Thomas K. Graves, DVM, PhD, DACVIM
Department of Veterinary Clinical Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, USA
With constantly changing information and availability, it can be difficult to make the best choices for insulin in veterinary patients. The characteristics of available insulin products, including NPH, lente, PZI, glargine, and detemir, as well as the rapid-acting insulins lispro, aspart, and glulisine, are reviewed in this article. The new human insulin analog, degludec, will be introduced briefly. Strategies for monitoring insulin therapy will also be reviewed. Oral medications used for treatment of type 2 diabetes will also be reviewed.
Types of Insulin
There are many different types of insulin that vary with species of origin and with chemical modifications and formulations that affect onset and duration of action (Table 1). Porcine insulin, which is identical to canine insulin in its amino acid structure, is available for use in dogs in some countries, but, unfortunately, no specific feline insulin formulation is currently available. Human, bovine, or porcine insulins are used in treating both diabetic cats and dogs. Data concerning the pharmacokinetics and pharmacodynamics of insulin in dogs and cats are difficult to interpret. Most published studies have been conducted in normal animals, and some have been done in animals with diabetes. In either case, it is difficult to determine the effects of endogenous versus exogenous insulin. Determinations of potency, time to peak activity, and duration of activity—factors that influence choice of doses and dosing intervals—vary widely from animal to animal. In fact, there is no reasonable way to predict the kinetics of a given insulin preparation in any given patient.
Table 1. Insulin products
Humulin R (Lilly)
Novalin R (Novo Nordisk)
Novalog (Novo Nordisk)
Protamine zinc insulina
Porcine insulin zincb
Humulin N (Lilly)
Novalin N (Novo Nordisk)
ProZinc (Boehringer Ingelheim)
Levemir (Novo Nordisk)
In cats, neutral protamine Hagedorn (NPH) insulin (Humulin-N, Eli Lilly; Novalin N (Novo Nordisk), porcine lente insulin (Vetsulin), and protamine zinc insulin (ProZinc, PZI) have been used commonly to treat diabetes mellitus. In recent years, insulin glargine (Lantus) has probably become the most commonly used insulin preparation in cats, despite relatively little published evidence supporting its use. Recently, another insulin analog, insulin detemir (Levemir), has received some attention among veterinary researchers and feline practitioners. In dogs, porcine lente insulin and NPH are used most commonly.
NPH and Lente
NPH is considered an intermediate-acting insulin and is available as a human recombinant product. NPH is used commonly in animals with diabetes, and is typically given subcutaneously twice daily. Lente insulin uses zinc as a positively charged ion on which to base insulin polymerization. Polymers are absorbed and metabolized slowly so that the onset and duration of lente insulin are extended beyond those of regular insulin. Human recombinant lente insulin has been removed from the United States market and is no longer available for use. Porcine lente insulin, however, has gained in popularity in some countries and is currently marketed and labeled for in cats. The only currently available veterinary product is Vetsulin® (the tradename is Caninsulin in other countries). This insulin product was withdrawn from the U.S. Market, but has recently become available again. While identical to canine insulin, porcine insulin is dissimilar in amino acid sequence when compared with feline insulin, but it is no more divergent (by three amino acids) than is human insulin. Lente is typically given twice daily by subcutaneous injection, and studies show it is a reasonable choice for treating diabetic cats.15 A recent study suggested the duration of porcine lente is shorter than either PZI or glargine in cats.8
Protamine Zinc Insulin
PZI has been used extensively in feline diabetes. It is typically given subcutaneously twice daily, with a starting dose of 1–3 U/cat. This insulin preparation was widely available, but was largely removed from the human market in the 1990s. Recently, PZI preparations marketed for use in cats have once again become available. A human recombinant protamine zinc insulin product approved for feline diabetics was introduced in 2009. This insulin was the subject of a multi-center clinical trial reported by Nelson et al.11 In that study of 133 diabetic cats, “good” glycemic control was defined as an average blood glucose concentration below 200 mg/dL during a 9-h blood glucose curve. A glucose nadir of less than 150 mg/dL was also considered good, as was a serum fructosamine concentration lower than 450 µmol/L. In that study, after 45 days of twice-daily treatment with PZI, 60% of cats exhibited good glycemic control based on the glucose nadir. Seventy-five percent of cat owners reported improved polydipsia and 79% reported improved polyuria.
Glargine and Detemir
Insulin glargine is a genetically engineered insulin analog that has hormonal action identical to native insulin, has no known immunogenicity, and achieves long-lasting glycemic control while minimizing fluctuations in blood glucose concentration in many human diabetics. Glargine is based on human recombinant insulin with a few amino acid substitutions. Glycine is substituted for an asparagine residue at the amino terminal of the A chain, and two arginine residues are added to the end of the B chain. The result is a shift in the isoelectric point of the insulin molecule so that it is completely soluble at a low pH (around pH 4). The pH of interstitial fluid is approximately 7.4, and when glargine is injected into a patient, the insulin precipitates into hexamers that are inactive. These insulin hexamers are slowly broken down in the body to form active insulin monomers. The result is that the onset is gentle and the duration is long-lived. Because of the difference in pH, glargine cannot be mixed with other insulin formulations. Experience with using glargine in cats is growing,15,21 and many clinicians have had good success with its use. Glargine is best used twice daily subcutaneously. A study of glargine use in dogs was reported recently.4 In that study, 12 client-owned diabetic dogs were treated with insulin glargine BID for 24 weeks, and treatment resulted in significantly lower mean blood glucose concentrations and improvement in polyuria/polydipsia in 91 percent of dogs.
Insulin detemir has been used in Europe for several years and was just recently approved for use in the US. Rather than having amino acid substitutions (like insulin glargine), insulin detemir is acylated with myristic acid, which still allows some hexamers to form at neutral pH, but, more importantly, allows the insulin to bind to albumin. This results in a very slow, smooth delivery of insulin such that once-daily delivery is all that is needed in many human patients.
We have done some work comparing the pharmacodynamics of glargine, detemir, and regular insulin in cats.5 In the cats we have studied we have found detemir to be relatively “peakless” and its duration of action is similar to that of glargine. The effect of detemir seems to be more predictable and consistent between cats, based on our studies, but this bears further investigation. The mean duration of action of insulin detemir was 13.5 h, while glargine’s mean duration of action was 11.3 h. The difference between the two was not significant, and it was concluded that either analog might be useful as a once-a-day insulin treatment in a given cat, while others would need more frequent dosing.
Another recent study has demonstrated similar effects of glargine and detemir in cats, although the dose of detemir needed to achieve the same clinical effect was less than with glargine.17 One advantage of detemir might be that, because it is bound to albumin, it reaches higher concentrations in organs with fenestrated capillaries, especially the liver. This more closely mimics circulation of insulin in normal physiology. We have observed no toxic effects of detemir in cats we have studied, and it seems reasonable that detemir could be tried if adequate glycemic control is not achieved with other types of insulin.
There are very few clinical data available to guide the use of glargine and/or detemir in dogs. We have used both insulin preparations in dogs, however, especially in situations in which we suspect short duration of action of NPH as an underlying reason for poor glycemic control. Glargine can be dosed similarly to other insulin types, but not detemir. Detemir contains four times more insulin per unit than other preparations because of its lower affinity for the human insulin receptor, an affinity that may be greater in dogs, and that would increase the risk of hypoglycemia. In fact, one study showed a very high potency of detemir in dogs compared with other insulins.18 Detemir should be used when the patient is large enough that dosing at 0.1 to 0.2 U/kg is possible using U-100 insulin, when there is evidence that NPH is very short-acting, if other insulin preparations have been tried without success, or any combination of these factors. Detemir can be used once a day in some dogs and may be needed twice a day at starting doses of 0.1 to 0.2 U/kg/dose.
Lispro, Aspart, and Glulisine
Historically, a combination of regular insulin and an intermediate-acting insulin was used to replace postprandial insulin in human diabetic patients. The action profile of regular insulin after subcutaneous (SC) injection may be inadequate, however, for the treatment of diabetes because its absorption is relatively slow and the duration of action is too long (about 5–8 h in people, about 5 h in cats and dogs).14,16 Insulin lispro was the first rapid-acting analog to be approved for use in people.20 The amino-acid sequence of insulin lispro consists of a reversal of proline at the B28 position and lysine at the B29 position. This small change greatly decreases the tendency for association and enhances the rate of absorption. In insulin aspart the proline residue at B28 is replaced with an aspartic acid residue. In insulin glulisine, lysine at B29 is replaced by glutamic acid and on position B3 asparagine is replaced by lysine. Insulin aspart and insulin glulisine have pharmacokinetic and pharmacodynamic profiles similar to insulin lispro.14
There are no reports on the use of rapid-acting analogs in the chronic treatment of diabetes in cats. Insulin lispro has been successfully used in dogs to treat diabetic ketoacidosis (DKA).19 In that study insulin lispro was administered intravenously and had similar efficacy as the traditionally used regular insulin. No adverse reactions were seen. There is no clear rationale for preferring insulin lispro over regular insulin for use in constant-rate intravenous insulin infusions. The biochemical alteration in insulin lispro confers greater dissociation and faster absorption of insulin injected subcutaneously, but both insulin lispro and regular insulin should dissociate immediately when delivered intravenously.
Insulin degludec is the newest insulin analog to be garnering attention in the clinical diabetes community. Degludec is an ultra-long-acting human insulin analog in which the insulin molecule is conjugated to a hexadecanedioic acid residue linked to a gamma-L-glutamyl spacer at the end of the B-chain. This product is currently not available commercially in the United States, but its manufacturer reportedly plans to begin marketing the product within the next 2 years. In people, this insulin analog has an extremely long half-life, and is being evaluated for a three-times-weekly dosing scheme.7 Should this treatment regimen be successful in other species, it could potentially revolutionize insulin therapy for veterinary patients.
Monitoring Insulin Therapy
Blood glucose curves have been used extensively to monitor insulin therapy in dogs and cats. Conventional wisdom has dictated that adjustments in insulin therapy, whether changes in dose, frequency, or type of insulin, only be made based on serial blood glucose measurements. The validity of blood glucose curves has been called into question by several recent studies. A European group of investigators studied the differences in blood glucose curves performed in-hospital vs. those performed by owners at home.1 Owners were instructed in blood sample collection using a simple micro-lance technique, and blood glucose concentrations were determined using portable meters. When interpreting blood glucose curves in the hospital, veterinarians commonly consider that there may be a falsely elevated degree of hyperglycemia because of the stress of hospitalization. Surprisingly, Casella and colleagues found the opposite to be true. Following repeated curves in the same dogs multiple times at home and in the hospital, it was found that mean blood glucose concentrations were significantly higher when dogs were tested at home. Recommendations for changes in insulin therapy were different 42% of the time based on the at-home vs. in-hospital curves.
Another study of blood glucose curves found more startling results. Investigators in Australia performed repeated blood glucose curves in dogs two weeks apart with no change in insulin therapy protocols, and compared differences between the curves.3 Minimum, mean, and maximum blood glucose concentrations, as well as morning and evening pre-insulin glucose concentrations, and the time to reach the nadir in blood glucose were compared. Variations in these measurements within the same dog were striking. For example, the coefficient of variation for the maximum glucose concentration was 100%, and for the time-to-nadir the CV was 98%. These results tell us that blood glucose curves may not be as helpful as we once thought. Certainly, changes in insulin regimens should be made conservatively when based on a glucose curve. Also, other methods of assessing glycemic control might be more important. Clinical signs of hyperglycemia (polyuria, polydipsia, appetite changes, weight loss, cataracts, etc.) and serum concentrations of fructosamine may be more useful indicators of glycemic control than blood glucose curves in many dogs.
How then are blood glucose curves useful? Veterinary academicians argue as to the existence of the Somogyi phenomenon, but whether some dogs experience true insulin-induced hyperglycemia (from a rebound effect in response to hypoglycemia), or whether it is a simple question of insulin kinetics, wide swings reminiscent of the Somogyi effect are possible. For example, a dog receiving insulin might respond quickly with profound hypoglycemia. If the dog experiences insulin-induced hyperglycemia or if the action of the insulin is short-lived, the dog can return quickly to a hyperglycemic state. In this instance, fructosamine would remain elevated (from persistent hyperglycemia), and clinical signs of hyperglycemia would persist. Without the information gained from a blood glucose curve, the dose of insulin might be mistakenly increased with disastrous consequences. Unfortunately, while they can only give a rough estimate of the duration of action of insulin or the appropriateness of a given dose, blood glucose curves may be the best means of identifying hypoglycemia, and they may have other uses as well. Still, the best strategy for monitoring insulin therapy may be by clinical response (resolution of polyuria/polydipsia, improvement in body condition and appetite, lack of signs of hypoglycemia).
Oral Hypoglycemic Drugs
Oral medications have been available for many years for treatment of hyperglycemia in type 2 diabetic patients. It is important to note that oral hypoglycemic drugs are not useful for the treatment of type 1 diabetes, with the possible exception of acarbose (see below). Glipizide is the most commonly used of these drugs in veterinary medicine, but its use is limited to cats. Most oral hypoglycemic agents rely on a certain level of endogenous insulin-secreting ability in the patient, and they act either to increase the amount of insulin secreted, by sensitizing target cells to insulin, or by decreasing gluconeogenesis so that hyperglycemia is curbed. One exception is the drug acarbose.
Acarbose is an alpha-glucosidase inhibitor that inhibits digestion of complex carbohydrates at the small intestinal brush border. This action slows the absorption of small carbohydrates (such as glucose), and blunts the post-prandial rise in blood glucose seen in many diabetic patients. The major side effects of the drug are flatulence and diarrhea, although I have not observed these effects in several dogs treated with the drug. Dogs seem to tolerate the drug well, and it can effectively treat post-prandial hyperglycemia in some diabetic dogs. In one study of acarbose in 5 diabetic dogs, insulin therapy combined with acarbose resulted in better control of hyperglycemia than insulin alone, although diarrhea developed in 3 of the 5 dogs when treated with acarbose12. The use of acarbose has also been reported in cats.10 In that study, diabetic cats were fed a low-carbohydrate diet and treated with acarbose in addition to insulin or glipizide. Diabetes went into “remission” in 11 of 18 cats, but the effect of acarbose was not studied per se, controls were not included, and glycemic control was not documented. No adverse effects were observed.
Glipizide (Glucotrol®), a sulfonylurea drug, works by stimulating insulin secretion from the pancreatic beta cell, and it has been used fairly extensively in cats, usually in combination with a low-carbohydrate diet. In cats, it is given at a dose of 5 mg/kg BID, and reported side effects have included vomiting (15% of patients), hypoglycemia, and cholestatic liver disease.2 Glyburide is another commonly prescribed sulfonylurea drug used in human diabetes (sold under several tradenames), but it has not been evaluated in veterinary diabetic patients.
Biguanide drugs work by decreasing hepatic glucose production. The prototypical biguanide, metformin (Glucophage®, Bristol-Myers Squib), has been used extensively for decades in the treatment of human diabetes and pre-diabetes, but its use in veterinary medicine has been limited. Because the drug does not stimulate insulin secretion, hypoglycemia is a rare side effect, but vomiting and diarrhea can occur with the use of this drug. In one small study of 5 diabetic cats treated with metformin, hyperglycemia was controlled in only one, and one cat died shortly after starting treatment. Although no biochemical abnormalities were observed, side effects included lethargy, inappetance, vomiting, and weight loss Based on this small study, metformin should be used with caution in veterinary patients.13 In human diabetes patients, metformin is sometimes combined with other oral hypoglycemic drugs.
Thiazolidinediones have received considerable attention in treating type 2 diabetes in human patients, and these drugs act by increasing tissue sensitivity to insulin. This is accomplished by binding to the nuclear receptor peroxisome proliferator-activated receptor gamma, resulting in increased sensitivity to insulin in fat and muscle cells. Troglitazone (Rezulin®), the first thiazolidinedione approved for use in the U.S., was removed from the U.S. market in 2000 due to the increased risk of hepatotoxicity in people, but rosiglitazone (Avandia®, GlaxoSmithKline) and pioglitazone (Actos®, Takeda) are still available. The use of these drugs has also been associated with increased risk of heart disease and other adverse events, and they should be used with caution in veterinary patients, if at all. This class of drugs has not been evaluated in cats with diabetes, but one study showed improved insulin sensitivity in obese laboratory cats treated with darglitazone.6
1. Casella, M., G. Wess, M. Hässig, and C.E. Reusch. 2003. Home monitoring of blood glucose concentration by owners of diabetic dogs. J. Small Anim. Pract. 44:298–305.
2. Feldman, E.C., R.W. Nelson, and M.S. Feldman. 1997. Intensive 50-week evaluation of glipizide administration in 50 cats with previously untreated diabetes mellitus. J. Am. Vet. Med. Assoc. 210:772–777.
3. Fleeman, LM., and J.S. Rand. 2003. Evaluation of day-to-day variability of serial blood glucose concentration curves in diabetic dogs. J. Am. Vet. Med. Assoc. 222:317–321.
4. Fracassi, F., F.S. Boretti, N.S. Sieber-Ruckstuhl, and C.E. Reusch. 2012. Use of insulin glargine in dogs with diabetes mellitus. Vet. Rec. 170:52.
5. Gilor, C., T.K. Ridge, K.J. Attermeier, and T.K. Graves. 2010. Pharmacodynamics of insulin detemir and insulin glargine assessed by an isoglycemic clamp method in healthy cats. J. Vet. Intern. Med. 24:870–874.
6. Hoenig, M. and D.C. Ferguson. 2003. Effect of darglitazone on glucose clearance and lipid metabolism in obese cats. Am. J. Vet. Res. 64:1409–1413.
7. Keating, G.M. 2013. Insulin degludec and insulin degludec/insulin aspart: a review of their use in the management of diabetes mellitus. Drugs. 73:575–593.
8. Marshall, R.D., J.S. Rand, and J.M. Morton. 2008. Glargine and protamine zinc insulin have a longer duration of action and result in lower mean daily glucose concentrations than lente insulin in healthy cats. J. Vet. Pharmacol. Ther. 31:205–212.
9. Martin, G.J, and J.S. Rand. 2001. Pharmacology of a 40 IU/ml porcine lente insulin preparation in diabetic cats: findings during the first week and after 5 or 9 weeks of therapy. J. Feline Med. Surg. 3:23–30.
10. Massaferro, E.M., D.S. Greco, A.S. Turner and M.J. Fettman. 2003. Treatment of feline diabetes mellitus using an alpha-glucosidase inhibitor and a low-carbohydrate diet. J. Feline Med. Surg. 5:183–189.
11. Nelson, R.W., K. Henle, and C. Cole. 2009. Field safety and efficacy of protamine zinc recombinant human insulin for treatment of diabetes mellitus in cats. J. Vet. Intern. Med. 23:787–793.
12. Nelson, R.W., J. Robertson, E.D. Feldman and C. Briggs. 2000. Effect of the α-glucosidase inhibitor acarbose on control of glycemia in dogs with naturally acquired diabetes mellitus. J. Am. Vet. Med. Assoc. 216:1265–1269.
13. Nelson, R., D. Spann, D. Elliott, A. Brondos and R. Vulliet. 2004.Evaluation of the oral antihyperglycemic drug metformin in normal and diabetic cats. J. Vet. Intern. Med. 18:18–24.
14. Plum, A., H. Agerso, and L. Andersen. 2000. Pharmacokinetics of the rapid-acting insulin analog, insulin aspart, in rats, dogs, and pigs, and pharmacodynamics of insulin aspart in pigs. Drug. Metab. Dispos. 28:155–160.
15. Rand, J. Editorial: glargine, a new long-acting insulin analog for diabetic cats. J Vet Intern Med. 2006;20:219–220.
16. Rave K., E. Potocka, L. Heinemann, A.H. Boss, M. Marino, D. Costello, and R. Chen. 2009. Pharmacokinetics and linear exposure of AFRESA compared with the subcutaneous injection of regular human insulin. Diabetes Obes. Metab. 11:715–720.
17. Roomp, K., and J. Rand. 2009. Intensive blood glucose control is safe and effective in diabetic cats using home monitoring and treatment with glargine. J. Feline Med. Surg. 11:668–682.
18. Sako, T., A. Mori, P. Lee P, H. Oda, K. Saeki, Y. Miki, M. Kurishima, K. Mimura, S. Nozawa, H. Mizutani, Y. Makino, K. Ishioka, and T. Arai. 2011. Time-action profiles of insulin detemir in normal and diabetic dogs. Res. Vet. Sci. 90:396–403.
19. Sears, K.W., K.J. Drobatz, and R.W. Hess. 2009. Use of lispro insulin for treatment of dogs with diabetic ketoacidosis (abstract #40). J. Vet. Intern. Med. 23:696.
20. Sheldon, B., D. Russell-Jones, and J. Wright. 2009. Insulin analogues: an example of applied medical science. Diabetes Obes. Metab. 11:5–19.
21. Weaver, K.E., E.A. Rozanski, O.M. Mahony, D.L. Chan, and L.M. Freeman 2006. Use of glargine and lente insulins in cats with diabetes mellitus. J. Vet. Intern. Med. 20:234–238.