride, I try to carry enough water, and drink enough of it, to counteract this, but it's a losing battle. I can replace enough lost fluid to keep me going, but when I get home and weigh myself it is not unusual to find that I've lost five pounds, if the exercise lasted more than an hour and a half and the weather was at all warm. It takes a while to rehydrate yourself after something like that, so my weight tends to read low for a day or two. I always fall for it. That is, I always jump to the conclusion that my weight is down because my recent heroic exercise effort has burned a few pounds of fat off my body. Funny how that fat always comes back after I drink a lot of water.

To figure out how much you ought to weigh, you can calculate your Body Mass Index (BMI).

  1. First, figure out your height in inches (for me, that's 70).
  2. Second, square that result (for me: 70 times 70 equals 4900).
  3. Third, divide your weight in pounds by that result (for me, as of March 10, 2008: 186 pounds divided by 4900 equals 0.3796).
  4. Fourth, multiply that result by 703 (for me: 0.3796 times 703 equals a BMI of 26.7).
  5. Get upset because a BMI of 25 or more indicates you are overweight, and a BMI of 30 or more indicates you are obese.

If you live outside the United States, you can do the same calculation in metric units, using weight in kilograms and height in meters -- but skip step 4, which is only needed for those of us using the antique measurement units we favor on these shores.

By this calculation I need to get down to 170 pounds in order to escape from the "overweight" category. My doctor was satisfied with getting me out of the "obese" category, and hasn't been pressuring me to lose more weight, but I'm sure there will be advantages to getting my BMI below 25, and I ought to work on it.