I wrote about my use of the dual wave/ combo bolus on another forum. I have copied and pasted my post below:
I have been doing something similar for about the last 3 years.
From what I have read in this thread, it sounds like most of you who are using this technique are doing so by calculating a carb bolus based strictly on the carbs in the meal and stretching out some or all of the carb bolus using the square wave/ extended bolus.
I use a bolus calculation technique that is usually referred to as TAG (Total Available Glucose). The premise behind TAG is that a portion of the protein and fat content of the meal will also contribute to carb loading above and beyond the actual carbs in the meal. Whereas 100% of the carbs are accounted for when calculating the bolus, only a percentage of the protein and fats are expected to be eventually converted to glucose. The carb bolus is delivered immediately and the protein and fat bolus is delivered using the square wave/ extended bolus.
I have found through trial and error that for my metabolism, approximately 40% of ingested protein and fats are eventually converted to carbs (my percentages have changed over time as I played with the technique). I total the grams of protein and fat and multiply by 40%. As an example, if the protein and fat total to 80 grams, I would multiply by .40 and come up with 32 grams. With an i:c ratio of 1:10, this 32 grams of protein and fat carb loading would need to be covered by 3.2 units of insulin, in addition to the carbs in the meal. Since proteins and fats are digested more slowly than carbs, the 3.2 units have to be delivered using a square wave/ extended bolus to
prevent a person from dropping too low.
How do I time the square wave/ extended bolus? Again, through much trial and error I find that a delivery rate of approximately 1 - 1.2 units of insulin per hour delivery rate for the square wave/ extended bolus keeps me from dropping too low or spiking too high. So in this example, I would deliver the 3.2 units over a time span of 3 hours (3.2 units per hr/ 3 hrs = delivery rate of 1.06 units per hr, which is within the 1 - 1.2 target delivery rate. The 1 - 1.2 delivery rate is used assuming I was in BG target range to begin with. If my BG is above target, I would use a delivery rate higher than this range or transfer part of the protein/ fat bolus to the carb bolus for immediate delivery. If my BG is below range, I would use a delivery rate which is lower.
As you all have noticed, I also experience a much smaller spike after eating a meal when using this technique. Typically, I see my BG spike as little as 20 - 30 points after eating a pasta meal which will have over 100 grams of carbs. After such meals it is common to see 1 hr post prandials of 100 - 120 and 2 hr post prandials of less than 100. If I delivered the insulin correctly, I will be at or near target several hours after the meal, without ever going low.
TAG is NOT taught or recognized by most medical professionals I have talked with. My TAG percentages are essentially a personalized algorithm for my metabolism of a meal and should not be seen as percentages that anyone else can pick up and run with.
Why does the technique work? Better absorption may be one reason the technique works. My explanation however is that the square wave/ extended bolus on top of a carb bolus simulates secondary phase insulin release, which is how a non-diabetic pancreas would handle a meal. Having once tried Symlin, this technique works better to control my post-meal spikes. However, if you are not sensitive to protein/ fat carb loading, the technique may not work well for you. My understanding is that TAG will work best for someone whose second phase insulin release is less than satisfactory. Unfortunately, most of us are simply given a catch all diagnosis of "diabetes" without any further classification of inadequate first and/ or second phase insulin release.
You can read a lay person's article about second phase insulin release at the following link: