HbA1c variability predicts retinopathy, nephropathy in teens with type 1 diabetes
Maria E. Craig, MBBS, PhD, FRACP, professor of pediatric e ndocrinology, The Children’s Hospital at Westmead and The University of Sydney in Australia, and colleagues evaluated data from 1,706 adolescents (median age, 16.9 years) with type 1 diabetes examined for complications at The Children’s Hospital at Westmead between January 1990 and May 2014. Participants eligible for inclusion had diabetes duration of at least 5 years (median duration, 8.1 years) and had available data on more than five consecutive HbA1c measurements since time of diagnosis.
The researchers evaluated glycemic control through calorimetric measurements of HbA1c (before February 1994) and high performance liquid chromatography (after February 1994). The researchers calculated each participant’s intrapersonal mean and standard deviation (SD) of all recorded glycemic control measurements, and the SD-HbA1c was designated as a measure of glycemic variability. The researchers also calculated coefficient of variation, a normalized parameter of glycemic variability.
Retinopathy was evaluated by seven-field stereoscopic fundus photography, and albumin secretion rate was determine using the mean of three consecutive timed overnight urine collections. The researchers evaluated peripheral nerve function using thermal threshold testing for hot and cold sensation at the dorsum of the left foot, and vibration threshold testing at the left malleolus and left great toe. Cardiac autonomic neuropathy was evaluated through measures of heart rate variability acquired from analyses of 10-minute continuous electrocardiogram readings. The researchers defined peripheral neuropathy as a test score above the 95th percentile on either a vibration or thermal threshold test , and cardiac autonomic neuropathy was defined as a measurement below the fifth percentile on heart rate variability.
After adjusting for established risk factors , including HbA1c, diabetes duration, blood pressure and lipids, the researchers used generalized estimating equations to assess the relationship between complication outcomes and HbA1c variability.
Multivariable analysis revealed an association between SD-HbA1c and retinopathy (OR = 1.32; 95% CI, 1-1.73), albuminuria (OR = 1.81; 95% CI, 1.04-3.14), increased log10 albumin excretion rate (OR = 1.1; 95% CI, 1.05-1.15) and cardiac autonomic neuropathy (OR = 2.28; 95% CI, 1.23-4.21).
Each 1- unit increase in SD-HbA1c correlated with 32% higher odds of retinopathy, 81% higher odds of albuminuria, 128% higher odds of cardiac autonomic neuropathy and 10% increase in log10 albumin excretion rate.
No association was seen between peripheral neuropathy and glycemic variability, in analyses of both SD-HbA1c and coefficients of variation.
“Our most novel finding was the significant association between glycemic variability and [cardiac autonomic neuropathy],” the researchers wrote. “A 1-unit increase in SD-HbA1c more than doubled the odds of [cardiac autonomic neuropathy]. The effect size was greater than that observed for any other microvascular complication, and it is of substantial clinical significance considering that [cardiac autonomic neuropathy] is linked with increased mortality and a higher risk of sudden cardiac death.”
– by Jennifer Byrne
