Progression to insulin for patients with diabetes mellitus on dual oral antidiabetic therapy using the US Department of Defense Database


Correspondence to: Karen Rascati, PhD, Health Outcomes and 2409 University Ave STOP A1900, Austin, TX 78712-1113, USA.




To compare ‘progression to insulin’ for three cohorts on oral antidiabetic medication combinations: metformin/sulphonylurea (Met/SU), metformin/thiazolidinedione (Met/TZD) and sulphonylurea/thiazolidinedione (SU/TZD).


Retrospective cohort analysis design was used. The subjects were US nationwide members of military and their families. A total of 5608 patients who were on antidiabetic monotherapy for at least 1 year before adding a second agent to their medication regimen between October 2001 and September 2008 participated in this study. Mean age ranged from 64 to 71 years among the cohorts. Cox regression compared the progression to insulin, adjusting for demographics, months of follow-up and co-morbidities [measured with Chronic Disease Score (CDS)].


By the end of the 2- to 6-year follow-up period, 14.3% of the Met/TZD cohort, 23.6% of the Met/SU cohort and 28.2% of the SU/TZD cohort had insulin added to their regimen. Those in the Met/SU cohort had a 1.8 times higher probability of progression to insulin than those in the Met/TZD cohort [odds ratio (OR) = 1.80, 95% confidence interval (CI) = 1.51–2.14), and those in the SU/TZD cohort had a 2.5 times higher probability of progression to insulin than those in the Met/TZD cohort (OR = 2.51, 95% CI = 2.04–3.08).


When sensitizers were paired (Met/TZD), a lower percentage of patients progressed to insulin during the study period, as opposed to patients who used a combination of a secretagogue with a sensitizer (SU/TZD or Met/SU).