Table 1

Risk of glycemic failure with thiazolidinediones and sulfonylureas in ADOPT and RECORD by sex- and obesity-defined subgroup

Patients (n)Events (n)Absolute 5-year risk (%)Hazard ratio (95% CI)P value
TZDSUTZDSUTZDSU(TZD vs. SU)
ADOPT monotherapy failure
 Nonobese males373395476321.721.90.78 (0.54–1.14)0.21
 Obese males4023874410815.043.80.32 (0.23–0.46)<0.001
 Nonobese females208174163410.931.50.34 (0.19–0.62)<0.001
 Obese females407379319311.642.20.23 (0.16–0.35)<0.001
RECORD dual-therapy failure
 Nonobese males240228667033.634.01.00 (0.72–1.40)0.94
 Obese males3613569213230.741.40.72 (0.55–0.94)0.02
 Nonobese females137127264520.738.80.52 (0.32–0.84)0.01
 Obese females3793947214222.740.50.52 (0.38–0.68)<0.001
  • Failure was defined according to original trial protocol: ADOPT trial (monotherapy), defined as fasting plasma glucose ≥180 mg/dL; RECORD (dual therapy with metformin), defined as HbA1c ≥8.5%. SU, sulfonylureas; TZD, thiazolidinediones.