Published: Oct 31, 2013 | Updated: Oct 31, 2013
By Salynn Boyles, Contributing Writer, MedPage Today

Full Story:  http://www.medpagetoday.com/Endocrinology/Diabetes/42610

Counting carbohydrates is an accepted strategy among patients with type 1 diabetes for managing postprandial blood sugar levels, but high-quality studies showing the practice to be effective arelimited, researchers reported.

A newly published meta-analysis that included six studies in adults and one in children found no significant improvement in HbA1cconcentrations overall in carb counters versus control- or usual care groups (-0.35% [-3.9 mmol/mol], 95% CI minus 0.70-0.06; P=0.096).

Five of the studies found carbohydrate counting to be beneficial, while two suggested that less quantitative methods of determining insulin needs were equally effective or superior to the practice, researcher and biochemistry professorJennie Brand-Miller of the University of Sydney, in Sydney, Australia, and colleagues wrote online in The Lancet Diabetes & Endocrinology.

They concluded that alternative methods of establishing prandial insulin dosing should be explored.

“It is possible that other methods of matching insulin with food are not being studied because of the belief that carbohydrate counting is a well founded, evidence-based therapy,” the researchers concluded. “Indeed, this meta-analysis shows the scarcity of high-level evidence.”

Carb counting involves targeting bolus insulin dosing to the total carbohydrate content of a meal, based on the premise that carbohydrates are the predominant macronutrient contributing to the rise in postprandial glycemia, the researchers wrote.

It is widely recommended even though its efficacy compared to other methods for assessing insulin needs has not really been shown, Brand-Miller told MedPage Today.

In an effort to better understand the available evidence, Brand-Miller and colleagues conducted a systematic review and meta-analysis of randomized controlled trials comparing carbohydrate counting with general or alternative dietary advice in type 1 diabetics.

Their search of all relevant databases, including Medline, Embase, the Cumulative Index to Nursing and Allied Health Literature, Web of Knowledge, and the Cochrane Central Register of Controlled Trials identified 311 studies, with 18 considered potentially relevant.

Of these studies, eight were excluded due to the lack of a control group, and three were excluded for other reasons. The remaining seven studies totalled 703 type 1 diabetics (599 adults and 104 children).

Six of the studies had parallel designs and one had a cross-over design with three interventions. Just one involved children, who ranged in age from 8 to 13.

An assessment of the quality of the studies showed an average score of 7.6 on a 1 to 13 point scale. All seven studies used appropriate randomization strategies. but none had adequate intervention allocation concealment and none masked the intervention from the outcome assessor, the researchers noted.

A subanalysis of five parallel design studies in adults did show carbohydrate counting to be more useful for lowering HbA1c than standard care (0.64% point [7.0 mmol/mol] HbA1c reduction; 95% CI minus 0.91-minus 0.37: P<0.0001).

“This was a significant difference in HbA1c that most people would consider clinically important,” Brand-Miller said. “But the question remains, ‘Is five studies with flawed study designs enough to make carbohydrate counting the gold standard?'”

She added that the evidence is clear that carbohydrate counting can provide modest improvement in glycemic control over usual care, but it is not clear if other strategies — such as eating a low-glycemic index diet — may lead to better patient outcomes.

Preliminary data from the study were presented in Chicago in June at the 73rd Scientific Session of the American Diabetes Association.