Human Insulin as an Antidote to the High Cost of Insulin: Clinical Insignificance of Pharmacokinetic/Pharmacodynamic Differences

Now that Congress has been unable to pass legislation capping the monthly cost of insulin for non-Medicare patients at $35, alternative solutions need to be found for this serious, increasing problem. A survey of 199 insulin-requiring people with diabetes published in 2019 revealed that one-fourth of respondents were underusing insulin because of costs, and these individuals were three times more likely to have an A1C level >9.0% than those who were able to take their prescribed insulin doses (1). The 2021 National Health Interview carried out by the Centers for Disease Control and Prevention revealed that >1 million Americans ration insulin (2). A survey of 2,000 individuals using insulin published in 2022 by Charity Rx revealed that 62% underused insulin, 80% took on credit card debt averaging $9,000 to afford their average $400 monthly cost of insulin, and 38% had a hospital admission for ≥1 day because of insulin rationing (3), the latter suggesting that most of these individuals probably had type 1 diabetes.

The two largest clinical challenges to using insulin are inconsistent eating patterns of insulin users and a 20–30% day-to-day intraindividual variability in the response to insulin (4,5). The combination of these two factors supersedes the small pharmacokinetic (PK)/pharmacodynamic (PD) differences between human and analog insulins. Human insulins (NPH and regular) are just as effective as analog insulins (6). In head-to-head comparisons between rapid-acting insulin analogs and regular insulin in 26 studies involving 9,509 participants, the average difference in A1C was 0.05%. In head-to-head comparisons between basal insulin analogs and bedtime NPH insulin involving 10,512 participants, also in 26 studies, the average A1C difference was 0.04%. There was a small but statistically significant increase in overnight hypoglycemia with NPH insulin. However, small routine bedtime snacks, which mitigate this risk, were not recommended in any of the studies. There were no differences in daytime or severe hypoglycemia in either of these head-to-head comparisons. A more recent Cochrane Database systemic review and meta-analysis comparing randomized controlled trials (RCTs) of rapid-acting analogs with regular insulin in type 2 diabetes also confirmed no difference in efficacy or hypoglycemia rates (7). Finally, in three studies, there were no differences in change in A1C between individuals taking premixed insulin analogs and those taking premixed human insulin (8).

In addition to the head-to-head comparisons between human and analog insulins showing that small differences in PK/PD characteristics do not affect clinical effectiveness, other RCTs comparing only human insulins and only analog insulins also confirm that these small differences do not affect clinical effectiveness. Regarding human insulin, because it takes 30 minutes before blood insulin concentrations start to increase after injecting regular insulin, classical teaching for many decades has been to administer it 20–30 minutes before a meal. Yet, when injected just before a meal or 20 minutes before a meal over 6 weeks, mean pre- and postprandial glucose checks seven times per day yielded virtually identical results (9). Regarding rapid- and ultra-rapid-acting analog insulins, there were no differences in change in A1C between insulin aspart and faster-acting insulin aspart (10–15), insulin lispro and ultra-rapid-acting insulin lispro (16), or insulin aspart and inhaled technosphere insulin (17).

There are also relatively small differences in PK/PD characteristics among the long-acting basal insulins, which have their major activity during the overnight period. As with the short-, rapid-, and ultra-rapid-acting insulins, these PK/PD differences have no effect on changes in A1C when comparing insulin detemir with insulin glargine (18–20), insulin lispro protamine suspension (21) with insulin degludec (22), insulin glargine with insulin lispro protamine suspension (23), and insulin glargine with insulin degludec (24–36). There were some significant differences in overnight hypoglycemia in some of the studies (27,28,32–34), but as in studies comparing bedtime NPH and basal insulins, small bedtime snacks were not routinely recommended.

What about insulin regimens composed entirely of combinations of analog insulins versus combinations of human insulins? Comparing basal/bolus regimens composed of basal and rapid-acting analog preprandial insulins with bedtime human NPH and preprandial regular insulin, three studies showed no difference (37–39), and one favored the analogs (40). Even when the intensive insulin regimens were different, there was no difference between human and analog insulins. A self-mixed/split regimen of NPH and regular insulin was just as effective as a basal-bolus regimen of analog insulins (41).

Differences in PK/PD characteristics in insulins of the same class do not translate into differences in clinical effectiveness, as shown by this very large number of studies. Real-world experience corroborates this conclusion. There was no change in A1C levels in insulin-requiring people over 30 years between 1988 (when only human insulins were available) and 2020 (after analogs started to become available in the late 1990s), with ∼28% achieving the American Diabetes Association’s A1C target of <7.0% and ∼15% having an A1C >10% (42).

This means that all insulin regimens (i.e., basal alone, basal-bolus, self-mixed/split, and premixed regimens) can use either human or analog insulin formulations without compromising effectiveness. This is extremely good news for individuals for whom the cost of insulin presents a problem because human insulins are available for $25–30 per vial at Walmart pharmacies. Few patients or clinicians seem to be aware of this. It behooves patients with financial constraints to bring this fact to the attention of their clinicians and for the clinicians to act on it. There is no need for the cost of insulin to be an impediment to patients taking their prescribed doses of insulin.