Is Insulin Therapy Safe?

Background: After 98 years of insulin therapy, issues of safety remain of concern.
Areas of Uncertainty: Uncertainty has been expressed variously in regard of arterial cell wall proliferation, promotion of proliferative retinopathy, promotion of tumor growth, and for pregnancy. Immunological issues have been little studied since the advent of highly purified insulins in the 1970s. A specific topic is whether hypoglycemia, severe or otherwise, might promote cardiac thrombotic or dysrhythmic events.
Data Sources: A literature review in these areas is difficult because nearly all clinical trials with insulin refer to adverse events. However, the specific topics aforementioned allow for some informed literature searching supplemented by finger-searching of published articles, notably in connection with the insulin analogues.
Therapeutic Understandings: Safety data for pregnancy are weak because of power problems, but there are no signals for added maternal or fetal risk. Clinical-outcome trials that assess insulin against other glucose-lowering therapies or with significantly different insulin preparations in different arms are few and are sometimes conducted at modest dosage but fail to suggest promotion of arterial disease. Concern over growth-promoting activity of insulin glargine turned out to be ill-founded when the circulating moiety after injection was noted to have a lower IGF-1:insulin activity than human insulin, and a direct study of retinopathy progression or meta-analysis of malignancy incidence failed to show signals of concern. It does seem that severe hypoglycemia can cause death in some people with type 1 diabetes, although the tissue mechanism is unknown, but reducing severe hypoglycemia in type 2 diabetes does not protect against arterial events. Both symptomatic and severe hypoglycemia can however be reduced by use of more recently marketed insulin analogues, and this improves tolerability if not safety.
Conclusions: In conclusion, although insulin therapy clearly gives health benefits, the evidence for long-term harm is absent or weak.