Insulin in Type 2 Diabetes


Synopsis: Insulin therapy in Type 2 diabetics is associated with lower HbA1C levels and higher resource use, but few patients achieve tight control.

Source: Hayward RA, et al. JAMA 1997;278:1663-1669.

Late adult onset, or type 2, diabetes is one of the most common conditions seen by all primary care physicians. The risk association between Type 2 diabetes and macrovascular disease (coronary and cerebral vascular disease) and microvascular disease (blindness and end stage renal disease) is well known. An active debate is ongoing regarding whether the benefits of tight glucose control in preventing these complications observed in clinical trials of young, Type 1 diabetics can be applied to the community practice of the typical older, often frail, Type 2 diabetics.

Hayward and associates, as part of a Patient Outcomes Research Team Study, have assessed the longitudinal care of a cohort of Type 2 diabetics who were members of a single closed panel HMO. This assessment provides insights into the level of effectiveness, using hemoglobin A1C as an intermediate marker of diabetic control, and resources; using hospitalization, office visits, and laboratory testing; and in the typical practice of medicine. A prominent theme is the differences between patients receiving sulfonylureas compared to insulin. Selected findings from the plethora of results are listed in the table.

Insulin therapy was particularly effective when started for those with poor control (HbA1C > 10%), but few patients with any form of therapy achieved tight control (HbA1C < 8%).


Assessment of care of Type 2 diabetics in a closed-panel HMO

Measure Findings

Type of therapy 17% diet only; 49% sulfonylureas; 34% insulin

Insulin schedule 66% twice daily; average dose, and dose 55 U/day

Patients with Sulfonylureas, 25%; hemoglobin A1C > 9% insulin, 35%

12 mo. Pre vs. 12 mo. Post Outpatient visits 9.7 vs. 12.6 Starting InsuliLab test RVUs 86 vs. 107 Hemoglobin A1C 9.3 vs. 8.4


I interpret these results differently from the newspaper summaries that condemn insulin. Insulin therapy is most effective in those with the poorest glucose control. Still, even in this predominantly white, married, insured cohort, few patients achieved tight control. Translating these findings to my practice is limited by the lack of any information about patient body weight. I suspect that the failure to lose weight (or longer term limit future weight gain) is more important than the type of hypoglycemic therapy used in the failure to achieve tight control in most patients.