Bolus advisor masterclass part 1: Counting fat and protein

Seville, November 2016

It's been a few weeks since I attended this course, and it's time to assemble my thoughts and learning points. It was a terrific day and I certainly learned quite a lot, most of it very relevant to my work with people with Type 1 diabetes. In fact, two days after the course I was passing on some of the information to the group I was teaching at the time.

It is important to note that the day was hosted by one particular pump manufacturer, Roche, so there is a likelihood of bias. Having said that, I think the majority of information supplied was correct - it would be fairly straightforward to check, although I haven't done so. The first topic covered in the day was a comparison between the specifications and capabilities of the different insulin pumps on the market. Then we focussed on the algorithm that each manufacturer uses to guide the user in the amount of insulin to give in various circumstances: for exercise, for high fat and protein meals, and to correct high blood glucose levels after a meal or snack.

Exercise

The guidance around exercise wasn't very different from what we already advise - if you're exercising within 90-120 minutes after a meal you could give less insulin for the meal; if not then you'll probably have to eat or drink some carbohydrate to prevent blood glucose dropping, and if you're using an insulin pump you've also got the option of reducing background (basal) insulin. I've written extensively and comprehensively on the knotty topic of exercise and Type 1 diabetes. It's a challenging area, and management.is very individual. All the pumps work in a similar way, although the Roche handset has some features that help with the mathematics of percentage reductions.

High fat and protein meals

OK, this is going to start getting technical (although nowhere near as nerdy as the section on post-prandial correction doses in part 2).

For people with Type 1 diabetes, the evidence suggests that best management of blood glucose levels, and therefore long-term health and freedom from diabetes-related complications, comes from matching insulin injected and carbohydrates consumed. (Just for contrast, the approach for people with Type 2 diabetes in the first instance is weight loss).

The first point that hit home during the course was that although we focus on counting the carbohydrate in a meal, there is a contribution to blood glucose that comes from the protein and the fat in a meal. When we focus on carb counting, there is an unspoken (and for me until now, unrecognised) assumption that the meal is constituted of a 'normal' proportion of carbs, fat and protein. I actually know this to be true, because we have found when someone with Type 1 goes on a very low carb diet they need more rapid insulin with meals than the carbohydrate content would suggest.

Fat and protein have another effect alongside their contribution to blood glucose levels - they also slow down the digestion of carbohydrates. The action profile of insulin can't be adjusted to suit the meal composition - rapid insulin has a fixed onset, peak and acting time that doesn't change, so giving insulin in the standard way before a meal when it's a high fat/protein meal doesn't work very well, and post-prandial blood glucose often ends up way higher than one would like.

Up to now, conventional wisdom suggested that for a high fat meal carbs should be counted as usual, but the matching insulin dose should be delayed and/or split (if on injections) or spread over a longer time period (if on a pump), to account for the delay in digestion and later peak in post-prandial blood glucose. This course not only suggested that more insulin is needed because of the greater contribution of fat and protein to blood glucose, but gave some useful guidelines on how much more is needed, and how it should be delivered.

Six different high fat/protein meals were listed: fish and chips, Indian and Chinese takeaway, pizza, pasta with creamy sauce and fast food (McDonalds, KFC etc).  The following procedure was recommended separately for each.

As a first attempt, you should give 25% more insulin than you need for the carbs and deliver 50% at the start of the meal, and 50% an hour later (if on injections) or over 2.5 hours (if on a pump). Then, to see if these percentages are right, you should monitor blood glucose at 2.5 hours and 6 hours without having any more food or insulin. This will work best if your blood glucose level is within the normal range before the meal.

The 2.5 hour test is to find out whether the 50/50 percentage split is right. If blood glucose at this stage is more than 4 mmol/L higher than it was before the meal, then more insulin is needed up front - they suggest increasing by 20% at a time (i.e. switching to a 70/30 split next time). Conversely, if blood glucose is lower at this point than before the meal, the split should be changed to 30/70.

The 6 hour test is to find out whether the 25% extra insulin is right. If blood glucose at this stage is between 2 and 6 mmol/L higher than before the meal, next time add on another 10% - instead of 25% extra, add 35%. If at 6 hours blood glucose is more than 6 mmol/L higher than before the meal, next time add on 20% (to 45% extra). Conversely, if blood glucose is lower at this point, then next time knock the percentage down by 10% to 15% extra.

Multiple experiments may be needed to get the best results! Apparently parents are often taken aback when their Type 1 children come home from a carb counting course and assert that they've been given homework that requires them to eat fish and chips, takeaways and fast food.

Coming soon: Part 2 will contain even more technical stuff about how to manage post-prandial correction of high blood glucose levels.