The impact of frequent and unrecognized hypoglycemia on mortality in the ACCORD study

Seaquist ER, Miller ME, Bonds DE; for the ACCORD Investigators.
Diabetes Care. 2012;35(2):409-414. 

This overview was created under the auspices of KnowledgePoint360 Group, LLC, and was not associated with funding via an educational grant or a promotional interest. 

Seaquist and colleagues examined data from an Action to Control Cardiovascular Risk in Diabetes (ACCORD) substudy to explore the relationship between frequent and unrecognized hypoglycemia and mortality. Subjects in the ACCORD study had type 2 diabetes; previous cardiovascular (CV) event, evidence of atherosclerosis, left ventricular hypertrophy, albuminuria, and at least two additional CV risk factors; and A1C 7.5%–11%. Subjects were randomized to standard (achievement of A1C 7.0%–7.9%) or intensive (achievement of A1C <6.0%) therapy. Those with frequent or severe hypoglycemic events excluded from the study population. ACCORD’s primary endpoint was a composite of CV mortality, nonfatal myocardial infarction, or nonfatal stroke. 

In this subanalysis of ACCORD, data from 10,096 subjects who had follow-up information related to hypoglycemia and mortality were examined. At all study visits, subjects were queried regarding episodes of low blood glucose, and symptomatic, severe hypoglycemia that required medical assistance; an assessment of the incidence of hypoglycemic episodes that did not require medical assistance and hypoglycemia unawareness at each 4-month visit was also initiated. Responses from these questions were analyzed, and hazard ratios that related risk of death to the updated annualized number of hypoglycemic episodes and the updated annualized number of intervals of hypoglycemia unawareness were obtained using Cox proportional hazards regression models. These models examined four variables: (1) annualized number of hypoglycemic episodes, (2) previous hypoglycemia requiring medical attention, (3) interaction term between previous hypoglycemia requiring medical attention and annualized number of hypoglycemic episodes, and (4) annualized number of intervals of hypoglycemia unawareness. Each of these models controlled for the baseline covariates that predicted mortality in the ACCORD cohort. These covariates included demographic and anthropometric characteristics (ie, age, body mass index [BMI], educational level, and race), medical history (ie, smoking, CV disease, congestive heart failure, or previous amputation), and laboratory and clinical measures (ie, A1C level, serum creatinine, and urine albumin-to-creatinine ratio). 

Findings:  

  Standard (n=5,051) Intensive (n=5,045)
≥1 hypoglycemic event requiring medical assistance n=215 n=565
Mean number of hypoglycemic episodes in 7 days preceding 4-month visit (SMBG <70 mg/dL) 0.29 (SD, 0.49) 1.06 (SD, 0.98)
Hypoglycemia unawareness reported 2.6% of visits 5.8% of visits
Hazard ratio for models including mortality and frequency of hypoglycemic episodes 0.98 (95% CI,
0.91–1.06); P=0.615
0.93 (95% CI,
0.90–0.97); P<0.001

 

 

 CI=confidence interval; SD=standard deviation; SMBG=self-monitoring of blood glucose.  

 

  • The hazard ratios for models including mortality and hypoglycemic episodes are shown above. There were no statistically significant findings in either the standard or intensive groups when hazard ratios for mortality were examined in other models (including unrecognized hypoglycemia)  
  • An inverse relationship was observed between mortality and both (1) the updated annualized number of hypoglycemic events and (2) the updated annualized number of 4-month intervals where subjects were unaware of hypoglycemia when the results were controlled for the baseline covariates that predicted mortality in the ACCORD cohort  

 

Don’t miss these slides on ACCORD in our Slide Library: 

ACCORD Substudy: Impact of Frequent and Unrecognized Hypoglycemia on Mortality—Design   
ACCORD Substudy: Impact of Frequent and Unrecognized Hypoglycemia on Mortality—Results
   

 

 This overview was created under the auspices of KnowledgePoint360 Group, LLC, and was not associated with funding via an educational grant or a promotional interest. 

 

Last Modified: 11/18/2013