Do insulin analogues differ in effectiveness and safety for the treatment of diabetes?

This study assessed the effectiveness and safety of different long-acting and short-acting insulins for the treatment of patients with diabetes. The data supported the use of long-acting insulins instead of neutral protamine Hagedorn (NPH) insulin (Humulin N; Novolin N), as basal insulin for patients with type 2 diabetes (T2D).

Insulin may be required by patients with T2D if average blood glucose levels over 2 to 3 months (HbA1c) are above 9%, with symptoms. However, the risk of hypoglycemia (very low blood glucose levels) with insulin therapy, is often a concern.

Evidence suggests that there is a lower hypoglycemic risk with the use of long-acting insulin analogues (laboratory-made forms of insulin) like glargine (Lantus) U-100 and U-300, detemir (Levemir), and degludec (Tresiba), compared to NPH insulin. Additionally, short-acting insulin analogues are useful in providing better control for a high rise in blood glucose levels after meals (postprandial hyperglycemia) over normal human insulin.

Limited studies directly compare two or more types of long-acting or short-acting insulin. There is a need to compare the effectiveness and safety of long-acting and short-acting insulin analogues for treating T2D.

This analysis included 50 trials that compared insulin analogues such as lispro (Humalog), apart (NovoLog), glulisine (Apidra), detemir, glargine, and degludec) with glargine U-100, or human (NPH) insulin. A total of 25,554 patients with T2D were included. HbA1c levels were assessed at 24 weeks, 52 weeks and 103 weeks.

At 52 weeks, determir was less effective than glargine U-100. Compared to NPH insulin, basal (long-acting) analogues had significantly reduced HbA1c levels at 24 weeks (by 0.1%), along with the risk of total (by 30%) and night (by 43%) hypoglycemia. Glargine U-300 and degludec were associated with a significant reduction in nocturnal hypoglycemia risk.

Overall, prandial (short-acting) analogues did not have significant differences in blood glucose control, hypoglycemia, or weight gain.

The study supports the use of long-acting analogues as basal insulin for T2D over NPH insulin. The were no significant differences among short-acting analogues.

This study included a limited number of small trials and in some cases, indirect data comparisons were made. Further randomized studies are needed.