Diabetes Carbo-Counting for Prolonged Activity

Article for IDAA Newsletter December 14, 1999

Calculating an Insulin-Carbohydrate Ratio for Prolonged Exercise

Over the last few years I have struggled to find a way to manage my insulin and blood sugar levels while engaging in prolonged exercise in the backcountry. At last, I have managed to devise a method that is applicable to all insulin-dependent diabetics, regardless of whether they are infusion pump users or multiple needle injection users.

My form of exercise tends to be prolonged activity (e.g., day hikes, backpacking, scrambling, backcountry skiing, mountaineering trips, etc.) that lasts for 6 - 14 hours. I have tried many different insulin reduction strategies through the advice of dieticians and endocrinologists (e.g., "eat 15 g of carbohydrate every hour" or "try reducing your lunch time insulin bolus by 3 units") but none of these methods have worked very well for me. My blood sugars would often fluctuate between extremes and I never would know whether my blood sugar levels were "on their way up" or "going down". Eating 15 g of carbohydrate every hour was nearly impossible since often the weather conditions were unsuitable, the location at that moment was hazardous (e.g., high avalanche risk), the party was all roped up so everyone would have had to stop, the rest of the group may have gone on ahead, or there was that usual "scenic" viewpoint (great place to take a break) just 20 minutes ahead.

For the last few years I have practised basic carbohydrate counting, instead of following a set meal plan consisting of meal exchanges. At present, my typical daytime insulin to carbohydrate requirement works out to be approximately 0.9 units of insulin for every 10 g of carbohydrate. The medical staff seem to discuss it in the reverse order (e.g., as 1 unit insulin per 11 g carbohydrate) but I prefer to simply multiply my grams of carbohydrate by a ratio figure of 0.9 units/10 g carbohydrate. For example, if my lunch consisted of a bagel (typically 45 g of carbohydrate) and an apple (typically 20 g of carbohydrate), the total amount of carbohydrate for that meal would be 65 g. To calculate the required amount of insulin for the meal, I would multiply the 65 g carbohydrate by 0.9 units insulin/10 g carbohydrate which equals 5.9 units of insulin (65 x 0.9/10 = 5.9).

Since I can eat any amount of food at any time of the day in my regular daily lifestyle and maintain good blood glucose control, I have been convinced that I can also use this carbohydrate-counting method for outdoor physical activities. Through trial and error, I have determined that my required insulin ratios (both basal and bolus) during prolonged exercise should be about 50% of my normal daily requirements. My hourly basal rate is reduced to 0.3 units/hour (from the regular rate of 0.6 units/hour) and my mealtime bolus ratio is reduced to 0.5 units insulin/10 g carbohydrate for snacks or meals (as opposed to my regular bolus ratio of 0.9 units/10 g carbohydrate). For example, if a snack worth 20 g of carbohydrate was eaten during prolonged exercise (see chart below), the required insulin bolus would be 1.0 unit insulin (20 g carbohydrate x 0.5 units/10 g carbohydrate = 1.0 unit), instead of the usual 1.8 units of insulin (20 g carbohydrate x 0.9 units/10 g carbohydrate = 1.8 units).

Typical Scramble/Climb on April 25, 1999:

Time	Blood Sugar	Carbo-hydrate consumed (g)	Analysis of Situation	Insulin Boluses	Hourly Basal Rate on Infusion Pump
9 am	11.4 mmol/L (. 205 mg/dL)	55 g	About 55 g carb. required to start off my exercise and counteract any earlier insulin "still in my system".	-	Reduced to 0.3 units/hour for next 9.5 hours
10 am	5.9 mmol/L (. 106 mg/dL)	20 g	Group took first break. Ate 20 g of carbohydrate. (20 g carb. x 0.5 units/10 g carb = 1.0 unit insulin)	1.0 unit
11 am	7.9 mmol/L (. 142 mg/dL)	30 g	Early lunch break. Ate 30 g carbohydrate. (30 g carb. x 0.5 units/10 g carb. = 1.5 units insulin)	1.5 units
2.30 pm	5.1 mmol/L (. 92 mg/dL)	60 g	Snack before final stretch to summit. Ate 60 g of carbohydrate. Since I’m on a sliding scale, subtract 1 more unit since blood sugar is down to 5.1 mmol/L. (60 g x 0.5 units/10 g carb. - 1.0 unit = 2.0 units insulin)	2.0 units
4.30 pm	6.4 mmol/L (. 115 mg/dL)	50 g	Snack after getting down from summit. Ate 50 g of carbohydrate. (50 g x 0.5 units/10 g carb. = 2.5 units insulin)	2.5 units
6.30 pm	4.7 mmol/L (. 85 mg/dL)	50 g	End of day so no need to reduce insulin. Ate 50 g of carbohydrate. Since I’m on a sliding scale, subtract 1 more unit since blood sugar is down to 4.7 mmol/L. (50 g x 1.0 units/10 g carb. - 1.0 unit = 4.0 units insulin)	4.0 units

Using this carbohydrate counting method, I take many small doses of insulin throughout the day, rather than a few large doses, so my blood glucose levels tend to be much more stable with less extreme fluctuations. What this method does require, however, is frequent blood glucose monitoring (ideally every 1 - 3 hours) so that, in fact, one is constantly "correcting" the blood sugar level if it is slightly high or low.

Also, I have managed to calculate different insulin/carbohydrate ratios for different levels of prolonged physical activity. On a "regular" day I need 0.9 units insulin/10 g carbohydrate; if walking around the city I need 0.7 units insulin/10 g carbohydrate; and if doing highly aerobic backcountry exercise I need 0.5 units/10 g carbohydrate. The level of aerobic activity determines the amount of reduction in the basal and bolus insulin/carbohydrate ratios.

In summary, by using the carbo-counting method with an appropriate insulin/carbohydrate ratio, I now have fairly stable blood sugar levels for a variety of prolonged physical activities. Ultimately, each person must determine the absolute insulin/carbohydrate ratio reductions for him or herself. The intent of this article was to suggest a method of dealing with exercise and insulin levels and may not be appropriate for everybody.

"Calculating a Carbo-Insulin Ratio for Prolonged Exercise." was published in: The Challenge: Newsletter of the IDAA. Volume XIII, Number 4. Winter 1999. pp.12 - 13.