3064 — Differential Impact of Poorly Controlled Diabetes on Mortality by Race/Ethnicity and Medication Status in a National Cohort of Veterans with Diabetes
Hunt KJ, Gebregziabher M, Lynch CP, Echols C, and Egede LE, Center for Disease Prevention and Health Interventions for Diverse Populations, Ralph H. Johnson Veterans Affairs Medical Center, Charleston, SC;
We examined the association between HbA1c and mortality stratified by race/ethnicity in a national cohort of Veterans with type 2 diabetes. Moreover, we considered medication use and adherence as potential modifiers of the association between HbA1c and mortality.
892,223 Veterans with diabetes in 2002 were followed through December 2006. HbA1c category was the main exposure [i.e., HbA1c less than 7%, HbA1c 7 to 8%(reference), HbA1c 8 to 9% and HbA1c greater than 9%]. Covariates included age, sex, marital status, rural/urban residence, geographic region, number of comorbidities, and diabetes medication status [i.e., defined by medication use and adherence (i.e., medication possession ratio of at least 80%)]. Both HbA1c and medication status varied over time and Cox-regression that accounts for time varying variables was used.
The study population was 61.51% non-Hispanic white (NHW), 12.14% non-Hispanic black (NHB), 13.86% Hispanic, and 12.48% other (OTH), with a mean age 66.2 ± 11.2 years in 2002. During follow-up, 20.83% died. In diabetes medication non-users there was a graded association between increasing HbA1c category and the risk of mortality within each race/ethnicity. Mortality hazard ratios (HR) for HbA1c greater than 9% were 1.55 (95% confidence interval: 1.43, 1.69) in NHW, 1.58 (1.34, 1.87) in NHB, 2.22 (1.75, 2.82) in Hispanics and 2.70 (2.35, 3.10) in OTH. In contrast, in non-adherent medication users, HbA1c levels less than 7% predicted higher mortality risk in NHB [HR = 1.12 (1.05, 1.20)], but lower mortality risk in OTH [HR = 0.89 (0.81, 0.98)]. Moreover, HbA1c levels greater than 9% were only predictive of mortality in NHW [HR = 1.11 (1.06, 1.16)] and OTH [HR = 1.17 (1.02, 1.35)]. Finally, in adherent medication users, HbA1c levels less than 7% predicted higher mortality risk in NHB [HR = 1.18 (1.07, 1.31)] and OTH [HR = 1.31 (1.16, 1.48)], while HbA1c levels greater than 9.0% predictive higher mortality risk across all race/ethnic groups.
We found evidence for racial/ethnic differences in the association between glycemic control (i.e., tight and poor) and morality which varied by medication use/adherence.
The differential impact of poor and tight glycemic control on mortality by race/ethnicity, as well as results from recent clinical trials should be considered when setting target HbA1c levels.