If we never ate, basal, or background, insulin would meet our needs just fine. But we do eat, and food makes our blood glucose go up, which is why many insulin users now use a basal–bolus insulin regimen, in which boluses of insulin are taken before each meal or substantial snack.
Bolus insulin doses should be tailored to the amount of carbohydrate in a meal or snack, because carbohydrate in food is what makes your blood glucose level rise. For the most part, carbohydrate is also the body’s preferred source of energy, so some is necessary in the diet. Carbohydrate is found in most starchy foods (such as breads, cereals, potatoes, rice, pasta, and beans) as well as fruits, juices, milk, sweets, and to a lesser extent, nonstarchy vegetables.
Fat’s direct impact on blood glucose levels is minimal. However, consuming large amounts of fat can produce a temporary state of insulin resistance, in which the body’s cells become resistant to the effects of insulin and take up glucose from the blood less readily. This temporary, fat-induced insulin resistance can cause a gradual blood glucose rise over many hours.
Protein’s effect on blood glucose is minimal when it is included as part of a complex meal. But when protein is consumed in the absence of carbohydrate, upward of 50% of the protein may be converted into glucose within a few hours, resulting in a moderate blood glucose rise.
However, since it is uncommon for most people to consume purely protein or excessive amounts of fat in all their meals, bolus insulin doses are generally calculated to offset the impact of carbohydrate on blood glucose levels.
Covering the carbohydrate
Appropriate bolus insulin dosing requires individualization and adjustment. For starters, different amounts of carbohydrate require different-size bolus doses. To determine what size bolus dose is needed, first count up the grams of carbohydrate you are about to eat, then use something called an “insulin-to-carbohydrate ratio.” This ratio specifies how many grams of carbohydrate are “covered” by each unit of rapid-acting insulin. For example, a 1-unit-per-10-grams-of-carbohydrate (1:10) ratio means that one unit of insulin covers 10 grams of carbohydrate. A 1:20 ratio means that each unit covers 20 grams.
Calculating a meal or snack bolus becomes simple when you know your insulin-to-carbohydrate ratio. Simply divide the number of grams of carbohydrate you plan to eat by the second number in your ratio. If each unit of insulin covers 10 grams of carbohydrate and you consume 65 grams of carbohydrate, you will need 6.5 units of insulin (65/10 = 6.5).
The beauty of knowing your insulin-to-carbohydrate ratio is that it gives you the flexibility to eat as much or as little carbohydrate as you choose while still maintaining good blood glucose control. However, it is common to have different insulin-to-carbohydrate ratios at different times of day due to changes in hormone levels (which affect insulin sensitivity), physical activity (which enhances insulin sensitivity) and the amount of basal insulin overlapping with the bolus. For most people, insulin sensitivity tends to be a bit lower in the morning than later in the day. For example, I require a 1:10 ratio at breakfast, 1:12 at lunch, and 1:15 at dinner and in the evening.
Setting your ratios
There are two commonly used methods for determining initial, or starting, insulin-to-carbohydrate ratios: the 500 rule, and the weight method. Whichever method you choose, it is best to begin with a conservative approach to prevent hypoglycemia, then fine-tune as needed.
The 500 rule. This approach is based on the assumption that the average person consumes (in meals and snacks) and produces (by the liver) approximately 500 grams of carbohydrate daily. (The liver normally secretes a small amount of glucose at all times, and it produces larger amounts when blood glucose levels fall.) By dividing 500 by the average number of units of insulin you take daily (basal plus bolus), you should get a reasonable approximation of your insulin-to-carbohydrate ratio.
For example, if you take a total of 25 units of insulin in a typical day, each unit should cover approximately 20 grams of carbohydrate (500/25 = 20). If you take 60 units daily, your insulin-to-carbohydrate ratio would be 1 unit per 8 grams of carbohydrate (500/60 = 8).
The following list shows approximate insulin-to-carbohydrate ratios based on average total daily units of insulin (including both basal and bolus doses):
- 8–11 units 1:50
- 12–14 units 1:40
- 15–18 units 1:30
- 19–21 units 1:25
- 22–27 units 1:20
- 28–35 units 1:15
- 36–45 units 1:12
- 46–55 units 1:10
- 56–65 units 1:8
- 66–80 units 1:6
- 81–120 units 1:5
- >120 units 1:4
The obvious weakness to this approach is that it assumes all people eat about the same amount of food and produce the same amount of glucose each day. People who are heavy or who tend to eat relatively large amounts of carbohydrate will underestimate their insulin requirement with this approach; those who are lean or active or who eat relatively little will overestimate their requirements.
The weight method. This approach is based on the supposition that insulin sensitivity diminishes as body mass increases; hence each unit of insulin will cover less carbohydrate in a heavier person than in a lighter person. The following list shows approximate insulin-to-carbohydrate ratios based on weight (in pounds):
- <60 lb 1:30
- 60–80 lb 1:25
- 81–100 lb 1:20
- 101–120 lb 1:18
- 121–140 lb 1:15
- 141–170 lb 1:12
- 171–200 lb 1:10
- 201–230 lb. 1:8
- 231–270 lb. 1:6
- >270 lb 1:5
One of the potential problems with this system is that it fails to consider body composition. A person who weighs 250 pounds but is very muscular will be much more sensitive to insulin than a person of similar weight who has a great deal of body fat. It also fails to account for a person’s degree of insulin resistance. This presents more of a problem for people with Type 2 diabetes, who tend to be insulin-resistant, than for people with Type 1, who tend not be insulin-resistant.
It is best to establish your basal insulin levels before attempting to fine-tune your mealtime and snacktime bolus ratios. Any basal insulin changes made after fine-tuning your boluses will require additional bolus adjustments.
Fine-tuning bolus ratios is best done empirically, or through trial and error. You should verify your insulin-to-carbohydrate ratio for each meal and snack separately, since they can vary considerably.
Keep detailed written records when trying out different insulin-to-carbohydrate ratios. Check and note your blood glucose level before each meal and then again 3–4 hours later (to give the insulin a chance to work fully). Do not snack, exercise, or take more bolus doses between the two blood glucose readings. It is best to eliminate factors other than the food in the meal that might affect the results of your readings. For example, do not include data collected during or immediately after strenuous exercise. Don’t count data collected during an illness or major emotional stress, at the start of a menstrual cycle, or after an episode of low blood glucose (hypoglycemia). Meals with very high fat content or unknown carbohydrate content (such as restaurant meals) should not be used as part of your analysis.
Because strange things can happen to a person on any given day, I like to consider 10–14 days of data when deciding on the insulin-to-carbohydrate ratio for a given meal. This table shows what 12 days of data might look like. Based on the information in the table, I would assign an insulin-to-carbohydrate ratio of 1 unit per 12 grams of carbohydrate for this person’s breakfast. A ratio of greater than 1:12 tends to produce a blood glucose rise; less than 1:12 tends to produce a drop. When used, 1:12 held the blood glucose fairly steady, with lunch readings within 30 mg/dl of breakfast readings. I would throw out the data on 6/3 due to the low reading prior to breakfast. I would also throw out the data on 6/8; it is inconsistent with every other result, and the meal was much larger than usual.
Fine-tuning insulin-to-carbohydrate formulas can be a challenging proposition, so the more detailed your records, the better. You might discover a variety of factors that have a subtle influence on your blood glucose levels. Look for variations by day of the week, work or school schedules, time of the month, physical or recreational activities, changes in pump infusion sets or insulin vials or cartridges, injection or infusion sites, dining in versus eating out, and even social engagements.
Betty, for instance, had high readings every Sunday at lunchtime but normal readings the rest of the week. The reason? Church, most likely. Betty is very passionate about prayer. The lack of movement in church (she sits for several hours) coupled with the adrenaline surge she gets from the service is likely producing a consistent blood glucose rise. The solution: Use her usual 1:10 breakfast formula during the week, but increase it to 1:6 on Sundays.
Dan was experiencing very inconsistent blood glucose levels prior to dinner despite having the same lunch each day and using a consistent 1:15 bolus ratio at lunchtime. In reviewing his records, he found that most of his dinnertime lows were preceded by morning workouts; most of his dinnertime highs were preceded by no workout. The solution: Use 1:10 at lunch after sedentary mornings, but decrease it to 1:20 following morning exercise.
Given the complexities of determining bolus formulas, it is worth having a second set of eyes look over your records. Don’t hesitate to ask your physician or diabetes educator to review your data and help you to form reasonable conclusions.