Current issues of ACP Journal Club are published in Annals of Internal Medicine


Awareness of hypoglycemia did not differ between human and porcine insulins

ACP J Club. 1992 Nov-Dec;117:83. doi:10.7326/ACPJC-1992-117-3-083

Source Citation

Colagiuri S, Miller JJ, Petocz P. Double-blind crossover comparison of human and porcine insulins in patients reporting lack of hypoglycaemia awareness. Lancet. 1992 Jun 13;339:1432-5.



To investigate whether the frequency and characteristics of hypoglycemic episodes differ in response to human and porcine insulins among patients who have reported reduced hypoglycemic awareness when using human insulin.


Double-blind, cross-over trial with patients randomly assigned to an order of four 1-month treatment periods.


University endocrinology outpatient department in Australia.


67 patients, who reported hypoglycemic unawareness after changing from porcine to human insulin, were referred by physicians across Australia and entered the study. 50 patients (30 men, mean age 47 y, duration of diabetes 20 y, median duration of insulin treatment 13.5 y) completed the study.


Patients were assigned to 2 periods of human insulin and 2 periods of porcine insulin in random order, in the manner in which they usually took it (isophane insulin or insulin zinc suspension with or without soluble insulin, or biphasic suspension). All insulins were monocomponent, 100 U/mL (Novo Nordisk A/S, Denmark). Patients continued their usual daily activities and diabetes management regimen.

Main outcome measures

Frequency, timing, and characteristics of hypoglycemic events, reported by patients in a record that also included routine glucose self-monitoring reports. Hypoglycemia was defined as a symptomatic episode or altered consciousness that was corrected by the administration of glucose or glucagon, or a blood glucose measurement < 3.5 mmol/L, or both.

Main results

Mean frequency of hypoglycemia was 2.6 episodes per week for both human and porcine insulin treatment periods (difference 0.004 [SD 0.134] episodes/d). The order of the 4 treatment periods had no effect on the frequency of hypoglycemic events. The time of day when hypoglycemia occurred did not differ for the insulin types (45% and 50% of hypoglycemic events occurred from before breakfast to before lunch for treatment with human and porcine insulin, respectively). The frequency of reduced or absent hypoglycemia awareness did not differ between the 2 insulin types (difference human-porcine = -0.015 [SD 0.131] episodes/d; P > 0.2). Glycemic control was at least acceptable for 70% of patients on either treatment, and blood glucose concentrations did not differ between groups.


Type of insulin (human or porcine) did not affect frequency, timing, severity, or awareness of hypoglycemia.

Source of funding: Novo Nordisk A/S (insulin preparations).

For article reprint: Dr. S. Colagiuri, Department of Endocrinology, Diabetes, and Metabolism, Prince of Wales Hospital, High Street, Randwick, New South Wales 2031, Australia.


The debate on human insulin safety has been intense. The key question is, "Does human insulin cause more hypoglycemia unawareness than nonhuman insulin?" Although this question can be studied in several ways, the best evidence comes from studies such as that of Colagiuri and colleagues (double-blind randomized trials of diabetic patients, lasting several months or longer, with hypoglycemia as a predetermined study outcome). By studying only high-risk patients (with previous hypoglycemic unawareness on human insulin), the investigators maximized their ability to detect any real difference in hypoglycemia unawareness between the insulin types. As opposed to 2 other small clinical trials (1), hypoglycemic awareness did not differ by insulin species, and, in common with 6 other trials, hypoglycemia did not occur more often during human insulin therapy.

Unfortunately, the controversy may not be over. This trial and others are not large enough to clearly exclude increased risk for hypoglycemia with human insulin at levels (e.g., a doubling in risk) that would yield a large burden. Such a burden would be needless because human insulin offers no advantage over nonhuman insulin for most patients. Finally, the social and legal problems arising from the controversy will continue without further evidence from clinical trials that human insulin is safe. Thus, a large clinical trial comparing the incidence of severe hypoglycemia with human and nonhuman insulin may be inevitable.

In the meantime, what should physicians prescribe? I think it is reasonable to leave patients doing well on nonhuman or human insulin alone; to start all new patients on human insulin; to comply with patients' requests to switch to human insulin, keeping in mind that the nature of their hypoglycemia may change; and to switch patients on human insulin to nonhuman insulin if they are having problematic hypoglycemia not explained by another mechanism.

Jeffrey Mahon, MD
University of Western OntarioLondon, Ontario, Canada

Jeffrey Mahon, MD
University of Western Ontario
London, Ontario, Canada


1. Egger M, Smith GD, Teuscher AU, Teuscher A. BMJ. 1992;303:622-6.