Basal-bolus insulin therapy for Type 2 Diabetes   (July 12, 2008)

The journal Diabetes Care this month has a fascinating study about basal-bolus insulin therapy for T2DM, titled Adjust to Target in Type 2 Diabetes: Comparison of a simple algorithm with carbohydrate counting for adjustment of mealtime insulin glulisine. The authors compared two algorithms for adjusting mealtime rapid-acting (bolus) insulin and background (basal) insulin in 273 patients with type 2 diabetes over a period of 24 weeks. The simple algorithm group was provided set doses of bolus insulin to take before each meal. The other group was provided an insulin-to-carbohydrate ratio to use for each meal and adjusted their bolus dose based on the amount of carbohydrate consumed ("carb counting").

Before you say "Huh?", let me do some explaining about basal-bolus insulin therapy.

Most people are aware that a single shot of a single type of insulin given once daily isn't enough to accurately mimic the pancreas' output of insulin. So programs of insulin administration for the past thirty years or more have emphasized the need for multiple daily injections of different types of insulin. Most of these programs are set in concrete, with little variability in doses possible from one day to the next. Sometimes two shots a day are given, mixing two types of insulin into each shot. Sometimes three shots a day, with a mixture in the morning, a supper dose of rapid-acting insulin, and a bedtime dose of long-acting insulin. Usually, doses are adjusted every few weeks or months, based on overall trends of blood glucose and A1C. These programs, and many other variations, are still widely used for type 2 diabetes, where it's assumed there's still some residual pancreatic production of insulin that can help maintain the blood glucose levels.

But in type 1 diabetes, where the pancreatic beta cells have failed almost completely, and can't make and release any worthwhile amount of insulin, it's become standard therapy to use what's called "basal-bolus" insulin therapy. Basal-bolus insulin can be given by using rapid-acting insulin in an insulin pump, or by using injections of rapid-acting insulin (Humalog, Novolog, or Apidra) at every meal, together with long-acting insulin (Lantus, Levemir, or sometimes NPH  or Ultralente) injected once or twice daily.

The basal dose is designed to mimic the background, minute-by-minute release of insulin by a normal pancreas. The basal dose is usually kept stable from day-to-day, and adjusted periodically based upon general trends and specific markers, such as middle-of-the-night and/or dawn blood glucose levels.

The bolus dose is designed to mimic the release of insulin in response to meals by a normal pancreas.

In basal-bolus therapy, the bolus dose that is given may vary from meal to meal, based on numerous factors, mainly the amount of carbohydrate in that meal, but also the blood glucose level at that moment. Sophisticated dose adjustments might also take into account more factors, such as how much insulin is on board from a previous meal, any planned upcoming exercise, stress levels, and even time of day. Sounds pretty complex, and indeed it is.

And simplification of the rules for bolus insulin administration can be utilized. The simplest guide would be to give a small amount of insulin if eating a small amount of carb, a moderate amount of insulin if eating a moderate amount of carb, and a large amount of insulin if eating a large dose of carb. And no insulin if no carb. Deciding on what's "small," "moderate," or "large" would be individually decided by the patient together with the physician. This simplification is a good place to start when introducing the system to new patients.

Now, back to the study. Although basal-bolus therapy is pretty standard in T1DM, there's not been much written about using basal-bolus insulin programs in type 2 diabetes. The authors decided to try evaluating a standard carb-counting vs. a simplified version of basal-bolus treatment in people with T2DM who were already taking insulin, with or without metformin. Over 24 weeks, they found that both programs work to lower A1C. The A1C had been in the range of 7–10% at screening, and the average A1C levels at week 24 were down to 6.70% (simple algorithm) and 6.54% (carb counting).

On the downside, severe hypoglycemia did happen, although rates were considered to be low and equal in the two groups (53 episodes of severe hypoglycemia in 19 patients on the simple algorithm, and the carb count group had 37 episodes in 19 patients). And weight increased in both groups: simple algorithm 3.6 kg (3.4%) and carb count 2.4 kg (2.3%).

I think this study is fascinating, as it demonstrates that basal-bolus insulin therapy can be an alternative for people with T2DM who need to start insulin therapy, and unsurprisingly, it shares the same concerns as any other program of insulin therapy: the risks of hypoglycemia and weight gain. The reported improvement in A1C over the short term (24 weeks) in people who were failing on other insulin programs was dramatic. How much of the improvement was due to basal-bolus therapy, vs. other parameters (such as increased adherance to meal plans and blood-glucose monitoring schedules when in a study) is unclear, but it's dramatic none-the-less.

My advice, if you are taking insulin for either T1DM or T2DM: if you are not on a basal-bolus program, and want better glucose control, and you're willing to do the extra work (giving extra shots, doing frequent blood-glucose testing, paying attention to the carbs in the food you eat, and learning to adjust insulin doses based on the carb content of the meal), it's worth thinking about switching to a basal-bolus program.