Carbohydrate reference tables

Leamington Spa, March 2013

People with diabetes who take insulin can do so in a number of different ways. The most common regimens are either a twice-daily injection, a fixed basal-bolus schedule of four injections, or basal-bolus with 'carb-counting'.

The twice-daily injection is 'mixed' insulin, where a proportion of the shot is long-acting and the rest is short-acting insulin. The amount is fixed, so with this regimen it is important to eat regular amounts at regular times, because if you skip a meal or a snack or don't consume enough carbohydrate, you can experience low blood glucose - a 'hypo'.

The fixed basal-bolus regimen consists of a daily shot of long-acting insulin, and three fixed injections of rapid acting insulin, one with each main meal. This is a little bit more flexible than using the mixed insulin, because you can tweak it a little by adding in or omitting some extra rapid-acting insulin as a 'correction' to pre-meal blood glucose levels that are too high or too low. It is still necessary to eat regular and consistent amounts of carbohydrate.

'Carbohydrate counting' is the most flexible option, where you still need the once- or twice-daily long-acting insulin, but you match the rapid-acting insulin to the amount of carbohydrate you eat. It allows you to skip meals, eat extra meals, eat early or late, and eat as much or as little carbohydrate as you want, but in return you have to do some work. In order to match carbs and insulin, you have to estimate how much carbohydrate is in your food, convert that to units of insulin, sometimes add in or take away a correction dose, and that's what you inject.

Estimating carbohydrate content of food can be done in a few ways. Food labels are the best estimate, as long as the food has a label and you can work out how much of it you ate. A whole margarita pizza is 337g and the label says it contains 29.1g carbohydrate per 100g, and you ate about a third of it with salad - call it 30g carbs. But if you cooked six handfuls of raw macaroni and added tomato and cheese and shared it with your partner who ate a bit more than half and there's some left over, and the macaroni packet says there's 73g of carbohydrate per 100g raw macaroni and the tin of tomatoes is 400g at 3g carbs per 100g, and then you had a medium apple, then how much carbohydrate did you have?

There are now pictorial guides and smartphone apps that will help you to estimate visually how much carbohydrate is in various foods, but they cost money, and many people don't have smartphones either. So we give patients carbohydrate reference tables that we have compiled ourselves.

For historical reasons, our hospital has two different carb reference booklets. Neither is particularly comprehensive, and I have some major concerns about their inconsistency. I don't much like either of them, so when it was suggested that I take over the project to bring them together in a single updated version, I thought that would be a great idea.

I have spent a lot of time on four different supermarket websites. They are great; nowadays you can find out all sorts of nutritional information without actually having to pick up a packet of food and look at the label. I am discovering some interesting things - for example, the carbohydrate content of sliced bread is a common starting point, and it's easy to remember that a thick slice is about 20g carbohydrate, a medium slice is about 15g and a thin slice is 10g. What I discovered in my 'research' is that supermarkets no longer offer thin sliced bread. Who knew!

I've also been using the pictorial guides, and the Bible for nutritional information: McCance and Widdowson's The Composition of Foods (Sixth Edition) which contains not only the carbohydrate content but a million other details of the nutritional composition of almost anything edible that you can think of. I must be about a third of the way through the job of revising the tables, and there are some interesting questions that I'll take to the team for a decision. For example, I think we've already decided that, unlike previous tables, we won't be rounding the figures to the nearest 5g. But how many different breakfast cereals should be included, given that cereals always have labels with nutritional information? Should we bother to include any cereals at all?

It's an interesting exercise, it's useful for me to focus on carb content of all sorts of foods, and a handy task to have when patients don't turn up for their appointments. I'm looking forward to having an agreed final version that we can all trust.