Type 2 Diabetes: Is Carb Counting Unnecessary?

You’ve got type 2 diabetes. A few years ago, you started using a long-acting insulin once a day, and your fasting glucose levels and your A1c came down. But now your A1c is creeping back up. Your doctor tells you that you need to add a mealtime insulin to your plan.

And oh, by the way, your doctor says, you should learn advanced carbohydrate counting. Then she explains that it’s a little complicated and advises you to see a dietitian. The dietitian will teach you to estimate how much carbohydrate is in each of your meals, and you’ll adjust your insulin dose according to an insulin-to-carb ratio, which might be different at breakfast, lunch, and dinner. It’s what people with type 1 diabetes use to get good control, so it must be the best way for people with type 2 diabetes, right?

Maybe not.

A new study published in the July issue of Diabetes Care suggests that pattern management, with weekly adjustments of mealtime insulin doses, might work as well for people with type 2 diabetes as using insulin-to-carbohydrate ratios.

Researchers at four sites in the United States recruited 273 adults with type 2 diabetes to try one of two insulin-adjustment plans. The participants were already using two or more injections of insulin a day, and about a third were also taking metformin, but they still had blood glucose levels above goal. Their average age was 55, and most were overweight or obese.

All the participants were switched to multiple daily injections: one shot of a long-acting insulin (glargine) plus a rapid-acting insulin (glulisine) before each meal. Those who were taking metformin kept taking it.

The participants agreed to check their blood glucose levels four times a day and record the results. The goals were:

  • fasting (before breakfast): less than 95 mg/dl
  • before lunch and dinner: less than 100 mg/dl
  • at bedtime: less than 130 mg/dl

By random assignment, half the participants used pattern management and half used advanced carbohydrate counting.

All the participants started with a dose of glargine equal to 50% of their total pre-study insulin dose. Glargine was adjusted weekly to get to the fasting glucose goal of less than 95 mg/dl.

Pattern Management Group

In the pattern management group, the remaining 50% of the total insulin dose was split to cover three meals: 50% for the largest (most carbohydrate) meal, 33% for the middle-sized meal, and 17% for the smallest meal.

This group used these starting doses for a week. If there was a pattern of blood glucose levels out of range at the same time of day—more than half the values from the week were either too high or too low—the dose of insulin for the meal that came before was changed for the next week. (A person’s blood glucose level before lunch reflects whether the dose of insulin given at breakfast was in balance with the amount of carb in the breakfast.) With this plan, a person might take, for example, 8 units of glulisine at lunch every day for a week, no matter what he ate. If his blood glucose levels before dinner were too high four nights out of seven, he’d raise his lunchtime insulin dose to 9 units for the next week. (See details below.)

Carb Counting Group

The other participants used carb counting to figure their glulisine dose at every meal. They also used correction doses: If their blood glucose level before a meal was a little low or a little high, they lowered or raised the dose of insulin by a given amount.

They also used pattern management weekly: If there was a pattern of blood glucose levels out of range, they adjusted their insulin-to-carb ratio for the meal that came before and used the new ratio for the next week. For example, if a person was using a ratio of 1 unit of mealtime insulin for every 20 grams of carb and there was a pattern of highs before the next meal, she would increase her ratio to 1 unit for 15 grams.


The study lasted six months. Average A1c and fasting glucose levels decreased steadily in both groups. At the end, the average A1c of the people in the carb counting group dropped from 8.3% to 6.5%. Average A1c in the pattern management group dropped from 8.1% to 6.7%. About 70 percent of participants in each group reached an A1c of less than 7%, and almost half in each group reached an A1c of less than 6.5%. More people in the pattern management group finished the study: over 90 percent versus 80 percent of the carb counting group.

Maggie Powers, PhD, RD, CDE, a researcher at the International Diabetes Center at Park Nicollet in Minneapolis, says one of the lessons from this study is that patients and doctors need to “aggressively titrate” (adjust) doses of insulin “and not do it just every three or four months.” Says Powers, “If you titrate insulin weekly, you can get to goal A1c.”

On average, people in the pattern management group gained about eight pounds, while people in the carb counting group gained about five pounds. “It’s important to note that this was not a weight loss study,” says Powers. “The participants could eat whatever they wanted. They could have been eating dessert every night; no one told them not to.”

Not for Everyone

The researchers don’t suggest that everyone with type 2 diabetes will do well with this pattern management plan. They note it’s possible that this group of people just happened to be very consistent in their eating habits, with about the same amount of carb at a given meal. Or perhaps the participants consciously or unconsciously adjusted their carbohydrate intake based on the feedback they were getting from the blood glucose checks.

And the study results don’t mean that pattern management will work as well as insulin-to-carb ratios for people with type 1 diabetes. Indeed, the researchers write: “Insulin delivery based on carbohydrate counting is the gold standard for improving glycemic control in type 1 diabetes…”

People with established type 1 (in general, those who have had type 1 for more than a year) make little to no insulin of their own. To hit their blood glucose goals, they may need to be more precise in the doses of insulin they give themselves.

But many people with type 2 diabetes still make some insulin of their own. It may be that injected insulin just needs to get them in the general vicinity, and then their pancreases can do the fine tuning at mealtime.

Marie McCarren is the author of Carb Counting Made Easy, ADA Guide to Insulin & Type 2 Diabetes, and A Field Guide to Type 2 Diabetes.

Insulin Adjustments Used By the Pattern Management Group

Adjust glargine dose once a week.

Average of fasting glucose for previous 3 days (mg/dl)

Add ____ units to your glargine dose.

higher than 180

+ 8


+ 6


+ 4


+ 2


No change


If average fasting blood glucose of past three days is lower than 70 mg/dl, decrease dose of glargine by same number of units that glulisine dose is being raised that week.

Adjust mealtime insulin once a week.

If your mealtime (rapid-acting insulin) dose is:

In previous week, if there was a pattern of blood glucose levels below target, subtract ___units from your dose.*

In previous week, if there was a pattern of blood glucose levels above target, add ___ units to your dose.*

10 units or less

– 1


11-19 units

– 2


20 units or more

– 3



* The dose you adjust is the one for the meal before the pattern.

Before-lunch pattern: adjust breakfast dose

Before-dinner pattern: adjust lunch dose

Before-bed pattern: adjust dinner dose

(For a before-breakfast pattern, you adjust your long-acting insulin.)

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.