Skip to main content

09-04-2014 | Diabetes | Article

Baseline HbA1c is the best determinant of insulin response


Free abstract

medwireNews: Baseline glycated haemoglobin (HbA1c) level is the strongest predictor of the change in HbA1c that patients with diabetes will experience when either starting or switching insulin therapy, research shows.

While many other clinical, biochemical and demographic factors were also significantly associated with change in HbA1c, they tended to be interdependent and have poor predictive power (r2<0.010) and thus were of poor clinical utility, report Philip Home (Newcastle University, UK) and colleagues in Diabetes Care.

Home and team assessed the ability of more than 20 variables to predict and explain the change in HbA1c among 66,727 participants of the A1chieve trial, which examined the safety and efficacy of three different insulin analogue therapies in insulin-naïve and insulin-experienced patients with diabetes.

They found that, in univariate analyses, baseline HbA1c accounted for 68% of the variance in the mean 23 mmol/mol improvement in HbA1c levels observed at 24 weeks in patients starting insulin therapy. In addition, HbA1c levels explained 57% of the variance in the mean 19 mmol/mol improvement observed in those switching insulin therapy.

These associations improved to 70% and 58%, respectively, in multivariate analyses.

Baseline fasting and postprandial glucose levels, lipid markers, starting insulin dose and the use of oral glucose-lowering drugs also had some predictive power in univariate analysis (r2 = 0.011–0.116), but this power either markedly reduced or disappeared when other variables were accounted for in the multivariate analysis.

To assess the explanatory power of each variable, the researchers added within-study and end-of-study measurements to their statistical models. In this analysis, baseline HbA1c explained 69% and 54% of the variance in change in HbA1c levels for insulin-naïve and insulin-experienced patients, respectively.

The only other variable of any significance was the end-of-trial level of fasting plasma glucose, with an explanatory power of 9% for insulin-naïve patients and 11% for insulin-experienced patients.

Of note, body weight did not predict change in HbA1c. This is “clinically important”, say the researchers, because weight gain “limits insulin dose titration and thus attainment of improved glucose control.”

Hypoglycaemia and factors associated with long duration of diabetes also showed no association with the change in HbA1c level.

“Clinically, it seems that physicians can expect from these results that all markers of poor metabolic control may predict that larger improvements in glucose control can be achieved when starting insulin therapy, although, ultimately, it is enough to look at HbA1c level”, Home and co-authors conclude.

medwireNews ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2014

By Laura Cowen, medwireNews Reporter