Meeting Coverage ADA: Weak Support for Carb Counting in T1D

CHICAGO -- The standard practice of counting carbohydrates to better manage postprandial glycemic control in type 1 diabetes may not have substantial evidence behind it, researchers reported here.

Action Points
This study was published as an abstract and presented at a conference. These data and conclusions should be considered to be preliminary until published in a peer-reviewed journal.

The standard practice of counting carbohydrates to better manage postprandial glycemic control in type 1 diabetes may not have substantial evidence behind it.

Note that the review found no changes in insulin dosing, weight, or fasting plasma glucose.

In a review and meta-analysis of six randomized, controlled trials, carbohydrate counting slightly improved glycemic control with a mean drop in glycated hemoglobin (HbA1c) of -0.3%, but it wasn't significant (P=0.185), Kirstine Bell, a PhD candidate at the University of Sydney, reported at the American Diabetes Association meeting.

"There's limited evidence to recommend carbohydrate counting over other dietary interventions for improving glucose control in type 1 diabetes," Bell said. "And it's concerning that there are only six studies that could be identified when this is such an integral element of diabetes management in type 1 diabetes. It affects the management of millions of people around the world."

Carbohydrate counting is the best known method for matching insulin dosing to meals, and is the recommended dietary strategy for achieving glycemic control in type 1 diabetes, though that recommendation has been largely based on expert consensus, Bell said.

To assess the state of the literature, she and colleagues conducted a review and meta-analysis of carbohydrate counting trials in adults and children with type 1 diabetes. They ultimately included six studies with a total of 563 adults and 104 children conducted between January 1980 and April 2012.

Overall, there were five adult studies and one pediatric trail that included children, ages 8 to 13. Five trials had a parallel group design while one involved a cross-over.

All trials lasted at least 3 months and compared carbohydrate counting with usual care, which consisted of either general nutrition advice or low dietary glycemic index (GI) advice.

For their analysis, Bell and colleagues set the primary outcome of change in HbA1c, and secondary outcomes included hypoglycemia, insulin dose, weight change, fasting plasma glucose, and quality of life.

Studies were generally of good to high quality, with an overall quality score of 7.7 out of 13, the researchers reported.

Bell and colleagues found a decrease in mean HbA1c with carbohydrate counting compared with usual care (-0.3%), but it wasn't significant (P=0.185).

They were unable to conduct an analysis of secondary outcomes because of the low number of studies and inconsistencies in their reporting metrics.

They did, however, find a trend toward reduced risk of hypoglycemia and improved quality of life with carbohydrate counting.

On the other hand, there were no changes in insulin dosing, weight, or fasting plasma glucose.

After I calm down, I'll probably have more to say on this subject, but for now: keeping close track of carbohydrates, and matching them to insulin, directly affects both my insulin dosages and my postprandial readings, and I have the measurements to prove it.

I totally agree, David. Makes you wonder who the heck they were studying!?

No worries, don't buy a single word of this. Conducting studies is the bread & butter for clinicians regardless of whether the conclusions actually make sense. I'm sure they think they've done a good job, lol!

I'm sure they do. You've nailed their psychology precisely. Obviously they've overlooked -- or failed to consider -- some rather important variables somewhere along the way. This perfectly illustrates why it's important not to take any single study of anything as gospel.

Perhaps a short restatement. The meta-analysis found a real decrease in A1c but we can't be sure if carb counting was the reason because it wasn't statistically significant. The study found no effect on fasting blood sugars - perhaps because fasting blood sugars are measured when there isn't any food on board. The study didn't find any effect on weight change but then it didn't find any effect on shoe size either. But carb counting did reduce the risk of hypoglycemia and improved quality of life. It sounds like a ringing endorsement to me :-)

Can't entirely concur.

The standard practice of counting carbohydrates to better manage postprandial glycemic control in type 1 diabetes may not have substantial evidence behind it.

That doesn't quite rise to the level of "ringing endorsement" for me.

My own postprandial results run counter to that statement, as mentioned above.

I was being ironic/sarcastic. The summary sounds like it was written by someone who never had to think like a pancreas.

ROFL. I'm one of the most sarcastic people I know, and it went completely over my head. DUH.

What exactly is the alternative to carbohydrate counting? Going back to exchanges? In either case, people have to pay attention to what they eat and nutritional values of food and calculate an insulin dose.

I think reducing the risk of hypoglycemia and improving quality of life should be huge positives. Who cares about weight and insulin dose. Unless someone is trying to lose weight, why would those even be considered important outcomes?

Sometimes I don't understand why these types of studies are even conducted.

The one criticism of carbohydrate counting I do have is that there is no meal planning whatsoever, which I think was really helpful. As a kid I used to see the dietician once a year to go over meal plans and exchanges; as an adult I've seen the dietician once, and it was at my own request because I was trying to lose weight. Even if a dietician doesn't know everything, it can be helpful to sit down and talk about your diet with someone on a regular basis.

Sounds like a flawed study... I just wonder what those other methods are for dosing your insulin? Do you go by calories or just pull it out of your hat?

The old system used before the 2000s was food exchanges — basically you had a list of foods divided into categories such as starches/grains, dairy, fruit/vegetables, sweets (I can't remember all of them, but I believe there were seven) with each category containing foods listed in 15g servings. You would get a meal plan with so many "exchanges" from each group. For example, breakfast might be two starch exchanges (30g carbs), one milk exchange (15g) and one fruit exchange (15g). So then you could decide if you wanted toast with a glass of milk and half an orange, or oatmeal with milk and fruit, or cereal with milk and some fruit, and so on for each meal. Essentially, you took the same insulin dose for each meal, but had to make sure you had the right type and amount of exchanges.

To correct high blood sugar, you would use a sliding scale where you got a scale of blood sugar numbers and corresponding extra units to give on top of the meal dose based on which range your pre-meal blood sugar fell into. For example, you might give 1 extra unit if you were 150-200, 2 extra units if you were 200-250, 3 extra units for 250-300, and so on.

It still required just as much work and calculation as carbohydrate counting, in my opinion, but was less flexible and less precise.

1 Like

You mean I shouldn't eat at McDonald's, lol. Seriously, though, you are helping me see some possible value out of the study, the idea being that A1C isn't strictly about carb counting but also the type of carbs i.e.healthy vs junk food. I'll stop there, I am in danger of giving the study too much credit!

Take a deep breath.

Everybody here who counts carbs probably does it as one part of a complete diabetes treatment plan to get the results that we do. If you really think that counting carbs all by it's lonesome should have an impact on blood sugar control, try a simple test:

Count up the carbs in a bag of potato chips then try your best to dose for it with insulin. See how that works out for your Blood Sugar control.

That's basically all the analysis says. It confirms everything we discuss on a daily basis regarding the dedication it takes to properly control diabetes and also confirms, to me anyway, that he typical poster here is way out on the nether ends of the bell curve as far as dedication goes.

We do are best to control a gazillion variables a day to maintain BG control. You can't expect any one thing to be the silver bullet.

I don't take issue with the findings because the findings actually make sense. I do take issue with the conclusions quoted by the PI, as well as the sensationalism of the title and headlines. The findings alone, by no means, illegitimizes carb counting or suggests you should go back to the exchange system. It just confirms what we already know. You have to do a lot more to see good results.

That is all true as far as it goes. My fundamental problem with this is that it's too glib by half and far too lacking in intellectual rigor. It as much as admits that there are many variables not accounted for, then goes right ahead anyway, implying conclusions about what was measured. Presuming that someone was paid to do this, we deserve something more exacting.

Thanks Jen,

I think the dietician at my endo mentioned something about exchanges being listed on foods too and various people seem to think I'm going to be on a sliding scale and I had no idea what they meant really. I just tell them I'm calculating carbs etc. At the hospital, my second time, when I was on injections, they weren't going to give me any insulin unless my bg went above 140, I ended up just doing all my own testing and bolusing and they did my basal. It does sound a lot less precise to have a one size fits all correction number since everyone is so different. I could take 2 units sometimes and only come down 40 points.


But carb counting did reduce the risk of hypoglycemia and improved quality of life. It sounds like a ringing endorsement to me :-) NUff said still_young!!

Sarcasm well

God bless

▼ Reply

You have to do a lot more to see good results So true , David.And most of us here on tu d are willing to do that "Lot more", unlike whoever the diabetics they used in the study Good control is our goal, in the DOC. Most diabetics I know of out side of the DOC do not desire to do that much, nor have they been given the tools/education to get good control. The "advice" that they are being told by their health care team is that sliding scale and exchanges are the only options.
God bless,

The results of this study should obviously be taken with a grain of salt. Think about it, the basal/bolus regime was only invented in 1972 and this study involved studies conducted between 1980 and 2012. The usual standard of care for many of these patients probably was conventional insulin therapy with a 70/30 mix and no adjustment of mealtime insulin. At best, the usual are probably involved mealtime doses. And carb counting is meaningless when you are using fixed doses. Faced with that problem, the results cannot provide us with any measure of how effective carb counting is when used in intensive insulin therapy.

That being said, the author Kristine Bell is doing some interesting work. At the same meeting she presented her work on a food insulin index that accounts for fat and protein in meals. Many of us do account for fat and protein in our meals. There is a group here devoted to Total Available Glucose (the TAGers) and Bernstein for years has recommended counting up to half of protein as available glucose.

For more information on this study as well as an interview with the author, see this report over at Clinical Endochronolgy News.