Efficacy and safety of biphasic insulin aspart 70/30 in type 2 diabetes patients of different race or ethnicity (INITIATEplus trial)

Objective: To determine if self-titration using biphasic insulin aspart 70/30 (BIAsp 30) had a different impact on efficacy and safety across different racial/ethnic subgroups. Research design/methods: This was an exploratory, post hoc analysis by race (White vs. Black/African-American) and ethnicity (Hispanic/Latino vs. non-Hispanic/Latino) of data from the INITIATEplus trial. Participants were treated twice-daily with BIAsp 30 over 24 weeks. Trial registration: NCT00101751. Main outcome measures: Efficacy endpoints included reductions in mean glycated hemoglobin (A1C) and fasting plasma glucose (FPG). Safety endpoints included hypoglycemia rates (events/patient-year) and adverse events. Body weight changes were also measured. Results: Glycemic control improved by a similar extent for all demographic groups. Observed mean decreases in A1C ranged from 2.4% to 2.6% after 24 weeks' treatment. Baseline-adjusted mean A1C decreases for White vs. Black/African-American subjects were 2.56% and 2.13% (p<0.0001), and for Hispanic/Latino vs. non-Hispanic/Latino subjects were 2.45% and 2.42% (p=0.677), respectively. Final FPG values were similar among all groups (141-146 mg/dL [7.83-8.10mmol/L]), and baseline-adjusted FPG decreases were not significantly different (p>0.025) between groups. Hypoglycemia was low for White, Black/African-American, Hispanic/Latino, and non-Hispanic/Latino subjects (0.08, 0.04, 0.03, and 0.07 major events/patient-year, with 0.60, 0.30, 0.37, and 0.52 minor events/patient-year, respectively). Body weight increases were 3.17 and 3.06 kg (White vs. African-American) and 2.69 and 3.19 kg (Hispanic/Latino vs. non-Hispanic/Latino). Final weight-adjusted total daily insulin doses were 0.60 U/kg for Black/African-American subjects vs. 0.78 U/kg for White subjects (p<0.0001), and 0.71 U/kg for Hispanic/Latino subjects vs. 0.74 U/kg for non-Hispanic/Latino subjects (p=0.42). Limitations: The trial was not designed or powered for comparisons across racial or ethnic groups, subjects were not stratified for pre-baseline medication regimens between each race and ethnic group, and unequal subject numbers and baseline A1C disparities existed between the pairs of groups being compared. Conclusions: Diabetes self-management with BIAsp 30 using an easily followed self-titration algorithm produced low hypoglycemia rates. All subgroups achieved A1C reductions > 2.1% and FPG declines > 82 mg/dL that were similar across groups, demonstrating that self-titration of BIAsp 30 can successfully be pursued in a primary care setting by patients who had previously failed to meet ADA A1C targets under oral antidiabetes therapy, with race or ethnicity not an obstacle to achieving better glycemic control.