Hypoglycemia Happens at All HbA1c Levels

Diabetes patients at all levels of glycemic control reported experiencing severe hypoglycemia, researchers found.

Incidence of severe hypoglycemia was common in 10.8% of diabetic patients surveyed, with rates of hypoglycemia ranging from 9.3% to 13.8% across all levels of HbA1c, according to Kasia Lipska, MD, of Yale University, and colleagues.

Compared to a baseline HbA1c of 7% to 7.9%, no level of HbA1c -- ranging from less than 6% to 9% or higher -- had a significantly different relative risk of hypoglycemia in a model fully adjusted for age, diabetes duration, and category of diabetes medication, they wrote online inDiabetes Care.

The authors noted that prior research has shown that severe hypoglycemia risks were inversely associated with HbA1c levels. However, another recent analysis also showed that poorer glycemic control was common to those with higher HbA1c levels.

In addition, the investigators pointed out, few studies have indicated any conclusive relationship between patients in usual care settings and changes in glycemic control.

Lipska and colleagues studied the relationship between HbA1c levels and hypoglycemia through a sample of 9,094 type 2 diabetes patients ages 30 to 77, who were treated with glucose-lowering medications in a usual care setting and who participated in the DISTANCE (Diabetes Study of Northern California) survey.

The authors hypothesized that severe hypoglycemia risks were elevated in patients with either very high or very low HbA1c levels, and that lower HbA1c levels were more strongly associated with hypoglycemia risk in older versus younger patients and patients with longer versus shorter disease duration. They also expected variance in HbA1c and hypoglycemia risks by glucose-lowering regimen.

The survey assessed incidence of diabetes complications, including self-reported severe hypoglycemia, and sociodemographic factors. These data were linked with clinical, pharmacy, and laboratory data from participants' electronic medical records.

The primary outcome was occurrence of at least one event of severe hypoglycemia in the year prior to the survey. Researchers also used data on the last HbA1c level collected in the 12 to 24 months prior to the window for the severe hypoglycemic event. HbA1c levels were categorized as near-normal glycemia (less than 6%), very good (6% to 6.9%), good (7% to 7.9%), suboptimal (8% to 8.9%), and very poor (9% or greater).

Duration of diabetes was based on self-report and was categorized as "10 years or less" or "more than 10 years."

Participants were evenly split between men and women. Whites made up the largest ethnic group (25.7%). The majority of patients were older than 50, with 32.2% of participants ages of 50 to 59 and 33.7% of participants ages 60 to 69.

Of the 10% of patients reporting hypoglycemia, most were female (55.4%). These patients tended to be older (35.1% were ages 60 to 69) and were most often Latino (22.9%), white (22.3%), or were categorized as "multiracial/other/missing" (17.6%).

Participants with diabetes for more than 10 years were significantly more likely to report severe hypoglycemia than those with a duration of 10 years or less (13.9% versus 8.3%,P<0.001).

Hypoglycemia was most common in patients receiving insulin therapy (19%) and least common among those using nonsecretagogue glucose-lowering therapies (5.8%). Despite those differences, there was no significant relationship between HbA1c and age, diabetes duration, or category of diabetes medication.

Though prior to adjustment there was a U-shaped curve for risk of severe hypoglycemia based on level of HbA1c, the adjusted relative risk showed there was no significant difference between level of HbA1c for risk of severe hypoglycemia.

"Contrary to conventional wisdom, hypoglycemia occurs just as frequently among those with poor glycemic control as it does in those achieving near-normal glycemia," they concluded.

The authors noted that their study was limited by lack of confirmation of hypoglycemia, the retrospective study design, age cutoff of 77, and lack of causal pathways for associations in the study outcomes.

http://www.medpagetoday.com/Endocrinology/Diabetes/40764

Graham