Low Glycemic Level Linked to Higher Mortality in Elderly

Lead author Elbert S. Huang, MD, MPH, associate professor of medicine at the University of Chicago in Illinois, stated that, "We saw increased mortality and complications, as anticipated, among those with very high blood sugars, but we also saw a modestly increased risk of death among those with very low levels of blood sugar." "In our study, we found the best overall outcomes among those with the intermediate levels of control, those with A1cs below 8 percent but above 6 percent. We observed similar patterns for those in their 60s, 70s, and over 80."

The goal of the study was to determine the range of glycemic levels in older diabetic patients associated with the lowest complications and mortality rates.

The study sample consisted of 71,092 patients with Type 2 diabetes, at least 60 years old, who were enrolled in Kaiser Permanente Northern California from 2004 to 2008. Associations between baseline A1c level and subsequent outcomes, including nonfatal acute metabolic, microvascular, and cardiovascular events, and mortality, were evaluated with Cox proportional hazards models.

At baseline, mean age was 71.0 ± 7.4 years, and mean A1c level was 7.06 ± 1.2%. For levels of A1c of more than 6.0%, the risk for any nonfatal complication rose monotonically, with adjusted hazard ratios (HRs) of 1.09 (95% confidence interval [CI], 1.02 - 1.16) for A1c levels of 6.0% to 6.9% and 1.86 (95% CI, 1.63 - 2.13) for A1c levels of more than 11.0%.

There was a U-shaped association of mortality with A1c level. Compared with the mortality risk with A1c level of less than 6.0%, the risk was lower for A1c levels between 6.0% and 9.0% (e.g., HR, 0.83 [95% CI, 0.76 - 0.90] for A1c levels 7.0% - 7.9%) and higher at A1c levels of 11.0% or more (HR, 1.31; 95% CI, 1.09 - 1.57). At A1c levels of 8.0% or higher, the risk was significantly higher for any endpoint (complication or death).

"We cannot say whether this unexpected finding is due to the very low blood sugar itself, the treatments used to control blood sugars, or to some other factors not directly related to the care of diabetes," said senior study author Andrew J. Karter, PhD, from the Kaiser Permanente Division of Research in Oakland, California. "It may be that the sickest patients at high risk of dying simply had low blood sugars to start with, rather than anything directly associated with the care of diabetes increasing the risk of death. Further research will be focused on identifying the mechanisms that underlie the somewhat increased mortality among those with very low A1c."

Across age groups (60 - 69 years, 70 - 79 years, and ≥ 80 years), the patterns of findings were generally consistent. The investigators therefore recommended using caution when setting a target A1c level of less than 8.0% for older patients, because A1c levels of less than 6.0% were associated with increased mortality risk.

Diabetes Care. Published online April 19, 2011