Carb Counting

According to Teresa Love, diabetes education dietitian at Howard County General Hospital, figuring out how much Humalog insulin to take with each meal is not terribly difficult. You take the number of grams of carbohydrates you are about to eat and divide it by the denominator of the appropriate insulin-to-carb ratio.

For example, if you are going to eat 50 grams of carbs and your I-to-C ratio is 1:8, you divide 50 by the denominator, 8, and get (rounding upward) 7 units of insulin. It is also permissible to round downward, to 6 units, if your blood sugar tends to drop into the hypoglycemic range with your given I-to-C ratio. (But see More on Insulin-To-Carb Ratios for a discussion of how your I-to-C ratio may vary from meal to meal.)

It is accordingly highly desirable to know how many carbs are in that meal you are anticipating eating.

Teresa recommended Calorie King Calorie Fat and Carbohydrate Counter 2007 for looking up carbs. It gives carb counts for generic foods (for instance, eggs), brand-name foods (for instance, Cheerios), and restaurant foods (for instance, spicy chicken stir-fry in a Thai restaurant). And it has carb counts for specific menu items in fast-food and chain restaurants such as McDonald's and Olive Garden.

The generic and brand-name foods it covers fall in a wide range that includes all sorts of dairy products, including ice cream and yogurt; red meats, including deli; white meats, including poultry; fish and shellfish; frozen and packaged foods, including pizza; things made from grains, including bread, breakfast cereals, pancakes, waffles, desserts, donuts, etc.; sweets and snacks, including candy, chocolate, chewing gum, nuts, granola bars, and the like; beverages, including coffee, tea, soda, sports drinks; and even alcoholic libations, including beer and wine.


You can also find out how many carbs are in packaged foods simply by reading the nutrition label. First you determine what a "serving" is; it may be different from the amount you intend to eat. For example, a "serving" of Cheerios is officially 1 cup, which also happens to be 1 oz. It has 22 grams of carbs. If you intend to eat 1.5 cups, you'll have half again as many carbs, which come to 33 g.

If one serving of a given food has more than 5 g of dietary fiber, you'll subtract the amount of dietary fiber. Since a serving of Cheerios has only 3 g of fiber, you can ignore this. If you were eating All-Bran, you would honor it. If you were eating, say, kidney beans, the high fiber content would likewise cause you to subtract it out and work with "net carbs" instead of total carbs. For all other foods, ignore the net-carbs adjustment.


Some books on carb counting work with "exchange lists." Teresa gave me Carb Counting and Exchange Lists: Tools To Help You Plan Your Meals. An "exchange" is something like a "serving." Different types of foods have different numbers of grams of carbohydrates per exchange. Most cereal- and grain-based ("starchy") foods use 15 g per exchange, as do most fruits, "starchy" vegetables, sweets, desserts, and so forth. Non-starchy veggies of the type you would put in a garden salad use 5 g per exchange, one-third that of starchy foods. Milk: 8 g per exchange, one-half of a starch exchange. Meats: 0 g. Fatty foods such as nuts and oils: 0 g.

If you eat at a Chinese restaurant, Teresa told me, you can think of each 1/3 cup of rice that you eat with your main course as 1 exchange = 15 g. The main dish (if not breaded) is apt to be a mixture of meat (or seafood) and vegetables, and you will likely eat enough of it to account for 2 exchanges. If you eat it with a cup of rice, that's 5 exchanges. Each large egg roll appetizer is 2 exchanges. Small egg rolls are 1 exchange apiece. Most Chinese soups are 0, but wonton soup will give you 1 exchange for every 4 wontons, and egg drop soup with noodles is 1 exchange. Once you know the total number of exchanges in the meal, you can multiply that number by 15 grams to compute your carbs.

Thats where your insulin-to-carbs ratio comes in. If it is 1:8, and if your meal will comprise 7 exchanges (105 g), you will inject 13 or 14 units of insulin, depending on whether you round up or down.

More on Insulin-To-Carb Ratios

Teresa Love is a dietitian who works at Howard County General Hospital's Diabetes Education Center along with nurse Mike Taylor, who is also trained in diabetes education. I consulted Teresa yesterday. From her I learned that it is not unusual for a patient to need different insulin-to-carb ratios for different meals of the day.

The liver becomes active overnight pouring sugar into the blood, resulting in elevated blood sugar around dawn. This is the "dawn phenomenon," and its causes are yet more complex than that quick summary.

So the insulin-to-carb ratio for breakfast will need to be a "tight" one. Such a ratio, for me personally, seems to be 1:5, which means one unit of insulin for every 5 grams of breakfast carbs. It is "tight" because the second number is relatively low, compared to the rest of the day.

Teresa recommended 1:8 I-to-C ratios, for me, for both lunch and dinner. (These numbers are specific to my own case. Other diabetics will typically have different numbers.) But, judging by the fact that I went slightly hypoglycemic last night after dinner, I am now thinking 1:10 for that meal, not 1:8.

For dinner last night I consumed what I estimated to be 55 grams of carbs: five squares of a small pizza at 7 grams per square (if my estimate is correct); 1 bottle of Samuel Adams Honey Porter at approximately 20 grams. Total carbs (55 grams) divided by 8 was roughly 7, so I took 7 units of Humalog insulin. That lowered my blood sugar reading from 113 before the meal to 87 three hours later. Not terribly bad, but not ideal.

The Area-Under-The-Curve Theory

I have been told by Mike Taylor, diabetes education nurse, to use a insulin-to-carb ratio of 1:5. For every 5 grams of carbs, inject 1 unit of insulin. That worked nicely for today's breakfast, but I injected 10 units instead of 12 for a high-carb lunch and, though I still was too high in my blood sugar after 2 hrs., I was too low after 4 1/2 hrs.

That may be because my lunch was high-glycemic load. According to Mike's area-under-the-curve explanation, more sugar than the Humalog insulin I injected prior to eating could offset was in my bloodstream shortly after I ate. The sugar curve was higher than the insulin curve.

Then at some point the sugar was gone but the insulin was not. Now the insulin curve was higher than the sugar curve, and I went a little hypoglycemic.

That's why I need to balance high-glycemic load carbs with low-glycemic load carbs in a meal. The glycemic load of the meal as a whole has to match what the Humalog expects. The bread I ate for breakfast on Mike's advice was intended to avoid the area-under-the-curve problem.

Also, the Humalog apparently does not leave my system in just 3 or 4 hours as advertised. It lasts for at least 4 1/2 hrs.

Insulin/Carb Ratios

Today I visited Mike Taylor, the diabetic education nurse at Howard County General Hospital in Columbia., Maryland. Mike looked at my log of Humalog injections, carbs eaten, and blood sugar readings and agreed that my results have been somewhat inconsistent.

The main problem, we both agreed, seems to be that my "basal" blood sugar levels are uncontrolled. The Humalog I am taking, which is intended to control post-meal ("bolus") elevations in blood sugar, has to fight in addition the changes in my blood sugar that occur independently of mealtimes. That is throwing off the results I am getting with mealtime insulin.

Mike will contact my doctor, Dr. Moore, with a suggestion that I be put on basal insulin as well as Humalog. He feels I need to inject perhaps 18 units of basal insulin each morning. Then I would also continue using Humalog, but in smaller mealtime doses. At that point, he suggested, I would start by using an insulin/carb ratio of 1:15 — 1 unit of insulin for every 15 grams of carbs.

That ratio might later have to be adjusted. Proper I/C ratios for "bolus" insulin such as Humalog are generally between 1:5 and 1:15, depending on the individual.

In addition, at each meal there would possibly be a "correction factor" needed: 1 extra unit of Humalog for every 50 mg/dL by which my blood sugar exceeds the midpoint of my target range for basal blood sugar level. For instance, if my target range was 70-100 mg/dL, the midpoint would be 85. If the measured level prior to a meal was 135-184, I would add 1 unit of insulin to my pre-meal dose. If it were 185-234, 2 extra units. And so on.

Mike also told me that things like eggs and bacon, which I often eat for breakfast, count as 0, carb-wise, even though for purposes of calorie counting they do have some carbohydrates in them. So a bacon-and-eggs breakfast would require no insulin injection, if I skipped having orange juice with it!

O.J. is not a good idea, Mike said, because it gets into the blood stream faster than even fast-acting insulin like Humalog can match it. Better I should eat a muffin, a bagel, or a piece of bread with my bacon and eggs, because those complex carbs take longer to enter the blood stream.

Mike set up an appointment for me with the dietitian in his office three days from now.

For now, Mike suggested I use an I/C ratio of 1:5, with the proviso that I count only those carbs that really count, diabetes-wise.

More on Humalog

As I said in my last post, Humalog is a fast-acting insulin that I inject from a handy "pen" before or just after each meal. It is a non-hexameric insulin generically called Lispro. Actually, it is an insulin analogue, whatever that means. It seems to have been engineered through recombinant DNA technology to make it quick to take effect (and quick to exit one's system).

I am new to Humalog and new to insulin in general, and I seem to have been getting erratic results with it.

I have been carefully logging my pre-prandial (immediately before a meal) and post-prandial (2 hrs. after a meal) blood sugar readings, along with the number of insulin units I have injected and the amount of carbohydrate I have consumed. There has seemed to be little predictability with respect to how these numbers relate to each other.

Today at breakfast I paid more attention than usual to the mechanics of injecting myself, while at the same time I was intentionally boosting my dosage because lower doses had seemed to be reducing my blood sugar too little. I noticed this morning that I had to exert quite a bit of pressure on the button or plunger — the gizmo on the end of the injector pen opposite the needle — in order to get the button to go in all the way and deliver the full dosage of insulin.

It struck me that in the past I may not have been pushing that little button in all the way, every time!

You are supposed to push "firmly" on the button until a diamond shape appears centered in a little window that is positioned along the shaft of the pen. I hadn't been consistently mindful of that, I realized somewhat sheepishly today.

You also need to carefully prime the pen before injecting — by squirting some insulin in the air prior to setting the dosage you want to inject. I have always been doing that properly, I think.

You additionally need to leave the needle in the injection site for a slow count of five after finishing pushing in the button, so that all the injected insulin has time to enter your body. I believe I have always been doing that properly as well.

But I probably haven't been pressing the button all the way in each and every time. The resistance to the pressure of one's thumb on the button goes way up in the last couple of millimeters of button travel, and I may not have been exerting sufficient pressure to overcome it.

From now on, I'll make a concerted effort to push the button all the way in until I see the diamond centered in the little window!


This means all my results to date are suspect. Rats! I went to see my general practitioner two days ago with complaints of inconsistent correlations between insulin, carbs, and blood sugar. I said there was some unknown X-factor interfering with my progress. Though he tried not to show it, I think he was a bit exasperated with it all. He didn't know what to tell me, so he referred me to a diabetes education nurse at a local hospital. I'll be seeing her next week.

If I have not been using the Humalog pen properly, it might well account for the inconsistency I've experienced. Only time will tell, then, whether I can gain the desired consistency by being more mindful of the proper injection technique.

Humalog

I have in the last week started injecting Humalog, the first insulin product I have tried. Humalog is fast-acting insulin to be injected just before a meal or immediately after. The amount to be injected varies with the individual and with the meal that is to be eaten, or has just been consumed. The insulin offsets the blood sugar level that otherwise develops when the carbohydrates in the meal are digested and are turned into sugar. The effect lasts only 3 or 4 hours ... by which time you're probably ready for your next meal.

My doctor has me logging my injections, the carbs I eat, and my pre-meal and (2-hr.-later) post-prandial blood glucose measurements.

If you inject too little Humalog, your post-prandial blood sugar reading will be too high: above the target range of 100 to 140 mg/dL.

If you inject too much Humalog, you could end up with low blood sugar, a.k.a. hypoglycemia, with its unpleasant symptoms, or even insulin shock (going into a coma). There is real danger in that case, so it's a good idea to have glucose tablets on hand to speed extra sugar into your blood stream if your blood sugar drops too far.

Humalog is "bolus" insulin, as opposed to "basal" insulin: it does its work quickly and exits your system rapidly. You use it only at mealtimes. As a result, it doesn't "fix" too-high glucose levels during fasting periods or between meals. My doctor has told me that after I get accustomed to Humalog use, I'll probably have to augment it with a longer-acting basal insulin, possibly Lantus.

I have been given Humalog already loaded in an injection pen, rather than using a syringe with a separate vial of insulin. The pen is designed to let me easily set the number of units of insulin to deliver through its short, fine needle into the subcutaneous fat of, say, the belly. I find it works well. There is virtually no pain when I jab myself with the needle. I have to be careful not to press on the area of the injection after removing the needle, though, lest some of the payload leak out.


Before going on Humalog, I tried three oral medications: Metformin (glucophage), Actos, and Januvia. All seemed to make me feel "funky." Januvia was the worst, seeming to give me a headache within 24-48 hrs., every time I tried to start taking it — even with the dosage level halved!

Metformin took longer — 2 or 3 days — to start making me wish I didn't have to take it. No headache, but I was feeling decidedly out of sorts.

Actos took even longer to put me in a blue funk ... but after 3 wks. it hadn't brought my blood sugar down appreciably.

Both my general practitioner and the endocrinologist I consulted were of the opinion that my body — especially my brain — was reacting to lower blood sugar with false symptoms of hypoglycemia, which would go away if I kept taking the medicine. I felt so bad on Januvia, however, that I just couldn't get over that hump.

On Humalog, I'm having few if any such side effects.

But when my blood sugar spikes after a meal, presumably because I injected too little Humalog, I feel "funky." Which makes me think some of the problems I subjectively identified as side effects of the oral medications I was using may have been because they didn't lower my blood sugar enough.